DIABETIC NEUROPATHY

DIABETIC NEUROPATHY

Edited by

Mitra Tavakoli
Exeter Centre of Excellence for Diabetes Research, NIHR Exeter Clinical Research Facility, and
Institute of Biomedical and Clinical Sciences, University of Exeter Medical School, Exeter, United Kingdom
Elsevier
Radarweg 29, PO Box 211, 1000 AE Amsterdam, Netherlands
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
Copyright © 2022 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photo-
copying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how
to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as
the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted
herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding,
changes in research methods, professional practices, or medical treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, meth-
ods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety
and the safety of others, including parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/
or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any meth-
ods, products, instructions, or ideas contained in the material herein.

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
A catalog record for this book is available from the Library of Congress
ISBN: 978-0-12-820669-0

For Information on all Elsevier publications

visit our website at https://www.elsevier.com/books-and-journals

Publisher: Stacy Masucci

Acquisitions Editor: Ana Claudia A. Garcia
Editorial Project Manager: Pat Gonzalez
Production Project Manager: Maria Bernard
Cover Designer: Christian J. Bilbow
Typeset by MPS Limited, Chennai, India
 Dedication

To All Patients Who Suffer from Diabetes and its Complications!

 Contents

List of contributors xiii Do the vascular alterations induce ischemic/hypoxic

neuropathy in diabetes? 26
About the editor xvii Resistance to ischemia in diabetic nerve and ischemia/
Foreword xix reperfusion injury 26
Preface xxi Pathology of nerve microenvironment 26
Skin biopsy 27
Acknowledgments xxiii Conclusion 28
Declaration of competing interests 28
1. Classification, risk factors, and clinical References 28
presentation diabetic neuropathy 1
DILEK GOGAS YAVUZ 3. Electrodiagnostic evaluation in diabetic
neuropathy 35
Introduction 1
DEEPAK MENON AND VERA BRIL
Classification 2
Risk factors 3 Introduction 35
Symptoms and signs of diabetic neuropathy 4 Pathophysiology of diabetic sensorimotor
Conclusion 8 polyneuropathy 35
References 8 Nerve conduction parameters and correlation with
pathology 36
2. Pathologic basis for diabetic neuropathy Physiological changes in nerve function in diabetic
in humans 11 sensorimotor polyneuropathy and correlation with
SOROKU YAGIHASHI nerve conduction studies 36
Which nerves to choose and what parameters to test 37
Introduction 11 How to diagnose diabetic sensorimotor polyneuropathy
Classification of diabetic neuropathy 12 with nerve conduction studies 38
Basic pathologic changes in several types of Evolution from subclinical to clinical diabetic
neuropathy 13 sensorimotor polyneuropathy 38
Sensory nerve pathology 14 Nerve conduction studies features in diabetic
Distal nerve fiber loss 14 sensorimotor polyneuropathy: type 1 versus type 2
Metabolic and vascular basis for nerve fiber loss and diabetes mellitus 39
heterogeneity of neuropathy 16 Impact of treatment in nerve conduction studies 39
Demyelination and ultrastructure of Schwann cells 18 Atypical features in nerve conduction studies—
Impaired regeneration and related factors 20 demyelinating features in diabetic sensorimotor
Motor nerve pathology 20 polyneuropathy and when to suspect chronic
Autonomic nerve pathology 22 inflammatory demyelinating
Implication of microangiopathy in diabetic polyradiculoneuropathy 40
polyneuropathy 22 Nerve conduction studies in predicting prognosis 41
Distinct vascular supply in peripheral nerve 22 Vibration perception thresholds in diabetic
Microvessel changes in diabetic nerve 23 neuropathy 42

vii
viii Contents

Electromyography in diabetic sensorimotor How and where to take a skin punch biopsy following
polyneuropathy 42 EFNS/PNS guidelines 81
Limitations of nerve conduction studies 42 Freezing the biopsy 82
Conclusion 43 Staining protocol for IENFD analysis 82
References 43 IENFD quantification 84
Diagnostic accuracy of IENFD analysis in DN 84
4. Advances in screening, early diagnosis, Relationship between skin biopsy outcomes and
symptoms or signs of DN 85
and accurate staging of diabetic Other morphological changes 86
neuropathy 47 Alternative markers used on skin biopsies 86
JOSIE CARMICHAEL, HASSAN FADAVI, FUKASHI Biomarkers of pain pathophysiology 86
ISHIBASHI, ANGELA C. SHORE AND MITRA TAVAKOLI Autonomic structures 87
Conclusion 87
Introduction 47
References 88
Types of nerve fibers 48
Screening 49
Diagnostic tests for diabetic peripheral neuropathy 50
6. Central nervous system involvement in
Symptoms 51 diabetic peripheral neuropathy 91
Signs 51 JOYCE LIM, SOLOMON TESFAYE AND
Large-fiber tests 54 DINESH SELVARAJAH
DPNCheck 55
Painful diabetic neuropathy 91
Small-fiber tests 55
Central nervous system changes in response to peripheral
Quantitative sensory testing 59
nerve injury and how these changes result in chronic
Corneal nerves as a biomarker for diabetic peripheral
pain in diabetes 92
neuropathy 60
Changes in the spinal cord 93
Corneal confocal microscopy for the detection of
Changes in the brain—structural and functional 93
diabetic peripheral neuropathy 61
Recent studies 94
Diagnostic performance for clinical diabetic peripheral
Maladaptive responses 95
neuropathy 61
Clinical implications 96
Early detection of neuropathy 65
Future direction 96
Langerhans cells in diabetic peripheral neuropathy 66
Conclusions 98
Comparing corneal confocal microscopy and
References 99
intraepidermal nerve fiber density 67
Longitudinal studies for application of corneal confocal
microscopy for assessment diabetic peripheral 7. Spinal cord involvement in diabetic
neuropathy 68 neuropathy and neuropathic pain 103
Corneal confocal microscopy application in clinical ANDREW G. MARSHALL, ANNE MARSHALL AND
trials 69 NIGEL A. CALCUTT
Conclusions 70
References 71 Introduction 103
Glucotoxicity in the spinal cord 104
Spinal cord pathology associated with diabetes and
5. Skin biopsy analysis in diabetic neuropathy 105
neuropathy 79 Studies in animal models of diabetes 106
PÁLL KARLSSON Spinal cord pathology 106
Spinal cord electrophysiology 106
Introduction 79 Spinal cord cellular pathology 108
Advantages and disadvantages of skin biopsies 80 Clinical studies of spinally mediated pain in
IENFD quantification in nondiabetic etiologies 81 diabetes 110
 Contents ix
The spinal cord and painful diabetic neuropathy 111 Foot muscle atrophy—an early sign of motor deficits in
Peripheral input 111 DSPN 148
Spinal mechanisms 112 Compositional and structural muscle changes of the
Conclusion 116 lower extremity 149
Acknowledgments 116 Emerging quantitative MRI parameters for muscle
References 116 assessment 151
Ultrasonomyopgrahy 153
8. The conundrum and enigma of painful Muscle energy properties 154
and painless neuropathy 123 Management and interventions 155
References 157
JOHAN RØIKJER AND NIELS EJSKJAER

Introduction 123 10. Diabetic autonomic neuropathy,

Clinical presentations and highly variable patterns of measurement and management; part 1:
neuropathy 123 measurement 163
Risk factors for developing diabetic peripheral and JOSEPH COLOMBO
neuropathic pain 124
Huge unmet clinical needs and treatment Autonomic neuropathy 163
challenges 125 A brief history 163
Current status for detecting painful and painless Autonomic function 165
neuropathy 125 Autonomic dysfunction and morbidity and mortality
Early detection and prevention 126 risk 166
Similarities in treatment of painful and painless A history of assumption and approximation
neuropathy 127 quantified 167
Deep sensory profiling and mechanism-based treatment Diabetes: also considered a model of the effect of chronic
in painful and painless neuropathy 127 disease on P&S activity over time 170
Novel clinical endpoints and research Natural history of autonomic decline 170
methodologies 129 Earlier detection and earlier intervention 173
Perspectives 130 Small fiber disorder 173
References 130 Postural change P&S dysfunction 174
Autonomic neuropathy scoring 177
9. Motor dysfunction in diabetes 135 Diagnosis of autonomic dysfunction 178
ANDERS STOUGE AND HENNING ANDERSEN Acknowledgments 180
References 180
Introduction 135
Motor nerve dysfunction 136 11. Diabetic neuropathy in children and
Electroneuronography 136 youth 185
Electromyography 137
GULCIN AKINCI, MASHA G. SAVELIEFF,
Muscle strength and functional impairments 138 GARY GALLAGHER, BRIAN C. CALLAGHAN AND
Muscle strength evaluation 138 EVA L. FELDMAN
Diabetic myopathy 141
Diabetic neuropathy and muscle impairment 141 Background 185
Muscle impairments in diabetes independent of Definitions and classifications 186
neuropathy 145 Epidemiology 186
Muscle strength, balance, gait, and falls 147 Clinical presentations 189
Muscular composition and structure 147 Diabetic neuropathy 189
Introduction—technical advances and muscle Diabetic autonomic neuropathy 189
composition 147 Risk factors 191
Clinical assessment of muscle atrophy 148 Role of puberty and height 191
x Contents

Sex 191 14. Diabetic foot 223

Race and ethnicity 191 FRANK LEE BOWLING, KEELEY JANE FOLEY AND
Glycemic control 193 ANDREW J.M. BOULTON
Obesity and other metabolic syndrome components 193
Smoking 194 Introduction 223
Eating disorders 194 Epidemiology of diabetic foot problems 223
Screening and diagnosis 194 Risk assessment 224
Diabetic neuropathy, quality of life, and mortality 195 Diabetic peripheral neuropathy 225
Managing diabetic neuropathy 195 Neurological assessment 226
Emerging risk factors 195 Treatment 227
Painful neuropathy 195 Peripheral neuropathy and autoimmune disease 227
Diabetic autonomic neuropathy 195 Charcot neuroarthropathy 228
Conclusion 196 Classification of Charcot neuroarthropathy 228
Acknowledgments 196 Charcot neuroarthropathy treatment 229
References 196 Management of diabetic foot ulceration 229
Wound classification 229
12. Treatment-induced painful diabetic Neuropathic and neuroischemic ulcers 229
neuropathy 201 Neuropathic and neuroischemic ulcer treatment 230
Antibiotic-resistant organisms 230
MILLA ROSENGÅRD-BÄRLUND AND HANNA HARNO
Treatment of infection in diabetic foot disease 231
Introduction 201 Antibiotic therapy 231
Potential mechanisms and pathogenesis 202 Soft tissue infection 231
Epidemiology 202 Osteomyelitis 231
Risk factors 202 Debridement 232
Clinical presentation 203 Offloading 232
Diagnostic methods 204 References 232
Management of symptoms associated with treatment-
induced neuropathy of diabetes 206 15. Management of diabetic foot
Outcomes of treatment-induced neuropathy of diabetes
disease 235
and association with other microvascular
PRASHANTH R.J. VAS AND VENU KAVARTHAPU
complications 207
Discussion 207
Introduction 235
Declaration of competing interests 209
Management of diabetic foot ulceration 235
Acknowledgments 209
Multidisciplinary diabetic foot service 236
References 209
Management from ulceration to healing 236
Advanced therapies in diabetic foot ulceration 244
13. Health economics of diabetic foot Metabolic control and treatment of associated
disease: costs of diabetic neuropathy and comorbidities 248
diabetic foot 211 Charcot neuroarthropathy “Charcot foot” 250
BEATRIZ RODRÍGUEZ-SÁNCHEZ AND ALAN SINCLAIR Surgical management of the diabetic foot 251
Surgical management of infected diabetic foot ulcer and
Background 211 infection 251
Cost-of-illness studies on diabetic foot disease 212 Surgical management of a deformed diabetic
Cost-effectiveness and cost-utility of diabetic foot foot 251
disease-related interventions 214 Future outlook 252
Conclusion 219 Conclusion 252
References 219 References 253
 Contents xi
16. Strategies for the prevention or Nerve conduction parameters 288
reversal of neuropathy 259 Intraepidermal nerve fiber density 290
Alternative treatments 290
LINDSAY A. ZILLIOX, KRISH CHANDRASEKARAN AND
JAMES W. RUSSELL Targeting balance 290
Targeting neuropathic pain 290
Introduction 259 Discussion 291
Improved glycemic control 260 Conclusion 292
Somatic neuropathy in type 1 diabetes mellitus 260 References 292
Somatic neuropathy in type 2 diabetes mellitus 261
Diabetic autonomic neuropathy 263
Lifestyle interventions 264 18. Treatment of diabetic
Bariatric surgery 269 polyneuropathy 299
Dietary interventions 270
AMANDA C. PELTIER
Multifactorial treatment 270
Treatment based on pathogenic concepts 271 Epidemiology and historical trials in diabetic
Alpha-lipoic acid 272 polyneuropathy 299
Sirtuin 1 (SIRT1) and diabetic polyneuropathy 272 Glucose control and diabetic polyneuropathy 300
Role of NAD1 analogs in diabetic Treatment of pain in diabetic polyneuropathy 301
polyneuropathy 273 Gabapentanoids 301
Benfotiamine 274 Serotonin norepinephrine reuptake inhibitors 301
Actovegin 274 Tricyclic antidepressants 303
Epalrestat 275 Antiepileptic drugs 303
Antiinflammatory drugs 275 Opioids 303
Conclusion 276 Supplements and alternative treatments for painful
Acknowledgments 276 diabetic polyneuropathy 304
References 276 α-Lipoic acid 304
Benfotiamine 304
Cannabidiol and cannabis-derived treatments 304
17. Treatment of diabetic peripheral Exercise and nonpharmacologic treatments 305
neuropathy: technologies, exercise, and Emerging treatments and recent trials 305
alternative treatments 283 Conclusion 306
References 307
KALLIOPI PAFILI AND NIKOLAOS PAPANAS

Introduction 283
Technologies 283 19. Diabetic autonomic neuropathy,
Technologies targeting improvement in quality of life measurement and management; part II:
and medical adherence 283 management 313
Technologies to identify high-risk for impendent diabetic JOSEPH COLOMBO
foot ulceration 284
Technologies targeting personalized offloading 284 Autonomic neuropathy 313
Technologies to increase adherence to therapeutic Earlier detection and earlier intervention 314
footwear 285 Disease initiation and progression of neuropathy 314
Technologies to monitor physical activity 285 Oxidative stress 316
Exercise 286 Treating diabetic autonomic dysfunction and
Quality of life, neuropathic symptoms, and neuropathy 321
deficits 286 Antioxidants 325
Physical capacity 287 Acknowledgements and disclaimers 329
Balance 288 References 329
xii Contents

20. Translating a treatment for diabetic 21. Bedside clinical features and
peripheral neuropathy from rodents to translational snapshots of diabetic
humans: can a case be made for fish oil and polyneuropathy 349
salsalate? 337 BRENDAN N. PUTKO AND DOUGLAS W. ZOCHODNE
MARK YOREK
Introduction 349
Diabetic peripheral neuropathy 337 Symptoms 350
Omega-3 polyunsaturated fatty acid; eicosapentaenoic Examination 354
acid and docosahexaenoic acid 338 Complications 358
Omega-3 polyunsaturated fatty acid and cardiovascular Neuropathy scales 358
health 339 Clinical trials in DPN, a recent snapshot 359
Omega-3 polyunsaturated fatty acid and vascular Translational opportunities 361
dysfunction 341 Conclusions 362
Omega-3 polyunsaturated fatty acid and peripheral References 363
neuropathy related to obesity and insulin
resistance 341 Index 367
Omega-3 polyunsaturated fatty acids and diabetic
peripheral neuropathy 342
References 344
 List of contributors

Gulcin Akinci Department of Neurology, Krish Chandrasekaran Department of Neurology,

University of Michigan, Ann Arbor, MI, United University of Maryland and Maryland
States VA Healthcare System, Baltimore, MD, United
Henning Andersen Department of Neurology, States
Aarhus University Hospital, Aarhus, Denmark Joseph Colombo Franklin Cardiovascular
Andrew J.M. Boulton Division of Endocrinology, Associates, PA and Autonomic Dysfunction and
Diabetes and Gastroenterology, Faculty of POTS Center, Sewell, NJ, United States; Physio
Medicine, Biology and Health, University of PS, Inc., Atlanta, GA, United States;
Manchester, Manchester, United Kingdom; Neurocardiology Research Center, Sicklerville,
Manchester Royal Infirmary, Manchester, United NJ, United States
Kingdom; Diabetes Research Institute, University Niels Ejskjaer Steno Diabetes Center North
of Miami, Miami, FL, United States Denmark, Aalborg University Hospital, Aalborg,
Frank Lee Bowling Division of Endocrinology, Denmark; Departments of Endocrinology and
Diabetes and Gastroenterology, Faculty of Clinical Medicine, Aalborg University Hospital,
Medicine, Biology and Health, University of Aalborg, Denmark
Manchester, Manchester, United Kingdom; Hassan Fadavi Peripheral Neuropathy Group,
Manchester Royal Infirmary, Manchester, United Imperial College, London, United Kingdom
Kingdom; Vascular Surgery and Diabetes,
Central Manchester University Hospital – NHS Eva L. Feldman Department of Neurology,
Foundation Trust, Manchester, United Kingdom University of Michigan, Ann Arbor, MI, United
States
Vera Bril Ellen and Martin Prosserman Centre for
Neuromuscular Diseases, University Health Keeley Jane Foley Vascular Surgery and Diabetes,
Network, University of Toronto, Toronto, ON, Central Manchester University Hospital – NHS
Canada Foundation Trust, Manchester, United Kingdom
Nigel A. Calcutt Department of Pathology, School Gary Gallagher Department of Neurology,
of Medicine, University of California San Diego, University of Michigan, Ann Arbor, MI, United
La Jolla, CA, United States States

Brian C. Callaghan Department of Neurology, Hanna Harno Clinical Neurosciences, Neurology,

University of Michigan, Ann Arbor, MI, United Helsinki University Hospital and University of
States Helsinki, Helsinki, Finland

Josie Carmichael Exeter Centre of Excellence for Fukashi Ishibashi Ishibashi Medical and Diabetes
Diabetes Research, NIHR Exeter Clinical Centre, Hiroshima, Japan
Research Facility and Institute of Biomedical and Páll Karlsson Danish Pain Research Center,
Clinical Sciences, University of Exeter Medical Department of Clinical Medicine, Aarhus
School, Exeter, United Kingdom University, Aarhus, Denmark

xiii
xiv List of contributors

Venu Kavarthapu Diabetes Foot Clinic, King’s Johan Røikjer Steno Diabetes Center North
College Hospital, London, United Kingdom; Denmark, Aalborg University Hospital, Aalborg,
Department of Orthopaedics, King’s College Denmark; Center for Sensory-Motor Interaction,
Hospital, London, United Kingdom Aalborg University, Aalborg, Denmark
Joyce Lim Department of Oncology and Human Milla Rosengård-Bärlund Division of
Metabolism, Medical School, University of Endocrinology, Abdominal Center, Helsinki
Sheffield, Sheffield, United Kingdom University Hospital and University of Helsinki,
Helsinki, Finland
Andrew G. Marshall Musculoskeletal Biology,
Institute of Life Course and Medical Sciences, James W. Russell Department of Neurology,
University of Liverpool, Liverpool, United University of Maryland and Maryland VA
Kingdom; Department of Clinical Healthcare System, Baltimore, MD, United States
Neurophysiology, The Walton Centre, Liverpool, Masha G. Savelieff Department of Neurology,
United Kingdom; Division of Neuroscience and University of Michigan, Ann Arbor, MI, United
Experimental Psychology, School of Biological States
Sciences, Faculty of Biology, Medicine and Dinesh Selvarajah Department of Oncology and
Health, University of Manchester, Manchester, Human Metabolism, Medical School, University
United Kingdom of Sheffield, Sheffield, United Kingdom
Anne Marshall Division of Diabetes, Angela C. Shore Exeter Centre of Excellence for
Endocrinology and Gastroenterology, School of Diabetes Research, NIHR Exeter Clinical
Medical Sciences, Faculty of Biology, Medicine Research Facility and Institute of Biomedical and
and Health, University of Manchester, Clinical Sciences, University of Exeter Medical
Manchester, United Kingdom School, Exeter, United Kingdom
Deepak Menon Ellen and Martin Prosserman Alan Sinclair Foundation for Diabetes Research in
Centre for Neuromuscular Diseases, University Older People, King’s College, London, United
Health Network, University of Toronto, Toronto, Kingdom
ON, Canada
Anders Stouge Department of Neurology, Aarhus
Kalliopi Pafili Diabetes Centre-Diabetic Foot University Hospital, Aarhus, Denmark
Clinic, Second Department of Internal Medicine, Mitra Tavakoli Exeter Centre of Excellence for
Democritus University of Thrace, University Diabetes Research, NIHR Exeter Clinical
Hospital of Alexandroupolis, Alexandroupolis, Research Facility and Institute of Biomedical and
Greece Clinical Sciences, University of Exeter Medical
Nikolaos Papanas Diabetes Centre-Diabetic Foot School, Exeter, United Kingdom
Clinic, Second Department of Internal Medicine, Solomon Tesfaye Academic Department of
Democritus University of Thrace, University Diabetes and Endocrinology, Sheffield Teaching
Hospital of Alexandroupolis, Alexandroupolis, Hospitals NHS Foundation Trust, Sheffield,
Greece
United Kingdom
Amanda C. Peltier Neuromuscular Division, Prashanth R.J. Vas Diabetes Foot Clinic, King’s
Department of Neurology and Medicine, College Hospital, London, United Kingdom;
Vanderbilt University Medical Center, Nashville, King’s Health Partners’ Institute of Diabetes,
TN, United States Endocrinology and Obesity, London, United
Brendan N. Putko Division of Neurology, Kingdom
Department of Medicine and the Neuroscience Soroku Yagihashi Department of Exploratory
and Mental Health Institute, University of Medicine on Nature, Life and Man, Toho
Alberta, Edmonton, Canada University School of Medicine, Inage-cho, Inage-
Beatriz Rodrı́guez-Sánchez Department of ku, Chiba, Japan; Department of Pathology and
Applied Economics, Public Economics and Molecular Medicine, Hirosaki University
Political Economy, Faculty of Law, University Graduate School of Medicine, Hirosaki, Aomori,
Complutense of Madrid, Madrid, Spain Japan
 List of contributors xv
Dilek Gogas Yavuz Section of Endocrinology and Lindsay A. Zilliox Department of Neurology,
Metabolism, Marmara University School of University of Maryland and Maryland VA
Medicine, Istanbul, Turkey Healthcare System, Baltimore, MD, United States
Mark Yorek Department of Internal Medicine, Douglas W. Zochodne Division of Neurology,
University of Iowa, Iowa City, IA, United States; Department of Medicine and the Neuroscience
Department of Veterans Affairs Iowa City Health and Mental Health Institute, University of
Care System, Iowa City, IA, United States; Alberta, Edmonton, Canada
Fraternal Order of Eagles Diabetes Research
Center, University of Iowa, Iowa City, IA, United
States
 About the editor

This book represents a career-long interest intervention at rare models of neuropathies ini-
by Dr. Tavakoli in advancing our knowledge tiated a new direction of research.
of diabetic neuropathy, which has been her Dr. Tavakoli has been the lead investigator
passion for the past 20 years. in several national and international multisite
Dr. Mitra Tavakoli is an associate professor clinical trials and a chief investigator in major
of medicine at the University of Exeter and is clinical trials. Her research is funded mainly by
distinguished for her significant contribution Industry, National Institute of Health Research
to our understanding of the small fiber neurop- (NIHR), and National Institute of Health
athy at diabetes and establishing the technique (NIH). She has published over 75 peer-
of “Corneal Confocal Microscopy.” She is an reviewed scientific papers and textbook
internationally recognized researcher in diabe- chapters and in excess of 100 abstracts. She
tes care, with a significant focus on innovation regularly speaks at international and national
and bringing new ideas to the field. conferences and meetings. In addition, she is
Her recent work elucidated the role of inter- widely respected for training scientists across
vention for the regeneration of nerves at Type the world. She spearheads new technologies
1 and Type 2 diabetes. Her ongoing research and contributes to the major UK and interna-
into understanding the role of novel tional collaborations.

xvii
 Foreword

At the beginning of the 21st century, we are but noncommunicable diseases such as diabetes.
experiencing a global epidemic of diabetes with His words were very true, and of all the late
approximately 500 million people with the con- complications of diabetes, neuropathies are the
dition across the world [1]. The International commonest. It is the neuropathies that lead to
Diabetes Federation (IDF) publishes an Atlas of significant morbidity and mortality and drive
the global epidemiology of diabetes every 2 the cost of diabetes across the world. However,
years, and the recent one published in at present we are experiencing a global pan-
December 2019 estimated that the number of demic of Covid-19. There has therefore been a
people with diabetes would increase to .700 massive global effort in devising new treatments
million in the next 20 years or so [1]. The cur- for this all-too-often fatal viral illness as well as
rent prevalence of diabetes in the United States rolling out a global vaccination program. As
is approximately 13.3% affecting around 34 mil- President of the IDF, it is our fear that this pan-
lion individuals. There are certain ethnicities demic might detract from the recent increased
that have higher prevalence of diabetes, and in interest in the importance of noncommunicable
the United States this includes native diseases such as diabetes, heart disease, and can-
Americans, Hispanics, and Blacks although cer across the world. During 2020, there was
recent evidence suggests that in certain of these already evidence that the management of numer-
groups the prevalence has decreased in recent ous very ill patients had necessarily led to the
years [2,3]. Elsewhere in the world there is even neglect of many conditions particularly noncom-
higher prevalence with Mauritius at approxi- municable diseases such as diabetes [5]. In the
mately 25% rising in some Polynesian islands area of diabetes complications the threat to rou-
to 33% [1]. A population-based study from tine diabetes care has been already outlined in
Pakistan suggests that more than one in four several publications [6 8]. The IDF also fears
adults in that country has diabetes [4]. that the difficulties in diabetes management dur-
Elsewhere in the world, prevalence is particu- ing the pandemic, including canceling clinical
larly high in the Indian subcontinent, Middle appointments, occasional virtual telephone calls,
East, and many of the Caribbean islands. fear of patients attending hospital, will lead to a
A few years ago I had the honor to share the tsunami of late complications in the coming
stage at a large meeting in Bangalore, South years. There are already anecdotal reports from a
India, with the then Prime Minister, Sri number of countries suggesting that lower limb
Narendra Modi. In a passionate speech, he spoke amputations are increasing as a consequence of
about the problems faced by Indian population: neuropathy and vascular disease going
he commented that the biggest threat to the untreated during the Covid-19 pandemic.
health of the Indian population and indeed the It is therefore vital to raise the global profile
Indian subcontinent was no longer communica- and importance of diabetes and particularly
ble diseases such as malaria and tuberculosis, the late complications. Another grim statistic

xix
xx Foreword

was recorded in 2019 when for the first time Royal Infirmary; University of Miami, Miami, FL,
diabetes was listed in the top 10 causes of United States 2International Diabetes Federation;
death across the world by the NHS [9]. Among European Association for the Study of Diabetes,
males, diabetes represented the large rise of
Manchester Diabetes Centre, Manchester Royal
deaths among the top ten conditions with an
increase of 80% since 2000. Infirmary, Manchester, United Kingdom
As noted above, neuropathies are among the
most common long-term complications of dia-
betes affecting up to 50% of patients [10,11]. References
The clinical features of the neuropathies vary
[1] IDF Diabetes Atlas, 9th edition 2019, Brussels.
immensely, and patients may present to a wide
Available at: www.diabetesatlas.org.
spectrum of specialties from dermatology to [2] Cheng YJ, Kanaya AM, Araneta MRG, et al.
podiatry, for example, or from urology to car- Prevalence of diabetes by race and ethnicity in the
diology. The commonest somatic neuropathy, United States, 2011-2016. JAMA 2019;322:2389 98.
chronic sensorimotor neuropathy, affects up to [3] Bullock A, Sheff K, Hora I, et al. Prevalence of diag-
nosed diabetes in American Indian and Alaska Native
50% of older patients with Type 2 diabetes and
adults, 2006-2017. BMJ Open Diabetes Res Care 2020;8
is a major contributory factor in the pathogene- (1):e001218. Available from: https://doi.org/10.1136/
sis of foot ulceration which should be emi- bmjdrc-2020-001218.
nently preventable. Similarly, the autonomic [4] Basit A, Fawwad A, Qureshi H, Shera AS. Prevalence
neuropathies are also common and can affect of diabetes, pre-diabetes and associated risk factors:
second national diabetes survey of Pakistan. BMJ
any area receiving autonomic innovation.
Open 2018;8:e020961. Available from: https://doi.
Unfortunately, the epidemiology and natu- org/10.1136/bmjopen-2017-020961.
ral history of the neuropathies remain poorly [5] Godlee F. Surviving the long road ahead. BMJ
defined partly because of poor patient selection 2020;369:m1840.
and variable criteria that have been used in [6] Shin L, Bowling FL, Armstrong DG, Boulton AJM.
Saving the diabetic foot during the COVID-19 pan-
studies of the neuropathies. Most important to
demic: a tale of two cities. Diabetes Care
remember is that sensorimotor neuropathy 2020;48:1704 9.
may be asymptomatic in up to 50% of cases [7] Boulton AJM. Diabetic foot disease during the
and therefore every person with diabetes, COVID-19 pandemic. Medicina 2021;57(2):97.
regardless of type, should have a careful clini- Available from: https://doi.org/10.3390/
Medicina57020097.
cal examination of the lower extremities and [8] Rogers LC, Lavery LA, Joseph WS, Armstrong DG. All
feet at least annually. feet on deck—the role of podiatry during the COVID-
The editor of this book should be congratu- 19 pandemic: preventing hospitalizations in an over-
lated for assembling many experts in the somatic burdened health care system, reducing amputation
and autonomic neuropathies from across the and death in people with diabetes. J Am Podiatr Med
Assoc. 25 March 2020 [Epub ahead of print]. doi:10.
world and from multiple specialties. Each is an
7547/20-051.
expert in their own area and they provide an [9] WHO. The top ten causes of death. 2020. Available at:
excellent overview of all aspects of the diabetic who.int/news-room/fact-sheets/detail/the-top-10-
neuropathies which I am sure that readers from causes-of-death.
any specialty will benefit greatly from reading. [10] Boulton AJM, Malik RA, Sosenko JS, Arezzo JC.
Diabetic somatic neuropathies. Technical review.
Diabetes Care 2004;27:1458 86.
Andrew J.M. Boulton1,2 [11] Pop-Busui R, Boulton AJM, Feldman EL, et al.
1 Diabetic neuropathy: a position statement by the
University of Manchester, Division of Diabetes, American Diabetes Association. Diabetes Care
Endocrinology and Gastroenterology; Manchester 2017;40:136 54.
 Preface

With the current epidemic of obesity and medical care and the curtailing of health ser-
diabetes, which has worsened during the vices, including screening programs.
COVID-19 pandemic, the prevalence and con- During the COVID-19 pandemic, telemedi-
sequences of diabetic neuropathy are set to cine and remote monitoring became remark-
increase globally. ably important. The deployment of
Diabetic peripheral neuropathy (DPN) is the telemedicine, devices, and digital applications
most common diabetes-related microvascular is used to increase screening of individuals and
complication and can significantly increase monitor the progression of diabetes complica-
morbidities, such as chronic pain, foot ulcera- tions, which needs to be easily accessible to
tions and amputations, and mortality. Despite general practitioners and patients. In the
these significant consequences, effective absence of any screening program, this is an
screening and treatment strategies are lacking area that requires further investments and
at present, unlike other diabetes-related micro- investigations for diabetic neuropathy.
vascular complications such as retinopathy One of the biggest lessons we have learned
and nephropathy. This usually results in a from the COVID-19 pandemic over the last 18
delay in the diagnosis of DPN till it is well months is that the scientific community work-
established and more difficult to treat, while ing together can do some pretty amazing
retinopathy and nephropathy can be detected things. The process may have changed forever
early using current screening strategies such as the way drugs are developed. I hope the same
retinal images and urinary assessments, which will be applied to DPN, and we will be able to
allow early interventions to prevent the pro- develop better treatments for diabetic neuropa-
gression of the disease. thy to relieve the burden of this awful disease.
Glucose control is still the only main Throughout the book, the authors presented
disease-modifying therapy for diabetic neurop- the latest findings and practical information.
athy. Several disease modifications and clinical Each author has considerable expertise in their
trials for diabetic neuropathy have failed due own area, and they provide an excellent over-
to a lack of sensitive biomarkers. There exists view of all aspects of diabetic neuropathy.
an urgent need to identify the most accurate This textbook aims to provide a platform for
early biomarker of nerve damage to better advance in basic and clinical science in the
diagnose DPN in the clinical care of patients field of diabetic neuropathy with the hope to
and, in particular, to permit an accurate evalu- translate it to improved patient care.
ation of future therapies in clinical trials.
The current focus on the COVID-19 pan-
demic has invariably led to compromises in

xxi
 Acknowledgments

This book is the result of a huge team effort. It was a great pleasure to work with the
I extend my deepest gratitude and thanks to all Elsevier team throughout. Special thanks to Pat
colleagues who have contributed to the writing Gonzalez for her support, enthusiasm, and
of this book. practical assistance.
I would like to thank all colleagues who con-
tributed to reviewing some of the chapters.

xxiii
 C H A P T E R

1
Classification, risk factors, and clinical
presentation diabetic neuropathy
Dilek Gogas Yavuz
Section of Endocrinology and Metabolism, Marmara University School of Medicine, Istanbul, Turkey

Introduction Morbidity and mortality associated with dia-

betes primarily result from chronic vascular
Diabetes mellitus is a major metabolic dis- complications. Complications of diabetes such
order. Nearly half a billion people are living as retinopathy, nephropathy, and cardiovascular
with diabetes worldwide. International Diabetes diseases are leading to reduced quality of life,
Federation estimated that over four million peo- increased need for medical care, disability and
ple aged 20 79 years died from diabetes- decreased life expectancy in diabetic patients [1].
related causes in 2019 [1]. The global impact of The overall prevalence of ay diabetic neu-
diabetes is increasing considerably; the number ropathy was estimated to be 35% in diabetic
of diabetic patients is expected to increase to patients [4]. Chronic kidney disease resulting
578 million (10.2%) in 2030 and 700 million in diabetic nephropathy and other associated
(10.9%) in 2045 [1]. conditions was reported to be 25% %36 in
Type 2 diabetes is the most common form people with diabetes [5]. Diabetic foot compli-
of diabetes, accounting for around 90% of all cations are associated with neurological disor-
diabetes worldwide. The prevalence of type 2 ders and peripheral vascular disease in the
diabetes is rising across all continents, associated lower limbs [6].
with population aging, increasing urbanization, The global prevalence of diabetic foot com-
and economic development [1,2]. Although plications varies between 3% and 13%, with a
there is considerable variation by region, the global average of 6.4%. Lower limb amputation
incidence of type 1 diabetes is increasing world- in people with diabetes is 10 to 20 times more
wide as well [3]. common than those without diabetes [1,7].
The long-term micro and macrovascular There is a lack of understanding in the
complications of diabetes can be present at management and treatment of diabetic foot
diagnosis in type 2 diabetes and can be complications among healthcare professionals.
detected as early as five years after the onset of Less than one-third of physicians recognize
type 1 diabetes [1]. the signs of diabetic peripheral neuropathy.

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00014-1 1 © 2022 Elsevier Inc. All rights reserved.
2 1. Classification, risk factors, and clinical presentation diabetic neuropathy

Increasing awareness of diabetic foot complica- Thomas originally proposed the classifica-
tions, risk assessments, and foot care by a tion of neuropathy in diabetic patients in 1997,
multidisciplinary team can reduce foot compli- which most authors generally accept. Thomas
cations and amputations by up to 85% [1,8]. divided diabetic neuropathies into four main
Diabetes is the leading cause of neuropathy groups: (1) Hyperglycemic neuropathy, (2)
in the Western world. Diabetic neuropathies Symmetric polyneuropathy (Sensory/autonomic
are the most common among the long-term polyneuropathy, Acute painful diabetic neuro-
complications of diabetes, affecting up to 50% pathy), (3) Focal and multifocal neuropathy
of patients [1]. (Cranial neuropathy, Thoraco-abdominal neu-
Epidemiologic studies indicate that neurop- ropathy, Focal limb neuropathies, Diabetic
athy in diabetes patients is approximately amyotrophy), (4) Mixed forms [11].
20% in community patients [9]. An early study Recently the American Diabetes Association
reported that incidence of neuropathy was 50% classified diabetic neuropathies into three main
for patients with diabetes duration more than categories: (1) diffuse symmetric (distal sym-
25 years [10]. metric polyneuropathy and autonomic), (2)
Diabetic neuropathy can cause significant mononeuropathy (mononeuropathy, mono-
morbidity and mortality. The major problem neuritis multiplex, atypical forms), (3) radiculo-
with Diabetic neuropathy is implicated in pathy, or polyradiculopathy [12].
50% 75% of nontraumatic amputations [11]. Classification of diabetic neuropathy according
Therefore it is essential to detect symptoms or to clinical presentation is shown in Table 1.1.
signs of diabetic neuropathy and to determine
risk factors as early as possible to implement TABLE 1.1 Classification of diabetic neuropathies
interventions and prevent further neuronal according to clinical presentation.
damage. A. Generalized neuropathies
In clinical practice, detection of diabetic neu-
• Distal symmetric polyneuropathy
ropathy is challenging due to the lack of a clear
• Autonomic neuropathy
consensus concerning the definition of the opti- (cardiovascular, gastrointestinal, urogenital, sudomotor)
mal clinical assessment criteria to diagnose • Acute sensory neuropathy
neuropathy in diabetic patients. Treatment-induced neuropathy, Diabetic neuropathic
cachexia
Classification
B. Focal neuropathies

Chronic hyperglycemia can affect any part • Isolated Cranial neuropathies (e.g., Cranial nerves III,
of the nervous system. Sensory, autonomic, VII, VI, IV)
• Isolated peripheral nerve neuropathies (e.g., ulnar,
and motor neurons of the peripheral nervous
median, femoral, peroneal nerve)
system are vulnerable to the adverse effects of • Entrapment neuropathies (e.g., Carpal tunnel syndrome,
hyperglycemia in diabetic patients [10]. Ulnar neuropathy)
Various classifications for diabetic neuro- • Other mononeuropathies (Phrenic mononeuropathy)
pathy have been proposed in recent years. • Mononeuropathy multiplex
Nevertheless, there is no universally accepted
C. Multifocal neuropathies
classification. Some of the suggested classifica-
tions are based on etiology, pathological or Diabetic radiculoplexus neuropathies
topographical features, and a classification lumbosacral radiculoplexus neuropathy
thoracic radiculoneuropathy
based upon the clinical manifestation is most
cervical radiculoplexus neuropathy
useful in clinical practice.

Diabetic Neuropathy
 Classification 3
Clinically, it is essential to distinguish obesity, dyslipidemia might have attenuated
between symmetric and asymmetric neuro- the beneficial effects of glucose control.
pathies because they require different diagnos- Diabetic neuropathy can be increased by
tic and therapeutic approaches. Although the cardiovascular risk factors such as dyslipide-
most common neuropathy type is distal sym- mia, hypertriglyceridemia, low high-density
metric polyneuropathy in type 1 and type 2 lipoprotein, hypertension, abdominal obesity,
diabetes, different forms of Diabetic neuro- independently from glycemic control. Obesity
pathy usually coexist in the same patient [13]. is associated with diabetic neuropathy in
It is also essential to keep in mind that dia- patients with Type 2 diabetes and in selected
betes may not necessarily cause neuropathy in type 1 diabetic patients [17,18].
all diabetic patients. Other etiologic factors Population base studies showed that obesity
rather than diabetes need to be excluded before is common in patients with neuropathy. Also,
diagnosing diabetic neuropathy [13]. clinical studies revealed neuropathy was more
prevalent in obese normoglycemic individuals
than lean controls, suggesting obesity is an
independent risk factor for neuropathy [19].
Risk factors Observational studies highlighted that weight,
Hyperglycemia is the primary risk factor for Body mass index (BMI), and waist circumference
diabetic neuropathy. Improved glycemic con- are all associated with neuropathy. Obesity is
trol can prevent the progression of diabetic the second most influential metabolic risk factor
neuropathy in type 1 diabetes but is not effec- for neuropathy after diabetes [19,20].
tive in preventing distal polyneuropathy in Hypertension is a risk factor for diabetic
type 2 diabetes [14]. Clinical trials have demon- neuropathy—it can be developed with diabetes
strated the benefit of tight glucose control by progression or stimulate progression of chronic
slowing the progression of diabetic peripheral kidney disease. Essential hypertension can
neuropathy in type 1 and type 2 diabetic occur in diabetes patients as comorbidity,
patients [15]. In an observational study long- or diabetic nephropathy can cause secondary
term, persistent benefits of strict glycemic con- hypertension. Hypertension can aggravate poly-
trol were demonstrated in type 1 diabetic neuropathies in diabetic patients as blood pres-
patients [15]. sure control is emphasized clinically for distal
The duration of diabetes is a major risk symmetric polyneuropathy prevention [21].
factor for neuropathy, and it is well recognized Dyslipidemia is a common cardiovascular
that the risk of developing neuropathy increases risk factor in diabetes. Hypertriglyceridemia
with the duration of diabetes and age. As was found to be an important independent risk
hyperglycemia takes time for nerve damage, factor corresponding to a 2.1-fold increase in
diabetic neuropathy is more common in older distal symmetric polyneuropathy occurrence
adults ( . 50 years) [16]. [22]. According to the FIELD study, hypertri-
Diabetic neuropathy should be suspected in glyceridemia was a risk factor for new-onset
all patients with type 2 diabetes and those who DPN, and fenofibrate treatment reduced new-
have had type 1 diabetes for more than five onset neuropathy by approximately 18% and
years [13]. reversed baseline neuropathy [23].
Patients with type 2 diabetes develop distal Height is considered as a marker for neuro-
symmetric polyneuropathy (DSPN) despite nal length, and it was reported as an indepen-
adequate glycemic control. The presence of dent predictor of neuropathy among patients
multiple comorbidities such as hypertension, with type 1 diabetes and type 2 diabetes [24].

Diabetic Neuropathy
4 1. Classification, risk factors, and clinical presentation diabetic neuropathy

TABLE 1.2 Risk Factors of Diabetic Neuropathy. accounting for about 75% of the diabetic neu-
ropathies [12]. The disease’s course is chronic
Unmodifiable risk factors
and progressive. DSPN predisposes diabetic
Advanced Age patients to variable degrees of pain, motor dys-
Duration of diabetes function, postural instability, gait abnormalities,
nerve palsies, ulcers, burns, infections, gan-
Height
grene, and Charcot’s disease. DSPN is the most
Modifiable risk factors common cause of foot ulceration and lower
Poor glucose control (HbA1c, FPG) extremity amputation. It is also a major contrib-
utor to falls and fractures [12,13].
Obesity (BMI, weight)
Neuropathic pain can cause physical and psy-
Abdominal Obesity chosocial impairment, disability, and reduced
Dyslipidemia (high LDL, hypertriglyceridemia Low HDL) health-related quality of life. A minority number
of patients are seen with anorexia, depression,
Hypertension
and weight loss [12,13].
Smoking The development of DSPN is insidious, usu-
Heavy alcohol intake ally starting from the distal and progressing
proximally to the extremities. Toes and under
BMI, Body mass index; FPG, fasting plasma glucose; LDL, low-density
cholesterol; HbA1c: hemoglobin A1c; HDL, high-density cholesterol.
feet affected first, gradually ascending the leg.
Fingertips and hand symptoms usually occur
when distal extremity symptoms reach at knee
Cigarette smoking and alcohol consumption level. Characteristic "stocking-glove" like distri-
are all considered independent risk factors bution of neuropathic symptoms present in
for diabetic neuropathy [24]. Most risk factors long-term diabetic patients [25].
for diabetic neuropathy are modifiable and can Early symptoms occur in the DSPN disease
be controlled or treated, while age, duration course due to small nerve involvement
of diabetes, and height are not. Risk factors followed by large nerve dysfunction. Small-
for diabetic neuropathy are summarized in and large-fiber dysfunction most commonly
Table 1.2. coexist.
DSPN manifests with either pain or gives
rise to negative symptoms, but many patients
Symptoms and signs of diabetic neuropathy experience both [25,26].
Common symptoms associated with small
The phenotype of diabetic neuropathy is het-
nerve dysfunction are pain and dysesthesias
erogeneous with diverse clinical manifestations;
(unpleasant sensations of burning). Painful
some patients with diabetic neuropathy have
DSPN may also be described as a sensation of
few complaints, but their physical examination
electricity, shooting pain, contact hyperalgesia,
reveals mild to moderately severe sensory loss.
burning, lancinating, tingling, which tend to
At the other end, patients may have a severe
occur or worsen most at night [12,16]. Pain
neuropathic deficit with no symptoms [12,25].
evoked by contact, for example, with socks,
shoes, and bedclothes (allodynia). Small fiber
Generalized diabetic neuropathy
involvement might also give rise to negative
Distal symmetric polyneuropathy symptoms, with selective loss of temperature
Distal symmetric polyneuropathy is the most and pain sensation [25]. Large-fiber dysfunc-
common form among diabetic neuropathies tion is characterized by may cause numbness,

Diabetic Neuropathy
 Classification 5
tingling without pain, and loss of protective promote dysphagia, regurgitation, esophageal
sensation is a risk factor for diabetic foot ulcer- erosion, and strictures [16]. Delayed gastric
ation. Patients frequently complain that their emptying and gastric retention can cause early
feet feel like they are wrapped in wool or walk satiety, cramping, bloating, epigastric pain
on thick socks. Large-fiber dysfunction is a risk (heartburn), nausea and vomiting, and loss of
factor of falls as the gait might be insecure, appetite to the point of anorexia [10,30].
either wide-based or high stepping [16,25]. Delayed gastric emptying and gastric reten-
tion may impair medication absorption and
Autonomic neuropathy prone patients to hypoglycemia [27]. Colon
Autonomic neuropathy may be observed at abnormalities may result in severe constipa-
any stage of diabetes. Usually, it develops in tion, diarrhea, and fecal incontinence [30].
patients who have had the disease for 20 years. Symptoms of gastrointestinal autonomic neu-
The sympathetic, parasympathetic, and enteric ropathy do not typically occur until later in the
nerves are affected in diabetic autonomic neu- course of diabetes.
ropathy [12]. Genitourinary autonomic neuropathy affects
Cardiovascular autonomic neuropathy is a sexual function and bladder problems which
crucial cause of morbidity and mortality in negatively affect quality-of-life issues in
type 1 and type 2 diabetic patients. Its associa- diabetes.
tion with increased incidence of malignant Erectile dysfunction can also be mediated by
arrhythmia and sudden death was demon- autonomic dysfunction and occurs in approxi-
strated in diabetes [27]. mately 35% to 90% of men with diabetes [10]. In
In type 1 diabetes, autonomy imbalance women, diabetic neuropathy may cause vaginal
may result in a prolonged Q.T. interval on dryness, dyspareunia, and reduced libido [27].
the electrocardiogram, which may predispose Bladder dysfunction is more frequently
the patient to life-threatening cardiac arrhyth- observed in type 1 diabetic patients than type 2
mias and sudden death [27]. Diabetic neuropa- diabetes.
thy also can reduce appreciation of ischemic Impaired bladder sensation may result in
pain, which may delay the diagnosis of silent retention and incomplete voiding of urine,
or asymptomatic myocardial infarction [28]. overflow incontinence, and urinary tract infec-
Clinical findings include exercise intolerance, tions [10]. The most common symptoms of
fatigue, syncope or dizziness, persistent sinus bladder dysfunction include dysuria, fre-
tachycardia, decreased heart rate variability, quency, urgency, nocturia, incomplete voiding,
bradycardia, supine hypertension. Vasomotor and urinary incontinence.
neuropathy frequently causes orthostatic hypo- Diabetic Sudomotor Dysfunction: Diabetic
tension [28]. Orthostatic blood pressure autonomic neuropathy initially results in a loss
measurements may be used to evaluate cardio- of thermoregulatory sweating, resulting in
vascular autonomic dysfunction [12,29]. global anhidrosis. Diabetic autonomic neuropa-
Gastrointestinal autonomic neuropathy may thy can also cause hyperhidrosis, Gustatory
cause paresis anywhere in the gastrointestinal sweating, and abnormal sweat production that
system. The most common symptoms are appears over the face, head, neck, shoulders,
gastroparesis, dysphagia, diabetic diarrhea, and chest after consuming food [31].
and constipation [30]. Delayed esophageal Symptoms and consequences of distal sym-
transit and gastroparesis occur in approxi- metric diabetic polyneuropathy and diabetic
mately 50% and 40% of long-standing autonomic neuropathy are summarized in
diabetics, respectively. Gastroparesis may Table 1.3.

Diabetic Neuropathy
6 1. Classification, risk factors, and clinical presentation diabetic neuropathy

TABLE 1.3 Symptoms and deficits of distal symmetric diabetic polyneuropathy and diabetic autonomic neuropathy.
Symptoms Deficits

Distal symmetric Sensory symptoms: pain, paresthesia, numbness, tingling, Charcot Or Diabetic Foot
polyneuropathy burning, electric shocks, stabbing, ramping, nighttime falls, Syndrome
antalgic gait Foot Ulcers
Muscular symptoms: muscle weakness, atrophy, balance Lower Extremity Amputation
problems, ataxic gait Falls And Fractures
Depression
Cardiovascular Tachycardia, Supine hypertension, Exercise intolerance, Syncope, Painless myocardial infarction
autonomic lightheadedness, Arrhythmia, fatigue, sustained heart rate, Malignant arrhythmias
neuropathy dizziness, lightheadedness Sudden death
Gastrointestinal Dysphagia, bloating, nausea and Gastroesophageal reflux
autonomic vomiting, cramping, diarrhea, constipation, epigastric pain disease
neuropathy (heartburn), loss of appetite, early satiety, fecal incontinence Esophageal erosion and
strictures
Anorexia
Delaying the absorption of
medication
Increased variability in blood
sugar, and hypoglycemia
Genitourinary Nocturnal frequency and urgency Urinary hesitancy, urinary Reduced quality of life
autonomic incontinence Prone to Urinary tract
neuropathy erectile dysfunction decreased libido, retrograde ejaculation, infections
decreased sexual desire

Rare forms of generalized neuropathy Diabetic neuropathic cachexia is a rare type of

Acute sensory diabetic neuropathy is a rare symmetrical painful neuropathy that mainly
form of neuropathy that consists of two sub- affects the lower limbs and the lower trunk. It
types: treatment-induced neuropathy and dia- is characterized by unintentional severe (weight
betic neuropathic cachexia. loss 10% of baseline body weight) [32].
Treatment-induced neuropathy is an iatro-
genic form of neuropathy. It typically manifests Diabetic mononeuropathies
within days or weeks after rapid glycemic con- Cranial mononeuropathies occur in older
trol with severe burning pain, hyperalgesia, or diabetic patients with a long duration of diabe-
allodynia, and symptoms and signs of autonomic tes and are extremely rare. Cranial neuropathy
dysfunction. Rapid glycemic control, a decrease commonly involves the oculomotor, facial,
in hemoglobin (HbA1c) 2% (22 mmol/mol) per 3 and trochlear nerves in diabetic patients. Third
months associated acute sensory symptoms. nerve palsy with pupillary sparing is the hall-
Decreasing HbA1c 5% (55 mmol/mol) increases mark of diabetic oculomotor palsy [32].
the risk of developing treatment-induced neu- Mononeuropathies have an acute onset, are
ropathy by 90% [32]. Rapid glycemic control on associated with pain, and have a self-limiting
any therapeutic regimen can be associated with course resolving in two months. Mononeuro-
treatment-induced neuropathy. pathies can involve median, ulnar, radial, and

Diabetic Neuropathy
 Classification 7
common peroneal nerves. Mononeuropathies neuropathy. Radiculoplexus neuropathies share
often emerge during periods of transition in the similar clinical, neurophysiological, and neuro-
diabetic illness, for example, after an episode of pathological characteristics can occur alone or
hyper- or hypoglycemia, when insulin treatment in combination in people with diabetes [32].
is initiated or adjusted, or when there has been Painful unilateral or multiple asymmetrical
rapid weight loss [33]. neuropathies tend to occur in older patients
When multiple nerves are involved, the with relatively mild or even unrecognized dia-
term "mononeuropathy multiplex" is used. betes in radiculoplexus neuropathies pain fol-
Vasculitis should be ruled out as a cause of the lowed by muscle weakness. Although pain is
symptoms [29]. initially the worst symptom, weakness and
Entrapment neuropathies and their associa- atrophy become the main problems [33].
tion with diabetes have been well established. Diabetic lumbosacral radiculoplexus neu-
It has been estimated that up to 1/3 of people ropathy, also known as diabetic amyotrophy, is
with diabetes may have entrapment syndrome. the most frequent subtype [35]. It is a rare type
However, entrapment neuropathies associated of neuropathy affecting approximately 1% of
with diabetes are often asymptomatic and dis- middle-aged people with type 2 diabetes
covered by alterations at nerve conduction with mild impairment of glucose metabolism
studies [32]. The most common entrapment and absence of other microvascular complica-
neuropathies in diabetes are carpal tunnel syn- tions. It usually begins abruptly and presents
drome (30% of people with diabetic polyneuro- with severe, deep thigh pain, progressing
pathy and 14% of people without diabetic to weakness and muscle atrophy. Symptoms
polyneuropathy), ulnar neuropathy with an present abruptly, unilaterally, and proximally.
observed incidence of 2.1% [34]. Lumbosacral radiculoplexus neuropathy is
Carpal tunnel syndrome presents with characterized by severe and often debilitating
numbness, pain, or tingling in the area inner- motor involvement manifesting with muscle
vated by the median nerve while motor weak- weakness and atrophy. It is associated with
ness is uncommon, thenar muscle wasting significant morbidity and increased psycholog-
may be observed, especially in the elderly ical burden due to intense pain and anxiety
Electrophysio-logical studies show reduced of disease progression, leading to depression
median nerve sensory and motor conduction [33,35].
velocity [34]. Prognosis is favorable; it is a monophasic
The presence of diabetic polyneuropathy disease that improves over time usually lasts
has been associated with an increased risk of two years with a steady but slow improve-
developing entrapment syndromes. ment. The diagnosis of diabetic lumbosacral
Rarely phrenic nerves may be affected in the radiculoplexus neuropathy is predominantly
course of diabetes mellitus; There are case clinical [33].
reports of bilateral phrenic neuropathy leading Diabetic thoracic radiculoneuropathy affects
to diaphragmatic paralysis, where subjects pre- motor, sensory, and autonomic nerve fibers and
sented with orthopnea and respiratory failure represents a rare cause of chronic abdominal
[32,34]. pain with unknown prevalence. It primarily
affects people with type 2 diabetes, irrespec-
Diabetic Radiculopathies tively of their glycemic control and the concomi-
The spectrum of diabetic radiculoplexus tance of other microvascular complications [33].
neuropathies consists of three subtypes: lum- Diabetic cervical radiculoplexus neuropathy
bosacral, thoracic, and cervical radiculoplexus manifests with pain accompanied by weakness

Diabetic Neuropathy
8 1. Classification, risk factors, and clinical presentation diabetic neuropathy

and may affect the entire brachial plexus or the [7] Mishra SC, Chhatbar KC, Kashikar A, Mehndiratta A.
upper limbs. Presentation is usually unilateral Diabetic foot. BMJ. 2017;359(Suppl 1):j5064.
[8] Williams R, Karuranga S, Malanda B, Saeed P, Basit
and may progress to bilateral in some cases. A, Besançon S, et al. Global and regional estimates
Diabetic cervical radiculoplexus neuropathy and projections of diabetes-related health expendi-
affects motor, sensory and autonomic nerve ture: Results from the International Diabetes
fibers and shares clinical and neurophysiologi- Federation Diabetes Atlas, 9th edition. Diabetes Res
cal findings [36]. Clin Pract 2020;162:108072.
[9] Shaw JE, Zimmet PZ. The epidemiology of diabetic
neuropathy. Diabetes Rev 1999;7:245 52.
[10] Duby JJ, Campbell KR, Setter MS, White RJ,
Conclusion Rasmussen KA. Diabetic neuropathy: an intensive
review. Am J Health Syst Pharm 2004;61(2):160 73.
Diabetic neuropathy is the most debilitating [11] Thomas PK. Classification of the diabetic neuropa-
thies. In: Gries FA, Cameron NA, Low P, Ziegler D,
complication that can cause significant morbid- editors. Textbook of diabetic neuropathy. Stuttgart:
ity and mortality in type 1 and type 2 diabetic Thieme; 2003. p. 175 7.
patients. It is essential to detect symptoms or [12] Pop-Busui R, Boulton AJM, Feldman EL, Bril V,
signs of diabetic neuropathy and determine Freeman R, Malik RA, et al. Diabetic neuropathy:
a position statement by the American Diabetes
risk factors as early as possible to implement
Association. Diabetes Care 2017;40(1):136 54.
interventions and to prevent further neuronal [13] Vinik AI, Nevoret ML, Casellini C, Parson H. Diabetic
damage. Increasing awareness and knowledge neuropathy. Endocrinol Metab Clin North Am 2013;42
among healthcare professionals about the man- (4):747 87.
agement of diabetic foot complications is vital [14] Callaghan BC, Little AA, Feldman EL, Hughes RAC.
Enhanced glucose control for preventing and treating
to prevent lower leg amputations.
diabetic neuropathy. Cochrane Database Syst Rev
2012;6:1 60.
[15] Ang L, Jaiswal M, Martin C, Pop-Busui R. Glucose
References control and diabetic neuropathy: lessons from recent
[1] Saeedi P, Petersohn I, Salpea P, Malanda B, Karuranga large clinical trials. Curr Diab Rep 2014;14(9):1 15
S, Unwin N, et al. Global and regional diabetes preva- 528.
lence estimates for 2019 and projections for 2030 [16] Zakin E, Abrams R, Simpson DM. Diabetic neuropa-
and 2045: Results from the International Diabetes thy. Semin Neurol 2019;39(5):560 9.
Federation Diabetes Atlas, 9th edition. Diabetes Res [17] Callaghan BC, Xia R, Banerjee M, Rekeneire N, Harris
Clin Pract 2019;157:107843. TB, Newman AB, et al. Metabolic syndrome compo-
[2] American Diabetes Association. Classification and nents are associated with symptomatic polyneuropa-
Diagnosis of Diabetes: Standards of Medical Care in thy independent of glycemic status. Diabetes Care
Diabetes. Diabetes Care 2021;44(Supplement 1):S15 S3. 2016;39:801 7.
[3] Atkinson MA, Eisenbarth GS, Michels AW. Type 1 [18] Feldman EL, Callaghan BC, Pop-Busui R, Zochodne
diabetes. Lancet. 2014;383:69 82. DW, Wright DE, Bennett DL, et al. Diabetic neuropa-
[4] Lee R, Wong TY, Sabanayagam C. Epidemiology of thy. Nat Rev Dis Prim 2019;5(1):1 19.
diabetic retinopathy, diabetic macular edema and [19] Callaghan BC, Xia R, Reynolds E, Banerjee M, Rothberg
related vision loss. Eye Vis 2015;2:17. AE, Burant CF, et al. Association between metabolic syn-
[5] Adler AI, Stevens RJ, Manley SE, Bilous RW, Cull CA, drome components and polyneuropathy in an obese
Holman RR, et al. Development and progression of population. JAMA Neuro l73 2016;12:1468 76.
nephropathy in type 2 diabetes: the United Kingdom [20] Callaghan BC, Gallagher G, Fridman V, Feldman EL.
prospective diabetes study (UKPDS 64). Kidney Int Diabetic neuropathy: what does the future hold?
2003;63(1):225 32. Diabetologia 2020;63(5):891 7.
[6] Davies M, Brophy S, Williams R, et al. The preva- [21] Forrest KY, Maser RE, Pambianco G, Becker DJ,
lence, severity, and impact of painful diabetic Orchard TJ. Hypertension as a risk factor for diabetic
peripheral neuropathy in type 2 diabetes. Diabetes neuropathy: a prospective study. Diabetes. 1997;46
Care 2006;29:1518 22. (4):665 70.

Diabetic Neuropathy
 References 9
[22] Lee KA, Park TS, Yong Jin H. Non-glucose risk factors [29] Aring AM, Jones DE, Falko JM. Evaluation and pre-
in the pathogenesis of diabetic peripheral neuropathy. vention of diabetic neuropathy. Am Fam Physician
Endocrine. 2020;70(3):465 447. 2005;71:2123 8.
[23] Ansquer JC, Foucher C, Aubonnet P, Le Malicot K. [30] Wolosin JD, Edelman SV. Diabetes and the gastroin-
Fibrates and microvascular complications in diabetes- testinal tract. Clin Diabetes 2000;18:48 51.
insight. Curr Pharm Des 2009;15(5):537 52. [31] Deli G, Bosnyak E, Pusch G, Komoly S, Feher G.
[24] Adler AI, Boyko EJ, Ahroni JH, Stensel V, Forsberg Diabetic neuropathies: diagnosis and management.
RC, Smith DG. Risk factors for diabetic peripheral sen- Neuroendocrinology 2013;98:267 80.
sory neuropathy. Diabetes Care 1997;20:1162 7. [32] Samakidou G, Eleftheriadou I, Tentolouris A, Papanas
[25] Gylfadottir SS, Weeracharoenkul D, Toft Andersen S, N, Tentolouris N. Rare diabetic neuropathies: It is not
Niruthisard S, Suwanwalaikorn S, Staehelin Jensen T. only distal symmetrical polyneuropathy. Diabetes Res
Painful and non-painful diabetic polyneuropathy: clin- Clin Pract 2021;177:1 10.
ical characteristics and diagnostic issues. J Diabetes [33] 46 Ropper AH, Samuels MA, editors. Peripheral nerve
Investig 2019;10(5):1148 57. diseases in Adams and Victor’s Principles of
[26] Tesfaye S, Boulton AJM, Dyck PJ, Freeman R, Horowitz Neurology. New York: McGraw-Hill; 2009.
M, Kempler P, et al. Diabetic neuropathies: update [34] Smith BE. Focal and entrapment neuropathies.
on definitions, diagnostic criteria, estimation of severity, In: Zochodne DW, Malik RA, editors. Handbook of
and treatments Toronto Diabetic Neuropathy Expert Clinical Neurology Elsevier, 126. 2014. p. 31 43.
Group Diabetes Care 2010;33(10):2285 93. [35] Llewelyn D, Llewelyn JG. Diabetic amyotrophy: a
[27] Vinik AI, Maser RE, Mitchell BD, Freeman R. Diabetic painful radiculoplexus neuropathy. Pract Neurol Apr
autonomic neuropathy. Diabetes Care 2003;26:1553 79. 2019;19(2):164 7.
[28] Ziegler D. Cardiovascular autonomic neuropathy: clin- [36] Laughlin RS, Dyck PJ. Diabetic radiculoplexus neurop-
ical manifestations and measurement. Diabetes Rev athies. Handb Clin Neurol 2014;126:45 52.
1999;7:342 57.

Diabetic Neuropathy
 C H A P T E R

2
Pathologic basis for diabetic neuropathy
in humans
Soroku Yagihashi1,2
1
Department of Exploratory Medicine on Nature, Life and Man, Toho University School of Medicine,
Inage-cho, Inage-ku, Chiba, Japan 2Department of Pathology and Molecular Medicine, Hirosaki
University Graduate School of Medicine, Hirosaki, Aomori, Japan

Introduction descriptive term meaning of a demonstrable

disorder, either clinically evident or subclini-
Dr. Max Ellenberg, who dedicated his life cal, that occurs in the setting of diabetes
to the clinical care for patients with diabetic without other causes for peripheral neuropa-
neuropathy in the mid 20th century, con- thy” [2]. This definition was succeeded by
cluded that neuropathy is not a complication subsequent panel conferences, and the entity
of diabetes but a concomitant with the same of neuropathy includes all manifestations in
[1]. This foresight has been substantiated by the somatic and/or autonomic parts of the
the following clinical evidence that (1) neu- peripheral nervous system [3]. Hence neu-
ropathy may be the initial manifestation of ropathy may not simply be “one of the com-
diabetes, (2) neuropathy is the most preva- plications” but is likely to represent diabetes
lent among individuals with comorbidities or itself.
various complications, (3) neuropathy is Patients with neuropathy suffer from
often present in prediabetic stage, (4) neurop- intractable pain, uncomfortable sensation,
athy occurs in the absence of overt hypergly- severe autonomic imbalance, and lowered qual-
cemia or during good control, (5) signs and ity of life. Nearly half of the patients with diabe-
symptoms of neuropathy are not parallel tes are affected with such neuropathic
with severity of diabetes or duration of dia- complications and often require expensive limb
betes. Despite the advancement in knowledge amputation [4]. Care for neuropathy is critical
concerning diabetic neuropathy, its complex- to the management of diabetes not only because
ity continues to be highly perplexing. In of its intractable symptoms but also of serious
February 1988, San Antonio consensus panel impact on life expectancy [5,6]. With longer
announced that “diabetic neuropathy is a duration of diabetes, the incidence of

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00003-7 11 © 2022 Elsevier Inc. All rights reserved.
12 2. Pathologic basis for diabetic neuropathy in humans

neuropathy is increased [7,8]. Neuropathy man- neuropathy. This type is commonly accompa-
ifests early in prediabetic stage in type 2 diabe- nied by involvement of autonomic nervous
tes (T2D). Because of the insidious nature of its system of both sympathetic and parasympa-
symptoms, accurate pathologic features in early thetic nerves [15]. The presence of cardiac auto-
neuropathy are not well established, although nomic neuropathy markedly reduces the
involvement of small nerve fibers has recently quality of life and shortens life expectancy
been proposed. Currently, there is no effective [16,17]. Distal sensory predominant neuropa-
means for the prevention and suppression of thy usually accompanies autonomic neuropa-
the development of neuropathy except for thy, and this combination is collectively called
meticulous blood glucose control. For a better diabetic polyneuropathy (DPN).
understanding of the pathophysiology of dia- Focal type may be subclassified into mono-
betic neuropathy and determination of the treat- neuropathy and multiple mononeuropathies
ment design, it will be useful to review the (mononeuropathy multiplex). Mononeuropathy
pathologic basis of diabetic neuropathy. is typical for cranial neuropathies (oculomotor
nerve, abducens nerve, etc.) presenting ophthal-
moplegia [18]. Less clearly differentiated are
Classification of diabetic neuropathy truncal neuropathy and proximal motor neurop-
athy of the arm or the thigh (diabetic amyotro-
Diabetic neuropathy is a complex disorder phy) from multiple mononeuropathies. The
with metabolic and vascular basis. Long-term latter is now regarded as lumbosacral radiculo-
hyperglycemia or hyperlipidemia exerts plexus neuropathy (LSRPN) [13,19,20]. Mode of
chronic injurious and reparative remodeling of the onset may be divided into acute, subacute, or
the peripheral nerve tissues. In this setting, chronic, the former of which is atypical com-
degenerative, regenerative, inflammatory pared to chronically developed DPN. Motor type
lesions develop diffusely in centripetal manner of cranial neuropathy occurs as sudden or rapid
on the basis of metabolic aberration [9,10]. In onset of painful oculomotor palsy [18]. In con-
contrast to metabolic deficits, compromised trast, proximal motor neuropathy manifests
vascular supply triggers acute infarction or acute or subacute focal painful muscle atrophy
chronic ischemic changes, resulting in mainly in the arm, thigh, or thorax [20,21]. In cases with
focal or multifocal pathological phenotype obvious muscle atrophy in extremities or trunk,
[11 13]. Thomas divided diabetic neuropathy it was used to have the term of diabetic amyotro-
into two major types of diffuse and focal neu- phy [22]. This terminology is not used nowadays
ropathy [14] (Table 2.1). Diffuse type repre- because it does not indicate the accurate causa-
sents distal symmetric sensory predominant tion. Proximal motor neuropathy is associated
with epineurial vasculitis with a pattern of multi-
ple mononeuropathies [19,23,24]. In this case,
TABLE 2.1 Classification of diabetic neuropathy. neurologic deficits involve proximal radiculo-
plexus neuropathy with motor deficits. The asso-
1. Diffuse type neuropathy
(1) Diabetic sensori-motor (distal symmetric) ciation of vasculitis with muscle atrophy
neuropathy (DSN) suggests immunogenic basis in its pathogenesis
(2) Autonomic neuropathy (DAN) [19,23]. There remain some ambiguities and com-
2. Focal type neuropathy plexities for the clinical phenotype and their
(1) Mononeuropathy; Cranial nerve plasy
backgrounds. Nevertheless, efforts for the sepa-
(2) Multiple mononeuropathy complex;
Lumboscarcalradiculoplexoneuropathy (LSRPMN) ration of clinical phenotype of neuropathy are
meaningful and important for the understanding

Diabetic Neuropathy
 Basic pathologic changes in several types of neuropathy 13
of the disease and for the decision of treatment autonomic nerves and its reactive growth
design. resembling traumatic neuroma [32]. Most of
the changes were not specific for diabetes but
merely reactive process to chronic injury.
Basic pathologic changes in several types Later systematic investigations of autopsy
of neuropathy materials embedded in epon after fixation with
glutaraldehyde and osmium provided more
Pathological basis of DPN appears to be dis- precise spatial distribution of nerve lesions in
tinct in each type of neuropathy. There may be the peripheral nervous system. On the semi-
some overlap or transient form of the disorder. thin cross-sections of the sciatic, tibial, and sur-
For the convenience of understanding, repre- al nerves, nerve fiber loss with thickened
sentative features will briefly be introduced in endoneurial vascular walls was consistently
each type, and the implication of the lesions demonstrated [12,33,34]. The results from these
related to the clinical phenotype will be studies were informative, but the subjects
discussed. investigated were limited and often compli-
Classical pathological studies on diffuse cated with uremia, and therefore causal rela-
type of DPN were conducted on the materials tionships of nerve lesions with clinical
obtained at the time of autopsy [25 28]. Most phenotype were not well addressed [34]. Such
conspicuous changes were detected in the systematic studies are extremely valuable but
peripheral nerve showing loss of nerve fibers, too laborious to collect the data from sufficient
demyelination/remyelination and fibrosis. The number of subjects to draw an appropriate
changes were more robust in the dorsal root conclusion.
compared to ventral root, suggesting sensory Following the autopsy studies, investiga-
predominance of the lesion [26]. Dorsal column tions on the biopsied sural nerve samples from
underwent demyelination and gliosis, but patients with clinical information on neuropa-
there were not any substantial changes in dor- thy were commenced at the light and electron
sal root ganglia [27]. Sympathetic and para- microscopic levels. To this end, freshly
sympathetic nerves contained nerve fibers with obtained sural nerve tissues were applied to
degeneration and demyelination [29,30]. teased nerve fiber studies and light micro-
Ganglion cells in celiac and mesenteric ganglia scopic observations on the semithin thick
were vacuolated and markedly swollen [31]. epon sections stained with toluidine blue or
Autonomic nerve alterations supplying the phenylenediamine [35 38]. The ultrathin sec-
internal organs also indicated the long-term tions were investigated using electron micro-
presence of structural derangement in autopsy scope after uranyl acetate staining. Teased fiber
cases of diabetic patients [31,32]. Despite the method was developed for the identification of
presence of apparent pathological changes myelinated fibers for axonal degeneration,
such as neuronal loss and reactive gliosis in the demyelination, and remyelination [35,36,38,39].
spinal cord or dorsal root ganglia, it was not Semithin sections were suited for the morpho-
clear whether the changes really represented metric analysis of myelinated fibers [38],
the alterations caused by diabetes or other fac- whereas unmyelinated nerve fibers and fine
tors because of the lack of precise clinical infor- subcellular changes were depicted only by
mation. Common appearance of neuromatous electron microscopic observations [40,41].
changes showing bundle of proliferative nerve Despite considerable information obtained
fasciculi in gastrointestinal walls or vascular from the biopsied samples, this method is not
walls suggested the repeated injury to internal recommended any more in clinical practice

Diabetic Neuropathy
14 2. Pathologic basis for diabetic neuropathy in humans

because of the invasiveness and postsurgical repeated investigations [51 53]. The topic on
complication of sensory loss in the lateral side CCM will be precisely described in another
of the foot, except for the exceptional occasion chapter in this book (Chapter 4). General out-
of necessity for differential diagnosis or special line of pathological characteristics will be
research purpose [42,43]. Alternative site of described in each type of neuropathy in the fol-
biopsy with less troublesome complications lowing sections.
was also introduced in case of the necessity of
structural analysis of the peripheral nerve [44].
Currently, less invasive methods for patho-
logical investigations on the skin biopsy have
Sensory nerve pathology
evolved as a standard for detection of neuropa-
thy in diabetes [45 48]. Immunohistochemical
Distal nerve fiber loss
staining with antibodies against protein G Salient feature of the peripheral nerve in
product 9.5 (PGP9.5) as well as the use of mar- DPN is nerve fiber loss with distal predomi-
kers for basement membranes or microvessels nance [13,20,38] (Fig. 2.1). Remnant fibers also
enable the evaluation of small nerve fibers (C- undergo axonal degeneration and regenera-
fibers in the epidermis and of Aδ fibers in the tion, segmental demyelination, and remyelina-
dermis). With the advent of this method, early tion. With regard to the length, distal sensory
reduction of intraepidermal nerve fiber density sural nerve originating from the sciatic nerve
(IENFD) as an index of neuropathy was con- usually displays the most severe changes
firmed in subjects with prediabetes [49,50]. [20,34]. Deficit of regenerative capacity acceler-
More recently, corneal confocal microscopic ates fiber loss in diabetes.
(CCM) observations on the cornea have devel- Thomas [36] first attempted to connect the
oped for the detection of small nerve fiber pathologic basis with the clinical signs and mea-
abnormalities in patients with diabetes. This sures of neuropathy. He concluded that the fiber
method is not invasive and feasible for changes did not well explain the clinical signs

FIGURE 2.1 Sural nerve pathology in a patient

with diabetic neuropathy. Cross-sectional view of
sural nerve obtained from 57 year-old man with
overt diabetic neuropathy, showing marked reduc-
tion of myelinated nerve fibers. Remaining nerve
fibers show active axonal degeneration (blue arrow)
and small cluster of regenerating nerve fibers (green
arrow). Walls of endoneurial microvessels are thick-
ened with narrowing of the lumen (red arrow).

Diabetic Neuropathy
 Sensory nerve pathology 15
and symptoms but well corresponded to the con- even in the absence of the nerve fibers (phantom
duction delay. Behse et al. [38] provided a good pain) [62] (Fig. 2.2). The nerve pathology depicted
correlation of myelinated fiber loss in the biop- so far, however, is only cross-sectional in nature,
sied sural nerve with the severity of nerve con- and therefore no sequential information is avail-
duction deficits. Loss of nerve fibers was not able regarding the way of progression.
limited to specific population of nerve fibers but There are some exceptional studies employing
involved large and small myelinated and unmy- repeated sural nerve biopsy to explore the treat-
elinated fibers referred to as panmodal nerve ment effects on DPN by measuring MNFD prior
fiber loss. They considered that 20% 30% of to and after the trial [63,64]. Without treatment,
the conduction delay was not explained by the there was approximately 1% 2% decline of
nerve fiber loss. Teased fiber studies showed MNFD after 52 weeks in diabetic subjects. Since
predominantly segmental demyelination and percentage change of MNFD (ΔMNFD) included
remyelination [35], once led to the hypothesis lost nerve fibers (5% 6%) along with regenerat-
that diabetic neuropathy is Schwann cell disease. ing fibers (4% 5%), set off number of changed
With advancement of nerve physiology technol- nerve fibers among remaining fibers was approx-
ogy, fiber loss well corresponded with the reduc- imately 1% 2% either increased or decreased
tion of evoked amplitudes (sensory nerve action (which is equal to the sum of the number of
potentials, i.e., SNAP or compound muscle regenerating nerve fibers 2 lost myelinated nerve
action potentials, i.e., CMAP) in diabetic patients fibers). As results, they demonstrated significant
with overt neuropathy [54 56]. It was shown suppressive effects of progressive MNFD decline
that even under good blood glucose control there in diabetic patients treated with aldose reductase
occurred progressive decline of CMAP or SNAP inhibitor (ARI) sorbinil compared to placebo
over 4 7 years of observation period [56,57]. group [63]. In those studies, they found 3.8 times
There was no significant correlation, however, increase in the fraction of regenerated nerve
between measures of nerve fiber size, g-ratio fibers with 33% increase in total. It is thus con-
(axon/fiber size ratio), electrophysiologic para- ceivable that the effects of sorbinil were mainly
meters, and clinical indices of sensory tests or the promotion of the nerve fiber regeneration
painful symptoms [58]. [63]. Unfortunately, sorbinil was dropped out
The degree of fiber loss became more pro- from the market because of serious adverse
nounced with advancing stage of neuropathy effects. In contrast, another ARI (zenarestat) treat-
[59 61]. The extent of nerve fiber loss was quanti- ment did not exhibit any significant increase in
tatively expressed by the reduction of the MNFD with no difference between treated and
myelinated nerve fiber density (MNFD), which untreated group[64]. Since precise analysis of
represented the number of myelinated nerve fibers nerve fiber caliber spectrum was not presented, it
within the unit area of each fascicle expressed as is not clear whether small regenerating fibers
#/mm2. This measure well correlated with clinical were in fact increased or not or whether loss of
staging of diabetic neuropathy [59]. The nerve large fibers was protected.
fiber loss appeared to clinically manifest with the Nerve biopsies are too invasive merely for the
appearance of negative symptoms such as loss of evaluation of treatment effects and are thus not
protective sensation or numbness [62]. In contrast, recommended nowadays. However, recent
positive subjective symptoms such as pain or par- microarray or proteomics analyses on the stored
esthesia triggered by the signals released from samples obtained from clinical trials provided
stimulated fibers, or spinal cord, or brain in valuable information on the question whether
chronic state, appeared at any time when the there are responders or nonresponders or how
impulse was evoked from remnant nerve fibers or progressive fiber loss is related to genetic,

Diabetic Neuropathy
16 2. Pathologic basis for diabetic neuropathy in humans

FIGURE 2.2 Natural history of neuropathy in diabetes. Distal predominant nerve fiber loss is a consistent feature of dia-
betic neuropathy. In concert with the advancement of neuropathy with increasing duration of diabetes, nerve fibers progres-
sively disappear. Dependent on the nerve fiber loss, loss of sensation in the distal limb may be recognized as a sign of negative
symptoms which patients do not complain. In contrast, positive symptoms such as pain and paresthesia which patients com-
plain appear any time during the course when injured nerve fibers are stimulated to release impulses to spinal cord and brain.

inflammatory, or metabolic factors [65,66]. Details sciatic nerve suggested a vascular basis for dia-
of the findings are reported in the literature, but betic neuropathy [13]. The focality of the lesions
the relationships of genes and molecules to path- was more prominent in the proximal nerve but
ological features are yet to be explored. less obvious in the distal side [67]. Such distribu-
tion of nerve fiber loss was postulated to be
ascribed to the occlusive or obstructive lesions of
Metabolic and vascular basis for nerve fiber nutrient vasa nervorum [68,69]. It was also spec-
loss and heterogeneity of neuropathy ulated that such distribution reflected ischemic
In contrast to generalized loss of myelinated lesions of myelinated fibers undergoing axonal
fibers in distal sural nerve, the spatial distribu- degeneration in the proximal site and summa-
tion of focal nerve fiber loss in the proximal tion of the focality became gradually diffuse fiber

Diabetic Neuropathy
 Sensory nerve pathology 17
loss in the distal portion (Fig. 2.3). To confirm this patchy distribution of fiber loss was compara-
possibility, peripheral nervous system was exam- bly found in a series of sural nerve biopsies in
ined in a rat model of microsphere infarction, type 1 hereditary motor and sensory neuropa-
which displayed vasculitic neuropathy, showing thy (Charcot Marie Tooth disease), which
spatial distribution of nerve lesions, similar to had a genetic basis free from vascular influ-
those depicted in diabetic patients [70]. By means ences, thereby questioning the vascular causa-
of a series of experiments and observations, the tion of diabetic neuropathy [72]. To reconcile
researchers concluded that fiber loss was primary this discrepancy, it was interpreted that Dyck
and vascular in origin, and ischemia/hypoxia was et al. examined mostly older subjects much
a cause of diabetic neuropathy [13,71]. influenced with vascular alterations [68,69],
There were some disagreements against the whereas Thomas and his group examined
vascular hypothesis derived from the focality mainly younger subjects with type 1 diabetes
of fiber loss lesions in diabetic neuropathy. (T1D) [71]. Alternatively, the pattern of fiber
Thomas and his colleagues described that the loss is likely to be different between T1D and

<Proximal>
Compression/Edema
Dorsal root ganglia
Focal fiber loss &
Impaired
Segmental demyelinaon
nerve blood
flow

Wallerian
degeneraon <Median>

Ischemia/ Focal fiber loss &

Sciac nerve trunk reperfusion Segmental demyelinaon

Compression/Edema
Peroneal nerve
<Distal>
Impaired blood flow Diffuse
Posterior bial nerve
fiber loss
Trauma
Hypoxia
Malnutrion
etc
Sural nerve Distal axonal
degeneraon
FIGURE 2.3 Spatial pattern of nerve fiber loss in diabetic neuropathy. Three-dimensional pattern of nerve fiber loss in dia-
betic. In the proximal portion of the sciatic nerve, nerve fiber loss distributes focally starting at mid-thigh region related to endo-
neurial microangiopathy. Fibers distal to the site of ischemia nerve fibers undergo Wallerian degeneration and secondary
demyelination. Summation of focal nerve fiber loss display diffuse nerve fiber loss in the distal side of the nerve (sural nerve)
which also shows many fibers with axonal degeneration and regeneration. Along with the long nerve fasciculi of sciatic nerve,
nerve fiber loss is accompanied by expanded endoneurial space (nerve edema) and increased endoneurial pressure which
makes diabetic nerve susceptible to compression or hypoxia. Source: Modified from Dyck PJ, Lais A, Karnes JL, O’Brien P, Rizza
R. Fiber loss is primary and multifocal in sural nerves in diabetic polyneuropathy. Ann Neurol 1986; 19: 425 439.

Diabetic Neuropathy
18 2. Pathologic basis for diabetic neuropathy in humans

T2D, showing nonuniform pathological fea- and needs to be clarified by future investiga-
tures in diabetes. Sima and their colleagues tions [77].
demonstrated significant differences in the dis-
tribution of nerve fiber loss in the sural nerve
between T1D and T2D when subjects were
Demyelination and ultrastructure of
adjusted to comparable age, duration of diabe-
Schwann cells
tes, severity of hyperglycemia [73]. In the
study, focality of fiber loss was determined by It is worth noting that fiber loss was not
the coefficient variation (CV) of MNFD from always evident as an early pathological feature
area to area of nerve fasciculi, and there was a in diabetic neuropathy. Not in frequently,
significantly greater CV in T2D than that in demyelination and remyelination were com-
T1D. They concluded that the focal nerve fiber monly encountered in teased fiber studies
loss was considered to have a vascular origin [38,78,79]. Thomas and Lascelles reported fre-
in T2D, whereas the changes in T1D were dif- quent appearance of fibers undergoing seg-
fuse, suggesting metabolic origin (Table 2.2). It mental demyelination in addition to the loss of
is worth noting that not limited to hyperglyce- nerve fibers and Schwann cell proliferation in
mia, hyperlipidemia may be invoked in the biopsied radial nerve from eight cases with
development of neuropathy in T2D [74]. To diabetes. Chronic injury to Schwann cells was
confirm the implication of hyperglycemia, suspected to be mainly implicated in the dia-
meticulous blood glucose control effectively betic nerve pathology. Further investigation on
inhibited the progression of neuropathy in sural nerves obtained from 19 relatively young
Diabetes Control and Complication Trial diabetic patients (age: 20 , 60 years) with
(DCCT) study [75]. In contrast to this result, painful and autonomic neuropathy showed
failure of tight blood glucose control for the marked loss of large and small nerve fibers
suppression of neuropathy progression accompanied by fibers with axonal regenera-
observed in patients with T2D in UK tion and remyelination [60]. They found diffi-
Prospective Diabetes Study group investigation culty to connect the pathological findings to
[76] indicated the difference in the pathoge- clinical symptoms. Said et al. studied sural
netic processes of neuropathy between T1D nerve pathology from five young (age:
and T2D. Differences in neuropathy between 20 , 30 years) subjects with T1D with severe
T1D and T2D are still speculative, however, early onset neuropathy [78]. They found 20%

TABLE 2.2 Heterogeneity of diabetic neuropathy.a

Nerve fiber Vascular

Major Fiber Axoglial Cell Basement

cause loss Patternb Atrophy contactc Endothel junctiond membrane

T1D Metabolic diffuse Axonal Yes Impaired Nonoccluded Cohesive Mildly thickened
degeneration

T2D Vascular Focal Wallerian No Equivocal Occluded Separated Thick and

degeneration multiplicated
a
Differences in the structural changes between type 1 (T1D) and type 2 diabetes (T2D) proposed by Sima et al. [73].
b
Pattern; major type of nerve fiber degeneration.
c
Putative structure of axoglial dysjunction proposed by Sima et al. [73].
d
Contact between two endothelial cells with tight junctions by Sima et al. [128].

Diabetic Neuropathy
 Sensory nerve pathology 19
reduction of myelinated nerve fibers, and some showing degenerative, reactive, regenerative, and
with dying-back type, accompanied by 33% proliferative changes [84]. Invagination of
fibers of segmental demyelination and 11% Schwann cell cytoplasm into the axolemma and
axonal degeneration. Often, regenerative and disruption or delamination of myelin lamellae
remyelinated fibers occupied considerable per- were also found. Such disruption of myelin struc-
centage area of nerve fasciculi [80]. Such de- ture probably reflected the process of demyelin-
and remyelination changes may probably be ation [86]. Schwann cell cytoplasm contained
secondary to axonal degeneration because dis- degradation products of disrupted myelin frag-
tal to the site of axonal degeneration, segmen- ments or autophagic vacuoles. Occasionally, mul-
tal demyelination ensues. It may also be tilayered Schwann cell cytoplasmic processes
possible that the changes of myelin sheath surrounded regenerated nerve axons, showing
occur primarily in face of increased endoneur- onion bulb appearances, often called hypertro-
ial pressure or compression effects because phic changes. The regenerative Schwann cells
endoneurial edema is common in diabetic accompanied thickened and elongated projections
peripheral nerve [81,82]. of basement membranes [35,37], which appeared
Using electron microscopy (EM), axonal to result from the renewal or replicated Schwann
degeneration was characterized by depletion of cells containing several small regenerating axons
axonal organelles such as neurofilaments, micro- (Fig. 2.4). Vacuolar cytoplasmic organelles are
tubules, and smooth endoplasmic reticulum (ER), often misinterpreted as pathological signs, but
as well as mitochondria [40] (Fig. 2.4). Oftentimes, they may frequently be artifacts during the pro-
there appeared concurrent aggregation of glyco- cess of fixation. Most of the changes detected by
gen granules, autophagic vacuoles, and phago- EM were not specific nor pathognomonic, simply
somes, as well as engulfed myelin debris. expressed the pathological signs. Variety of axo-
Glycogen accumulation enclosed by limited nal disintegration represented multiple facets of
membrane was considered to be a sign of hypoxic ischemic nerve injury [83,84]. No causal relation-
changes [41,83 85]. Structural changes of ships were obtained for subcellular changes, not
Schwann cells also assumed a variety of patterns indicative of etiology.

FIGURE 2.4 Ultrastructure of sural nerve in a

patient with diabetic neuropathy. Cross sectional
view of sural nerve obtained from a patient with
diabetes. Combination of active nerve fiber degener-
ation and microangiopathic changes of endoneurial
microvessels is characteristic for the feature of dia-
betic neuropathy. Endoneurial microvessel exhibits
occluded lumen by swollen cytoplasm of endothelial
cells (1), and attachment of inflammatory cells and
macrophages. Vascular wall shows thickening and
multiplication of basement membranes (2) in which
atrophic degenerated pericytes are buried with vac-
uolation (3). Unmyelinated nerve fibers also
undergo vacuolation (4) and loss of axons, leaving
elongated basement membranes of atrophic
Schwann cells (5), indicating affected small nerve
fibers.

Diabetic Neuropathy
20 2. Pathologic basis for diabetic neuropathy in humans

Sural nerve biopsy is not routinely con- and intermediate, with dramatically increased,
ducted any more for the purpose of diagnosis, decreased, and steady ΔMFD%, respectively.
but on the special occasion, nerve biopsy is still HbA1c at baseline was the only factor signifi-
required for the diagnosis and decision of cantly different between regenerator and
treatment. When differential diagnosis from degenerator (OR 0.68; P , .01). Microarray
amyloidosis, chronic inflammatory demyelinating analysis revealed that upregulated genes in the
polyradiculoneuropathy (CIDP), and Charcot regenerator group were enriched with cell
Marie Tooth diseases is required, special diag- cycle and myelin sheath function, while down-
nostic procedures such as immunostaining and regulated genes were enriched in immune/
genetic analyses may be needed on the tissue sec- inflammatory responses [66]. These data sug-
tions [13]. gest that HbA1c levels predict myelinated
nerve fiber regeneration and degeneration in
patients with DN.
Impaired regeneration and related factors
Progressive nerve fiber loss in diabetes may
also be attributed to impaired regeneration of Motor nerve pathology
nerve fibers. Evaluation of density of regenera-
tive nerve fiber cluster was undertaken to esti- Large fiber dysfunction is a constant feature
mate the balance between degeneration and of DPN. Previous studies disclosed degenera-
regeneration of myelinated fibers [38,78,80]. In tive changes and nerve fiber loss in the periph-
a sural nerve biopsy study, the number of eral motor nerve [33,83]. Indeed, patients with
regenerating fibers, identified using light long-term diabetes or sustained metabolic
microscopy, was found to decline in propor- aberration usually displayed delayed nerve
tion to the reduction in total MNFD. In con- conduction and lowered action potentials in
trast, number of regenerating fibers relative to both sensory sural nerve and motor tibial nerve
total myelinated fibers was significantly [36,38,54]. The severity of DPN was in fact
greater in patients with T1D compared to T2D represented by the lowered action potentials of
[63,80]. This may be accounted for by younger CMAP which corresponded with the extent of
age population in T1D compared to T2D, while reduced nerve fiber density [34,36]. Consistent
regenerative capacity is less potent in T2D. On with the changes of sensory nerves, motor
the other hand, there was no significant corre- nerves of sciatic nerve or tibial nerve showed
lation between the number of regenerative fibers undergoing active axonal degeneration,
clusters and painful symptoms [60,80]. segmental demyelination and fiber loss [12,
Recently, Feldman’s group studied the 33,81]. The motor nerve involvement in DPN
genetic role in the impaired regeneration of may be reflected by the presence of muscle atro-
nerve fibers in DPN. Hur et al. identified risk phy with distal predominance [87,88].
factors associated with sural nerve fiber loss in Degenerative changes of motor nerves resulted
patients with diabetic neuropathy. In their in progressive muscle atrophy. Motor unit
study, diabetic patients with neuropathy trea- denervation of muscle and decreased muscle
ted with acetyl-L-carnitine for 52 weeks were cell size and diminished capillaries were also
retrospectively examined. Sural nerve biopsy described [89]. Instability, gait disturbances, and
was performed before and after the trial [65]. muscle weakness were major clinical signs of
In accordance with the change of myelinated motor nerve involvement. In the classic litera-
fiber density (ΔMFD%), subjects were divided ture, motor nerve endings in contact with mus-
into three groups of regenerator, degenerator, cles often underwent denervation and synaptic

Diabetic Neuropathy
 Motor nerve pathology 21
degeneration [90]. Disruption of muscle spindles frequent association of microvasculitis in these
and intrafusal fibers were also reported [91]. patients [19,94]. Said et al. documented the
Autopsy-based study disclosed the presence of presence of vasculitis around perineurium in
central chromatolytic neurons in the anterior 10 cases with T2D and painful proximal neu-
horn and reactive gliosis, suggesting retrograde ropathy with muscle atrophy. With vasculitic
nerve fiber degeneration [26,27]. It is not still lesions, peripheral nerves showed nerve fiber
clear, however, whether the changes were pri- loss with features of axonal degeneration and
marily induced by diabetes or were secondary myelin breakdown, indicative ischemic effects
to the wasting of muscle tissues. on the nerve [19] (Fig. 2.5). In their subsequent
Motor predominance of neuropathy may be studies on the diabetic patients with painful
characteristic in some variants of DPN, such as multifocal neuropathy, they found marked reduc-
mononeuropathy or multiple mononeuropathies. tion of nerve fiber density, associated with a nec-
Neuropathy with motor predominance should rotizing vasculitis of perineurial and endoneurial
strictly be differentiated from sensory- blood vessels in the root, plexus, and nerve trunk,
predominant type of DPN because treatment further supporting the immunogenic basis of this
option is different. Mononeuropathy was typi- disorder [20]. Similar findings were detected by
cally characterized by sudden onset of painful Dyck’s laboratory collecting a relatively large
ophthalmoplegia in cases of cranial neuropathy, cohort of patients [23,94] and also reported by
which usually resolved spontaneously without other group [24]. Painless cases were also encoun-
any treatment [92,93]. Thrombotic occlusion of tered [95].
nutrient arteries was considered to be the cause Distinction of CIDP from other motor pre-
by histological observations on the affected nerve. dominant neuropathies is often diagnostic
Motor predominant neuropathy with pain- challenge in patients with diabetes [23]. Both
ful muscle atrophy occurs in elderly people types may exhibit hypertrophic onion bulb
with or without diabetes. In cases with diabe- structures of Schwann cells with marked nerve
tes, this condition was used to be called as dia- fiber loss. Clinical studies disclosed signifi-
betic proximal motor neuropathy or diabetic cantly increased prevalence of CIDP in patients
amyotrophy. Currently, this condition is called with diabetes compared to nondiabetic patients
diabetic lumbosacral radiculoplexus neuropa- [96 98]. Bril et al. reported that the prevalence
thy (DLSRPN) [23]. Biopsy studies disclosed of CIDP with DM was nine times greater
FIGURE 2.5 Microvasculitis of epineurial micro-
vessels encountered in diabetic proximal neuropa-
thy. Proximal motor neuropathy is marked by
robust muscle atrophy with pain in the thigh and
buttocks, now called “diabetic lumbosacral radiculo-
plexoneuropathy.” This type frequently exhibits
small round cell infiltration in epineurial arterial
walls, indicating the pattern of multiple microvascu-
litis. Immunological and inflammatory mechanisms
are suspected to play a role in the pathogenesis.
Sural nerve biopsy usually demonstrates nerve fiber
loss, segmental de- and remyelination.

Diabetic Neuropathy
22 2. Pathologic basis for diabetic neuropathy in humans

compared to those without DM [97]. The rea- axons to sympathetic and parasympathetic
son for this high combination is not clear. postganglionic neurons. Preganglionic sympa-
Other study could not confirm the high preva- thetic fibers constitute synaptic contacts with post-
lence of CIDP in patients with diabetes [99]. It ganglionic neurons where they in turn emanate
may be speculated, however, that deranged postganglionic adrenergic fibers to target organs
immune regulatory system is exerted in the [106 109]. Distal predominance of axonal degen-
presence of chronic metabolic aberration simi- eration is a salient feature accompanied by distal
lar to the association of T1D with the autoim- predominant fiber loss. Changes are comparable
mune disorders such as hyperthyroidism, between sympathetic and parasympathetic ner-
Hashimoto’s thyroiditis or Sjogren’s syndrome vous system. The findings from animal models of
as well. Pathological study demonstrated the diabetes were translated to humans which have
presence of characteristic features of diabetic comparable pathology between rodents and
microangiopathy as well as the prominence of humans [106,107,109]. Basically, most distal por-
fibers with demyelination/remyelination in tions of postganglionic fibers underwent first
patients combined with DPN and CIDP [97,99]. degenerative changes followed by preganglionic
fibers. Submicroscopically, nerve terminals in the
sympathetic nervous system showed swollen
Autonomic nerve pathology axons containing aggregation of microtubules,
smooth endoplasmic reticula and mitochondria as
Involvement of autonomic nerves in both sym- well as neurofilaments and microtubules. These
pathetic and parasympathetic nerves is common features were consistent with axonal dystrophy, a
and typical feature of DPN [10,15]. Fine structural phenomenon of aborted attempts of nerve fiber
studies on autonomic nerves were conducted in regeneration [110] (Fig. 2.6). The axonal dystrophy
a systematic manner in a variety of organs and tis- was frequently encountered in the nerve term-
sues [27,31,100]. Appenzeller and Ogin demon- inals contacting with postganglionic neurons.
strated degeneration of myelinated fibers of Dystrophic changes also manifested in dendrites
paravertebral sympathetic chains [29]. Low et al. and neurites. Increased frequency of dystrophic
depicted degenerative and demyelinating changes axons well correlated with the loss of synaptic
of preganglionic nerve fibers in paraspinal sympa- contacts, reflecting denervation status.
thetic nerve trunk [30]. As to the parasympathetic In some cases with T1D, autoantibodies to
nerves, loss of vagal nerve fibers was demon- ganglion cells were detected [111,112]. The pres-
strated in the wall of esophagus in patients with ence of such autoantibodies was often associated
esophageal dysmotility like achalasia [101]. with symptoms of orthostatic hypotension, car-
Serious condition of painless myocardial infarction diac arrythmia, abnormal sweating and reduced
was accounted for by denervation of cardiac mus- heart rate variability (low CVR-R).
cles [102]. Greater splanchnic nerve underwent
preganglionic sympathetic nerve degeneration Implication of microangiopathy in
[103]. More recently, sweat gland denervation was diabetic polyneuropathy
clearly illustrated in diabetic patients [104]. In gas-
trointestinal tracts, in addition to autonomic nerve
Distinct vascular supply in peripheral
changes, intravisceral myenteric ganglia were
involved with degenerative changes of interstitial
nerve
(Cajal) cells [105]. Vascular supply of peripheral nerve is differ-
Autonomic nervous system originates in the ent from those in central nervous system or mus-
spinal cord or brain stem to send their long cle tissues, showing sparse distribution, wider

Diabetic Neuropathy
 Implication of microangiopathy in diabetic polyneuropathy 23
FIGURE 2.6 Axonal dystrophy in autonomic
neuropathy (mesenteric nerve of diabetic BB rat).
Both sympathetic and parasympathetic nerves are
affected in diabetes and distal nerve terminals fre-
quently undergo degenerative changes with fiber
loss. Axonal dystrophy denotes markedly swollen
axon containing accumulated axoplasmic organelles
commonly encountered in preganglionic nerve fiber
terminals in the sympathetic ganglia with loss of
synaptic contacts. The swollen axons contain numer-
ous tubulovesicular networks of endoplasmic reticu-
lum and microtubular structures often mixed with
membranous bodies (see lower high magnified
view). This change is known to reflect the failure
axonal regrowth (aborted regeneration) and typical
of sympathetic neuropathy in diabetes but not spe-
cific to diabetes but also found in aging processes.
Source: This picture is quoted by Yagihashi, S and Sima,
AA. Diabetic autonomic neuropathy. The distribution of
structural changes in sympathetic nerves of the BB rat.
Am J Pathol 1985; 121: 138 147.

caliber of the lumina [113], and lack of autoregu- diabetes because microangiopathic changes are
lation of blood flow [114]. Endoneurial microves- salient pathological features. This categorization
sels run proximo-distally in the central zone of may not be necessarily valid because not all
the fasciculi, whereas microvessels distributed in facets of neuropathy follow the rule of microan-
the peripheral area of the fasciculi are derived giopathy, in particular, in cases of autonomic
from transperineurial arterioles through perineu- neuropathy. Fagerberg [117] first described in
rium originated from epineurial artery. The detail on the alteration of endoneurial microves-
arterioles are innervated with sympathetic and sels as a causation of DPN. He found consis-
peptidergic nerves which end in perineurial area tently thickened and hyalinous vascular wall of
[115,116] (Fig. 2.7). When endoneurial tissues endoneurial microvessels stained positive with
undergo edema with increased endoneurial pres- periodic acid Schiff (PAS) in diabetic patients
sure, transperineurial arterioles are likely to be with neuropathy [117]. Ultrastructural changes
compressed in the layered perineurium, result- of endoneurial microvessels consisted of endo-
ing in compromised blood flow [116]. Thus lack thelial swelling, hyperplasia, loss of pericytes,
of blood flow autoregulation and compression of vacuolation of endothelial cells ad thickening,
arteriolar walls gives rise to peripheral nerve and multiplication of basement membranes
hypoxia often with arteriovenous shunting. [12,13,40,41].
Infrequently, microthrombi were identified
in endoneurial microvessels in patients with
Microvessel changes in diabetic nerve diabetic neuropathy [11,118]. Foci of actual
Diabetic neuropathy is usually categorized as infarction with focal ischemic necrosis were
one of the microvascular complications of not well detected in previous literature

Diabetic Neuropathy
24 2. Pathologic basis for diabetic neuropathy in humans

FIGURE 2.7 Nerve vascular supply. Vascular system in peripheral nerve is different from central nervous system or
muscles. Distribution of vessels is sparse where it lacks autonomous blood flow regulation. In the center of nerve fasciculus,
only a few central microvessels with wide luminal space run proximo-distally, whereas the peripheral area is supplied from
capillaries derived from transperineurial arterioles originated from epineurial artery. Sympathetic and peptidergic nerves
terminate at the middle of transperineurial arterioles, thus nerve regulation of blood flow is missing in the endoneurium.
Sympathetic activation of the nerve often constricts epineurial arteriole, resulting in nerve ischemia. In cases of treatment
induced neuropathy in diabetes, arterio-venous shunting is often observed [Tesfaye, S., Malik, R., Harris, N., Jakubowski, J.J.,
Mody, C., Rennie, I.G., Ward, J.D.]. Arteriovenous shunting and proliferating new vessels in acute painful neuropathy of
rapid glycaemic control (insulin neuritis). Diabetologia 1996; 39: 329 335]. Source: Modified from Yagihashi, S., Mizukami, H.,
Sugimoto, K. Mechanism of diabetic neuropathy: where are we now and where to go? J Diabetes Investig 2011; 2:18 32.

although such thrombotic mechanisms may indicative of diabetes [124]. It has a critical role
underlie the endoneurial ischemia/hypoxia. for the prediction of the progression of neurop-
Normally, blood nerve barrier is well main- athy and outcome of the disease. Multiplicated
tained and there is no fenestration of endothe- basement membranes were considered to
lial cells. Fenestrated vessels were often result from repeated cell death and renewal of
encountered in cases with overt neuropathy pericytes or endothelial cells leaving basement
[119]. Because of swollen endothelial cells, membrane materials [124].
closed capillaries were often frequent and cor- It was also shown that endoneurial microves-
related with severity of neuropathy [120,121]. sel changes preceded clinical onset of diabetes
These results were not confirmed by other and suspected to be an early harbinger of neu-
investigators and were not specific for diabetes ropathy [122]. Over 10 year observation period
[122,123]. Attachment of inflammatory cells on after sural nerve biopsy, prediabetic patients
the vascular walls was also common, but its with microvessel changes developed clinical
implication was not well discussed. Giannini onset of diabetes and thereafter followed the pro-
and Dyck characterized thickening of basement gression of neuropathy [125]. Importance of
membranes of endoneurial microvessels as an microangiopathy in the neuropathy was also

Diabetic Neuropathy
 Implication of microangiopathy in diabetic polyneuropathy 25
introduced by Johnson et al., who examined in a dystrophic changes in the axons were supposed
systemic way the spinal ganglion, sciatic nerve to be secondary to ischemia [41].
and distal tibial and sural nerves obtained from In contrast to the implication of occluded ves-
autopsy subjects mostly aged with diabetes [12]. sels proposed by Dyck et al. and Bradley et al.
They found proximodistal gradient of nerve fiber [122] could not find any significant difference in
loss which showed focal distribution. The focal endothelial area, pericyte area, capillary density,
nerve fiber loss was striking from lumbosacral to or percentage of closed capillaries except for
posterior tibial nerve which was associated with basement membrane (basal lamina) area in
vascular proliferation near the perineurium of patients with diabetes, casting critique on the
sural nerve. In their preceding study, they found role of closed capillaries in the development of
thickened perineurial basement membranes neuropathy [122]. This discrepancy may be
[126], which they considered to compress the explained by the differences in age of diabetic
wall of transperineurial arterioles, resulting in patients studied, because Dyck et al.’s cases were
wedge-shaped focal ischemic lesions of endo- mostly aged, while they were much younger in
neurium. Deficient nerve supply of epi- and Bradley et al.’s study. Sima et al. [128] found an
transperineurial arterioles in diabetic subjects increased thickness of endoneurial basement
was also postulated to contribute to the nerve membranes in T2D with neuropathy and loss of
ischemia in diabetes [115,116]. Powell and his endothelial cell junctions [128]. These changes
colleagues [41] confirmed the endoneurial and were refractory to the treatment with ARI while
transperineurial vascular changes with marked nerve fibers were repaired, indicating different
nerve fiber loss in sural nerves of diabetic mechanisms of neuropathy between T1D and
patients. Together with perineurial vascular T2D. Our observation also supported the signifi-
growth, epineurial artery showed narrowing the cant relationship between basement membrane
lumen with thickened intima in patients with thickness and the extent of loss of myelinated
diabetes [127]. Glycogen accumulation or fibers [129] (Fig. 2.8).

FIGURE 2.8 Relationship between

basement membrane thickness of endo-
neurial microvessels and myelinated nerve
fiber density. Thickening of basement
membrane is characteristic in diabetic
microangiopathy and endneurial micro-
vessels are not the exception.
Morphometric analysis well demonstrates
a significant inverse correlation between
basement membrane thickness of endo-
neurial capillaries and myelinated nerve
fiber density, indicating causative role of
microangiopathy in the progression and
severity of neuropathy. There is no such
relationship in Charcot Marie Tooth dis-
ease despite marked reduction of nerve
fiber density.

Diabetic Neuropathy
26 2. Pathologic basis for diabetic neuropathy in humans

Do the vascular alterations induce dorsal root ganglia and axons in nerves. It is
ischemic/hypoxic neuropathy in diabetes? known that both ischemic nerve and diabetic
nerve are resistant to further worsening of con-
Dyck summarized in his memorial lecture duction block when additional ischemia is over-
for the Wartenberg award pathological charac- loaded [131]. The reason for this refractoriness is
teristics of acute ischemic neuropathy as fol- not clear, but suspected to be preconditioning of
lows [71]: (1) nerve ischemia resulted in focal saved energy in the setting of hypoxia while pre-
or multifocal central fascicular or sector axonal served ATP energy may be used in diabetic nerve
degeneration, which tended to begin at mid [131,132]. In contrast to preserved functional
upper arm or mid thigh levels; (2) the spatial capacity, diabetic nerves were much more suscep-
distribution of these ischemic cores suggested tible to structural damage [133,134]. Diabetic sub-
that they were watershed zones of poor perfu- jects, therefore, are likely to suffer from severe
sion; (3) occlusion of focal regions of individual neuropathic damage even by the modest expo-
epineurial arterioles was not likely to cause sure to normally tolerable ischemia and reperfu-
fiber degeneration; (4) axons were more sion. Structurally, reperfusion injury caused
vulnerable to ischemia than Schwann cells or demyelination, intramyelinic edema and endo-
myelin; and (5) the neuropathologic site responsi- neurial edema, in addition to axonal degeneration
ble for clinical symptoms and deficits attributed [135]. Reperfusion in peripheral nerves elicited an
to a single nerve. He indicated that the focal nerve acute inflammatory response and the generation
fiber loss was not ascribed to infarcted lesion of of free radicals [134,136]. In this setting, topo-
endoneurial microvessels, but more reasonable to graphic distribution of nerve damage indicated
be due to ischemia. Experiments with micro- that nerve injury was generated in the watershed
sphere injection in rats yielded multiple foci of area by ischemia/reperfusion.
ischemic core lesion from proximal sciatic to distal
tibial and sural nerve [70]. In this study, central
fascicular fiber degeneration with relative sparing Pathology of nerve microenvironment
of the subperineurial fibers was found in the sci-
atic nerve. The ischemic core began at the mid- Nerve fiber microenvironment is crucial for
thigh level of sciatic nerve, presumably anatomic the tissue remodeling for peripheral nerve
watershed zones of poor perfusion, and extended pathology [137]. Basement membranes of
more diffusely into the tibial, peroneal and sural Schwann cells provide scaffold for axon growth
nerves. Consistent with the Dyck’s consideration, and elongation of regenerative fibers. In the
Malik et al. also escalated implication of hypoxia milieu of interstitial collagens, oxygen, nutrients,
in the pathogenesis of diabetic neuropathy and and growth factors are supplied in the endoneur-
strengthened the influence of hypoxia on the pro- ium for promotion of the neurite growth and for
gression of neuropathic changes [130]. protection from fiber degeneration. Behse et al.
[138] reported increased interstitial area occupy-
Resistance to ischemia in diabetic nerve ing 35% 45% of endoneurial area consisting of
densely packed collagen fibrils filling the space
and ischemia/reperfusion injury for lost nerve fibers. Fine measurement of colla-
What is then the difference between ischemic/ gen fibrils disclosed thickened diameter of fibrils
hypoxic neuropathy and diabetic neuropathy? It in diabetic nerve with increased content of type
may be underlined that there is a dissociation IV and V collagen [139]. Interestingly, Schwann
between functional deficit and structural severity. cell basement membranes exhibited circular in
Endoneurial hypoxia targets the cell bodies in diabetic nerve compared with corrugated

Diabetic Neuropathy
 Skin biopsy 27
appearance in nondiabetic nerves [140]. penetrating the width of 1 mm basement mem-
Increased nonenzymatic glycation of basement brane of epidermis (IENFD) is used for the
membrane collagens and production of grading of neuropathy. More recently, addition
advanced glycation end-products may be major of special stains with neurofilaments, regenera-
contributors to increased rigidity and durability tive molecule of GAP43, or mitochondrial mar-
of basement membrane structure [141]. On the kers, as well as Schwann cell signal markers
other hand, Bradley et al. found that basement was attempted for more efficient evaluation of
membrane changes were not uniform and thick- neuropathy [142]. In addition, repeated biopsy
ened perineurial basement membranes were is now applied to explore the regenerative pro-
more specific to diabetes rather than those of cess of nerve fibers in vivo.
endoneurial microvessels [139]. The reason for We also recently disclosed that the suppressed
this discrepancy remains to be explored, but it is vascular endothelial growth factor (VEGF)
likely that such alterations of environment influ- expression on the vascular wall in the dermis cor-
ence the progression of neuropathy. related with nerve fiber loss in diabetic patients
[143]. It was speculated that lowered VEGF in the
nerve tissues was involved in impaired nerve
Skin biopsy fiber regeneration. Despite the minimal invasive-
ness, skin biopsy is not routinely conducted, but
As outlined briefly in the previous section, only applied to the research purpose in diabetes
skin biopsy is employed for the evaluation of clinic. Nevertheless, strong correlation of IENFD
small nerve fibers in the field of neurology. with clinical staging of neuropathy in diabetes
Usually, skin tissues with 3 mm diameter indicates the usefulness of skin biopsy for further
obtained from punch biopsy are submitted to clarification of underlying mechanisms and
the analysis of nerve fibers in both epidermis future development of effective treatment design
and upper dermis. Number of nerve fibers [48] (Fig. 2.9).

FIGURE 2.9 Distribution of intradermal nerve fibers in diabetic neuropathy. Reduction of intraepidermal nerve fiber
density (IENFD) is an index of small fiber neuropathy and commonly found in prediabetic patients. The reduction of
IENFD corresponds with severity of neuropathy. Recently, we found reduced expression of vascular endothelial growth
factor (VEGF) in dermal capillaries, indicating the relationship between disturbed nerve fiber regeneration and reduced
IENFD. Left; nondiabetic control (IENFD 21/mm), Center; Prediabetic (impaired glucose tolerance) subject (IENFD 9/mm),
Right; Diabetic subject with overt neuropathy (IENFD 0/mm).

Diabetic Neuropathy
28 2. Pathologic basis for diabetic neuropathy in humans

Conclusion addition, there emerge exciting new theories

related to impaired energy utilization and
It has been evidenced by findings of various small nerve fiber alterations in the pathogene-
studies that the combination of long-term sis of diabetic neuropathy [146 148]. It is
metabolic aberration and vascular deficits hoped that appropriate translation of the find-
is a major cause of diabetic neuropathy. ings to humans will lead to the final success to
Understanding of the underlying pathological overcome this difficult disorder.
background is crucial for proper clinical diag-
nosis and management of DPN. Vast majority
of neurological complications of diabetes are Declaration of competing interests
polyneuropathy with distal symmetric sensori-
motor neuropathy and autonomic neuropathy. The author have no conflict to declare for
Its pathological features are reflected by this manuscript.
length-dependent and distal predominant axo-
nal degeneration of nerve fibers and fiber loss. References
Concurrently, microvessels supporting periph-
[1] Ellenberg M. Diabetic neuropathy: evaluation of fac-
eral nerve tissues also undergo thickening and tors in onset. Ann N Y Acad Sci 1959;82:245 50.
multiplication of vascular wall, resulting in [2] American diabetes association consensus panel.
nerve ischemia/hypoxia. In clinical practice, Report and recommendations of the San Antonio con-
the progression of neuropathy may likely be ference on diabetic neuropathy. Diabetes Care
1988;11:592 7.
more complicated overlapped with hyperten-
[3] American diabetes association consensus panel.
sion, lipidemia, and uremia on the top of Report and recommendations of the San Antonio con-
hyperglycemia. There emerges recent evidence ference on diabetic neuropathy. Diabetes Care 1992;15
that neuropathic patterns appear to be different (Suppl 2):62 7.
in T1D and T2D, where lipid abnormalities [4] Vinik AI, Liuzzi FJ, Holland MT, Liuzzi FJ, Stansberry
KB, Colen B. Diabetic neuropathies. Diabetes Care
overly the basis of neuropathy. On the other
1992;15:1926 75.
hand, although much less frequent but more [5] Bjerg L, Hulman A, Carstensen B, Charles M, Witte
rapidly, mononeuropathy and radiculoplexus DR, Jørgensen ME. Effect of duration and burden of
neuropathy cause ophthalmoplegia and proxi- microvascular complications on mortality rate in type
mal limb neuropathy, respectively, in which 1 diabetes: an observational clinical cohort study.
Diabetologia 2019;62:633 43.
vascular occlusion and immune-mediated vas-
[6] Hicks CW, Selvin E. Epidemiology of peripheral neu-
culitis are involved. ropathy and lower extremity disease in diabetes. Curr
Finally, it may not be too much emphasized Diab Rep 2019;19:86 PMID31456118.
to value the basic studies on diabetic neuropa- [7] Pirart J. Diabetes mellitus and its degenerative compli-
thy to explore the effective prevention and cations: a prospective study of 4,400 patients observed
between 1947 and 1973 (1978) Diabetes Care
treatment of DPN. At baseline, augmented glu-
1978;1:168 88.
cose flux exerts an increased collateral glyco- [8] Partanen J, Niskanen L, Lehtinen J, Mervaala E,
lytic pathway with activation of polyol Siitonen O, Uusitupa M. Natural history of peripheral
pathway, hexosamine pathway, glycation of neuropathy in patients with non-insulin-dependent
structural proteins and enhanced oxidative diabetes mellitus. N Engl J Med 1995;332:89 94.
[9] Thomas PK. Diabetic neuropathy: mechanisms and
stress [144,145]. In this setting, downstream
future treatment options. J Neurol Neurosurg
alterations of molecular signals susceptible to Psychiatry 1999;67:277 9.
ischemia may first be listed to the target of the [10] Pop-Busui R, Boulton AJM, Feldman EL, Bril V,
correction of neuropathy development. In Freeman R, Malik RA, et al. Diabetic neuropathy: a

Diabetic Neuropathy
 References 29
position statement by the american diabetes associa- diabetics. II. Spinal Cord Peripheral nerves Diabetologia
tion. Diabetes Care 2017;40:136 54. 1968;4:34 43.
[11] Timperley WR, Ward JD, Preston FE, Duckworth T, [29] Appenzeller O, Ogin G. Myelinated fibers in human
O’Malley B. Clinical and histological studies in dia- paravertebral sympathetic chain: white rami commu-
betic neuropathy. A reassessment of vascular factors nicantes in alcoholic and diabetic patients. J Neurol
in relation to intravascular coagulation. Diabetologia Neurosurg Psychiatry 1974;37:1155 61.
1976;12:237 43. [30] Low PA, Walsch JC, Huang CY, McLeod JG. The sym-
[12] Johnson PC, Doll SC, Cromey DW. Pathogenesis of pathetic nervous system in diabetic neuropathy. A
diabetic neuropathy. Ann Neurol 1986;19:450 7. clinical and pathological study. Brain 1975;98:341 56.
[13] Dyck PJ, Giannini C. Pathologic alterations in the dia- [31] Duchen CW, Anjorin MB, Watkins MD, Mackay JD.
betic neuropathies of humans: a review. J Neuropathol Pathology of autonomic neuropathy in diabetes melli-
Exp Neurol 1996;55:1181 93. tus. Ann Intern Med 1980;92:301 3.
[14] Thomas PK. Classification, differential diagnosis, and [32] Budzilovich GN. Diabetic neuropathy complex.
staging of diabetic peripheral neuropathy. Diabetes Virchows Arch A Pathol Anatand Histol 1970;350:
1997;46(Supple 2):S54 7. 105 22.
[15] Watkins PJ. Diabetic autonomic neuropathy. N Engl J [33] Sugimura K, Dyck PJ. Multifocal fiber loss in proximal
Med 1990;322:1078 9. sciatic nerve in symmetric distal diabetic neuropathy.
[16] Maser RE, Mitchell BD, Vinik AI, Freeman R. The J Neurol Sci 1982;53:501 9.
association between cardiovascular autonomic neu- [34] Ohnishi A, Harada M, Tateishi J, Ogata J, Kawanami
ropathy and mortality in individuals with diabetes: a S. Segmental demyelination and remyelination in lum-
meta-analysis. Diabetes Care 2003;26:1895 901. bar spinal roots of patients dying with diabetes melli-
[17] Vinik AI, Ziegler D. Diabetic cardiovascular auto- tus. Ann Neurol 1 1983;3:541 8.
nomic neuropathy. Circulation 2007;115:387 97. [35] Thomas PK, Lascelles RG. Schwann-cell abnormalities
[18] Raff MC, Asbury AK. Ischemic mononeuropathy and in diabetic neuropathy. Lancet 1966;1:1355 7.
mononeuropathy multiplex in diabetes mellitus. N [36] Thomas PK. The morphological basis for alterations in
Engl J Med 1968;279:17 21. nerve conduction in peripheral neuropathy. Proc R
[19] Said G, Goulon-Goeau C, Lacroix C, Moulonguet A. Soc Med 1971;64:295 8.
Nerve biopsy findings in different patterns of proxi- [37] Bischoff A. Ultrastructural pathology of peripheral
mal diabetic neuropathy. Ann Neurol 1994;35:559 69. nervous system in early diabetes. Adv Metab Disord
[20] Said G. Diabetic neuropathy—a review. Nat Clin Pract 1973;2(Supple 1):441 9.
Neurol 2007;3:331 40. [38] Behse F, Buchthal F, Carlsen F. Nerve biopsy and con-
[21] Ellenberg M. Diabetic truncal mononeuropathy—a duction studies in diabetic neuropathy. J Neurol,
new clinical syndrome. Diabetes Care 1978;1:10 13. Neurosurg Psychiatry 1977;40:1072 82.
[22] Barohn RJ, Sahenk Z, Warmolts JR, Mendell JR. The [39] Said G, Slama G, Selva J. Progressive centripetal
Bruns-Garland syndrome (diabetic amyotrophy). degeneration of axons in small fiber diabetic poly-
Revisited 100 years later. Arch Neurol 1991;48:1130 5. neuropathy. A clinical and pathological study. Brain
[23] Dyck PJ, Windebank AI. Diabetic and non-diabetic 1983;106:791 807.
lumbosacral radiculoplexus neuropathies: new insights [40] Yagihashi S, Matsunaga M. Ultrastructural pathology
into pathophysiology and treatment. Muscle Nerve of peripheral nerves in patients with diabetic neuropa-
2002;25:477 91. thy. Tohoku J Exp Med 1979;129:357 66.
[24] Kelkar P, Parry G. Mononeuropathy multiplex in diabe- [41] Powell HC, Rosiff J, Myers RR. Microangiopathy in
tes mellitus: evidence for underlying immune pathogen- human diabetic neuropathy. Acta Neuropathol
esis. J Neurol Neurosurg Psychiatry 2003;74:803 6. 1985;68:295 305.
[25] Woltman HW, Wilder RM. Diabetes mellitus: pathoge- [42] Perry JR, Bril V. Complications of sural nerve biopsy
netic changes in the spinal cord and peripheral nerves. in diabetic vs non-diabetic patients. Can Neurol Sci
Arch Intern Med 1929;44:576 603. 1994;21:34 7.
[26] Dolman CL. The morbid anatomy of diabetic neuropa- [43] Campbell CA, Turza KC, Morgan RF. Postoperative
thy. Neurology 1963;13:135 42. outcomes and reliability of “sensation-sparing” sural
[27] Greenbaum D, Richardson PC, Salmon HV, Urich H. nerve biopsy. Muscle Nerve 2009;40:603 9.
Pathological observations on six cases of diabetic neu- [44] Thomsen NO, Mojaddidi M, Malik RA, Dahlin LB.
ropathy. Brain 1964;87:201 14. Biopsy of the posterior interosseous nerve: a low mor-
[28] Reske-Nielsen E, Lundbaek K. Pathological changes in the bidity method for assessment of peripheral nerve dis-
central and peripheral nervous system of young long-term orders. Diabet Med 2009;26:100 4.

Diabetic Neuropathy
30 2. Pathologic basis for diabetic neuropathy in humans

[45] Herrmann DN, Griffin JW, Hauer P, Cornblath DR, [58] Malik RA, Veves A, Walker D, Siddque I, Lye RH,
McArthur JC. Epidermal nerve fiber density and sural Schady W, et al. Sural nerve fiber pathology in dia-
nerve morphometry in peripheral neuropathies. betic patients with mild neuropathy: relationship to
Neurology 1999;53:1634 40. pain, quantitative sensory testing and peripheral nerve
[46] Wendelschafer-Crabb G, Kennedy WR, Walk D. electrophysiology. Acta Neuropathol 2001;101:367 74.
Morphological features of nerves in skin biopsies. J [59] Dyck PJ, Hansen S, Karnes J, O’Brien P, Yasuda H,
Neurol Sci 2006;242:15 21. Windebank A, et al. Capillary number and percentage
[47] Lauria D, Devigili G. Skin biopsy as a diagnostic tool closed in human diabetic sural nerve. Proc Nat Acad
in peripheral neuropathy. Nat Clin Pract Neurol Sci U S A 1985;82:2513 17.
2007;3:546 57. [60] Britland ST, Young RJ, Sharma AK, Clarke BF.
[48] Arimura A, Deguchi T, Sugimoto K, Uto T, Nakamura Association of painful and painless diabetic polyneuropa-
T, Arimura Y, et al. Intraepidermal nerve fiber density thy with different patterns of nerve fiber degeneration
and nerve conduction study parameters correlate with and regeneration. Diabetes 1990;39:898 908.
clinical staging of diabetic polyneuropathy. Diabetes [61] Llewelyn JG, Gilbey SG, Thomas PK, Watkins PJ,
Res Clin Pract 2013;99:24 9. Muddle JR. Sural nerve morphometry in diabetic auto-
[49] Sumner CJ, Sheth S, Griffin JW, Cornblath DR, Polydefkis nomic and painful sensory neuropathy. A clinico-
M. The spectrum of neuropathy in diabetes and impaired pathological study. Brain 1991;114:867 92.
glucose tolerance. Neurology 2003;60:108 11. [62] Yagihashi S, Yamagishi S, Wada R. Pathology and
[50] Smith AG, Russel JW, Feldman EL, Goldstein J, Peltier pathogenetic mechanisms of diabetic neuropathy: cor-
A, Smith S, et al. Lifestyle intervention for pre-diabetic relation with clinical signs and symptoms. Diabetes
neuropathy. Diabetes Care 2006;29:1294 9. Res Clin Pract 2007;77(Supple 1):S184 9.
[51] Malik RA, Kallinikos P, Abbot CA, van Schie CHM, [63] Sima AA, Bril V, Nathaniel V, McEwen TA, Brown
Morgan P, Efron N, et al. Corneal confocal micros- MB, Lattimer SA, et al. Regeneration and repair of
copy: a non-invasive surrogate of nerve fiber damage myelinated fibers in sural-nerve biopsy specimens
and repair in diabetic patients. Diabetologia from patients with diabetic neuropathy treated with
2003;46:683 8. sorbinil. N Engl J Med 1988;319:548 55.
[52] Quattrini C, Tavakoli M, Jeziorska M, Kallinikos P, [64] Greene DA, Arezzo JC, Brown MB. Effect of aldose
Tesfaye S, Finnigan J, et al. Surrogate markers of small reductase inhibition on nerve conduction and mor-
fiber damage in human diabetic neuropathy. Diabetes phometry in diabetic neuropathy. Zenarestat Study
2007;56:2148 54. Group Neurol 1999;53:580 91.
[53] Tavakoli M, Quattrini C, Abbot C, Kalliknios P, [65] Hur J, Sullivan KA, Pande M, Hong Y, Sima AA,
Marshall A, Finnigan J, et al. Corneal confocal micros- Jagadish HV, et al. The identification of gene expres-
copy: a novel non-invasive test to diagnose and strat- sion profiles associated with progression of human
ify the severity of human diabetic neuropathy. diabetic neuropathy. Brain 2011;134:3222 35.
Diabetes Care 2010;33:1792 7. [66] Hur J, Sullivan KA, Callaghan BC, Pop-Busui R,
[54] Veves A, Malik RA, Lye RH, Masson EA, Sharma AK, Feldman EL. Identification of factors associated with
Schady W, et al. The relationship between sural nerve sural nerve regeneration and degeneration in diabetic
morphometric findings and measure of peripheral neuropathy. Diabetes Care 2013;36:4043 9.
nerve function in mild diabetic neuropathy. Diabet [67] Dyck PJ, Karnes JL, O’Brien P, Okazaki H, Lais A,
Med 1991;8:917 21. Engelstad J. The spatial distribution of fiber loss in
[55] Russel JW, Karnes JL, Dyck PJ. Sural nerve myelinated diabetic polyneuropathy suggests ischemia. Ann
fiber density differences associated with meaningful Neurol 1986;19:440 9.
changes in clinical and electrophysiologic measure- [68] Dyck PJ, Lais A, Karnes JL, O’Brien P, Rizza R.
ments. J Neurol Sci 1996;135:114 17. Fiber loss is primary and multifocal in sural nerves
[56] Gibbons CH, Freeman R, Tecilazich F, Dinh T, Lyons in diabetic polyneuropathy. Ann Neurol
TE, Gnardellis C, et al. The evolving natural history of 1986;19:425 39.
neurophysiologic function in patients with well con- [69] Dyck PJ, Karnes J, O’Brien P, Nukada H, Lais A, Low
trolled diabetes. J Peripher Nerv Syst 2013;18:153 61. P. Spatial pattern of nerve fiber abnormality indicative
[57] Kobori M, Yagihashi S, Shiina N, Shiozawa N, of pathologic mechanisms. Am J Pathol 1984;
Haginoya N, Ojima M, et al. Four-year sequential 117:225 38.
nerve conduction changes since first visit in Japanese [70] Nukada H, Dyck PJ. Microsphere embolization of
patients with early type 2 diabetes. J Diabetes Invest nerve capillaries and fiber degeneration. Am J Pathol
2017;8:369 76. 1984;115:275 87.

Diabetic Neuropathy
 References 31
[71] Dyck PJ. Hypoxic neuropathy: does hypoxia play a in atherosclerotic peripheral vascular disease. Brain
role in diabetic neuropathy? The 1988 robert warten- 1996;119:1449 60.
berg lecture. Neurology 1988;319:542 8. [85] Mizisin AP. Mechanisms of diabetic neuropathy;
[72] Llewelyn JG, Thomas PK, Gilbey SG, Watkins PJ, Schwan cells. In: Zochodne DW, Malik RA, editors.
Muddle JR. Pattern of myelinated fiber loss in Handbook of Clinical Neurology (Diabetes and the
the sural nerve in neuropathy related to type 1 nervous system, 126. Elsevier; 2014. p. 401 22.
(insulin-dependent) diabetes. Diabetologia [86] Kalichman MW, Powell HC, Mizisin AP. Reactive,
1988;31:162 7. degenerative, and proliferative Schwann cell
[73] Sima AA, Nathaniel V, Bril V, McEwen TA, Greene responses in experimental galactose and human dia-
DA. Histopathological heterogeneity of neuropathy in betic neuropathy. Acta Neuropathol 1998;95:47 56.
insulin-dependent and non-insulin-dependent diabe- [87] Andersen H, Gadeberg PC, Brock B, Jakobsen J.
tes, and demonstration of axo-glial dysjunction in Muscular atrophy in diabetic neuropathy: a stereologi-
human diabetic neuropathy. J Clin Invest cal magnetic resonance imaging study. Diabetologia
1988;81:349 64. 1997;40:1062 9.
[74] Wiggin TD, Sullivan KA, Pop-Busui R, Amato A, Sima [88] Andersen H, Gjerstad MD, Jakobsen J. Atrophy of foot
AAF, Feldman EL. Elevated triglycerides correlate muscles: a measure of diabetic neuropathy. Diabetes
with progression of diabetic neuropathy. Diabetes Care 2004;27:2382 5.
2009;58:1634 40. [89] Andreassen CS, Jensen JM, Jakobsen J, Uløj BP,
[75] DCCT Research Group. The effect of intensive treat- Andersen H. Striated muscle fiber size, composition,
ment of diabetes on the development and progression and capillary density in diabetes in relation to neuropa-
of long-term complications in insulin-dependent dia- thy and muscle strength. J Diabetes 2014;6:462 71.
betes mellitus. N Engl J Med 1993;329:977 86. [90] Reske-Nielsen E, Gregersen G, Harmsen A, Lundbaek
[76] UK Prospective Diabetes Study (UKPDS) group. K. Morphological abnormalities of the terminal neuro-
Intensive blood-glucose control with sulphonylureas muscular apparatus in recent juvenile diabetes.
or insulin compared with conventional treatment and Diabetologia 1970;6:104 9.
risk of complications in patients with type 2 diabetes [91] Cazzato G, Walton JN. The pathology of the muscle
(UKPDS 33). Lancet 1998;352:837 53. spindle. A study of biopsy material in various muscu-
[77] Callaghan BC, Hur J, Feldman EL. Diabetic neuropathy: lar and neuromuscular diseases. J Neurol Sci
one disease or two? Curr Opin Neurol 2012;25:536 41. 1968;7:15 70.
[78] Said G, Goulon-Goeau C, Slama G, Tchobroutsky G. [92] Dreyfus PM, Hakim S, Adams RD. Diabetic ophthal-
Severe early-onset polyneuropathy in insulin- moplegia. Arch Neurol Psychiat 1957;77:337 49.
dependent diabetes mellitus. A clinical and pathologi- [93] Asbury AK, Aldredge H, Herschberg R, Fisher CM.
cal study. N Engl J Med 1992;326:1257 63. Oculomotor palsy in diabetes mellitus: a clinicopatho-
[79] Malik RA, Tesfaye S, Newrick PG, Walker D, logical study. Brain 1970;93:555 66.
Rajbhandari SM, Siddique I, et al. Sural nerve pathol- [94] Dyck PJ, Norell JE, Dyck PJ. Microvasculitis and ische-
ogy in diabetic patients with minimal but progressive mia in diabetic lumbosacral radiculoplexus neuropa-
neuropathy. Diabetologia 2005;48:578 85. thy. Neurology 1999;53:2113 21.
[80] Bradley JL, Thomas PK, King RH, Muddle JR, Ward [95] Garces-Sanchez M, Laughlin RS, Dyck PJ, Engelstad
JD, Tesfaye S, et al. Myelinated nerve fibre regenera- JN, Norell JE, Dyck PJB. Painless diabetic motor neu-
tion in diabetic sensory polyneuropathy. Acta ropathy: a variant of diabetic lumbosacral radiculo-
Neuropathol 1995;90:403 10. plexus neuropathy? Ann Neurol 2011;69:1043 54.
[81] Jakobsen J. Peripheral nerves in early experimental [96] Stewart JD, McKelvey R, Durcan L, Carpenter S,
diabetes: expansion of the endoneurial space as a Karpati G. Chronic inflammatory demyelinating poly-
cause of increased water content. Diabetologia neuropathy (CIDP) in diabetics. J Neurol Sci
1978;14:113 19. 1996;142:59 64.
[82] Powell HC, Costello ML, Myers RR. Endoneurial fluid [97] Bril V, Blanchette CM, Noone JM, Runken MC,
pressure in experimental model of diabetic neuropa- Gelinas D, Russel JW. The dilemma of diabetes in
thy. J Neuropathol Exp Neurol 1981;40:613 24. chronic inflammatory demyelinating polyneuropathy.
[83] Nukada H, Dyck PJ. Acute ischemia causes axonal sta- J Diabetes Compl 2016;30:1401 7.
sis, swelling, attenuation, and secondary demyelin- [98] Rajabally YA, Stettner M, Kiesseier BC, Hartung H-P,
ation. Ann Neurol 1987;22:311 18. Malik RA. CIDP and other inflammatory neuropathies
[84] Nukada H, van Rij AM, Packer SG, McMorran PD. in diabetes-diagnosis and management. Nat Rev
Pathology of acute and chronic ischaemic neuropathy Neurol 2017;13:599 611.

Diabetic Neuropathy
32 2. Pathologic basis for diabetic neuropathy in humans

[99] Laughlin RS, Dyck PJ, Melton 3rd LJ, Leibson C, [115] Grover-Johnson NM, Baumann FG, Imparato AM,
Ransom J, Dyck PJB. Incidence and prevalence of Kim GE, Thomas PK. Abnormal innervation of lower
CIDP and the association of diabetes mellitus. limb epineurial arterioles in human diabetes.
Neurology 2009;73:39 45. Diabetologia 1981;20:31 8.
[100] Hensley GT, Soergel KH. Neuropathologic findings [116] Beggs J, Johnson PC, Olafsen A, Watkins CJ.
in diabetic diarrhea. Arch Pathol 1968;85:587 97. Innervation of the vasa nervorum: changes in human
[101] Smith B. Neuropathology of the oesophagus in diabe- diabetes. J Neuropathol Exp Neurol 1992;51:612 29.
tes mellitus. J Neurol Neurosurg Psychiatry [117] Fagerberg SE. Diabetic neuropathy: a clinical and his-
1974;37:1151 4. tological study on the significance of vascular affec-
[102] Faerman I, Faccio E, Milei J, Nuñez R, Jadzinsky M, tions. Acta Med Scand. (Suppl) 1959;345:1 80.
Fox D, et al. Autonomic neuropathy and painless myo- [118] Williams E, Timperley WR, Ward JD, Duckworth T.
cardial infarction in diabetic patients. Histologic evi- Electron microscopical studies of vessels in diabetic
dence of their relationship. Diabetes 1977;26:1147 58. peripheral neuropathy. J Clin Pathol 1980;33:462 70.
[103] Huang CY, Walsh JC. Pathologic changes in the greater [119] Malik RA, Newrick PG, Sharma AK, Jennings A, Ah-
splanchnic nerves of subjects with diabetic peripheral See AK, Mayhew TM, et al. Microangiopathy in
neuropathy. Proc Austr Assoc Neurol 1973;10:13 18. human diabetic neuropathy: relationship between
[104] Gibbons CH, Illigens BM, Wang N, Freeman R. capillary abnormalities and the severity of neuropa-
Quantification of sudomotor innervation: a compari- thy. Diabetologia 1989;32:92 102.
son of three methods. MuscleNerve 2010;42:112 19. [120] Dyck PJ, Karnes JL, Daube J, O’Brien P, Service F.
[105] He CL, Soffer EE, Ferris CD, Walsh RM, Szurszewski Clinical and neuropathological criteria for the diag-
JH, Farrugia G. Loss of interstitial cells of Cajal and nosis and staging of diabetic polyneuropathy. Brain
inhibitory innervation in insulin-dependent diabetes. 1985;105:861 80.
Gastroenterology 2001;121:427 34. [121] Yasuda H, Dyck PJ. Abnormalities of endoneurial
[106] Yagihashi S, Sima AA. Diabetic autonomic neuropathy microvessels and sural nerve pathology in diabetic
in the BB rat. Ultrastructural and morphometric changes neuropathy. Neurology 1987;37:20 8.
in sympathetic nerves. Diabetes 1985;34:558 64. [122] Bradley J, Thomas PK, King RH, Llewelyn JG,
[107] Yagihashi S, Sima AA. Diabetic autonomic neuropathy Muddle JR, Watkins PJ. Morphometry of endoneurial
in BB rat. Ultrastructural and morphometric changes in capillaries in diabetic sensory and autonomic neurop-
parasympathetic nerves. Diabetes 1986;35:733 43. athy. Diabetologia 1990;33:611 18.
[108] Schmidt RE, Plurad SB, Parvin CA, Roth KA. Effect [123] Malik RA, Veves A, Masson EA, Sharma AK, Ah-See
of diabetes and aging on human sympathetic auto- AK, Schady W, et al. Endoneurial capillary abnor-
nomic ganglia. Am J Pathol 1993;143:143 53. malities in mild human diabetic neuropathy. J
[109] Schmidt RE. Neuropathology and pathogenesis of Neurol Neurosurg Psychiat 1992;55:557 61.
diabetic autonomic neuropathy. Int Rev Neurobiol [124] Giannini C, Dyck PJ. Basement membrane reduplica-
2002;50:257 92. tion and pericyte degeneration precede development
[110] Lampert PW. A comparative electron microscopic study of diabetic polyneuropathy and are associated with
of reactive, degenerating, regenerating, and dystrophic its severity. Ann Neurol 1995;37:498 504.
axons. J Neuropathol Exp Neurol 1967;26:345 68. [125] Thrainsdottir S, Malik RA, Dahlin LB, Wiksell P,
[111] Zanone MM, Peakman M, Purewal T, Watkins P, Vergani Eriksson KF, Rosén I, et al. Endoneurial capillary
D. Autoantibodies to nervous tissue structures are associ- abnormalities presage deterioration of glucose intol-
ated with autonomic neuropathy in type 1 (insulin-depen- erance and accompany peripheral neuropathy in
dent) diabetes mellitus. Diabetologia 1993;36:564 9. man. Diabetes 2003;52:2615 22.
[112] Vernino S, Low PA, Fealey RD, Stewart JD, Farrugia [126] Johnson PC, Brendel K, Meezan E. Human diabetic
G, Lennon VA. Autoantibodies to ganglionic acetyl- perineurial cell basement membrane thickening. Lab
choline receptors in autoimmune autonomic neurop- Invest 1981;44:265 70.
athies. N Engl J Med 2000;343:847 55. [127] Korthals JK, Gieron MA, Dyck PJ. Intima of epineur-
[113] Lundborg G. Structure and function of the intraneur- ial arterioles is increased in diabetic polyneuropathy.
al microvessels as related to trauma, edema forma- Neurology 1988;38:1582 6.
tion, and nerve function. J Bone Jt SurgAm Vol [128] Sima AAF, Nathaniel V, Prashar A, Bril V, Greene
1975;57:938 48. DA. Endoneurial microvessels in human diabetic
[114] Smith DR, Kobrine AI, Rizzoli HV. Absence of auto- neuropathy. Endothelial cell dysfunction and lack of
regulation in peripheral nerve blood flow. J Neurol treatment effect by aldose reductase inhibitor.
Sci 1977;33:347 52. Diabetes 1991;40:1090 9.

Diabetic Neuropathy
 References 33
[129] Yagihashi S. Pathology and pathogenetic mechan- Schwann cell and perineurial basal laminae.
isms of diabetic neuropathy. Diabetes Metab Rev Implication for diabetic vasculopathy. Neuropathol
1995;11:193 225. Appl Neurobiol 1989;15:339 55.
[130] Malik RA, Masson EA, Sharma AK, Lye RH, Ah-See [141] Sugimoto K, Nishizawa Y, Horiuchi S, Yagihashi S.
AK, Compton AM, et al. Hypoxic neuropathy: rele- Localization in human diabetic peripheral nerve of N
vance to human diabetic neuropathy. Diabetologia (epsilon)-carboxymethyllysine-protein adducts, an
1990;33:311 18. advanced glycation endproduct. Diabetologia
[131] Low P. Recent advances in the pathogenesis of dia- 1997;40:1380 7.
betic neuropathy. MuscleNerve 1987;10:121 8. [142] Bönhof GJ, Strom A, Püttgen S, Ringle B,
[132] Day TJ, Schmelzer JD, Low PA. Aortic occlusion and Brüggemann J, Bódis K, et al. Patterns of cutaneous
reperfusion and conduction, blood flow, and the nerve fiber loss and regeneration in type 2 diabetes
blood-nerve barrier of rat sciatic nerve. Exp Neurol with painful and painless polyneuropathy.
1989;103:173 8. Diabetologia 2017;60:2495 503.
[133] Zochodne DW, Cheng C, Sun H. Diabetes increases [143] Sugimoto K, Murakami H, Deguchi T, Arimura A,
sciatic nerve susceptibility to endothelin-induced Daimon M, Suzuki S, et al. Cutaneous microangiopa-
ischemia. Diabetes 1996;45:627 32. thy in patients with type 2 diabetes: impaired vascu-
[134] Nukada H, McMorran PD, Baba M, Ogasawara S, lar endothelial growth factor expression and its
Yagihashi S. Increased susceptibility to ischemia and correlation with neuropathy, retinopathy and
macrophage activation in STZ-diabetic rat nerve. nephropathy. J Diabetes Investig 2019;10:1318 31.
Brain Res 2011;1373:172 82. [144] Yagihashi S, Mizukami H, Sugimoto K. Mechanism
[135] Nukada H, McMorran PD. Perivascular demyelin- of diabetic neuropathy: where are we now and where
ation and intramyelinic oedema in reperfusion nerve to go? J Diabetes Investig 2011;2:18 32.
injury. J Anat 1994;185:259 66. [145] Yagihashi S. Glucotoxic mechanisms and related
[136] Nagamatsu M, Schmelzer JD, Zollman PJ, Smithson therapeutic approaches. Int Rev Neurobiol
IL, Nickander KK, Low PA. Ischemic reperfusion 2016;127:121 49.
causes lipid peroxidation and fiber degeneration. [146] Feldman EL, Nave KA, Jensen TS, Bennet DLH. New
MuscleNerve 1996;19:37 47. horizons in diabetic neuropathy: mechanisms, bioen-
[137] Hill R. Extracellular matrix remodelling in human ergetics, and pain. Neuron 2017;93:1296 313.
diabetic neuropathy. J Anat 2009;214:219 25. [147] Beirowski B, Babetto E, Golden JP, Chen Y-J, Yang K,
[138] Behse F, Buchthal F, Carlsen F, Knappeis GG. Gross RW, et al. Metabolic regulator LKB1 is crucial
Endoneurial space and its constituents in the sural nerve for Schwann cell-mediated axon maintenance. Nat
of patients with neuropathy. Brain 1974;97:773 84. Neurosci 2014;17:1351 61.
[139] Bradley JL, King RH, Muddle JR, Thomas PK. The [148] Viader A, Sasaki Y, Kim S, Strickland A, Workman
extracellular matrix of peripheral nerve in diabetic CS, Yang K, et al. Aberrant Schwann cell lipid
neuropathy. Acta Neuropathol 2000;99:539 46. metabolism linked to mitochondrial deficits leads
[140] King RHM, Llewelyn JG, Thomas PK, Gilbey SG, to axon degeneration and neuropathy. Neuron
Watkins PJ. Diabetic neuropathy: abnormalities of 2013;77:886 98.

Diabetic Neuropathy
 C H A P T E R

3
Electrodiagnostic evaluation in diabetic
neuropathy
Deepak Menon and Vera Bril
Ellen and Martin Prosserman Centre for Neuromuscular Diseases, University Health Network,
University of Toronto, Toronto, ON, Canada

Introduction reliable than clinical examination and accurately

identify subclinical neuropathy before the onset
The diagnosis of diabetic sensorimotor poly- of symptoms. NCS also provide an objective, reli-
neuropathy (DSP) depends heavily on the able and noninvasive assessment tool and bio-
demonstration of objective electrophysiological marker in clinical trials of DSP and also help in
nerve abnormalities through nerve conduction excluding alternate causes of neuropathy. In fact,
studies (NCS) along with the assessment of NCS provide insight into the pathophysiological
clinical signs and symptoms [1]. NCS are and functional properties of large nerve fibers
highly reproducible and the most reliable, non- helping demonstrate whether the pathology is
invasive method for assessing DSP, especially axonal, demyelinating, or a combination of both.
in the subclinical stages [2,3]. The American Hence electrophysiological studies should be con-
Diabetes Association (ADA) recommends sidered as an extension of the physical examina-
screening for diabetic neuropathy in those with tion and are valuable in the evaluation of DSP.
type 2 diabetes mellitus (DM) at diagnosis, and
5 years after diagnosis with subsequent annual
evaluations in those with type 1 DM [4]. While Pathophysiology of diabetic sensorimotor
NCS are not recommended for routine screen- polyneuropathy
ing for DSP, various studies have shown that
as many as 30% 40% of those referred with DSP is a chronic, symmetric, predominantly
presumptive DSP have alternate diagnoses sensory, length-dependent, large and small
such as lumbosacral radiculopathies, focal neu- fiber polyneuropathy with progressive axonal
ropathies, or no neuropathy at all after degeneration and endoneurial microangiopa-
electrophysiological testing, indicating the thy constituting the basic pathology [6,7]. The
potential inaccuracy of performing only clinical increased glucose levels channel metabolism into
evaluation [5]. NCS are more sensitive and alternate polyol and hexosamine pathways that

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00005-0 35 © 2022 Elsevier Inc. All rights reserved.
36 3. Electrodiagnostic evaluation in diabetic neuropathy

generate increased oxidative free radicals and F-wave latencies are determined by the fastest
the formation of advanced glycosylation end conducting nerve fibers. These conduction
(AGE) products. These, in turn, modify gene parameters depend on the degree of myelina-
expression resulting in an inflammatory cascade tion, cross sectional nerve area and nodal and
[8]. The pathological changes that ensue are internodal integrity.
observed as inflammatory mononuclear infil- According to classical teaching, an axonal
trates, complement deposition, microvasculitis neuropathy is characterized by reduced CMAP
and necrotizing arteritis resulting in an ischemic and SNAP amplitudes with relative preserva-
neuropathy [9,10]. Axonal degeneration and tion of conduction velocities and distal laten-
regeneration are more frequent than segmental cies. In demyelinating polyneuropathies,
demyelination and remyelination, and progres- CMAP and SNAP amplitudes are preserved
sive axonal loss is considered to be the chief and conduction velocities reduced, and distal
pathological hallmark in diabetic polyneuropa- latencies and F-wave latencies are prolonged.
thy [6,11]. Differences in nerve pathology However, reduced amplitudes alone are not
between type 1 and type 2 DSP have been conclusive proof of axonopathy since reduced
described: in type 1 diabetes, dysfunction is ini- amplitudes can be observed in demyelinating
tially focused at the nodes of Ranvier and milder pathology due to a combination of temporal
axonal changes are observed whereas in type dispersion and conduction block. Similarly,
2 DM, nerve fiber loss with focal and vascular slowed conduction velocity or prolonged distal
pathology are more common [12]. latency are not definitive evidence of demye-
lination as these findings are observed with
loss of the larger and faster conducting fibers
Nerve conduction parameters and in axonopathy [13]. A combination of para-
correlation with pathology meters, including F-wave latencies, conduction
block, temporal dispersion and proximal and
The various parameters assessed in NCS distal segment conduction velocities and
represent distinct nerve physiological proper- amplitudes, needs to be considered when
ties. The sensory nerve action potential defining the electrophysiological pathology
(SNAP) amplitude reflects the number of sen- [13]. Reference values for NCS parameters are
sory nerve fibers while the compound muscle determined in each laboratory and cut-offs for
action potential (CMAP) amplitude is ulti- axonal versus demyelinating changes can be
mately a representation of the number and variable [14 16].
size of muscle fibers in the activated motor
units which in turn depends on the number of
axons or neurons in the corresponding periph- Physiological changes in nerve function
eral nerve. So, loss of sensory or motor nerve in diabetic sensorimotor polyneuropathy
fibers leads to loss of corresponding ampli- and correlation with nerve conduction
tudes. In addition, a loss of synchronicity of studies
firing of motor fibers due to variably reduced
conduction velocities leads to temporal dis- In DSP, the nerves undergo both functional
persion or conduction block and these reduce and structural changes. From a physiological
the CMAP amplitude. Loss of synchronicity in point of view, early in the disease process,
sensory nerve firing leads to reduced SNAP there occurs a combination of microangiopa-
amplitudes. The motor and sensory nerve con- thy, oxidative stress and energy failure in the
duction velocities (NCV), onset latencies and neurons resulting in impaired axonal transport,

Diabetic Neuropathy
 Which nerves to choose and what parameters to test 37
dysregulated Na/K ion channels, increased Na may also reflect other pathological processes
currents and alteration of membrane excitabil- including nerve entrapments [19].
ity [17]. The transmembrane ion channels
located in the nodal and paranodal regions
have a vital role in normal impulse propaga- Which nerves to choose and what
tion and responsiveness of nerves. The altered parameters to test
distribution of these ion channels and widen-
ing of the nodal gap may be responsible for the When selecting the protocol for NCS in
reversible nerve conduction slowing in newly DSP, factors such as balancing sensitivity with
diagnosed and poorly controlled diabetes, par- specificity, technical ease and patient comfort,
ticularly type 1 [7,18]. Action potential trans- reproducibility, availability of normative data,
mission occurs in a saltatory fashion from one comparison across studies, etc. have to be con-
node of Ranvier to the next, and the thickness sidered. Being a diffuse and symmetrical pro-
of myelin and internodal distance determine cess, unilateral NCS are sufficient in DSP [20].
the conduction velocity. During the process of Studies differ in the selection of nerves and
remyelination, the number of nodes increase, the methods of nerve testing. In the largest
internodal intervals decrease, and the myelin report of more than sixty thousand patients
thickness relative to axon diameter decreases. with DSP, the most common nerve conduction
All of these factors result in reduced conduction abnormalities were abnormalities in sural
velocity [2]. As the disease progresses, more SNAP amplitudes (62.7%) and latencies
apparent structural changes evolve, from myelin (58.3%), followed by peroneal F-wave laten-
swelling to axonal pathology, especially at the cies (33.6%), CMAP amplitudes (32.6%), and
nodes of Ranvier, and frank segmental demyelin- distal motor latencies (14%) [5]. This clearly
ation can be observed. Eventually, there is a loss confirms the wisdom of choosing lower limb
of nerve regeneration culminating in Wallerian nerves, that is sural and peroneal nerves, for
degeneration, inexorable progressive axonal loss, testing in diabetic neuropathy [5]. An earlier
and reduction in nerve fiber density. Sensory report emphasized the need for lower limb
nerves are the most sensitive in DM, and the sen- NCS abnormalities to make a diagnosis of
sory NCS are the earliest to show changes with polyneuropathy for research purposes, and
loss of SNAP amplitudes beginning distally in 75% of the articles that formed the basis of
the lower limb. Being a length-dependent pro- this report were on DSP [19]. In the Diabetes
cess, NCS changes would be observed first in the Control and Complications Trial (DCCT),
legs and then later in the hands [18]. It is possible median, peroneal, and sural NCS were tested
to find axonal changes in the lower limbs when [20]. In another report, distal tibial CMAP
demyelinating features are present in the upper amplitudes correlated well with the severity
limbs, but this does not imply that demyelination of DSP [21]. The Toronto expert panel on DSP
occurs first in DSP as upper limb nerves suffer recommended testing six nerve conduction
compression frequently at carpal tunnels and attributes—peroneal, tibial and ulnar motor
cubital tunnels , and this produces demyelinat- NCV, sural SNAP amplitudes, peroneal and
ing change that clouds the diagnosis of DSP [19]. ulnar F-wave latencies [22 24]. Tibial NCS
At the later stages of DSP with marked axonal can be challenging due to technical limitations
loss, a marked drop in CMAP amplitudes is of adequate proximal stimulation at the popli-
observed with some slowing of motor NCV. So, teal fossa. Ulnar nerve parameters can be
it is important to keep in mind that the NCS compromised by compression at the cubital
besides showing pathological changes of DSP, tunnel. We recommend routine testing of

Diabetic Neuropathy
38 3. Electrodiagnostic evaluation in diabetic neuropathy

upper and lower limb motor and sensory abnormal findings in one or more parameters in
nerves in the evaluation of DSP: sural SNAPs two or more lower limb nerves are sufficient for
and conduction velocities, peroneal CMAPs, diagnosis [19,23].
conduction velocities and F-wave latencies,
and median motor and sensory nerve para-
meters. Ulnar and tibial nerve studies can be Evolution from subclinical to clinical
added as deemed necessary in the electrodiag- diabetic sensorimotor polyneuropathy
nostic evaluation although both have technical
challenges compared to median and peroneal Subclinical diabetic neuropathy is defined as
NCS. Various studies have examined the util- the absence of signs or symptoms of neuropa-
ity of medial and lateral plantar sensory thy in the presence of abnormal NCS or other
nerves, dorsal sural nerve, superficial radial validated objective measures of neuropathy [1].
nerve, and sural/radial SNAP ratio [25]. Asymptomatic DSP refers to those who have
While each of these studies has potential bene- abnormalities on examination and abnormal
fits, technical difficulties, lack of ease of repro- objective tests, but no symptoms of DSP. NCS
ducibility, and less-than-expected sensitivity can reveal abnormalities well before symptom-
limit their clinical utility and confine them to atic manifestations [3,28] and the sensitivity of
select situations or to the research setting [26]. NCS is further increased by using composite
scores of amplitude, velocity, and latency from
various lower limb nerves [3]. As DSP pro-
gresses from the asymptomatic to symptomatic
How to diagnose diabetic sensorimotor stages, serial NCS changes have been noted. In
polyneuropathy with nerve conduction the presymptomatic stage, motor nerve con-
studies duction velocity slowing and reduction in
SNAP amplitudes are seen [29]. The reduced
A strict adherence to standardized technical SNAP amplitudes at this stage might be a function
protocols is a prerequisite for accurate measure- of demyelination resulting in temporal dispersion
ment and interpretation of NCS results. These rather than a loss in nerve fibers [30] although
include appropriate skin preparation and elec- pathological studies have shown loss of fibers cor-
trode placements, ensuring supramaximal stimu- relating with loss of sural SNAP amplitudes and
lation, monitoring limb temperature, accurate clinical parameters indicating the primacy of axo-
measurements including stimulator electrode nal loss [31,32]. The conduction velocity gradually
distances, limb positions and cursor placements diminishes at around 0.5 m/s/year [30]. As the
on the wave forms. In the Rochester Neuropathy disease process evolves, the electrophysiological
Study, a combination of nerve conduction abnor- changes progress: prolongations in distal motor,
mality in two or more nerves and quantitative F-wave and sensory latencies, slowing of sensory
autonomic examination was the most sensitive, and motor NCV, and reduction in CMAP and
objective, and best suited protocol for detection SNAP amplitudes. Both conduction velocities and
of subclinical neuropathy, whereas the sum- amplitudes are significantly lower in symptomatic
mated CMAP amplitude of ulnar, peroneal, and patients, and these changes in NCS are associated
tibial nerves was best for judging severity [27]. A with abnormal electromyographic signals. These
composite score of nerve conduction parameter findings suggest that early changes in nerve func-
deviation from normal, was specific and sensi- tion and myelin progress to axonal pathology by
tive in the diagnosis of diabetic neuropathy [23]. the time patient starts having symptomatic poly-
However, in the clinical and research settings, neuropathy [32]. A differential pattern between

Diabetic Neuropathy
 Impact of treatment in nerve conduction studies 39
TABLE 3.1 Evolution of pathophysiological changes in diabetic neuropathy and their electrophysiological correlates.
Stage Pathophysiological changes Clinical characteristics Electrophysiological correlate

I Small fiber dysfunction Burning, lancinating pain, Conventional NCS normal

hyperalgesia, allodynia

II Large fiber involvement; ion channel Pins and needle, tingling Mild slowing of sensory conduction parameters;
dysfunction at nodal and paranodal regions paresthesia can be reversible with treatment

III Beginning of structural changes, reduced myelin Paresthesias and dysesthesia Slowing of nerve conduction velocities and other
thickness, increased nodal number and reduced progressing in a glove and features of demyelination; early reduction in
internodal interval, segmental demyelination, stocking pattern SNAP and CMAP amplitudes
early axonal pathology

IV Wallerian degeneration, progressive axonal loss Paresthesia could be Reduction in SNAP more than CMAP, lower
and reduction in nerve fiber density replaced by numbness, gait, limbs more than upper limbs; can have additional
and balance issues emerge features of entrapment neuropathies especially in
upper limbs

CMAP, compound muscle action potential; NCS, nerve conduction studies; SNAP, sensory nerve action potential.

upper and lower limb nerve conduction with neuropathy in type 1 and type 2 DM [35].
an axonal pattern of injury being more common in More severe disease, in terms of NCS, was
the legs and a “demyelinating” pattern in the seen in type 2 DM in some studies, but the
arms is also noticeable although upper limb impact of age and other confounders make this
nerves are susceptible to nerve compressions that association questionable [36]. One notable dif-
can confound the diagnosis. This difference is ference in NCS between types 1 and 2 DM has
more apparent in younger patients [33]. As the been the response to treatment in the early
symptoms worsen and disease progresses, more stages of DSP. Isolated nerve conduction slow-
deterioration in NCS is observed. In a 10-year ing that is prominently associated with worse
study of 36 patients with type 2 DM, the preva- glycemic control and is partially reversible
lence of NCS abnormalities increased from 8.3% at with treatment can be observed in type 1 DM,
baseline to 16.7% at 5 years and 41.9% at 10 years but this has not been found in type 2 DM [14].
[30]. The deterioration was in SNAP and motor In all other aspects, NCS parameters cannot
CMAP amplitudes rather than in conduction differentiate type 1 from type 2 DSP.
velocities indicating predominantly axonal degen-
eration with time and this worsening correlated
with the level of glycemic control [34]. An over- Impact of treatment in nerve conduction
view of the evolution of pathological and studies
electrophysiological changes is given in Table 3.1.
The concept that DSP is fundamentally irre-
versible may not be entirely accurate, particu-
Nerve conduction studies features in larly in the early stages. Nerve conduction
diabetic sensorimotor polyneuropathy: slowing and predominant demyelinating fea-
type 1 versus type 2 diabetes mellitus tures, out of proportion to axonal changes, are
seen with poor glycemic control and may be
Differences in the pathophysiology of type 1 reversible [14].
and type 2 DSP have been described although Type 1 DM patients in the DCCT who
extensive electrophysiological testing has not received 5 years of intensive therapy had a
found any differences between subclinical 64% reduced risk of DSP, and the NCV

Diabetic Neuropathy
40 3. Electrodiagnostic evaluation in diabetic neuropathy

remained stable (motor) or improved (sensory improved nerve function and or ion channel
and some motor) [37]. The beneficial effect of function rather than nerve axon structure or
treatment with pancreatic transplant was number.
observed in type I DM patients who had signif- Although earlier studies on NCS showed
icant improvement in motor NCV in upper more favorable changes, later studies have not
and lower limbs, increased CMAP amplitudes confirmed these responses, and overall the
in upper limbs and improved median sensory effects of improved glycemic control in type 2
NCV and SNAP amplitudes as early as 1 year DSP are less robust [42]. The magnitude of any
after surgery [38]. Similar improvements in NCS changes is small, perhaps only 1 3 m/s in
NCS were observed in a Japanese population nerve conduction velocity in 3-year studies [43].
of type 2 DM treated with multiple insulin These changes in NCV reflect the improved met-
injections [35]. These patients had improved abolic state and ion channel activity at the nodes
median motor and sensory NCV compared to of Ranvier as discussed previously. Early meta-
those on conventional therapy who experi- bolic dysfunction of myelinated fibers can be
enced a decline in median NCV [39]. directly related to the Na1/K1-ATPase pathway,
A perplexing finding has been the improve- whereas the impact of impaired endoneurial
ment of symptoms and even NCS in placebo- blood flow is seen later and would be less
treated patients in different intervention reversible.
studies for DSP. While the symptomatic Finally, a meta-analysis examined the effects
improvement may be explained partially by of enhanced glucose control in type 1 DM, and
the “Hawthorne effect”—behavioral alteration showed prevention of clinical neuropathy and
due to study participation—improvements in also reduction in nerve conduction abnormali-
NCS are more difficult to explain. This might ties [40]. However, similar benefits on clinical
be explained by short-term improvements in neuropathy were not apparent in type 2 DM,
glycated hemoglobin and/or triglyceride levels although there was significant reduction in
after enrolling in a clinical trial that lead to a nerve conduction abnormalities [44]. The pres-
more sustained improvement in sensory nerve ence of multiple additional vascular and meta-
conduction parameters [40]. Such placebo bolic risk factors, in addition to hypoglycemia
responses might be avoided by having a pla- and weight gain, might offset the benefits of
cebo run-in period at the start of clinical trials tight glycemic control in type 2 DM [42].
in DSP.
Another interesting observation is that a few
studies have showed rapid control of hypergly- Atypical features in nerve conduction
cemia in both types of DM to favorably alter studies—demyelinating features in
nerve conduction parameters without any signif- diabetic sensorimotor polyneuropathy and
icant clinical improvement [37]. In patients who when to suspect chronic inflammatory
had preserved sural SNAP amplitudes at base- demyelinating polyradiculoneuropathy
line, there was a significant improvement in con-
duction velocities, distal latencies and F-wave Some clinical features raise the suspicion of
latencies from baseline. The improvements were an etiology alternate to diabetes as the cause of
less marked in patients with low baseline sural neuropathy [41]. These include prominent
SNAP amplitudes, but remained significant in asymmetry, upper limb onset or prominent
the median nerve parameters [41]. Motor and upper limb symptoms, significant motor signs
sensory amplitudes did not change indicating and symptoms, rapid progression and a posi-
that the improved parameters were due likely to tive family history [45]. Besides these clinical

Diabetic Neuropathy
 Nerve conduction studies in predicting prognosis 41
factors, certain demyelinating findings on NCS response was similar [52]. A combination of clin-
suggest a nondiabetic etiology, although DSP is ical and electrophysiological features needs to be
likely the result of a mixed axonal and demyelin- considered before making a diagnosis of CIDP in
ating process similar to most other neuropathies. DM. Electrophysiologically, prominent demye-
Nerve conduction slowing independent of the linating features with significant slowing of con-
CMAP amplitude reduction has been described duction velocities as well as conduction blocks
in DSP, and this is found mainly early in the and temporal dispersion would increase the pos-
course of DSP in poorly controlled type 1 patients sibility of CIDP [23]. Proximal segment NCS
[46]. Demyelinating features such as conduction would also be beneficial since the changes in
block and temporal dispersion are not common DSP affect distal more than proximal segments
in DSP unless the patient has coexisting condi- as this is a length-dependent process, whereas in
tions such as focal entrapment neuropathies, CIDP conduction velocity slowing in proximal
MGUS-associated neuropathy or chronic inflam- segments can be seen. Also, compared to CIDP
matory demyelinating polyradiculoneuropathy alone, patients with CIDP 1 DM have more axo-
(CIDP) [47]. nal features [50]. Strict electrophysiological crite-
CIDP has a higher prevalence in patients ria for CIDP are highly specific, but may fail to
with diabetes with odds ranging from nine to identify CIDP in DM and therefore miss a
eleven times higher compared to nondiabetics treatable neuropathy in diabetes patients. For
[48,49]. There was no difference between type clinical practice, the evolution of new-onset pro-
1 and 2 DM in terms of association with CIDP. gressive symmetric proximal and/or distal
Other studies have questioned such an associa- weakness in a patient with DSP, associated with
tion between CIDP and DM, however these significant demyelinating features on NCS
included only patients with “classical” CIDP should lead to the suspicion of possible superim-
based on both clinical and electrophysiological posed CIDP, and a trial of immunomodulatory
criteria and excluded patients if they satisfied treatment. The presence of more prominent
only demyelinating features by NCS, limiting demyelinating features in patients with sus-
the sensitivity for the diagnosis of CIDP and pected CIDP and DSP makes a favorable
therefore the number of patients [50]. response to treatment more likely [53]
Diagnosing CIDP in diabetes subjects is chal- (Table 3.2).
lenging especially since demyelinating features
may be found in DSP and the preexisting dia-
betic neuropathy can mask the demyelinating Nerve conduction studies in predicting
changes in CIDP. Significant differences were prognosis
noted when comparing patients with CIDP alone
to those with DSP having demyelinating features In a 6-year prospective study, motor NCV
(D-DSP). CIDP patients were older, had better were an independent risk factor and the best
glycemic control, a shorter duration of diabetes, predictor of new foot ulcers in patient with dia-
and more diffuse reflex changes and nerve inju- betes [50]. The common peroneal nerve con-
ries than patients with D-DSP pointing to a dif- duction values were 36.7 m/s, 34.6 m/s, and
ferent pathology of demyelination in the two 31 m/s for DSP, diabetic foot ulcer and diabetic
conditions [51]. On the other hand, patients with Charcot arthropathy respectively, whereas
CIDP and diabetes (CIDP 1 DM) compared to peroneal motor NCV was 43.4 m/s in diabetic
CIDP without diabetes (CIDP-DM) had a more subjects without symptoms of neuropathy.
severe neuropathy, more proximal weakness Other than creatinine levels and skin oxygen
and impairment in gait, though the treatment levels, peroneal motor nerve conduction

Diabetic Neuropathy
42 3. Electrodiagnostic evaluation in diabetic neuropathy

TABLE 3.2 Atypical electrophysiological features for diabetic neuropathy.

Predominant upper limb more than lower limb findings—alternate etiologies such as entrapment neuropathies—
hereditary or acquired, amyloid neuropathies
Prolonged distal CMAP duration, temporal dispersion and conduction block at noncompression site—CIDP or its variants
Nonlength-dependent pattern with slowing of proximal more than distal segments—CIDP or its variants
Serial nerve conduction studies showing worsening of demyelinating more than axonal features—CIDP or its variants
Motor findings more than sensory—CIDP or its variants
Absent sensory with normal motor—pure sensory CIDP or ganglionopathy

CIDP, chronic inflammatory demyelinating polyradiculoneuropathy; CMAP, compound muscle action potential.

velocity also was a predictor of mortality in diagnoses such as a focal neuropathy, diabetic
this cohort [54]. lumbosacralradiculoplexusneuropathy or a coex-
isting myopathy. Fibrillation potentials are the
most sensitive indicator of axonal involvement
Vibration perception thresholds in and confirm a length-dependent process when
diabetic neuropathy observed in the distal foot muscles, but lack
specificity for DSP [58]. Fibrillation potentials can
VPTs are a quantitative way to measure sen- even predate NCS abnormalities confirming
sory function of large nerve fibers. An increase early axonal involvement [25]. With longer dura-
in VPT is one of the sensitive tests for detecting tion of DSP, reduced recruitment due to loss of
and assessing severity of DSP [55]. Among var- motor units is observed. Increased duration and
ious methods, measurement using a neurothe- polyphasic motor unit potentials can be seen
siometer was most sensitive for DSP, especially both in proximal and distal muscles, but these
in mild cases [23,52]. VPT testing by the findings correlate less with disease duration [58].
method-of-limits offers a quick, relatively inex- Single-fiber EMG is more sensitive than NCS
pensive and painless patient-friendly method and abnormalities can predate NCS changes in
for testing large nerve fibers, and this can be about 20% patients. Interestingly, increased jitter
used both for screening and monitoring of correlates with higher glycated hemoglobin in
DSP. VPT shows a good association with both type 1 DM but such a relationship is not present
sural SNAP amplitude and conduction veloc- in type 2 DM [43]. Overall, SFEMG and routine
ity, and VPT obtained using the neurothesi- EMG are less useful than NCS in the monitoring
ometer can be used reliably in clinical research of DSP as both are time-consuming and painful.
trials [56,57].

Limitations of nerve conduction studies

Electromyography in diabetic
sensorimotor polyneuropathy NCS assess only large-diameter nerve fibers
and cannot provide information on small nerve
Electromyography (EMG) provides informa- fibers that may be involved early in DSP before
tion additional to NCS in patients with DSP, and large nerve fiber changes are apparent [59].
helps confirm active axonal involvement. EMG The assessment of small nerve fibers depends
is useful when there is a suspicion of other on measures such as quantitative thermal

Diabetic Neuropathy
 References 43
threshold testing, laser Doppler flare area [2] Bril V. Role of electrophysiological studies in diabetic
imaging, sudomotor testing, corneal confocal neuropathy. Can J Neurol Sci J Can Sci Neurol
1994;21:S8 12.
microscopy, and intraepidermal nerve fiber [3] Dyck PJ, O’Brien PC, Litchy WJ, Harper CM, Klein
thickness by skin biopsy. Moreover, NCS para- CJ, Dyck PJB. Monotonicity of nerve tests in
meters of conduction velocity and latency diabetes: subclinical nerve dysfunction precedes
depend on the largest, fastest conducting nerve diagnosis of polyneuropathy. Diabetes Care
fibers and do not indicate how many fibers 2005;28:2192 200.
[4] Pop-Busui R, Boulton AJM, Feldman EL, Bril V,
have been lost; the latter depending on the Freeman R, Malik RA, et al. Diabetic neuropathy: a
motor and sensory nerve amplitudes. [60]. position statement by the American Diabetes
Other techniques, such as threshold tracking Association. Diabetes Care 2017;40:136 54.
which assesses nerve excitability irrespective of [5] Kong X, Lesser EA, Potts FA, Gozani SN. Utilization
myelin integrity, may be potential candidates of nerve conduction studies for the diagnosis of poly-
neuropathy in patients with diabetes: a retrospective
to diagnose and monitor DSP, but are not analysis of a large patient series. J Diabetes Sci
widely available or used [61]. Technol 2008;2:268 74.
[6] Dyck PJ, Giannini C. Pathologic alterations in the dia-
betic neuropathies of humans: a review. J Neuropathol
Exp Neurol 1996;55:1181 93.
Conclusion [7] Boulton AJM, Malik RA, Arezzo JC, Sosenko JM.
Diabetic somatic neuropathies. Diabetes Care
NCS are the gold standard for diagnosis of 2004;27:1458 86.
[8] Feldman EL, Callaghan BC, Pop-Busui R, Zochodne
DSP and sural SNAP amplitudes are the best
DW, Wright DE, Bennett DL, et al. Diabetic neuropa-
indicators of DSP [22]. As NCS are more sensitive thy. Nat Rev Dis Prim 2019;5(1):42.
than the clinical examination, they can [9] Dyck PJ, Karnes JL, O’Brien P, Okazaki H, Lais A,
be used in patients with diabetes to measure Engelstad J. The spatial distribution of fiber loss in
underlying nerve function and pathophysiology. diabetic polyneuropathy suggests ischemia. Ann
Neurol 1986;19:440 9.
Clinicians should be aware that axonal changes
[10] Younger DS. Diabetic neuropathy: a clinical and
are fundamental in DSP although there is a spec- neuropathological study of 107 patients. Neurol Res
trum including demyelination, and the findings Int 2010;2010:140379.
will depend on the duration of diabetes and level [11] Yagihashi S, Mizukami H, Sugimoto K. Mechanism of
of glycemic control. Electrophysiological testing diabetic neuropathy: where are we now and where to
go? J Diabetes Investig 2011;2:18 32.
also allows exclusion of other diagnoses that can
[12] Sima AA, Cherian PV. Neuropathology of diabetic
cause nerve damage in diabetes subjects. Finally, neuropathy and its correlations with neurophysiology.
the presence of major demyelinating features Clin Neurosci (New York, NY) 1997;4:359 64.
should raise the suspicion of a coexisting [13] Tankisi H, Pugdahl K, Johnsen B, Fuglsang-
immune-mediated neuropathy. NCS also provide Frederiksen A. Correlations of nerve conduction mea-
sures in axonal and demyelinating polyneuropathies.
a prognostic marker and can help identify
Clin Neurophysiol 2007;118:2383 92.
patients who are at risk of imminent foot ulcers [14] Dunnigan SK, Ebadi H, Breiner A, Katzberg HD,
and early mortality. Lovblom LE, Perkins BA, et al. Conduction slowing in
diabetic sensorimotor polyneuropathy. Diabetes Care
2013;36:3684 90.
[15] Van den Bergh PYK, Hadden RDM, Bouche P,
References Cornblath DR, Hahn A, Illa I, et al. European
[1] Tesfaye S, Boulton AJM, Dyck PJ, Freeman R, Horowitz Federation of Neurological Societies/Peripheral Nerve
M, Kempler P, et al. Diabetic neuropathies: update on Society guideline on management of chronic inflam-
definitions, diagnostic criteria, estimation of severity, matory demyelinating polyradiculoneuropathy: report
and treatments. Diabetes Care 2010;33:2285 93. of a joint task force of the European Federation of

Diabetic Neuropathy
44 3. Electrodiagnostic evaluation in diabetic neuropathy

Neurological Societies and the Peripheral Nerve [28] Walter-Höliner I, Barbarini DS, Lütschg J, Blassnig-
Society - first revision. Eur J Neurol 2010;17:356 63. Ezeh A, Zanier U, Saely CH, et al. High prevalence
[16] Breiner A, Brannagan TH. Comparison of sensitivity and incidence of diabetic peripheral neuropathy in
and specificity among 15 criteria for chronic inflam- children and adolescents with type 1 diabetes mellitus:
matory demyelinating polyneuropathy. Muscle Nerve results from a five-year prospective cohort study.
2014;50:40 6. Pediatr Neurol 2018;80:51 60.
[17] Deli G, Bosnyak E, Pusch G, Komoly S, Feher G. [29] de Souza RJ, de Souza A, Nagvekar MD. Nerve con-
Diabetic neuropathies: diagnosis and management. duction studies in diabetics presymptomatic and
Neuroendocrinology 2013;98:267 80. symptomatic for diabetic polyneuropathy. J Diabetes
[18] Arezzo JC, Zotova E. Electrophysiologic measures of Complications 2015;29:811 17.
diabetic neuropathy: mechanism and meaning. Int [30] Arezzo JC. The use of electrophysiology for the assess-
Rev Neurobiol 2002;50:229 55. ment of diabetic neuropathy. Neurosci Res Commun
[19] England JD, Gronseth GS, Franklin G, Miller RG, 1997;21:13 23.
Asbury AK, Carter GT, et al. Distal symmetric poly- [31] Mohseni S, Badii M, Kylhammar A, Thomsen NOB,
neuropathy: a definition for clinical research: report Eriksson K, Malik RA, et al. Longitudinal study of
of the American Academy of Neurology, the neuropathy, microangiopathy, and autophagy in sural
American Association of Electrodiagnostic nerve: implications for diabetic neuropathy. Brain
Medicine, and the American Academy of Physical Behav 2017;7:e00763.
Medicine and Rehabilitation. Neurology [32] Valls-Canals J, Povedano M, Montero J, Pradas J.
2005;64:199 207. Diabetic polyneuropathy. Axonal or demyelinating?
[20] Diabetes Control and Complications Trial Research Electromyogr Clin Neurophysiol 2002;42:3 6.
Group, Nathan DM, Genuth S, Lachin J, Cleary P, [33] Bagai K, Wilson JR, Khanna M, Song Y, Wang L,
Crofford O, et al. The effect of intensive treatment of Fisher MA. Electrophysiological patterns of diabetic
diabetes on the development and progression of long- polyneuropathy. Electromyogr Clin Neurophysiol
term complications in insulin-dependent diabetes mel- 2008;48:139 45.
litus. N Engl J Med 1993;329(14):977 86. [34] Partanen J, Niskanen L, Lehtinen J, Mervaala E,
[21] Hasegawa O, Matsumoto S, Iino M, Mori I, Arita T, Siitonen O, Uusitupa M. Natural History of peripheral
Baba Y. [Interrelationship among nerve conduction neuropathy in patients with non-insulin-dependent
velocity, amplitudes of compound muscle and com- diabetes mellitus. N Engl J Med 1995;333(2):89 94.
pound nerve action potentials in diabetic neuropathy]. [35] Hendriksen PH, Oey PL, Wieneke GH, Bravenboer B,
No To Shinkei 1999;51:863 6. Banga JD. Subclinical diabetic neuropathy: similarities
[22] Dyck PJ, Albers JW, Andersen H, Arezzo JC, Biessels between electrophysiological results of patients with
G-J, Bril V, et al. Diabetic polyneuropathies: update on type 1 (insulin-dependent) and type 2 (non-insulin-
research definition, diagnostic criteria and estimation dependent) diabetes mellitus. Diabetologia
of severity: diabetic polyneuropathies. Diabetes Metab 1992;35:690 5.
Res Rev 2011;27:620 8. [36] Albers JW, Brown MB, Sima AAF, Greene DA. Nerve
[23] Albers JW, Pop-Busui R. Diabetic neuropathy: conduction measures in mild diabetic neuropathy in
mechanisms, emerging treatments, and subtypes. Curr the Early Diabetes Intervention Trial: the effects of
Neurol Neurosci Rep 2014;14:473. age, sex, type of diabetes, disease duration, and
[24] Dyck PJ, Carter RE, Litchy WJ. Modeling nerve con- anthropometric factors. Neurology 1996;46:85 91.
duction criteria for diagnosis of diabetic polyneuropa- [37] The Diabetes Control and Complications Trial
thy. Muscle Nerve 2011;44:340 5. Research Group. The effect of intensive diabetes ther-
[25] Perkins B, Bril V. Electrophysiologic testing in dia- apy on the development and progression of neuropa-
betic neuropathy. Handb Clin Neurol thy. Ann Intern Med 1995;122:561 8.
2014;126:235 48. [38] Navarro X, Sutherland DER, Kennedy WR. Long-term
[26] Barnett C, Perkins BA, Ngo M, Todorov S, Leung R, effects of pancreatic transplantation on diabetic neu-
Bril V. Sural-to-radial amplitude ratio in the diagnosis ropathy. Ann Neurol 1997;42:727 36.
of diabetic sensorimotor polyneuropathy. Muscle [39] Ohkubo Y, Kishikawa H, Araki E, Miyata T, Isami S,
Nerve 2012;45:126 7. Motoyoshi S, et al. Intensive insulin therapy prevents
[27] Dyck PJ, Karnes JL, O’Brien PC, Litchy WJ, Low PA, the progression of diabetic microvascular complica-
Melton LJ. The Rochester Diabetic Neuropathy Study: tions in Japanese patients with non-insulin-dependent
reassessment of tests and criteria for diagnosis and diabetes mellitus: a randomized prospective 6-year
staged severity. Neurology 1992;42:1164. study. Diabetes Res Clin Pract 1995;28:103 17.

Diabetic Neuropathy
 References 45
[40] Perkins BA, Dholasania A, Buchanan RA, Bril V. diabetes patients with “demyelinating” diabetic senso-
Short-term metabolic change is associated with rimotor polyneuropathy to those diagnosed with
improvement in measures of diabetic neuropathy: a 1- CIDP. Brain Behav 2013;3:656 63.
year placebo cohort analysis: short-term metabolic [52] Dunnigan SK, Ebadi H, Breiner A, Katzberg HD,
change and improved nerve function. Diabet Med Barnett C, Perkins BA, et al. The Characteristics of
2010;27:1271 9. chronic inflammatory demyelinating polyneuropathy
[41] Kikkawa Y, Kuwabara S, Misawa S, Tamura N, Kitano in patients with and without diabetes an observa-
Y, Ogawara K, et al. The acute effects of glycemic con- tional study. PLoS One 2014;9:e89344.
trol on nerve conduction in human diabetics. Clin [53] Abraham A, Alabdali M, Qrimli M, Albulaihe H,
Neurophysiol 2005;116:270 4. Breiner A, Barnett C, et al. Treatment responsiveness
[42] Ang L, Jaiswal M, Martin C, Pop-Busui R. Glucose in CIDP patients with diabetes is associated with high-
control and diabetic neuropathy: lessons from recent er degrees of demyelination. PLoS One 2015;10.
large clinical trials. Curr Diab Rep 2014;14:528. [54] Carrington AL, Shaw JE, Abbott CA, Vileikyte L,
[43] Bril V, Werb MR, Greene DA, Sima AAF. Single-fiber Boulton AJM. Can motor nerve conduction velocity
electromyography in diabetic peripheral polyneuropa- predict foot problems in diabetic subjects over a 6-
thy. Muscle Nerve 1996;19:2 9. year outcome period? Diabetes Care 2002;25:6.
[44] Callaghan BC, Little AA, Feldman EL, Hughes RAC. [55] Martin CL, Waberski BH, Pop-Busui R, Cleary PA,
Enhanced glucose control for preventing and treating Catton S, Albers JW, et al. Vibration perception thresh-
diabetic neuropathy. Cochrane Database Syst Rev old as a measure of distal symmetrical peripheral neu-
2012;CD007543. ropathy in type 1 diabetes: results from the DCCT/
[45] Vinik A., Casellini C., Nevoret M.-L. (2000). Diabetic neu- EDIC study. Diabetes Care 2010;33:2635 41.
ropathies. In: Feingold K.R., Anawalt B., Boyce A., [56] Bril V, Kojic J, Ngo M, Clark K. Comparison of a neu-
Chrousos G., Dungan K., Grossman A., et al., editors. rothesiometer and vibration in measuring vibration
Endotext, South Dartmouth (MA): MDText.com, Inc. perception thresholds and relationship to nerve con-
,https://www.ncbi.nlm.nih.gov/books/NBK279175/.. duction studies. Diabetes Care 1997;20:1360 2.
[46] Wilson JR, Stittsworth JD, Kadir A, Fisher MA. [57] Bril V, Perkins BA. Comparison of vibration percep-
Conduction velocity versus amplitude analysis: evi- tion thresholds obtained with the Neurothesiometer
dence for demyelination in diabetic neuropathy. and the CASE IV and relationship to nerve conduction
Muscle Nerve 1998;21:1228 30. studies. Diabet Med 2002;19:661 6.
[47] Abu-Shakra SR, Cornblath DR, Avila OL, Chaudhry [58] Lamontagne A, Buchthal F. Electrophysiological stud-
V, Freimer M, Glass JD, et al. Conduction block in dia- ies in diabetic neuropathy. J Neurol Neurosurg
betic neuropathy. Muscle Nerve 1991;14:858 62. Psychiatry 1970;33:442 52.
[48] Bril V, Blanchette CM, Noone JM, Runken MC, [59] Breiner A, Lovblom LE, Perkins BA, Bril V. Does the
Gelinas D, Russell JW. The dilemma of diabetes in prevailing hypothesis that small-fiber dysfunction pre-
chronic inflammatory demyelinating polyneuropathy. cedes large-fiber dysfunction apply to type 1 diabetic
J Diabetes Complications 2016;30:1401 7. patients? Diabetes Care 2014;37:1418 24.
[49] Sharma KR, Cross J, Farronay O, Ayyar DR, Shebert [60] Bril V, Ellison R, Ngo M, Bergstrom B, Raynard D, Gin
RT, Bradley WG. Demyelinating neuropathy in diabe- H. Electrophysiological monitoring in clinical trials.
tes mellitus. Arch Neurol 2002;59:758 65. Neurology 1998;6.
[50] Laughlin RS, Dyck PJ, Melton LJ, Leibson C, Ransom J, [61] Shibuta Y, Nodera H, Mori A, Okita T, Kaji R.
Dyck PJB. Incidence and prevalence of CIDP and the Peripheral nerve excitability measures at different tar-
association of diabetes mellitus. Neurology 2009;73:39 45. get levels: the effects of aging and diabetic neuropa-
[51] Dunnigan SK, Ebadi H, Breiner A, Katzberg HD, thy. J Clin Neurophysiol Publ Am Electroencephalogr
Lovblom LE, Perkins BA, et al. Comparison of Soc 2010;27:350 7.

Diabetic Neuropathy
 C H A P T E R

4
Advances in screening, early diagnosis,
and accurate staging of diabetic
neuropathy
Josie Carmichael1, Hassan Fadavi2, Fukashi Ishibashi3,
Angela C. Shore1 and Mitra Tavakoli1
1
Exeter Centre of Excellence for Diabetes Research, NIHR Exeter Clinical Research Facility and
Institute of Biomedical and Clinical Sciences, University of Exeter Medical School, Exeter,
United Kingdom 2Peripheral Neuropathy Group, Imperial College, London, United Kingdom
3
Ishibashi Medical and Diabetes Centre, Hiroshima, Japan

Introduction There were 4.8 million people with diabetes

in the United Kingdom alone in 2019, and this
Diabetes is one of the fastest-growing health is rising. Figures from Diabetes UK shows that
challenges of the 21st century, with the number someone is diagnosed with diabetes every two
of adults living with diabetes having more minutes, with 5.3 million expected to be living
than tripled over the past 20 years [1]. The with the condition by 2025 [3].
International Diabetes Federation reported that Diabetes is strongly associated with both
in 2019, the prevalence of diabetes was 9.3% micro- and macrovascular complications. As a
(463 million people worldwide), with a pre- result, 10% of global health expenditure, equal to
dicted rise to 10.9% (700 million people) by USD 760 billion, is directed toward diabetes and
2045 [2]. Furthermore, it has been shown that its complications [2]. Microvascular changes lead
over 1.1 million children and adolescents below to nephropathy, retinopathy and neuropathy.
20 years have type 1 diabetes. On top of these Among these diabetes-associated complications,
staggering figures are the number of people diabetic peripheral neuropathy (DPN) is the
with impaired glucose tolerance (IGT) or meta- most common and costly, occurring in around
bolic syndrome at 373.9 million in 2019 (7.5%) 50% of individuals with diabetes [4]. Distal sym-
and a predicted rise to 548.4 million (8.6%) by metric polyneuropathy (DSPN) [5] typically fol-
2045 [2]. lows a distal-proximal course and results in

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00018-9 47 © 2022 Elsevier Inc. All rights reserved.
48 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

symmetrical symptoms and signs between the prevalence of diabetes, there is a huge strain
body’s left and right sides. Common symptoms on NHS budgets that are unable to keep up.
include burning, numbness, tingling, pain, Autonomic neuropathies are a class of DPN
and/or weakness starting in the distal lower that share similar diffuse pathophysiology with
extremities, which progress into more extreme DSPN but differ by being largely nonsensory
symptoms of neuropathic pain in around [4]. These typically affect the cardiovascular,
10% 30% of affected patients [6,7]. Symptoms urogenital, and gastrointestinal systems. Patients
may be sporadic or constant but can be debili- may also suffer from sudomotor dysfunction,
tating and in many people lead to depression, hypoglycemia obliviousness and abnormal
sleep disorders and overall reduced quality of pupillary function [5]. Rare forms of DPN
life [8]. include mononeuropathies, polyradiculopathies,
The true prevalence of DPN is underesti- and treatment-induced neuropathies [5]. These
mated as its assessment is challenging. However, atypical forms are generally self-limiting and
DPN is recognized as the most common compli- resolve with medical management and physical
cation of diabetes. therapy, usually over several months [11].
DPN is the strongest initiating risk factor In clinical settings, there are several different
for diabetic foot ulceration (neuropathic ulcer) approaches to assess DPN, and the choice of
[9,10], and existing ulcers may be further the test will depend on the aim of testing. It is
exacerbated from damage to sensory neurons. usually sufficient in a busy clinic to establish
Resultant limb numbness causes ulcers to whether a patient is symptomatic, particularly
remain undetected for longer periods [10]; thus of painful DPN [12], and whether or not they
corrective actions are not taken nor advice are at high risk of foot ulceration, typically
sought at the early stages of the disease. Often through monofilament testing. However, to
the first sign that a person has DPN is a foot fully assess the damage and phenotype of
ulcer, which may lead to irreversible tissue DPN, sensory deficits must be detected early.
damage, lower limb amputation, and signifi- Those accurate biomarkers are available for
cant morbidity. monitoring of DPN and for use in clinical trials
People with diabetes account for more than of potential new treatments.
40% of hospitalizations for major amputations Currently, there are no simple markers for
and 73% of emergency admissions for minor early detection of DPN in routine clinical prac-
amputations in the United Kingdom. A single tice. The measures we use are crude and detect
diabetes-related foot ulcer can take over 240 the disease very late in its natural history. Even
days to put into remission and costs d8,000 per the benefits gained by standardizing clinical
annum (pa) to treat. Ulcers frequently recur and assessment with scored clinical assessments
eventually may require the amputation of a remain subjective being heavily reliant on the
lower limb. DPN is hugely costly to our NHS examiners’ interpretations.
( . d1.1 billion pa in direct medical costs) and In this chapter, we review the current
the wider UK economy (Bd4 billion) and is par- knowledge and the optimal approaches for
ticularly debilitating and distressing for patients diagnosis and screening of DPN.
and their families. It can also lead to untimely
death [11,12], with 5-year mortality ranging
from 52% to 80% after major amputation [13].
Types of nerve fibers
Furthermore, with diabetes-related lower Peripheral nerve fibers can be classified using
limb amputations increasing at the rate of Erlanger and Gasser’s classification, which defines
almost 20% pa in line with the increasing nerves based on diameter, conduction speed,

Diabetic Neuropathy
 Introduction 49
TABLE 4.1 Classification of nerve fibers in the peripheral nervous system according to modified Erlanger and Gasser.
Diameter Conduction
Classification Myelination (um) velocity (m/s) Type Function

Aα (alpha) Yes 12 22 70 120 Sensory/motor Proprioception, touch sensory, somatic

motor to extrafusal muscles
Aβ (beta) Yes 5 12 30 70 Sensory/motor Proprioception, touch/pressure sensory,
somatic motor to intrafusal muscles
Aδ (delta) Yes 1 5 5 30 Sensory Touch and cold thermoreceptors,
nociception
Aγ (gamma) Yes 2 8 15 30 Motor Somatic motor to intrafusal muscles
B Yes ,3 3 15 Autonomic Visceral afferent fibers and preganglionic
efferent fibers
C No 0.1 1.3 0.6 2 Sensory/autonomic Temperature (warm receptors), pain
perception, nociception, itching

and myelination level (Table 4.1). A-fibers have diagnosis of DPN and loss of protective sensa-
the largest diameter, with the thickest myelina- tion using simple standard tests for the identifi-
tion and fastest conduction speed, and act as cation of those at risk for diabetic foot ulcer
sensory and motor fibers within the somatic [15,16]. It is recommended that all patients with
nervous system. They may be further divided type 2 diabetes mellitus (T2DM) be screened for
into large nerve fibers that have sensory and DPN at diagnosis, and for type 1 diabetes melli-
motor functions (Aα and Aβ), and small nerve tus (T1DM), the screening should begin 5 years
fibers (Aγ which has motor functions, and Aδ, postdiagnosis [17]. After this initial screening,
which may be autonomic or sensory fibers) [14]. all patients should be reviewed annually.
Group B-fibers are small, with moderate Nerve conduction studies (NCS) are consid-
myelination and slower conduction velocities ered the gold standard for the diagnosis of
than A-fibers. B-fibers act mainly as general vis- large-fiber neuropathy. The Toronto consensus
ceral afferent and pre-ganglionic fibers and are [18] recommended the use of abnormal NCS
found only in the autonomic nervous system. with a symptom or sign to diagnose DPN.
Group C-fibers have a small diameter, low However, the need for specialist examiners
conduction velocity, and are the only unmyelin- and equipment renders NCS inappropriate as
ated group. They act as somatic, afferent fibers a screening test. Thus it is used only to confirm
that carry sensory information relating to tem- any possible/probable DN picked up postscre-
perature and pain and have autonomic functions ening using other measures [17].
such as the stimulation of the sweat glands [14]. More commonly, screening for DPN involves
history taking for neuropathic symptoms and
examination of the feet, along with a screening
Screening test [17]. Traditional screening tests benefit
The American Diabetes Association (ADA) from being quick and easy; however, such as
and the International Working Group on the NCS, these only assess large-fiber function and
Diabetic Foot (IWGDF) recommend regular are unable to detect any early changes in small
examination of people with DM for the nerve fibers. Furthermore, two systematic

Diabetic Neuropathy
50 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

reviews focusing on the use of monofilament clinical practice is ‘the presence of symptoms
testing, a commonly used screening test for and/or signs of peripheral nerve dysfunction
DPN (Fig. 4.1D), described a variation of diag- in people with diabetes after excluding other
nostic value in the current literature and a lack causes’ [5]. Based on this, the diagnostic tests
of consistency in recommended test procedure are focused on assessing the symptoms and
and interpretation [19]. Sensitivity for peripheral signs of nerve dysfunctions.
nerve fiber damage ranged from 43% to 93% There are numerous testing methods avail-
when using NCS as a reference standard [20]. able to assess the peripheral nervous system’s
Both review papers did not recommend the sole structure and function, with each test having
use of monofilament testing to diagnose periph- its own advantages and disadvantages.
eral neuropathy [20]. This is just one example of Bedside tests used to aid diagnosis of DPN—
the shortcomings of current screening tests. including the 10 g monofilament (Fig. 4.1D),
the Ipswich Touch Test, and vibration per-
ception threshold testing with the Vibratip
Diagnostic tests for diabetic peripheral (Fig. 4.1E), a tuning fork, or automated devices
neuropathy such as Neurobiothesiometer, are not only reli-
ant on patients’ subjective response but are
While there is no single accurate definition also mainly used to identify the loss of pro-
of diabetic neuropathy, a simple definition for tective foot sensation and risk of ulceration.

FIGURE 4.1 Examples of ten different tests for diabetic peripheral neuropathy (DPN). (A) Physical examination and
Neuropathy Disability Score (NDS). (B) Nerve conduction studies being performed on the lower leg. (C) DPNcheck device
to test sural nerve conduction performed on the lower leg. (D) Monofilament screening test sights and procedure. (E)
Vibration perception threshold testing using Vibratip. (F) Medoc TSAII quantitative sensory testing device for thermal per-
ception threshold. (G) Sudoscan equipment. Hand and feet sensor plates with displaying test results. (H) Neuropad test
demonstrated original blue color. (I) A punch skin biopsy to collect samples needed for IENFD measurement. (J) An image
of the corneal subbasal nerves using corneal confocal microscopy and the CCM probe positioning during corneal scanning.

Diabetic Neuropathy
 Diagnostic tests for diabetic peripheral neuropathy 51
As such, these tests tend to diagnose DPN other causes. Clinicians commonly recognize
when it is already well established. Late diag- pain descriptors that patients with neuropathic
nosis hampers the potential benefits of intensi- pain use. The McGill pain questionnaire was
fied multifactorial intervention at an early the first questionnaire designed to offer a
stage of the disease, which could prevent the multidimensional assessment of pain, which
sequelae of DPN. Unfortunately, by the time included assessing severity or intensity, emo-
DPN is detected with the crude tests currently tional impact, and significance to the pain suf-
used, it is often very well established and con- ferer [65]. This questionnaire is one of the most
sequently impossible to reverse or halt the commonly used multidimensional pain scales
inexorable neuropathic process. Early diagno- globally. A short-form is available for screen-
sis and timely intervention are thus essential in ing, which has shown good agreement with
preventing the development of DPN. the original version [66].
Some of the most common tests and meth-
ods for diagnosis of DN have been summa-
rized in Table 4.2. Signs
The Neuropathy Disability Score (NDS) is a
commonly used clinical examination method
Symptoms that assesses neuropathy signs (Fig. 4.1A).
Thirty-five items are used for both sides, evalu-
Various clinical scoring systems are avail-
ating cranial nerve damage, muscle strength,
able for DPN screening, including symptom
sensation loss and reflex delay/loss [67].
scoring, sign scoring, or both (Table 4.3). These
However, some of the items have demon-
systems may enhance diagnostic accuracy
strated only a weak relationship with DPN,
through a composite score of different com-
and the full scoring system is too long to be
bined tests and are useful tools for aiding
used in clinical practice. Therefore a revised
diagnosis of DPN, along with quantitative
NDS has been created. This system is more
measures. Each questionnaire has a scoring
commonly used and tests for four main neu-
system that can diagnose, and in some, stratify
ropathy signs; ankle reflex, vibration, pinprick,
disease severity. Table 4.3 presents a summary
and temperature sensation at both sides of the
of the most commonly used questionnaires for
largest toes. A maximum score is 10, and usu-
assessing DPN.
ally, more than 6 is considered abnormal [68].
The Neurological Symptom Score (NSS) is
a 17-question, interview-based assessment of
sensory, motor, and autonomic function used Composite scoring systems
to screen for DPN [64]. Still, it is considered The reliance on symptoms or signs alone
too extensive to be used efficiently in clinical may lead to low diagnostic accuracy for the
practice. The diabetic neuropathy score (DNS) presence of DPN, and a combination of both
is an adaptation of the NSS that is a much allows a more thorough assessment. Several
quicker screening method, with only four ques- scoring systems assess both signs and symp-
tions and still offering moderate sensitivity toms of DPN to produce a composite score.
(79%) and specificity (78%), but with slightly The Toronto clinical neuropathy score (TCNS)
lower reliability for diagnosing DPN [22] when consists of three parts: symptom scores, reflex
using a diagnostic score of 1 or more. test scores and sensory test scores. The maxi-
Other symptom scoring systems focus only mum score is 19, and the test is able to stratify
on pain and differentiating neuropathic from patients into no DPN, mild DPN, moderate

Diabetic Neuropathy
52 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

TABLE 4.2 Diagnostic tests available for assessing diabetic peripheral neuropathy.
Nerve fibers
assessed Advantages Limitations

Symptoms Questionnaires Large (Aβ-fibers) Easy to administer. Used for Lack of sensitivity, accuracy and
and signs and small (Aδ- monitoring symptoms [21] reproducibility, subjective [22]
and C-fibers)
NDS Large (Aβ-fibers) Does not require specialist Not sensitive or reproducible, low
and small (Aδ- equipment, assesses large- and correlation with small-fiber
and C-fibers) small-fiber function [23] quantitative tests [24]

10-g Large (Aβ-fibers) Simple, quick, and inexpensive No standardization of methods.

monofilament [25]. Cannot detect early neuropathy [25].
Ipswich Touch Large (Aβ-fibers) Simple. Requires no specialist Can only detect advanced
Test equipment [26]. neuropathy. [25]
Can test at home.
QST (Thermal Large (Aβ-fibers) Measures small and large-fiber Unable to differentiate between
and vibration and small (Aδ- function. [21] peripheral and central abnormalities.
thresholds) and C-fibers) Good repeatability [27] [28]
High interoperator variability [29]
Large-fiber DPNCheck Large, sural Quick, easy to perform, good Relies on the accessibility of sural
tests nerve (Aβ-fibers) sensitivity (92% 95%) compared nerve [32]. Validation studies had
to NCS [30,31] small patient numbers [30,31]
NCS Large (Aβ-fibers) A sensitive measure of large nerve Doesn’t assess small fibers,
function [33], reproducible [34] uncomfortable [35], does not assess
early neuropathic changes
Small-fiber Skin biopsy Small (C-fibers) Gold standard for SNF, Invasive, risk of infection,
tests (IENFD) quantitative, good sensitivity, repeatability, requires trained
detects early nerve changes [36,37] personnel and special labs [21]
CCM Small (Aδ- and C- Noninvasive, good reproducibility Relatively expensive, requires
fibers) [38], rapid and objective [39,40] specialist equipment and personnel,
manual analysis is time-consuming
[21]
Autonomic Neuropad Small (C-fibers) Can be self-administered, Varied interpretation of the results
tests suitable for screening [31,41 44]
Noninvasive
Good sensitivity [31,41 46]
Sudoscan Small (C-fibers) Noninvasive, easy to perform Unclear if measuring sudomotor
function
Variable specificity (53% 92%)
[47 51]
QSART Small (C-fibers) Sensitive for SFN (82%) [52], gold Time-consuming, Requires specialist
standard for measuring equipment and trained personnel
sudomotor function [53]
Uncomfortable

CCM, corneal confocal microscopy; IENFD, intraepidermal nerve fiber density; NCS, nerve conduction studies; NDS, Neuropathy Disability
Score; QSART, quantitative sudomotor axon reflex test; QST, quantitative sensory testing; SFN, small-fiber neuropathy.

Diabetic Neuropathy
 Diagnostic tests for diabetic peripheral neuropathy 53
TABLE 4.3 Summary of questionnaires available for assessing diabetic peripheral neuropathy.
Questionnaire Assessed Type of administration Scoring

Symptoms NSP Symptoms of neuropathy Clinician administered 34 categories (women)

36 categories (men)

DNS Symptoms of diabetic Clinician administered 4 for symptoms (Total 4)

peripheral neuropathy
NSS Symptoms of neuropathy Clinician administered 8 for muscle weakness
5 for sensory disturbances
4 for autonomic symptoms
(Total 17)
NPQ Symptoms of neuropathic Completed by the patient Total of 12
pain

NPSI Symptoms of neuropathic Completed by patient 10 descriptors, 2 duration

pain (Total 12)

McGill Pain Multidimensional Clinician administered Subclass 1—sensory

Questionnaire symptoms of pain Subclass 2—affective
Subclass 3—evaluative
Subclass 4—miscellaneous
(Total 78)

CNE Signs of peripheral Clinician administered 21 for sensory testing

neuropathy 8 for muscle strength
4 for ankle reflex (Total 33)
Signs NDS Signs of peripheral Clinician administered 2 for vibration sensation
neuropathy 2 for temperature sensation
2 for pinprick
4 for ankle reflex (Total 10)

DNE Signs of peripheral Clinician administered 4 for muscle strength

neuropathy 2 for reflex responses
10 for sensory testing
(Total 16)
NIS-LL Signs of neuropathy in the Clinician administered 64 for muscle strength
lower limbs 16 for sensory testing
8 for reflex responses
(Total 88)
MNDS Signs of peripheral Clinician administered 12 for sensory tests
neuropathy 18 for muscle strength
16 for reflex testing
(Total 46)
UENS Signs of peripheral Clinician administered 11 for each side (Total 22)
neuropathy

Symptoms DN4 Symptoms and signs of Clinician administered 7 for symptoms

and signs neuropathic pain 3 for signs (Total 10)

(Continued)

Diabetic Neuropathy
54 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

TABLE 4.3 (Continued)

Questionnaire Assessed Type of administration Scoring

LANSS Symptoms and signs of Clinician administered 5 for symptoms

neuropathic pain 2 for signs (Total 7)

TCNS Signs and symptoms of Clinician administered 6 for symptoms

peripheral neuropathy 5 for sensory tests
8 for reflex tests (Total 19)
MNSI Signs and symptoms of Symptoms by patient Foot 15 for symptoms
peripheral neuropathy examination by a clinician 8 for foot examination
(Total 23)

CNE, clinical neurological examination; DN4, Douleur Neuropathique en 4; DNE, diabetic neuropathy examination; DNS, diabetic
neuropathy symptom; LANSS, Leeds Assessment of Neuropathic Symptoms and Signs; MNDS, Michigan Neuropathy Disability Score;
MNSI, Michigan Neuropathy Screening Instrument; NSP, neuropathy symptoms profile; NPQ, neuropathic pain questionnaire;
NSS, neuropathy symptom score; NPSI, neuropathic pain symptom inventory; NDS, Neuropathy Disability Score; NIS-LL, neuropathy
impairment score in the lower limbs; TCNS, Toronto Clinical Neuropathy Score; UENS, Utah Early Neuropathy Scale [24,54 63].

DPN, and severe DPN depending on the over- the patient has signs and symptoms other than
all score [69]. Testing has proven validity and those rated by the scoring test.
reliability for diagnosing and staging DPN
compared to electrophysiology measures [69].
The Michigan neuropathy screening instru- Large-fiber tests
ment (MNSI) is another commonly used compos-
ite scoring system that includes a questionnaire Nerve conduction studies
and a foot examination [54]. Neuropathy can be The current “gold standard” for clinical
defined as seven or more positive responses to diagnosis of DPN is through NCS by a trained
this symptoms section alone [54]. The foot exam- neurophysiologist (Fig. 4.1B). In 2010, the
ination is more frequently used and encom- Toronto Consensus, by an expert panel, recom-
passes foot appearance (including ulcers), ankle mended that one abnormal finding as part of
reflex, and the 128-Hz tuning fork test [54]. One NCS, combined with a symptom or sign of
study [70] found a range of sensitivity (35% neuropathy, should be used to confirm DPN
79%) and specificity (65% 94%) in comparison [18]. NCS has also demonstrated an ability to
to NCS, depending on the cut-off value used for predict future DPN [71].
abnormality in MNSI. The higher specificity For reliable NCS results, close attention
values indicate a potential high diagnostic must be paid to factors such as filter setting,
impact for MNSI scoring; however, the lower limb temperature, and recording location, as
sensitivity range indicates that milder DPN outcomes can be vulnerable to variations.
cases are likely not to get picked up. Trials have demonstrated that NCS consistently
Scoring of symptoms and signs is conve- demonstrate excellent intraobserver agreement
nient and easy to perform as a method of [35,72]; however, a poor interobserver agree-
screening for DPN. These tests are easily inter- ment between expert clinical neurophysiolo-
preted, making them a useful tool in support- gists is common [35] when no standardized,
ing decisions on which patients should be specific technique is followed. One study [72]
referred for specialist assessment. Quantitative, assessed the results of four neurophysiologists
objective measures should be considered when from four different centers. Specific assessment

Diabetic Neuropathy
 Diagnostic tests for diabetic peripheral neuropathy 55
methods were provided in a specially prepared DPNCheck
syllabus, and a training session was provided
To overcome some of the shortcomings of
beforehand. The outcome was a significant
standard NCS testing, a novel point-of-care
improvement in interobserver agreement with
nerve conduction device (POCD), DPNCheck
a standardized approach, and although not
(Neurometrix Inc., Waltham, MA) has been
entirely eliminated, levels of disagreement
developed with the potential to serve as an
were consequently considered clinically insig-
acceptable proxy to standard NCS which are
nificant for medical practice [72].
time-consuming, expensive, and often require
Conversely, when considering the use of
patients to be seen in specialist clinics (Fig.
NCS in therapeutic clinical trials, even small
4.1C). This test for sural nerve conduction
interobserver variability may be significant
velocity and amplitude is much quicker (3 min-
enough to impact results through impacting
utes) to perform than conventional electrodiag-
the statistical power of a study and thus the
nostic testing. It has been validated in type 1
trial’s outcomes. This may partially explain
and 2 diabetes populations through comparison
why previous clinical trials have used NCS as
with the NDS [32] and standard NCS [30,31].
a primary outcome to detect treatment efficacy
These studies have reported a high sensitivity
and have reported failed outcomes [73 75].
of 92% 95% for detecting abnormalities
Evidence supports the use of a single observer
(Table 4.4). However, these studies’ cohorts
to repeat electrophysiological tests on each
have been small, with two of the three studies
patient in these trials.
assessing very low numbers of patients with
Furthermore, standard NCS testing is not
T1DM [30,31,85]. Furthermore, the DPNCheck
easily applicable as a screening tool for DPN
device is dependent on the presence of an
since it is time-consuming, requires a specialist
accessible sural nerve which can be anatomi-
operator and can be uncomfortable for the
cally absent in up to 9% of healthy subjects [32].
patient [35]. Electrodiagnostic studies have also
been identified as one of the largest drivers of Small-fiber tests
health care costs related to neuropathy evalua-
tion [76]. Results are often found to be normal Punch skin biopsy
in patients with diabetes who have early or The evidence strongly suggests that in DPN,
small-fiber predominant neuropathy. damage to small fibers precedes damage to

TABLE 4.4 Summary studies for validity for four potential screening tests for diabetic peripheral neuropathy.
Test Fibers assessed Validated against Sensitivity Specificity

DPNCheck [30,31] Large (Aβ-fibers) NCS 92% 95% 82% 89%

Neuropad [41,43 46] Small (C-fibers) NCS, NDS, VPT 70% 97.8% 50% 67%
Sudoscan [47 51] Small (C-fibers) NCS, clinical examination, 70% 87.5% 53% 92%
UENS, VPT, NSS
CCM [39,77 84] Small (Aδ- and C-fibers) NCS, clinical examination, CASS 59% 86% (CNFL) 61% 84% (CNFL)
65% 82% (CNFD) 41% 79% (CNFD)
17% 100% (CNBD) 45% 96% (CNBD)

CASS, composite autonomic scoring scale; CCM, corneal confocal microscopy; CNBD, corneal nerve branch density; CNFD, corneal nerve
fiber density; CNFL, corneal nerve fiber length; NCS, nerve conduction studies; NDS, Neuropathy Disability Score; NSS, Neuropathy
Symptoms Score; UEN, Utah Early Neuropathy Scale; VPT, vibration perception threshold.

Diabetic Neuropathy
56 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

large fibers [86,87] and punch skin biopsy is identify DPN in a diabetic cohort. Other stud-
currently considered the gold-standard single ies have found a decrease in IENFD correlating
test for diagnosing small-fiber neuropathy [88]. with the progression of neuropathy and dura-
A measure of intraepidermal nerve fiber tion of diabetes [95,96], with reports that
density (IENFD) can be quantified from these IENFD may also be lower in patients with
biopsies, which is a method of documenting painful DPN compared to painless DPN [97].
the density of terminal branches of peripheral A 5-year follow-up study investigating the
nerves within the epidermis (no./mm2). The progression of DPN in T1DM and T2DM
European Federation of Neurological Societies reported a significant reduction of IENFD
has published guidelines for its use in diagnos- in T2DM patients, with IENFD measurement
ing peripheral neuropathies [89] (Fig. 4.1I). being the single most abnormal parameter [98].
Two immunostaining methods have become Overall, the reduction in IENFD was not signif-
the most widely used in IENFD measurement: icant in T1DM subjects. However, the lower
indirect immunofluorescence (IF) and bright- number of patients in the T1DM group may
field immunohistochemistry (BFI). Although IF explain this finding, as it would make it more
is considered a slightly more sensitive tech- challenging to prove statistically significant
nique due to the higher signal/noise ratio [90], changes [98].
the two methods have excellent correlation The main issue with IENFD measurements
[91], and both can comparably detect SFN [90]. as a biomarker for small-fiber neuropathy is
At present, age-related normative values exist that it is an invasive procedure. Obtaining a
only for BFI, published by a multinational biopsy can cause side effects such as a mild
group of eight centers [92]. infection due to improper wound management
For both IF and BFI techniques, IENFs are or, less commonly, excessive bleeding. Even
typically counted directly through an epifluor- though reported side effects are rare (1.9/1000)
escence microscope’s oculars by focusing [92], the nature of this technique limits its prac-
through the optical planes [90]. For IF only, the tical use, particularly when a repeat biopsy
more precise but time-consuming technique is required in longitudinal studies or clinical
confocal microscopy (CM) can analyze optical intervention trials.
sections of three-dimensional images using From a screening perspective, although
computer software [90]. The two techniques IENFD measurement from a lower-limb skin
have shown excellent correlation [90], and the biopsy is considered the gold standard for the
latter is usually used when the more complex, diagnosis of small-fiber neuropathy, it is inva-
second-level analysis is needed. sive and therefore not suitable for routine
IENFD measurements have been shown to screening [88,99].
detect small-fiber neuropathy with depletion of
IENFD detected in patients with normal NCS Quantitative sudomotor axon reflex test
and no clinical signs or symptoms of neuro- The assessment of sudomotor nerve (sweat)
pathy [36,37]. A recent study reported low sen- function has also been used to assess small
sitivity of just 61% when using a cut-off of 4.5 autonomic C-fibers, as anhidrosis can be char-
fibers/mm IENFD to diagnose clinical DPN in acteristic of the presence of peripheral auto-
T1DM patients [77]. Earlier studies have pub- nomic neuropathy.
lished significantly higher values for sensitivity The reference standard for measuring sudo-
(80%) [93] and specificity (95%) [94]; however, motor function is the quantitative sudomotor
these studies were comparing healthy controls axon reflex test (QSART). This test uses local
to DPN patients rather than the test’s ability to sweat production, measured as a change of

Diabetic Neuropathy
 Diagnostic tests for diabetic peripheral neuropathy 57
relative humidity over time, during and after large-fiber diagnostic tests, the strong correla-
skin activation. Special software is used to digi- tion between Neuropad and NDS [41,46],
talize, plot, and analyses the temporal resolu- IENFD [42], CCM [102], Sudoscan [103], and
tion, latency, magnitude, and duration of the measures of sweat gland dysfunction [104]
sudomotor response [100]. However, due to have been reported. It has also been identified
highly technical demands and relative discom- as a useful tool for staging the severity of
fort of the examination, QSART remains mostly neuropathy in patients with type 2 diabetes,
limited to research centers and is not consid- demonstrating excellent agreement with the
ered a potential screening tool for DPN [53]. Michigan classification system [105]. Another
significant advantage of Neuropad is its high
Neuropad NPV, making it ideal for serving as a screening
Neuropad is a patented 10-minute screening test primarily to exclude DPN [45,106,107].
test for the early detection of diabetic foot syn- However, studies are not consistent in terms
drome and can be used as a triage test [101]. It of the position of the Neuropad on the foot
is a unique, noninvasive, painless, and simple and the NDS cut-off value chosen to indicate
diagnostic screening test employing a chemical clinical DPN presence. Furthermore, some
reaction to minute quantities of sweat as a bio- studies graded the Neuropad color change as a
marker for much earlier signs of DPN. percentage [43] or score out of 1 [42], whereas
The test has been created to assess the sweat others simply classified the results as normal
function (small autonomic C-fibers) in the or abnormal [41,44]. Standardization of elapsed
feet of patients with suspected neuropathy. An time before test result analysis is also necessary
adhesive pad containing cobalt salts is stuck as extending the observation period to 15 min-
onto the foot’s plantar aspect and changes utes may provide greater diagnostic usefulness
color from blue to pink within 10 minutes if [108]. This highlights a need for software
the sudomotor function is normal [44]. If there development that can consistently grade each
is a decreased function, the pad remains blue test’s color change over time to enable continu-
or turns patchy in color. There is a strong asso- ous and more accurate monitoring of sudomo-
ciation between skin dryness, sudomotor dys- tor dysfunction.
function, and diabetic foot ulcer and the To address these issues and increase both
function of Neuropad. An abnormal Neuropad the sensitivity and specificity of Neuropad
response is associated with sympathetic dys- screening and create a continuous output, a
function and clinical neuropathy (Fig. 4.1H). smartphone software app and internet-based
This test’s main advantage is that patients image-processing system has been developed.
can self-administer at home, reducing clinical Neurometrics-Diab is a digital therapeutics
contact time and aiming to reinforce abnormal (DTx) smartphone app that uses the Neuropad
results visually. Instructions have been con- as a biomarker to produce a continuous record
firmed as clear for patients to follow, and the of a person’s neuropathy see if it is improving,
test is easy to use for most patients [41]. is stable or is worsening with trend-lines help-
However, due to older age, visual and kinetic ing to predict outcomes. Using a smartphone
problems, a fifth of patients still needed help camera, patients can take a photo of their test
when self-testing. result, which is then automatically sent to a
It has been reported as having good-to- web server where the photo is run through a
excellent (70% 97.8%) [41 46] sensitivity for proprietary image-processing algorithm result-
DPN detection (Table 4.4). When comparing ing in a recorded percentage score. Over time
Neuropad to a range of different small- and a trend can be calculated. The DTx app is

Diabetic Neuropathy
58 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

currently at the advanced prototype stage. however, these studies’ findings emphasize the
Versions for other medical conditions are importance of profile matching different sub-
under development. ject groups for a weight that did not occur in
some validation studies [48,51].
Sudoscan Validation studies have reported consis-
Sudoscan (Impeto Medical) is another quick, tently good values for sensitivity (70% 87.5%)
simple, and noninvasive test that aims to assess (Table 4.4) when using foot ESC results to
sudomotor function using “reverse iontophore- screen for DPN [47 49,51,109]. However, there
sis” [109,110] to measure electrochemical skin are inconsistencies in the ESC cut-off values
conductance (ESC) of sweat in the hands and used for identifying sudomotor dysfunction,
feet. Compared to age-corrected standard data, ranging from 52 uS [47] to 77 uS [49]. This vari-
a reduced ESC result may indicate degenera- ation and inconsistencies in the neuropathy
tion of small C-fibers that innervate the sweat tests being used as a reference standard are the
glands and, therefore lead to reduced sweat likely cause of the extensive range in reported
gland function [48] (Fig. 4.1G). specificity of between 53% and 92% [47 49,51].
The ESC measurements from the feet are It highlights the need for standardization of the
considered more sensitive for the detection of classification criteria used. Patient cohorts also
DPN than the hands [49], with less variation in differed in their severity of DPN, with partici-
results [111]. This is likely due to a fluctuation pants in one study [51] having significantly
in the hands’ contact on the electrodes. In con- more advanced DPN than those in the study
trast, the feet are aided by gravity to maintain by Smith and colleagues [48]. Therefore the test
constant pressure on the electrodes throughout performed better in the former.
the test. Lower ESC at the feet was also signifi- Overall, Sudoscan appears to be a promising
cantly associated with increasing symptoms in DPN screening test that is noninvasive, easy to
a large cohort of patients with T2DM [112]. perform and eliminates the subjective compo-
Reference values in healthy subjects are nent of clinician error, demonstrating a good
available from a global collaborative analysis correlation with IENFD [114]. However, there
comparing different ethnic groups, age, and is some doubt as the current evidence does not
gender [113]. This study noted a significantly strongly support ESC to distinguish between
lower hands and feet ESC for African patients with DPN and control individuals
American, Indian, and Chinese populations [115]. Therefore longitudinal and more exten-
than the Caucasian population, highlighting sive cohort validation studies are needed,
the need to match ethnicity groups in ESC along with standardization of diagnostic crite-
studies. The same study also observed no sig- ria before Sudoscan can be used as a screening
nificant difference between women and men at tool for small-fiber neuropathy.
the hands or feet and a weak decline in ESC Progress has been made in developing
with increased age. point-of-care devices (POCDs), which may
ESC measurements may also be associated diagnose DPN early before clinical signs are
with subjects’ weight [114], perhaps due to a apparent. Neuropad, DPNCheck, and Sudoscan
weight-dependent change in sensitivity of the are newer screening tests that have demon-
stainless-steel electrodes or sweat gland den- strated potential for early detection; however,
sity when the subject is in the standing posi- validation studies, thus far, have reported a
tion. This could also be due to the correlation range of sensitivities and specificities depend-
between higher weight and larger feet only ing on cohort and test used for comparison
[114]. These hypotheses are yet to be assessed; (Table 4.3).

Diabetic Neuropathy
 Diagnostic tests for diabetic peripheral neuropathy 59

Quantitative sensory testing findings. Malmström et al. [126] failed to detect

differences between obese and other groups for
Quantitative sensory testing (QST) has cold and warm thresholds at the suprailiac site
become a common method for evaluating [126]. In contrast, Price and colleagues [127]
small nerve fiber function using thermal found that obese participants had significantly
threshold and thermal pain measurements and higher cold and warm detection thresholds
large-fiber function using vibration thresholds than normal BMI participants on the abdomen.
[29]. The most common commercial system is Two psychophysical algorithms can be used
the Medoc TSA-II NeuroSensory Analyser to determine thermal thresholds. These are the
(Medoc Advanced Medical Systems, Israel) method of limits and the method of level
which is used to determine thermal thresholds (described in detail elsewhere) [27,128], with
(Fig. 4.1F). In recent years a cheaper, more por- the method of limits used more commonly
table device has been designed, NerveCheck due to it being less time-consuming [123].
(Phi Med Europe S.L., Barcelona, Spain), which Measurements determined using limits have
has shown good reproducibility (ICC values 5 been reported as significantly higher than
0.79, 0.71, and 0.86 for vibration, warm and those measured by level, irrespective of test
cold sensation, respectively) and comparable location [29]. However, the two methods corre-
diagnostic accuracy (86%, 72%, and 79% for late well with each other [29], and the 2013
vibration, warm and cold sensation testing, consensus concluded that both were reliable
respectively) to established QST equipment for [123]. The major difference between these two
the diagnosis of DPN [116]. methods is the effect of reaction time. For the
Cold thresholds can be used to evaluate method of limits, a patient has a longer reac-
myelinated Aδ-fiber function, whereas warm tion time due to age or height (causing a more
thresholds are used to assess the function of extended sensory pathway), which may erro-
unmyelinated C-fibers. Published normative neously give a higher threshold.
data sets are available for heat threshold detec- Both warm and cold thresholds can be
tion [117 122], and recommendations for affected in patients with DPN, irrespective
conducting QST in both clinical practice and of how long the course of diabetes is, but the
research have previously been published by frequency of abnormal warm thresholds is
The International Association for the Study of significantly higher [118]. A study found that
Pain (NeuPSIG) [123]. cold detection thresholds significantly reduced
QST has been found to have reasonable in DM patients with no evidence of preclinical,
repeatability [27]; however, interoperator and subclinical, and clinical DPN, respectively
interpatient variability depend on several fac- [129]. A longitudinal study also found a signifi-
tors. Training of both examiner and patient, the cant positive correlation between deterioration
methodology of assessment, baseline skin tem- of cold detection thresholds and pain intensity
perature, stimulus characteristics, location and in painful DN, with warm detection thresholds
number of stimuli sites and duration of intervals also correlating at nonsignificant value [130].
between tests have all affected QST measure- One major issue with the use of QST is
ments [29]. Using standardized methodology that it cannot differentiate between peripheral
with extensive training has significantly reduced and central temperature perception causes.
interobserver variability [124,125]. However, this It involves sensory receptors, spinal cord
may be too time-consuming to be implemented. pathways and termination sites in the thala-
When it comes to the effects of body fat on mus. This means that if there is poor concen-
thermal detection thresholds, there are conflicting tration, a language barrier or cognitive defect,

Diabetic Neuropathy
60 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

subjects’ results may affect their subjective These properties allow CCM to acquire high-
nature [28]. quality images of the corneal C-fibers in the
subbasal nerve plexus. Considering the known
relationship between damage to these fibers
Corneal nerves as a biomarker for
and DPN, the potential for their use as a surro-
diabetic peripheral neuropathy gate biomarker for DPN has been identified.
Anatomically and developmentally, the eye When analyzing the subbasal nerve plexus,
can be considered an extension of the central most studies report results from four morpho-
nervous system (CNS). The cornea is the most logical parameters: Corneal nerve fiber density
densely innervated tissue in the body. It is (CNFD), which is the total number of main
richly supplied by a large number of sensory nerve fibers per mm2, corneal nerve fiber
nerve fibers and a lesser number of autonomic length (CNFL), which is the sum of the length
fibers [131]. The cornea possesses small unmy- of all nerve fibers and branches (mm/mm2),
elinated C-fibers and myelinated Aδ-fibers for tortuosity coefficient (TC) which is a unitless
sensory innervation. These are derived from measurement that uses deviation from a
the trigeminal nerve’s ophthalmic division and straight line to measure the tortuosity of the
enter the corneal stroma at its periphery, in a main nerve fibers independent of their orienta-
radial fashion parallel to the corneal surface. As tion, and corneal nerve branch density (CNBD)
the fibers run forward toward the cornea center, which is defined as the number of branches
they lose their myelin sheath; a necessary step emanating from all main nerve fibers. There
to maintain corneal transparency [131]. is, however, a discrepancy in how this can be
Corneal C-fibers form a delicate three- quantified between studies with the estab-
dimensional network known as the “subbasal lished protocol for these parameters described
nerve plexus” [132], which is located beneath elsewhere [39].
the basal layer of the corneal epithelium. Of these four parameters, CNFL has been
Mapping of the cornea [133,134] has shown the most frequently used parameter for DPN,
that this plexus forms a spiral or “whorl-like” with one study reporting superior reliability
pattern. The spiral center, often called the vor- than other parameters [137]. Some studies have
tex, is located approximately 2 3 mm inferior assessed the diagnostic performance of CCM
and nasal to humans’ corneal apex. Due to this for DPN and reported the results for CNFL
arrangement, subbasal nerves in the superior only [78,82]. Hertz et al. [137] reported that
and human apical cornea are oriented verti- CNFL produced the highest intra- and interob-
cally. In contrast, subbasal nerves in other cor- server reproducibility (ICC of 0.72 and 0.73,
neal regions may be orientated horizontally or respectively), with TC demonstrating the low-
obliquely, consistent with their locations within est (0.23 and 0.29, respectively).
the whorl-like plexus [135]. Two other parameters that have been
Corneal confocal microscopy (CCM) is a reported in research studies are nerve reflectiv-
noninvasive, in vivo ophthalmic imaging tech- ity [138] and nerve fiber beading (no./100 μm)
nique that allows a detailed examination of [139,140]. Nerve fiber reflectivity is usually
the cornea, at high magnification, on a cellular assessed using grades as first outlined by
level (Fig. 4.1J) [136]. By capturing multiple Oliveira-Soto and Efron [138], whereby classifi-
two-dimensional images at different depths, cation can be split into four grades according
CCM imaging can delineate the corneal to a comparison with reference images. The
layers [136], providing superior magnification number of beadings is defined as the number
compared to standard slit-lamp biomicroscopy. of beadings in a length of 100 um of subbasal

Diabetic Neuropathy
 Diagnostic tests for diabetic peripheral neuropathy 61
nerves within a frame [141]. Both parameters Cross-sectional studies have shown an
have demonstrated changes in dry eye condi- increase in the densities of LCs in the central
tions, where patients with Sjogren’s syndrome cornea related to conditions such as dry eye
have demonstrated significantly higher bead- with and without contact lens wear [149,151],
ing than dry eye patients of other primary bacterial keratitis [152], thyroid eye disease
causes [142]. However, both measures require [153], and diabetes [154,155].
subjective judgment. Quantifying the beading
is challenging, and it does require special soft-
ware [141]. Corneal confocal microscopy for the
More recently, newer corneal parameters detection of diabetic peripheral
have been investigated. These include inferior neuropathy
whorl length (IWL) [143], defined as the length
In the early 2000s, a novel study by
of the nerves at the inferior whorl of the super-
Rosenberg and colleagues [156] reported the
ficial nerve plexus, nerve fiber width (CNFW)
correlation between increasing severity of DPN,
[144] and nerve fiber area (CNFA) [145]. These
corneal sensitivity and progressive loss of cor-
new measures have previously shown signifi-
neal subbasal nerve fibers [156]. This was closely
cant differences between the nonneuropathic
followed by a similar small study published in
and clinically neuropathic groups in DM [146]
2003 [157] which found that CCM was able
with CNFW and CNFA, demonstrating 74%
to detect abnormalities in the corneal nerves
and 66% sensitivity specificity equal error rate
of 18 patients with diabetes deemed to have
point, respectively when identifying nonneuro-
only mild neuropathy using conventional tests.
pathic patients compared to control subjects
Similarly, Midena and colleagues [158] reported
[146]. This indicates that these new measures
a significant decrease in corneal nerve fiber and
may have the capacity to identify individuals
branch number, along with decreased beading
with early neuropathy; however, research into
in patients with diabetes. It should be noted that
these new parameters is currently limited.
these three studies used a light corneal confocal
Another type of cell found in the subbasal
microscope, which is the first commercially
layer that has been of interest in DPN research
available generation of the confocal imaging
are dendritic cells (DCs). These antigen-
device with inferior image quality in compari-
presenting cells of the cornea are of paramount
son to the methods now commonly used.
importance. They play a critical role in activat-
Since then, the use of CCM for rapid, nonin-
ing other immune systems in the ocular surface,
vasive clinical assessment of corneal nerves has
influencing both suppression and induction of
grown substantially, especially in recent years.
inflammation [147,148].
It has proven to be particularly useful as a
Langerhans cells (LCs) are usually up to
diagnostic marker for detecting diabetic neu-
15 μm in diameter and can be seen in various
ropathy and a range of other peripheral neu-
forms in the cornea [149]. In their immature
ropathies [39,77 83,159 164]. Some of them
form, these cells have small dendritic pro-
are reviewed in this paper.
cesses or lack dendrites completely and are
mainly located in the peripheral cornea’s epi-
thelium [150]. In pathological states, LCs
Diagnostic performance for clinical
mature, form interlocking dendritic pro-
cesses, which may comprise a net-like struc-
diabetic peripheral neuropathy
ture, and migrate from the periphery into the Several cross-sectional studies have evalu-
central cornea [150]. ated the ability of CCM to diagnose clinical

Diabetic Neuropathy
62 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

levels of DPN in comparison to a range of CNFL was the most commonly reported
other diagnostic tests (Table 4.5). It must be nerve parameter for these studies, with all nine
noted that most of these studies assessed assessing its diagnostic ability and finding
patients with T1DM only, meaning there is lim- significant differences between patients with
ited published data available for the diagnostic and without DPN. A range of sensitivity values
sensitivity and specificity values when asses- between 59% and 86% were found and a speci-
sing patients with T2DM. ficity range of 61% 84%, depending on the
These studies used a cut-off point for the cut-off value used for diagnosis. The earliest
reference neuropathy test/combination of tests of these studies [39] examined patients using a
to determine whether a patient had a DPN. Tomey confoscan CCM. It is well known that
However, the reference test and cut-off points these images are of more inferior quality, mak-
varied between studies meaning there were ing it more challenging to identify nerve fibers
no universal diagnostic reference criteria. Some during analysis. This is likely the explanation
studies validated CCM against a single test for the significantly lower diagnostic threshold
of NCS [77,78] or NDS [39]. In contrast, other value reported in this study compared to the
studies used a combination of the two [79] or others presented (Table 4.5).
NCS and clinical examination [81,83]. A combi- For CNFD, six of the cross-sectional studies
nation of diagnostic tests will likely increase (Table 4.5) reported a significant reduction in
the efficiency of detecting DPN compared to DM patients with DPN compared to both DM
one test used alone. This is significant as some patients without DPN and healthy controls
studies compare CCM to one single test, which [39,77,79,81,83,84]. These studies reported sen-
is not the gold standard in the case of NDS. sitivity and specificity ranges as 65% 82% and
Additionally, NCS only measures large-fiber 41% 79%, respectively. A significantly higher
function, which is affected later than small cut-off point of 39.2 CNFD no./mm2 was
nerve fibers in DPN. One study [166], demon- defined in T2DM patients in the consortium
strated that diagnostic ability of CNFL measure- study, resulting in an increased sensitivity
ment in DM patients is significantly worse if value to 69% [81]. This may explain why its
using clinical signs and symptoms as a refer- specificity is the lowest value of only 41%, as a
ence standard in comparison to electrophysiol- higher cut-off value may create more false-
ogy, plus one sign/symptom as per the Toronto positive results. It is notable that based on their
consensus guidelines, which highlights the cohort. Scarr et al. [83] defined the lowest
importance of a standardized diagnostic refer- thresholds for diagnosis for both CNFD and
ence [166]. CNFL out of the studies using the Heidelberg
To explore which of the many measurements retinal tomograph (HRT) (III) RCM-CCM [83].
derived from CCM could best distinguish This is likely due to their significantly older-
patients with and without clinical DPN, as part aged cohort compared to the other cross-
of each study, the same patients were examined sectional studies as CNFD and CNFL have
using CCM, and all nerve parameters were been shown to reduce with age [165].
derived. For each nerve parameters tested, ROC For corneal nerve branch density (CNBD),
curves were plotted to determine a CCM cut-off six cross-sectional studies [39,77,79,81,83,84]
point used to distinguish between patients with reported a significant reduction in DM with
and without DPN in the diabetic cohort only. A DPN than without DPN. For diagnostic value,
range of cut-off points were studied to identify the sensitivity (17% 100%) [39,79] and speci-
the best sensitivity/specificity value for diag- ficity (45% 96%) [39,79] values were signi-
nosing DPN for each nerve parameter. ficantly more varied, suggesting that this

Diabetic Neuropathy
TABLE 4.5 Summary of studies assessing the clinical utility of corneal nerve parameters for the diagnosis of clinical levels of diabetic neuropathy compared to
chosen reference standards.
CNFL
Disease threshold CNFD CNBD
Number Type of Diabetes duration Type of Validated Sensitivity Specificity (mm/ threshold threshold CNFT
Study subjects CCM device type Age (years) (years) neuropathy against (%) (%) mm2 ) (no./mm2 ) (no./mm2 ) (TC)

Cross- Tavakoli 118 (101 Tomey 1,2 55 6 4.8 (HC) 10.7 6 1.82 DSPN NDS 64 79 3.39 27.81 13.89
sectional et al. [39,102] DM, 17 ConfoScan 55 6 1.9(DPN-) (DPN-) (CNFL) (CNFL)
HC)
P4 58 6 2.1(mild 15.5 6 2.08 82 52
DPN) (mild (CNFD) (CNFD)
59 6 2.5(mod DPN) 91 45
DPN) 18.6 6 3.06 (CNBD) (CNBD)
61 6 2.05(sev (mod
DPN) DPN)
19.3 6 2.85
(sev DPN)
Ahmed et al. 153 (89 HRT (II) 1 38.9 6 17.6 17.6 6 DSPN NCS, 85 84 14
[78] DM, 63 (HC) 14.0(DPN- clinical
HC) examination
34.9 6 14.8 )
(DPN-) 31.4 6
50.0 6 14.3 13.5
(DPN 1 ) (DPN 1 )
Tavakoli 52 (34 HRT (III) 1,2 42 6 0 (DAN-) 12 6 3 DAN Composite 86 78 4.8 23.3 19.5
et al. [84,165] DM, 18 44 6 3(DAN 1 ) (DAN-) autonomic (CNFL) (CNFL)
HC) scoring
26 6 2 scale
86 79
(DAN 1 ) (CASS) (CNFD) (CNFD)
100 56
(CNBD) (CNBD)
Chen et al. 89 (63 HRT (III) 1 44 6 15 (HC) 23 6 16 DSPN NCS, DNS/ 59 74 16.5 24 15
[79] DM, 26 44 6 13 (DPN-) (DPN-) NDS (CNFL) (CNFL)
HC)
59 6 11 (DPN 1 ) 39 6 14 82 71
(DPN 1 ) (CNFD) (CNFD)
17 96
(CNBD) (CNBD)
Alam et al. 88 (61 HRT (III) 1 41 6 114.9(HC) 17.2 6 12.0 DSPN NCS, 74 61 16.8 25 36.5
[77] DM, 27 38.8 6 12.5 (DPN-) clinical (CNFL) (CNFL)
HC) examination
(DPN-) 37.2 6 13.1 77 79
53.3 6 11.9 (DPN 1 ) (CNFD) (CNFD)
(DPN 1 )

(Continued)
TABLE 4.5 (Continued)
CNFL
Disease threshold CNFD CNBD
Number Type of Diabetes duration Type of Validated Sensitivity Specificity (mm/ threshold threshold CNFT
Study subjects CCM device type Age (years) (years) neuropathy against (%) (%) mm2 ) (no./mm2 ) (no./mm2 ) (TC)

67 58
(CNBD) (CNBD)
Scarr et al. 137 HRT(III) 1 64 6 8 (HCs) 52 58 DSPN NCS, 73 75 13.7 18.8 15.6
[83] (67DM, 65 6 7 (T1DM) (median 54) clinical (CNFL) (CNFL)
69HC) examination
76 75
(CNFD) (CNFD)
44 75
(CNBD) (CNBD)
Perkins et al. 998 (516 Tomey 1 42 6 19 21 6 15 DSPN NCS, 71 67 16.4 28 37.6
[81] T1DM, Confoscan clinical (CNFL) (CNFL)
(Consortium) 482 examination
P4, HRT 65 75
T2DM)
(II), HRT (CNFD) (CNFD)
(III) 67 72
(CNBD) (CNBD)
2 62 6 10 12 6 9 65 69 16.3 39.2 44.8
(CNFL) (CNFL)
69 41
(CNFD) (CNFD)
69 63
(CNBD) (CNBD)
Longitudinal Pritchard 90 (T1 HRT (III) 1 42 6 16(DPN-) 15 6 12 DSPN NCS, 63 74 14.1
et al. [82] DM) 51 6 14(DPN 1 ) (DPN-) DNSS/NDS

29 6 16
(DPN 1 )
Lovblom 65 (T1 HRT (III) 1 34 6 15 (DPN-) 17 6 12 DSPN NCS, TCNS 82 69 14.9 41.7 36.1 15.9
et al. [80] DM) 38 6 16 (DPN 1 ) (DPN-) (CNFL) (CNFL)
21 6 9 55 59
(DPN 1 ) (CNFD) (CNFD)
82 50
(CNBD) (CNBD)
73 72
(CNFT) (CNFT)

CNBD, corneal nerve branch density; CNFD, corneal nerve fiber density; CNFL, corneal nerve fiber length; DM, diabetes mellitus; DPN, diabetic peripheral neuropathy; DSPN, distal symmetrical polyneuropathy; NDS, Neuropathy Disability
Score; TCNS, Toronto Clinical Neuropathy Score.
The studies presented are all published studies. Data presented as standard units or mean 6 standard deviation.
 Diagnostic tests for diabetic peripheral neuropathy 65
parameter has shown the least promise for than the T2DM group without neuropathy.
DPN diagnosis until now. Another limitation of two of these studies
There are several strengths to each of the [78,83] was that only one image per eye was
cross-sectional studies. Three used profile- used for analysis. One criterion for choosing
matched healthy controls [78,79,83], meaning that this image in the Ahmed et al. [78] study was
differences in measurements between the two the most nervous frame. Using this method to
groups due to age should have been accounted choose one image per eye instead of calculating
for, giving a better representation of changes that an average of three images or more may be less
have occurred due to DPN. Ahmed et al. [78] time consuming for analysis; however, it is
also looked at the option of combining two cor- likely to give the false elevation of measure-
neal nerve parameters for the identification of ments per patient instead of representing an
neuropathy. Automated software for analysis accurate average.
CCM images would be required if CCM were to Another significant issue with these studies
be introduced in large-scale screening [79,83], is that most of them use the Toronto consensus
which is significant as automated software for as to the diagnostic criteria for DPN [77 79,
analysis would be required if CCM were to be 81,83], that is one abnormal finding as part of
introduced in large-scale screening. NCS, in combination with a symptom or sign
Perkins et al. [81], in a consortium multicen- of neuropathy [18]. As mentioned previously,
ter study funded by NIH, assessed data from a NCS assesses large-fiber function whereas
large cohort of 998 subjects. This large cohort CCM assesses small-fiber function.
of different ethnicities and T1DM and T2DM Despite the variation in results and limita-
gave a more accurate representation of the tions of the studies, these findings supported
population of people with diabetes instead the expanded role of CCM in the assessment of
of focusing on one specific subgroup [81]. diagnosis DPN as a supplement to the vast
Another strength of this study was that it sug- array of neurological tests currently in use.
gested an alternative approach of using one
lower value chosen to more confidently rule
in the presence of neuropathy (maximize speci-
Early detection of neuropathy
ficity) and one higher value determined to As there are currently no therapeutic agents
simultaneously, more confidently rule out the approved for DPN treatment, early detection
presence of neuropathy (maximize sensitivity). is essential to modify any risk factors. Several
This combination of decision criteria aims studies have specifically investigated CCM
to minimize false-positive and false-negative findings in the early stages of DM and mild
results. The study found that using this crite- levels of DPN.
rion increased their sensitivity to 88% and The published baseline characteristics of
specificity to 89% using manual methods of T1DM patients as part of the LANDMark study
analysis. However, this method caused 57.8% [167] were that CNFL was reduced in patients
of the subjects to be unclassified as they fell without clinical neuropathy, based on the
between the two limits. Toronto criteria. Another paper written from
There were several limitations to these the same study [168] assessed the use of CCM
cross-sectional studies. Some did not match the for distinguishing between control patients and
clinical profiles of their patients to the control DM patients (156 T1DM, 75 T2DM) with and
subjects. For example, the patient group with without clinical DPN. For the patients with
neuropathy in Alam et al. [77] was significantly DPN, all cases were defined as mild (as defined
older, with significantly longer disease duration by QST plus neurophysiology). This study

Diabetic Neuropathy
66 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

reported a significant reduction in CNFL when to control subjects. This agreed with a decrease
comparing patients with and without mild neu- in IENFD and significantly higher warm
ropathy, suggesting that CNFL changes may thresholds and vibration perception thresholds
occur early in the course of the disease. in the same cohort.
One study [169] assessed the corneal subba-
sal plexus in patients with recently diagnosed
Langerhans cells in diabetic peripheral
T2DM (mean duration 2.1 6 1.6 years). This
study reported significant differences between
neuropathy
CNFD, CNBD and CNFL parameters when The dominant antigen-presenting cells in the
comparing the patient cohort to the control cornea and ocular surfaces are LCs and DCs,
group, with CNFD emerging as the most sensi- which are derived from the bone marrow
tive in detecting corneal nerve pathology; and can stimulate both primary and secondary
indeed, 21% of the patients fell below the 2.5th T- and B-cell responses [147]. It has been shown
percentile of the control group. For this study, that CCM provides a noninvasive means to
high-adapted software produced an image readily demonstrate LCs in the cornea of
composed of an image stack. It reconstructed a healthy subjects and a range of inflammatory
combined mosaic image with an expanded ophthalmic conditions (Fig. 4.2) [152,171]. Some
field of view compared to standard imaging studies demonstrated that the number of LCs
using CCM. This software is also able to cor- increases in diabetic neuropathy [155,172 174];
rect for artifacts. As this method is not widely however, the LCs activation mechanism is still
used, there is no direct comparison to other unclear.
studies. To our knowledge, no other studies Zhivov et al. [154] assessed the corneal basal
are assessing recently diagnosed patients with epithelial layer and the subbasal nerve plexus
DM (,2 years duration). It must also be con- for the presence of LCs in healthy subjects and
sidered that in this study, even though patients found that 31% of subjects had LCs present.
were diagnosed newly, there may have been a Tavakoli and colleagues [155] were the first to
delay in diagnosis, which could have varied assess LC density with differing severities of
between patients. diabetic neuropathy (based on NDS scoring
Another study assessing early nerve changes compared to controls). This study found a sig-
assessed patients with IGT [170]. This study nificant increase in the proportion of indivi-
reported evidence that CCM may detect duals with LCs in patients with T1DM and
changes in nerve parameters before established T2DM (73.8%) compared to control subjects
diabetes. Asghar and colleagues [170] reported (46.1%). The study also found that LC density
that in patients with IGT, CNFD and CNBD was significantly increased in the patients with
were significantly reduced with 40.5% of sub- diabetes (17.73 6 1.45) compared to control
jects with IGT having significant small-fiber subjects (6.94 6 1.58). However, with progres-
damage based on CNFD reduction compared sion of neuropathy, patients with moderate

FIGURE 4.2 Images from Bowman’s layer of the cornea

at (A) T2DM with mild neuropathy; (B) moderate neuropa-
thy; (C) healthy control subject (yellow arrows show LCs
and red arrows indicate main corneal C nerve fibers).

Diabetic Neuropathy
 Diagnostic tests for diabetic peripheral neuropathy 67
and severe neuropathy showed a reduction in those greater than 50 μm with dendritic struc-
the LC density in comparison to patients with tures were considered as mature cells. A signif-
mild neuropathy and were not significantly icant increase in mature and immature cells
different from control subjects. This may sug- was found, and a correlation existed between
gest that LCs have a role in the early phase of CNFD and LCs density [176]. However, this
nerve damage. This study only focused on study only assessed a specific age-group of
Bowman’s layer which has been shown to have the diabetic cohort, so it does not represent
a lower density of LCs in comparison to the the whole diabetic population. Overall, studies
epithelial layer [175], so it is not an accurate investigating LCs density in patients with dia-
representation of overall LC density in the cen- betes are still limited, and more information
tral cornea. Another limitation of the study is required to conclude the effect of diabetes
was that the Tomey Confoscan CCM was used on LCs.
for imaging which has been shown to underes-
timate LCs density compared to the newer
Comparing corneal confocal microscopy
Heidelberg HRT III CCM [154], and cannot dif-
ferentiate mature from immature LCs [154].
and intraepidermal nerve fiber density
A more recent study [176], used the HRT Studies have found CCM to be comparable
(III) CCM to assess the density of LCs in a with measures of IENFD from biopsies in their
cohort of children and adolescents with diabe- diagnostic performance for detecting patients
tes and found a higher percentage of patients with clinical levels of DPN [77,79] (Fig. 4.3).
(85.9%) and controls (69.1%) with LCs present Both studies found no significant difference
when compared to the previous two studies in their diagnostic efficacy in patients with
[154,155]. This study was also able to distin- T1DM.
guish between mature and immature cells by An older study using the Tomoscan light
classing LCs of less than 50 μm in length, with- confocal microscope [87] also concluded that
out dendritic structures as immature cells, and both IENFD and CCM assessment accurately

FIGURE 4.3 Corneal confocal microscopy images of the corneal, subbasal nerves (A) (C). Healthy control (A) shows
numerous corneal main nerve fibers (green arrowheads) with branching nerves (blue asterisks). CCM images of patients
with mild (B) and severe (C) diabetic neuropathy demonstrate reduced corneal nerves and branches. Skin biopsies (D) and
(E) immunostained. Healthy control (D) shows numerous intraepidermal nerve fibers (red arrowheads) with a well-
developed subepidermal nerve plexus (yellow arrowheads). A patient with diabetes (E) demonstrates reduced subepider-
mal and minimal intraepidermal nerve fibers. Scale bar 5 100 mm. Source: (D) and (E) Adapted from Tavakoli M, Mitu-
Pretorian M, Petropoulos IN, Fadavi H, Asghar O, Alam U, et al. (2013). Corneal confocal microscopy detects early nerve regeneration
in diabetic neuropathy after simultaneous pancreas and kidney transplantation. Diabetes; 62(1):254 60.

Diabetic Neuropathy
68 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

quantify small nerve fiber damage in patients contrast, the subbasal nerve plexus is made up
with diabetes. Intraepidermal and corneal of C-fibers only. This means that a direct com-
nerve fiber lengths were also both further parison cannot be made between the two mea-
reduced in patients with painful compared surements. Although the Aδ-fibers only make
with painless diabetic neuropathy. up 10% of the total number in the epidermal
In comparison, one study’s findings, using layer, they may be affected differently in DPN
HRT (III) CCM, were notably different [169]. than the unmyelinated C-fibers, affecting the
This study reported that CCM and IENFD overall results.
were both able to detect nerve fiber loss in
recently diagnosed type 2 diabetes, but mainly
in different patients. They, therefore suggested Longitudinal studies for application of
a possible patchy manifestation pattern of
corneal confocal microscopy for
small-fiber neuropathy. Only 2.7% of the
patients had both abnormal CNFD and IENFD.
assessment diabetic peripheral neuropathy
Abnormal CCM with normal IEND was noted Longitudinal studies suggest that CCM
in 20.5% of the diabetic group and 11.0% for has a good predictive value for subsequent
vice versa. No correlation between the CCM DPN [166,178]. Longitudinal analysis of a
measures and IENFD was observed. There are T1DM cohort showed a mean 1-year change
possible explanations for these contradictory in CNFL of 1.6% in patients with unstable
findings. First, the cohort of patients in this T1DM, whereas healthy volunteers showed a
study was all patients with T2DM, all of who 5% increase per year [178].
had been newly diagnosed (known diabetes As part of a 4-year follow-up study, a study
duration of 2.1 6 1.8 years). The disease dura- [80] (Table 4.5) found that three corneal nerve
tion was significantly less than that of Chen parameters were all significant predictors for
et al. [79] (DPN 1 39 6 14 DPN 23 6 15 years) the development of DPN, with a baseline
and Alam et al. [77] (DSPN 1 37.2 6 13.1 CNFL of ,14.9 mm/mm2 being the strongest
DSPN 17.2 6 12.0 years). These two studies single predictor when compared to 11 other
also used comparisons between patients with small and large-fiber tests. Other studies
and without clinical DPN to compare IENFD [82,179] also reported an association between
and CCM, whereas Ziegler et al. [169] only lower baseline CNFL and DPN development.
compared patients with T2DM to healthy con- Pritchard et al. [82] (Table 4.5) found a signifi-
trols. Finally, Ziegler et al. used a different cant association with longer diabetes duration,
location for the IENFD biopsy. This was taken higher triglycerides, worsening retinopathy
from the lateral calf in comparison to the dor- and nephropathy, impaired sensation to tem-
sum of the foot. This more proximal site perature and vibration, and slower peroneal
may have been at less risk of IENFD changes and sural nerve conduction velocities. However,
or may present a different pattern of loss, as studies with larger cohorts and patients with
DSPN is known to follow a distal-proximal type 2 diabetes are needed to confirm these
course. studies’ findings and a more extended period
One issue with the comparison of IENFD of monitoring. Studies should also ensure a set
with analysis of the corneal subbasal nerve number of follow-ups over a set period; as for
plexus is that intraepidermal nerves consist of Lovblom et al. [80], more than half of the
both unmyelinated C-fibers (90%) and myelin- patients had just one follow-up visit, meaning
ated Aδ-fibers (10%) [177], which are both that true progression is statistically challenging
included in the measurement for IENFD. In prove.

Diabetic Neuropathy
 Diagnostic tests for diabetic peripheral neuropathy 69
Another prospective study specifically transplant and found a significant improve-
looked at a group of patients with IGT at the ment in all CCM parameters at 12 months.
first visit [180]. They found that in subjects Symptoms, neurophysiology, QST and skin
with IGT, lower baseline CNFD, CNBD, CNFL, biopsy results remained unchanged in the
and lower mean dendritic length of IENF were same patients. A similar, earlier study using
the strongest predictors of progression to an older CCM model also reported similar
T2DM over 3 years. Although significance findings, with CNFD and CNFL increasing sig-
was not recorded, there appeared to be very nificantly after just 6 months [182]. These stud-
similar baseline HbA1c measures between ies may demonstrate that CCM can provide a
those patients who remained IGT versus those novel noninvasive means to evidence early
developing T2DM over the 3 years follow-up nerve repair missed by currently advocated
(42.8 6 1.2 and 42.4 6 1.0, respectively assessment techniques. However, an alterna-
(mmol/mol)), suggesting that corneal nerve tive interpretation of this data could be that
parameters may have been stronger predictors corneal nerves respond well to the restoration
of conversion to T2DM in comparison to base- of insulin and normoglycemia. In contrast,
line HbA1c. Those subjects who returned to other peripheral nerves do not; therefore CCM
normoglycemia showed a significant improve- may be measuring something unique that is
ment in their CCM parameters while IENF not an accurate biomarker of the condition of
length continued to decline during the same peripheral nerves.
period. These findings may suggest that cor-
neal nerve fibers regenerate quicker than IENF
when glycemic control is improved [180].
Corneal confocal microscopy application
Another observational follow-up study [181]
examined a small cohort of patients with dia-
in clinical trials
betes (15 T1DM and 10 T2DM) at baseline and Several DPN intervention trials have focused
follow-up at 2 years. At follow-up, an improve- on large-fiber function and have generally had
ment in glycemic control, cholesterol levels ineffective outcomes. More recently, some stud-
and blood pressure were found and increased ies have instead focused on CCM measures as
CNFD, with a significant correlation between a markers for clinical trials of potential new treat-
decrease in HbA1c and increased CNFD. This ments. In a recent pilot trial of seal oil omega-3,
demonstrated that improvements in HbA1c polyunsaturated fatty acid supplementation in
might lead to morphological repair of corneal patients with type 1 diabetes (disease duration
nerve fibers; however, due to the small sample 27 6 14 years) over 12 months [183], there was
size and mixing of T1DM and T2DM in analy- a significant increase (30.4%) in CNFL, with no
sis, it is unclear if these differences are occur- change found in NCS velocity or sensory func-
ring in both types of diabetes. It must also be tion. Those subjects at high risk for future
noted that this was not planned as an interven- DPN and those with already diagnosed DPN
tional study, meaning there were no placebo (as determined by a TCNS of $ 1) showed the
controls or randomization, which would need best treatment response. This study was a
to take place to confirm or reject these findings. single-arm, open-label, proof of concept trial;
CCM has been used to investigate the therefore no placebo group was used, which is
subbasal nerve plexus changes in patients with necessary to reduce a trial’s bias.
T1DM postsimultaneous pancreas and kidney Another study to determine whether the
(SPK) transplant. Tavakoli et al. [164] assessed peptide, ARA 290, improves metabolic control
15 patients at 6 and 12 months postSPK and neuropathic pain in patients with type 2

Diabetic Neuropathy
70 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

diabetes used CCM measurements as a coprim- them at risk of having the first presentation
ary endpoint. This study found that ARA 290 with an active foot ulcer in A&E. Early diagno-
treatment was associated with increased CNFD sis of DPN is critical since early damage to the
correlated with changes in neuropathic symp- small nerve fibers in the feet of people with
toms [184]. This study was a double-blind, diabetes can be stopped from progressing and
placebo-controlled investigator-initiated phase even reversed while late signs of DPN and in
II clinical trial whose inclusion criteria were particular lack of vital protective sensation in
patients with T2DM who also had small-fiber the feet cannot be reversed. Although a com-
neuropathy symptoms. Whether allocation to mon and much-feared complication, over 30%
the treatment and placebo groups was random- of people with diabetes remain unaware.
ized was not discussed in the article. This Diabetes United Kingdom consistently states
study’s limitation was that patients assigned to that 80% of diabetes-related foot complications
both groups generally had excellent metabolic are preventable, yet no practical solution to this
control (HbA1c 5 7.3 6 0.4 and 6.9 6 0.2 for huge problem has been proposed. The only pri-
treatment and placebo groups, respectively), mary care test recommended by the National
which does not truly represent the clinical pop- Institute of Health and Care Excellence (NICE)
ulation of patients with T2DM. It may be that is the 10 g Semmes Weinstein monofilament
this treatment is less or more effective for examination (SWME), which is a crude, inaccu-
patients with poor metabolic control, compara- rate, and subjective test for lack of sensation in
tively. Finally, disease duration was also not the feet of people with diabetes—defined as
mentioned, so it was unclear if there was a sig- “late neuropathy” which cannot be reversed. Its
nificant difference between the two groups. evidence base is poor, and outside NICE, it is
These trials may be evidence that, such as regarded as a particularly poor test for DPN.
small-fiber damage occurring before large-fiber Early diagnosis and timely intervention are
damage, small fibers are also the first to start thus essential in preventing DPN development.
regenerating after damage. Trials over a longer Whereas measurement of urinary albumin
period, including other small-fiber neuropathy excretion and fundoscopic examinations serve
measures, are required before these findings as objective tests for early nephropathy and ret-
can be confirmed. inopathy, respectively, a comparably objective,
accurate test which is unbiased by the patient’s
subjective response is lacking for DPN.
Conclusions Currently advocated diagnostic tests either
focus on large nerve fibers, thus are not sensitive
There is an unmet need for a simple, reli- to early abnormalities, are too time-consuming
able, and accurate test for the early detection of and/or are too invasive to be used for repeated
DPN, which may help reduce the incidence of measures.
ulcers and amputations in people with diabe- More recently, a number of noninvasive
tes. Current tests for DPN in primary care tests have been developed as surrogate mea-
require HCPs to conduct and detect late neu- sures of DPN. Of these, CCM has shown great
ropathy. Many people with diabetes do not potential for detecting small-fiber neuropathy,
have an annual foot check despite it being one the earliest manifestation of DPN. CCM has
of the most important of the eight care pro- also demonstrated promising prognostic utility
cesses mandated by NICE. In fact, over 500,000 and has demonstrated early nerve regeneration
people in England alonenever have annual post-SPK surgery and as part of several clinical
diabetes-related foot examinations. This places trials.

Diabetic Neuropathy
 References 71
Given that CCM is a rapid and noninvasive [5] Pop-Busui R, Boulton AJM, Feldman EL, Bril V,
test, it is suitable for large-scale screening for Freeman R, Malik RA, et al. Diabetic neuropathy: a
position statement by the American Diabetes Association.
DPN, and advancements in automated analysis Diabetes Care 2017;40(1):136.
software would further improve its promising [6] Sloan G, Shillo P, Selvarajah D, Wu J, Wilkinson ID,
potential. Tracey I, et al. A new look at painful diabetic neuropa-
In conclusion, there is no optimal biomarker thy. Diabetes Res Clin Pract 2018;144:177 91.
and ideal endpoint available for DPN at the [7] Albers JW, Herman WH, Pop-Busui R, Feldman EL,
Martin CL, Cleary PA, et al. Effect of prior intensive
current time. Hence, there is an urgent need to insulin treatment during the Diabetes Control and
identify the most accurate early biomarker of Complications Trial (DCCT) on peripheral neuropathy
nerve damage to diagnose DPN in patients’ in type 1 diabetes during the Epidemiology of
clinical care better and, in particular, to permit Diabetes Interventions and Complications (EDIC)
a precise evaluation of future therapies in Study. Diabetes Care 2010;33(5):1090 6.
[8] Kioskli K, Scott W, Winkley K, Kylakos S, McCracken
clinical trials. The global effort among scientists LM. Psychosocial factors in painful diabetic neuropa-
and clinicians, and researchers in the field thy: a systematic review of treatment trials and survey
should address these shortcomings to reduce studies. Pain Med 2019;20(9):1756 73.
the incidence of complications and achieve [9] Singh N, Armstrong DG, Lipsky BA. Preventing
this; the search should continue for better and foot ulcers in patients with diabetes. JAMA 2005;293
(2):217 28.
sensitive tests, screening, and early detection. [10] Ndip A, Ebah L, Mbako A. Neuropathic diabetic foot
We need to improve our systematic evaluation ulcers - evidence-to-practice. Int J Gen Med 2012;5:129 34.
of the evidence and promote from each transla- [11] Smith BE. Chapter 3 - Focal and entrapment neuropa-
tional step to the next—the biomarkers with thies. In: Zochodne DW, Malik RA, editors. Handbook
the best evidence and performance at different of clinical neurology, 126. Elsevier; 2014. p. 31 43.
[12] National Institute for Health and Care Excellence
populations. This will require evaluation of the (NICE). Diabetic foot problems: prevention and manage-
wider biomarker research agenda. Such evalu- ment; (2015) [updated 11th October 2019]. Available
ation may also benefit more from fostering from: ,https://www.nice.org.uk/guidance/ng19..
international collaborations rather than from [13] Thorud JC, Plemmons B, Buckley CJ, Shibuya N,
the fragmented efforts of small, opportunistic Jupiter DC. Mortality after nontraumatic major ampu-
tation among patients with diabetes and peripheral
studies. “We must learn to measure what we vascular disease: a systematic review. J Foot Ankle
value rather than valuing what we can easily Surg 2016;55(3):591 9.
measure.” [14] Manzano GM, Giuliano LMP, Nóbrega JAM. A brief
historical note on the classification of nerve fibers. Arq
Neuro-Psiquiatr 2008;66:117 19.
References [15] American Diabetes Association. 11. Microvascular
[1] Neeland IJ, Patel KV. Chapter 4 - Diabetes: key mar- complications and foot care: standards of medical
kers of injury and prognosis. In: Nambi V, editor. care in diabetes-2020. Diabetes Care 2020;43(Suppl 1):
Biomarkers in cardiovascular disease. Elsevier; 2019. 135 51.
p. 41 51. [16] Bakker K, Apelqvist J, Schaper NC. Practical guide-
[2] International Diabetes Federation. IDF diabetes atlas; lines on the management and prevention of the diabetic
(2019). Available from: ,https://diabetesatlas.org/ foot 2011. Diabetes Metab Res Rev 2012;28(Suppl 1):
en/sections/worldwide-toll-of-diabetes.html.. 225 31.
[3] Diabetes UK. Us, diabetes and a lot of facts and stats; [17] Yang Z, Zhang Y, Chen R, Huang Y, Ji L, Sun F, et al.
(2019). Available from: ,https://www.diabetes.org. Simple tests to screen for diabetic peripheral neuropa-
uk/resources-s3/2019-02/1362B_Facts%20and%20stats% thy. Cochrane Database Syst Rev 2014;7(2):CD010975.
20Update%20Jan%202019_LOW%20RES_EXTERNAL. [18] Tesfaye S, Boulton AJM, Dyck PJ, Freeman R,
pdf.. Horowitz M, Kempler P, et al. Diabetic neuropathies:
[4] Hicks CW, Selvin E. Epidemiology of peripheral neu- update on definitions, diagnostic criteria, estimation
ropathy and lower extremity disease in diabetes. Curr of severity, and treatments. Diabetes Care 2010;33
Diabetes Rep 2019;19(10):86 98. (10):2285 93.

Diabetic Neuropathy
72 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

[19] Feng Y, Schlösser FJ, Sumpio BE. The Semmes [32] Killian JM, Foreman PJ. Clinical utility of dorsal sural
Weinstein monofilament examination as a screening nerve conduction studies. Muscle Nerve 2001;24
tool for diabetic peripheral neuropathy. J Vasc Surg (6):817 20.
2009;50(3):675 82. [33] Shabeeb D, Najafi M, Hasanzadeh G, Hadian MR,
[20] Dros J, Wewerinke A, Bindels PJ, van Weert HC. Musa AE, Shirazi A. Electrophysiological measure-
Accuracy of monofilament testing to diagnose periph- ments of diabetic peripheral neuropathy: a systematic
eral neuropathy: a systematic review. Ann Fam Med review. Diabetes Metab Syndr: Clin Res Rev 2018;12
2009;7(6):555 8. (4):591 600.
[21] Petropoulos IN, Ponirakis G, Khan A, Almuhannadi H, [34] Bril V. Role of electrophysiological studies in diabetic
Gad H, Malik RA. Diagnosing diabetic neuropathy: neuropathy. Can J Neurological Sci / J Canadien des
something old, something new. Diabetes Metab J Sci Neurologiques 1994;21(S4):S8 12.
2018;42(4):255 69. [35] Dyck PJ, Overland CJ, Low PA, Litchy WJ, Davies JL,
[22] Meijer JW, Smit AJ, Sonderen EV, Groothoff JW, Dyck PJB, et al. Signs and symptoms versus nerve
Eisma WH, Links TP. Symptom scoring systems conduction studies to diagnose diabetic sensorimotor
to diagnose distal polyneuropathy in diabetes: the polyneuropathy: Cl vs. NPhys trial. Muscle Nerve
Diabetic Neuropathy Symptom score. Diabet Med 2010;42(2):157 64.
2002;19(11):962 5. [36] Ragé M, Van Acker N, Knaapen MW, Timmers M,
[23] Young MJ, Boulton AJ, MacLeod AF, Williams DR, Streffer J, Hermans MP, et al. Asymptomatic small
Sonksen PH. A multicentre study of the prevalence of fiber neuropathy in diabetes mellitus: investigations
diabetic peripheral neuropathy in the United with intraepidermal nerve fiber density, quantitative
Kingdom hospital clinic population. Diabetologia. sensory testing and laser-evoked potentials. J Neurol
1993;36(2):150 4. 2011;258(10):1852 64.
[24] Zilliox LA, Ruby SK, Singh S, Zhan M, Russell JW. [37] Løseth S, Stålberg E, Jorde R, Mellgren SI. Early
Clinical neuropathy scales in neuropathy associated diabetic neuropathy: thermal thresholds and intraepi-
with impaired glucose tolerance. J Diabetes dermal nerve fibre density in patients with normal
Complications 2015;29(3):372 7. nerve conduction studies. J Neurol 2008;255(8):
[25] Sharma S, Kerry C, Atkins H, Rayman G. The Ipswich 1197 202.
Touch Test: a simple and novel method to screen [38] Ostrovski I, Lovblom LE, Farooqi MA, Scarr D,
patients with diabetes at home for increased risk of Boulet G, Hertz P, et al. Reproducibility of in vivo cor-
foot ulceration. Diabet Med 2014;31(9):1100 3. neal confocal microscopy using an automated analysis
[26] Rayman G, Vas PR, Baker N, Taylor Jr. CG, Gooday C, program for detection of diabetic sensorimotor poly-
Alder AI, et al. The Ipswich Touch Test: a simple and neuropathy. PLoS One 2015;10(11):e0142309.
novel method to identify inpatients with diabetes at risk [39] Tavakoli M, Quattrini C, Abbott C, Kallinikos P,
of foot ulceration. Diabetes Care 2011;34(7):1517 18. Marshall A, Finnigan J, et al. Corneal confocal micros-
[27] Zaslansky R, Yarnitsky D. Clinical applications of copy: a novel noninvasive test to diagnose and stratify
quantitative sensory testing (QST). J Neurological Sci the severity of human diabetic neuropathy. Diabetes
1998;153(2):215 38. Care 2010;33(8):1792 7.
[28] Themistocleous AC, Ramirez JD, Serra J, Bennett DL. [40] Tavakoli M, Malik RA. Corneal confocal microscopy:
The clinical approach to small fibre neuropathy a novel non-invasive technique to quantify small
and painful channelopathy. Pract Neurol 2014;14 fibre pathology in peripheral neuropathies. J Vis Exp
(6):368 79. 2011;47:2194.
[29] Lin Y-H, Hsieh S-C, Chao C-C, Chang Y-C, Hsieh S-T. [41] Tentolouris N, Achtsidis V, Marinou K, Katsilambros N.
Influence of aging on thermal and vibratory thresh- Evaluation of the self-administered indicator plaster
olds of quantitative sensory testing. J Peripher Nerv neuropad for the diagnosis of neuropathy in diabetes.
Syst 2005;10(3):269 81. Diabetes Care 2008;31(2):236.
[30] Lee JA, Halpern EM, Lovblom LE, Yeung E, Bril V, [42] Quattrini C, Jeziorska M, Tavakoli M, Begum P,
Perkins BA. Reliability and validity of a point-of-care Boulton AJM, Malik RA. The Neuropad test: a visual
sural nerve conduction device for identification of dia- indicator test for human diabetic neuropathy.
betic neuropathy. PLoS One 2014;9(1):e86515. Diabetologia 2008;51(6):1046 50.
[31] Perkins BA, Grewal J, Ng E, Ngo M, Bril V. Validation [43] Ponirakis G, Petropoulos IN, Fadavi H, Alam U,
of a novel point-of-care nerve conduction device for Asghar O, Marshall A, et al. The diagnostic accuracy
the detection of diabetic sensorimotor polyneuropa- of Neuropad for assessing large and small fibre dia-
thy. Diabetes Care 2006;29(9):2023. betic neuropathy. Diabet Med 2014;31(12):1673 80.

Diabetic Neuropathy
 References 73
[44] Papanas N, Papatheodorou K, Christakidis D, [57] Bril V, Tomioka S, Buchanan RA, Perkins BA, mTCNS
Papazoglou D, Giassakis G, Piperidou H, et al. Study Group. Reliability and validity of the modified
Evaluation of a new indicator test for sudomotor func- Toronto Clinical Neuropathy Score in diabetic sensori-
tion (Neuropad) in the diagnosis of peripheral neurop- motor polyneuropathy. Diabet Med 2009;26(3):240 6.
athy in type 2 diabetic patients. Exp Clin Endocrinol [58] Gupta H, Jin KH, Nguyen HQ, McCann MT, Unser M.
Diabetes 2005;113(4):195 8. CNN-based projected gradient descent for consistent
[45] Liatis S, Marinou K, Tentolouris N, Pagoni S, CT image reconstruction. IEEE Trans Med Imaging
Katsilambros N. Usefulness of a new indicator test for 2018;37(6):1440 53.
the diagnosis of peripheral and autonomic neuropathy [59] Hawker GA, Mian S, Kendzerska T, French M.
in patients with diabetes mellitus. Diabet Med 2007;24 Measures of adult pain: Visual Analog Scale for Pain
(12):1375 80. (VAS Pain), Numeric Rating Scale for Pain (NRS
[46] Kamenov ZA, Petrova JJ, Christov VG. Diagnosis of Pain), McGill Pain Questionnaire (MPQ), Short-Form
diabetic neuropathy using simple somatic and a new McGill Pain Questionnaire (SF-MPQ), Chronic Pain
autonomic (neuropad) tests in the clinical practice. Grade Scale (CPGS), Short Form-36 Bodily Pain Scale
Exp Clin Endocrinol Diabetes 2010;118(4):226 33. (SF-36 BPS), and Measure of Intermittent and
[47] Yajnik CS, Kantikar VV, Pande AJ, Deslypere JP. Constant Osteoarthritis Pain (ICOAP). Arthritis Care
Quick and simple evaluation of sudomotor function Res (Hoboken) 2011;63(Suppl 11):S240 52.
for screening of diabetic neuropathy. ISRN Endocrinol [60] Herman WH, Pop-Busui R, Braffett BH, Martin CL,
2012;2012:103714. Cleary PA, Albers JW, et al. Use of the Michigan
[48] Smith AG, Lessard M, Reyna S, Doudova M, Neuropathy Screening Instrument as a measure of distal
Singleton JR. The diagnostic utility of sudoscan for dis- symmetrical peripheral neuropathy in type 1 diabetes:
tal symmetric peripheral neuropathy. J Diabetes results from the diabetes control and complications
Complications 2014;28(4):511 16. trial/epidemiology of diabetes interventions and compli-
[49] Selvarajah D, Cash T, Davies J, Sankar A, Rao G, cations. Diabet Med 2012;29(7):937 44.
Grieg M, et al. SUDOSCAN: a simple, rapid, and [61] Meijer J-WG, Bosma E, Lefrandt JD, Links TP, Smit AJ,
objective Method with potential for screening for dia- Stewart RE, et al. Clinical diagnosis of diabetic poly-
betic peripheral neuropathy. PLoS One 2015;10(10): neuropathy with the diabetic neuropathy symptom
e0138224. and diabetic neuropathy examination scores. Diabetes
[50] Krieger S-M, Reimann M, Haase R, Henkel E, Care 2003;26(3):697.
Hanefeld M, Ziemssen T. Sudomotor testing of diabe- [62] Meijer JW, van Sonderen E, Blaauwwiekel EE, Smit
tes polyneuropathy. Front Neurol 2018;9:803. AJ, Groothoff JW, Eisma WH, et al. Diabetic neuropa-
[51] Casellini CM, Parson HK, Richardson MS, Nevoret thy examination: a hierarchical scoring system to diag-
ML, Vinik AI. Sudoscan, a noninvasive tool for detect- nose distal polyneuropathy in diabetes. Diabetes Care
ing diabetic small fiber neuropathy and autonomic 2000;23(6):750 3.
dysfunction. Diabetes Technol Therapeut 2013;15(11): [63] Spallone V, Morganti R, D’Amato C, Greco C,
948 53. Cacciotti L, Marfia GA. Validation of DN4 as a screen-
[52] Thaisetthawatkul P, Fernandes Filho JA, Herrmann ing tool for neuropathic pain in painful diabetic poly-
DN. Contribution of QSART to the diagnosis of small neuropathy. Diabet Med 2012;29(5):578 85.
fiber neuropathy. Muscle Nerve 2013;48(6):883 8. [64] Asad A, Hameed MA, Khan UA, Ahmed N, Butt MU.
[53] Buchmann SJ, Penzlin AI, Kubasch ML, Illigens BM, Reliability of the neurological scores for assessment of
Siepmann T. Assessment of sudomotor function. Clin sensorimotor neuropathy in type 2 diabetics. J Pak
Auton Res 2019;29(1):41 53. Med Assoc 2010;60(3):166 70.
[54] Feldman EL, Stevens MJ, Thomas PK, Brown MB, [65] Melzack R, Katz J. The McGill Pain Questionnaire:
Canal N, Greene DA. A practical two-step quantitative appraisal and current status. Handbook of pain assess-
clinical and electrophysiological assessment for the ment. 2nd ed. New York, NY: The Guilford Press;
diagnosis and staging of diabetic neuropathy. 2001. p. 35 52.
Diabetes Care 1994;17(11):1281 9. [66] Melzack R. The short-form McGill pain questionnaire.
[55] Bennett M. The LANSS Pain Scale: the Leeds assess- Pain 1987;30(2):191 7.
ment of neuropathic symptoms and signs. Pain [67] Dyck PJ, Sherman WR, Hallcher LM, John Service F,
2001;92(1-2):147 57. O’Brien PC, Grina LA, et al. Human diabetic endo-
[56] Bril V. NIS-LL: the primary measurement scale for neurial sorbitol, fructose, and myo-inositol related
clinical trial endpoints in diabetic peripheral neuropa- to sural nerve morphometry. Ann Neurol 1980;8(6):
thy. Eur Neurol 1999;41(Suppl 1):8 13. 590 6.

Diabetic Neuropathy
74 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

[68] Abbott CA, Carrington AL, Ashe H, Bath S, Every LC, [80] Lovblom LE, Halpern EM, Wu T, Kelly D, Ahmed A,
Griffiths J, et al. The North-West Diabetes Foot Care Boulet G, et al. In vivo corneal confocal microscopy
Study: incidence of, and risk factors for, new diabetic and prediction of future-incident neuropathy in type 1
foot ulceration in a community-based patient cohort. diabetes: a preliminary & longitudinal analysis. Can J
Diabet Med 2002;19(5):377 84. Diabetes 2015;39(5):390 7.
[69] Bril V, Perkins BA. Validation of the Toronto Clinical [81] Perkins BA, Lovblom LE, Bril V, Scarr D, Ostrovski I,
Scoring System for diabetic polyneuropathy. Diabetes Orszag A, et al. Corneal confocal microscopy for iden-
Care 2002;25(11):2048 52. tification of diabetic sensorimotor polyneuropathy: a
[70] Moghtaderi A, Bakhshipour A, Rashidi H. Validation pooled multinational consortium study. Diabetologia
of Michigan neuropathy screening instrument for dia- 2018;61(8):1856 61.
betic peripheral neuropathy. Clin Neurol Neurosurg [82] Pritchard N, Edwards K, Russell AW, Perkins BA,
2006;108(5):477 81. Malik RA, Efron N. Corneal confocal microscopy pre-
[71] Weisman A, Bril V, Ngo M, Lovblom LE, Halpern EM, dicts 4-year incident peripheral neuropathy in type 1
Orszag A, et al. Identification and prediction of dia- diabetes. Diabetes Care 2015;38(4):671 5.
betic sensorimotor polyneuropathy using individual [83] Scarr D, Lovblom LE, Lovshin JA, Boulet G, Farooqi
and simple combinations of nerve conduction study MA, Orszag A, et al. Lower corneal nerve fibre
parameters. PLoS One 2013;8(3):e58783. length identifies diabetic neuropathy in older adults
[72] Litchy WJ, Albers JW, Wolfe J, Bolton CF, Walsh N, with diabetes: results from the Canadian Study of
Klein CJ, et al. Proficiency of nerve conduction using Longevity in Type 1 Diabetes. Diabetologia 2017;60
standard methods and reference values (Cl. NPhys (12):2529 31.
Trial 4). Muscle Nerve 2014;50(6):900 8. [84] Tavakoli M, Begum P, McLaughlin J, Malik RA.
[73] Wahren J, Foyt H, Daniels M, Arezzo JC. Long-acting Corneal confocal microscopy for the diagnosis of dia-
C-peptide and neuropathy in type 1 diabetes: a 12- betic autonomic neuropathy. Muscle Nerve 2015;52
month clinical trial. Diabetes Care 2016;39(4):596 602. (3):363 70.
[74] Ruggenenti P, Lauria G, Iliev IP, Fassi A, Ilieva AP, [85] Sharma S, Vas PR, Rayman G. Assessment of diabetic
Rota S, et al. Effects of manidipine and delapril in neuropathy using a point-of-care nerve conduction
hypertensive patients with type 2 diabetes mellitus: device shows significant associations with the
the delapril and manidipine for nephroprotection LDIFLARE method and clinical neuropathy scoring. J
in diabetes (DEMAND) randomized clinical trial. Diabetes Sci Technol 2015;9(1):123 31.
Hypertension 2011;58(5):776 83. [86] Umapathi T, Tan WL, Loke SC, Soon PC, Tavintharan S,
[75] Dyck PJ, Norell JE, Tritschler H, Schuette K, Chan YH. Intraepidermal nerve fiber density as a
Samigullin R, Ziegler D, et al. Challenges in design of marker of early diabetic neuropathy. Muscle Nerve
multicenter trials: end points assessed longitudinally 2007;35(5):591 8.
for change and monotonicity. Diabetes Care 2007;30 [87] Quattrini C, Tavakoli M, Jeziorska M, Kallinikos P,
(10):2619 25. Tesfaye S, Finnigan J, et al. Surrogate markers of small
[76] Callaghan BC, Burke JF, Rodgers A, McCammon R, fiber damage in human diabetic neuropathy. Diabetes
Langa KM, Feldman EL, et al. Expenditures in the 2007;56(8):2148 54.
elderly with peripheral neuropathy: where should we [88] Lauria G, Hsieh ST, Johansson O, Kennedy WR, Leger
focus cost-control efforts? Neurol Clin Pract 2013;3 JM, Mellgren SI, et al. European Federation of
(5):421 30. Neurological Societies/Peripheral Nerve Society
[77] Alam U, Jeziorska M, Petropoulos IN, Asghar O, Fadavi Guideline on the use of skin biopsy in the diagnosis of
H, Ponirakis G, et al. Diagnostic utility of corneal confo- small fiber neuropathy. Report of a joint task force of
cal microscopy and intra-epidermal nerve fibre density the European Federation of Neurological Societies and
in diabetic neuropathy. PLoS One 2017;12(7):e0180175. the Peripheral Nerve Society. Eur J Neurol 2010;17
[78] Ahmed A, Bril V, Orszag A, Paulson J, Yeung E, Ngo M, (7):903 12.
et al. Detection of diabetic sensorimotor polyneuropathy [89] Lauria G, Cornblath DR, Johansson O, McArthur JC,
by corneal confocal microscopy in type 1 diabetes: a con- Mellgren SI, Nolano M, et al. EFNS guidelines on the
current validity study. Diabetes Care 2012;35(4):821 8. use of skin biopsy in the diagnosis of peripheral neu-
[79] Chen X, Graham J, Dabbah MA, Petropoulos IN, ropathy. Eur J Neurol 2005;12(10):747 58.
Ponirakis G, Asghar O, et al. Small nerve fiber quanti- [90] Provitera V, Nolano M, Stancanelli A, Caporaso G,
fication in the diagnosis of diabetic sensorimotor poly- Vitale DF, Santoro L. Intraepidermal nerve fiber
neuropathy: comparing corneal confocal microscopy analysis using immunofluorescence with and with-
with intraepidermal nerve fiber density. Diabetes Care out confocal microscopy. Muscle Nerve 2015;51
2015;38(6):1138 44. (4):501 4.

Diabetic Neuropathy
 References 75
[91] Nolano M, Biasiotta A, Lombardi R, Provitera V, 2020 eine Kurzübersicht des Geförderten Dan Ziegler.
Stancanelli A, Caporaso G, et al. Epidermal innerva- Diabetologie und Stoffwechs 2020;15(06):450 3.
tion morphometry by immunofluorescence and [104] Ishibashi F, Kojima R, Kawasaki A, Yamanaka E,
bright-field microscopy. J Peripher Nerv Syst 2015;20 Kosaka A, Uetake H. Correlation between sudomotor
(4):387 91. function, sweat gland duct size and corneal nerve
[92] Lauria G, Bakkers M, Schmitz C, Lombardi R, Penza P, fiber pathology in patients with type 2 diabetes melli-
Devigili G, et al. Intraepidermal nerve fiber density at tus. J Diabetes Investig 2014;5(5):588 96.
the distal leg: a worldwide normative reference study. [105] Papanas N, Giassakis G, Papatheodorou K,
J Peripheral Nerv Syst 2010;15(3):202 7. Papazoglou D, Monastiriotis C, Christakidis D, et al.
[93] Vlčková-Moravcová E, Bednařı́k J, Dušek L, Toyka Use of the new indicator test (Neuropad s) for the
KV, Sommer C. Diagnostic validity of epidermal assessment of the staged severity of neuropathy
nerve fiber densities in painful sensory neuropathies. in type 2 diabetic patients. Exp Clin Endocrinol
Muscle Nerve 2008;37(1):50 60. Diabetes 2007;115:58 61.
[94] Chien HF, Tseng TJ, Lin WM, Yang CC, Chang YC, [106] Manes C, Papanas N, Exiara T, Katsiki N,
Chen RC, et al. Quantitative pathology of cutaneous Papantoniou S, Kirlaki E, et al. The indicator test
nerve terminal degeneration in the human skin. Acta Neuropad in the assessment of small and overall
Neuropathol 2001;102(5):455 61. nerve fibre dysfunction in patients with type 2 diabe-
[95] Shun CT, Chang YC, Wu HP, Hsieh SC, Lin WM, Lin tes: a large multicentre study. Exp Clin Endocrinol
YH, et al. Skin denervation in type 2 diabetes: corre- Diabetes 2014;122(3):195 9.
lations with diabetic duration and functional impair- [107] Papanas N, Boulton AJ, Malik RA, Manes C, Schnell O,
ments. Brain 2004;127(Pt 7):1593 605. Spallone V, et al. A simple new non-invasive sweat
[96] Lauria G, Morbin M, Lombardi R, Borgna M, indicator test for the diagnosis of diabetic neuropathy.
Mazzoleni G, Sghirlanzoni A, et al. Axonal swellings Diabet Med 2013;30(5):525 34.
predict the degeneration of epidermal nerve fibers in [108] Spallone V, Morganti R, Siampli M, Fedele T,
painful neuropathies. Neurology 2003;61(5):631. D’Amato C, Cacciotti L, et al. Neuropad as a diagnos-
[97] Sorensen L, Molyneaux L, Yue DK. The relationship tic tool for diabetic autonomic and sensorimotor neu-
among pain, sensory loss, and small nerve fibers in ropathy. Diabet Med 2009;26(7):686 92.
diabetes. Diabetes Care 2006;29(4):883 7. [109] Mayaudon H, Miloche PO, Bauduceau B. A new sim-
[98] Løseth S, Stålberg EV, Lindal S, Olsen E, Jorde R, ple method for assessing sudomotor function: rele-
Mellgren SI. Small and large fiber neuropathy in vance in type 2 diabetes. Diabetes Metab 2010;36(6,
those with type 1 and type 2 diabetes: a 5-year Part 1):450 4.
follow-up study. J Peripheral Nerv Syst 2016;21 [110] Gin H, Baudoin R, Raffaitin CH, Rigalleau V,
(1):15 21. Gonzalez C. Non-invasive and quantitative assess-
[99] Malik RA, Veves A, Tesfaye S, Smith G, Cameron N, ment of sudomotor function for peripheral diabetic
Zochodne D, et al. Small fibre neuropathy: role in the neuropathy evaluation. Diabetes Metab 2011;37
diagnosis of diabetic sensorimotor polyneuropathy. (6):527 32.
Diabetes Metab Res Rev 2011;27(7):678 84. [111] Bordier L, Dolz M, Monteiro L, Névoret M-L,
[100] Illigens BMW, Gibbons CH. Sweat testing to evaluate Calvet J-H, Bauduceau B. Accuracy of a rapid and
autonomic function. Clin Autonomic Res 2008;19 non-invasive method for the assessment of small
(2):79 87. fiber neuropathy based on measurement of electro-
[101] Tsapas A, Liakos A, Paschos P, Karagiannis T, chemical skin conductances. Front Endocrinol
Bekiari E, Tentolouris N, et al. A simple plaster 2016;7:7 18.
for screening for diabetic neuropathy: a diagnostic [112] Mao F, Liu S, Qiao X, Zheng H, Xiong Q, Wen J, et al.
test accuracy systematic review and meta-analysis. Sudoscan is an effective screening method for asymp-
Metabolism 2014;63(4):584 92. tomatic diabetic neuropathy in Chinese type 2 diabe-
[102] Tavakoli M, Quattrini C, Begum P, Fadavi H, tes mellitus patients. J Diabetes Investig 2017;8
Boulton A, Malik R. Neuropad and corneal confocal (3):363 8.
microscopy: new indicators for human diabetic neu- [113] Vinik AI, Smith AG, Singleton JR, Callaghan B,
ropathy. Diabetologia 2010;53(S1112):443 4. Freedman BI, Tuomilehto J, et al. Normative values
[103] Ziegler D. Untersuchung von Polymorphismen for electrochemical skin conductances and impact
in Kandidatengenen für periphere und kardiale of ethnicity on quantitative assessment of sudomo-
autonome Neuropathie in einer bevölkerungsbezogenen tor function. Diabetes Technol Ther 2016;18
Kohorte Hans-Christian-Hagedorn-Projektförderung (6):391 8.

Diabetic Neuropathy
76 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

[114] Novak P. Electrochemical skin conductance corre- possible biomarkers for persistent pain? J Rehabil Med
lates with skin nerve fiber density. Front Aging 2016;48(1):43 7.
Neurosci 2016;8. [127] Price RC, Asenjo JF, Christou NV, Backman SB,
[115] Rajan S, Campagnolo M, Callaghan B, Gibbons CH. Schweinhardt P. The role of excess subcutaneous fat
Sudomotor function testing by electrochemical skin in pain and sensory sensitivity in obesity. Eur J Pain
conductance: does it really measure sudomotor func- 2013;17(9):1316 26.
tion? Clin Autonomic Res 2019;29(1):31 9. [128] Meier PM, Berde CB, DiCanzio J, Zurakowski D,
[116] Ponirakis G, Odriozola MN, Odriozola S, Petropoulos Sethna NF. Quantitative assessment of cutaneous
IN, Azmi S, Fadavi H, et al. NerveCheck: an inexpen- thermal and vibration sensation and thermal pain
sive quantitative sensory testing device for patients detection thresholds in healthy children and adoles-
with diabetic neuropathy. Diabetes Res Clin Pract cents. Muscle Nerve 2001;24(10):1339 45.
2016;113:101 7. [129] Lysy Z, Lovblom LE, Halpern EM, Ngo M, Ng E,
[117] Yarnitsky D, Granot M, Nahman-Averbuch H, Orszag A, et al. Measurement of cooling detection
Khamaisi M, Granovsky Y. Conditioned pain modu- thresholds for identification of diabetic sensorimotor
lation predicts duloxetine efficacy in painful diabetic polyneuropathy in type 1 diabetes. PLoS One 2014;9
neuropathy. Pain 2012;153(6):1193 8. (9):e106995.
[118] Rolke R, Magerl W, Campbell KA, Schalber C, [130] Krämer HH, Rolke R, Bickel A, Birklein F. Thermal
Caspari S, Birklein F, et al. Quantitative sensory test- thresholds predict painfulness of diabetic neuropa-
ing: a comprehensive protocol for clinical trials. Eur J thies. Diabetes Care 2004;27(10):2386 91.
Pain 2006;10(1):77 88. [131] Müller LJ, Marfurt CF, Kruse F, Tervo TMT. Corneal
[119] Rolke R, Baron R, Maier C, Tölle TR, Treede RD, nerves: structure, contents and function. Exp Eye Res
Beyer A, et al. Quantitative sensory testing in the 2003;76(5):521 42.
German Research Network on Neuropathic Pain [132] Marfurt C, Anokwute MC, Fetcko K, Mahony-Perez E,
(DFNS): standardized protocol and reference values. Farooq H, Ross E, et al. Comparative anatomy of the
Pain 2006;123(3):231 43. mammalian corneal subbasal nerve plexus. Invest
[120] Magerl W, Krumova EK, Baron R, Tölle T, Treede R-D, Ophthalmol Vis Sci 2019;60(15):4972 84.
Maier C. Reference data for quantitative sensory test- [133] Ismail-Beigi F, Craven T, Banerji MA, Basile J, Calles J,
ing (QST): refined stratification for age and a novel Cohen RM, et al. Effect of intensive treatment of
method for statistical comparison of group data. Pain hyperglycaemia on microvascular outcomes in type 2
2010;151(3):598 605. diabetes: an analysis of the ACCORD randomised
[121] Dyck PJ, Dyck PJ, Kennedy WR, Kesserwani H, trial. Lancet 2010;376(9739):419 30.
Melanson M, Ochoa J, et al. Limitations of quantita- [134] Kalteniece A, Ferdousi M, Petropoulos I, Azmi S,
tive sensory testing when patients are biased toward Adam S, Fadavi H, et al. Greater corneal nerve loss at
a bad outcome. Neurology 1998;50(5):894 5. the inferior whorl is related to the presence of dia-
[122] Dyck PJ, Bushek W, Spring EM, Karnes JL, Litchy betic neuropathy and painful diabetic neuropathy.
WJ, O’Brien PC, et al. Vibratory and cooling detection Sci Rep 2018;8(1):3283.
thresholds compared with other tests in diagnosing [135] Tavakoli M, Petropoulos IN, Malik RA. Corneal con-
and staging diabetic neuropathy. Diabetes Care focal microscopy to assess diabetic neuropathy: an
1987;10(4):432 40. eye on the foot. J Diabetes Sci Technol 2013;7
[123] Backonja MM, Attal N, Baron R, Bouhassira D, (5):1179 89.
Drangholt M, Dyck PJ, et al. Value of quantitative [136] Tavakoli M, Hossain P, Malik RA. Clinical applica-
sensory testing in neurological and pain disorders: tions of corneal confocal microscopy. Clin Ophthalmol
NeuPSIG consensus. Pain 2013;154(9):1807 19. (Auckland, NZ) 2008;2(2):435 45.
[124] Attal N, Fermanian C, Fermanian J, Lanteri-Minet M, [137] Hertz P, Bril V, Orszag A, Ahmed A, Ng E, Nwe P,
Alchaar H, Bouhassira D. Neuropathic pain: are there et al. Reproducibility of in vivo corneal confocal
distinct subtypes depending on the aetiology or ana- microscopy as a novel screening test for early
tomical lesion? Pain 2008;138(2):343 53. diabetic sensorimotor polyneuropathy. Diabet Med
[125] Attal N, Bouhassira D, Gautron M, Vaillant JN, Mitry E, 2011;28(10):1253 60.
Lepère C, et al. Thermal hyperalgesia as a marker of [138] Oliveira-Soto L, Efron N. Morphology of corneal
oxaliplatin neurotoxicity: a prospective quantified sen- nerves using confocal microscopy. Cornea 2001;20
sory assessment study. Pain 2009;144(3):245 52. (4):374 84.
[126] Malmström EM, Stjerna J, Högestätt ED, Westergren H. [139] Maddaloni E, Sabatino F, Del Toro R, Crugliano S,
Quantitative sensory testing of temperature thresholds: Grande S, Lauria Pantano A, et al. In vivo corneal

Diabetic Neuropathy
 References 77
confocal microscopy as a novel non-invasive tool to [151] Machetta F, Fea AM, Actis AG, de Sanctis U,
investigate cardiac autonomic neuropathy in type 1 Dalmasso P, Grignolo FM. In vivo confocal micro-
diabetes. Diabet Med 2015;32(2):262 6. scopic evaluation of corneal langerhans cells in dry
[140] Ishibashi F, Kojima R, Taniguchi M, Kosaka A, eye patients. Open Ophthalmol J 2014;8:51 9.
Uetake H, Tavakoli M. The Expanded bead size of [152] Su P-Y, Hu F-R, Chen Y-M, Han J-H, Chen W-L.
corneal C-nerve fibers visualized by corneal confocal Dendritiform cells found in central cornea by in-vivo
microscopy is associated with slow conduction veloc- confocal microscopy in a patient with mixed bacterial
ity of the peripheral nerves in patients with type 2 keratitis. Ocul Immunol Inflamm 2006;14:241 4.
diabetes mellitus. J Diabetes Res 2016;7:e3653459. [153] Wu LQ, Cheng JW, Cai JP, Le QH, Ma XY, Gao LD,
[141] Labbé A, Alalwani H, Van Went C, Brasnu E, et al. Observation of corneal Langerhans cells by
Georgescu D, Baudouin C. The relationship between in vivo confocal microscopy in thyroid-associated
subbasal nerve morphology and corneal sensation in ophthalmopathy. Curr Eye Res 2016;41(7):927 32.
ocular surface disease. Invest Ophthalmol Vis Sci [154] Zhivov A, Stave J, Vollmar B, Guthoff R. In vivo con-
2012;53(8):4926 31. focal microscopic evaluation of Langerhans cell den-
[142] del Castillo JMBT, Wasfy MAS, Fernandez C, Garcia- sity and distribution in the normal human corneal
Sanchez J. An in vivo confocal masked study on cor- epithelium. Graefe’s Archive Clin Exp Ophthalmol
neal epithelium and subbasal nerves in patients with 2005;243(10):1056 61.
dry eye. Investig Ophthalmol Vis Sci 2004;45 [155] Tavakoli M, Boulton AJM, Efron N, Malik RA.
(9):3030 5. Increased Langerhan cell density and corneal nerve
[143] Petropoulos IN, Ferdousi M, Marshall A, Alam U, damage in diabetic patients: role of immune mechan-
Ponirakis G, Azmi S, et al. The inferior whorl for isms in human diabetic neuropathy. Cont Lens
detecting diabetic peripheral neuropathy using cor- Anterior Eye 2011;34(1):7 11.
neal confocal microscopy. Investig Ophthalmol Vis [156] Rosenberg ME, Tervo TM, Immonen IJ, Müller LJ,
Sci 2015;56(4):2498 504. Grönhagen-Riska C, Vesaluoma MH. Corneal struc-
[144] Kowtharapu BS, Winter K, Marfurt C, Allgeier S, ture and sensitivity in type 1 diabetes mellitus. Invest
Kohler B, Hovakimyan M, et al. Comparative quanti- Ophthalmol Vis Sci 2000;41(10):2915 21.
tative assessment of the human corneal sub-basal [157] Malik RA, Kallinikos P, Abbott CA, van Schie CHM,
nerve plexus by in vivo confocal microscopy and his- Morgan P, Efron N, et al. Corneal confocal micros-
tological staining. Eye (Lond) 2017;31(3):481 90. copy: a non-invasive surrogate of nerve fibre damage
[145] Brines M, Culver DA, Ferdousi M, Tannemaat MR, and repair in diabetic patients. Diabetologia 2003;46
van Velzen M, Dahan A, et al. Corneal nerve fiber (5):683 8.
size adds utility to the diagnosis and assessment of [158] Midena E, Brugin E, Ghirlando A, Sommavilla M,
therapeutic response in patients with small fiber neu- Avogaro A. Corneal diabetic neuropathy: a confocal
ropathy. Sci Rep 2018;8(1):4734. microscopy study. J Refract Surg 2006;22:1047 52.
[146] Chen X, Graham J, Dabbah MA, Petropoulos IN, [159] Bitirgen G, Turkmen K, Malik RA, Ozkagnici A,
Tavakoli M, Malik RA. An automatic tool for quantifi- Zengin N. Corneal confocal microscopy detects cor-
cation of nerve fibers in corneal confocal microscopy neal nerve damage and increased dendritic cells in
images. IEEE Trans Biomed Eng 2017;64(4):786 94. Fabry disease. Sci Rep 2018;8(1):12244.
[147] Dana MR. Corneal antigen-presenting cells: diversity, [160] Che N-N, Yang H-Q. Potential use of corneal confo-
plasticity, and disguise the Cogan lecture. Investig cal microscopy in the diagnosis of Parkinson’s dis-
Ophthalmol Vis Sci 2004;45(3):722 7. ease associated neuropathy. Transl Neurodegener
[148] Kalogeropoulos D, Papoudou-Bai A, Lane M, 2020;9(1):28.
Goussia A, Charchanti A, Moschos M, et al. Antigen- [161] Cruzat A, Qazi Y, Hamrah P. In vivo confocal
presenting cells in ocular surface diseases. Int microscopy of corneal nerves in health and disease.
Ophthalmol 2020;40:1603 18. Ocul Surf 2017;15(1):15 47.
[149] Alzahrani Y, Colorado LH, Pritchard N, Efron N. [162] Tavakoli M, Marshall A, Banka S, Petropoulos IN,
Longitudinal changes in Langerhans cell density of Fadavi H, Kingston H, et al. Corneal confocal micros-
the cornea and conjunctiva in contact lens-induced copy detects small-fiber neuropathy in Charcot-
dry eye. Clin Exp Optom 2017;100(1):33 40. Marie-Tooth disease type 1A patients. Muscle Nerve
[150] Resch MD, Marsovszky L, Németh J, Bocskai M, 2012;46(5):698 704.
Kovács L, Balog A. Dry eye and corneal langerhans [163] Tavakoli M, Marshall A, Thompson L, Kenny M,
cells in systemic lupus erythematosus. J Ophthalmol Waldek S, Efron N, et al. Corneal confocal micros-
2015;2015:543835. copy: a novel noninvasive means to diagnose

Diabetic Neuropathy
78 4. Advances in screening, early diagnosis, and accurate staging of diabetic neuropathy

neuropathy in patients with fabry disease. Muscle to accumulation of dendritic cells and nerve fiber
Nerve 2009;40(6):976 84. damage of the subbasal nerve plexus in the cornea.
[164] Tavakoli M, Mitu-Pretorian M, Petropoulos IN, Invest Ophthalmol Vis Sci 2014;55(6):3603 15.
Fadavi H, Asghar O, Alam U, et al. Corneal confocal [175] Hamrah P, Zhang Q, Liu Y, Dana MR. Novel charac-
microscopy detects early nerve regeneration in dia- terization of MHC class II-negative population of res-
betic neuropathy after simultaneous pancreas and ident corneal Langerhans cell-type dendritic cells.
kidney transplantation. Diabetes 2013;62(1):254 60. Invest Ophthalmol Vis Sci 2002;43(3):639 46.
[165] Tavakoli M, Ferdousi M, Petropoulos IN, Morris J, [176] Ferdousi M, Romanchuk K, Mah JK, Virtanen H,
Pritchard N, Zhivov A, et al. Normative values for Millar C, Malik RA, et al. Early corneal nerve fibre
corneal nerve morphology assessed using corneal damage and increased Langerhans cell density in
confocal microscopy: a multinational normative data children with type 1 diabetes mellitus. Sci Rep 2019;9
set. Diabetes Care 2015;38(5):838 43. (1):8758.
[166] Halpern EM, Lovblom LE, Orlov S, Ahmed A, Bril V, [177] Basantsova NY, Starshinova AA, Dori A, Zinchenko
Perkins BA. The impact of common variation in the YS, Yablonskiy PK, Shoenfeld Y. Small-fiber neuropa-
definition of diabetic sensorimotor polyneuropathy thy definition, diagnosis, and treatment. Neurol Sci
on the validity of corneal in vivo confocal microscopy 2019;40(7):1343 50.
in patients with type 1 diabetes: a brief report. J [178] Halpern EM, Lovblom LE, Orlov S, Bril V, Perkins
Diabetes Complications 2013;27(3):240 2. BA. editors. The existence of rapid small fiber decline
[167] Pritchard N, Edwards K, Dehghani C, Fadavi H, in type 1 diabetes: a potential proxy for neuropathy
Jeziorska M, Marshall A, et al. Longitudinal assess- progression using in vivo corneal confocal micros-
ment of neuropathy in type 1 diabetes using novel copy. Diabetes; In Proceedings of the twenty-second
ophthalmic markers (LANDMark): study design annual meeting of the diabetic neuropathy study
and baseline characteristics. Diabetes Res Clin Pract group. Dresden 2013.
2014;104(2):248 56. [179] Edwards K, Pritchard N, Dehghani C, Vagenas D,
[168] Edwards K, Pritchard N, Vagenas D, Russell A, Russell A, Malik RA, et al. Corneal confocal micros-
Malik RA, Efron N. Utility of corneal confocal copy best identifies the development and progression
microscopy for assessing mild diabetic neuropathy: of neuropathy in patients with type 1 diabetes. J
baseline findings of the LANDMark study. Clin Exp Diabetes Complications 2017;31(8):1325 7.
Optometry 2012;95(3):348 54. [180] Azmi S, Ferdousi M, Petropoulos IN, Ponirakis G,
[169] Ziegler D, Papanas N, Zhivov A, Allgeier S, Winter K, Alam U, Fadavi H, et al. Corneal confocal micros-
Ziegler I, et al. Early detection of nerve fiber loss by cor- copy identifies small-fiber neuropathy in subjects
neal confocal microscopy and skin biopsy in recently with impaired glucose tolerance who develop type 2
diagnosed type 2 diabetes. Diabetes 2014;63(7):2454 63. diabetes. Diabetes Care 2015;38(8):1502 8.
[170] Asghar O, Petropoulos IN, Alam U, Jones W, [181] Tavakoli M, Kallinikos P, Iqbal A, Herbert A, Fadavi H,
Jeziorska M, Marshall A, et al. Corneal confocal micros- Efron N, et al. Corneal confocal microscopy detects
copy detects neuropathy in subjects with impaired glu- improvement in corneal nerve morphology with an
cose tolerance. Diabetes Care 2014;37(9):2643 6. improvement in risk factors for diabetic neuropathy.
[171] Kawamoto K, Chikama T, Takahashi N, Nishida T. Diabet Med 2011;28(10):1261 7.
In vivo observation of Langerhans cells by laser con- [182] Mehra S, Tavakoli M, Kallinikos PA, Efron N,
focal microscopy in Thygeson’s superficial punctate Boulton AJ, Augustine T, et al. Corneal confocal
keratitis. Mol Vis 2009;15:1456 62. microscopy detects early nerve regeneration after
[172] Dauch JR, Bender DE, Luna-Wong LA, Hsieh W, pancreas transplantation in patients with type 1 dia-
Yanik BM, Kelly ZA, et al. Neurogenic factor- betes. Diabetes Care 2007;30(10):2608 12.
induced Langerhans cell activation in diabetic mice [183] Lewis EJH, Perkins BA, Lovblom LE, Bazinet RP,
with mechanical allodynia. J Neuroinflammation Wolever TMS, Bril V. Effect of omega-3 supplementa-
2013;10:64. tion on neuropathy in type 1 diabetes: a 12-month
[173] Davidson EP, Coppey LJ, Holmes A, Lupachyk S, pilot trial. Neurology 2017;88(24):2294 301.
Dake BL, Oltman CL, et al. Characterization of dia- [184] Brines M, Dunne AN, van Velzen M, Proto PL,
betic neuropathy in the Zucker diabetic Sprague- Ostenson C-G, Kirk RI, et al. ARA 290, a nonerythro-
Dawley rat: a new animal model for type 2 diabetes. poietic peptide engineered from erythropoietin,
J Diabetes Res 2014;e714273. improves metabolic control and neuropathic symp-
[174] Leppin K, Behrendt AK, Reichard M, Stachs O, toms in patients with type 2 diabetes. Mol Med
Guthoff RF, Baltrusch S, et al. Diabetes mellitus leads (Cambridge, Mass) 2015;20(1):658 66.

Diabetic Neuropathy
 C H A P T E R

5
Skin biopsy analysis in diabetic
neuropathy
Páll Karlsson
Danish Pain Research Center, Department of Clinical Medicine, Aarhus University, Aarhus, Denmark

Introduction (small fiber) neuropathy. In fact, DN was one

of the first diseases to be shown to have
Nociceptor cell bodies are located in the decreased IENFD as shown with skin biopsy
trigeminal ganglion or the dorsal root ganglion, [2]. The skin biopsy method is only suitable to
and their nerve endings terminate in the skin. assess unmyelinated or thinly myelinated nerve
These nerve fibers are present in nerve bundles fibers. To detect damage to the myelinated fibers,
in the subepidermal neural plexus and some neurophysiological examinations such as nerve
of the free nerve endings penetrate the conduction studies (NCS) should be used, which,
epidermal dermal layer. Damage to the nerve in turn, is unable to detect damage or loss of
fibers due to diabetes may give rise to symp- unmyelinated fibers.
toms or signs of diabetic neuropathy (DN). The The presence of nociceptor nerve endings in
damage manifests typically as a decrease of the the epidermis has been known for over 150 years,
unmyelinated or thinly myelinated nociceptive but they were first described in 1868 [3]. The first
nerve fibers in the skin, but other morphologi- study to visualize the epidermal nerve fibers was
cal abnormalities can also be seen. This loss is published in 1989, where the authors collected
quantified by taking a small punch skin biopsy skin samples from themselves and immunos-
(typically 3 mm in diameter) and by estimating tained frozen sections using protein gene product
the density of nerve fibers penetrating the epi- 9.5 (PGP 9.5) that is a protein expressed on the
dermal layer (the topmost layer of the skin [1]), nerves [4]. It was, however, only in this century it
frequently referred to as intraepidermal nerve became possible to easily and reliably immunos-
fiber density (IENFD). This morphometric tain the nerve fibers [5,6]. Since then, the staining
analysis and the decrease of IENFD are often protocol has been optimized, general counting
considered being the pathological gold stan- rules have been established, and the skin biopsy
dard of small fiber damage when diagnosing technique and the quantification of IENFD have
DN, as it is the method that offers the most been introduced in clinical practice, with norma-
reliable, quantitative, and objective analysis of tive reference values available at the distal leg,

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00013-X 79 © 2022 Elsevier Inc. All rights reserved.
80 5. Skin biopsy analysis in diabetic neuropathy

both for bright field and indirect immunofluores- alterations that can be assessed and quantified,
cence microscopy [7 11]. including axonal swellings or branching den-
While the skin biopsy is considered mini- sity [14 19]. Skin biopsies and especially
mally invasive, there is a less invasive alterna- IENFD analysis have been used to follow effi-
tive to skin biopsies to assess the IENFD. The cacy of interventions by taking repeated biop-
“skin blister” method, where negative pressure sies, for example, during lifestyle changes [20].
vacuum is used to form a blister on the skin, However, there are a number of disadvantages
can be performed to separate the epidermis to the skin biopsies and IENFD analysis. These
from the rest of the skin [12]. Studies have include a low correlation between IENFD and
shown that there is a minimal difference of the other (more subjective) test results, such as clini-
IENFD values between the two methods and a cal symptoms reported by the patients or mea-
high correlation between them [13]. However, sures from quantitative sensory testing (QST). In
while the skin blister method is less invasive, addition, IENFD analysis is unable to determine
there are some disadvantages, including that it the etiology of the neuropathy [21 23]. Low
is time consuming as it can take up to 90 min- IENFD is seen in different etiologies of neuropa-
utes, and because it is superficial in nature, it is thy, possibly due to a final pathway that is
not possible to perform morphological analysis shared by all neuropathies regardless of the
of the deeper skin tissue, for example, the der- underlying disease, or that we are yet to discover
mis. Therefore, skin blisters are not frequently a suitable marker capable of discriminating
used, neither in clinical practice nor for between different etiologies [24]. The lack of cor-
research purposes. relation between IENFD and other measures
such as symptoms or QST may either be due to
the fact that while IENFD is an objective measure,
Advantages and disadvantages of skin symptom-reporting is highly subjective and QST
biopsies is psychophysical in nature, or because IENFD
analysis only quantifies damage in the peripheral
Taking a skin biopsy is a process that only nervous system and not in the central nervous
takes a few minutes, it is minimally invasive system, which also plays an important role in
with the resulting wound heals in 7 14 days, sensory and pain perception.
the procedure can be repeated to follow dis- The main disadvantages of the skin biopsies
ease progression over time, and gives repro- are perhaps that immunohistological stainings
ducable and reliable results. The procedure is are difficult to standardize, and processing and
therefore being used more and more as a diag- staining protocols are heterogeneous. This means
nostic tool in the clinic. Additional advantage that fixation methods, staining protocols, and
of skin biopsies is the variability of different section thickness often vary between laboratories,
immunostainings and information that can be making it difficult to compare results between
extracted from them. While quantification of laboratories. Quantification of IENFD or other
IENFD is the gold standard, given that reliable measures is also subjective in nature and the
antibodies are reliable, there are no limitations quality of the quantification depends both on
of alternative immunostainings that can be per- the quality of the tissue and the staining, and on
formed. Some of the analyses may hold the the person performing the analysis. It is therefore
potential to explain why some patients with important that the entire process is as standard-
diabetes develop symptoms of DN, such as ized as possible. One way to ensure this is to fol-
sensory dysfunction or neuropathic pain. low a published and recommended guideline
There are also a number of morphological from the European Federation of Neurological

Diabetic Neuropathy
 How and where to take a skin punch biopsy following EFNS/PNS guidelines 81
Societies (EFNS) and the Peripheral Nerve alcohol-induced neuropathy [17,34]), in fibro-
Society (PNS) [7]. This is especially important for myalgia [35 38] and in idiopathic cases with
diagnostic purposes and when comparing results no identifiable cause [39,40]. In rare cases, a
from patients with already available normative sural or peroneal nerve biopsy can be acquired.
reference values obtained following the guide- The nerve biopsy was often used as a diagnos-
lines. For research purposes, there can be various tic tool before the skin biopsy became widely
reasons for deviating from the guidelines. These available, but is now only used if it adds
should be described and the findings should not needed information and other tests are unable
be compared with results obtained using other to provide, for example, in amyloidosis, atypi-
protocols. Ideally, each skin biopsy laboratory cal chronic inflammatory neuropathies and in
should have their own dataset with normative nonsystemic vasculitis, where nerve biopsy is
reference values for both sexes and each age, the only way to make a definite diagnosis
divided by decade, as IENFD decreases with [41 43].
older age and male sex.

How and where to take a skin punch

IENFD quantification in nondiabetic biopsy following EFNS/PNS guidelines
etiologies
Skin biopsies can be taken from practically
Quantification of IENFD can be used to anywhere from the body, for example, where
assess nerve fiber damage in neuropathies of the patient complains about pain or other
nondiabetic etiologies as well. Examples symptoms. For diagnostic purposes and in
include hereditary neuropathies (e.g., Fabry majority of research studies, the gold standard
disease [25 27] and hereditary sensory neu- is to take a skin biopsy from the distal leg,
ropathy type 1 [28]), autoimmune neuropathies 10 cm above the lateral malleolus (see Fig. 5.1).
(e.g. Sjögren’s syndrome [29]), entrapment neu- Here we describe where the biopsies should
ropathies (e.g., carpal tunnel syndrome [30]), be taken and how they should be processed
neuropathy of infectious (e.g., HIV [31 33]), or when following the recommended EFNS/PNS
toxic causes (e.g., following chemotherapy or guidelines [7].

FIGURE 5.1 The skin biopsy is typically taken

from the distal leg, 10 cm above the lateral
malleolus.

Diabetic Neuropathy
82 5. Skin biopsy analysis in diabetic neuropathy

The biopsy should be taken following local is more frequently used, especially in the clinic.
anesthesia using 0.5 1.0 mL lidocaine, under ster- Immunofluorescence staining is typically used
ile conditions using 3 5 mm disposable biopsy when performing more advanced stainings, for
punch. The biopsy should be immediately trans- example, when using other antibodies than PGP
ferred to a fixative. There are several types of fixa- 9.5 for IENFD analysis and when performing
tives available, but both the type of fixative and double- or triple stainings. Below is a description
the incubation time may influence the subsequent of a commonly used immunohistochemistry pro-
immunohistological staining. The biopsy should tocol for staining of skin biopsy sections for
be fixed in either Zamboni’s fixative (2% parafor- IENFD analysis.
maldehyde, picric acid) or using freshly made 2%
periodate-lysine-paraformaldehyde. Very impor-
tantly, the biopsy should be fixed over-night at
cold temperature (4 C), but the fixation should not Staining protocol for IENFD analysis
exceed 24 hours to avoid over-fixation, which will
influence the subsequent immunohistological The first step of the staining protocol is at
stainings. After fixation, the biopsy should be least 1 hour of blocking incubation using
washed in either neutral buffer like PBS or in cryo- Tritron X-100, powdered milk, and normal
protectant for 2 3 5 minutes and then stored at serum made in the same animal as the second-
cold temperature in cryoprotectant over night. ary antibody that will be used later in the pro-
tocol. After three times wash, the skin sections
are incubated in PGP 9.5 as primary antibody
Freezing the biopsy overnight on a slow-rotating shaker table at
4 C. Following another round of three times
The biopsy is now ready to be frozen down. wash, the skin sections are incubated in sec-
Freezing methods vary, but many laboratories ondary antibody for 1 hour at room tempera-
use a cryostat. The biopsy should be frozen ture. The antibodies can be diluted in a weak
down submerged in optimal cutting temperature blocking buffer. The sections are washed three
compound lying on the side so both epidermis times and then submerged in a solution con-
and dermis are facing upward (see Fig. 5.2) and taining 33.3% methanol and 3.3% hydrogen
then stored at 220 C for a short-term storage peroxide diluted in PBS for 30 minutes. After
(,1 year) or 280 C ( . 1 year). The biopsy can another round of wash, the biopsies are incu-
be sectioned using a cryostat where 50 µm thick bated in avidin biotin complex for 1 hour and
sections are cut so that both epidermis and der- lastly, the nerve fibers are visualized using a
mis are cut. At least three sections should be cut peroxidase substrate kit until desired darkness
and they transferred directly to a neutral buffer has been reached. How long time that takes
(e.g., PBS) in a tissue culture plate allowing a depends on the concentrations of the primary
free-floating staining, ensuring that both sides of and secondary antibodies. It is important that
the sections are exposed to the reagents used the incubation time is long enough to visualize
during the staining protocol. This is important as even the thinnest nerve fibers, but not too long
the section thickness is too thick for a fixed stain- to avoid high background staining that may
ing on slides with a risk of incomplete penetra- cover the fibers and make them difficult to
tion. The sections can either be immunostained count or even make them invisible. The sec-
using immunohistochemistry or immunofluores- tions are submerged in gelatine solution and
cence. While there now exist normative reference lined on a slide to air dry. The last step of the
values for both methods, immunohistochemistry staining protocol is to briefly counterstaine the

Diabetic Neuropathy
 Staining protocol for IENFD analysis 83

(A) FIGURE 5.2 Schematic illus-

tration of the most commonly
used biopsy site for diagnostic
and research purposes (A). The 3-
mm punch biopsy is taken from
the distal leg, 10 cm proximal
from the lateral malleolus. Upon
fixation and cryoprotection, the
biopsy should be frozen sub-
merged in optical cutting temper-
ature (OCT) lying down, so that
the 50-µm thick sections will be
cut perpendicular to the direction
of the epidermis (A). Once the sec-
tions have been immunostained
and PGP 9.5-positive nerve fibers
visualized, only fibers crossing the
epidermal layer from the dermis
are counted (B). Only nerve fibers
that branch in the epidermis and
cross the epidermal-dermal border
are counted (c, d), while fibers
that branch within the epidermis
are not (b). Similarly, nerve fibers
that are only visible in one skin
layer are not counted (g, e).

(B)

1 0
2 2
1 1 0

a b c d e f g

Diabetic Neuropathy
84 5. Skin biopsy analysis in diabetic neuropathy

sections (e.g. in eosin), dehydrate them and clinical DN criteria, for example, the one pro-
then coverslip. posed by the American Diabetes Association
(ADA), which is “the presence of symptoms
and/or signs of peripheral nerve dysfunction
IENFD quantification in people with diabetes after the exclusion of
other causes” [44]. Another often used classifi-
The IENFD quntification should be performed cation is the Toronto classification from 2010,
at high magnification using high power objective which divides DN into three levels of certainty
of at least 40X under live microscopy, to ensure of the diagnosis: possible, probable, and defi-
that all unmyelinated nerve fibers are clearly visi- nite [45].
ble. Quantification under lower magnification Quantitative information on nerve fiber
increases the risk of not seeing all nerve fibers damage is not needed in clinical practice in
entering the epidermis. The EFNS/PNS guide- cases where there is no doubt about the diag-
lines also include a set of “counting rules” to nosis. Neither skin biopsy nor NCS can con-
ensure uniformity between laboratories when firm or identify the underlying cause of the
counting. As the name suggests, only dermal neuropathy and nor do they influence treat-
nerve fibers that cross the epidermal dermal ment strategies [46]. Even so, IENFD is the
boarder should be counted. This means that fibers pathologic gold standard of small nerve fiber
that are only visible in either the dermis or the damage in DN [42] and is an important diag-
epidermis should not be counted (see Fig. 5.2). nostic tool in cases where there is uncertainty
Because the section thickness is 50 µm, which is about the neuropathy-diagnosis. IENFD analy-
relatively thick, it is important to analyze the sis has generally been shown to have good
entire thickness (Z-axis) of the section, which diagnostic sensitivity and specificity in DN,
requires focusing up and down through the entire with good positive and negative predictive
section thickness. The nerve fibers are so thin that values [47]. Using the Toronto definition of
the individual fibers are visible at different depths probable neuropathy, area under the curve
and not all fibers are visible at the same time or analysis for IENFD to detect neuropathy has
depth. Nerve fibers that branch from another been reported to be 0.76, and slightly higher
nerve are also counted, but only if the branching for SFN or 0.82 [46,48]. However, a study that
point is located in the dermis and the branching estimated the diagnostic utility of IENFD anal-
fiber crosses the epidermal dermal boarder (see ysis and NCS in 366 patients with diabetes
Fig. 5.2). The total number of counted nerve fibers using several gold-standard definitions to
is divided with the length of the outer skin layer determine the sensitivity and specificity [49]
(measured in millimeters) to estimate the IENFD found that IENFD analysis had a sensitivity of
(expressed as number of fibers per mm). It is 81% but a specificity of 41%, while NCS had
recommended to estimate the IENFD in at least lower sensitivity of 40% but higher specificity
three sections and to take the mean to obtain a of 95%. It is unclear if small fiber damage pre-
more precise estimation. cedes large fiber damage, but that hypothesis
may explain these differences. Devigili and col-
leagues compared the diagnostic utility of skin
Diagnostic accuracy of IENFD analysis biopsy, QST, and clinical examination for neu-
in DN ropathies, and found that IENFD analysis out-
ranked other tests with both sensitivity and
Skin biopsies, just like NCS, are not often specificity rates of around 90%, in patients
needed in patients with diabetes who meet the with mixed etiologies [50]. Similar comparisons

Diabetic Neuropathy
 Relationship between skin biopsy outcomes and symptoms or signs of DN 85
have been made between IENFD analysis from and transient receptor potential cation channel
skin biopsy and corneal confocal microscopy subfamily V member 1 (TRPV1) at both time
(CCM), where the tests showed strong and points and that the annual rate of IENF loss
comparative findings. IENFD analysis and was 23.76 fibers/mm (61.46 fibers/mm).
CCM are therefore promising methods, albeit
bigger and more robust longitudinal studies
are needed to truly reveal their clinical useful- Relationship between skin biopsy
ness [51 54]. The diagnosis of DN cannot, in outcomes and symptoms or signs of DN
other words, rely on a single test, but on clini-
cal judgment using several tests including clin- As mentioned earlier, there is only a low
ical history and examination. Apart from the association between IENFD and presence of
usual biopsy site at the distal leg, an additional pain and especially pain intensity. A system-
biopsy can be taken from the lateral proximal atic review from 2019 looked into this relation-
thigh (7 10 cm below the greater trochanter) ship, and noted that of all neuropathy
to determine if the nerve damage is length- etiologies, DN is the most studied entity or a
dependent or nonlength-dependent [55,56]. total of 40 studies at that time [23]. The review
DN most often presents as length-dependent included studies who reported at least one
where the distal leg biopsy shows lower finding from skin biopsy and at least one addi-
IENFD compared with the biopsy from the lat- tional parameter related to symptom or sign of
eral proximal thigh. neuropathy or nerve fiber function. The review
One advantage of skin biopsies is that they reported that only eight out of 18 of the
can be repeated over time, for example, to fol- included studies that reported correlation anal-
low disease progression. Even so, not many ysis between IENFD and patient-reported pain
longitudinal studies have been performed with found a significant inverse association, for
IENFD analysis as an outcome. A study in example, lower IENFD with higher pain inten-
2016 compared the longitudinal rate of IENFD sity. Of these eight studies, there were three
decline in healthy controls, in DN, patients studies that reported association between
with impaired glucose tolerance and in IENFD and patient-reported pain, one of which
patients with idiopathic small fiber neuropathy found a significant negative association. In
[40]. The study participants were evaluated at addition, less than one-third of the studies
baseline and again 24 39 months later. The found a significant association between IENFD
study found that the IENFD was reduced in all and pain intensity. It seems therefore unlikely
three neuropathy groups at all three biopsy that low IENFD determines the presence or
sites along the leg, but not in healthy controls. severity of painful symptoms in DN, even if it
The yearly rate of IENFD decline was esti- is a key pathological finding. In terms of asso-
mated to be 21.42, 21.59, and 22.8 fibers/mm ciation between IENFD and QST measures, the
for the distal leg, distal thigh and proximal review reported that 22 out of 29 studies (76%)
thigh, respectively, for the patient groups. A found significant association between IENFD
similar study from 2012 took biopsies from DN and at least one QST parameter. However,
patients at baseline and again after 6 months to when solely looking at studies that included
identify biomarkers to follow disease progres- DN or impaired glucose tolerance patients, all
sion and, in future studies, to assess efficacy of eight studies found significant association
disease-modifying agents [57]. The study between the two parameters, hereof five that
found that intraepidermal and subepidermal found significant association between IENFD
nerve fibers were reduced in DN for PGP 9.5 and all tested QST parameters. At the same

Diabetic Neuropathy
86 5. Skin biopsy analysis in diabetic neuropathy

time, only 21% of studies of non-DN etiology swellings and type 2 diabetes, neuropathy and
found the same association. These findings pain. The study found that the swellings were
indicate that the etiology may be an important related to type 2 diabetes rather than neuropa-
factor when examining the role of the IENFs of thy or pain, and postulated that the swellings
sensory dysfunction or pain, albeit other may be an early indicator of a disease affecting
mechanisms play an important role as well, for the small nerve fibers and may predict future
example, the fact that QST is a cognitive com- nerve fiber loss [65]. Analysis of axonal swell-
ponent and central mechanisms of sensitization ing ratio may even increase the diagnostic per-
also likely play a role in nociception. formance of skin biopsies [16,65].

Other morphological changes Alternative markers used on skin biopsies

The estimation of intraepidermal nerve fiber One of the advantages that skin biopsies have
density can be considered being relatively to offer is the amount of additional information
crude and simple measure that only takes into that can be extracted from them. It is possible to
account the number of nerve fibers penetrating use virtually any specific, reliable, and relevant
the epidermal dermal boarder. Several studies antibody on them to get quantitative and mor-
have quantified other morphological changes phological information about different subpopu-
in DN, including epidermal and dermal nerve lations of nerve fibers as well as other cells and
fiber length density (NFLD) using stereology structures that are in the skin. A few examples of
that may improve the diagnostic accuracy these analyses will be discussed below.
[17,58,59], quantification of nerve fibers in dif-
ferent skin layers that are effected in DN
[57,60], and branching density that correlates Biomarkers of pain pathophysiology
with severity of DN [61]. Another morphologi-
cal change that has been studied is axonal As already discussed, there is a weak associa-
swellings—enlargements that are visible on tion between IENFD and presence of pain or pain
some of the nerve fibers, and are typically severity. It is, therefore, of interest to identify “bio-
expressed as a swelling ratio (number of swel- markers of pain pathophysiology,” or a marker
lings per IENFD) [14,58,62,63]. Axonal swel- that can readily distinguish between painful and
lings are believed to represent degenerative nonpainful DN that would hold a potential to be
changes on the nerve fibers and have been a novel treatment target for painful DN. The noci-
shown to contain watery axoplasm, neurofila- ceptors in the skin can be divided into two groups
ments, and abnormal mitochondria [16]. Many according to peptide expression. They are either
studies have shown that axonal swelling ratio peptidergic fibers that release calcitonin gene-
is higher in patients with neuropathy com- related peptide or substance P, or they are non-
pared with those without neuropathy peptidergic fibers expressing receptor tyrosine
[14 16,64], including patients with DN kinase [66]. Recent studies have shown that
[62,63,65]. The clinical significance of these patients with painful DN have increased pepti-
swellings is unclear, as they have both been dergic fiber density compared with healthy parti-
associated with pain [14,62] and to have no cipants without diabetes and diabetes patients
association with pain [63,65]. A recent study without neuropathy [67,68]. Furthermore, the pep-
including a total of 249 participants looked fur- tidergic correlated with pain intensity while the
ther into the association between axonal IENFD did not. Similarly, growth-associated

Diabetic Neuropathy
 Conclusion 87
protein 43, a marker of regenerating nerve allow for assessment of autonomic structures
sprouts, has been shown to be higher in patients in the skin, such as vascular, sudomotor, vaso-
who describe their pain as burning [69]. motor, and pilomotor nerve fibers. It is there-
Biomarkers identifying other structures than fore possible to perform density analysis of
nerve fibers may also be of interest. Immune cells specific subpopulations of autonomic nerve
such as macrophages have been shown to be fibers, nerve fiber innervation in sweat glands
increased in the skin of patients with painful neu- and their density and morphology, and arrec-
ropathies because of diabetes and chemotherapy, tor pili muscle, responsible for hair erection
compared with patients without pain and healthy upon contraction. The vast majority of sudo-
controls [70,68]. Another interesting pain-related motor fibers (cholinergic fibers) can be visual-
structure that can be quantified using skin biop- ized using the vasoactive intestinal peptide
sies is the “nociceptive” Schwann cell. In a recent antibody while pilomotor fibers (e.g., arrector
publication in science, an until-now-unknown pili muscle, predominantly noradrenergic) can
organ likely playing an essential physiological be visualized using dopamine beta hydroxy-
role in sensing painful stimuli was identified in lase antibody. Gibbons and colleagues have
animal model of pain [71]. The paper describes shown a significant decrease of sudomotor
specialized cutaneous Schwann cells that initiate innervation in patients with diabetes compared
pain sensation and surround themselves around with healthy controls, including sweat gland
the nerves in the skin and feed them with noci- nerve fiber density (SGNFD), and have com-
ceptive information. A recent paper by the same pared different methods to quantify this mea-
group followed up on the initial work and sure [77,78]. The study showed that there was
showed that these cells are also present in the a significant and relatively strong correlation
human skin [72]. However, they still remain to be between IENFD and SGNFD and that patients
quantified and compared between patients with with more severe DN also had lower SGNFD.
and without painful DN and it is unclear what Although not studied in DN, detailed assess-
role they play in the human nociception. Taken ment of the autonomic innervation in skin
together, these markers show promise and may biopsies taken from patients with Parkinson’s
represent novel treatment targets for painful DN. disease has shown abnormalities in autonomic
Several other markers have been studied and nerves to vessels, sweat glands, and arrector
compared between patients with and without pili muscle, and even Meissner corpuscles in
DN or painful DN, including the TRPV1 that is glabrous skin [79]. There is a need to further
responsible for heat sensation greater than 43 C assess the autonomic structures in the skin in
[39,57,73], growth-associated protein 43 (GAP-43) DN, and combining this analysis with IENFD
that is believed to be a marker of regenerating estimation may even increase the specificity of
nerve sprouts and has been associated with burn- IENFD in diagnosing DN.
ing pain sensation [57,62,68,74], and Langerhans
cells that also have been implicated with painful
symptoms in painful DN [75,76]. Conclusion
Skin biopsies allow for a quantification of
Autonomic structures IENFD and assessment of nociceptive damage
in DN. They are, however, not needed in cer-
A subset of DN patients has autonomic neu- tain clinical cases as they do not change the
ropathy and may have symptoms such as treatment plan. The skin biopsies represent an
reduced sweat function. The skin biopsies also important research tool due to their ability to

Diabetic Neuropathy
88 5. Skin biopsy analysis in diabetic neuropathy

detect quantifiable molecular changes that [12] Kennedy WR, et al. A skin blister method to study epi-
increase our knowledge of why some, but not dermal nerves in peripheral nerve disease. Muscle
Nerve 1999;22(3):360 71.
all, DN patients develop neuropathic pain and [13] Panoutsopoulou IG, et al. Skin blister and skin biopsy
may even contribute to the development of to quantify epidermal nerves: a comparative study.
mechanism-based treatment. Disadvantages of Neurology 2009;72(14):1205 10.
skin biopsies include that they require special- [14] Lauria G, et al. Axonal swellings predict the degenera-
ized and trained staff, they are invasive, and tion of epidermal nerve fibers in painful neuropathies.
Neurology 2003;61(5):631 6.
the quality of the analysis differs between labo- [15] Gibbons CH, et al. The utility of skin biopsy for pre-
ratories and is dependent on staining quality. diction of progression in suspected small fiber neurop-
athy. Neurology 2006;66(2):256 8.
[16] Ebenezer GJ, et al. Denervation of skin in neuropathies:
the sequence of axonal and Schwann cell changes in skin
References biopsies. Brain 2007;130(Pt 10):2703 14.
[17] Karlsson P, et al. Structural and functional characteri-
[1] Lauria G, Devigili G. Skin biopsy as a diagnostic tool zation of nerve fibres in polyneuropathy and healthy
in peripheral neuropathy. Nat Clin Pract Neurol subjects. Scand J Pain 2016;10:28 35.
2007;3(10):546 57. [18] Isak B, et al. Quantitative sensory testing and struc-
[2] Levy DM, et al. Immunohistochemical measurements tural assessment of sensory nerve fibres in amyo-
of nerves and neuropeptides in diabetic skin: relation- trophic lateral sclerosis. J Neurol Sci 2017;373:329 34.
ship to tests of neurological function. Diabetologia [19] Andersen ST, et al. Corneal confocal microscopy as a
1992;35(9):889 97. tool for detecting diabetic polyneuropathy in a cohort
[3] Langerhans P. Uber die nerven der menschlichen with screen-detected type 2 diabetes: ADDITION-
haut. Arch Patholog Anat 1868;44:325 37. Denmark. J Diabetes Complicat 2018;32(12):1153 9.
[4] Dalsgaard CJ, Rydh M, Haegerstrand A. Cutaneous inner- [20] Smith AG, et al. Lifestyle intervention for pre-diabetic
vation in man visualized with protein gene product 9.5 neuropathy. Diabetes Care 2006;29(6):1294 9.
(PGP 9.5) antibodies. Histochemistry 1989;92(5):385 90. [21] Koskinen MJ, et al. Intraepidermal nerve fibre density
[5] Sommer C, Lauria G. Skin biopsy in the management in cancer patients receiving adjuvant chemotherapy.
of peripheral neuropathy. Lancet Neurol 2007;6 Anticancer Res 2011;31(12):4413 16.
(7):632 42. [22] Karlsson P, et al. Structural and functional assessment
[6] Polydefkis M, et al. Skin biopsy as a tool to assess dis- of skin nerve fibres in small-fibre pathology. Eur J
tal small fiber innervation in diabetic neuropathy. Pain 2015;19(8):1059 70.
Diabetes Technol Ther 2001;3(1):23 8. [23] Karlsson P, et al. Structural, functional, and symptom
[7] Lauria G, et al. European Federation of Neurological relations in painful distal symmetric polyneuropa-
Societies/Peripheral Nerve Society Guideline on the thies: a systematic review. Pain 2019;160(2):286 97.
use of skin biopsy in the diagnosis of small fiber neu- [24] Myers MI, Peltier AC, Li J. Evaluating dermal myelin-
ropathy. Report of a joint task force of the European ated nerve fibers in skin biopsy. Muscle Nerve 2013;47
Federation of Neurological Societies and the (1):1 11.
Peripheral Nerve Society. Eur J Neurol 2010;17 [25] Moller AT, et al. Autonomic skin responses in females
(7):903 12 e44-9. with Fabry disease. J Peripher Nerv Syst 2009;14
[8] Wang WJ. Clinical features of cutaneous amyloidoses. (3):159 64.
Clin Dermatol 1990;8(2):13 19. [26] Uceyler N, et al. Impaired small fiber conduction in
[9] Kennedy WR, Wendelschafer-Crabb G. The innerva- patients with Fabry disease: a neurophysiological
tion of human epidermis. J Neurol Sci 1993;115 case-control study. BMC Neurol 2013;13:47.
(2):184 90. [27] Liguori R, et al. Small fiber neuropathy in female patients
[10] McCarthy BG, et al. Cutaneous innervation in sensory with fabry disease. Muscle Nerve 2010;41(3):409 12.
neuropathies: evaluation by skin biopsy. Neurology [28] Fridman V, et al. Natural history and biomarkers in
1995;45(10):1848 55. hereditary sensory neuropathy type 1. Muscle Nerve
[11] Provitera V, et al. A multi-center, multinational age- 2015;51(4):489 95.
and gender-adjusted normative dataset for immuno- [29] Sene D, et al. Sjogren syndrome-associated small fiber
fluorescent intraepidermal nerve fiber density at the neuropathy: characterization from a prospective series
distal leg. Eur J Neurol 2016;23(2):333 8. of 40 cases. Med (Baltim) 2013;92(5):e10 18.

Diabetic Neuropathy
 References 89
[30] Schmid AB, et al. The relationship of nerve fibre [47] Gwathmey KG, Pearson KT. Diagnosis and manage-
pathology to sensory function in entrapment neuropa- ment of sensory polyneuropathy. BMJ 2019;365:l1108.
thy. Brain 2014;137(Pt 12):3186 99. [48] Callaghan BC, et al. Better diagnostic accuracy of neu-
[31] Obermann M, et al. Correlation of epidermal nerve ropathy in obesity: a new challenge for neurologists.
fiber density with pain-related evoked potentials in Clin Neurophysiol 2018;129(3):654 62.
HIV neuropathy. Pain 2008;138(1):79 86. [49] Smith, AG,Lessard, M, andSingleton, JR. The diagnos-
[32] Polydefkis M. Skin biopsy findings predict develop- tic utility of nerve conduction studies and skin biopsy
ment of symptomatic neuropathy in patients with for diabetic neuropathy (DPN): a Bayesian analysis.
HIV. Nat Clin Pract Neurol 2006;2(12):650 1. 2015 [cited 2021 January 14]. Available from https://
[33] Polydefkis M, et al. Reduced intraepidermal nerve n.neurology.org/content/84/14_Supplement/S42.007.
fiber density in HIV-associated sensory neuropathy. [50] Devigili G, et al. The diagnostic criteria for small fibre
Neurology 2002;58(1):115 19. neuropathy: from symptoms to neuropathology. Brain
[34] Kroigard T, et al. Characterization and diagnostic 2008;131(Pt 7):1912 25.
evaluation of chronic polyneuropathies induced by [51] Schaldemose EL, et al. An unbiased stereological method
oxaliplatin and docetaxel comparing skin biopsy to for corneal confocal microscopy in patients with diabetic
quantitative sensory testing and nerve conduction polyneuropathy. Sci Rep 2020;10(1):12550.
studies. Eur J Neurol 2014;21(4):623 9. [52] Schaldemose EL, et al. Improved sampling and analy-
[35] Kosmidis ML, et al. Reduction of intraepidermal nerve sis of images in corneal confocal microscopy. J
fiber density (IENFD) in the skin biopsies of patients Microsc 2017;268(1):3 12.
with fibromyalgia: a controlled study. J Neurol Sci [53] Alam U, et al. Diagnostic utility of corneal confocal
2014;347(1-2):143 7. microscopy and intra-epidermal nerve fibre density in
[36] Giannoccaro MP, et al. Small nerve fiber involvement diabetic neuropathy. PLoS One 2017;12(7):e0180175.
in patients referred for fibromyalgia. Muscle Nerve [54] Chen X, et al. Small nerve fiber quantification in the
2014;49(5):757 9. diagnosis of diabetic sensorimotor polyneuropathy:
[37] Caro XJ, Winter EF. Evidence of abnormal epidermal comparing corneal confocal microscopy with intraepi-
nerve fiber density in fibromyalgia: clinical and immuno- dermal nerve fiber density. Diabetes Care 2015;38
logic implications. Arthritis Rheumatol 2014;66(7):1945 54. (6):1138 44.
[38] Uceyler N, et al. Small fibre pathology in patients with [55] Provitera V, et al. The role of skin biopsy in differenti-
fibromyalgia syndrome. Brain 2013;136(Pt 6):1857 67. ating small-fiber neuropathy from ganglionopathy.
[39] Wilder-Smith EP, et al. Epidermal transient receptor Eur J Neurol 2018;25(6):848 53.
potential vanilloid 1 in idiopathic small nerve fibre [56] Zhou L. Small fiber neuropathy. Semin Neurol 2019;39
disease, diabetic neuropathy and healthy human sub- (5):570 7.
jects. Histopathology 2007;51(5):674 80. [57] Narayanaswamy H, et al. A longitudinal study of sen-
[40] Khoshnoodi MA, et al. Longitudinal assessment of sory biomarkers of progression in patients with dia-
small fiber neuropathy: evidence of a non-length- betic peripheral neuropathy using skin biopsies. J Clin
dependent distal axonopathy. JAMA Neurol 2016;73 Neurosci 2012;19(11):1490 6.
(6):684 90. [58] Karlsson P, et al. Epidermal nerve fiber length density
[41] Uceyler N, et al. Non-systemic vasculitic neuropathy: estimation using global spatial sampling in healthy
single-center follow-up of 60 patients. J Neurol subjects and neuropathy patients. J Neuropathol Exp
2015;262(9):2092 100. Neurol 2013;72(3):186 93.
[42] Terkelsen AJ, et al. The diagnostic challenge of small [59] Karlsson P, et al. Dermal innervation in healthy sub-
fibre neuropathy: clinical presentations, evaluations, jects and small fiber neuropathy patients: a stereologi-
and causes. Lancet Neurol 2017;16(11):934 44. cal reappraisal. J Peripher Nerv Syst 2013;18(1):48 53.
[43] Sommer C. Nerve and skin biopsy in neuropathies. [60] Vlckova-Moravcova E, et al. Diagnostic validity of epi-
Curr Opin Neurol 2018;31(5):534 40. dermal nerve fiber densities in painful sensory neu-
[44] Pop-Busui R, et al. Diabetic neuropathy: a position ropathies. Muscle Nerve 2008;37(1):50 60.
statement by the American Diabetes Association. [61] Quattrini C, et al. Surrogate markers of small fiber
Diabetes Care 2017;40(1):136 54. damage in human diabetic neuropathy. Diabetes
[45] Tesfaye S, et al. Diabetic neuropathies: update on defi- 2007;56(8):2148 54.
nitions, diagnostic criteria, estimation of severity, and [62] Cheng HT, et al. Increased axonal regeneration and
treatments. Diabetes Care 2010;33(10):2285 93. swellings in intraepidermal nerve fibers characterize
[46] Callaghan BC, et al. Diabetic neuropathy: what does painful phenotypes of diabetic neuropathy. J Pain
the future hold? Diabetologia 2020;63(5):891 7. 2013;14(9):941 7.

Diabetic Neuropathy
90 5. Skin biopsy analysis in diabetic neuropathy

[63] Cheung A, et al. Epidermal axonal swellings in pain- [71] Abdo H, et al. Specialized cutaneous Schwann cells
ful and painless diabetic peripheral neuropathy. initiate pain sensation. Science 2019;365(6454):695 9.
Muscle Nerve 2015;51(4):505 13. [72] Puneet R, et al. Demise of nociceptive Schwann cells
[64] Herrmann DN, et al. Epidermal nerve fiber density, causes nerve retraction and pain hyperalgesia. Pain
axonal swellings and QST as predictors of HIV 2021;162(6):1816 27. Available from https://doi.org/
distal sensory neuropathy. Muscle Nerve 2004;29 10.1097/j.pain.0000000000002169.
(3):420 7. [73] Okuse K. Pain signalling pathways: from cytokines to
[65] Karlsson P, et al. Axonal swellings are related to type ion channels. Int J Biochem Cell Biol 2007;39(3):490 6.
2 diabetes, but not to distal diabetic sensorimotor [74] Bursova S, et al. Expression of growth-associated pro-
polyneuropathy. Diabetologia 2021. tein 43 in the skin nerve fibers of patients with type 2
[66] Snider WD, McMahon SB. Tackling pain at the source: diabetes mellitus. J Neurol Sci 2012;315(1 2):60 3.
new ideas about nociceptors. Neuron 1998;20(4):629 32. [75] Dauch JR, et al. Neurogenic factor-induced
[67] Karlsson P, et al. Increased peptidergic fibers as a Langerhans cell activation in diabetic mice with
potential cutaneous marker of pain in diabetic small mechanical allodynia. J Neuroinflam 2013;10:64.
fiber neuropathy. Pain 2020. [76] Casanova-Molla J, et al. Epidermal Langerhans cells in
[68] Gylfadottir, et al. Neurol Neuroimmunol small fiber neuropathies. Pain 2012;153(5):982 9.
Neuroinflamm 2022; Available from https://doi.org/ [77] Gibbons CH, et al. Quantification of sudomotor inner-
10.1212/NXI.0000000000001111. vation: a comparison of three methods. Muscle Nerve
[69] Galosi E, et al. A pain in the skin. Regenerating nerve 2010;42(1):112 19.
sprouts are distinctly associated with ongoing burning [78] Gibbons CH, et al. Quantification of sweat gland
pain in patients with diabetes. Eur J Pain 2018;22 innervation: a clinical-pathologic correlation.
(10):1727 34. Neurology 2009;72(17):1479 86.
[70] Shepherd AJ, et al. Angiotensin II triggers peripheral [79] Nolano M, et al. Loss of cutaneous large and small
macrophage-to-sensory neuron redox crosstalk to fibers in naive and l-dopa-treated PD patients.
elicit pain. J Neurosci 2018;38(32):7032 57. Neurology 2017;89(8):776 84.

Diabetic Neuropathy
 C H A P T E R

6
Central nervous system involvement in
diabetic peripheral neuropathy
Joyce Lim1, Solomon Tesfaye2 and Dinesh Selvarajah1
1
Department of Oncology and Human Metabolism, Medical School, University of Sheffield, Sheffield,
United Kingdom 2Academic Department of Diabetes and Endocrinology, Sheffield Teaching Hospitals
NHS Foundation Trust, Sheffield, United Kingdom

Painful diabetic neuropathy processing. For example, chronic pain patients

are defined as “difficult patients” in that they
Painful distal symmetrical polyneuropathy often have neuropsychological changes that
(painful diabetic neuropathy) affects about 20% include changes in affect and motivation or
of people with diabetes [1] and is defined as changes in cognition, all of which rarely predate
“pain as a direct consequence of abnormalities their pain condition. Chronic pain may also arise
in the peripheral somatosensory system in peo- after the onset of depression, even in patients
ple with diabetes” [2]. It is associated with a without a prior history of pain or depression.
high healthcare burden and reduced produc- Taken together, these clinical insights suggest an
tivity [3]. In the United Kingdom, the average integral role of the central nervous system (CNS)
annual healthcare cost per patient with painful in the pathogenesis and maintenance of the
diabetic neuropathy is d2511 [3]. A cross- chronic pain state.
sectional study, also in the United Kingdom, The management of painful diabetic neu-
revealed that approximately 60% of working ropathy generally involves optimization of
patients with painful diabetic neuropathy felt glycaemic control, modification of cardiovas-
being less productive at work [3]. The pain cular risk factors, and symptomatic relief
experienced by patients can considerably inter- [5 8]. Currently, the mainstay of symptom-
fere with their daily activities, work, mood, atic treatment for diabetic neuropathy is
and sleep, and is associated with a reduced pharmacotherapy and this is covered in
quality of life [3]. Patients with painful diabetic greater detail in other chapters. It is clear that
neuropathy are also more likely to suffer from the CNS is the main site of action of many of
depression and emotional distress [4]. Pain is a the agents used to manage neuropathic pain,
complex multidimensional process, which not thus also hints at its important role in painful
only affects sensory but also emotional/cognitive diabetic neuropathy.

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00001-3 91 © 2022 Elsevier Inc. All rights reserved.
92 6. Central nervous system involvement in diabetic peripheral neuropathy

Despite recent advances and considerable but is also involved in the perception of pain or
research, treatment for neuropathic pain is nociceptive inputs [2]. Changes in the CNS such
largely ineffective. A meta-analysis that evaluated as descending inhibition of pain, neurotransmit-
the medical management of neuropathic pain ter imbalance, central sensitization, and cortical
found overall only modest outcomes, with the reorganization may be implicated in the patho-
number needed to treat to achieve 50% pain genesis of neuropathic pain in diabetes [2]. The
relief ranging from 4 to 10 [9]. In addition, Phase structural and functional changes in the CNS in
III trials of disease-modifying therapies such as painful conditions, including painful diabetic
α-lipoic acid and benfotiamine have largely neuropathy, have also been supported by vari-
shown equivocal results [5]. At present, there is ous imaging studies [2].
no pathogenetic treatment found to provide suf-
ficient pain relief or prevent the development of
neuropathic pain in diabetes [10]. One possible Central nervous system changes in
reason for this is because the pathogenesis of response to peripheral nerve injury and
painful diabetic neuropathy is not fully under- how these changes result in chronic pain
stood. More robust and better designed studies in diabetes
are also required to further assess the benefits of
future novel therapies [11]. There is an emerging Neuropathic pain is a combination of sen-
theory, which postulates that the wide variability sory loss and maladaptive changes in the
in treatment response may in part be due to an peripheral and CNSs [2,18,19] (Fig. 6.1A,B).
underlying heterogeneity in clinical pain pheno- The molecular/ion channel alterations have been
types [12]. Using quantitative sensory testing covered in greater detail in previous chapters.
(QST), distal symmetrical painful neuropathy The changes in the ion channels and ectopic
(DSP) patients can be broadly subdivided into activity at the periphery lead to enhanced synap-
two phenotypes: “irritable” (IR, with relatively tic transmission within the dorsal horn of the spi-
preserved sensory function associated with ther- nal cord and amplification of nociceptive
mal and/or mechanical hyperalgesia) and “non- information, a process known as central sensiti-
irritable” (NIR, dominated by thermal and zation [5,20]. Central sensitization and the altered
mechanical sensory loss) [13]. Subsequent studies balance between inhibitory interneurons and the
appear to suggest that some treatments are seem- descending pain modulatory system can either
ingly more effective in those with the IR com- facilitate or inhibit the transmission of nocicep-
pared to the NIR nociceptor phenotype (for a tive information at spinal levels [5,21]. The
review, see [14] and for examples, see [15 17]). altered balance within these systems (gain of
Consequently, pain phenotyping may, in the excitation or loss of inhibition) is believed to also
future, become important in guiding individual contribute to the development of neuropathic
patients’ treatment although the exact approach pain. Conditioned pain modulation is a tech-
is heavily under debate. This chapter will focus nique to test for the integrity of the descending
on the role for CNS imaging in phenotyping pain modulatory system; a painful conditioning
patients with painful Diabetic peripheral neurop- stimulus applied to one body site reduces pain
athy (DPN) for personalized therapy. caused by a different stimulus in another body
There is growing evidence suggesting that site. It is found to be impaired in patients with
maladaptive alterations in the CNS, not just the various pain syndromes, including painful dia-
peripheral nervous system, play a role in the betic neuropathy [5,20,21]. Variability in the pain
generation and maintenance of neuropathic pain modulation mechanisms among individuals may
[2]. The CNS not only detects and localizes pain, partly explain why some patients experience

Diabetic Neuropathy
 Central nervous system changes in response to peripheral nerve injury and how these changes result in chronic pain in diabetes 93

FIGURE 6.1 (A) An overview of peripheral and central mechanisms involved in painful diabetic neuropathy (PDN).
(B) Overview of peripheral and central mechanisms contributing to neuropathic pain. Source: Adapted from (A) Sloan G,
et al. A new look at painful diabetic neuropathy. Diabetes Res Clin Pract 2018;144:177 191; (B) Meacham K, Shepherd A, Mohapatra
DP, Haroutounian S. Neuropathic pain: central vs. peripheral mechanisms. Curr Pain Headache Rep 2017;21:28.

more debilitating pain than others and why some diabetes causes a generalized, concomitant insult
patients respond to pharmacological and non- to both the peripheral and CNSs [23].
pharmacological therapies but not others [21].
Changes in the brain—structural and
Changes in the spinal cord functional
In recent years, there have been a number of Further investigations into CNS involvement
neuroimaging studies that have exploited in diabetic neuropathy have demonstrated struc-
advances in multimodal magnetic resonance tural changes in the brain. Patients with diabetic
imaging (MRI) to gain new insights into CNS neuropathy displayed abnormalities in the periph-
involvement in diabetic neuropathy. Eaton eral (pre- and postcentral gyrus) and deep gray
et al. revealed significant reduction in the spi- matter (GM) nuclei (caudate, putamen, medial
nal cord cross-sectional area in the cervical and pallidum, thalamus, and ventral nuclear) in
upper thoracic regions in patients with regions involved with somatosensory/nociception
advanced diabetic neuropathy compared to perception. Moreover, there were alterations in the
nondiabetic controls [22,23]. A subsequent, white matter (WM) tracts (spinothalamic tract, cor-
larger study confirmed these findings but also ticospinal tract, and thalamocortical projecting
revealed that spinal cord atrophy was present tracts) indicative of the involvement of somatosen-
in subjects with early, subclinical diabetic neu- sory, motor, and pain-related pathways in diabetic
ropathy. Moreover, subjects with hereditary neuropathy [25]. Specific alterations in patients
sensorimotor neuropathy, an inherited neuropathy with painful diabetic neuropathy have been
localized to the peripheral nerves, had preserved reported in the cingulate cortex, insular cortex,
cord cross-sectional area. This argues against the prefrontal lobe, thalamus, periaqueductal WM,
“dying-back” phenomenon [24], that is, peripheral and external capsule [25]. The involvement of the
nerve injury leading to subsequent cord atrophy anterior cingulate cortex (ACC) and prefrontal
in diabetic neuropathy rather it suggests that cortex has been implicated in the unpleasant

Diabetic Neuropathy
94 6. Central nervous system involvement in diabetic peripheral neuropathy

perception associated with allodynia [25] (pain pain conditions, demonstrate dynamic neuronal
due to a stimulus that normally does not provoke changes that have profound effects on the brain
pain) [26]. Both patients with painful and painless in patients with diabetic neuropathy. More
diabetic neuropathy were also found to have recently, it has been demonstrated how these
reduced GM volume localized to regions involved structural and functional changes are related to
in somatosensory perception compared to control painful DPN clinical phenotypes.
patients with diabetes and no diabetic neuropathy Functional magnetic resonance imaging
and healthy volunteers [27]. Using proton MR (fMRI) is a modality of MRI, which measures
spectroscopy to examine neurochemical alterations brain activity by detecting changes in blood
in the thalamus, patients with painless diabetic flow [bold oxygen level dependent (BOLD) sig-
neuropathy demonstrated abnormal thalamic neu- nal] that is coupled with neuronal activation,
ronal integrity compared to those with painful dia- that is, when an area of the brain is in use,
betic neuropathy. This was supported by blood flow to that region also increases and
structural MRI that demonstrated that thalamic this is assessed using fMRI. Using this modal-
GM volume was reduced in subjects with painless ity of MRI, several studies have identified
diabetic neuropathy but not painful diabetic neu- in vivo the neurological signature for both
ropathy, suggesting that preservation of thalamic physical and neuropathic pain (capsaicin
neuronal integrity may play a key role in the per- model). The brain regions activated in response
ception of pain in diabetic neuropathy [27]. The to a nociceptive stimulus can be divided into
thalamus acts as a relay for the transmission of somatic region responsible for the localization
nociceptive information from the periphery to of pain and coding intensity such as the ven-
other areas of the brain [28]. Animal models of trolateral thalamus, somatosensory cortex, dor-
painful diabetic neuropathy have demonstrated sal posterior insula, and the emotional/
increased spontaneous activity from the thalamic affective regions responsible for emotional
nuclei [29]. Magnetic resonance perfusion imaging pain processing such as the anterior insula,
of the brain has shown increased thalamic vascu- dorsal lateral prefrontal cortex, and ACC [32].
larity in patients with painful diabetic neuropathy
[23]. These findings imply the possible involve-
ment of the thalamus in central amplification of
somatosensory signals [28]. Similar changes have
Recent studies
also been observed in the ACC [30]. Patients with A case-controlled, multimodal MRI study of
painful diabetic neuropathy displayed increased carefully phenotyped patients demonstrated a
cerebral blood flow in the ACC, suggesting that pathophysiological relationship between ana-
ACC activation is implicated in the development tomical and functional changes in the brain
of central sensitization in response to peripheral and sensory phenotypes. Subjects with insen-
neuropathic pain [30]. In addition, patients sate painful diabetic neuropathy (i.e., painful
with painful diabetic neuropathy have altered “painless” diabetic neuropathy) had the lowest
functional connectivity in the ventrolateral S1 cortical thickness and greatest S1 cortical
Periaqueductal grey (PAG), which correlates with functional reorganization with expansion of
their pain intensity and cerebral blood flow in the area representing pain in the lower limb
response to tonic heat stimulation [31]. This sup- region to include face and lips regions [33]
ports the idea that abnormalities in the PAG may compared to painful diabetic neuropathy sub-
result not only in reduced inhibition, but also in jects with relatively preserved sensation.
facilitation of pain [31]. Collectively, these findings, Furthermore, the extent to which S1 cortical
supported by numerous studies in other chronic structure and function are altered was related

Diabetic Neuropathy
 Central nervous system changes in response to peripheral nerve injury and how these changes result in chronic pain in diabetes 95
to the severity of neuropathy and the magni- connectivity between the insular cortex and corti-
tude of self-reported pain. These data suggest colimbic system compared to nonresponders
a dynamic plasticity of the brain in diabetic [36]. The insula [37] cortex plays a pivotal role in
neuropathy, driven by the neuropathic process processing the emotion and cognitive dimen-
and may ultimately determine an individual’s sions of the chronic pain experience. The cortico-
clinical pain phenotype [33]. limbic circuits have also long been implicated in
Over the last decade, resting-state fMRI (RS- reward, decision making, and fear learning.
fMRI)—a quick and simple, noninvasive tech- Hence, these findings suggest that this network
nique—has become an increasingly appealing may have a role in determining treatment
method to examine spontaneous brain activity response in painful DPN. Using advanced multi-
in individuals without relying on experimental modal MR neuroimaging, a number of studies
external stimulation tasks. During a typical RS- have demonstrated alterations in pain-processing
fMRI examination, the hemodynamic response brain regions that relate to clinical pain pheno-
to spontaneous neuronal activity (BOLD) sig- type, treatment response, and behavioral/psy-
nal is acquired while subjects are instructed to chological factors impacted by pain. Taken
simply rest in the MRI scanner [34]. Data together, these assessments could serve as a pos-
acquired are used in brain mapping to evaluate sible Central Pain Signature for painful DPN.
regional interactions or functional connectivity The challenge now is to apply this potential pain
that occur in a resting state. Most studies use a biomarker at an individual level in order to dem-
machine learning approach to identify patterns onstrate clinical utility. To this end, the applica-
of functional connectivity that differentiate tion of a machine learning approach [35] to
patients from controls. RS-fMRI experiments in classify individual patients into different clinical
painful DPN have reported greater thalamic- pain phenotypes using brain imaging features
insula functional connectivity and decreased taken from a quick, 6-minute RS-fMRI scan is
thalamic-somatosensory cortical functional con- appealing.
nectivity in patients with the IR nociceptor phe-
notype compared to the NIR nociceptor
phenotype. There was a significant positive cor-
relation between thalamic-insula functional con-
Maladaptive responses
nectivity with self-reported pain scores [35]. The perception of pain depends on the com-
Conversely, there was a greater reduction in plex interaction between cognitive, emotional,
thalamic-somatosensory cortical functional con- socio-cultural, and physical factors [38]. Emotions
nectivity in those with more severe neuropathy. and cognitive factors such as attention and mem-
This demonstrates how RS-MRI measures of ory influence how patients perceive pain; nega-
functional connectivity relate to both the somatic tive emotional states can cause increased pain
and nonsomatic assessments of painful DPN. whereas positive states result in reduced pain
Using a machine learning approach to integrate [39]. Depression and anxiety are frequently pres-
anatomical and functional connectivity, data ent in patients with neuropathic pain and are
achieved an accuracy of 92% and sensitivity of associated with catastrophizing, which is charac-
90% indicating good overall performance [35]. terized by feeling of helplessness, inability to
Multimodal MRI combining structural and RS- cope with the pain and propensity to exaggerate
fMRI has also been used to predict treatment the threat value of pain [40]. This can in turn lead
response in painful DPN. Responders to intrave- to maladaptive behavior such as avoidance of
nous lidocaine treatment have significantly activities, which promotes disability and persis-
greater S1 cortical volume and greater functional tence of pain [4]. The significance of this is also

Diabetic Neuropathy
96 6. Central nervous system involvement in diabetic peripheral neuropathy

evident in the therapeutic context, as a patient’s new therapeutic interventions and alter the man-
enhanced or diminished response to pharmaco- agement approach for painful diabetic neuropa-
therapy may be influenced by his or her expecta- thy. Preliminary findings from the studies
tion, emotional state, and attention [39]. It also conducted to date suggest their utility as diag-
highlights the important role of the CNS in pain- nostic, prognostic, and pharmacodynamic bio-
ful diabetic neuropathy. markers that should be carefully considered for
In conclusion, neuronal damage leads to mal- possible inclusion in clinical trials of pain treat-
adaptive responses in the nociceptive pathway ments. There are of course limitations and the
and the nervous system, which is the driver of usefulness of biomarkers remains a controversial
persistent and chronic pain [41]. QST profiles topic. Focused research is now needed to stan-
have revealed three distinct phenotypes in dardize the application of neuroimaging biomar-
patients suffering from neuropathic pain, namely kers to address the heterogeneity in their
sensory loss, thermal hyperalgesia, and mechani- application in terms of methodological approach,
cal hyperalgesia, which reflect the underlying outcomes examined, and time points used for
pathophysiological mechanisms [42]. For exam- assessments. The direct clinical application of
ple, peripheral sensitization is the predominant brain imaging in practice will also be dependent
mechanism in thermal hyperalgesia while central on a robust health economic analysis of screen-
sensitization is mainly involved in mechanical ing and interventions with study designs that
hyperalgesia [42]. Patients with diabetic neuropa- allow relevant and appropriate cost comparisons.
thy also manifest these phenotypes, with sensory A crucial element here is the patient’s perspec-
loss being the most common (83%), followed by tive on the usefulness, acceptance, and feasibility
thermal hyperalgesia (75%), and mechanical of these novel imaging technologies in the provi-
hyperalgesia (34%) [43]. Further stratification of sion of future services. We focused on the appli-
patients based on their phenotypes and key patho- cation of CNS neuroimaging on people with
physiological driver(s) is necessary to provide painful diabetic neuropathy throughout the
individualized treatment to optimize outcomes chapter as the vast majority of studies performed
[20,44]. More research is required to determine the were in this group. The clinical implications of
causality of CNS changes in patients with diabetic CNS involvement in patients with painless dia-
neuropathy to help direct management as changes betic neuropathy, on the other hand, remain
in the architecture of the brain and spinal cord, much to be understood—for example, the associ-
especially early in the disease process may imply ation and possible mechanisms of neuropathy on
irreversibility and may be associated with poorer cognitive decline and the indifference observed
outcomes [27]. A better understanding of the in some patients with recurrent diabetic foot
underlying mechanisms for maladaptive plasticity ulcers when it comes to observing preventative
may help to identify specific therapeutic targets to foot care practices despite of the increased risk of
prevent the development of neuropathic pain and major amputations.
to normalize function in patients with established
neuropathic pain [41].
Future direction
Clinical implications Future research has a significant role in
addressing the interrelationships between the
As highlighted in this chapter, brain imaging peripheral and CNS in painful diabetic neuropa-
has the potential to fast-track the development of thy. Future research directions can be divided into

Diabetic Neuropathy
 Future direction 97
TABLE 6.1 Some of the benefits and limitations of various neuroimaging modalities used in the investigation of
central nervous system lesions [45 51].

Neuroimaging
techniques Benefits Limitations

Functional • Detects local increased cerebral blood flow and • Low temporal resolution
MRI changes in blood oxygen concentration (BOLD • Signals may be affected by scanner’s loud noise
contrast) and physiologic noise
• Two primary types: task-based fMRI (utilization • Neurovascular uncoupling: false negative BOLD
of visual, auditory, or other stimuli to induce responses due to interference of coupling
different cognitive states in subjects) and resting- between neuronal activity and adjacent
state fMRI (emphasis on spontaneous BOLD microvasculature
signal fluctuation)
• Relatively high spatial resolution
• Noninvasive
PET scan • Neuronal metabolic activity measured using 18F- • Involves the use of radioactive tracers
FDG can act as a marker of nerve injury and • Lower spatial resolution compared to fMRI
neuropathic pain
• Noninvasive
1
H-MRS • Noninvasive • Low sensitivity
• Does not involve ionizing radiation • Lack of standardization and quality assurance of
• Able to investigate a wide range of biological data acquisition and analysis methods
processes
• Combination of spectroscopy and imaging
enables the acquisition of metabolic,
physiological, and anatomical data in a single
experiment

three key areas: mechanistic or pathophysiological, controls, thus providing diagnostic clues of the
performing clinical trials relevant to an individual brain patterns in pain conditions and the underly-
with painful diabetic neuropathy and influencing ing mechanisms involved in the generation, main-
federal agencies/commissioners of diabetes and tenance, and exacerbation of chronic pain [14]. As
pain services. Brain imaging techniques such as a prognostic tool, brain imaging can be used to
fMRI, positron emission tomography (PET), and identify individuals who are more “vulnerable” or
proton magnetic resonance spectroscopy (1H- “resilient” against developing persistent pain after
MRS) provide neurochemical, structural, or being exposed to factors that can trigger the onset
functional information on the processing and of chronic pain [14]. For example, increased func-
modulation of nociceptive inputs in the brain, tional connectivity of the nucleus accumbens with
which result in the perception of pain [14]. prefrontal cortex predicts transition from acute
Table 6.1 summarizes some of the benefits and pain to chronic pain in subjects with subacute
limitations of these imaging modalities [45 51]. back pain [52]. Brain imaging can be used to pre-
Neuroimaging studies have shown differences in dict treatment response, hence may enable stratifi-
the areas of brain activated in acute pain and cation of participants in clinical trials [14]. fMRI
chronic pain as well as differences in functional data derived from visual stimulation have shown
connectivity between affected patients and healthy a greater than 80% accuracy in distinguishing

Diabetic Neuropathy
98 6. Central nervous system involvement in diabetic peripheral neuropathy

whether a patient with fibromyalgia was adminis- are the cause or consequence of the chronic pain
tered pregabalin or placebo [53]. Brain imaging state. However, it is evident from studies in pain-
can be a useful diagnostic, prognostic, and predic- ful diabetic neuropathy that the brain is ailing
tive tool and its use as a potential biomarker in both in terms of functional reorganization and
chronic pain clinical trials requires further extensive neurodegeneration.
research focussing on standardization of out-
comes, validation, reproducibility, and evalua-
tion of outcomes [14]. Conclusions
The next important step in delineating CNS
involvement in diabetic neuropathy is to develop In conclusion, clinicians and patients
methods that allow noninvasive assessment of acknowledge that painful diabetic neuropathy
the functional changes in the spinal cord in vivo. is an important complication of diabetes that
As detailed earlier, there have been a number of leads to significant comorbidity and suffering.
anatomical studies that have demonstrated Hitherto considered a disease of the peripheral
increased cord atrophy in diabetic neuropathy— nervous system, research over the past decades
however, accurate assessment of (1) neuronal has outlined the clinical features and brain/spi-
activity within the spinal cord in relation to nal cord imaging correlates of painful diabetic
evoked/spontaneous pain and (2) top-down neuropathy. Insights gained from clinical
inhibition of afferent nociceptive inputs will com- research are beginning to be translated into
plete the picture of CNS modulation of neuro- applications but there are still gaps in our
pathic pain. This poses an enormous imaging knowledge. Current challenges include replica-
challenge because the small size of the spinal tion to provide further support for the neuroim-
cord results in low signal to noise and there is aging tools. It is important to acknowledge that
considerable motion artifact from respiration, most of the studies described have been rela-
cardiac cycle, and cerebrospinal fluid pulsa- tively small, cohort studies that were explor-
tions [54]. These challenges are now being atory that could lead to false positive results as
addressed with high-field strength MRI at 7 a consequence of an inflated risk of type 1 error.
and 14 T [55]—opening up this part of the neu- Increase assay sensitivity in clinical trial—
ral axis for future mechanistic studies and exposing fewer patients to risk of new drugs
clinical applications. and reducing costs of Phase 3 studies. Another
Another promising future frontier imaging important aspect of neuroimaging research in
innovation that will significantly impact on clini- the context of chronic pain is the need to stan-
cal and experimental pain research is the avail- dardize the use of neuroimaging tools and
ability of PET ligands for opioid [56] and postprocessing methods. Significant steps have
dopamine receptors [57]. This has allowed the been made toward addressing this limitation
study of these receptor systems in a number of with the publication of various consensus state-
clinical pain states but presently remains solely a ments [60,61]. However, emerging approaches to
research tool. Preliminary studies have showed imaging analyses such as machine learning algo-
abnormal receptor binding in a number of differ- rithms (e.g., support vector machines, convolu-
ent neuropathic pain states [58] and reversal of tional neural networks, etc.) may also prove
this abnormality following the resolution of pain- useful in addressing concerns about replication
ful symptoms [59]. The main limitation as yet to [32,53,62]. These methods enable pooling of MR
be resolved with many of these cohort, observa- data from studies utilizing different imaging
tional studies is whether the changes described equipment and analysis procedures and have

Diabetic Neuropathy
 References 99
produced informative inferences about brain implications for other chronic pain etiologies
structure and function in diabetic neuropathy outside the field of diabetes and hence are
[35,60,61]. These are essential steps toward gain- highly relevant.
ing regulatory approvals from federal agencies
for the use of neuroimaging as a validated bio-
marker in diabetic painful neuropathy [14]. References
This would lead to increased assay sensitivity
[1] Veves A, Backonja M, Malik RA. Painful diabetic neu-
in later-stage clinical trials (e.g., Phase 2b or 3 ropathy: epidemiology, natural history, early diagno-
studies), which will minimize patient exposure sis, and treatment options. Pain Med 2008;9:660 74.
to experimental drugs and has the potential to [2] Sloan G, et al. A new look at painful diabetic neuropa-
reduce the overall costs of drug development. thy. Diabetes Res Clin Pract 2018;144:177 91.
Disease-modifying treatments that target the [3] Alleman CJM, et al. Humanistic and economic burden
of painful diabetic peripheral neuropathy in Europe: a
pathogenesis of diabetic neuropathy and effec- review of the literature. Diabetes Res Clin Pract
tive symptoms control remain the highest unmet 2015;109:215 25. Available from: https://doi.org/
needs. As of yet, standard recommendations to 10.1016/j.diabres.2015.04.031.
optimize cardiovascular risk factors effectively [4] Selvarajah D, et al. The contributors of emotional dis-
(in some countries where there is local regulatory tress in painful diabetic neuropathy. Diabetes Vasc
Dis Res 2014;11:218 25.
approval for the use of agents such as aldose [5] Feldman EL, et al. Diabetic neuropathy. Nat Rev Dis
reductase inhibitors, alpha-lipoic acid, and ben- Prim 2019;5.
fothiamine) and symptom control are the corner- [6] Alam U, Sloan G, Tesfaye S. Treating pain in diabetic
stones in practice [63]. In the future, these neuropathy: current and developmental. Drugs Drugs
recommendations could be augmented by per- 2020;. Available from: https://doi.org/10.1007/
s40265-020-01259-2.
sonalized, targeted approaches toward restoring [7] Singh R, Rao HK, Singh TG. Neuropathic pain in dia-
neurotransmitter imbalance, cortical neuronal betes mellitus: challenges and future trends. Obes
degeneration/functional reorganization, and fur- Med 2020;18.
ther potential factors addressed in this chapter [8] Knezevic NN, Tverdohleb T, Knezevic I, Candido KD.
that might represent treatment targets. As in The role of genetic polymorphisms in chronic pain
patients. Int J Mol Sci 2018;19:E1707.
other microvascular complications for diabetes, [9] Finnerup NB, et al. Pharmacotherapy for neuropathic
for example, diabetic nephropathy with microal- pain in adults: systematic review, meta-analysis and
buminuria or retinopathy with digital retinal updated NeuPSig recommendations. Lancet Neurol
photography, the identification of a robust bio- 2015;14:162 73.
marker for painful diabetic neuropathy will lead [10] Kioskli K, Scott W, Winkley K, Kylakos S, McCracken
LM. Psychosocial factors in painful diabetic neuropa-
to early identification and also accelerate the thy: a systematic review of treatment trials and survey
development of targets of the specific mecha- studies. Pain Med 2019;20:1756 73.
nism that leads to chronic pain. Moreover, the [11] Bönhof GJ, et al. Emerging biomarkers, tools, and
disease processes that are not unique to diabetes treatments for diabetic polyneuropathy. Endocr Rev
but also contribute to the persistence and resis- 2018;40:153 92.
[12] Finnerup NB, Sindrup SH, Jensen TS. The evidence
tance to treatment of the chronic pain condition for pharmacological treatment of neuropathic pain.
will be elucidated further. For example, the CNS Pain 2010;150:573 81.
processes described in this chapter in patients [13] Backonja MM, et al. Value of quantitative sensory test-
with painful diabetic neuropathy are probably ing in neurological and pain disorders: NeuPSIG con-
still important contributors to chronic pain in sensus. Pain 2013;154:1807 19.
[14] Smith SM, et al. The potential role of sensory testing,
people without diabetes. Thus neuroimaging skin biopsy, and functional brain imaging as biomar-
led developments in the etiological treatments kers in chronic pain clinical trials: IMMPACT consid-
for people with diabetes will have wider erations. J Pain 2017;18:757 77.

Diabetic Neuropathy
100 6. Central nervous system involvement in diabetic peripheral neuropathy

[15] Mainka T, et al. Presence of hyperalgesia predicts [33] Selvarajah D, et al. Structural and functional abnor-
analgesic efficacy of topically applied capsaicin 8% in malities of the primary somatosensory cortex in dia-
patients with peripheral neuropathic pain. Eur J Pain betic peripheral neuropathy: a multimodal MRI study.
(UK) 2016;20:116 29. Diabetes 2019;68:796 806.
[16] Edwards RR, et al. Patient phenotyping in clinical [34] Lv H, et al. Resting-state functional MRI: everything
trials of chronic pain treatments: IMMPACT recom- that nonexperts have always wanted to know. Am J
mendations. Pain 2016;157:1851 71. Neuroradiol 2018;39:1390 9.
[17] Wasner G, Kleinert A, Binder A, Schattschneider J, Baron [35] Teh K, Wilkinson ID, Heiberg-Gibbons F, Awadh M,
R. Postherpetic neuralgia: topical lidocaine is effective in Kelsall A, Pallai S, et al. Network functional connectiv-
nociceptor-deprived skin. J Neurol 2005;252:677 86. ity differentiates clinical pain phenotypes in diabetic
[18] Jensen TS, Finnerup NB. European Journal of Pain neuropathy. Diabetologia 2021;64:1412 21.
Supplements Neuropathic pain: peripheral and central [36] Wilkinson ID, et al. Determinants of treatment
mechanisms. Eur J Pain Suppl 2009;3:33 6. response in painful diabetic peripheral neuropathy: a
[19] Meacham K, Shepherd A, Mohapatra DP, combined deep sensory phenotyping and multimodal
Haroutounian S. Neuropathic pain: central vs. periph- brain mri study. Diabetes 2020;69:1804 14.
eral mechanisms. Curr Pain Headache Rep 2017;21:28. [37] Lu C, et al. Insular cortex is critical for the perception,
[20] Feldman EL, Nave KA, Jensen TS, Bennett DLH. New modulation, and chronification of pain. Neurosci Bull
horizons in diabetic neuropathy: mechanisms, bioen- 2016;32:191 201.
ergetics, and pain. Neuron 2017;93:1296 313. [38] Torta R, Ieraci V, Zizzi F. A review of the emotional
[21] Colloca L, et al. Neuropathic pain. Nat Rev Dis Prim aspects of neuropathic pain: from comorbidity to co-
2017;3. pathogenesis. Pain Ther 2017;6:11 17.
[22] Eaton SEM, et al. Spinal-cord involvement in diabetic [39] Bushnell MC, Čeko M, Low LA. Cognitive and emo-
peripheral neuropathy. Lancet 2001;358:35 6. tional control of pain and its disruption in chronic
[23] Selvarajah D, Wilkinson ID, Davies J, Gandhi R, pain. Nat Rev Neurosci 2013;14:502 11.
Tesfaye S. Central nervous system involvement in dia- [40] Radat F, Margot-Duclot A, Attal N. Psychiatric co-
betic neuropathy. Curr Diab Rep 2011;11:310 22. morbidities in patients with chronic peripheral neuro-
[24] Selvarajah D, et al. Early involvement of the spinal pathic pain: a multicentre cohort study. Eur J Pain
cord in diabetic peripheral neuropathy. Diabetes Care (UK) 2013;17:1547 57.
2006;29:2664 9. [41] Costigan M, Scholz J, Woolf CJ. Neuropathic pain: a
[25] Zhang Y, et al. Sensorimotor and pain-related altera- maladaptive response of the nervous system to dam-
tions of the gray matter and white matter in Type 2 age. Annu Rev Neurosci 2009;32:1 32.
diabetic patients with peripheral neuropathy. Hum [42] Baron R, et al. Peripheral neuropathic pain: a
Brain Mapp 2020;41:710 25. mechanism-related organizing principle based on sen-
[26] Jensen TS, Finnerup NB. Allodynia and hyperalgesia sory profiles. Pain 2017;158:261 72.
in neuropathic pain: clinical manifestations and [43] Vollert J, et al. Stratifying patients with peripheral neuro-
mechanisms. Lancet Neurol 2014;13:924 35. pathic pain based on sensory profiles: algorithm and
[27] Selvarajah D, et al. Magnetic resonance neuroimaging sample size recommendations. Pain 2017;158:1446 55.
study of brain structural differences in diabetic [44] Themistocleous AC, et al. The Pain in Neuropathy
peripheral neuropathy. Diabetes Care 2014;37:1681 8. Study (PiNS): a cross-sectional observational study
[28] Shillo P, et al. Painful and painless diabetic neuropa- determining the somatosensory phenotype of
thies: what is the difference? Curr Diab Rep 2019;19:32. painful and painless diabetic neuropathy. Pain
[29] Freeman OJ, Evans MH, Cooper GJS, Petersen RS, 2016;157:1132 45.
Gardiner NJ. Thalamic amplification of sensory input in [45] Glover GH. Overview of functional magnetic reso-
experimental diabetes. Eur J Neurosci 2016;44:1779 86. nance imaging. Neurosurg Clin N Am 2011;22:133 9.
[30] Watanabe K, et al. Altered cerebral blood flow in the [46] Chen JE, Glover GH. Functional magnetic resonance
anterior cingulate cortex is associated with neuropathic imaging methods. Neuropsychol Rev 2015;25:289 313.
pain. J Neurol Neurosurg Psychiatry 2018;89:1082 7. [47] Agarwal S, Sair HI, Pillai JJ. Limitations of rsfMRI in
[31] Segerdahl AR, Themistocleous AC, Fido D, Bennett the setting of focal brain lesions. Neuroimag Clin N
DL, Tracey I. A brain-based pain facilitation mecha- Am 2017;27:645 61.
nism contributes to painful diabetic polyneuropathy. [48] Slough C, Masters SC, Hurley RA, Taber KH. Clinical
Brain 2018;141:357 64. positron emission tomography (PET) neuroimaging:
[32] Wager TD, et al. An fMRI-based neurologic signature advantages and limitations as a diagnostic tool. J
of physical pain. N Engl J Med 2013;368:1388 97. Neuropsychiatry Clin Neurosci 2016;28:1 6.

Diabetic Neuropathy
 References 101
[49] Behera D, Jacobs KE, Behera S, Rosenberg J, Biswal S. emission tomography. J Cereb Blood Flow Metab
18F-FDG PET/MRI can be used to identify injured 1999;19:803 8.
peripheral nerves in a model of neuropathic pain. J [57] Hagelberg N, et al. Striatal dopamine D2 receptors in
Nucl Med 2011;52:1308 12. modulation of pain in humans: a review. Eur J
[50] Chatham JC, Blackband SJ. Nuclear magnetic reso- Pharmacol 2004;500:187 92.
nance spectroscopy and imaging in animal research. [58] Harris RE, et al. Decreased central μ-opioid receptor
ILAR J 2001;42:189 208. availability in fibromyalgia. J Neurosci 2007;27:10000 6.
[51] Öz G, et al. Clinical proton MR spectroscopy in central [59] Jones AKP, et al. Changes in central opioid receptor
nervous system disorders. Radiology 2014;270:658 79. binding in relation to inflammation and pain in
[52] Baliki MN, et al. Corticostriatal functional connectivity patients with rheumatoid arthritis. Rheumatology
predicts transition to chronic back pain. Nat Neurosci 1994;33:909 16.
2012;15:1117 19. [60] Huang L, et al. Brain white matter changes associated
[53] Harte SE, et al. Pharmacologic attenuation of cross- with urological chronic pelvic pain syndrome: multi-
modal sensory augmentation within the chronic pain site neuroimaging from a MAPP case-control study.
insula. Pain 2016;157:1933 45. Pain 2016;157:2782 91.
[54] Brooks JCW, et al. Physiological noise modelling for [61] Kairys AE, et al. Increased brain gray matter in the
spinal functional magnetic resonance imaging studies. primary somatosensory cortex is associated with
Neuroimage 2008;39:680 92. increased pain and mood disturbance in patients with
[55] Maieron M, et al. Functional responses in the interstitial cystitis/painful bladder syndrome. J Urol
human spinal cord during willed motor actions: 2015;193:131 7.
evidence for side- and rate-dependent activity. J [62] López-Solà M, et al. Towards a neurophysiological
Neurosci 2007;27:4182 90. signature for fibromyalgia. Pain 2017;158:34 47.
[56] Jones AKP, et al. Measurement of changes in opi- [63] Sloan G, Selvarajah D, Tesfaye S. Pathogenesis, diag-
oid receptor binding in vivo during trigeminal neu- nosis and clinical management of diabetic peripheral
ralgic pain using [11C]diprenorphine and positron neuropathy. Nat Rev 2021;17:400 20.

Diabetic Neuropathy
 C H A P T E R

7
Spinal cord involvement in diabetic
neuropathy and neuropathic pain
Andrew G. Marshall1,2,3, Anne Marshall4 and Nigel A. Calcutt5
1
Musculoskeletal Biology, Institute of Life Course and Medical Sciences, University of Liverpool,
Liverpool, United Kingdom 2Department of Clinical Neurophysiology, The Walton Centre, Liverpool,
United Kingdom 3Division of Neuroscience and Experimental Psychology, School of Biological
Sciences, Faculty of Biology, Medicine and Health, University of Manchester, Manchester, United
Kingdom 4Division of Diabetes, Endocrinology and Gastroenterology, School of Medical Sciences,
Faculty of Biology, Medicine and Health, University of Manchester, Manchester, United Kingdom
5
Department of Pathology, School of Medicine, University of California San Diego, La Jolla, CA,
United States

Introduction Each of the 31 segments of the spinal cord (8

cervical, 12 thoracic, 5 lumbar, 5 sacral, and 1
The spinal cord is a critical site for integra- coccygeal) receives peripheral sensory input
tion of peripheral sensory information en from a pair (1 on each lateral side) of dorsal (in
route to the brain and for regulated dispersal quadrupeds)/posterior (in bipeds) spinal roots
of responses via the motor and autonomic divi- that derive from the dorsal root ganglia that
sions of the peripheral nerve system (PNS). house perikarya of peripheral primary sensory
Like the brain, the spinal cord is protected neurons. Motor outflow from the spinal cord is
physically by its location within the spinal via a pair of ventral (in quadrupeds)/anterior
canal of the vertebral column, where it is (in bipeds) spinal roots containing axons that
ensheathed by the dura mater and pia mater project from motor cell bodies located in the
membranes and also by the blood:spinal cord ventral horn of the spinal cord. The spinal
barrier (BSCB) which regulates access of blood cord itself is fundamentally divided into gray
borne materials. A brief review of the organiza- and white matter (Fig. 7.1). The centrally located
tion of the spinal cord may help set the stage gray matter is butterfly shaped with pairs of
for consideration of how diabetes impacts spi- dorsal and ventral horns connected by the gray
nal cord structure and function. commissure arranged around the central canal.

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00010-4 103 © 2022 Elsevier Inc. All rights reserved.
104 7. Spinal cord involvement in diabetic neuropathy and neuropathic pain

FIGURE 7.1 Anatomy of the spinal cord showing major tracts of the white matter (left panel) and Rexed laminae I-X of
the gray matter (right panel). Ascending tracts are labeled in green/italicized text, and descending tracts in red text. Line
drawing courtesy of Daniel Mizisin.

Gray matter consists of neuronal cell bodies, The spinal cord is bathed in cerebrospinal
their dendrites and axons, associated neuroglia fluid (CSF) of the subarachnoid space that
and blood vessels. It is also the place where the flows down a pressure gradient from the ven-
central projections of peripheral sensory neu- tricles of the brain and exits via continuities
rons terminate to form synapses with second with the endoneurial space of the dorsal/pos-
order neurons and is divided into 10 zones terior (sensory) and ventral/anterior (motor)
termed the laminae of Rexed (Fig. 7.1, right roots at the root entry zone. CSF obtained via a
panel). Gray matter is surrounded by white spinal tap or intrathecal catheters allows indi-
matter consisting of axons, their ensheathing rect assessment of both local environmental
glial cells (astrocytes, oligodendrocytes, micro- conditions such as glucose levels and also of
glia) and blood vessels. White matter is spinal cord activity via assays of neurotrans-
broadly divided into anterior, lateral, and pos- mitters and metabolites that spill over into the
terior funiculi, which are further subdivided CSF from synapses in the gray matter.
into tracts of axons all traveling in the same
direction (Fig. 7.1, left panel). The ascending
tracts (Fig. 7.1, left panel, green/italicized text) Glucotoxicity in the spinal cord
consist of axons that send information to the
brain, whereas the descending tracts (Fig. 7.1, Existing within the BSCB gives the spinal cord
left panel, red text) consist of axons that origi- a different exposure to blood borne systemic
nate from neuronal perikarya located in the insults such as the hyperglycemia and dyslipide-
cerebral cortex, midbrain, medulla, and vestib- mia of diabetes compared with peripheral nerve.
ular apparatus. Tracts cross from one side of For example, glucose levels in the CSF are
the spinal cord to the other via the anterior around two-thirds that of peripheral blood and
white commissure. the endoneurial fluid of peripheral nerves.

Diabetic Neuropathy
 Spinal cord pathology associated with diabetes and neuropathy 105
Moreover, although CSF glucose concentrations thinning/demyelination in the posterior columns,
increase during diabetes [1], they remain notably lateral (spinocerebellar) tracts, and ventral col-
lower than those to which peripheral axons and umns. In the gray matter, shrinkage, chromatoly-
sensory ganglia are exposed via the endoneurial sis, and loss of anterior horn cells (cell bodies of
fluid. Consequently, pathogenic mechanisms peripheral motor neurons) with attendant gliosis
that rely on tissue exposure to increased absolute are most frequently described, with occasional
glucose concentrations such as nonenzymatic reports of similar damage in the posterior horns.
glycation of proteins may not be as toxic to the Interestingly, both white and gray matter lesions
spinal cord as to peripheral nervous tissue, are more frequent in the lower segments of the
although cells of the spinal cord will still be cord. This is suggestive of a length-dependent
exposed to glucose concentrations that are abnor- pathogenic mechanism, as it is commonly impli-
mally high in relative terms. Another feature of cated in diabetic peripheral neuropathy. Functional
the anatomy of the spinal cord and associated correlates of spinal pathology may be detected by
peripheral nerves is that motor neuron perikarya, electrophysiology. For example, somatosensory
located in the ventral horn of the spinal cord, are evoked potential recordings indicate involvement
exposed to lower glucose concentrations during of dorsal column pathways in diabetic subjects
diabetes than are their own axons that project with peripheral neuropathy but the alterations are
into the periphery. Conversely, the cell bodies of inconsistent with only very minor delays in the
sensory neurons located in the dorsal root gan- spinal conduction of large fiber sensory inputs
glia, their peripheral projections and most of having been described [8,9].
their central projections are exposed to more The development of high-quality noninvasive
extreme hyperglycemia than the terminals of magnetic resonance imaging (MRI) has enabled
their central projections that synapse in the spinal in vivo assessment of the gross anatomy of the
dorsal horn and are thus within an environment spinal cord of patients with diabetes, thereby
protected by the BSCB. The sensory-predominant precluding concerns about tissue preservation
characteristics of some chemotherapy-induced that accompany autopsy studies. Volumetric
peripheral neuropathies have been attributed to MRI studies have documented a reduction in
better access of the neurotoxic drugs to sensory cross-sectional spinal cord area in patients with
ganglia compared to the ventral spinal cord and confirmed diabetic neuropathy when compared
similar considerations may also apply in diabetes to age-matched healthy controls and to patients
when hyperglycemia or other blood-borne factors with diabetes and no neuropathy [10,11]. Spinal
are proposed as the initiating toxic insult. cord atrophy was most marked in patients with
clinically detectable polyneuropathy compared
to those with subclinical disease [11]. This reduc-
Spinal cord pathology associated with tion in spinal cord area was seen even at the
diabetes and neuropathy early stages of neuropathy, suggesting that
spinal cord atrophy is an early event in the
Evidence of demyelination, axonal loss, and progression of disease [11]. Unfortunately,
gliosis has been described in postmortem reports current imaging technologies are not able to
of human spinal cord extending back over a cen- detail involvement of specific Rexed laminae
tury or more [2 7], although caveats concerning or white matter tracts or even to identify dif-
tissue source, postmortem delay, handling, and ferential structural involvement of the gray
fixation have caused such studies to be viewed and white matter of the spinal cord and
with caution. The most commonly reported white advances in instrumentation will be necessary
matter lesions include axonopathy and myelin to provide such detail.

Diabetic Neuropathy
106 7. Spinal cord involvement in diabetic neuropathy and neuropathic pain

The underlying reason for the reduction in shorter length of nerves may attenuate patholo-
spinal cord cross-sectional area measured by gies arising from physical length dependent
MRI and whether it is a consequence of periph- processes. Both of these features may contribute
eral nerve pathology or a parallel spinal patho- to the limited number of reports of spinal injury
logical process is unclear. However, there was in preclinical studies using diabetic rodents and
a correlation between spinal cord area and the relatively mild damage they describe.
sural nerve sensory conduction velocity, as
well as with a composite neuropathy score,
suggesting that the degree of peripheral and Spinal cord pathology
spinal pathology was related [11] and promot-
Within white matter, neuroaxonal dystrophy
ing speculation that there may also be shared
has been described in the dorsal columns, spe-
pathogenic mechanisms. Postmortem findings
cifically ascending sensory fibers of the gracile
of microvascular disease in the spinal cord are
fascicle, of spontaneously diabetic rats [20]
similar to those seen in peripheral nerves [4,6]
with a progressive myelinated fiber atrophy
suggesting a shared sensitivity to insults that
and later Wallerian degeneration that pre-
promote vascular injury. Ischemic hypoxia
sented in a proximal to distal gradient [21].
resulting from microvascular disease has been
Autonomic nerves within the cord were also
widely studied as a pathogenic mechanism for
affected, with atrophy of sympathetic postgan-
diabetic peripheral neuropathy [12] and could
glionic fibers of the gray ramus despite normal
also contribute to the pathogenesis of spinal
preganglionic fibers in the white ramus [21].
cord lesions [157]. However, the functional sig-
Gray matter pathology is otherwise limited to
nificance of such spinal cord pathologies and if
a report of reduced perikaryal volume of ven-
they relate to structural and functional altera-
tral horn motor neurons in the absence of overt
tions in the brain during diabetes [13 17]
degeneration or numerical loss that recapitu-
remains to be established.
lated the mild pathology of motor axons in
peripheral nerve [22,23]. Indeed, the absence of
overt axonal degeneration and demyelination
Studies in animal models of diabetes in the spinal cord of diabetic rodents is gener-
ally consistent with the picture in the PNS of
Many of the genetic, chemical, and dietary
these animals, where large fiber axonal atrophy
animal models of diabetes exhibit multiple indi-
and myelin thinning are the main features seen
ces of peripheral neuropathy and are widely
in nerve trunks following many months of dia-
used to investigate pathophysiological mechan-
betes. The overall pattern of neuropathy in dia-
isms that may be pertinent to the human condi-
betic rodents is one of a slow distal axonal
tion [18,19]. With regard to neuropathology,
degeneration with spared perikarya [24].
rodent models of diabetes offer the advantage of
controlled tissue acquisition and processing for
optimal preservation of structures in the spinal
cord, which reduces artifacts and allows spinal
Spinal cord electrophysiology
cord sections cut from resin blocks to be viewed Conduction slowing has been reported in
by light and electron microscopy. However, it both the ascending and descending tracts of
must also be noted that exposure to diabetes is diabetic rats, with delayed onset compared to
much shorter in rodents that have a life span of slowing of peripheral nerve conduction [25,26].
only 1 2 years and are frequently studied after Spinal conduction slowing in type 1 diabetic
only 1 6 months of diabetes, while the much rats was prevented by spinal infusions of

Diabetic Neuropathy
 Studies in animal models of diabetes 107
insulin or IGF-1 that was unlikely to modulate nerve conduction velocity in rodents, in which
systemic hyperglycemia, suggesting a direct stimulation of the sciatic nerve via a needle
neuroprotective effect [27,28]. electrode evokes two waveforms in muscles of
Wide dynamic range neurons, second-order the ipsilateral paw; a direct nerve to muscle
neurons that receive input from a broad range M-wave and the longer latency trans-spinally
of sensory stimuli and project via the spinotha- mediated H-wave (aka H-reflex: Fig. 7.2).
lamic tract to the brain, show abnormal sponta- Rather than recording M and H waves evoked
neous activity in diabetic rodents [29,30] with at two different stimulation sites and using the
distinct patterns of dysfunction between differ- peak to peak latency of each wave and distance
ent models of diabetes [31]. These neurons also between stimulation sites to calculate conduction
exhibit changes to the shape and organization velocity [34], a train of stimuli from a single loca-
of their dendritic spines suggesting a structural tion can be used to assess function of the spinal
malorientation that may contribute to abnor- synapse(s) that mediate the H wave. It is well
mal sensory processing [32]. Spinal wind-up, a established that stimulus trains of certain fre-
frequency-dependent increase in the excitabil- quencies, while having no impact on M wave
ity of spinal neurons following repeated electri- amplitude, produce a decrease in amplitude of
cal stimulation of C fibers, is also enhanced in the H wave—phenomenon variously known as
diabetic mice [33], further supporting the idea rate-dependent depression (RDD or HRDD)
that spinal mediation of sensory processing is paired pulse depression or frequency-dependent
altered by diabetes. depression [35 37] and that this phenomenon
Synaptic function in the spinal cord can be can be lost following physical- [36,38] or disease-
assessed using modification of the protocols related [39,40] injury to the spinal cord. Loss of
widely used to measure motor and sensory RDD accompanied damage to spinal descending

FIGURE 7.2 Representation of the circuits involved in generation of the M and H waves following nerve stimulation.
When stimulating at 1 Hz, M wave amplitude remains constant whereas there is a decline in H wave amplitude termed
rate-dependent depression (H-RDD 5 (H1 2 H2)/H1 3 100).

Diabetic Neuropathy
108 7. Spinal cord involvement in diabetic neuropathy and neuropathic pain

inhibitory systems and interneurons with result- general degenerative phenotype. Prolonged release
ing spasticity [41] that was subsequently linked of prostaglandin E was recorded in the same
to disrupted spinal GABAergic inhibition [42] experimental system, suggesting an extended
and specifically to failure of the inhibitory func- spinal sensitization response [56]. Moreover,
tions of the spinal GABAA [43] and ultimately both basal and stimulus evoked release of GABA
GABAB receptors [44]. are also increased in diabetic rats [53,57]. Excess
RDD is impaired in the spinal cord of dia- spinal GABA may therefore drive a spinally
betic rodents [43,45 47] and humans [48]. In mediated hyperalgesia instead of serving its
rodents, this has been ascribed to reduced usual inhibitory function. This is supported by
expression and activity of the postsynaptic ion the counterintuitive efficacy of GABA antago-
pump potassium chloride cotransporter 2 nists to alleviate behavioral indices of pain in
(KCC2) which disrupts the chloride reversal diabetic rats [43,44].
potential and causes the GABAA receptor, a
chloride ion channel, to lose its inhibitory func-
tion and develop excitatory function [49]. Spinal cord cellular pathology
Excess spinal BDNF has been linked to
In the absence of overt degenerative pathol-
reduced KCC2 expression [47]. This mecha-
ogy in the spinal cord of diabetic rodents, inter-
nism is not unique to diabetes and has been
est has focused on morphological changes to
advanced as a basic mechanism of spinal sensi-
cells and any accompanying shifts in their
tization following peripheral nerve injury
expression of receptors, cytokines, chemokines,
[50 52]. Because both RDD and behavioral
and other molecules. Such changes may indi-
indices of neuropathic pain in diabetic rodents
cate early reactive responses to ongoing lesions
respond to the same pharmacological interven-
that precede degeneration or be part of patho-
tions and also segregate from neuropathic pain
genic cascades that contribute to dysregulation
in models of peripherally driven neuropathic
of spinal sensory and motor processing.
pain, it has been suggested that RDD can be
Indeed, there is a growing appreciation that
used as a biomarker to identify neuropathic
spinal glia, comprising oligodendrocytes, astro-
pain arising from spinal sensitization in indi-
cytes, and microglia, form a network that can
vidual diabetic patients [47]. This mechanism
orchestrate a coordinated response to insults
is discussed in detail below.
that includes interaction with the vascular and
Loss of GABAA receptor inhibitory function
immune systems and leads to modulation of
may also address another physiological abnor-
neuronal synaptic function. This network can
mality in the spinal cord diabetic rodents.
also be triggered to driving an inappropriate
Studies using spinal microdialysis in conscious
proinflammatory state in the spinal cord that has
unrestrained rats to measure real-time release of
been linked to spinally mediated neuropathic
neurotransmitters during painful stimuli and
pain in many conditions, including diabetes.
subsequent evoked behaviors initially showed
that hyperalgesia during the formalin test was
unexpectedly accompanied by decreased spinal Oligodendrocytes
release of the primary afferent derived excitatory Despite lack of overt demyelination in white
neurotransmitters glutamate and substance P matter tracts of diabetic rodents, there is evidence
[53]. Reduced substance P release is consistent that oligodendrocytes are vulnerable to insults
with the reduced synthesis axonal transport and associated with diabetes. Oligodendrocytes show
evoked release of neuropeptides in peripheral evidence of oxidative damage to myelin lipids
sensory nerves of diabetic rodents [54,55] and their [58] and there is reduced expression of myelin

Diabetic Neuropathy
 Studies in animal models of diabetes 109
basic protein [59], a protein that interacts with not surprising given the emerging association
lipids to maintain correct myelin organization. between diabetes-induced macro- and micro-
Oligodendrocyte responses to hyperglycemia vascular disease, increased BBB permeability,
may also promote spinal modification of sensory and cognitive dysfunction [67]. However, data
processing. For example, oligodendrocytes, like obtained in the brain may not directly extrapolate
their peripheral nerve counterparts (Schwann to the spinal cord given that diabetes-induced
cells), express both aldose reductase [60] and BBB permeability shows regional variations even
cyclooxygenase 2 (COX-2) [61]. Aldose reductase within the brain [68]. Indeed, increased albumin
has been widely implicated in the pathogenesis of leak through the BBB was reported in the peri-
diabetic neuropathy and other complications due ventricular area of the brain of diabetic mice, but
to its activity metabolizing excess glucose during not in the spinal cord of the same animals [69].
hyperglycemia [62], while COX-2 produces pros- Whether this reflects a permanent regional
taglandins that evoke inflammatory and some difference or a temporal variation as diabetes-
forms of neuropathic pain. Spinal glucose metabo- induced lesions evolve is not known. In contrast,
lism by aldose reductase during diabetes Evans blue extravasation into the spinal cord,
increases COX-2 expression and activity in oligo- another measure of BSCB permeability, has been
dendrocytes, which in turn drives prostaglandin reported to be markedly reduced in diabetic
release and promotes spinally mediated hyperal- rats—an unexpected finding that was attributed
gesia [56,61]. Thus, while the severity of hypergly- to reduced spinal vascular volume and thus
cemia and duration of diabetes that are achieved reduced blood flow [65]. Whether this is the result
in rodent models of diabetes may not be sufficient of loss of blood vessels or vasoconstriction has yet
to cause oligodendrocyte death or demyelination, to be confirmed by detailed histopathological
there are indications of early molecular stresses studies.
on oligodendrocytes that may lead to both mye- A number of studies have measured expres-
lopathy and promote neuropathic pain. sion of glial fibrillary acidic protein (GFAP) as an
index of astrocyte number and health status.
Astrocytes Reduced total spinal GFAP levels and numbers
Astrocytes are distributed throughout the of GFAP expressing astrocytes have been
spinal cord, where they associate with neuro- reported in both white and gray matter of
nal synapses in the gray matter to modulate streptozotocin-induced type 1 diabetic rats
neurotransmitters and their metabolites. [70 72], though whether this reflects reduced
Astrocytic foot processes also regulate the fluid expression or cell loss is unclear. Conversely,
microenvironment of the spinal cord by increased total spinal GFAP protein, GFAP
ensheathing the endothelial cells and pericytes immunostaining intensity, number of GFAP
of the blood vessel wall, which together forms expressing cells, and altered astrocyte morphol-
the blood:brain barrier (BBB) and BSCB. ogy have been recorded in the dorsal horn of
Electron microscopy studies have suggested type 1 and type 2 diabetic mice and rats [73 77],
that diabetes causes retraction of astrocytic foot along with increased expression of nNOS and
processes from the other cells of the BBB [63]. iNOS [74]. Such findings are generally inter-
Expression of proteins associated with junc- preted as reflecting initiation of astrogliosis to fill
tions between adjacent endothelial cells, such the void of degenerated cells and/or activation
as occluding and VE-cadherin [64 66], is also of astrocytes [78] as part of a more generalized
reduced in the spinal cord of diabetic rats. At spinal neuroinflammatory state that, in concert
present, the majority of studies of astrocytes in with microglia [79], contributes to inflammatory-
diabetes have focused on the brain, which is like spinally mediated neuropathic pain [80].

Diabetic Neuropathy
110 7. Spinal cord involvement in diabetic neuropathy and neuropathic pain

Microglia pain has been supported by multiple studies

Microglia are the resident macrophages of showing that disruption or depletion of spi-
the CNS and are increasingly recognized as nal microglia can prevent or reverse indices
prominent orchestrators of the interactions of neuropathic pain in diabetic rodents
between neurons, astrocytes, and the immune [76,77,87 95]. Reactive microglia release
system in the brain [79] and spinal cord [81]. multiple factors that can influence spinal
They are versatile immune cells that possess sensory processing, including BDNF (via
properties that in the vascular immune system regulation of KCC2 as discussed above) and
are distributed across multiple specialized cell proinflammatory cytokines such as IL-1β and
types, thereby obviating the need for infiltra- TNFα, as recently reviewed [80].
tion of hematogenous immune cells into the A proportion of the activated microglia in
CNS in response to pathogens, tissue injury, or the spinal cord of diabetic mice may derive
other lesions. Microglial activation was initially from infiltrating blood-derived monocytes [92],
defined as a change in cell morphology with while infiltration of neutrophils into the gray
thickening and reduction in the length and matter of diabetic rats has also been reported
complexity of their ramified processes and an and was attributed to increased levels of L-
increase in cell body size/volume that repre- selectin, an adhesion molecule expressed by
sents a shift from environmental surveillance leukocytes that is involved in their migration
(ramified) to injury-evoked response (reactive) from blood vessels into tissue parenchyma
[82]. More recently, it has become recognized [96]. Given the physical restrictions placed on
that the reactive microglial population also expansion of the spinal cord due to the sur-
exists in a plastic spectrum spanning the two rounding bone, it is not clear whether infiltrat-
polarities: proinflammatory and phagocytic M1 ing cells are replacing other degenerating or
and anti-inflammatory/neuroprotective M2, departing cells to maintain overall spinal cord
and that shifts in population size and ratio can volume or precipitating increased intraspinal
both reflect and drive changing environmental pressure that itself could have pathological
conditions [83]. consequences.
Diabetes has been widely reported to acti-
vate resident microglia in the spinal cord of
rodents, with definition of what constitutes an
activated (or more accurately reactive, as rami- Clinical studies of spinally mediated pain
fied microglia are also inherently active) cell in diabetes
varying from the morphological [84,85], though
the presence of generic signals of cell activity Preclinical studies have suggested a variety
[84,86] to expression of specific marker pro- of mechanisms that may contribute to neuro-
teins that identify cells as being in a specific pathic pain arising from peripheral, spinal, and
region of the polarization spectrum. Early supraspinal sites in models of diabetes that
reports linked microglial activation by proin- have recently been reviewed in detail else-
flammatory cytokines with kinin B1 receptor where [97]. The remainder of this chapter will
expression in the genesis of spinally mediated therefore focus on clinical evidence of potential
pain in diabetic rats [87]. Microglia associate central mechanisms that influence output from
with synapses and have the potential to regu- nociceptive projection neurons in the dorsal
late and modify spinal sensory processing. horn of the spinal cord in painful diabetic
Involvement in the genesis or amplification of neuropathy.

Diabetic Neuropathy
 Clinical studies of spinally mediated pain in diabetes 111

The spinal cord and painful diabetic in both the amputated and the intact leg, which
neuropathy was closely followed by a diagnosis of type 2 dia-
betes [115]. This suggests that the neuropathic
Up to one-third of patients with diabetes, suffer pain, at least for the amputated limb, originated
with neuropathic pain [98 101] that often leads to from somewhere other than the perceived loca-
sleep disturbance, poor quality of life, anxiety/ tion of the pain.
depression, and unemployment [102 105]. Data
suggest that 10-year mortality is higher in patients
with severe chronic pain [106]. The only treatment
option is pain management, which typically Peripheral input
involves a trial and error approach of prescribing Disorders of the PNS can result in hyperexcit-
antineuropathic pain medications, which them- ability of nociceptors; those results in an
selves have inconsistent therapeutic benefit [107]. increased nociceptive drive to the spinal cord.
Major unmet needs exist for the development of However, this does not always equate with pain.
reliable biomarkers that both facilitate individual- Indeed, there is often a disconnect between the
ized treatment of neuropathic pain and inform degree of peripheral nociceptive drive and the
mechanistic-based drug discovery. severity of perceived pain. It is well known that
Pain is a complex and subjective physiological the severity of perceived pain depends upon
and psychological experience and is therefore many higher CNS factors including level of
inherently difficult to study and treat [108]. Why arousal, anxiety, depression, expectation, and
some patients develop pain yet others do not, anticipation [116]. There are also neural pro-
despite apparently equivalent severity of neuropa- cesses that are mediated by, or are intrinsic to,
thy, remains an unanswered conundrum [109]. the spinal cord that may suppress or enhance
The individual variability in pain perception ascending nociceptive projections.
poses additional challenges to assessing and treat- Primary afferent activity, whether as a result
ing pain [110,111]. Sural nerve biopsy studies in of activation at the peripheral endings or, aber-
patients with diabetes with or without pain show rant spontaneous activity due to peripheral
no associations between pain and degeneration nerve dysfunction, inputs into the spinal cord,
and/or regeneration of large or small fibers [112]. with the majority of afferents arbourising and
However, recent data indicate an association making synaptic contacts in the dorsal horn.
between loss of small fiber peripheral terminals Inputs from particular afferent types terminate
in subjects with both neuropathy and pain when in distinct Rexed laminae, with nociceptor
compared to those with equivalent neuropathy afferents terminating in the superficial laminae
but no pain suggesting an association between of the dorsal horn [117,118]. The sensory infor-
the degree of small fiber neuropathy and pain mation resulting from pain-related afferent
[113,114]. The association of pain with distal fiber input is processed in a complex dorsal horn
loss also supports the long held idea that pain interneuronal network and subsequently fur-
perceived as being in the extremities is generated ther integrated by projection neurons in the
by metabolic or physical injury in the extremities. superficial and deep laminae of the spinal cord
However, this is clearly not the whole story. In [119]. Thereafter, pain-related signals ascend in
1999 Rajbhandari and colleagues described the distinct spinal cord tracts (see Fig. 7.1), most
case of a patient who had undergone a below notably via the contralateral spinothalamic tract
knee amputation as a teenager because of an arte- to the thalamus and brain [120]. There are also
riovenous malformation. In his 50s, the patient direct projections to the brainstem and hypothal-
developed typical distal bilateral neuropathic pain amus via the spinoreticular/spinomesencephalic

Diabetic Neuropathy
112 7. Spinal cord involvement in diabetic neuropathy and neuropathic pain

and spinohypothalamic tracts, respectively [121]. from the periphery, a situation that is most likely
Signal amplification or loss of inhibition (disinhi- to occur in the context of an acute nerve injury,
bition) could potentially occur at any point along such a mechanism is also theoretically unlikely to
these spinal systems resulting in enhanced noci- prominent in a chronic neuropathy. We will
ceptive transmission to the brain. therefore focus on evidence linking the remaining
It is also important to appreciate that the three potential mechanisms to pain associated
spontaneous and evoked neuropathic pain felt with diabetic neuropathy.
by patients, whether due to pathology/dysfunc-
tion/alterations affecting the PNS or indeed Wind-up
CNS, is likely dependent on some form of Pain wind-up, which is related to the psy-
peripheral input [122,123]. This does not imply chophysical phenomenon of pain temporal
that the stimulus has to be abnormal peripheral summation, refers to the facilitation of nocire-
activity generated by sensitized nociceptors, but sponsive spinal cord neurons, including spi-
rather that the CNS, even in the case of CNS nothalamic projections neurons as a result of
damage, does not autonomously generate neural from repeated low frequency stimulation of c-
activity that is perceived as pain without some fiber nociceptors. It has been proposed as a
peripheral afferent input. potential contributing mechanism of allodynia
and hyperalgesia [124]. Both wind-up of spinal
cord neurons in rodents [125] and temporal sum-
mation in human subjects [126] are reduced by
Spinal mechanisms
NMDA receptor antagonists, although the effect
There are a number of theoretical mechan- is partial and other nonglutamatergic mechan-
isms by which peripheral inputs to the spinal isms may also contribute.
cord could be subject to amplification (or There have been few reported investigations
reduced suppression) including: addressing wind-up in preclinical models of
diabetes, with a sole report that repetitive elec-
• Glial activation within the dorsal horn of the
trical stimuli targeting c-nociceptor afferents in
spinal cord
diabetic mice enhanced the nociceptive with-
• Facilitation due to synaptic plasticity in the
drawal reflex—an intrinsic polysynaptic circuit
nociceptive dorsal horn
that mediates removal of a limb from a noxious
• Wind-up
stimulus [33]. Temporal summation is assessed
• A reduction of tonic spinal inhibition (spinal
in humans using repetitive noxious thermal,
disinhibition)
electrical, or mechanical stimuli. The pain
• Alterations in the descending modulation of
intensity rating typically of the 10th stimulus is
spinal cord processing and outputs
compared to that associated with the first stim-
Glial activation, as discussed above, is cur- ulus. No differences in mechanical temporal
rently restricted to an interesting series of preclini- summation have been demonstrated between
cal observations that have yet to be validated in patients with either nonpainful or painful dia-
humans. There is also no evidence suggesting betic neuropathy and healthy controls [127 129]
that synaptic facilitation due to long-term potenti- or between painful and nonpainful diabetic neu-
ation is of importance in enhancing ascending ropathy [129] suggesting that excessive wind-up
nociceptive outputs from the dorsal horn in diabe- is not a major mechanism underlying painful
tes. Since the induction of long-term potentiation symptoms in humans. However, patients with a
along the pain pathways is classically dependent shorter duration of painful diabetic neuropathy
on high frequency bursts of nociceptive input have significantly more pronounced temporal

Diabetic Neuropathy
 Clinical studies of spinally mediated pain in diabetes 113
summation than patients with a longer duration, of RDD overlapped with both healthy controls
despite equivalent pain severity, indicating that and patients with painless diabetic neuropathy.
the spinal mechanisms involved in initiating and These findings support the hypothesis that
maintaining pain may evolve over time [127] as impaired RDD may serve as clinical biomarker
part of what is sometimes termed the chronifica- in a subset of patients where pain arises primar-
tion of pain [130]. ily from spinal disinhibition. Furthermore, loss of
RDD correlated with pain severity indicating
Impaired tonic spinal inhibition that more severe pain is associated with the pres-
(spinal disinhibition) ence of loss of RDD and, by inference, spinal dis-
A discussed in detail above, there is accumu- inhibition. Expanding these data from patients
lating preclinical evidence that loss of RDD and with type 1 diabetes, we have recently demon-
indices of neuropathic pain share a common strated that impaired RDD is also seen in
pathogenic mechanism involving spinal KCC2 patients with type 2 diabetes and painful neu-
depletion and disinhibition caused by loss of ropathy (unpublished observations). Similar to
GABAA receptor inhibitory function and possi- the findings in type 1 diabetes, not all subjects
bly conversion to excitatory functions [49]. As with type 2 diabetes and neuropathic pain dem-
both RDD and behavioral indices of neuropathic onstrated impairment of RDD. Approximately
pain exhibit common responses to spinally acting 60% of patients with diabetes will develop neu-
analgesics, it has been suggested that RDD status ropathy, 30% of those with neuropathy will
may be a viable biomarker for identifying the develop neuropathic pain and, from our explor-
dominant generator site in individual patients atory study, 40% of those will show RDD defi-
with painful diabetic neuropathy and for predict- cits. In contrast, diabetic rodents exhibit much
ing efficacy of therapeutic strategies that alleviate more homogeneous neuropathy, neuropathic
spinal disinhibition. pain, and impaired RDD phenotypes [43]. While
Loss/impairment of RDD has been demon- this may reflect a more complex aetiopathogen-
strated in spinal cord injured patients, likely esis of painful diabetic neuropathy in humans, it
attributable to disinhibition of sensory proces- is also plausible that this heterogeneity can be
sing [36]. To assess the translational potential used to enable definition of abnormal values and
of the preclinical findings, we measured the predict therapeutic response to medications that
magnitude of H-reflex RDD in patients with target spinal inhibition.
type 1 diabetes and painful or painless neurop- It is important to acknowledge that a pro-
athy and age-matched healthy controls using a portion of patients in our clinical studies of
minor modification of the protocol for perform- RDD were taking prescribed antineuropathic
ing nerve conduction studies [48]. We found pain medication and that therapeutic responses
that in patients with painful diabetic neuropa- were not studied in a systematic manner.
thy, there was loss of RDD compared to both Similarly, it is not known if antineuropathic
healthy controls and patients with painless dia- pain medication affects or normalizes RDD
betic neuropathy. The impairment of RDD was either in a general or drug-specific manner.
independent of measures of both large and Further larger scale studies that control for the
small fiber neuropathy indicating that is not use of antineuropathic pain medication and/or
merely a reflection of severity of neuropathy. systematically assess therapeutic response are
Loss of RDD was also independent of glycemic required to validate the use of RDD both as a
control. Not all subjects with pain had impaired biomarker of spinal disinhibition and a predic-
RDD (defined as a H3:H1 ratio .2 3 SD of tor of neuropathic pain medication efficacy in
the control group mean) and indeed the degree patients with painful diabetic neuropathy.

Diabetic Neuropathy
114 7. Spinal cord involvement in diabetic neuropathy and neuropathic pain

While the method is noninvasive and poten- neurons of the deep spinal cord with receptive
tially widely available, one potential drawback fields heterotopic to the conditioning stimulus
to using RDD as a biomarker is that the [140]. This results in a reduction in ascending
H-reflex can be absent or difficult to elicit in a nociceptive drive. In humans, the efficacy of
proportion of patients with advanced diabetic the DNIC system can be evaluated using a
neuropathy [131]. Loss of the H-reflex can also number of psychophysical paradigms, broadly
be caused by other commonly encountered termed conditioned pain modulation (CPM).
conditions such as radiculopathies affecting the This involves triggering the DNIC system by
S1 nerve root [132]. These may limit the use of delivering a noxious stimulus distant to a con-
RDD in patients with severe neuropathy or in trol site. For instance, CPM can be easily initi-
those with coexisting lumbosacral root disease. ated by immersing one hand in painfully cold
ice water while assessing ratings for a painful
Descending modulation of spinal cord stimulus on the contralateral side [141].
processing and outputs Impairment of CPM, indicative of altered des-
It is important to not view processing within cending modulation, has been demonstrated in
the spinal cord in isolation. There are multiple patients with diabetes and painful neuropathy
means by which abnormalities in PNS firing/ [142]. Furthermore, treatment of painful dia-
metabolism impact upon the complex intercon- betic neuropathy with duloxetine, a serotonin-
nected circuits within the dorsal horn. norepinephrine reuptake inhibitor thought to
Conversely, there is increasing recognition that augment monoaminergic descending inhibi-
top down descending inputs to the spinal cord tory pain control, has been shown to be a more
from the descending pain modulatory system effective treatment in patients with reduced
(DPMS) can dynamically modulate ascending CPM [142]. Similarly, treatment with tapenta-
sensory information. The DPMS involves a dol, which is also implicated in the modulation
complex brainstem subcortical cortical net- of descending pain inhibitory pathways
work that projects to the spinal dorsal horn through both activation the μ-opioid receptor
and can have either a facilitatory or inhibitory activation and inhibition of neuronal norepi-
influence on ascending nociceptive input to the nephrine reuptake, improved pain in patients
brain [133 136]. The major outputs of the with painful diabetic neuropathy and impaired
DPMS arise from the periaqueductal gray situ- CPM [143]. Interestingly, treatment with dulox-
ated in the midbrain, which descend via mu- etine or tapentadol, in parallel with their bene-
opioid sensitive neurons in the rostral ventral ficial therapeutic effects on pain, also activated
medulla to the spinal cord [137 139]. A paral- CPM [142,143].
lel projection also arises from the locus coeru- As well as deficient inhibition, enhanced
leus [138,139]. These descending pathways facilitation could also result in an increase in
chiefly involve the monoanimergic neurotrans- ascending nociceptive drive from the spinal
mitters noradrenaline and 5-hydroxytryptamine cord. A recent study that examined subjects
and can have pronociceptive or antinociceptive with painful diabetic neuropathy by fMRI and
effects on dorsal horn processing according to quantitative sensory testing found that ventro-
the receptor that is activated [138,139]. lateral PAG functional connectivity is altered
One major descending inhibitory system of in patients suffering from painful diabetic neu-
interest is the diffuse noxious inhibitory con- ropathy [144]. Furthermore, the magnitude of
trols (DNICs). DNIC is triggered by a noxious ventrolateral PAG connectivity to the rostral
stimulus that results in the activation of des- anterior cingulate cortex correlated with the
cending controls that inhibit WDR projection degree of evoked pain in response to a tonic

Diabetic Neuropathy
 Clinical studies of spinally mediated pain in diabetes 115
thermonoxious stimulus, suggesting that there receptor subtypes expressed by these inter-
was a facilitation of nociceptive drive [144]. neurons [147]. These interneurons act within a
While the spinal cord was not directly complex network that can modulate projection
assessed, the importance of the ventrolateral neurons, other interneurons and primary affer-
PAG as a key node of the DPMS implies that ent fibers through GABAA-mediated primary
the findings might relate to enhanced descend- afferent depolarization [148 150]. There is evi-
ing facilitation of nociceptive drive form the dence that monoamines act, in part, by enhanc-
spinal cord and that this is influenced by high- ing GABAA transmission in the dorsal horn of
er cortical centers. the spinal cord. For instance, the application of
It is not known whether pronociceptive altera- 5-HT to spinal cord slices has been shown to
tions in the DPMS precede the development of enhance GABA release [151,152]. Furthermore,
pain. There are no published prospective studies serotonin acting through 5-HT2 receptors [153]
that address this issue and, furthermore, the cur- directly modulates GABAA by potentiating
rent literature is biased toward the testing of GABA evoked inward currents. The antinoci-
patients with painful rather than painless dia- ceptive effects of serotonin, acting through
betic neuropathy. However, the possibility that 5-HT3 receptors on WDR projection neurons in
the reduced capacity for patients with diabetic the dorsal horn are blocked by the GABAA
neuropathy to activate DNIC might reflect an antagonist bicuculline, implying that the
early phenomenon was suggested in a cross- monoaminergic inhibitory effects are mediated
sectional study in patients with varying duration indirectly via inhibitory interneurons [154].
of neuropathic pain symptoms [127]. Despite the Similarly, as well as mediating peripheral noci-
persistence of pain, CPM was shown to become ceptive inputs via presynaptic inhibition [155],
more efficient with a longer duration of diabetic noradrenaline also directly excites dorsal horn
neuropathy. CPM values in patients with a lon- GABAergic interneurons [156].
ger duration ( . 2 years) of diabetic neuropathy As discussed above, preclinical models of
were no longer pronociceptive and, indeed, simi- painful diabetic neuropathy indicate that spi-
lar to those seen in healthy control participants. nal inhibitory dysfunction is driven by GABA,
This effect that was unrelated to antineuropathic where this normally inhibitory neurotransmit-
pain medication. This suggests that the influence ter, acting via ionotropic GABAA receptors,
of DPMS on spinal cord nociceptive processing becomes pronociceptive by causing depolariza-
in painful diabetic neuropathy is dynamic and tion rather than hyperpolarization of the post-
may return to equilibrium over time. Whether synaptic neuron [49]. If there was significant
the dynamic alterations in the DPMS that either convergence of GABAergic circuits underlying
enhance or suppress inhibition in the dorsal horn DPMS and spinal disinhibition, it is conceiv-
of the spinal cord are linked to the mechanisms able that otherwise normally functioning des-
underlying painful diabetic neuropathy-related cending pain inhibitory pathways become
spinal inhibitory dysfunction is not known. facilitatory when acting in the context of spinal
Top-down monoaminergic modulation of disinhibition by enhancing depolarizing GABA
ascending spinal nociceptive projections is transmission. Since both DNIC and spinal dis-
likely to exhibit significant complexity due to inhibition can be easily measured in humans,
multiple parallel descending pathways acting using CPM and HRDD, respectively, this
on a multiplicity of targets in the dorsal horn. could provide a window of opportunity to
For example, there is a wide diversity of dorsal study the potentially convergent interactions
horn inhibitory and excitatory interneuron sub- in dorsal horn inhibitory pathways in patients
types [145,146] and the patterns of monoamine with DPN.

Diabetic Neuropathy
116 7. Spinal cord involvement in diabetic neuropathy and neuropathic pain

Conclusion [2] Williamson RT. Changes in the spinal cord in diabetes

mellitus. Br Med J 1904;1(2246):122 3.
[3] Olsson Y, Säve-Söderbergh J, Sourander P, Angervall L.
Rather than being merely a passive conduit A patho-anatomical study of the central and peripheral
for motor and sensory tracts, the spinal cord nervous system in diabetes of early onset and long
forms a critical sensorimotor processing inter- duration. Pathol Eur 1968;3(1):62 79.
face between the brain and the periphery. [4] Reske-Nielsen E, Lundbaek K. Pathological changes in
the central and peripheral nervous system of young
Despite its protection by physical and physio-
long-term diabetics. II. The spinal cord and peripheral
logical barriers, the spinal cord is exposed to, nerves. Diabetologia 1968;4(1):34 43.
and damaged by, fundamental pathogenic fea- [5] Reske-Nielsen E, Lundbaek K, Gregersen G, Harmsen A.
tures of diabetes such as impaired insulin sig- Pathological changes in the central and peripheral nervous
naling and hyperglycemia. Building on system of young long-term diabetics. The terminal neuro-
muscular apparatus. Diabetologia 1970;6(2):98 103.
traditional pathological and electrophysiologi-
[6] Slager UT. Diabetic myelopathy. Arch Pathol Lab Med
cal approaches, increasing preclinical and clini- 1978;102(9):467 9.
cal evidence emerging from application of new [7] Compston A. On diabetic neuritis, with a clinical and
technologies is strengthening the argument pathological description of three cases of diabetic
that pathophysiological modifications to the pseudo-tabes By T. Davies Pryce MRCS, consulting
surgeon to the Nottingham General Dispensary. Brain
processing of sensory inputs to the spinal cord
1893: 16; 416 424. With Diabetic neuropathy. A clini-
in diabetic neuropathy contribute to symptoms cal study of 150 cases. By M. Mencer Martin. From the
such as pain. Department of Diabetes, King’s College Hospital,
Future studies in patients with diabetic neu- Denmark Hill, London, S.E., UK. Brain 1953: 76;
ropathy that combine deep phenotyping of 594 624 Brain 2011;134(Pt 11):3137 40.
[8] Ziegler D, Mühlen H, Dannehl K, Gries FA. Tibial
neuropathy and neuropathic symptoms along
nerve somatosensory evoked potentials at various
with neurophysiological and psychophysical stages of peripheral neuropathy in insulin dependent
measurements of spinal cord function are diabetic patients. J Neurol Neurosurg Psychiatry
likely to provide further important pathophysi- 1993;56(1):58 64.
ological insights. This approach, supplemented [9] Suzuki C, Ozaki I, Tanosaki M, Suda T, Baba M,
Matsunaga M. Peripheral and central conduction
by advancements in noninvasive imaging of
abnormalities in diabetes mellitus. Neurology 2000;54
the human spinal cord using fMRI, may have (10):1932 7.
direct clinical application by allowing patients [10] Eaton SE, Harris ND, Rajbhandari SM, Greenwood P,
with pain mediated by spinal pathophysiologi- Wilkinson ID, Ward JD, et al. Spinal-cord involvement
cal alterations to be treated with therapies that in diabetic peripheral neuropathy. Lancet 2001;358
(9275):35 6.
selectively target these mechanisms.
[11] Selvarajah D, Wilkinson ID, Emery CJ, Harris ND,
Shaw PJ, Witte DR, et al. Early involvement of the spi-
nal cord in diabetic peripheral neuropathy. Diabetes
Acknowledgments Care 2006;29(12):2664 9.
Supported in part by award number 1-17-ICTS-062 from [12] Van Dam PS, Cotter MA, Bravenboer B, Cameron NE.
the American Diabetes Association (AGM, NAC). Pathogenesis of diabetic neuropathy: focus on neurovas-
cular mechanisms. Eur J Pharmacol 2013;719(1 3):180 6.
[13] Selvarajah D, Wilkinson ID, Davies J, Gandhi R, Tesfaye
S. Central nervous system involvement in diabetic neu-
References ropathy. Curr Diabetes Rep 2011;11(4):310 22.
[1] Shram NF, Netchiporouk LI, Martelet C, [14] Selvarajah D, Wilkinson ID, Maxwell M, Davies J,
JaffrezicRenault N, Cespuglio R. Brain glucose: vol- Sankar A, Boland E, et al. Magnetic resonance neuro-
tammetric determination in normal and hyperglycae- imaging study of brain structural differences in dia-
mic rats using a glucose microsensor. Neuroreport betic peripheral neuropathy. Diabetes Care 2014;
1997;8(5):1109 12. 37(6):1681 8.

Diabetic Neuropathy
 References 117
[15] Selvarajah D, Hughes T, Reeves J, Boland E, Marques J, [28] Ishii DN, Guertin DM, Whalen LR. Reduced insulin-
Gandhi R, et al. A preliminary study of brain macro- like growth-factor-I messenger-RNA content in liver,
vascular reactivity in impaired glucose tolerance and adrenal-glands and spinal-cord of diabetic rats.
type-2 diabetes: quantitative internal carotid artery Diabetologia 1994;37(11):1073 81.
blood flow using magnetic resonance phase contrast [29] Chen SR, Pan HL. Hypersensitivity of spinothalamic
angiography. Diab Vasc Dis Res 2016;13(5):367 72. tract neurons associated with diabetic neuropathic
[16] Selvarajah D, Wilkinson ID, Fang F, Sankar A, Davies J, pain in rats. J Neurophysiol 2002;87(6):2726 33.
Boland E, et al. Structural and functional abnormalities [30] Pertovaara A, Wei H, Kalmari J, Ruotsalainen M. Pain
of the primary somatosensory cortex in diabetic periph- behavior and response properties of spinal dorsal horn
eral neuropathy: a multimodal MRI study. Diabetes neurons following experimental diabetic neuropathy in
2019;68(4):796 806. the rat: modulation by nitecapone, a COMT inhibitor with
[17] Manor B, Newton E, Abduljalil A, Novak V. The rela- antioxidant properties. Exp Neurol 2001;167(2):425 34.
tionship between brain volume and walking outcomes [31] Schuelert N, Gorodetskaya N, Just S, Doods H,
in older adults with and without diabetic peripheral Corradini L. Electrophysiological characterization of
neuropathy. Diabetes Care 2012;35(9):1907 12. spinal neurons in different models of diabetes type 1-
[18] Biessels GJ, Bril V, Calcutt NA, Cameron NE, Cotter and type 2-induced neuropathy in rats. Neuroscience
MA, Dobrowsky R, et al. Phenotyping animal models 2015;291:146 54.
of diabetic neuropathy: a consensus statement of the [32] Tan AM, Samad OA, Fischer TZ, Zhao P, Persson AK,
diabetic neuropathy study group of the EASD Waxman SG. Maladaptive dendritic spine remodeling
(Neurodiab). J Peripheral Nerv Syst 2014;19(2):77 87. contributes to diabetic neuropathic pain. J Neurosci
[19] Jin HY, Moon SS, Calcutt NA. Lost in translation? 2012;32(20):6795 807.
Measuring diabetic neuropathy in humans and ani- [33] Kimura S, Tanabe M, Honda M, Ono H. Enhanced
mals. Diabetes Metab J 2021;45(1):27 42. wind-up of the C-fiber-mediated nociceptive flexor
[20] Yagihashi S, Zhang WX, Sima AA. Neuroaxonal dys- reflex movement following painful diabetic neuropa-
trophy in distal symmetric sensory polyneuropathy of thy in mice. J Pharmacol Sci 2005;97(2):195 202.
the diabetic BB-rat. J Diabet Complications 1989;3 [34] Jolivalt CG, Frizzi KE, Guernsey L, Marquez A, Ochoa J,
(4):202 10. Rodriguez M, et al. Peripheral neuropathy in mouse
[21] Sima AA, Yagihashi S. Central-peripheral distal axo- models of diabetes. Curr Protoc Mouse Biol 2016;6
nopathy in the spontaneously diabetic BB-rat: ultra- (3):223 55.
structural and morphometric findings. Diabetes Res [35] Ho SM, Waite PM. Effects of different anesthetics on
Clin Pract 1985;5:289 98. the paired-pulse depression of the h reflex in adult rat.
[22] Jakobsen J. Axonal dwindling in early experimental dia- Exp Neurol 2002;177(2):494 502.
betes 0.1. study of cross sectioned nerves. Diabetologia [36] Ishikawa K, Ott K, Porter RW, Stuart D. Low fre-
1976;12(6):539 46. quency depression of the H wave in normal and spi-
[23] Jakobsen J. Early and preventable changes of nal man. Exp Neurol 1966;15(1):140 56.
peripheral-nerve structure and function in insulin- [37] Lloyd DP, Wilson VJ. Reflex depression in rhythmi-
deficient diabetic rats. J Neurol Neurosur Ps 1979;42 cally active monosynaptic reflex pathways. J Gen
(6):509 18. Physiol 1957;40(3):409 26.
[24] Ramji N, Toth C, Kennedy J, Zochodne DW. Does dia- [38] Angel RW, Hofmann WW. The H reflex in normal,
betes mellitus target motor neurons? Neurobiol Dis spastic, and rigid subjects. Arch Neurol 1963;9:591 6.
2007;26(2):301 11. [39] Metz J, Goode DJ, Meltzer HY. Descriptive studies of
[25] Biessels GJ, Cristino NA, Rutten GJ, Hamers FPT, H-reflex recovery curves in psychiatric patients.
Erkelens DW, Gispen WH. Neurophysiological changes Psychol Med 1980;10(3):541 8.
in the central and peripheral nervous system of [40] Sabbahi M, Etnyre B, Al-Jawayed I, Jankovic J. H-
streptozotocin-diabetic rats—course of development and reflex recovery curves differentiate essential tremor,
effects of insulin treatment. Brain 1999;122:757 68. Parkinson’s disease, and the combination of essential
[26] Terada MYHK R, Koyama N, Yokota T, Shigeta Y. tremor and Parkinson’s disease. J Clin Neurophysiol
Central-peripheral distal axonopathy in the spontane- 2002;19(3):245 51.
ously diabetic BB-rat: ultrastructural and morphomet- [41] Matsushita A, Smith CM. Spinal cord function in postis-
ric findings. J Neurol Sci 1993;115:58 66. chemic rigidity in the rat. Brain Res 1970;19(3):395 410.
[27] Carsten RE, Whalen LR, Ishii DN. Impairment of [42] Kakinohana O, Hefferan MP, Nakamura S,
spinal-cord conduction-velocity in diabetic rats. Kakinohana M, Galik J, Tomori Z, et al. Development
Diabetes 1989;38(6):730 6. of GABA-sensitive spasticity and rigidity in rats after

Diabetic Neuropathy
118 7. Spinal cord involvement in diabetic neuropathy and neuropathic pain

transient spinal cord ischemia: a qualitative and quan- [55] Tomlinson DR, Robinson JP, Willars GB, Keen P. Deficient
titative electrophysiological and histopathological axonal-transport of substance-P in streptozocin-induced
study. Neuroscience 2006;141(3):1569 83. diabetic rats—effects of sorbinil and insulin. Diabetes
[43] Jolivalt CG, Lee CA, Ramos KM, Calcutt NA. 1988;37(4):488 93.
Allodynia and hyperalgesia in diabetic rats are medi- [56] Freshwater JD, Svensson CI, Malmberg AB, Calcutt
ated by GABA and depletion of spinal potassium- NA. Elevated spinal cyclooxygenase and prostaglan-
chloride co-transporters. Pain 2008;140(1):48 57. din release during hyperalgesia in diabetic rats.
[44] Lee-Kubli CA, Zhou X, Jolivalt CG, Calcutt NA. Diabetes 2002;51(7):2249 55.
Pharmacological modulation of rate-dependent depression [57] Calcutt NA, Stiller C, Gustafsson H, Malmberg AB.
of the spinal h-reflex predicts therapeutic efficacy against Elevated substance-P-like immunoreactivity levels in
painful diabetic neuropathy. Diagnostics 2021;11:283 97. spinal dialysates during the formalin test in normal
[45] Hernandez-Reyes JE, Salinas-Abarca AB, Vidal-Cantu and diabetic rats. Brain Res 2000;856(1 2):20 7.
GC, Raya-Tafolla G, Elias-Vinas D, Granados-Soto V, [58] Lupachyk S, Shevalye H, Maksimchyk Y, Drel VR,
et al. alpha(5)GABA(A) receptors play a pronocicep- Obrosova IG. PARP inhibition alleviates diabetes-
tive role and avoid the rate-dependent depression of induced systemic oxidative stress and neural tissue 4-
the Hoffmann reflex in diabetic neuropathic pain and hydroxynonenal adduct accumulation: correlation
reduce primary afferent excitability. Pain 2019; with peripheral nerve function. Free Radic Biol Med
160(6):1448 58. 2011;50(10):1400 9.
[46] Jolivalt CG, Rodriguez M, Wahren J, Calcutt NA. [59] Pesaresi M, Giatti S, Calabrese D, Maschi O, Caruso D,
Efficacy of a long-acting C-peptide analogue against Melcangi RC. Dihydroprogesterone increases the gene
peripheral neuropathy in streptozotocin-diabetic mice. expression of myelin basic protein in spinal cord of
Diabetes Obes Metab 2015;17(8):781 8. diabetic rats. J Mol Neurosci 2010;42(2):135 9.
[47] Lee-Kubli CA, Calcutt NA. Altered rate-dependent [60] Jiang Y, Calcutt NA, Ramos KM, Mizisin AP. Novel
depression of the spinal H-reflex as an indicator of sites of aldose reductase immunolocalization in nor-
spinal disinhibition in models of neuropathic pain. mal and streptozotocin-diabetic rats. J Peripher Nerv
Pain 2014;155(2):250 60. Syst 2006;11(4):274 85.
[48] Marshall AG, Lee-Kubli C, Azmi S, Zhang M, [61] Ramos KM, Jiang Y, Svensson CI, Calcutt NA.
Ferdousi M, Mixcoatl-Zecuatl T, et al. Spinal disinhibi- Pathogenesis of spinally mediated hyperalgesia in dia-
tion in experimental and clinical painful diabetic neu- betes. Diabetes 2007;56(6):1569 76.
ropathy. Diabetes 2017;66(5):1380 90. [62] Oates PJ. Aldose reductase, still a compelling target
[49] Lee-Kubli C, Marshall AG, Malik RA, Calcutt NA. The for diabetic neuropathy. Curr Drug Targets 2008;9
H-reflex as a biomarker for spinal disinhibition in (1):14 36.
painful diabetic neuropathy. Curr Diabetes Rep [63] Salameh TS, Shah GN, Price TO, Hayden MR, Banks
2018;18(1):1. WA. Blood-brain barrier disruption and neurovascular
[50] Coull JA, Boudreau D, Bachand K, Prescott SA, Nault unit dysfunction in diabetic mice: protection with the
F, Sik A, et al. Trans-synaptic shift in anion gradient in mitochondrial carbonic anhydrase inhibitor topira-
spinal lamina I neurons as a mechanism of neuro- mate. J Pharmacol Exp Ther 2016;359(3):452 9.
pathic pain. Nature 2003;424(6951):938 42. [64] Hawkins BT, Lundeen TF, Norwood KM, Brooks HL,
[51] Doyon N, Ferrini F, Gagnon M, De Koninck Y. Egleton RD. Increased blood-brain barrier permeabil-
Treating pathological pain: is KCC2 the key to the ity and altered tight junctions in experimental diabetes
gate? Expert Rev Neurother 2013;13(5):469 71. in the rat: contribution of hyperglycaemia and matrix
[52] Mapplebeck JCS, Lorenzo LE, Lee KY, Gauthier C, metalloproteinases. Diabetologia 2007;50(1):202 11.
Muley MM, De Koninck Y, et al. Chloride dysregula- [65] Ved N, Lobo MED, Bestall SM, Vidueira CL, Beazley-
tion through downregulation of KCC2 mediates neu- Long N, Ballmer-Hofer K, et al. Diabetes-induced
ropathic pain in both sexes. Cell Rep 2019;28(3):590. microvascular complications at the level of the spinal
[53] Malmberg AB, O’Connor WT, Glennon JC, Cesena R, cord: a contributing factor in diabetic neuropathic
Calcutt NA. Impaired formalin-evoked changes of spi- pain. J Physiol-London 2018;596(16):3675 93.
nal amino acid levels in diabetic rats. Brain Res [66] Zhao YW, Liu XY, Yu AY, Zhou YJ, Liu BJ. Diabetes-
2006;1115(1):48 53. related alteration of occludin expression in rat blood-spinal
[54] Calcutt NA, Chen P, Hua XY. Effects of diabetes on cord barrier. Cell Biochem Biophys 2010;58(3):141 5.
tissue content and evoked release of calcitonin gene- [67] Hardigan T, Ward R, Ergul A. Cerebrovascular com-
related peptide-like immunoreactivity from rat sen- plications of diabetes: focus on cognitive dysfunction.
sory nerves. Neurosci Lett 1998;254(3):129 32. Clin Sci 2016;130(20):1807 22.

Diabetic Neuropathy
 References 119
[68] Banks W. The blood-brain barrier interface in diabetes [80] Tsuda M. Microglia-mediated regulation of neuro-
mellitus: dysfunctions, mechanisms and approaches to pathic pain: molecular and cellular mechanisms. Biol
treatment. Curr Pharm Des 2020;26(13):1438 47. Pharm Bull 2019;42(12):1959 68.
[69] Fujihara R, Chiba Y, Nakagawa T, Nishi N, Murakami R, [81] Borjini N, Paouri E, Tognatta R, Akassoglou K,
Matsumoto K, et al. Albumin microvascular leakage in Davalos D. Imaging the dynamic interactions between
brains with diabetes mellitus. Microsc Res Techniq immune cells and the neurovascular interface in the
2016;79(9):833 7. spinal cord. Exp Neurol 2019;322. Available from:
[70] Afsari ZH, Renno WM, Abd-El-Basset E. Alteration of https://doi.org/10.1016/j.expneurol.2019.113046.
glial fibrillary acidic proteins immunoreactivity in [82] Streit WJ, Graeber MB, Kreutzberg GW. Functional
astrocytes of the spinal cord diabetic rats. Anat Rec plasticity of microglia—a review. Glia 1988;1(5):301 7.
2008;291(4):390 9. [83] Mesquida-Veny F, Del Rio JA, Hervera A.
[71] Renno WM, Alkhalaf M, Afsari Z, Abd-El-Basset E, Macrophagic and microglial complexity after neuronal
Mousa A. Consumption of green tea alters glial fibrili- injury. Prog Neurobiol 2020;200. Available from:
ary acidic protein immunoreactivity in the spinal cord https://doi.org/10.1016/j.pneurobio.2020.101970.
astrocytes of STZ-diabetic rats. Nutr Neurosci 2008;11 [84] Tsuda M, Ueno H, Kataoka A, Tozakif-Saitoh H, Inoue K.
(1):32 40. Activation of Dorsal Horn Microglia contributes to
[72] Shayea AMF, Mousa AMA, Renno WM, Nadar MS, diabetes-induced tactile allodynia via extracellular signal-
Qabazard B, Yousif MHM. Chronic treatment with regulated protein kinase signaling. Glia 2008;56(4):378 86.
hydrogen sulfide donor GYY4137 mitigates microglial [85] Wodarski R, Clark AK, Grist J, Marchand F,
and astrocyte activation in the spinal cord of Malcangio M. Gabapentin reverses microglial activa-
streptozotocin-induced diabetic rats. J Neuropath Exp tion in the spinal cord of streptozotocin-induced dia-
Neur 2020;79(12):1320 43. betic rats. Eur J Pain 2009;13(8):807 11.
[73] Barragan-Iglesias P, Oidor-Chan VH, Loeza-Alcocer E, [86] Daulhac L, Mallet C, Courteix C, Etienne M, Duroux E,
Pineda-Farias JB, Velazquez-Lagunas I, Salinas-Abarca Privat AM, et al. Diabetes-induced mechanical hyperal-
AB, et al. Evaluation of the neonatal streptozotocin gesia involves spinal mitogen-activated protein kinase
model of diabetes in rats: evidence for a model of neu- activation in neurons and microglia via N-methyl-D-
ropathic pain. Pharmacol Rep 2018;70(2):294 303. aspartate-dependent mechanisms. Mol Pharmacol
[74] Dauch JR, Yanik BM, Hsieh W, Oh SS, Cheng HLT. 2006;70(4):1246 54.
Neuron-astrocyte signaling network in spinal cord [87] Talbot S, Chahmi E, Dias JP, Couture R. Key role for spi-
dorsal horn mediates painful neuropathy of type 2 nal dorsal horn microglial kinin B1 receptor in early dia-
diabetes. Glia 2012;60(9):1301 15. betic pain neuropathy. J Neuroinflammat 2010;7(1):36.
[75] El-Sherbeeny NA, Ibrahiem AT, Ali HS, Farag NE, [88] Koshimizu H, Ohkawara B, Nakashima H, Ota K,
Toraih EA, Zaitone SA. Carbamazepine conquers spi- Kanbara S, Inoue T, et al. Zonisamide ameliorates neu-
nal GAP43 deficiency and sciatic Nav1.5 upregulation ropathic pain partly by suppressing microglial activa-
in diabetic mice: novel mechanisms in alleviating allo- tion in the spinal cord in a mouse model. Life Sci
dynia and hyperalgesia. Arch Pharm Res 2020;43 2020;263. Available from: https://doi.org/10.1016/j.
(7):724 34. lfs.2020.118577.
[76] Reda HM, Zaitone SA, Moustafa YM. Effect of levetir- [89] Liu M, Gao LL, Zhang N. Berberine reduces neuro-
acetam vs gabapentin on peripheral neuropathy and glia activation and inflammation in streptozotocin-
sciatic degeneration in streptozotocin-diabetic mice: induced diabetic mice. Int J Immunopathol
influence on spinal microglia and astrocytes. Eur J Pharmacol 2019;33. Available from: https://doi.org/
Pharmacol 2016;771:162 72. 10.1177/2058738419866379.
[77] Tawfik MK, Helmy SA, Badran DI, Zaitone SA. [90] Morgado C, Pereira-Terra P, Cruz CD, Tavares I.
Neuroprotective effect of duloxetine in a mouse model Minocycline completely reverses mechanical hyperal-
of diabetic neuropathy: role of glia suppressing gesia in diabetic rats through microglia-induced
mechanisms. Life Sci 2018;205:113 24. changes in the expression of the potassium chloride
[78] Acioglu C, Li L, Elkabes S. Contribution of astrocytes co-transporter 2 (KCC2) at the spinal cord. Diabetes
to neuropathology of neurodegenerative diseases. Obes Metab 2011;13(2):150 9.
Brain Res 2021;1758. Available from: https://doi.org/ [91] Pabreja K, Dua K, Sharma S, Padi SSV, Kulkarni SK.
10.1016/j.brainres.2021.147291. Minocycline attenuates the development of diabetic
[79] Kwon HS, Koh SH. Neuroinflammation in neurode- neuropathic pain: possible anti-inflammatory and
generative disorders: the roles of microglia and astro- anti-oxidant mechanisms. Eur J Pharmacol 2011;661
cytes. Transl Neurodegen 2020;9(1):42 54. (1 3):15 21.

Diabetic Neuropathy
120 7. Spinal cord involvement in diabetic neuropathy and neuropathic pain

[92] Sun JJ, Tang L, Zhao XP, Xu JM, Xiao Y, Li H. peripheral neuropathy: the patients’ perspectives. J
Infiltration of blood-derived macrophages contri- Pain 2006;7(12):892 900.
butes to the development of diabetic neuropathy. J [104] Tölle T, Xu X, Sadosky AB. Painful diabetic neuropathy: a
Immunol Res 2019;2019. Available from: https://doi. cross-sectional survey of health state impairment and
org/10.1155/2019/7597382. treatment patterns. J Diabetes Complicat 2006;20(1):26 33.
[93] Sun JS, Yang YJ, Zhang YZ, Huang W, Li ZS, Zhang Y. [105] Zelman DC, Brandenburg NA, Gore M. Sleep
Minocycline attenuates pain by inhibiting spinal micro- impairment in patients with painful diabetic periph-
glia activation in diabetic rats. Mol Med Rep 2015;12 eral neuropathy. Clin J Pain 2006;22(8):681 5.
(2):2677 82. [106] Torrance N, Elliott AM, Lee AJ, Smith BH. Severe
[94] Suzuki N, Hasegawa-Moriyama M, Takahashi Y, chronic pain is associated with increased 10 year
Kamikubo Y, Sakurai T, Inada E. Lidocaine attenu- mortality. A cohort record linkage study. Eur J Pain
ates the development of diabetic-induced tactile allo- 2010;14(4):380 6.
dynia by inhibiting microglial activation. Anesth [107] Alam U, Sloan G, Tesfaye S. Treating pain in diabetic
Analg 2011;113(4):941 6. neuropathy: current and developmental drugs.
[95] Zhang TT, Xue R, Fan SY, Fan QY, An L, Li J, et al. Drugs 2020;80(4):363 84.
Ammoxetine attenuates diabetic neuropathic pain [108] Mackey S, Greely HT, Martucci KT. Neuroimaging-
through inhibiting microglial activation and neuroin- based pain biomarkers: definitions, clinical and research
flammation in the spinal cord. J Neuroinflamm 2018;15. applications, and evaluation frameworks to achieve per-
Available from: https://doi.org/10.1186/s12974-018- sonalized pain medicine. PAIN Rep 2019;4(4):e762.
1216-3. [109] Finnerup N, Sindrup S, Jensen T. Management of pain-
[96] Newton VL, Guck JD, Cotter MA, Cameron NE, ful neuropathies. Handb Clin Neurol 2013;279 90.
Gardiner NJ. Neutrophils infiltrate the spinal cord [110] Price DD, McGrath PA, Rafii A, Buckingham B. The
parenchyma of rats with experimental diabetic neu- validation of visual analogue scales as ratio scale
ropathy. J Diabetes Res 2017;2017. Available from: measures for chronic and experimental pain. Pain
https://doi.org/10.1155/2017/4729284. 1983;17(1):45 56.
[97] Calcutt NA. Diabetic neuropathy and neuropathic [111] Fillingim RB. Individual differences in pain: under-
pain: a (con)fusion of pathogenic mechanisms? Pain standing the mosaic that makes pain personal. Pain
2020;161(9):S65 86. 2017;158(Suppl. 1):S11 s8.
[98] Davies M, Brophy S, Williams R, Taylor A. The prev- [112] Llewelyn JG, Thomas PK, Gilbey SG, Watkins PJ,
alence, severity, and impact of painful diabetic Muddle JR. Pattern of myelinated fibre loss in the
peripheral neuropathy in type 2 diabetes. Diabetes sural nerve in neuropathy related to type 1 (insulin-
Care 2006;29(7):1518 22. dependent) diabetes. Diabetologia 1988;31(3):162 7.
[99] Veves A, Backonja M, Malik RA. Painful diabetic [113] Kalteniece A, Ferdousi M, Petropoulos I, Azmi S,
neuropathy: epidemiology, natural history, early Adam S, Fadavi H, et al. Greater corneal nerve loss at
diagnosis, and treatment options. Pain Med 2008;9 the inferior whorl is related to the presence of dia-
(6):660 74. betic neuropathy and painful diabetic neuropathy.
[100] Abbott CA, Malik RA, van Ross ER, Kulkarni J, Sci Rep 2018;8(1):3283.
Boulton AJ. Prevalence and characteristics of painful [114] Kalteniece A, Ferdousi M, Azmi S, Mubita WM,
diabetic neuropathy in a large community-based dia- Marshall A, Lauria G, et al. Corneal confocal micros-
betic population in the U.K. Diabetes Care 2011;34 copy detects small nerve fibre damage in patients with
(10):2220 4. painful diabetic neuropathy. Sci Rep 2020;10(1):3371.
[101] Alleman CJ, Westerhout KY, Hensen M, Chambers [115] Rajbhandari SM, Jarratt JA, Griffiths PD, Ward JD.
C, Stoker M, Long S, et al. Humanistic and economic Diabetic neuropathic pain in a leg amputated 44
burden of painful diabetic peripheral neuropathy in years previously. Pain 1999;83(3):627 9.
Europe: a review of the literature. Diabetes Res Clin [116] Brooks J, Tracey I. From nociception to pain percep-
Pract 2015;109(2):215 25. tion: imaging the spinal and supraspinal pathways. J
[102] Gore M, Brandenburg NA, Dukes E, Hoffman DL, Anat 2005;207(1):19 33.
Tai KS, Stacey B. Pain severity in diabetic peripheral [117] Todd AJ. Neuronal circuitry for pain processing in the
neuropathy is associated with patient functioning, dorsal horn. Nat Rev Neurosci 2010;11(12):823 36.
symptom levels of anxiety and depression, and sleep. [118] Abraira VE, Ginty DD. The sensory neurons of touch.
J Pain Symptom Manage 2005;30(4):374 85. Neuron 2013;79(4):618 39.
[103] Gore M, Brandenburg NA, Hoffman DL, Tai KS, [119] Koch SC, Acton D, Goulding M. Spinal circuits for touch,
Stacey B. Burden of illness in painful diabetic pain, and itch. Annu Rev Physiol 2018;80:189 217.

Diabetic Neuropathy
 References 121
[120] Craig AD. Pain mechanisms: labeled lines vs conver- [133] Basbaum AI, Fields HL. Endogenous pain control
gence in central processing. Annu Rev Neurosci mechanisms: review and hypothesis. Ann Neurol
2003;26:1 30. 1978;4(5):451 62.
[121] Willis WD, Westlund KN. Neuroanatomy of the pain [134] Porreca F, Ossipov MH, Gebhart GF. Chronic pain
system and of the pathways that modulate pain. J and medullary descending facilitation. Trends
Clin Neurophysiol 1997;14(1):2 31. Neurosci 2002;25(6):319 25.
[122] Haroutounian S, Nikolajsen L, Bendtsen TF, Finnerup [135] Gebhart GF. Descending modulation of pain.
NB, Kristensen AD, Hasselstrøm JB, et al. Primary affer- Neurosci Biobehav Rev 2004;27(8):729 37.
ent input critical for maintaining spontaneous pain in [136] Mason P. Medullary circuits for nociceptive modula-
peripheral neuropathy. Pain 2014;155(7):1272 9. tion. Curr Opin Neurobiol 2012;22(4):640 5.
[123] Haroutounian S, Ford AL, Frey K, Nikolajsen L, [137] Bannister K, Patel R, Goncalves L, Townson L,
Finnerup NB, Neiner A, et al. How central is central Dickenson AH. Diffuse noxious inhibitory controls
poststroke pain? The role of afferent input in post- and nerve injury: restoring an imbalance between
stroke neuropathic pain: a prospective, open-label descending monoamine inhibitions and facilitations.
pilot study. Pain 2018;159(7):1317 24. Pain 2015;156(9):1803 11.
[124] Herrero JF, Laird JMA, Lopez-Garcia JA. Wind-up of [138] Bannister K, Dickenson AH. What do monoamines
spinal cord neurones and pain sensation: much ado do in pain modulation? Curr Opin Support Palliat
about something? Prog Neurobiol 2000;61 Care 2016;10(2):143 8.
(2):169 203. [139] Bannister K, Dickenson AH. What the brain tells the
[125] Dickenson AH, Sullivan AF. Evidence for a role of the spinal cord. Pain 2016;157(10):2148 51.
NMDA receptor in the frequency-dependent potentiation [140] Le Bars D, Dickenson AH, Besson JM. Diffuse nox-
of deep rat dorsal horn nociceptive neurons following C- ious inhibitory controls (DNIC). I. Effects on dorsal
fiber stimulation. Neuropharmacology 1987;26(8):1235 8. horn convergent neurones in the rat. Pain 1979;6
[126] Warncke T, Stubhaug A, Jorum E. Ketamine, an (3):283 304.
NMDA receptor antagonist, suppresses spatial and [141] Yarnitsky D. Conditioned pain modulation (the dif-
temporal properties of burn-induced secondary hyper- fuse noxious inhibitory control-like effect): its rele-
algesia in man: a double-blind, cross-over comparison vance for acute and chronic pain states. Curr
with morphine and placebo. Pain 1997;72(1 2):99 106. Opanaesthesiol 2010;23(5):611 15.
[127] Granovsky Y, Nahman-Averbuch H, Khamaisi M, [142] Yarnitsky D, Granot M, Nahman-Averbuch H,
Granot M. Efficient conditioned pain modulation Khamaisi M, Granovsky Y. Conditioned pain modu-
despite pain persistence in painful diabetic neuropa- lation predicts duloxetine efficacy in painful diabetic
thy. Pain Rep 2017;2(3):e592. neuropathy. Pain 2012;153(6):1193 8.
[128] Softeland E, Brock C, Frokjaer JB, Brogger J, [143] Niesters M, Proto PL, Aarts L, Sarton EY, Drewes
Madacsy L, Gilja OH, et al. Association between AM, Dahan A. Tapentadol potentiates descending
visceral, cardiac and sensorimotor polyneuropa- pain inhibition in chronic pain patients with diabetic
thies in diabetes mellitus. J Diabetes Complicat polyneuropathy. Br J Anaesth 2014;113(1):148 56.
2014;28(3):370 7. [144] Segerdahl AR, Themistocleous AC, Fido D, Bennett
[129] Themistocleous AC, Ramirez JD, Shillo PR, Lees JG, DL, Tracey I. A brain-based pain facilitation mecha-
Selvarajah D, Orengo C, et al. The Pain in nism contributes to painful diabetic polyneuropathy.
Neuropathy Study (PiNS): a cross-sectional observa- Brain 2018;141(2):357 64.
tional study determining the somatosensory pheno- [145] Abraira VE, Kuehn ED, Chirila AM, Springel MW,
type of painful and painless diabetic neuropathy. Toliver AA, Zimmerman AL, et al. The cellular and
Pain 2016;157(5):1132 45. synaptic architecture of the mechanosensory dorsal
[130] Ossipov MH, Morimura K, Porreca F. Descending horn. Cell 2017;168(1 2):295 310. e19.
pain modulation and chronification of pain. Curr [146] Hughes DI, Todd AJ. Central nervous system tar-
Opin Support Palliat Care 2014;8(2):143 51. gets: inhibitory interneurons in the spinal cord.
[131] Millan-Guerrero R, Trujillo-Hernandez B, Isais- Neurotherapeutics 2020;17(3):874 85.
Millan S, Prieto-Diaz-Chavez E, Vasquez C, [147] Bardoni R. Serotonergic modulation of nociceptive cir-
Caballero-Hoyos JR, et al. H-reflex and clinical exam- cuits in spinal cord dorsal horn. Curr Neuropharmacol
ination in the diagnosis of diabetic polyneuropathy. J 2019;17(12):1133 45.
Int Med Res 2012;40(2):694 700. [148] Rudomin P, Schmidt RF. Presynaptic inhibition in
[132] Jerath N, Kimura J. F wave, A wave, H reflex, and the vertebrate spinal cord revisited. Exp Brain Res
blink reflex. Handb Clin Neurol 2019;160:225 39. 1999;129(1):1 37.

Diabetic Neuropathy
122 7. Spinal cord involvement in diabetic neuropathy and neuropathic pain

[149] Bardoni R, Takazawa T, Tong CK, Choudhury P, aminobutyric acid. Brain Res Bull 2000;52
Scherrer G, Macdermott AB. Pre- and postsynaptic (6):559 65.
inhibitory control in the spinal cord dorsal horn. Ann [154] Alhaider AA, Lei SZ, Wilcox GL. Spinal 5-HT3 recep-
N Y Acad Sci 2013;1279:90 6. tor-mediated antinociception: possible release of
[150] Betelli C, MacDermott AB, Bardoni R. Transient, activ- GABA. J Neurosci 1991;11(7):1881 8.
ity dependent inhibition of transmitter release from [155] Kawasaki Y, Kumamoto E, Furue H, Yoshimura M.
low threshold afferents mediated by GABAA receptors Alpha 2 adrenoceptor-mediated presynaptic inhibi-
in spinal cord lamina III/IV. Mol Pain 2015;11:64. tion of primary afferent glutamatergic transmission
[151] Abe K, Kato G, Katafuchi T, Tamae A, Furue H, in rat substantia gelatinosa neurons. Anesthesiology
Yoshimura M. Responses to 5-HT in morphologically 2003;98(3):682 9.
identified neurons in the rat substantia gelatinosa [156] Gassner M, Ruscheweyh R, Sandkühler J. Direct exci-
in vitro. Neuroscience 2009;159(1):316 24. tation of spinal GABAergic interneurons by nor-
[152] Xie DJ, Uta D, Feng PY, Wakita M, Shin MC, Furue H, adrenaline. Pain 2009;145(1 2):204 10.
et al. Identification of 5-HT receptor subtypes enhanc- [157] Ved N, Da Vittoria Lobo ME, Bestall SM, Vidueira C,
ing inhibitory transmission in the rat spinal dorsal Beazley-Long N, Ballmer-Hofer K, et al. Diabetes-
horn in vitro. Mol Pain 2012;8:58. induced microvascular complications at the level of
[153] Li H, Lang B, Kang JF, Li YQ. Serotonin potenti- the spinal cord: a contributing factor in diabetic neu-
ates the response of neurons of the superficial ropathic pain. Journal of Physiology 2018;596
laminae of the rat spinal dorsal horn to gamma- (16):3675 93.

Diabetic Neuropathy
 C H A P T E R

8
The conundrum and enigma of painful
and painless neuropathy
Johan Røikjer1,2 and Niels Ejskjaer1,3
1
Steno Diabetes Center North Denmark, Aalborg University Hospital, Aalborg, Denmark 2Center for
Sensory-Motor Interaction, Aalborg University, Aalborg, Denmark 3Departments of Endocrinology and
Clinical Medicine, Aalborg University Hospital, Aalborg, Denmark

Introduction in people with diabetes with a lifetime preva-

lence of around 50% [6] and it is associated
The conundrum of diabetic painful and pain- with diabetic painful neuropathy in up to 40%
less neuropathies was clinically described centu- of all cases [7,8]. Presence of diabetic painful
ries ago [1,2], more lately in 1983 on the natural neuropathy without simultaneous peripheral
history of acute painful neuropathy [3], and sensory neuropathy is practically nonexisting
then again in 1992 as follows: “Diabetic neurop- or very rare—and preexisting sensory neuro-
athies form a group of diverse conditions: it is pathy, therefore, seems to be a prerequisite for
notable that there are very marked distinctions the development of painful neuropathy [9]. It
between those which recover (acute painful neu- is unknown whether the two conditions purely
ropathies, radiculopathies, mononeuropathies) coexist, or a more distinct causative association
and those which progress (sensory and auto- exists [10].
nomic neuropathy). There are many clinical, Overall, the pathophysiology and natural
structural and physiological differences between history of diabetic painful and painless neuro-
these groups” [4]. Furthermore, one brilliant pathy remain an enigma to many clinicians,
case study in 1999 reported that peripheral pain- researchers, and persons with diabetes.
ful neuropathy could be a third neuron disease
[5]. Over time many dedicated clinicians have
secured a detailed description of the natural
history and signs and symptoms of painful Clinical presentations and highly variable
and painless peripheral neuropathy on which patterns of neuropathy
today’s researchers base their efforts.
Distal symmetrical sensory neuropathy is Sensory loss may go completely undetected
easily the most common pattern of neuropathy in diabetes, as there are literally no symptoms,

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00012-8 123 © 2022 Elsevier Inc. All rights reserved.
124 8. The conundrum and enigma of painful and painless neuropathy

whereas neuropathic pain often forces the per- study has demonstrated that a hyposensitivity
son with diabetes to contact the general practi- sensory phenotype together with an increased
tioner or hospital healthcare-professional for severity of neuropathy both are associated
relief of symptoms. It may present as sharp, with painful diabetic neuropathy [11].
shooting, deep pains of a burning, itching, Collectively, the evidence suggests that an
or deep musculoskeletal qualities, at times of increased severity of diabetic peripheral neu-
unbearable intensity, but fortunately it is more ropathy may be linked to a risk of developing
often represented by mere uncomfortable painful neuropathy, although painful neuropa-
symptoms of pins and needles and a tingling thy also sometimes exists in people with mild
sensation, and at times numbness that inter- or moderate diabetic peripheral neuropathy.
feres with daily activities and quality of life. However, the evidence also points toward that
The onset may be insidious symptoms gradu- severe diabetic peripheral neuropathy and pain-
ally changing and increasing in intensity—and ful neuropathy are not mutually exclusive [11].
symptoms tend to decrease after lengthy peri-
ods, often years, where after the person with
diabetes is left with the peripheral sensory loss. Risk factors for developing diabetic
The paradox of coexisting anesthesia of the peripheral and neuropathic pain
feet and at the same time excruciating pains is
intriguing as well as simultaneous numbness Risk factors for developing a medical condi-
and hypersensitivity (personal clinical observa- tion or complication are something that is fre-
tions). Adding to these observations, it has quently debated and researched in the field of
been shown that a smaller number of persons medicine, but while we are very aware of the
with painful diabetic neuropathy report “gain risk factors for developing, that is, a diabetic
of function” signs such as allodynia and hyper- foot ulcer [18] or painless diabetic peripheral
algesia [4,10]. In spite of these clear variabilities neuropathy, the evidence of risk factors for
in the individual clinical presentations, neuro- developing or progressing to painful diabetic
logical examinations tend to be nearly identical neuropathy remains sparse. Age, alcohol,
in persons with painful and painless neuropa- body mass index, ethnicity, sex, diabetes dura-
thy [11]. At a more personal level, the person tion, diabetes type, glycemic control, microan-
suffering moderate to severe neuropathic pain giopathic complications, waist circumference,
may experience a degree of physical disability, hypertension, and smoking [19,20] all emerge
depression, anxiety, insomnia, and a poorer as risk factors for developing diabetic periph-
quality of life than patients with painless neu- eral neuropathy in general, but very little data
ropathy [11 13]. It remains unclear whether are currently available regarding painful neu-
the overall severity of neuropathy is greater ropathy. The longitudinal National Health
when painful neuropathy is present. Some Service Survey in the United States from 1989
studies have shown an association linking reported an association between self-declared
severity of impairment of neuropathy with hyperglycemic events, glycosuria, and painful
presence and intensity of painful neuropathy symptoms [21], but successive epidemiological
[10,11,14,15] and others have shown the oppo- studies have failed to clearly establish the asso-
site [16,17]. Along the same lines, a detailed ciation, although most of these only consider
cross-sectional study confirmed that painful current glycemic status and do not asses these
neuropathy is associated with severity of sen- longitudinally. Some studies have indicated
sory neuropathy and thermal hyposensitivity that diabetes duration, age, or sex might be
[12], and a second detailed cross-sectional predictors for the development of painful

Diabetic Neuropathy
 Current status for detecting painful and painless neuropathy 125
diabetic peripheral neuropathy only for other in order to achieve any meaningful power
studies to find opposite results, although there and thereby to detect and evaluate preventive
appears to be an overweight of studies indicat- measures.
ing female gender as an individual risk factor.
Also, retinopathy, nephropathy, and peripheral
arterial disease have been associated with Huge unmet clinical needs and treatment
painful diabetic peripheral neuropathy in one challenges
or several studies, but again other studies have
failed to present similar findings. Of all the Diabetic peripheral neuropathy, painful and
risk factors tested in the literature, obesity painless, contributes to an array of down-
and waist circumference seem to be the most stream consequences affecting persons with
promising risk factors for developing painful diabetes physically, mentally, and with regard
diabetic peripheral neuropathy at least in to quality of life and the impact of living
people with type 2 diabetes. In particular, a with chronic or longstanding pain [22]. Loss
cross-sectional study, characterized by a multi- of sensation in the feet causes balance pro-
level approach to diagnosing painful diabetic blems increasing the risk of falls reinforced by
peripheral neuropathy and simultaneous care- diminished vision due to retinopathy, vestibu-
ful exclusion of other painful neuropathies, lar dysfunction, and motorneuropathy [23].
confirmed that body mass index was an inde- A dominant complication to peripheral neu-
pendent predictor of painful diabetic peripheral ropathy is the risk of foot ulceration and lower
neuropathy in a multiple logistic regression extremity amputations as more than 15% of all
analysis [15]. persons with diabetes in their lifetime will
Despite the promising findings regarding experience a foot ulcer and more than 80,000
obesity, the overall evidence regarding any amputations are performed yearly [24]. The
independent risk factors for developing financial costs to healthcare due to diabetic
painful diabetic peripheral neuropathy remains neuropathy are estimated to be as high as 13.7
extremely limited and cannot provide suffi- billion US dollars in the United States
cient information for neither clinicians nor alone and the majority attributed to type 2 dia-
researchers to predict the onset of neuropathic betes and foot complications [24]. Societal bur-
pain [19]. Therefore, like it is the case with dens and healthcare expenses are bound to
most comparisons between painful and pain- increase at high rates if early diagnosis and
less diabetic peripheral neuropathy, it is not treatment interventions are not optimized. As of
exactly clear why some people with diabetic now there exists no efficient, disease-modifying
peripheral neuropathy have neuropathic pain, treatment and diagnosis often takes place when
while others have not. Simultaneously, like it is clinical challenges are prominent. But above
the case with the development of diabetic all, the person with diabetes suffering complica-
foot ulcers, not all people with diabetes will tions to peripheral neuropathy is paying heavily
develop diabetic peripheral neuropathy despite due to foot ulcerations, amputations, falls, pain,
having all the affirmed risk factors, which pre- life expectancy, and quality of life.
sents researchers and clinicians with another
issue in predicting which people to intervene Current status for detecting painful and
on in both primary prevention studies and in a painless neuropathy
clinical setting. This limitation hampers the
much-needed research in primary prevention, Detection and monitoring of diabetic peri-
as thousands of participants would be needed pheral neuropathy in clinical practice are

Diabetic Neuropathy
126 8. The conundrum and enigma of painful and painless neuropathy

recommended in most clinical guidelines the condition, perhaps particularly so in clini-

including consensus statements from the cal practice. They are more widely used in clin-
Toronto Diabetic Neuropathy Expert Group [8] ical trials of diabetic neuropathies.
and the American Diabetes Association [25]. Detection and description of painful and
Despite the frequent occurrence, screening for painless peripheral neuropathy rely in clinical
diabetic peripheral neuropathy is at times practice to a great deal on clinical observation
neglected leading to a considerable diagnostic and experience by the healthcare professional,
delay and insufficient preventative measures. guided by simple bedside tests such as
The reasons are many, but most important are monofilaments or biothesiometry occasionally
the lack of quick and reliable screening meth- adapting neuropathy impairment question-
ods for painful and painless neuropathy [26]. naires. There exists an unmet need for high-
Current clinical practice is, therefore, often lim- sensitivity, high-specificity bedside tests and
ited to screening for severe large fiber damage questionnaires for clinical use for the earliest
and loss of protective sensation with either possible detection of peripheral neuropathy.
a 10-g monofilament or by testing vibration
sensation with either a tuning fork or using
biothesiometry as recommend by NICE [27,28]. Early detection and prevention
Furthermore, treatment is often limited to
enhanced glycemic control and if pain is pres- Early detection of diabetic peripheral neu-
ent, use of analgesic compounds with an effi- ropathy, painful and painless, remains a major
cacy of around 30% is very near to the efficacy challenge in clinical practice. Peripheral neu-
of placebo treatment [26]. ropathy is often first detected when permanent
Over the last decades, the interest in small impairments and/or pain is already present
fiber neuropathy has increased mostly due to a [27]. Current peripheral neuropathy screening
growing agreement that it is detectable years relies on subjective tests of large fiber nerve
in advance of large fiber damage [29,30]. function (light touch, vibration), sensory
Unfortunately, early detection of small fiber testing mainly [27,28,35]. Small fiber testing
neuropathy has proven to be a diagnostic chal- (pain, temperature) is not routinely performed,
lenge and is generally not carried out in clinical although damage to small fibers appears to
practice, as clinicians and researchers currently precede large fiber damage [36,37]. In direct
have no validated and robust clinical tests contrast, diabetic retinopathy and diabetic
available, although several promising, and nephropathy have established screening detect-
equipment-heavy, technologies are currently ing preclinical conditions and allowing preven-
being developed [26]. A number of composite tative measures [38,39]. Our clinical focus and
neurological scores and questionnaires have practice are directed toward large fibers,
also been developed and validated for clinical although it has been clearly described that dia-
assessment of diabetic painful and painless betic peripheral neuropathy, painful and pain-
neuropathy. Widely used are the Michigan less, early on affects small unmyelinated fibers,
Neuropathy Screening Instrument [31], the which are responsible for pain perception,
Neuropathy Disability Score [32,33], and the sweating, and blood flow, all crucial to diabetic
modified Toronto Clinical Neuropathy Score foot ulceration [6,27] and impairment of small
[34]. Common for all is that they mostly aim at fibers are believed to proceed large fiber
identifying or excluding the presence of dia- [7,27,40]. Adding to this, small fiber degenera-
betic peripheral neuropathy and are less help- tion is found in prediabetes before the onset of
ful in detecting improvement or worsening of type 2 diabetes [40,41].

Diabetic Neuropathy
 Deep sensory profiling and mechanism-based treatment in painful and painless neuropathy 127
Currently recommendations and conven- beneficial effect of tight glycemic control [44],
tional screening programs hamper the imple- but several larger studies including the
mentation of timely preventative measures UKPDS, ACCORD, ADDITION, Steno 2, and
in the individual person with diabetes and VADT studies [45 49] all failed to show
allow progression of complications. There is a significant results regarding diabetic peripheral
clear need for robust and validated test meth- neuropathy, although other microvascular
ods and clinical endpoints to detect painful complications improved. Previous clinical
and painless peripheral neuropathy separately trials in both people with type 1 and type 2
before it becomes clinically significant, and diabetes have demonstrated beneficial effects
more so to target and focus on small fiber neu- of angiotensin-converting enzyme inhibitors on
ropathy as the location of earliest detectable mainly autonomic neuropathy but also on dis-
neuropathy impairment [27]. Theoretically, this tal peripheral neuropathy in people with dia-
could be a key to better compare and under- betes, although larger studies are still needed
stand the associations, causative or reflective, to confirm the findings [50 52]. Also statins
between painful and painless neuropathies. and fibrates have been shown to halt the pro-
But, most of all it could prove to be invaluable gression of diabetic peripheral neuropathy
for early detection and prevention of periph- [53,54], but might rarely induce neuropathy on
eral neuropathy. their own [55]. The implications for painful
neuropathy on the interventions mentioned
earlier are unknown. It is safe to say that treat-
Similarities in treatment of painful and ment for painful neuropathy is solely directed
painless neuropathy toward symptom relief and improvement of
quality of life as the main outcome measures.
At present there are no disease-modifying Unfortunately, almost all pharmacological
treatments for diabetic peripheral neuropathy agents have side effects with small therapeutic
approved by the US food and Drug windows between beneficial effects and at
Administration or the European Medicines times incapacitating side effects. The overall
Agency, leaving optimization of glycemic con- average efficacy of treatment is around 30%
trol to prevent progression as the clinicians’ following guidelines and algorithms for use of
most important intervention. Several clinical drug classes [56]. There is clearly a need for
trials have shown beneficial effect of tight more efficient and safer treatment for painful
glycemic control in preventing peripheral neu- neuropathy, but our limited understanding of
ropathy in people with T1DM [42]. The the pathophysiology and interactions between
DCCT/EDIC study even demonstrated benefi- painful and painless neuropathy prohibits the
cial effect of intensified glycemic control years development of mechanism-based treatment
after terminating the intervention [43]. For peo- options.
ple with type 2 diabetes (T2DM), the evidence
of beneficial effects of tight glycemic control
in preventing painful and painless diabetic Deep sensory profiling and mechanism-
peripheral neuropathy is less convincing, but based treatment in painful and painless
the varying results might arise from the fact neuropathy
that the many people with newly diagnosed
type 2 diabetes might have gone unnoticed While no disease-modifying treatment for
for many years prior to diagnosis. A small painful diabetic neuropathy exists, several
Japanese study demonstrated a relatively small interesting agents are currently undergoing

Diabetic Neuropathy
128 8. The conundrum and enigma of painful and painless neuropathy

early clinical trial. These agents all target pro- function alongside paradoxical heat sensation,
posed mechanisms responsible for the onset (2) a group with preserved sensory function
of neuropathic pain, which classifies them combined with heat and cold hyperalgesia
alongside other pathogenetic drugs like α-lipoic and mild dynamic mechanical allodynia, and
acid, C-peptide, benfotiamine, aldose-reductase (3) a group with loss of small fiber function,
inhibitors, and the nerve growth factor- pinprick hyperalgesia, and mechanical allody-
modulator Tanezumab [57]. While none of nia. Although the authors did not test the effi-
these drugs have been approved by the US ciency of different drugs in the three groups,
food and Drug Administration or the European they were able to replicate these sensory pro-
Medicines Agency, some have achieved local files in an experimental setting, generating one
authorization in some countries, as they have of the first models for mechanism-based treat-
much milder side effects than their approved ment of neuropathic pain. Although this exact
counterparts and have shown some effect in model is purely relevant for neuropathic pain
smaller studies although several different meta- and not for peripheral diabetic neuropathy in
analyses have proven them no better than pla- general, the basic ideas are very much relevant
cebo. However, if the disappointing results for future research of both painful and painless
from the meta-analysis are indeed due to ineffi- distal polyneuropathy and might also help dis-
cient drugs or targets or if positive results tinguishing the two from one another.
are hidden, inside small subgroups remain When considering both treatment and pre-
unknown, as characterization of the type of vention of diabetic peripheral neuropathy and
pain (or painless polyneuropathy) critically is neuropathic pain, diabetes type should also be
not a part of usual clinical trials. This issue taken into consideration as part of the deep
would also become apparent even if some phenotyping. It is apparent to most that the
of the new agents might prove efficient in some pathophysiological mechanisms for the devel-
people, as clinicians are still presented with opment of distal diabetic polyneuropathy and
the issue that they have no guidelines or neuropathic pain are hardly the same in people
indications weather a drug has any effect on with type 1 and type 2 diabetes, although cur-
their particular patient. Therefore future treat- rent clinical guidelines do not distinguish the
ment of neuropathic pain, and probably two conditions from each other. From the fact
also prevention of onset of diabetic peripheral that mild polyneuropathy is seen even in peo-
neuropathy, should focus on personalized, ple with prediabetes, while it is more generally
mechanism-based medication/prevention aris- seen after years of disease in people with type
ing from deep-sensory profiling and in-depth 1 diabetes, it also becomes clear that one
characterization of each individual with neuro- single mechanism might not fit to describe the
pathic pain or threatening neuropathy onset natural history of the development of diabetic
[58,59]. Using this approach, researchers peripheral neuropathy and neuropathic pain.
from the German Neuropathic Pain Research Chronic low-grade inflammation, oxidative
Network, EUROPAIN, and NEUROPAIN con- stress, impaired perfusion, and accumulation of
sortia applied their standardized protocol for advanced-glycemic end products are all thought
Quantitative Sensory Testing in a large cohort to be contributing factors to the development
of 902 participants [60]. They then subgrouped of diabetic peripheral neuropathy [61], but even
them into three distinct groups using cluster a combination of these factors does not seem to
analysis and identified the sensory profile of give a clear picture of the natural history and
each subgroup. The subgroups consisted of (1) even less so the differences between those who
a group with loss of small and large fiber develop painful and painless neuropathy.

Diabetic Neuropathy
 Novel clinical endpoints and research methodologies 129
Another question that must be addressed marked reductions in levels of N-acetyl-aspar-
when talking differences between the sensory tate in the same region with painless distal dia-
profiles of those with painful and painless dia- betic polyneuropathy [66,67]. Therefore, to
betes neuropathy is the cause of the pain. A perform an in-depth sensory profiling precisely
prevailing theory is that the initial pain arises and adequately, and thus distinguish painful
from damage to the peripheral nerve endings from painless diabetic peripheral neuropathy,
of the small nerve fibers (dying-back neuropa- future researchers might also need to consider
thy) [41], but that persistent and chronic pain the role of the central nervous system, although
arises from central sensitization and altered current research methodologies only allow
brain plasticity rather than morphological glimpses into the complexity of the vast net-
changes in the peripheral nerves [62]. With this work of neurons that is the central nervous
in mind, people with distal peripheral diabetic system.
neuropathy with and without pain might not In summary, deep-sensory profiling and
even differ regarding their peripheral morphol- mechanism-based prevention and treatment of
ogy, but instead differ in regard to their central diabetic peripheral neuropathy and neuro-
interpretation of the missing or enhanced sig- pathic pain are a huge field, which we have
nals arising from the distal nerves. This theory only just begun to explorer. Future research
is supported by the fact that primary nerve into the area is very much needed, and maybe
afferents are sensitized in painful diabetic neu- even essential to achieve an understanding
ropathy in rats, inducing dorsal horn hyperac- of diabetic peripheral neuropathy with and
tivity and neuroplastic changes in central without pain, how they differ, and how they
sensory neurons [63]. Furthermore, the rather are alike. Furthermore, future randomized con-
common occurrence of allodynia in people trolled trials and other clinical studies should
with painful diabetic neuropathy likewise sup- build on the growing knowledge achieved by
ports the idea that central pain processing is rather basal research into deep-sensory profil-
indeed altered [64]. The possible implications ing and power accordingly to test the effect not
of this are huge, but nevertheless central sensi- only in a large population with neuropathic
bilization and brain plasticity are not consid- pain but also in subgroups with different pro-
ered or even less so implemented into any posed mechanisms of neuropathic pain.
clinical guidelines or even designs for random-
ized controlled trials regarding diabetic periph-
eral neuropathy or neuropathic pain. This is Novel clinical endpoints and research
mainly due to our lacking understanding of methodologies
not only the processes of the central nervous
system but also its interaction with the periph- Despite a considerable number of new and
eral nerves. Functional changes in the pain- promising methods for early detection of par-
processing areas of the brain have, however, ticularly small fiber neuropathy, but also large
been linked to diabetic peripheral neuropathy fiber neuropathy, the lack of well-established
and neuropathic pain at least in rodents, while gold standard methods to validate novel meth-
the translation to human studies have been a ods against have led to the fact that both
bit less certain [65]. Some studies in humans researchers and clinicians are yet to agree on
have linked diabetic peripheral neuropathy and standardize the use of many of the newer
to increased activity of several brain areas methods [26]. In order to perform meaningful
including thalamus after peripheral thermal clinical trials, the research community must
stimulation, while others have indicated first have access to agreeable, standardized,

Diabetic Neuropathy
130 8. The conundrum and enigma of painful and painless neuropathy

highly sensitive methods for quantification of clinical endpoints for description of the natural
nerve fiber function, so that a potentially bene- history, pathophysiology, and mechanism-
ficial effect of disease-modifying treatment can based approaches—allowing not only success-
be detected. Novel methods and approaches ful clinical trials but also optimization of clini-
are discussed elsewhere and comprise skin cal practice in terms of early detection,
biopsies of small fiber nervous tissue combin- prevention, diagnosis, and individualized
ing form and function [68], cornea confocal treatment for painful and painless diabetic
microscopy for early detection in changes of peripheral neuropathy.
small fiber morphology and pain perception
thresholds, and ion-channel functionalities in
small fiber diabetic neuropathy [69]. Biofluid References
markers of neurodegeneration and inflamma-
tion [70], (deep) sensory phenotyping [71,72], [1] Vergely P. Des troubles de la sensibilité aus members
inférieurs ched les diabétiques. De la dissociation syr-
genotyping, and epi-genetic description [73] ingomyélique de la sensibilité chex les diabétiques.
are all cutting-edge novel approaches for Gazette Hebdomadaire de Médecine et de Chirurgie
describing diabetic neuropathy—and fit nicely 1893;32:376 81.
in with our wish for a better understanding [2] Pavy FW. Address on diabetes, Washington International
of the pathophysiology. All is needed, in our Congress. Med News (Philadelphia) 1887;24:357 61.
[3] Archer AG, Watkins PJ, Thomas PK, Sharma AK,
opinion, not only to compare painful and pain- Payan J. The natural history of acute painful neuropa-
less neuropathies and describe neuropathy thy in diabetes mellitus. J Neurol Neurosurg
mechanisms, but also to develop efficient Psychiatry 1983;46(6):491 9.
preventative agents, disease-modifying agents, [4] Watkins PJ. Clinical observations and experiments in
diabetic neuropathy. Diabetologia 1992;35:2 11.
and compounds to treat pain in diabetic
[5] Rajbhandari SM, Jarratt JA, Griffiths PD, Ward JD.
neuropathy. Diabetic neuropathic pain in a leg amputated 44 years
previously. Pain 1999;83(3):627 9.
[6] Feldman EL, Callaghan BC, Pop-Busui R, Zochodne
DW, Wright DE, Bennett DL, et al. Diabetic neuropa-
Perspectives thy. Nat Rev Dis Primers 2019;5:41.
[7] Callaghan BC, Cheng HT, Stables CL, Smith AL,
Comparing and describing associations Feldman EL. Diabetic neuropathy: clinical manifestations
between painful and painless diabetic periph- and current treatments. Lancet Neurol 2012;11:521 34.
[8] Tesfaye S, Boulton AJM, Dyck PJ, Freeman R,
eral neuropathy at this time is a relatively easy Horowitz M, Kempler P, et al. Diabetic neuropathies:
undertaking, as very little is known about asso- update on definitions, diagnostic criteria, estimation
ciations in terms of pathophysiology, struc- of severity, and treatments. Diabetes Care 2010;33
tures, functions, and mechanisms responsible (10):2285 93.
for the development of these intercorrelated [9] Terkelsen AJ, Karlsson P, Lauria G, Freeman R,
Finnerup NB, Jensen TS. The diagnostic challenge of
complex conditions. small fibre neuropathy: clinical presentations, evalua-
At this time, it is not possible to detect, tions, and causes. Lancet Neurol 2017;16:934 44.
describe, and identify, which person with dia- [10] Shillo P, Sloan G, Greig M, Hunt L, Selvarajah D, Elliott
betes will progress to painful neuropathy, J, et al. Painful and painless diabetic neuropathies: what
independent of the presence of painless neu- is the difference? Current Diabetes Reports 2019;19:32.
[11] Themistocleous AC, Ramirez JD, Shillo PR, Lees JG,
ropathy, which coexists in practically all cases. Selvarajah D, Orengo C, et al. The pain in neuropathy
The way forward seems to be utilization of the study (PiNS): a crosssectional observational study deter-
list of novel high-tech methodologies men- mining the somatosensory phenotype of painful and
tioned earlier in order to construct robust painless diabetic neuropathy. Pain. 2016;157:1132 45.

Diabetic Neuropathy
 References 131
[12] Raputova J, Srotova I, Vlckova E, Sommer C, Üçeyler [24] Gordois A, Scuffham P, Shearer A, Oglesby A, Tobian
N, Birklein F, et al. Sensory phenotype and risk factors JA. The health care costs of diabetic peripheral neu-
for painful diabetic neuropathy: a cross-sectional ropathy in the US. Diabetes Care 2003;26(6):1790 5.
observational study. Pain 2017;158:2340 53. [25] Pop-Busui R, Boulton AJM, Feldman EL, Bril V,
[13] Gore M, Brandenburg NA, Dukes E, Hoffman DL, Tai Freeman R, Malik RA, et al. Diabetic neuropathy: a
KS, Stacey B. Pain severity in diabetic peripheral neu- position statement by the American diabetes associa-
ropathy is associated with patient functioning, symp- tion. Diabetes Care 2017;40:136 54.
tom levels of anxiety and depression, and sleep. J Pain [26] Røikjer J, Mørch CD, Ejskjaer N. Diabetic peripheral
Symptom Manag 2005;30:374 85. neuropathy: diagnosis and treatment. Curr Drug Saf
[14] Veves A, Young MJ, Manes C, Boulton AJ. Differences 2021;16(1):2 16.
in peripheral and autonomic nerve function measure- [27] Burgess J, Frank B, Marshall A, Khalil RS, Ponirakis G,
ments in painful and painless neuropathy. A clinical Petropoulos IN, et al. Early detection of diabetic
study. Diabetes Care 1994;17:1200 2. peripheral neuropathy: a focus on small nerve fibres.
[15] Spallone V, Morganti R, D’Amato C, Cacciotti L, Diagnostics 2021;11:165.
Fedele T, Maiello MR, et al. Clinical correlates of pain- [28] NICE, National Institute for Health and Care
ful diabetic neuropathy and relationship of neuro- Eexelence. Diabetic foot problems: prevention and
pathic pain with sensorimotor and autonomic nerve management, NICE Guideline (NG19).
function. Eur J Pain 2011;15:153 60. [29] Divisova S, Vlckova E, Hnojcikova M, Skorna M, Nemec
[16] Andersen ST, Witte DR, Dalsgaard EM, Andersen H, M, Dubovy P, et al. Prediabetes/early diabetes-
Nawroth P, Fleming T, et al. Risk factors for incident associated neuropathy predominantly involves sensory
diabetic polyneuropathy in a cohort with screen- small fibres. J Peripher Nerv Syst 2012;17:341 50.
detected type 2 diabetes followed for 13 years: [30] Umapathi T, Tan WL, Loke SC, Soon PC, Tavintharan
ADDITION-Denmark. Diabetes Care 2018;41:1068 75. S, Chan YH. Intraepidermal nerve fiber density as a
[17] Truini A, Spallone V, Morganti R, Tamburin S, marker of early diabetic neuropathy. Muscle Nerve
Zenette G, Schenone A, et al. A cross-sectional study 2007;35:591 8.
investigating frequency and features of definitely [31] Feldman EL, Stevens MJ, Thomas PK, Brown MB,
diagnosed diabetic painful polyneuropathy. Pain Canal N, Greene DA. A practical two-step quantitative
2018;159:2658 66. clinical and electrophysiological assessment for the
[18] Armstrong DG, Boulton AJM, Bus SA. Diabetic foot diagnosis and staging of diabetic neuropathy.
ulcers and their recurrence. N Engl J Med Diabetes Care 1994;17:1281 9.
2017;376:2367 75. [32] Young MJ, Boulton AJM, Macleod AF, Williams DRR,
[19] Spallone V, Greco C. Painful and painless diabetic Sonksen PH. A multicentre study of the prevalence
neuropathy: one disease or two? Curr Diab Rep of diabetic peripheral neuropathy in the United
2013;13:533 49. Kingdom hospital clinic population. Diabetologia
[20] Hébert HL, Veluchamy A, Torrance N, Smith BH. Risk 1993;36(2):150 4.
factors for neuropathic pain in diabetes mellitus. Pain [33] Dyck PJ, Sherman WR, Hallcher LM, John Service F,
2017;158:560 8. O’Brien PC, Grina LA, et al. Human diabetic endo-
[21] Sun JK, Keenan HA, Cavallerano JD, Asztalos BF, neurial sorbitol, fructose, and myo-inositol related to
Schaefer EJ, Sell DR, et al. Protection from retinopathy sural nerve morphometry. Ann Neurol 1980;8:590 6.
and other complications in patients with type 1 diabe- [34] Bril V, Tomioka S, Buchanan RA, Perkins BA.
tes of extreme duration: the Joslin 50-year medalist Reliability and validity of the modified Toronto
study. Diabetes Care 2011;34:968 74. Clinical Neuropathy Score in diabetic sensorimotor
[22] Van Acker K, Bouhassira D, De Bacquer D, Weiss S, polyneuropathy. Diabet Med 2009;26:240 6.
Matthys K, Raemen H, et al. Prevalence and impact on [35] Malik R. Neuropad: early diagnostic test for diabetic
quality of life of peripheral neuropathy with or with- peripheral neuropathy. Prescriber 2008;19:42 5.
out neuropathic pain in type 1 and type 2 diabetic [36] Breiner A, Lovblom LE, Perkins BA, Bril V. Does the
patients attending hospital outpatients clinics. prevailing hypothesis that small-fiber dysfunction pre-
Diabetes Metab 2009;35(3):206 13. cedes large-fiber dysfunction apply to type 1 diabetic
[23] Agrawal Y, Carey JP, Della Santina CC, Schubert MC, patients? Diabetes Care 2014;37:1418.
Minor LB. Diabetes, vestibular dysfunction, and falls: [37] Bansal V, Kalita J, Misra UK. Diabetic neuropathy.
analyses from the National Health and Nutrition Postgrad Med J 2006;82:95 100.
Examination Survey. Otol Neurotol 2010;31(9):1445 50 [38] Liew G, Michaelides M, Bunce C. A comparison of the
[PubMed: 20856157]. causes of blindness certifications in England and Wales

Diabetic Neuropathy
132 8. The conundrum and enigma of painful and painless neuropathy

in working age adults (16 64 years), 1999 2000 with converting-enzyme (ACE) inhibitor trandolapril on
2009 2010. BMJ Open 2014;4:e004015. human diabetic neuropathy: randomised double-blind
[39] Marshall SM. Diabetic nephropathy in type 1 diabetes: controlled trial. Lancet 1998;352:1978 81.
has the outlook improved since the 1980s? Diabetologia [51] Didangelos T, Tziomalos K, Margaritidis C,
2012;55:2301 6. Kontoninas Z, Stergiou I, Tsotoulidis S, et al. Efficacy
[40] Ziegler D, Papanas N, Zhivov A, Allgeier S, Winter K, of administration of an angiotensin converting enzyme
Ziegler I, et al. Early detection of nerve fiber loss by inhibitor for two years on autonomic and peripheral
corneal confocal microscopy and skin biopsy in recently neuropathy in patients with diabetes mellitus. J
diagnosed type 2 diabetes. Diabetes 2014;63:2454 63. Diabetes Res 2017;2017.
[41] Dobretsov M, Romanovsky D, Stimers JR. Early dia- [52] Ruggenenti P, Lauria G, Iliev IP, Fassi A, Ilieva AP,
betic neuropathy: triggers and mechanisms. World J Rota S, et al. Effects of manidipine and delapril in
Gastroenterol 2007;13:175 91. hypertensive patients with type 2 diabetes mellitus:
[42] Group TDC and CTR. The effect of intensive treatment the delapril and manidipine for nephroprotection in
of diabetes on the development and progression of diabetes (DEMAND) randomized clinical trial.
long-term complications in insulin-dependent diabetes Hypertension 2011;58:776 83.
mellitus. N Engl J Med 1993;329:977 86. [53] Davis TME, Yeap BB, Davis WA, Bruce DG. Lipid-
[43] Martin CL, Albers JW, Pop-Busui R. Neuropathy lowering therapy and peripheral sensory neuropathy
and related findings in the diabetes control and in type 2 diabetes: the Fremantle Diabetes Study.
complications trial/epidemiology of diabetes inter- Diabetologia 2008;51:562 6.
ventions and complications study. Diabetes Care [54] Nielsen SF, Nordestgaard BG. Statin use before diabe-
2014;37:31 8. tes diagnosis and risk of microvascular disease: a
[44] Ohkubo Y, Kishikawa H, Araki E, Miyata T, Isami nationwide nested matched study. Lancet Diabetes
S, Motoyoshi S, et al. Intensive insulin therapy pre- Endocrinol 2014;2:894 900.
vents the progression of diabetic microvascular [55] Gaist D, Jeppesen U, Andersen M, Garcı́a Rodrı́guez LA,
complications in Japanese patients with non- Hallas J, Sindrup SH. Statins and risk of polyneuropathy:
insulin-dependent diabetes mellitus: a randomized a case-control study. Neurology 2002;58:1333 7.
prospective 6-year study. Diabetes Res Clin Pract [56] Finnerup NB, Kuner R, Jensen TS. Neuropathic
1995;28:103 17. pain: from mechanisms to treatment. Physiol Rev
[45] Turner R. Intensive blood-glucose control with sul- 2021;101(1):259 301.
phonylureas or insulin compared with conventional [57] Javed S, Petropoulos IN, Alam U, Malik RA. Treatment
treatment and risk of complications in patients with of painful diabetic neuropathy. Ther Adv Chronic Dis
type 2 diabetes (UKPDS 33). Lancet 1998;352:837 53. 2015;6:15 28.
[46] Ismail-Beigi F, Craven T, Banerji MA, Basile J, Calles J, [58] Colloca L, Ludman T, Bouhassira D, Baron R,
Cohen RM, et al. Effect of intensive treatment of Dickenson AH, Yarnitsky D, et al. Neuropathic pain.
hyperglycaemia on microvascular outcomes in type 2 Nat Rev Dis Prim 2017;3.
diabetes: an analysis of the ACCORD randomised [59] Finnerup NB, Attal N, Haroutounian S, McNicol E,
trial. Lancet 2010;376:419 30. Baron R, Dworkin RH, et al. Pharmacotherapy for
[47] Charles M, Ejskjaer N, Witte DR, Borch-Johnsen K, neuropathic pain in adults: a systematic review and
Lauritzen T, Sandbaek A. Prevalence of neuropathy meta-analysis. Lancet Neurol 2015;14:162 73.
and peripheral arterial disease and the impact of [60] Baron R, Maier C, Attal N, Binder A, Bouhassira D,
treatment in people with screen-detected type 2 dia- Cruccu G, et al. Peripheral neuropathic pain: a
betes: the ADDITION-Denmark study. Diabetes mechanism-related organizing principle based on sen-
Care 2011;34:2244 9. sory profiles. Pain 2017;158:261 72.
[48] Gæde P, Vedel P, Larsen N, Jensen GVH, Parving HH, [61] Yagihashi S, Mizukami H, Sugimoto K. Mechanism of
Pedersen O. Multifactorial intervention and cardiovascular diabetic neuropathy: where are we now and where to
disease in patients with type 2 diabetes. N Engl J Med go? J Diabetes Investig 2011;2:18 32.
2003;348:383 93. [62] Schreiber AK. Diabetic neuropathic pain: physiopa-
[49] Duckworth W, Abraira C, Moritz T, Reda D, thology and treatment. World J Diabetes 2015;6:432.
Emanuele N, Reaven PD, et al. Glucose control and [63] Chen SR, Pan HL. Hypersensitivity of spinothalamic
vascular complications in veterans with type 2 diabe- tract neurons associated with diabetic neuropathic
tes. N Engl J Med 2009;360:129 39. pain in rats. J Neurophysiol 2002;87:2726 33.
[50] Malik RA, Williamson S, Abbott C, Carrington AL, [64] Aslam A, Singh J, Rajbhandari S. Pathogenesis of pain-
Iqbal J, Schady W, et al. Effect of angiotensin- ful diabetic neuropathy. Pain Res Treat 2014;2014.

Diabetic Neuropathy
 References 133
[65] Silva M, Amorim D, Almeida A, Tavares I, Pinto- [70] Okdahl T, Brock C, Fløyel T, Wegeberg AL, Jakobsen
Ribeiro F, Morgado C. Pronociceptive changes in the PE, Ejskjaer N, et al. Increased levels of inflammatory
activity of rostroventromedial medulla (RVM) pain factors are associated with severity of polyneuropathy
modulatory cells in the streptozotocin-diabetic rat. in type 1 diabetes. J Clin Endocrinol (Oxf) 2020;
Brain Res Bull 2013;96:39 44. 93(4):419 28.
[66] Sorensen L, Siddall PJ, Trenell MI, Yue DK. [71] Christensen DH, Knudsen ST, Gylfadottir SS,
Differences in metabolites in pain- processing brain Christensen LB, Nielsen JS, Beck-Nielsen H, et al.
regions in patients with diabetes and painful neuropa- Metabolic factors, lifestyle habits, and possible poly-
thy. Diabetes Care 2008;31:980 1. neuropathy in early type 2 diabetes: a nationwide
[67] Tseng MT, Chiang MC, Chao CC, Tseng WYI, Hsieh study of 5,249 patients in the Danish Centre for strate-
ST. fMRI evidence of degeneration-induced neuro- gic research in type 2 diabetes (DD2) cohort. Diabetes
pathic pain in diabetes: enhanced limbic and striatal Care 2020;43(6):1266 75.
activations. Hum Brain Mapp 2013;34:2733 46. [72] Vollert J, Maier C, Attal N, Bennett DLH, Bouhassira
[68] Karlsson P, Proviterac V, Caporasoc G, Stancanellic A, D, Enax-Krumova EK, et al. Stratifying patients with
Saltalamacchia AM, Borreca I, et al. Increased peptider- peripheral neuropathic pain based on sensory profiles:
gic fibers as a potential cutaneous marker of pain in algorithm and sample size recommendations. Pain
diabetic small fiber neuropathy. Pain 2021;162:778 86. 2017;158(8):1446 55.
[69] Poulsen AH, Tigerholm J, Andersen OK, Mørch CD. [73] D’Agnelli S, Gerra MC, Bignami E, Arendt-Nielsen L.
Increased preferential activation of small cutaneous Exosomes as a new pain biomarker opportunity. Mol
nerve fibers by optimization of electrode design para- Pain 2020;16 1744806920957800.
meters. J Neural Eng 2020;18(1):16 20.

Diabetic Neuropathy
 C H A P T E R

9
Motor dysfunction in diabetes
Anders Stouge and Henning Andersen
Department of Neurology, Aarhus University Hospital, Aarhus, Denmark

Introduction improvement can be seen following treatment

with immunoglobulins [6,7]. Interestingly, in
In diabetic neuropathy (DN), symptoms and contrast to the predominant sensory symptom-
signs depend on the severity of impairment of atology in DSPN, in nerve conduction studies
sensory, autonomic, and motor nerves, clinically (NCS), motor and sensory nerves are impaired
presenting in a symmetrical or asymmetrical pat- to a similar degree [8,9]. Minimal F-wave
tern [1]. In up to 50% of patients with long-term latency, being a motor conduction parameter,
diabetes the bilateral and symmetrical presenta- has in fact been found to be the most sensitive
tion, termed diabetic sensorimotor polyneuropa- conduction parameter to detect peripheral poly-
thy (DSPN), can be found, and since it is by far neuropathies, including DSPN [8,9]. These find-
the most common presentation of DN, the litera- ings indicate that the predominance of sensory
ture often uses DSPN synonymously with DN symptoms is not primarily the consequence of
[2,3]. During the course of DSPN, the clinical less severe degeneration in motor nerves.
presentation is dominated by symptoms and Resprouting of motor nerves and collateral rein-
signs of sensory nerve damage, and conse- nervation of motor units are believed to be the
quently a prevalent belief has been that the compensatory mechanisms alleviating the conse-
motor system is involved only in advanced cases quences of muscle denervation, forming the
of DSPN [3,4]. Leaving aside DSPN, motor basis for why motor symptoms and deficits are
involvement is frequently found in other types primarily observed later in the course of DSPN
of DN such as diabetic lumbosacral radiculo- [10,11]. This proposition is supported by recent
plexus neuropathy and several mononeuropa- studies by Kristensen et al. reporting decreased
thies [5]. With that being said, due to the much motor unit numbers and increased motor unit
less common presentation of DN types unrelated size in patients with DSPN [9,12]. Accelerated
to DSPN, other peripheral nerve disorders muscle weakness in diabetes is attributed to the
should be considered in patients with diabetes degree of DSPN, following denervation and pro-
with a clinical presentation predominated by gressive muscular atrophy [13,14]. Atrophy and
motor impairment, for example, immune- weakness of the foot muscles can initially be
mediated neuropathies, where significant observed; however, this often goes unnoticed by

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00009-8 135 © 2022 Elsevier Inc. All rights reserved.
136 9. Motor dysfunction in diabetes

patients contributing to the fact that motor dys- In this chapter, we will deal with motor dys-
function can remain unrecognized until later function in diabetes, including peripheral
stages of DSPN [15]. Recent studies applying motor neuropathy in DSPN, muscle weakness
imaging modalities such as magnetic resonance and functional deficits, and compositional and
imaging (MRI) suggest that fat infiltration and structural changes. Finally, we will consider
atrophy may be detectable in skeletal muscles of the management of motor deficits in DSPN
the lower extremities even in earlier stages of and diabetes, including the effects of exercise
DSPN [16,17]. Interestingly, MRI studies suggest interventions.
a less obvious distal proximal gradient of mus-
cular abnormalities regardless of the distribution
of DSPN, with a larger decrease in proximal
muscle strength and fat infiltration than previ-
Motor nerve dysfunction
ously believed [16,17]. It is well established that
DSPN results in distal atrophy and MRI signal
Electroneuronography
abnormalities within skeletal muscles. Still, the Pathological changes in DN include axonal
understanding of muscle dysfunction in patients degeneration and demyelination of the periph-
with diabetes is incomplete, and emerging evi- eral nerves, which can be observed in motor as
dence also suggests the presence of a general well as in sensory nerves. Histological studies
skeletal muscle dysfunction in patients with dia- with findings of decreased nerve fiber density
betes, termed diabetic myopathy, related to both are consistent with the clinical and electrophys-
neuropathic and metabolic deficits [18]. Fig. 9.1 iological impairments, in line with the belief
illustrates the proposed combined neuromuscu- that nerve fiber loss is the primary reason for
lar and functional muscular deficits observed in the neurological deficits seen in diabetes
relation to diabetes and DSPN. [19,20]. Slowing of nerve conduction and

FIGURE 9.1 Overview of the neuromuscular (orange), metabolic (blue), and functional (green) deficits of motor function
in patients with diabetes.

Diabetic Neuropathy
 Motor nerve dysfunction 137
decreased response amplitudes of motor and latency has not been part of the standard elec-
sensory nerves are frequent neurophysiological trodiagnostic battery of DSPN.
findings [4,21]. Consensus of electrodiagnostic Few studies have examined motor nerve
criteria for the presence of DSPN follows the conduction within the central nervous system
guidelines of Dyck requiring at least one in patients with diabetes. Increased latencies
abnormal electrodiagnostic finding in two sep- and abnormal cortical excitability within cen-
arate nerves, one of which is required to be the tral motor pathways have been described in
sural nerve [4]. Standard examination follows clinical and preclinical animal studies [24 28].
the guidelines of the American Association of Overall, the clinical significance of these find-
Neuromuscular and Electrodiagnostic ings is elusive and unlikely to contribute signif-
Medicine (AANEM) suggesting to start with icantly to signs and symptoms of clinical motor
the examination of the peroneal and sural dysfunction.
nerves on one leg, and in the case where
neurophysiological measures are normal, no
further action is recommended [22]. However,
recent studies implicate that DSPN may go
Electromyography
unnoticed using this strategy, since measures Electromyography (EMG) enables the detec-
of the tibial nerve have proven to be more sen- tion of neurogenic and myogenic factors for
sitive to motor involvement than peroneal evaluation of neuromuscular deficits such as
measures [8]. In addition, Tankisi et al. muscle weakness and atrophy. In patients with
observed that peroneal nerve abnormalities diabetes, early studies of EMG report a number
were more common in participants without of abnormalities, including fibrillation poten-
peripheral neuropathy, suggesting abnormali- tials, decreased recruitment pattern, larger
ties in the peroneal nerve to be less specific. motor unit potentials, and a higher number of
Despite frequent abnormalities in motor polyphasic potentials [29,30]. These findings
nerves, the sural nerve remains the most spe- emphasize the presence of a neurogenic motor
cific NCS measure and should be included in dysfunction in patients with DN induced by
the criteria to rule out other peripheral motor axonal loss and incomplete compensatory rein-
nerve disorders and identify patients without nervation observed in NCS. For sensory
DSPN [8]. Both the peroneal and tibial nerve nerves, skin biopsies can provide a direct esti-
may show abnormalities in subjects without mate of the number of distal nerve segments.
peripheral polyneuropathies, which may be Clearly no such direct estimates of the number
due to a larger variation in the conduction of of motor axons are possible. Yet, neurophysio-
motor nerves in healthy subjects. Tankisi et al. logical techniques enable an indirect estimation
observed that prolonged minimal F-wave of the degree of loss of motor axons. Early
latency, a measure of pure motor nerve func- studies implicated that computer-assisted
tion of the whole length of the motor axon, motor unit number counting and subtraction
was the most common motor nerve abnormal- methods may help to quantify denervation and
ity in participants with peripheral neuropa- reinnervation of motor units [31]. For many
thies, and more common in the tibial (72%), years, these techniques have been associated
than in the peroneal nerve (58%). This is in line with a number of limitations, and as a conse-
with previous studies where F-wave latency quence motor unit number estimation (MUNE)
has been found to be the most sensitive and methods have only sparsely been used [32].
reproducible NCS measure of motor neuropa- Recently, emerging methods, like MScanFit
thy [11,23]. Yet, until recently minimal F-wave MUNE, have proved to provide reliable

Diabetic Neuropathy
138 9. Motor dysfunction in diabetes

estimates of motor unit numbers covering the

range of 5 160 units with an overall error rate
of only 6.9% [33]. In recent clinical studies of
DSPN, MScanFit proved more sensitive in
detecting motor impairment than conventional
NCS [9,12]. Interestingly, MScanFit evaluating
motor unit numbers and size of the median
nerve detected a higher incidence of motor
abnormalities (AUC 0.84) than motor NCS of
the median, tibial, and peroneal nerve, includ-
ing measures of minimum F-wave latency
(AUC 0.79). This may be due to resprouting
and collateral reinnervation of motor units
resulting in conventional NCS not being able
to detect motor changes before substantial
motor dysfunction is present. This is supported
FIGURE 9.2 ROC curves of parameters from the motor
by the study by Daube et al. who found that unit number (MUNE) estimation method MScanFit and
decreased CMAP amplitudes cannot be nerve conduction studies (NCS) of the median nerve and
detected until about 50% of the motor units are their individual ability to discriminate between patients
lost [10]. Fig. 9.2 illustrates ROC curves of with DSPN and healthy controls. Motor unit number esti-
mation from MScanFit (MUNE). The smallest amplitude of
MScanFit and NCS parameters and their ability
the largest motor units making up 50% of the compound
to discriminate between healthy subjects and motor action potential amplitude calculated by MScanFit
patients with DSPN. In Fig. 9.3, loss of motor (A50%). Sensory nerve action potential amplitude in µV
units is presented from a patient with DSPN (SNAP). Compound muscle action potential in mV
and compared with a healthy subject. (CMAP). MUNE parameters provide higher AUC values
than CMAP and comparable accuracy to SNAP. Source:
Comparisons of MScanFit with NCS in the
Adopted from Fig. 3 of Kristensen AG, et al. Detection of early
lower extremity of patients with type 2 diabe- motor involvement in diabetic polyneuropathy using a novel
tes show similar findings as for the upper MUNE method - MScanFit MUNE. Clin Neurophysiol
extremity [9]. Kristensen et al. found a substan- 2019;130(10):1981 7, copyright Clinical Neurophysiology.
tial decrease in the motor unit numbers in the
anterior tibial muscle of patients with DSPN with an unusual presentation. However, with
(AUC 0.90) as compared to CMAP amplitudes the emerging evidence that MUNE methods
(AUC 0.77). Other studies applying similar provide motor measures with comparable sen-
MUNE methods in patients with type 1 and sitivity to sensory NCS parameters, future clin-
type 2 diabetes have shown similar results ical studies could determine the utility of
[34 36]. These emerging MUNE studies are MScanFit MUNE as a clinical tool in DSPN.
important to obtain a better understanding of
the pathophysiology in DSPN.
Based on the symptomatology a common
assumption is that DSPN primarily targets sen-
Muscle strength and functional
sory nerves [3]. Findings of studies on MUNE
impairments
methods clearly demonstrate that motor and
sensory nerves are equally affected [12]. Still,
Muscle strength evaluation
in the clinical setting EMG investigations are Motor examination may include (1) func-
often considered redundant, except in cases tional tests, (2) manual muscle testing, and (3)

Diabetic Neuropathy
 Muscle strength and functional impairments 139
FIGURE 9.3 Motor unit number
estimation (MUNE) from MscanFit of a
patient with DSPN (B) and a healthy
control subject (A) illustrating the loss
of motor units and incomplete reinner-
vation observed following DSPN.
CMAP, Compound motor action poten-
tial. Source: Courtesy by Hatice Tankisi.

dynamometric evaluation, including hand-held Functional tests are easily implemented with-
dynamometers and stationary dynamometers, out the need of testing equipment and are part
providing quantification of isokinetic or iso- of a standard neurological examination. In early
metric muscle strength. studies, the ability to walk on heels and toes has

Diabetic Neuropathy
140 9. Motor dysfunction in diabetes

often been applied as measures reflecting mus- healthy subjects [41,42]. Dynamometry pro-
cle strength of the ankle dorsal and plantar flex- vides clear advantages compared to manual
ors [37]. However, these tests are linked with and functional testing when it comes to quanti-
limitations. First, inability to walk on heels or fying muscle strength. Isometric and isokinetic
toes is only present in patients with severe dynamometries enable quantitative informa-
DSPN [37]. Second, since the test is dichoto- tion on the static and dynamic muscle strength
mous, with only two possible outcomes, grad- of all major muscle groups of the upper and
ing of motor dysfunction is not possible, lower extremities. Furthermore, with strictly
providing a poor measure for differentiation standardized testing procedures, low variation
and evaluation of changes in longitudinal stud- for repeated measures (,10%) and day-to-day
ies. Other functional motor tests include the 6- testing can be obtained [43]. Fig. 9.4 depicts the
minute walk test and 5-time sit-to-stand test. set-up for evaluation of knee flexors and exten-
These tests reflect the overall functional muscle sors with the use of the isokinetic dynamome-
capacity and endurance and provide reliable ter Biodex-system 3 PRO-dynamometer
measures in identifying the risk of falling in
patients with diabetes [38]. Moreover, functional
tests have proved valuable as prognostic indica-
tors of mortality in the elderly population as
well as for evaluation of skeletal muscle health,
a factor increasingly considered important for
the overall health of patients with diabetes
[18,39]. However, the exact relation between
abnormal functional test findings and muscle
strength remains unknown, and likely a number
of other causes may affect functional motor
capacity, including loss of proprioceptive sensa-
tion, poor vision, pain, overweight, arthritis,
ageing, and limited joint mobility.
Manual muscle testing is part of a standard
neurological examination and provides semi-
quantitative measures of muscle strength. As
with functional tests, it is easily implementable.
Still, manual muscle testing is highly
examiner-dependent with low intra- and inter-
rater reproducibility. Even in experienced
hands, it is insensitive compared to quantita-
tive assessments of muscle strength, especially
in strong muscle groups with a short lever arm
[40]. Thus manual muscle testing cannot reli-
ably detect even a 50% reduction of strength of
the plantar flexors.
Dynamometric examination of isokinetic and iso-
metric muscle strength has become increasingly
popular in research settings for the evaluation FIGURE 9.4 BioDex setup for evaluation of maximal
of muscle strength in patients as well as muscle strength of the knee flexors and extensors.

Diabetic Neuropathy
 Muscle strength and functional impairments 141
(BioDex Medical Systems, Inc., NY, United weakness, assessed by the presence of foot
States). Normative reference material of drop, was evident in a very small proportion
healthy participants exists applying the above of patients with diabetes (6% and 1% of
BioDex set-up enabling correction for the influ- patients with type 1 and 2 diabetes, respec-
ence of obvious explanatory variables of age, tively). Consequently, motor deficits were
sex, height, and weight [44]. believed to occur only in relation to more
Determination of handgrip strength can eas- severe DSPN [37,51]. In a large-scale study
ily be performed as it only requires a handgrip applying manual muscle testing of elderly
dynamometer. However, in most patients with people ( . 70 years), Hiltunen et al. observed
diabetes this is not relevant since the upper decreased muscle strength in participants
extremities are infrequently weakened and if with diabetes compared to those without dia-
so, may be due to mononeuropathies such as betes at the upper extremity (thenar and
entrapment of the median nerve rather than interosseous muscles) as well as the lower
DSPN. extremity (tibialis anterior, tibialis posterior,
and peroneus). In contrast to earlier studies,
these findings suggested a more widespread
Diabetic myopathy muscle impairment in patients with diabetes
The importance of muscle impairments in dia- [52].
betes to the overall health of patients is becom- Subsequent studies with quantitative muscle
ing increasingly evident [18]. Deteriorating strength assessment support the observations
muscle health is linked with postural instability, by Hiltunen et al. Studies applying isokinetic
dyslipidemia and dysregulated blood glucose dynamometry of extensor and flexor muscle
levels with increased risk of diabetic complica- groups have shown the loss of muscle strength
tions, morbidity, and ultimately increased mor- in patients with type 1 as well as type 2 diabe-
tality [38,45,46]. Loss of muscle strength in tes at both ankle and knee level [17,43,53]. In
relation to atrophy, following DN, is well estab- all studies, muscle strength proximally as well
lished. It is the general understanding that mus- as distally proved to be related to the presence
cle biomechanical deficits in patients with and severity of DSPN. Notably, neither in
diabetes are related to the presence and severity patients with type 1 nor type 2 diabetes did
of DN. However, studies have documented muscle impairment independently relate to the
impaired muscle function in patients with diabe- degree of other chronic diabetic complications
tes unrelated to DN also [47 50]. The notion (retinopathy or nephropathy) [43,53]. In Stouge
that muscle dysfunction in diabetes is affected et al., plantar flexors had the largest reduction
by metabolic as well as neuropathic factors is in muscle strength (238%), based on the stan-
gradually becoming more apparent. Still, the dard deviation from the expected strength of
effects of diabetes and obesity on muscle func- healthy reference material, followed by the
tion, irrespective of DN, are largely unknown. knee extensors (227%) and dorsiflexors
(219%) [17]. Muscle weakness of knee flexors
in patients with DSPN did not significantly dif-
Diabetic neuropathy and muscle fer from healthy subjects after correction for
impairment multiple testing. In line with Stouge et al.,
Andersen proved that plantar flexors were
Muscle strength weakened at least to the same degree as the
Preceding quantitative studies of muscle dorsiflexors [43]. These findings challenge the
strength in patients with DSPN, muscle belief that the dorsal flexors are more prone to

Diabetic Neuropathy
142 9. Motor dysfunction in diabetes

muscle damage following DSPN. Rather, foot extensors (232%) and plantar flexors (234%)
drop may not be a consequence of more severe in a study group of patients with less severe
involvement of the dorsal flexors, but a conse- DSPN [16]. Proximal muscle deficits may par-
quence of dorsiflexors being a relatively smal- tially be contributed to DSPN; however, meta-
ler muscle group more easily resulting in bolic factors of diabetes and insulin resistance
clinically detectable deficits. Fig. 9.5 depicts the per se may also contribute. In Stouge et al.,
loss of muscle strength observed in flexors and multiple regression analyses proved that MRI
extensors at knee and ankle level in patients detectable muscle changes at thigh level,
with DSPN as compared to patients without including fat infiltration, to a higher degree
DSPN and healthy subjects. were related to the loss of muscle strength and
Interestingly, recent studies have suggested higher BMI than DSPN per se, supporting the
a larger involvement of proximal muscle assumption that additional factors, irrespective
groups than previously expected [16,17]. The of DSPN, may be of importance for muscle
understanding of the pathology of DSPN with impairment in diabetes [17]. Thus a clear dis-
a clear distal proximal gradient of sensory tal proximal gradient of motor deficits may
symptoms suggests a similar pattern should be not be evident until later stages of DSPN.
detectable for motor symptoms. Indeed, previ- In a 7 8 years follow-up study, patients
ous studies of patients with severe DSPN have with type 1 and 2 diabetes and DSPN, enrolled
found that weakness of ankle muscle groups is in the initial studies by Andersen et al. [43,53],
more pronounced than at the knee level, sup- were observed to have accelerated the loss of
porting a distal proximal gradient [43]. In con- muscle strength as compared to patients with-
trast, Almurdhi et al. observed a similar out DPSN and healthy controls. This highly
reduction in muscle strength between knee suggests motor dysfunction in diabetes to be a

FIGURE 9.5 Isokinetic muscle strength of patients with type 2 diabetes and DSPN ( 1 DSPN), patients without DSPN
(2DSPN), and healthy control subjects (Healthy Controls). Muscle strength is expressed as percentage (%) of predicted
muscle strength from normative reference material of healthy subjects adjusting for the explanatory variables age, sex,
height, and weight [44]. The bar plot clearly illustrates the lower muscle strength observed at ankle level as well as knee
level in patients with DSPN as compared to healthy subjects and patients without DSPN. *Differences in muscle strength
of knee flexors were not statistically significant after adjusting for multiple testing. Source: Adopted from muscle strength data
for Stouge A, et al. MRI of skeletal muscles in participants with type 2 diabetes with or without diabetic polyneuropathy. Radiology
2020;297(3):608 19.

Diabetic Neuropathy
 Muscle strength and functional impairments 143
progressive disorder [13]. In a subsequent twitch average rates of rise (240%, Nm/ms),
study, a more than 50% reduction in the mus- prolonged contraction duration due to a com-
cle volume of the lower leg was observed dur- bination of longer time to peak tension and
ing a follow-up period of 9 12 years in half-relaxation time ( 1 22%, ms), as well as the
patients with type 1 diabetes [14]. Thus, in later decreased maximal rate of voluntary torque
stages of DSPN, progressive loss of muscle development (248%, Nm/s). Contractile
strength seems primarily to be due to acceler- impairment and slowing seem to explain the
ated muscle atrophy rather than decreased greater power reduction observed in patients
muscle quality. Recent MRI studies of patients with diabetes than what is explained from loss
with type 2 diabetes and DSPN support this of muscle strength alone. Fig. 9.6 clearly
notion with clear findings of fat infiltration sec- depicts the impairment of muscle power prop-
ondary to atrophy of the lower extremity erties observed in patients with DSPN in addi-
[16,17]. tion to the loss of muscle strength. Similarly,
Muscle strength at the upper extremity is Sacchetti et al. evaluated isometric and isoki-
affected to a much lesser degree [43,53]. This is netic contractions at different angular velocities
not surprising considering the distal proximal at both the upper and lower extremity in a
gradient of DSPN. Nevertheless, in patients mixed group of patients with type 1 or type 2
with type 1 diabetes, muscle strength of wrist diabetes [56]. Patients were divided into two
flexors has been found to be associated with groups based on the NCS of the peroneal
the severity of DSPN [43]. nerve. Isometric strength proved similar
between groups of diabetes and healthy seden-
Muscle power tary subjects, whereas dynamic muscle
In addition to the loss of muscle strength, strength was lower in patients with diabetes
patient may elicit reduced muscle power and proved more pronounced at higher angu-
related to the degree of DSPN. Muscle power lar velocities. Moreover, isokinetic muscle
is defined as the product of muscle strength strength was lower in patients with motor
and muscle contractile velocity (work per unit nerve dysfunction than patients without nerve
time), and since daily living involves dynamic dysfunction, correspondingly increasing with
muscle action, loss of muscle power could angular velocities. These findings suggest mus-
have a greater impact on functional muscle cular deficits in patients with diabetes is best
limitation than muscle weakness per se [54]. visualized by dynamic muscle work. The slow-
Hilton et al. observed a greater reduction in ing of skeletal muscle contractile properties
power (273% to 286%) in plantar flexors and may be affected by both myogenic and neuro-
dorsiflexor in obese patients with type 2 diabe- genic factors. A preferential loss of type 2 mus-
tes as compared to the loss of muscle strength cle fibers has been suggested as an explanation
alone (241% to 279%) [55]. Allen et al. found with collateral reinnervation of type 1 motor
lower strength of dorsiflexors (235%) as well neurons. This is supported by findings of
as slower contractile properties for both volun- inverse correlations between reduced MUNE
tary (248%) and evoked contractions (240%) measures and prolonged twitch half-relaxation
(percutaneous electrical stimulation of the fibu- time of dorsiflexors [34]. Other potential expla-
lar nerve) in patients with type 2 diabetes and nations include changes in the excita-
DSPN [34]. Patients with DSPN depicted simi- tion contraction coupling and Ca21, or
lar ability as healthy subjects to activate their alterations in muscle morphology and struc-
muscles (.95%). However, slower contractile ture with fascicular disruption reducing the
properties were observed with decreased numbers of sarcomeres and hereby contractile

Diabetic Neuropathy
144 9. Motor dysfunction in diabetes

FIGURE 9.6 Muscle power properties of ankle dorsal flexors in patients with DSPN and healthy control subjects. The
table shows the reduction in muscle strength as well as muscle power properties in patients with DSPN as compared to
healthy subjects. Values are mean 6 SD. Twitch are synonymous with evoked potentials. MVC, Maximal isometric volun-
tary contraction; RTD, rate of torque development; TPT, time to peak tension; HRT, half-relaxation time. *P , .05. Source:
Adopted from Table 2 of Allen MD, et al. Skeletal muscle morphology and contractile function in relation to muscle denervation in dia-
betic neuropathy. J Appl Physiol (1985) 2014;116(5):545 52, copyright Journal of Applied Physiology.

velocity [57]. Still, the assessment of muscle following fatigue decreased to a similar degree
power in patients with diabetes is sparse, and in patients with DSPN and healthy controls
the association to DSPN is poorly understood (220%). In DSPN patients, this was further fol-
and mostly limited to studies on dorsal and lowed by a 12% reduction in CMAP. Both in
plantar flexors. healthy participants and patients with DSPN
muscle strength recovered by 2 minutes postfa-
Muscle endurance tigue. In contrast to the findings of Allen et al., a
The understanding of muscle endurance in previous study did not find that weakness of
patients with diabetes is limited due to few the extensors and flexors at the ankle and knee
studies and contradictory results. A recent was accompanied by lower muscular endurance
study observed decreased muscle endurance of [59]. In fact, following 30 maximal reciprocal
dorsiflexors in a small population (10 partici- isokinetic movements, the endurance ratio was
pants) of patients with type 2 diabetes and higher for patients with type 1 diabetes. It has
DSPN [58]. Patients with DSPN fatigued 21% been suggested that metabolic changes and
more quickly than healthy control subjects per- alterations in muscle fiber type composition
forming a sustained isometric contraction, (more type 1 fibers) could explain this finding,
despite equal levels of voluntary muscle activa- favoring less fatigability. In contrast to
tion. Maximal twitch torque immediately Andersen et al., participants in Allen et al.

Diabetic Neuropathy
 Muscle strength and functional impairments 145
elicited a larger degree of DSPN, and measures isokinetic dynamometry of the knee extensors
of voluntary muscle activation were included. and isometric dynamometry of grip strength.
Muscular endurance in patients with diabe- In addition, dual-energy X-ray absorptiometry
tes is still not well understood. The contradic- was included for the evaluation of lean body
tory findings in the above studies may be mass of the lower and upper extremity. Men
explained by the different protocols and study with type 2 diabetes exhibited lower muscle
population assessed. Furthermore, fatigue may strength, and in both genders, patients with
primarily be a neurogenic property with neu- diabetes had lower muscle strength relative to
romuscular transmission failure and therefore the amount of muscle mass, entitled muscle
may not be present unless a relatively severe quality (strength per unit muscle mass). In line
degree of DSPN is observed. Future studies with the emerging belief of a general muscle
should examine fatigue in patients with diabe- dysfunction in patients with diabetes, irrespec-
tes applying both dynamic and isometric con- tive of DSPN, the authors reported that
ditions. Other explanatory factors have also reduced muscle quality was related to the
been suggested, including muscle oxygen duration of diabetes and poor glycemic control
kinetics and microvascular dysfunction [57]. (HbA1c . 8%). Unfortunately, assessment of
peripheral nerve dysfunction was not
Respiratory muscle function included, and therefore the relation to DSPN is
unknown. Even though the difference in mus-
In diabetes, respiratory muscles may be
cle quality was statistically significant, the
impaired in patients with DSPN [60]. Applying
magnitudes were small (7% 8% reduction of
nerve stimulation of the phrenic nerve, reduc-
leg and arm muscle quality in both men and
tions of inspiratory and expiratory pressures
women with diabetes), and the variation
were related to the severity of DSPN assessed
between subjects was considerable. A similar
by a neurological disability score. Whereas
study evaluated 874 Asian participants, of
patients without DSPN, based on clinical
whom 49% were known to have diabetes [49].
examinations, were found to have normal pres-
The authors observed a threefold increase in
sures. Emerging evidence suggests that diabe-
the prevalence of sarcopenia, defined as a skel-
tes is associated with decreased pulmonary
etal muscle index (total skeletal muscle mass/
function [61]. Axonal loss within the phrenic
total body weight) at least 2 SDs lower than a
nerves may contribute to the more adverse out-
reference population, of patients with diabetes
come following admission to the intensive care
as compared to participants without diabetes.
unit in patients with diabetes. Still, only a very
These observations were evident after correc-
limited number of studies exist evaluating the
tion for age, gender, smoking, alcohol, physical
effects of DSPN on respiratory function, and a
activity, medications, and BMI. In addition,
possible relation remains largely unknown.
patients with diabetes had lower muscle qual-
ity (i.e., strength per unit of muscle mass)
although they did not have higher BMI, which
Muscle impairments in diabetes
was in contrast to Park et al. This difference
independent of neuropathy may be due to ethnic differences.
The population-based health, aging, and A 3-year follow-up reevaluated 70% of the
body composition study by Park et al. evalu- participants from the initial studies of Park
ated 2618 participants aged 70 79 years [45], et al. [47]. During follow-up, a considerable
of which 485 had type 2 diabetes. Quantitative reduction (215%) in strength of knee extensors
muscle strength was evaluated by applying was observed in patients with diabetes

Diabetic Neuropathy
146 9. Motor dysfunction in diabetes

compared to participants without diabetes. deficits assessed by the peroneal nerve. This
Muscle mass declined only by 24%, resulting underscores that diabetes per se has a detri-
in a considerable reduction in muscle quality. mental effect on muscle function.
The rapid decline of leg muscle strength and Hyperglycemia and lower insulin sensitivity
quality was significant even after correction for may independently or combinedly contribute
the influence of ethnicity, gender, physical to muscle impairment in diabetes patients.
activity, drinking, smoking, and other chronic Nomura et al. found that muscle strength of
disorders. As in the initial study, no examina- knee extensors was inversely related to insulin
tion of peripheral nerve dysfunction was per- resistance in type 2 diabetes involving both
formed. Since the increased loss of muscle patients with and without DSPN [62]. Yet,
strength was only observed in the lower although none of the participants had severe
extremity and not in the upper extremity, DSPN, these findings may still be suspected to
DSPN may have contributed to the accelerated be influenced by DSPN. In a cross-sectional
loss of muscle strength. These large-scale stud- study on 1391 participants, Sayer et al.
ies suggest that diabetes, irrespective of DSPN, reported that muscle strength of the wrist,
is related to reduced muscle quality and a assessed by grip strength, was lower not only
greater degree of sarcopenia than what would in patients with known diabetes but in partici-
be expected from age-related sarcopenia alone. pants with impaired glucose tolerance as well.
Studies assessing the association between These findings extended across the normal
diabetes and muscle impairment have largely range of glucose concentration for men and
focused on functional muscle deficits assessing remained significant even following adjust-
gait and walking speed, with limited knowl- ment for physical activity [63]. In support of
edge of the implications and presence of altera- Sayer et al., a study examining a population of
tions in quantitative muscle strength [46]. nondiabetic otherwise well-functioning older
Volpato et al. found that impaired quantitative adults found that decreased strength per kilo-
strength measures in the lower limb, including gram of muscle mass of knee extensors was
lower muscle strength and quality, were inversely related to the degree of insulin resis-
important contributing factors to functional tance [48]. These findings not only suggest that
muscle deficits assessed by gait speed [50]. The insulin insensitivity is an independent risk fac-
study included 835 randomly selected partici- tor of muscle impairment in diabetes, but also
pants at the age of 65 or older, of whom 11% further imply that muscle deficits may occur
had type 2 diabetes. Muscle strength of eight prior to the diagnosis of diabetes. Few studies
muscle groups, including knee extensors, plan- have examined muscle function in relation to
tar flexors, and dorsiflexors, was assessed hyperglycemia, and further studies are needed
using a handheld dynamometer. Furthermore, to conclude if hyperglycemia, independently of
peripheral quantitative computerized tomogra- insulin resistance, leads to lower muscle
phy was applied for muscle volume and fat quality.
infiltration calculation based on a cross- A range of other patient characteristics,
sectional area of the calf. The study included inevitable in patients with diabetes, for exam-
NCS of the superficial peroneal nerve. In a ple, obesity, low physical activity, and higher
multivariable analysis, differences in muscle age, may also independently and synergisti-
strength, muscle quality, and walking speed cally impair muscle function in patients with
between participants with and without diabe- diabetes. Age-related sarcopenia is related to
tes remained significant even after adjusting loss of lean muscle mass with atrophy of the
for peripheral arterial disease and motor individual muscle fibers as well as decreased

Diabetic Neuropathy
 Muscular composition and structure 147
muscle quality [64]. Age- and diabetes-related reinnervation with subsequent increased motor
muscle dysfunction shares several common unit size, has been suggested to explain functional
mechanisms, with diabetes accelerating the muscle impairment [57]. Motor unit alterations
decline in muscle function seen with ageing following DSPN could affect the final fine-tuning
[46]. Similarly, synergistic effects of obesity and grading of motor-outputs. Moreover, pro-
and age-related sarcopenia have been observed posed slowing of muscle contractile properties
in the elderly population [65]. In addition, in and impaired proprioception following DSPN
the study by Volpato et al., although findings challenges the reaction response to sudden change
of reduced muscle quality remained signifi- of movements or positions in daily living [72,73].
cant, an attenuation of the observed lower Fear of falling is suspected to have a major impact
muscle strength and higher fat infiltration in on daily life in patients with diabetes, expected to
patients with diabetes was observed after lead to restriction of daily physical activity, with
adjustment for BMI. A possibly independent decreased mobility, isolation, and reduced quality
effect of obesity on muscle strength and func- of life [74]; however, studies are lacking evaluat-
tion could be expected but is as of yet largely ing the impact of fear of falling on daily living
unknown. and quality of life.
In summary, studies of large cohorts have Postural stability is a complex skill, which
documented that diabetes and impaired glu- cannot be explained by alterations in muscular
cose tolerance have a negative impact on skele- function alone [67,75]. Further discussion of
tal muscle function, irrespective of DSPN, with additional factors, for example, somatosensory
observations of reduced muscle strength, lower deficits, visual and vestibular function, and
muscle quality, and accelerated sarcopenia. altered joint mobility, is considered outside the
scope of this chapter.

Muscle strength, balance, gait, and falls

Muscular composition and structure
The risk of falling relies heavily on postural
control. Muscle weakness, reduced power, and
Introduction—technical advances and
muscle endurance may all independently and
muscle composition
synergistically contribute to functional impairment
with gait instability, altered balance, and ulti- Technical advances in imaging modalities
mately increased risk of falling. Moreover, falls have enabled improved visualization of skele-
and impaired balance are major contributors to tal muscle tissue and thereby provide objective
overall morbidity and mortality in the elderly and reliable measures for skeletal muscle com-
population [66]. Indeed, multiple studies have positional and structural assessment. Both
shown that patients with diabetes are at increased ultrasonomyography, X-ray computed tomog-
risk of falling [38], and this risk is expected to fur- raphy (CT) and MRI, have been applied for
ther increase with the presence of DSPN [67]. imaging of skeletal muscles. MRI is the pre-
Several studies have linked DSPN with postural ferred option, and by far the image modality
instability and poor balance [68 70]. Moreover, most extensively covered in the literature, pro-
gait instability has been observed with altered viding detailed high-resolution images of mus-
walking pace, a wider stance, and larger move- cle tissue with clear separation of muscle tissue
ments in the anterior/posterior plane as well as from noncontractile tissue, for example, fat and
laterally [71]. Impaired neuromuscular control, fol- connective tissue [76]. Emerging quantitative
lowing the loss of motor units and incomplete MRI parameters are increasingly considered

Diabetic Neuropathy
148 9. Motor dysfunction in diabetes

valuable in diagnosis as well as monitoring of can often be attributed to hypoglycemia-

disease progression or treatment response of induced neuropathy [81]. Clinical assessment
neuropathic and muscular changes following of muscle atrophy is very insensitive with dis-
neuromuscular diseases [77,78]. Concerns tinct changes only in severe cases of DSPN.
regarding the extension of MRI to the clinical Obviously, clinical assessment does not allow
setting are (1) costs, (2) time required, and (3) for the evaluation of compositional and struc-
for practical and safety reasons, the limited tural muscle changes.
ability to evaluate very obese subjects and MRI for evaluation of skeletal muscles in
patients with pacemakers, claustrophobia, and patients with diabetes has, and still is, primar-
metal implants. Methodology of newer MRI ily applied in the experimental setting. The
studies of skeletal muscles includes time- clinical application of MRI in diabetes is lim-
optimization with faster MRI spin-sequences ited to evaluation and diagnosis of the pres-
and semiautomated MRI data-analysis pro- ence of Charcot of the feet [82]. Less common,
grams [79]. Furthermore, such protocols have MRI can be useful in diabetic patients with
proven clinically feasible even in obese patients complaints of acute muscle pain or if there is
with type 2 diabetes [17]. CT has higher spatial soft tissue swelling at the clinical examination.
resolution compared with MRI, but lower con- In particular, MRI can be used to visualize vas-
trast resolution of soft tissue involves ionizing cular properties of diabetic myopathy, such as
radiation exposure. CT is, therefore, seldomly muscle ischemia and infarction with necrosis
used for the evaluation of skeletal muscles. of muscle tissue, which can be observed in
Advantages of ultrasonomyography compared patients with long-standing diabetes [83].
to MRI and CT are low-cost equipment and the
possibility for bed-side evaluation, whereas
drawbacks of ultrasonography include lower
Foot muscle atrophy—an early sign of
resolution and high observer dependency with
lower interrater reliability of skeletal muscles.
motor deficits in DSPN
Consistent with the distal proximal gradient
of DSPN, abnormalities of the feet with the loss
of foot muscle bulk are early signs of atrophy
Clinical assessment of muscle atrophy
following motor neuropathy in DSPN. Several
Clinically perceptible changes following MRI studies have indeed reported the loss of
atrophy are predominantly related to the feet, small foot muscles with more than a 50% reduc-
including a reduction in the bulk of the exten- tion of muscle volume in patients with DSPN
sor digitorum brevis muscle [80]. Other signs [15,84,85]. In fact, atrophy of foot muscles is
include prominent extensor tendons and meta- even observed in patients without clinical signs
tarsal heads, and foot deformities with pes of DSPN, underlining the presence of motor
cavus and clawing of the toes. Such changes changes in DSPN well in advance of clinical
may lead to callus formation, which are easily findings and complaints [85]. Similar findings
detectable and often developing at high- have been observed applying ultrasonography
pressure sites. As atrophy progresses more with reduced muscle volume in both neuro-
proximally, the lower legs will be involved pathic and nonneuropathic patients with diabe-
showing reduced bulk of the calf muscles and tes [86]. Diagnostic criteria of DSPN differ in the
a more prominent anterior border of the tibial different studies, with Greenman et al. only
bone. Atrophy of small hand muscles is an assessing clinical signs of DSPN, in contrast to
infrequent finding, and if found in isolation, others also including NCS [15,87]. Still, these

Diabetic Neuropathy
 Muscular composition and structure 149
findings suggest that atrophy of foot muscles emphasize the complexity of foot deformities.
occurs in earlier phases of DSPN. Normal age- Other plausible contributing factors include
ing is associated with stagnation of the sensori- alterations in connective tissue, joint capsules,
motor function with age-related atrophy of foot ligaments, and aponeurosis. Foot ulcers are
muscles [88]. Still, DSPN clearly accelerates the another common diabetic foot complication
age-related muscle abnormalities. Fig. 9.7 with severe implications known to develop at a
depicts the severe atrophy and fat infiltration much higher rate with the presence of DSPN
observed in relation to DSPN in foot muscles as with loss of sensory and motor function.
compared to an age-matched healthy control. Patients with diabetes and foot ulcerations have
Toe deformities are a frequent finding in up to 20 times higher risk of foot amputations
patients with diabetes. The relation between than patients without diabetes [90]. DSPN with
deformities of the feet and loss of foot muscle atrophy of intrinsic foot muscles may further
bulk has, however, proved to be complex. independently be a risk factor of foot ulcera-
Deformities of the feet have been shown to tions due to changes of the trophism and struc-
develop without the presence of DSPN, and ture of the feet with altered plantar pressures
even patients with severe DSPN and foot ulcers and less protection at pressure prone areas [91].
may not show significant signs of foot deformi- In summary, motor dysfunction contributes to
ties [89]. Bus et al. assessed 20 patients with diabetic foot complications, emphasizing the
DSPN, 10 with toe deformities, and another 10 importance of early diagnostics and interven-
without. Interestingly, neither intrinsic muscle tion for motor deficits in DSPN.
atrophy nor the ratio between muscle volumes
of the intrinsic and extrinsic foot muscles was
found to be significantly correlated with the
Compositional and structural muscle
degree of toe deformities. These findings
changes of the lower extremity
Muscular atrophy and substantial fat infiltra-
tion in dorsiflexors and plantar flexors have been
observed in relation to DSPN in patients with
type 1 diabetes as well as type 2 diabetes in
numerous studies [14,17,55,92]. In line with obser-
vations of loss of muscle bulk in the feet,
Andersen et al. found that atrophy was much
more pronounced in the lower legs in patients
with type 1 diabetes than what would be expected
from the clinical examination [92]. A nearly 50%
reduction in muscle volume was observed by
MRI in contrast to a subtle reduction of 10% in
the circumference at mid-calf level. In addition,
FIGURE 9.7 MR images of the foot muscles in a patient although atrophy was also visualized at knee
with diabetes and DSPN (A) and an age-matched healthy level, a clear distal proximal gradient was
control (B). White areas within the muscle tissue are non- observed with a higher degree of atrophy distally.
contractile tissue. Images are consecutive slices in proximal A subsequent 10-year follow-up study further
to distal direction starting just below ankle level (left) going
distally on the foot (right). Images clearly illustrate the concluded that DSPN is associated with an accel-
muscular atrophy following DSPN with secondary atrophic erated degree of atrophy of both dorsiflexors and
fat infiltration. Source: Courtesy by Henning Andersen. plantar flexors [14]. As previously discussed, this

Diabetic Neuropathy
150 9. Motor dysfunction in diabetes

highly suggests that muscle atrophy is the pri- infiltration in DSPN patients. These studies
mary reason for the progressive loss of muscle suggest that although the pathogenesis of
strength in patients with diabetes. In contrast to type 2 diabetes differs from type 1 diabetes,
Park et al., intrinsic muscle strength (muscle qual- similar findings of atrophy and fat infiltra-
ity) was not significantly reduced in patients with tion exist associated to the presence and
diabetes with or without DSPN [45]. Still, in severity of DSPN. In Stouge et al., observa-
Andreassen et al. the observed muscle atrophy tions of decreased muscle quality were lim-
proved to be paralleled by a greater reduction in ited to plantar flexors and only in patients
muscle strength, which suggests differences in with DSPN. This supports the concept that
muscle quality might have gone unnoticed due to progressive loss of strength is primarily
the much smaller study population. related to muscle atrophy. Still, the authors
In a similar fashion, muscle composition argue that the preserved muscle quality in
has been assessed in type 2 diabetes patients without DSPN, as compared to Park
[16,17,55]. A recent study examined the mus- et al., could be explained by the relatively
cle composition in the lower extremity of 20 short diabetes duration and better glycemic
patients with type 2 diabetes and DSPN, and control of participants enrolled in the study.
20 patients with type 2 diabetes without Interestingly, recent MRI studies evaluating
DSPN [17]. A 1.7 2.5-fold higher fat infiltrate the entire lower leg of patients with type 2 dia-
was observed at calf level as well as at the betes and DSPN suggest that muscle groups at
thigh level as compared to healthy control the knee are involved to a larger degree and
participants and patients without DSPN. In with a less obvious distal proximal gradient
addition, despite a higher BMI of patients than previously believed [16,17]. Almurdhi
with DSPN [34] as compared to patients et al. examined 20 patients with type 2 diabetes
without DSPN [30] and healthy participants (8 with DSPN) and 20 healthy control subjects
[27], muscle volume was similar between observing a similar loss of muscle strength
groups and patients with DSPN had lower proximally as well as distally. Proximal
muscle volume in the lower extremity when strength proved related to the degree of DSPN
accounting for the difference in BMI (muscle with a further significant loss of muscle vol-
volume per unit bodyweight). In multivariate ume evident in both knee extensors and flexors
multiple regression analyses, fat infiltration [16]. This is supported by findings of fat infil-
was associated with the degree of DSPN and tration and muscle strength affected to a simi-
BMI at thigh level, but only DSPN at calf lar degree in knee extensors as compared to
level. This suggests that fat infiltration in muscle groups of the lower leg in Stouge et al.
obese type 2 diabetes patients cannot be Fig. 9.8 with cross-sectional MR images of the
attributed to differences in BMI alone, but to thigh and calf depictures the increased fat infil-
a great extend to the degree of DSPN. Similar tration in a 64-year-old patient with DSPN as
findings have been made in a previous MRI compared to an age- and gender-matched
study in a small group of obese participants healthy control subject. In contrast to Andersen
with DSPN (six participants) [55]. Patients et al., Stouge and Almurdhi et al. included less
with DSPN had significantly reduced isoki- severe DSPN patients. As previously dis-
netic muscle strength at the ankle despite an cussed, diabetes and obesity may in itself affect
approximately 30% higher body weight. muscle function, not following a distal proxi-
Furthermore, muscle volumes proved similar mal gradient, supporting the notion that mus-
between patients and healthy controls cle impairment in diabetes is a complex matter
despite a 2.2-fold higher intermuscular fat not only affected by DSPN.

Diabetic Neuropathy
 Muscular composition and structure 151

FIGURE 9.8 Cross-sectional MRI of a patient with type 2 diabetes and DSPN (A 1 B) and an age- and gender-matched
healthy control (C 1 D). Images at midthigh level (A 1 C) and midcalf level (B 1 D). Dark areas of muscle tissue illustrate
the increased fat infiltration observed in patients with DSPN. MRI slices are created from opposed-phase images acquired
applying T1-weighted 4-point Dixon sequences.

Emerging quantitative MRI parameters for direct fat quantification, with more precise
for muscle assessment and reliable measures of fat infiltration than
standard proton-spin density or T1-weighted
A range of quantitative MRI parameters imaging. Previous studies have assessed fat
has proved to be valuable as objective nonin- infiltration based on increased signal intensi-
vasive measures for muscle assessment, for ties in muscle tissue [14,16,55,92], reflecting
example, DIXON, T2-mapping, and diffusion overall noncontractile tissue rather than fat
tensor imaging (DTI). DIXON methods allow infiltration per se [17]. Nonetheless, direct

Diabetic Neuropathy
152 9. Motor dysfunction in diabetes

calculations of fat infiltration in Stouge et al. and inflammation. Moore et al. examined the
seem to correspond well with previous stud- anterior tibial muscle of a small number of
ies of noncontractile tissue estimations. DTI patients with DSPN (nine subjects) and
enables the assessment of the microstructure observed an increase of 22% in T2 values as
of skeletal muscles [93,94]. In Edalati et al. a compared to healthy control subjects [97].
greater increase in diffusion (increased mean However, this study did not account for the
diffusivity) was observed in patients with dia- higher fat infiltration in patients with DSPN,
betes during isometric contraction when which is known to affect T2 values also.
changing from rest to plantar flexion as com- Stouge et al. observed increased T2 values at
pared to healthy control participants [95]. both thigh and calf levels in patients with
This may suggest that a greater contraction of DSPN [17]. As compared to Moore et al. this
muscle fibers is needed to produce a compa- study did account for fat infiltration, reflect-
rable amount of muscle force in patients with ing muscular water changes rather than fat
type 2 diabetes as compared to healthy sub- infiltration. In addition, Stouge et al. observed
jects. Stouge et al. also included DTI, but only increased T2 values in knee extensors of
at rest, observing a more unidirectional diffu- patients without DSPN when compared to
sion [a higher fractional anisotropy (FA)], healthy controls. T2 changes may indicate
which was associated with the degree of pathological muscle changes related to dia-
DSPN and had a modest effect size in knee betic myopathy, irrespective of DSPN, such as
extensors and plantar flexors [17]. The chronic inflammation of skeletal muscles.
authors suggest that a more unidirectional This is supported by Huang et al. describing
diffusion could be related to denervation and visually increased T2 signal of muscle groups
atrophy of muscle fibers. Other possible following fat suppression at both calf and
explanations include changes in the distribu- thigh levels. In Huan et al. changes were not
tion of muscle fibers (a higher type 1 fiber isolated but predominantly visualized in
composition), as previously proposed in rela- plantar flexors and knee extensors in a popu-
tion to muscle power changes [34]. Yet, altera- lation of diabetes patients with a neuropathy
tions of muscle fiber distribution in patients prevalence of 50% [83].
with diabetes remain speculative with con- MRI has further been applied for the assess-
flicting results, with most studies observing ment of microstructural changes in relation to
no significant differences in muscle fiber com- axonal degeneration in peripheral nerves of the
position [96]. Moreover, in Stouge et al. differ- lower extremity, termed magnetic resonance
ences in DTI parameters between groups neurography (MRN). Pham et al. assessed the
were not statistically significant after correct- sciatic and tibial nerve along the entire length
ing for multiple testing. For now, the applica- from spinal nerve to ankle level in a mixed
tion of DTI on skeletal muscles is novel, group of type 1 and 2 diabetes patients [98].
studies in diabetes very limited, and hence, The study observed a clear proximal to distal
the clinical importance of DTI is to be deter- gradient of nerve damage with the highest
mined. Future studies shall examine whether abundance of nerve lesions at thigh level
DTI is helpful in the assessment of properties increasing with the severity of DSPN. This sug-
being important for contractile deficits and gests that distal fiber loss responsible for the
loss of muscle power in patients with diabetes length-dependent symptomology of DSPN
(e.g., muscle morphology and structure) [57]. may be due to an accumulation of localized
T2 relaxometry allows for assessment of mus- nerve changes predominantly at the proximal
cle water changes, for example, muscle edema site, rather than an equally diffuse extension

Diabetic Neuropathy
 Muscular composition and structure 153
along the entire nerve length, irrespective of well as the sciatic nerve of patients with severe
the systemically distribution of diabetes. This DSPN as compared to patients with type 1 dia-
is in line with similar MRI findings of predomi- betes without DSPN. Like Vaeggemose, T2
nant proximal alterations of nerves in other values did not differ between groups in Pham
polyneuropathies with comparable symptom- et al. [98]. This is in contrast to patients with
atology, including familial amyloid polyneuro- type 2 diabetes where higher T2 values were
pathy [99]. Quantitative MRI parameters of the observed in patients with DSPN as compared
sciatic and tibial nerve have identified the pres- with healthy subjects, whereas no differences
ence of an increased endoneurial flow in line in proton-spin density were observed [101]. In
with the histological and neurophysiological an earlier pilot study by Pham et al., a higher
findings of axonal loss and demyelination in number of T2 lesions were also observed in a
DSPN [100,101]. Vaeggemose et al. examined mixed group of type 1 and 2 diabetes patients
the sciatic and tibial nerve of 49 type 1 diabetes [102]. Different observations of T2-relaxation
patients and 30 healthy control subjects, and proton-spin density between patients with
including the MRI quantitative measures DTI, type 1 and type 2 diabetes may be contributed
proton-spin density, and T2-relaxation. In ROC to the different pathophysiology, with type 2
analyses, FA proved sensitive to discriminate diabetes highly related to the metabolic syn-
not only between patients without DSPN and drome in addition to hyperglycemia.
severe DSPN (AUC 0.90 0.95), but also Furthermore, possible explanations of higher
between patients without DSPN and mild values of proton-spin density include struc-
DSPN (AUC 0.76 0.79) and patients with mild tural changes due to nerve damage with infil-
DSPN and severe DSPN (AUC 0.60 0.69) tration of fibrous and fatty tissue [98,100].
[100]. Similar findings were observed for the However, quantitative MRI changes have not
apparent diffusion coefficient, a measure of been directly correlated with histological find-
overall diffusion. On the contrary, no signifi- ings, and the proposed microstructural under-
cant differences were observed in T2-relaxation lying changes remain speculative.
between groups, suggesting a superior diag-
nostic accuracy of DTI parameters. A subse-
quent study on 10 type 2 diabetes patients with
DSPN, 10 patients without DSPN, and 20
Ultrasonomyopgrahy
healthy control subjects showed similar results To determine the size and composition of
with diagnostic superiority of DTI measures skeletal muscles, ultrasonomyography has only
with FA performing best in separating groups been used in a few studies of patients with dia-
[101]. In both studies, an increase and lower betes. Severinsen et al. evaluated the extensor
restriction of diffusion was observed highly digitorum brevis muscle and the muscles
related to the degree of DSPN. Since axonal between the first and second metatarsal bone
loss is the predominant nerve pathology for [86]. For both muscle groups, ultrasonomyo-
motor deficits in DSPN, FA is a promising non- graphy revealed substantial atrophy closely
invasive measure for the detection of DSPN. related to the degree of DSPN and to the
Pham et al. observed that increased proton- CMAP obtained from the peroneal and tibial
spin density was a significant predictor of motor nerves at NCS [87]. The estimated cross-
DSPN, with proton-spin density values sectional area of the extensor digitorum muscle
increasing with the severity of DSPN [98]. proved to correlate well with the volume of
Similarly, Vaeggemose et al. observed higher foot muscles determined by MRI [86]. High-
values of proton-spin density in the tibial as resolution ultrasonography has also been

Diabetic Neuropathy
154 9. Motor dysfunction in diabetes

applied for imaging of peripheral nerves. half-time (145%) following submaximal exercise
While ultrasonography provides an excellent in patients with type 2 diabetes [110,111]. Also,
assessment of superficial nerves, more deeply Schrauwen-Hinderling et al. observed excess
situated nerves, for example, the sciatic nerves, amounts of intramyocellular lipids (IMCL) in
are not easily visualized. Based on the existing obese participants with and without DM2.
studies, it seems reasonable to conclude that to Consequently, a prevalent theory suggests that
visualize nerve lesions MRN is a more reliable mitochondrial inflexibility, including reduced
method as compared to ultrasonography [103]. mitochondrial fatty acid oxidation, may lead to
Still, ultrasonography may provide a practical, the accumulation of IMCL [112]. This is referred
cheaper, and faster alternative in the daily clin- to as “lipotoxicity” believed to affect insulin sen-
ical practice as compared to MRI when asses- sitivity. Various other factors of mitochondrial
sing changes of small foot muscles and nerve impairment may also be of importance, includ-
lesions in patients with diabetes. In addition, ing excessive amounts of mitochondrial-derived
ultrasonomyography easily elicit dynamic reactive oxygen species [113]. The direction of
muscle assessment and has proven to have an the causation is, however, still debated and,
acceptable interobserver reliability for size hence, whether insulin resistance leads to mito-
measurements of intrinsic foot muscles chondrial dysfunction or the opposite remains
(ICC 5 0.90 0.97) [104]. unsettled.
Greenman et al. measured the concentra-
tions of inorganic phosphate (Pi) and phospho-
Muscle energy properties creatine (PCr), calculating the ratio of inorganic
Consistent with muscle functional deficits, phosphate to phosphocreatine (Pi/PCr) in foot
alterations to the metabolic profile of skeletal muscles of patients with diabetes [114]. In
muscles in patients with diabetes have been doing so, Greenman et al. observed lower
reported. A metabolic inflexibility is consid- energy reserves in patients with type 2 diabe-
ered central in diabetes with failure to adjust tes, further declining with the presence of
substrate oxidation to substrate supply [105]. DSPN. This was later confirmed with similar
Studies on skeletal muscle samples have dem- results in another study from the same group
onstrated impaired mitochondrial function and [115]. In this study, patients with DSPN had
oxidative capacity with reduced NADH2-O2 reduced PCr/Pi ratio as well as total concentra-
oxidoreductase activity and downregulation of tion of 31P, reflecting reduced energy reserves
genes involved in oxidative phosphorylation in and muscle volume. These studies suggest that
patients with type 2 diabetes as well as in energy reserves decline prior to but is further
individuals with a positive family history of augmented by atrophy following DSPN.
diabetes [106 108]. Magnetic resonance spec- Metabolic impairments with reduced
troscopy (MRS) offers a noninvasive technique energy reserves and oxidative capacity may
enabling the assessment of muscle metabolism. partially explain functional muscle deficits,
Measures of phosphocreatine resynthesis rates including muscle power and endurance, as
after submaximal exercise have been widely discussed in previous sections. No study has
accepted as a measure of mitochondrial function directly evaluated biomechanical muscle func-
[109]. In line with studies on skeletal muscle tion in relation to metabolic impairments in
biopsies, MRS of phosphocreatine resynthesis diabetic patients with or without DSPN, and a
rates has shown compromised oxidative capac- possible direct association therefore remains
ity, with prolonged phosphocreatine recovery unrecognized.

Diabetic Neuropathy
 Muscular composition and structure 155

Management and interventions nondeveloping motor nerve dysfunction as com-

pared to 17% in the control group. Also, a lower
Preventive strategies percentage of the exercise group developed sen-
Management of DSPN is largely focused sory nerve dysfunction and loss of vibration sen-
on preventive strategies. Central is pharmaco- sation. Strikingly, no significant changes of BMI,
logical treatment, including insulin therapy waist circumference, or metabolic parameters
and oral antihyperglycemic agents with rigor- were observed between the groups during
ous glycemic control [2,116]. Yet, opposed to follow-up. Similarly, Kluding et al. described
type 1 diabetes, in which improved glycemic decreased symptoms and signs of neuropathy
control is shown to prevent or delay the devel- following a 10-week exercise intervention com-
opment of DSPN, even long-term glucose con- bining aerobic and resistance training [120]. No
trol only has limited effect in delaying DSPN quantifiable changes were observed in any of the
in patients with type 2 diabetes [117]. In their NCS. However, despite the short duration of
meta-analysis, Callaghan et al. reported a statis- intervention, patients experienced significant
tically nonsignificant yearly risk reduction of improvements in symptoms of DSPN assessed
0.58% in delaying DSPN following glycemic by the Michigan neuropathy screening instru-
control in patients with type 2 diabetes, con- ment and subjective pain assessment. These
trasting a reduction in type 1 diabetes of 1.84% studies suggest that the implementation of aero-
per year. bic and resistance training can postpone the
With emerging evidence, motor nerve dys- development of DSPN and may further benefit
function occurs already in early stages of in patients who already have DSPN. None of the
DSPN, and muscle impairment may even initi- participants in the study by Kluding et al. had
ate prior to DSPN [8,9,18]. Moreover, muscle severe DSPN and how such exercise interven-
health is increasingly considered pivotal to not tions may affect patients with severe motor nerve
only glycemic control but also overall well- impairment remains unclear.
being and survival in the elderly diabetic In addition to glycemic control, resistance
population [18]. Consequently, additional strat- training seems to be valuable for overall neuro-
egies, including early interventions of exercise muscular function in patients with diabetes,
regimens, are gaining increasing interest. including muscle strength, volume, and quality
[96]. Improvements in muscle strength of
Interventional strategies approximately 42% of the upper limb and 28%
Exercise-based training, including aerobic, of the lower limb have been reported in a 6-
strength, and balance training, has all proved month follow-up study of resistance training
beneficial on glycemic control and may further assessed by 1- repetition maximums (1RM) [121].
elicit direct effects on the neuromuscular system A large-scale study randomized 250 participants
[96,118]. Balducci et al. examined the effects of with diabetes to aerobic training, resistance train-
mild aerobic exercise in a 4-year study period on ing, or combined intervention in a 26-week inter-
a mixed group of patients with type 1 and 2 dia- vention program [122]. After 6-month follow-up,
betes without clinical symptoms and signs of muscle strength significantly increased evaluated
DSPN at baseline [119]. At follow-up, the ratio of from seated rows, bench press, and leg press in
patients developing DSPN was significantly the resistance and combined intervention group,
lower in patients included in the exercise group and to a lesser degree in the aerobic group. In
as compared with controls. Patients who exer- addition, patients in the aerobic and combined
cised maintained motor conduction velocity of group had increased cardiorespiratory work
the peroneal nerve during follow-up with (VO2peak), maximal workload, and treadmill

Diabetic Neuropathy
156 9. Motor dysfunction in diabetes

time. A previous study by the same group using with studies suggesting decrements in dynamic
the same setup described improved glycemic muscle work and power to be the dominant con-
control following resistance training and aerobic cern in diabetic muscle impairment [34,56]. The
training but combining the two interventions sig- Graded Resistance Exercise And Type 2 Diabetes
nificantly increased this effect [123]. Another in Older adults (The Great2DO study) is an
study evaluated resistance training in elderly ongoing initiative comparing high-intensity pro-
participants with type 2 diabetes describing a gressive power training with a traditional low-
substantial improvement of muscle strength in intensity approach. Preliminary results from 12
the lower extremities of approximately 50% months follow-up indicate that high power train-
[124]. Muscle strength improvement was only ing is more effective in improving muscle mass
accompanied by a minor increase in muscle and metabolic parameters compared with low-
mass, and the authors, therefore, argued that intensity interventions [126]. Analyses of muscle
improvement in strength was primarily due to strength were not included in the preliminary
higher muscle quality. In summary, resistance results but may prove to show similar association.
training is effective and possibly superior com- Evaluation of resistance training in patients
pared to aerobic training in improving neuro- with DSPN is very limited. Praet et al. exam-
muscular function, including muscle strength ined 11 male patients with DSPN in a 10-week
and quality, whereas combined training inter- progressive resistance and interval exercise
ventions may be the best choice for improving protocol [127]. Following the protocol, muscle
overall health in patients with diabetes. strength improved with approximately 17% of
The optimal protocol for resistance exercise is proximal muscle groups at both the upper and
still debated with evidence suggesting high- lower extremities. Patients proved to have a
intensity protocols to be superior with emphasis moderate degree of DSPN from neurophysio-
on muscle power training compared to low- or logical studies, but no detailed description of
moderate-intensity training. Indeed, patients with clinical signs or symptoms was provided.
diabetes may respond less well to resistance train- Moreover, while the study reported promising
ing as compared to healthy subjects [125]. Ibanez results regarding the implementation of resis-
et al. examined 20 patients with type 2 diabetes tance training in DSPN, it is noteworthy that
applying a progressive resistance training pro- muscle improvement was not assessed in distal
gram. The protocol included two different inten- muscle groups more frequently affected by
sity regimens, the first one including 3 5 series of DSPN. Contrary to previous belief, exercise
10 15 repetitions at 50% 70% of 1RM, and the intervention does not increase the risk of foot
second regimen including 5 6 repetitions at injuries, for example, foot ulcers, in patients
70% 80% of 1RM as well as 3 4 series of 6 8 with DSPN, but may improve symptoms and
repetitions at 30% 50% 1RM with maximum neuroplasticity of peripheral nerves, with bene-
intensity emphasizing muscle power. Patients fits of exercise interventions clearly exceeding
with type 2 diabetes had less improvement in possible foot complications [128]. A recent ini-
muscle strength at both upper and lower extremi- tiative, the ADAPT study (Activity for Diabetic
ties as compared to their healthy counterparts. Polyneuropathy), aims to assess the effects of
Yet, muscle power improved by the same degree 18 months of supervised combined aerobic,
in patients with type 2 diabetes and healthy parti- balance, and resistance training as well as die-
cipants. This points to emphasizing contraction tary counseling on DSPN progression in
velocities in exercise interventions with a higher patients with mild to moderate DSPN [129].
volume and higher speed of contraction to coun- Primary endpoints include skin biopsies for
teract deficits of diabetes. This is in agreement intraepidermal nerve fiber density and scoring

Diabetic Neuropathy
 References 157
of quality of life. Other outcomes include tests [10] Daube JR. Motor unit number estimates from A to Z.
of functional muscle function, for example, 6- J Neurol Sci 2006;242(1 2):23 35.
[11] Andersen H, Stålberg E, Gjerstad MD, Jakobsen J.
minute walk test, gait speed, and timed-up and Association of muscle strength and electrophysiologi-
go test. Only grip strength and heel raise are cal measures of reinnervation in diabetic neuropathy.
included for muscle strength assessment. Muscle Nerve 1998;21:1647 54.
Similarly, a recent study examined the effects [12] Kristensen AG, et al. Detection of early motor involve-
of 12 weeks of resistance training in a study of ment in diabetic polyneuropathy using a novel MUNE
method - MScanFit MUNE. Clin Neurophysiol
patients with type 2 diabetes with and without 2019;130(10):1981 7.
DSPN including measures of isokinetic muscle [13] Andreassen CS, Johannes J, Andersen H. Muscle
strength (ClinicalTrials.Gov NCT03252132). weakness: a progressive late complication in diabetic
However, for both initiatives results are still distal symmetric polyneuropathy. Diabetes 2006;55
pending. Consequently, how DSPN responds (3):806 12.
[14] Andreassen CS, et al. Accelerated atrophy of lower leg
to long-term exercise regimens, including resis- and foot muscles a follow-up study of long-term dia-
tance training, remains largely unknown. betic polyneuropathy using magnetic resonance imag-
ing (MRI). Diabetologia 2009;52(6):1182 91.
[15] Andersen H, Gjerstad MD, Jakobsen J. Atrophy of
intrinsic foot muscles—a measure of diabetic neuropa-
References thy. Diabetes Care 2004;27:2382 5.
[1] Callaghan BC, et al. Diabetic neuropathy: clinical man- [16] Almurdhi MM, et al. Reduced lower-limb muscle
ifestations and current treatments. Lancet Neurol strength and volume in patients with type 2 diabetes
2012;11(6):521 34. in relation to neuropathy, intramuscular fat, and vita-
[2] Pop-Busui R, et al. Diabetic neuropathy: a position min D levels. Diabetes Care 2016;39(3):441 7.
statement by the American Diabetes Association. [17] Stouge A, et al. MRI of skeletal muscles in participants
Diabetes Care 2017;40(1):136 54. with type 2 diabetes with or without diabetic poly-
[3] Feldman EL, et al. New horizons in diabetic neuropa- neuropathy. Radiology 2020;297(3):608 19.
thy: mechanisms, bioenergetics, and pain. Neuron [18] Monaco CMF, Perry CGR, Hawke TJ. Diabetic myopa-
2017;93(6):1296 313. thy: current molecular understanding of this novel
[4] Dyck PJ, et al. Diabetic polyneuropathies: update on neuromuscular disorder. Curr Opin Neurol 2017;30
research definition, diagnostic criteria and estimation (5):545 52.
of severity. Diabetes Metab Res Rev 2011;27(7):620 8. [19] Sasaki H, et al. Spectrum of diabetic neuropathies.
[5] Dyck PJ, Windebank AJ. Diabetic and nondiabetic Diabetol Int 2020;11(2):87 96.
lumbosacral radiculoplexus neuropathies: new [20] Russell JW, Karnes JL, Dyck PJ. Sural nerve myelin-
insights into pathophysiology and treatment. Muscle ated fiber density differences associated with mean-
Nerve 2002;25(4):477 91. ingful changes in clinical and electrophysiologic
[6] Jann S, Beretta S, Bramerio MA. Different types of measurements. J Neurol Sci 1996;135(2):114 17.
chronic inflammatory demyelinating polyneuropathy [21] Bagai K, et al. Electrophysiological patterns of diabetic
have a different clinical course and response to treat- polyneuropathy. Electromyogr Clin Neurophysiol
ment. Muscle Nerve 2005;32(3):351 6. 2008;48(3 4):139 45.
[7] Zochodne DW, Isaac D, Jones C. Failure of immuno- [22] England JD, Gronseth GS, Franklin G, Miller RG,
therapy to prevent, arrest or reverse diabetic lumbosa- Asbury AK, Carter GT, et al. Distal symmetric poly-
cral plexopathy. Acta Neurol Scand 2003;107 neuropathy: a definition for clinical research - Report
(4):299 301. of the American Academy of Neurology, the
[8] Tankisi H, et al. Evidence-based recommendations for American association of electrodiagnostic medicine,
examination and diagnostic strategies of polyneuropa- and the American academy of physical medicine and
thy electrodiagnosis. Clin Neurophysiol Pract rehabilitation. Neurology 2005;64(2):199 207.
2019;4:214 22. [23] Shin JB, et al. The usefulness of minimal F-wave
[9] Kristensen AG, et al. MScanFit motor unit number latency and sural/radial amplitude ratio in diabetic
estimation and muscle velocity recovery cycle record- polyneuropathy. Yonsei Med J 2000;41(3):393 7.
ings in diabetic polyneuropathy. Clin Neurophysiol [24] Madsen JG, et al. Attenuation of cortically evoked
2020;131(11):2591 9. motor-neuron potential in streptozotocin-induced

Diabetic Neuropathy
158 9. Motor dysfunction in diabetes

diabetic rats: a study about the effect of diabetes upon ankle dorsal and plantar flexors and knee extensors
cortical-initiated movement. Biomed Res Int and flexors. Eur Neurol 1997;37(4):239 42.
2020;2020:1942534. [41] Frontera WR, et al. Muscle fiber size and function in
[25] Uzun N, et al. Evaluation of asymptomatic central elderly humans: a longitudinal study. J Appl Physiol
neuropathy in type I diabetes mellitus. Electromyogr (1985) 2008;105(2):637 42.
Clin Neurophysiol 2006;46(3):131 7. [42] Bohannon RW. Hand-grip dynamometry predicts
[26] Comi G. Evoked potentials in diabetes mellitus. Clin future outcomes in aging adults. J Geriatr Phys Ther
Neurosci 1997;4(6):374 9. 2008;31(1):3 10.
[27] Moglia A, et al. Central motor conduction after mag- [43] Andersen H, Poulsen PL, Mogensen CE, Jakobsen J.
netic stimulation in diabetes. Ital J Neurol Sci 1998;19 Isokinetic muscle strength in long-term IDDM patients
(1):10 14. in relation to diabetic complications. Diabetes 1996;45:
[28] Andersen H, et al. Motor pathway function in nor- 440 5.
moalbuminuric IDDM patients. Diabetologia 1995;38 [44] Harbo T, Brincks J, Andersen H. Maximal isokinetic
(10):1191 6. and isometric muscle strength of major muscle groups
[29] Gregersen G. Latency time, maximal amplitude and related to age, body mass, height, and sex in 178
electromyography in diabetic patients. Acta Med healthy subjects. Eur J Appl Physiol 2012;112
Scand 1968;183(1 2):55 60. (1):267 75.
[30] Braune HJ. Early detection of diabetic neuropathy: a [45] Park SW, et al. Decreased muscle strength and quality
neurophysiological study on 100 patients. Electromyogr in older adults with type 2 diabetes: the health, aging,
Clin Neurophysiol 1997;37(7):399 407. and body composition study. Diabetes 2006;55
[31] Ballantyne JP, Hansen S. A quantitative assessment of (6):1813 18.
reinnervation in the polyneuropathies. Muscle Nerve [46] Kalyani RR, Corriere M, Ferrucci L. Age-related and
1982;5(9s):S127 34. disease-related muscle loss: the effect of diabetes, obe-
[32] Rashidipour O, Chan KM. Motor unit number estima- sity, and other diseases. Lancet Diabetes Endocrinol
tion in neuromuscular disease. Can J Neurol Sci 2014;2(10):819 29.
2008;35(2):153 9. [47] Park SW, et al. Accelerated loss of skeletal muscle
[33] Bostock H. Estimating motor unit numbers from a strength in older adults with type 2 diabetes: the
CMAP scan. Muscle Nerve 2016;53(6):889 96. health, aging, and body composition study. Diabetes
[34] Allen MD, et al. Skeletal muscle morphology and con- Care 2007;30(6):1507 12.
tractile function in relation to muscle denervation in [48] Barzilay JI, et al. Insulin resistance is associated with
diabetic neuropathy. J Appl Physiol (1985) 2014;116 decreased quadriceps muscle strength in nondiabetic
(5):545 52. adults aged .or 5 70 years. Diabetes Care 2009;32
[35] Allen MD, et al. Motor unit loss and weakness in asso- (4):736 8.
ciation with diabetic neuropathy in humans. Muscle [49] Kim TN, et al. Prevalence and determinant factors of
Nerve 2013;48(2):298 300. sarcopenia in patients with type 2 diabetes: the
[36] Toth C, et al. Motor unit number estimations are smal- Korean Sarcopenic Obesity Study (KSOS). Diabetes
ler in children with type 1 diabetes mellitus: a case- Care 2010;33(7):1497 9.
cohort study. Muscle Nerve 2014;50(4):593 8. [50] Volpato S, et al. Role of muscle mass and muscle qual-
[37] Dyck PJ, et al. The prevalence by staged severity of ity in the association between diabetes and gait speed.
various types of diabetic neuropathy, retinopathy, and Diabetes Care 2012;35(8):1672 9.
nephropathy in a population-based cohort: the [51] Lord SR, et al. Sensori-motor function in older persons
Rochester Diabetic Neuropathy Study. Neurology with diabetes. Diabet Med 1993;10(7):614 18.
1993;43(4):817 24. [52] Hiltunen LA. Does glucose tolerance affect elderly
[38] Khan KS, et al. Falls in individuals with type 2 diabe- persons’ balance, gait or muscle strength? Cent Eur J
tes; a cross-sectional study on the impact of motor Public Health 2001;9(1):22 5.
dysfunction, postural instability and diabetic poly- [53] Andersen H, Nielsen S, Mogensen C, Jakobsen J. Muscle
neuropathy. Diabet Med 2021;38(9):e14470. Available strength in type 2 diabetes. Diabetes 2004;53(6).
from: https://doi.org/10.1111/dme.14470. [54] Reid KF, Fielding RA. Skeletal muscle power: a critical
[39] Stephanie Studenski SP, Patel K, Rosano C, Faulkner determinant of physical functioning in older adults.
K, Inzitari M, Brach J, et al. Gait speed and survival in Exerc Sport Sci Rev 2012;40(1):4 12.
older adults. JAMA 2011;305(1). [55] Hilton TN, Tuttle LJ, Bohnert KL, Mueller MJ,
[40] Andersen H, Jakobsen J. A comparative study of isoki- Sinacore DR. Excessive adipose tissue infiltration in
netic dynamometry and manual muscle testing of skeletal muscle in individuals with obesity, diabetes

Diabetic Neuropathy
 References 159
mellitus, and peripheral neuropathy: association with of foot complications. Clin Biomech (Bristol, Avon)
performance and function. Phys Ther 2008;88 2008;23(9):1183 91.
(11):1336 44. [71] Petrofsky J, Lee S, Bweir S. Gait characteristics in peo-
[56] Sacchetti M, et al. Neuromuscular dysfunction in dia- ple with type 2 diabetes mellitus. Eur J Appl Physiol
betes: role of nerve impairment and training status. 2005;93(5 6):640 7.
Med Sci Sports Exerc 2013;45(1):52 9. [72] Brown SJ, Handsaker JC, Bowling FL, Boulton AJ,
[57] Allen MD, et al. Physiology in Medicine: neuromuscu- Reeves ND. Diabetic peripheral neuropathy compro-
lar consequences of diabetic neuropathy. J Appl mises balance during daily activities. Diabetes Care
Physiol (1985) 2016;121(1):1 6. 2015;38(6):1116 22.
[58] Allen MD, et al. Decreased muscle endurance associ- [73] Handsaker JC, Brown SJ, Bowling FL, Cooper G,
ated with diabetic neuropathy may be attributed par- Maganaris CN, Boulton AJM, et al. Contributory fac-
tially to neuromuscular transmission failure. J Appl tors to unsteadniness during walking up and down
Physiol (1985) 2015;118(8):1014 22. stairs in patients with diabetic peripheral neuropathy.
[59] Andersen H. Muscular endurance in long-term IDDM Diabetes Care 2014;37(11):3047 53.
patients. Diabetes Care 1998;21(4). [74] Hewston P, Deshpande N. Fear of falling and balance
[60] Kabitz HJ, et al. Diabetic polyneuropathy is associated confidence in older adults with type 2 diabetes melli-
with respiratory muscle impairment in type 2 diabe- tus: a scoping review. Can J Diabetes 2018;42
tes. Diabetologia 2008;51(1):191 7. (6):664 70.
[61] Diez-Manglano J, Asin Samper U. Pulmonary function [75] Sawacha Z, et al. Diabetic gait and posture abnormali-
tests in type 2 diabetes: a meta-analysis. ERJ Open Res ties: a biomechanical investigation through three
2021;7(1). dimensional gait analysis. Clin Biomech (Bristol,
[62] Nomura T, Ikeda Y, Nakao S, Ito K, Ishida K, Suehiro Avon) 2009;24(9):722 8.
T, et al. Muscle strength is a marker of insulin resis- [76] Goodwin DW. Imaging of skeletal muscle. Rheum Dis
tance in patients with type 2 diabetes: a pilot study. Clin North Am 2011;37(2):245 51 vi-vii.
Endocr J 2007;54:791 6. [77] Morrow JM, Sinclair CDJ, Fischmann A, Machado PM,
[63] Sayer AA, Dennison EM, Syddall HE, Gilbody HJ, Reilly MM, Yousry TA, et al. MRI biomarker assess-
Phillips DIW, Cooper C. Type 2 diabetes, muscle ment of neuromuscular disease progression: a pro-
strength, and impaired physical function: the tip of spective observational cohort study. Lancet Neurol
the iceberg? Diabetes Care 2005;10. 2016;15:65 77.
[64] Thompson LV. Age-related muscle dysfunction. Exp [78] Sneag DB, Tan ET. Can quantitative MRI be used to
Gerontol 2009;44(1 2):106 11. differentiate physiologic changes behind muscle
[65] Bouchard DR, Dionne IJ, Brochu M. Sarcopenic/ weakness in type 2 diabetes mellitus? Radiology
obesity and physical capacity in older men and 2020;297(3):620 1.
women: data from the Nutrition as a Determinant [79] Schlaffke L, et al. Multi-center evaluation of stability
of Successful Aging (NuAge)-the Quebec longitu- and reproducibility of quantitative MRI measures
dinal Study. Obes (Silver Spring) 2009;17 in healthy calf muscles. NMR Biomed 2019;32(9):
(11):2082 8. e4119.
[66] Rubenstein LZ. Falls in older people: epidemiology, [80] Lithner F, Bergenheim T, Borssén B. Extensor digitor-
risk factors and strategies for prevention. Age Ageing um brevis in diabetic neuropathy: a controlled evalua-
2006;35(Suppl. 2):ii37 41. tion in diabetic patients aged 15 50 years. J Intern
[67] Hewston P, Deshpande N. Falls and balance impair- Med 1991;230(5):449 53.
ments in older adults with type 2 diabetes: thinking [81] Rosner L, Elstad R. The neuropathy of hypoglycemia.
beyond diabetic peripheral neuropathy. Can J Neurology 1964;14:1 6.
Diabetes 2016;40(1):6 9. [82] Ledermann HP, Morrison WB. Differential diagnosis
[68] Cavanagh PR, Simoneau GG, Ulbrecht JS. Ulceration, of pedal osteomyelitis and diabetic neuroarthropathy:
unsteadiness, and uncertainty: the biomechanical con- MR imaging. Semin Musculoskelet Radiol 2005;9
sequences of diabetes mellitus. J Biomech 1993;26 (3):272 83.
(Suppl. 1):23 40. [83] Huang BK, Monu JU, Doumanian J. Diabetic myopa-
[69] Simoneau GG, et al. Postural instability in patients thy: MRI patterns and current trends. Am J
with diabetic sensory neuropathy. Diabetes Care Roentgenol 2010;195(1):198 204.
1994;17(12):1411 21. [84] Bus SA, et al. Intrinsic muscle atrophy and toe deformity
[70] Kanade RV, et al. Investigation of standing balance in in the diabetic neuropathic foot: a magnetic resonance
patients with diabetic neuropathy at different stages imaging study. Diabetes Care 2002;25(8):1444 50.

Diabetic Neuropathy
160 9. Motor dysfunction in diabetes

[85] Greenman RL, Khaodhiar L, Lima C, Dinh T, Giurini [99] Kollmer J, et al. In vivo detection of nerve injury in
JM, Veves A. Foot small muscle atrophy is present familial amyloid polyneuropathy by magnetic reso-
before the detection of clinical neuropathy. Diabetes nance neurography. Brain 2015;138(Pt 3):549 62.
Care 2005;28(6):1425 30. [100] Vaeggemose M, et al. Magnetic resonance neurogra-
[86] Severinsen K, et al. Atrophy of foot muscles in dia- phy visualizes abnormalities in sciatic and tibial
betic patients can be detected with ultrasonography. nerves in patients with type 1 diabetes and neuropa-
Diabetes Care 2007;30(12):3053 7. thy. Diabetes 2017;66(7):1779 88.
[87] Severinsen K, Andersen H. Evaluation of atrophy of [101] Vaeggemose M, et al. Diffusion tensor imaging MR
foot muscles in diabetic neuropathy a comparative neurography detects polyneuropathy in type 2 diabe-
study of nerve conduction studies and ultrasonogra- tes. J Diabetes Complicat 2020;34(2):107439.
phy. Clin Neurophysiol 2007;118(10):2172 5. [102] Pham M, et al. Proximal neuropathic lesions in distal
[88] Inzitari M, et al. Risk and predictors of motor- symmetric diabetic polyneuropathy: findings of high-
performance decline in a normally functioning resolution magnetic resonance neurography.
population-based sample of elderly subjects: the Diabetes Care 2011;34(3):721 3.
Italian Longitudinal Study on Aging. J Am Geriatr [103] Pham M, Baumer T, Bendszus M. Peripheral nerves
Soc 2006;54(2):318 24. and plexus: imaging by MR-neurography and high-
[89] Bus SA, et al. Role of intrinsic muscle atrophy in the resolution ultrasound. Curr Opin Neurol 2014;27
etiology of claw toe deformity in diabetic neuropathy (4):370 9.
may not be as straightforward as widely believed. [104] Fernandez-Torres R, et al. Instruments of choice for
Diabetes Care 2009;32(6):1063 7. assessment and monitoring diabetic foot: a system-
[90] Rossboth S, Lechleitner M, Oberaigner W. Risk fac- atic review. J Clin Med 2020;9(2).
tors for diabetic foot complications in type 2 diabe- [105] Kelley DE, Mandarino LJ. Fuel selection in human
tes—a systematic review. Endocrinol Diabetes Metab skeletal muscle in insulin resistance: a reexamination.
2021;4(1):e00175. Diabetes 2000;49:677 83.
[91] Abri H, et al. Plantar pressure distribution in diverse [106] Kelley DE, He J, Menshikova EV, Ritov VB.
stages of diabetic neuropathy. J Diabetes Metab Dysfunction of mitochondria in human skeletal mus-
Disord 2019;18(1):33 9. cle in type 2 diabetes. Diabetes 2002;51:2944 50.
[92] Andersen H, Gadeberg HC, Brock B, Jakobsen J. [107] Mootha VK, Lindgren CM, Eriksson K, Subramanian
Muscular atrophy in diabetic neuropathy: a stereo- A, Sihag S, Lehar J, et al. PGC-1alpha-responsive
logical magnetic resonance imaging study. genes involved in oxidative phosphorylation are
Diabetologia 1997;40:1062 9. coordinately downregulated in human diabetes.
[93] Oudeman J, et al. Techniques and applications of Nature 2003;34(3):267 73.
skeletal muscle diffusion tensor imaging: a review. J [108] Patti ME, Butte AJ, Crunkhorn S, Cusi K, Berria R,
Magn Reson Imaging 2016;43(4):773 88. Kashyap S, et al. Coordinated reduction of genes of oxi-
[94] Froeling M, Oudeman J, Strijkers GJ, Maas M, Drost dative metabolism in humans with insulin resistance
MR, Nocilay K, et al. Muscle changes detected with and diabetes: potential role of PGC1 and NRF1.2003.
diffusion-tensor imaging after long-distance running. Proc Natl Acad Sci USA 2003;100(14):8466 71.
Radiology 2015;274(2). [109] Kemp GJ, Brindle KM. What do magnetic resonance-
[95] Edalati M, et al. Diffusion tensor imaging of the calf based measurements of Pi . ATP flux tell us about
muscles in subjects with and without diabetes melli- skeletal muscle metabolism? Diabetes 2012;61
tus. J Magn Reson Imaging 2019;49(5):1285 95. (8):1927 34.
Available from: https://doi.org/10.1002/jmri.26286. [110] Schrauwen-Hinderling VB, et al. Impaired in vivo
[96] Orlando G, et al. Neuromuscular dysfunction in type mitochondrial function but similar intramyocellular
2 diabetes: underlying mechanisms and effect of lipid content in patients with type 2 diabetes mellitus
resistance training. Diabetes Metab Res Rev 2016;32 and BMI-matched control subjects. Diabetologia
(1):40 50. 2007;50(1):113 20.
[97] Moore CW, et al. Reduced skeletal muscle quantity [111] Phielix E, et al. Lower intrinsic ADP-stimulated mito-
and quality in patients with diabetic polyneuropathy chondrial respiration underlies in vivo mitochondrial
assessed by magnetic resonance imaging. Muscle dysfunction in muscle of male type 2 diabetic
Nerve 2016;53(5):726 32. patients. Diabetes 2008;57(11):2943 9.
[98] Pham M, et al. Magnetic resonance neurography [112] Coen PM, Goodpaster BH. Role of intramyocelluar
detects diabetic neuropathy early and with proximal lipids in human health. Trends Endocrinol Metab
predominance. Ann Neurol 2015;78(6):939 48. 2012;23(8):391 8.

Diabetic Neuropathy
 References 161
[113] Goodpaster BH. Mitochondrial deficiency is associated [123] Sigal RJ, et al. Effects of aerobic training, resistance
with insulin resistance. Diabetes 2013;62(4):1032 5. training, or both on glycemic control in type 2 diabe-
[114] Greenman RL, et al. Early changes in the skin micro- tes: a randomized trial. Ann Intern Med 2007;147
circulation and muscle metabolism of the diabetic (6):357 69.
foot. Lancet 2005;366(9498):1711 17. [124] Brooks N, et al. Strength training improves muscle
[115] Dinh T, et al. Foot muscle energy reserves in diabetic quality and insulin sensitivity in Hispanic older
patients without and with clinical peripheral neurop- adults with type 2 diabetes. Int J Med Sci 2006;4
athy. Diabetes Care 2009;32(8):1521 4. (1):19 27.
[116] Tesfaye S, et al. Diabetic neuropathies: update on [125] Ibanez J, et al. Lower muscle strength gains in older
definitions, diagnostic criteria, estimation of severity, men with type 2 diabetes after resistance training. J
and treatments. Diabetes Care 2010;33(10):2285 93. Diabetes Complicat 2008;22(2):112 18.
[117] Callaghan B, Little AA, Feldman EL, Hughes RAC. [126] Mavros Y, Kay S, Anderberg KA, Baker MK, Wang
Enhanced glucose control for preventing and treating Y, Zhao R, et al. Changes in insulin resitance and
diabetic neuropathy (review). Cochrane Library 2012;6 HbA1c are related to exercise-mediated changes in
(6)CD007543. Available from: https://doi.org/10.1002/ body composition in older adults with type 2 diabe-
14651858.CD007543.pub2. tes: interim outcomes from the GREAT2DO trial.
[118] Streckmann F, et al. Exercise intervention studies in Diabetes Care 2013;36(8):2372 9.
patients with peripheral neuropathy: a systematic [127] Praet SF, et al. Long-standing, insulin-treated type 2
review. Sports Med 2014;44(9):1289 304. diabetes patients with complications respond well to
[119] Balducci S, et al. Exercise training can modify the short-term resistance and interval exercise training.
natural history of diabetic peripheral neuropathy. J Eur J Endocrinol 2008;158(2):163 72.
Diabetes Complicat 2006;20(4):216 23. [128] Kluding PM, Bareiss SK, Hastings M, Marcus RL,
[120] Kluding PM, et al. The effect of exercise on neuro- Sinacore DR, Mueller MJ. Physical training and
pathic symptoms, nerve function, and cutaneous activity in people with diabetic pheriperal neuropa-
innervation in people with diabetic peripheral neu- thy: paradigm shift. Am Phys Ther Assoc
ropathy. J Diabetes Complicat 2012;26(5):424 9. 2017;97:31 43.
[121] Dunstan DW, Daly RM, Owen N, Jolley D, Courten [129] Kluding PM, Singleton JR, Pasnoor M, Dimachkie
M, Shaw J, et al. High-intensity resistance training MM, Barohn RJ, Smith G, et al. Activity for Diabetic
improves glycemic control in older patients with Polyneuropathy (ADAPT): study design and protocol
type 2 diabetes. Diabetes Care 2002;25(10):1729 36. for a 2-site randomized controlled trial. Am Phys
[122] Larose J, et al. Effect of exercise training on physical Ther Assoc 2017;97:20 30.
fitness in type II diabetes mellitus. Med Sci Sports
Exerc 2010;42(8):1439 47.

Diabetic Neuropathy
 C H A P T E R

10
Diabetic autonomic neuropathy,
measurement and management; part 1:
measurement
Joseph Colombo1,2,3
1
Franklin Cardiovascular Associates, PA and Autonomic Dysfunction and POTS Center, Sewell, NJ,
United States 2Physio PS, Inc., Atlanta, GA, United States 3Neurocardiology Research Center,
Sicklerville, NJ, United States

Autonomic neuropathy cytokines, sleep apnea, high blood pressure (BP),

oxidative stress, etc.], leads to, and exacerbates,
A brief history P&S imbalance [6,7].
Autonomic Neuropathy is well established In Diabetes, these factors and symptoms are
as a “crystal ball” for morbidity and mortality mostly mediated by the Sympathetic nervous sys-
risks in patients diagnosed with Diabetes [1]. tem, causing or contributing to Sympathetic
Unfortunately, Autonomic Neuropathy is poorly Excess (SE). This SE is typically the beginning of
understood and perceived to be untreatable [2]. P&S imbalance in most patients diagnosed with
Yet significant morbidity and mortality risks may Diabetes. SE will cause a Parasympathetic Excess
now be attributable to autonomic imbalance (PE), as the still healthy P&S nervous systems
between the Parasympathetic and Sympathetic attempt to reestablish proper balance. Eventually,
(P&S) nervous systems and their regulation of prolonged SE prolongs PE, and leads to an accel-
cardiovascular and respiratory function [3], erated aging of the Parasympathetic nervous sys-
indeed the function of all organs and virtually tem, which is the beginning of Autonomic
every cell in the human body. Heart rate alone Dysfunction and its progression to Autonomic
[4], nor heart rate variability (HRV) alone [5], pro- Neuropathy. Granted, earlier and more aggressive
vides reliable diagnostic criteria for Autonomic treatment has become the norm for Diabetes,
Neuropathy. Diabetes, with its many factors based on published standards, diagnosis and
and symptoms [e.g., high blood sugar, low insu- intervention is usually not early enough and
lin, obesity, overproduction of inflammatory oftentimes therapy is too much. Since the typical

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00007-4 163 © 2022 Elsevier Inc. All rights reserved.
164 10. Diabetic autonomic neuropathy, measurement and management; part 1: measurement

measures of Diabetes are Sympathetic in nature, a threefolda or greater abnormality.b More

and the drive is to normalize the symptoms of SE, information is needed.
many times the Sympathetics are unintentionally Autonomic neuropathy references the pro-
overmedicated, exacerbating the PE and further gression from healthy to end stage. The end
accelerating the onset of Autonomic Neuropathy. stage is known as Cardiovascular Autonomic
More information is needed to improve Neuropathy (CAN, see Fig. 10.1). This terminol-
outcomes. ogy is unfortunate in that “neuropathy” is often
It has long been well established that restor- misleading. The perception of “neuropathy” is
ing proper autonomic balance reduces morbid- “dead nerves.” The teaching is that dead nerves
ity and mortality, thereby improving outcomes are not treatable. Therefore the logic is that
[8]. To do so, independent, simultaneous Autonomic Neuropathy is not treatable. This is
measurements of P&S activity is needed. compounded by the fact that the function of the
Unfortunately, virtually all autonomic mea- two individual Autonomic branches, the P&S
sures are measures of only total autonomic nervous systems, has been difficult to measure
activity (which is a measure of the combined independently and simultaneously, without
P&S effect), and the activity of the two auto- assumption and approximation. The result is
nomic branches is based on assumption and that Autonomic Neuropathy is largely unknown.
approximation or criteria that require pro- However, recently, patients have been seeking
longed suffering before a diagnosis (and there- solutions to long-standing conditions and have
fore treatment) may be considered. For driven an increased awareness.
example, it has been stated that Orthostatic Regardless of the terminology, and contrary
Hypotension (OH) is the most debilitating to the common misperception, Autonomic
symptom of Autonomic Neuropathy [9,10]. Neuropathy is treatable, even in the end stage [2];
The clinical definition of OH is a 20/10 mmHg P&S balance is the key. To help with the misper-
decrease in BP upon assuming an upright pos- ception, where appropriate, we will use
ture (i.e., sitting or standing up). A normal “Autonomic Dysfunction” or more specifically
change in BP upon standing is a 10% increase in “P&S Dysfunction” as the “new name” for
BP (e.g., approximately an 11/7 mmHg “Autonomic Neuropathy.” Hopefully, this will
increase, assuming a normal average resting also help with earlier detection. The earlier P&S
BP). Therefore a weak increase is already Dysfunction is detected, the greater the number of
abnormal, and a lack of change in BP upon available therapy options, including supplements
standing is already significantly abnormal. Yet and lifestyle modifications as well as pharma-
we are forced to make our patients wait (and ceutical therapies. While it is easier to correct
suffer) until the abnormal BP change becomes early-stage P&S Dysfunction compared with

a
Normal is a 10 mmHg increase in systolic BP, and the criterion for OH is a 20 mmHg decrease in systolic BP. The net is a
30 mmHg decrease. Comparing the abnormal decrease (30) to the normal increase (10), the result is a threefold abnormal
change, and that is just the threshold for a clinical diagnosis.
b
Waiting for the clinical OH criteria may accelerate the onset of Heart Failure. The decrease in BP upon standing may also
cause poor cardiac perfusion and reduced Diastolic pressure. The resulting poor brain perfusion often leads to
Sympathetic stimulation that increases Systolic pressure. If left undetected, the Systolic-Diastolic difference (Pulse
Pressure) may eventually grow to Hypertension and Heart Failure (two common comorbidities associated with Diabetes).
In these cases, when the Hypertension and Heart Failure are treated with Sympatholytics, the Sympatholytics may further
exacerbate the condition.

Diabetic Neuropathy
 Autonomic neuropathy 165
advanced-stage autonomic neuropathic damage, the P&S have been abnormally out of balance for
now that we are able to measure P&S Dysfunction considerably longer, but are still treatable.
in all stages, all stages are treatable [2]. By way of an example, consider the thermo-
stat on the wall of your house. Assume that
you have no knowledge of, and cannot access
Autonomic function or hear, the separate heating and cooling sys-
The main function of the Autonomic Nervous tems of your house. Further, assume that you
System (ANS) is to maintain homeostasis, like your house temperature to be 70 F (like
regardless of the conditions. The P&S dynami- heart rate at 70 bpm), and that is what your
cally adjust their output to maintain homeostasis thermostat is set to. As long as the thermostat
and apparent normalcy even in the face of reads “70,” everyone is happy. Now assume it
degraded P&S function or degraded end-organ is a hot, humid, August day in a southern US
function. The P&S nervous systems control and city, the cooling system is working well, but a
coordinate organ function. Organ dysfunction $5 heater switch fails, and the heater comes on
generates symptoms (typically not P&S dysfunc- full. What happens to the temperature in the
tion). The P&S nervous systems control and house as read on the thermostat? . . . Nothing!
coordinate normal organ function even when . . . The cooling system is working well, so it
they are in abnormal states.c (Note, if the two “amps-up” and compensates, not only for the
P&S branches are measured independently and ambient temperature, but also for the dysfunc-
simultaneously, it would enable the ability to tional heating system, and 70 is still the read-
detect dysfunction earlier.) The Sympathetics are ing. If temperature was heart rate, the heart
the reactionary branch and react to the threshold rate does not change even though both auto-
set by the Parasympathetics. With P&S nomic branches are overactive. This is more
Dysfunction, due to Diabetes for example, these information and potentially earlier detection.
P&S adjustments result in P&S imbalance and The question becomes, how long will the
begin to affect other systems within the body, dysfunctional situation last until there is a cata-
including the cardiovascular, renal, and immune strophic failure (costing $5000 or more) in one
systems, and compound the effects of the disease or both of the cooling and heating systems?
itself; Diabetes in this case. This is, in part, the Meanwhile, you are totally unaware (like your
basis for the constellation of symptoms known to heart rate is still normal, 70) until the cata-
degrade Quality of Life (QoL) in patients diag- strophic failure (like a heart attack), when it is
nosed with Diabetes (as well as many other too late. Monitoring the P&S nervous system
chronic diseases). By the time symptoms present independently and simultaneously will detect
as a result of end-organ dysfunction or failure, when the $5 switch fails. Unfortunately, most

c
For example, assume proper organ control requires a P&S activity of 4. Given normal conditions, with proper P&S
balance, that would be a P-activity of 2 and an S-activity of 2 (2 1 2 5 4). All is normal and there are no symptoms.
However, if S-activity becomes excessive (i.e., SE) and rises to a 3, and P-activity degrades to a 1, normal organ function is
still maintained because P&S activity is still a 4 (3 1 1 5 4), and there are yet to be any symptoms. An ultimate possibility is
when SE rises to a 4, and P-activity degrades to a 0. Normal organ function is still maintained because P&S activity is still
a 4 (4 1 0 5 4), and there are still no symptoms until the ultimate symptom, Sudden Cardiac Death, because there is no
longer sufficient P-activity to prevent a Sympathetically mediated Ventricular Tachyrhythm from progressing. Since we
know that Diabetes leads to P&S Dysfunction, then we should monitor P&S activity more frequently and right from the
beginning, even when asymptomatic, to detect the earliest P&S dysfunction and treat to reestablish and maintain P&S
balance.

Diabetic Neuropathy
166 10. Diabetic autonomic neuropathy, measurement and management; part 1: measurement

noninvasive, autonomic measures only measure highlighted here since Diabetes typically involves
total autonomic function. They only measure the earlier cardiovascular disease as an indication of
resulting temperature as read on the thermostat, increased mortality risk.
and the underlying abnormality is missed until it With P&S Monitoring, early-stage Dysfunction
is late in the progression. From monitoring the (the $5 switch failure in the thermostat model)
individual branches, therapy recommendations may now be detected early and clinically fol-
(as described herein) help keep the systems func- lowed, before the onset of CAN (the $5000 failure
tioning normally (healthy) and a proper balance in the thermostat model). When P&S balance is
between the cooling (Parasympathetic) and heat- restored, by normalizing the ratio of resting S-
ing (Sympathetic) systems maintained, to help activity to resting P-activity (known as
keep the “$5 switches” functioning normally, Sympathovagal Balance, or SB 5 S/P), the pro-
helping to maintain both health and wellness. gression of P&S Dysfunction is slowed or stayed.
The technique known as P&S Monitoring [5] As you know, aging is dictated by an individual’s
monitors the switches (and more) as well as the history, lifestyle, and genetic make-up. The nor-
thermostat. mal aging process (without the acceleration due to
chronic disease or disorder) causes autonomic
decline, ultimately leading to CAN. This normal
Autonomic dysfunction and morbidity
aging process is the target rate of autonomic
and mortality risk decline. This is one reason why even “healthy”
Again, CAN is end-stage P&S Dysfunction (see older people have greater morbidity and mortality
Fig. 10.1). CAN leads to the catastrophic failure in risk than younger people. CAN is inevitable for
the above model. It is approximated by a severe all and each individual is different.
lack of HRV, which is by definition a severe lack Even after the onset of CAN, normalizing SB
of resting Parasympathetic activity.d CAN places (see Fig. 10.1) slows the progression of P&S
patients at risk for increased morbidity and mortal- Dysfunction by limiting the morbidity [9] and
ity. The large cohort Framingham Heart Study [11] mortality [10] risk, optimizing QoL, and promot-
provided evidence in the general population that ing normal longevity for the individual.
lack of HRV is associated with increased mortality Establishing and maintaining normal SB, appro-
risk. Decreased HRV is associated with a number priate for the individual, also reduces secondary
of clinical conditions, including Diabetes [12 14] symptoms, which reduces medication load, pre-
and many of its comorbidities, such as Coronary vents hospitalizations and re-hospitalizations,
Artery Disease (CAD) [15 19], Hypertension and thereby lowers healthcare costs [37 39].
[20 23], altered Ventricular function [24 28], and CAN is associated with high risk of sudden
premature aging [29,30]. Mortality after a myocar- cardiac death and other Major Adverse
dial infarction (MI) is more strongly associated Cardiovascular End-points (MACE). In
with decreased Parasympathetic tone (decreased patients with Diabetes, Diabetic Autonomic
HRV) [31 34] than with poor systolic function Neuropathy (DAN) indicates an increased risk
[35]. Decreased Parasympathetic tone is an inde- of CAN and is diagnosed based on the early
pendent predictor of mortality in Congestive symptoms of neuropathy and P&S Dysfunction
Heart Failure [36]. Cardiovascular diseases are (see Fig. 10.1). DAN involves lightheadedness

d
CAN is defined as the Parasympathetic measure (RFa, or Respiratory Frequency area) , 0.1 bpm2 [5]. This indicates that
there is insufficient Parasympathetic activity to prevent, for example, a Sympathetically mediated Ventricular
Tachyrhythm from progressing. In other words, the mortality risk is high as in the Framingham Heart criteria.

Diabetic Neuropathy
 A history of assumption and approximation quantified 167
(or dizziness, arguably the worst of the symp- There are two forms of CAN: functional and
toms), as well as nervous system dysfunction, structural (see Fig. 10.1). Functional CAN patients
vascular dysfunction, gastrointestinal (GI) often find (temporary) relief of CAN once a
upset, BP dysfunction, vision decline (faster proper P&S balance is restored for them, the indi-
than presbyopia), Anhydrosis, dry and cold vidual patient. Structural CAN is not reversible.
skin, sex dysfunction, renal dysfunction, men- Both carry mortality risks. CAN risk is the same
tal and cognitive dysfunction, and cardiovascu- as for post-MI and post-coronary artery bypass
lar disease [40 43]. Much of the sensory and graft (CABG) patients, as documented in the
motor dysfunctions (including wound healing Framingham Heart Study [11], as well as for other
difficulties and amputation issues) are a func- MACE (including stroke) patients. CAN has been
tion of poor autonomic nerve control of the found to carry a 50% increase in the 2-year mortal-
vasculature. This causes vascular dysfunction, ity rate [50,51].
leading to poor tissue perfusion, predisposing CAN is typically normal for post-MI, post-
the patient to acquiring wounds and not hav- CABG, chronically ill, and geriatric patients.
ing the sensation of the wound’s existence CAN is a normal part of the aging process on
(sensory nerves need blood also), facilitating the ANS. Again, CAN simply indicates mortal-
infection (due to the poor perfusion and again ity risk. In other words, it is a sign of age, and
without sensation) and often ending in ampu- people with CAN, who tend to be older, have
tation. Clinical symptoms generally do not a greater risk of dying than people without
appear until long after the onset of Diabetes. CAN, who tend to be younger and healthier.
However, subclinical P&S Dysfunction may However, regardless of age, CAN with abnor-
occur within a year of diagnosis in type 2 mal P&S balance (SB) is not normal. CAN with
Diabetes and within two years in type 1 high SB is high risk [8] and is associated with
Diabetes [44]. In general, a diagnosis of DAN is other risk factors [52,53], including (1) low ejec-
made only after eliminating other causes of tion fraction [54,55]; (2) poor cardiac output
neuropathy. [56]; (3) arrhythmias [57,58]; (4) cardiomyopa-
CAN is defined as very low, resting thies [24,59], including chronic heart failure
Parasympathetic activity (see Fig. 10.1). CAN [60]; (5) poor circulation [61], including poor
indicates that the Parasympathetics maybe too cardiac circulation (Angina or CAD) [62]; (6)
weak to prevent a Sympathetically mediated greater mortality [10]; and (7) greater morbid-
Ventricular Tachyrhythm from progressing. ity [9], including silent MI and early cardiac
Thus CAN with high, resting Sympathetic activ- death [9,63]. CAN with low P&S balance is
ity relative to Parasympathetic activity is a mor- associated with “Broken Heart Syndrome”
tality as well as a morbidity risk indicator (e.g., [64], depression, and depression-anxiety syn-
silent MI, respiratory failure, and sudden death) dromes. Often, CAN leads to the need for car-
[45 47]. The mortality risk for DAN has been diac intervention or an implanted cardiac
estimated to be 44% within 2.5 years of diagnos- device. CAN affects longevity.
ing symptomatic autonomic neuropathy [48].
Meta-analysis data for the pooled prevalence rate
risk for silent myocardial ischemia was 1.96, A history of assumption and
with 95% confidence interval of 1.53 2.51 approximation quantified
(P , .001; n 5 1468 total subjects). Thus a consis-
tent association between autonomic neuropathy Unfortunately, early signs of P&S Dysfunction
and the presence of silent myocardial ischemia are often not recognized due to two main reasons.
was shown [49]. First, the current understanding of the effects of

Diabetic Neuropathy
168 10. Diabetic autonomic neuropathy, measurement and management; part 1: measurement

FIGURE 10.1 The progression of P&S Dysfunction. The Parasympathetic or Vagal measure is RFa (the Respiratory
Frequency area in bpm2) and the Sympathetic measure is LFa (the Low Frequency area in bpm2). The normal range for both is
0.5 10.0 bpm2. The ratio of the two (LFa/RFa, known as Sympathovagal Balance or SB) has a normal range of 0.4 3.0. The nor-
mal region (depicted in gray) is the best we get when we are born. Regardless of health condition, the P&S nervous systems
decline with age. Disease and disorder accelerate the decline. The slowest decline is to stay within the normal limits of Balance.
At the origin, with no energy left in either branch, life ceases. The stages of P&S Dysfunction start in the periphery when
patients appear normal at rest. The second stage is Diabetic Autonomic Neuropathy (DAN, or Advanced Autonomic
Dysfunction, or AAD, if not Diabetic). DAN indicates higher morbidity risk. The third stage is Cardiovascular Autonomic
Neuropathy (CAN). CAN indicates mortality risk. There are two possible states of CAN: structural and functional. If CAN is
structural, then only responses around the point A1 are possible with treatment, and while structural CAN is irreversible, treat-
ment to normalize SB minimizes or normalizes the mortality risk of CAN. If CAN is functional, then both the responses around
points A1 and A2 are possible with treatment, and functional CAN may be temporarily reversible. Treating functional by also
normalizing SB minimizes mortality risk. P&S, Parasympathetic and Sympathetic.

the ANS and its interaction with other physiologi- lightheadedness, fatigue, palpitations, difficult-
cal systems is incomplete and therefore few thera- to-manage hypertension, memory and cognitive
pies are designed for the ANS. Second, a reliable difficulties, sleep difficulties, and so on.
clinical tool to measure and monitor the P&S ner- Therapeutic intervention seems to improve out-
vous systems, independently and simultaneously, comes by balancing P&S activity for the individ-
without assumptions or approximations, did not ual and slowing or halting autonomic decline
exist until recently [5]. Prior to this, Autonomic and the associated disease progression.
Neuropathy could be clinically diagnosed [OH, The history of noninvasive ANS monitoring
Gastroparesis, Shy Drager, Postural Orthostatic in clinical practice is confusing. Traditionally, it
Tachycardia Syndrome (POTS), etc.] only at an has been based only on measures of cardiac
advanced stage with dramatic symptoms. At this beat-to-beat analyses (HRV, or beat-to-beat BP).
stage, it is usually much too late for anything but Measures of HRV, as defined in the 1996
treatment of the symptoms. With simultaneous, Circulation standards article [65,66], are mixed
independent measures of P&S branches, known or incomplete measures of the P&S. This is also
as P&S Monitoring (Physio PS, Inc., Atlanta, GA), true for beat-to-beat BP. This is not surprising, as
these patients can be identified even when they HRV or beat-to-beat BP by themselves, regard-
are still asymptomatic, or mildly symptomatic, a less of how many components are computed,
fairly common situation usually accompanied by are but one independent measure of a system

Diabetic Neuropathy
 A history of assumption and approximation quantified 169
(the ANS) that contains two components: P&S. This is an issue because, for the vast majority of
From a mathematical perspective, one measure tests and times of assessment, patients are at rest
is insufficient to fully characterize a two- (seated or supine). The average patient with
component system. If one measure changes, it is Dysautonomia has seen over a dozen doctors
impossible to determine which component chan- over years to decades with little to no relief. To
ged without making assumptions and approxi- help restore some faith in the healthcare system,
mations or without additional information. This it is explained to them that they have been
has resulted in a very low clinical acceptance assessed largely at rest, and they are shown their
rate for this method, except in extreme cases. P&S assessment results that demonstrate that the
The common assumption is that the patient is (typically) “normal” at rest. “They bet-
Parasympathetics are very, very low as com- ter be, they have had many doctors over many
pared with the Sympathetics, but this may not years working very hard to ensure they are nor-
be true after therapy, therefore follow-up testing mal . . . at rest.” Their problems typically arise
is not revealing or worse, misleading. HRV or from abnormal P&S responses while active.
beat-to-beat BP alone provides no additional
information. The use of HRV or beat-to-beat BP
alone merely indicates the obvious: that the
patient’s ANS is or is not functioning well.
Another consequence of a lack of proper mea-
surements is that there is little data on the P&S
individually, so there are very few medications
that are approved for the P&S. In fact, only two
medications are approved specifically for
Dysautonomia [Midodrine (ProAmatine) and
Northera (Droxidopa)] and their approval is lim-
ited to only Neurogenic Orthostatic Hypotension
(NOH). All other therapies for P&S Dysfunction
are off-label, supplements, or lifestyle modifica-
tions [67]. P&S Dysfunction may now be diag-
nosed earlier and followed serially [5] and treated,
including titrated based on the individual patient’s
measured responses [67]. It is hoped that newer,
improved protocols may be developed.
There is yet another poorly understood issue.
The P&S are dynamic systems. Even when an The basis for the P&S assessment test is four
individual is at rest, the P&S is not resting. While challenges separated by resting baselines. First,
the Parasympathetic nervous system is consid- there is the resting (baseline) challenge where
ered the “Rest and Digest” nervous system, it is the patient is relaxed and breathing normally,
not resting while we are resting. In fact, it may (typicallye) sitting back and relaxed in a station-
be argued that the P&S is more active when we ary, well supported chair. This is equivalent to
are sleeping than while we are awake and active. supine and establishes the patient as their own

e
Patients who cannot stand up are to lay supine during the seated phases of the test and then asked to sit with proper
posture during the stand phase of the test.

Diabetic Neuropathy
170 10. Diabetic autonomic neuropathy, measurement and management; part 1: measurement

control so that they may be treated starting the

same day. It also establishes the classical auto-
nomic measures for Autonomic Neuropathy
(DAN and CAN) and measures SB (see
Fig. 10.1). While still at rest, sitting or supine,
the Parasympathetics are independently chal-
lenged with a minute of paced breathing (deep
breathing) at 0.1 Hz or six-breathes per minute.
After a brief rest period of normal breathing, to
allow the P&S to return to baseline, the
Sympathetics are independently challenged
with a series of five short Valsalva maneuvers.
As you may remember, Valsalva maneuvers are
taught as strong Parasympathetic challenges,
and the long ones ( . 20 seconds) are indeed.
Short Valsalva maneuvers (,20 seconds) are
strong Sympathetic challenges. Of course, Diabetes: also considered a model of the
there is a tacit challenge to the opposing effect of chronic disease on P&S activity
autonomic branch inherent in these two over time
challenges as well, which may also highlight
Dysautonomia. After another brief rest Natural history of autonomic decline
period of normal breathing, to allow the P&S
to return to baseline, both P&S branches are Normal autonomic decline
challenged together in a postural change As mentioned above, autonomic decline is a
challenge: standing up or tilt-test, or if fact of life (see Fig. 10.2). Humans are born with
wheel chair bound, sitting-up from a supine as active a P&S nervous system as they will
position.f This not only helps to document have. Ultimately, at the end of life, there is no
any cause of lightheadedness (Orthostatic activity in the P or S nervous systems. The nor-
Dysfunction or Syncope), it also helps to mal aging process causes a gradual decline
document how well the two ANS branches over time. Fig. 10.2 displays the normal
coordinate. If they do not coordinate for Parasympathetic (broken, blue curve) and nor-
something as common as standing up, they mal Sympathetic (broken, red curve) changes
are not likely to coordinate well when it over time from 346 known healthy volunteers
comes to controlling the organs or organ from ages 3 to 96years (58% female). None of
systems. With this simple, safe, and these subjects are diagnosed with chronic disease
comfortable office-based clinical study, reli- and all demonstrate normal responses to the clin-
able and repeatable P&S data are available ical P&S assessment study. P&S activity is pre-
both from the resting patient and the chal- sented on the ordinate (in bpm2) and age (in
lenged patient for a greater diagnostic yield. years) is presented on the abscissa. The broken,

f
Patients that may need a cane or crutch or some other device to assist with stand are welcome to use it. If patients feel
lightheaded or dizzy while standing and do not feel they are able to continue, they are welcome to sit back down to avoid
fainting and finish the test seated. This may not only capture reasons for the lightheadedness, but also the recovery from
lightheadedness.

Diabetic Neuropathy
 Diabetes: also considered a model of the effect of chronic disease on P&S activity over time 171

FIGURE 10.2 An age-matched comparison between normal subjects and patients with Diabetes. See text for details.

black line at the 1.0 level is the normalized Diabetes effect on autonomic decline
threshold for DAN or Advanced Autonomic The effect of Diabetes Mellitus on the ANS
Dysfunction (AAD, if not diabetic), based on may be the most studied effect of all chronic
Framingham Heart Study findings. These curves diseases. Fig. 10.2 also displays the
offer several insights to normal autonomic Parasympathetic (solid, blue curve) and
decline. At both ends of the age range, Sympathetic (solid, red curve) changes over
Parasympathetic activity is normally higher than time from 1483 patients diagnosed with type 2
Sympathetic activity (SBg is in the low-normal Diabetes from ages 23 to 86 years (51%
range: 0.4 , SB , 1.0). In the early years, this Female). Diabetes’ effect on P&S function is to
reflects the Parasympathetic involvement in accelerate P&S decline (see Fig. 10.2, compare
development. In the later years, the low-normal normal and patient curves), possibly causing
SB reflects the added Parasympathetic activity earlier morbidity and mortality risk. The key
that has been shown to protect the organs, points from the patient curves in Fig. 10.2 are:
reduce morbidity and mortality, and improve
outcomes [8]. This has become the recommended • On average, P&S levels are approximately
P&S balance for geriatric patients. In the middle 50% depleted upon first diagnosis, and it is
years, the balance is approximately 1.0 (the asymptomatic.
curves overlie each other) or a little above 1.0. • This seems to reflect the American
This may reflect the need of young adults to Diabetes Association’s (ADA’s) indication
have the additional energy for child-rearing. that by the time of diagnosis, a type 2

g
SB 5 (resting Sympathetic activity)/(resting Parasympathetic activity), Normal range: 0.4 , SB , 3.0; Preferred normal for
geriatric and chronically ill: 0.4 , SB , 1.0; Preferred normal for younger patients: 1.0 , SB , 3.0.

Diabetic Neuropathy
172 10. Diabetic autonomic neuropathy, measurement and management; part 1: measurement

Diabetes patient has had the disease for • Diabetes seems to cause late-stage (relative)
up to five years [68,69]. Sympathetic dominance (SB . 2.5). The
• This confirms the ADA recommendation Sympathetic (red) curve is higher than the
for P&S Monitoring within two years of a Parasympathetic (blue) curve for the patients
diagnosis of type 2 Diabetes [68,69]. after approximately age 72, whereas, for the
• The seven years (five plus two) of the normal subjects, the Parasympathetic (blue)
disease and the prediabetic conditions curve is higher than the Sympathetic (red)
have already significantly depleted the curve after approximately age 68. Again, high
P&S, before the disease is even detected. SB with CAN is known to increase morbidity
• On-going research is looking for and mortality, including Sudden Cardiac Death
earlier indications (e.g., Prediabetes [47]. The age-matched, older, normal subjects
and Metabolic Syndrome) to indicate demonstrate low-normal SB (0.4 , SB , 1.0).
even earlier monitoring and Low-normal SB is known to reduce morbidity
intervention. and mortality [70]. Low-normal SB may be
• On average, by time symptoms present, P&S titrated [5,67]. The choice of therapy is
levels are approximately 80% depleted, and dependent on the patient’s history.
the decline was asymptomatic. • Intervention slows P&S decline.

FIGURE 10.3 A graphic demonstration of the possible effects of early detection of autonomic imbalance and inter-
vention to correct the imbalance. See text for details. Diabetes, like aging, precipitates a similar cascade of secondary
symptoms: sleep and GI disturbance, dizziness or lightheadedness, secondary hypertension, urogenital dysfunction,
higher risk of cardiovascular disease, and more. This is a recurring theme and may help to lend more insight to patho-
genesis, identification, diagnosis, and treatment of the effects of Diabetes on the P&S. P&S decline, regardless of rate,
seems to follow the same progression whether it is caused by healthy aging or chronic disease. GI, Gastrointestinal;
P&S, Parasympathetic and Sympathetic.

Diabetic Neuropathy
 Earlier detection and earlier intervention 173
• Therapy to establish and restore normal The effect of autonomically active therapy
SB, especially early, is known to slow (e.g., beta-blockers, antihypertensives, antide-
autonomic decline. Note how flat the pressants) is to help establish and maintain
patients’ curves (Fig. 10.2) are after “first normal SB. Therapy should be titrated to nor-
presentation” and intervention around malizing SB to relieve the comorbidities. Then
age 45, until decline continues (in parallel the patient becomes more stable, and the phy-
with normals) around age 65, when aging sician may be more aggressive toward the pri-
apparently catches up. mary disease.
• If therapy is adopted earlier (Fig. 10.3), it
seems as if therapy could translate the
patient’s P&S response curves to the more Earlier detection and earlier intervention
normal curves of the healthy subjects,
minimizing morbidity and mortality, and As mentioned above, the P&S nervous sys-
improving outcomes. tems are always active. Typically, the earliest
• However, on average, even with earlier changes are to the dynamic state, while the
intervention, the older patients’ SB resting state seems to remain normal. The earli-
remains greater than 1.0, but is less than est deficits are often missed but result in Small
2.5, which may be why there is a decrease Fiber Disease, which affects peripheral nerves
in mortality risk, but not a decrease in and vasculature, leading to poor peripheral cir-
morbidity risk (patients still present with culation, compounding these effects and lead-
more symptoms than normal subjects) [5]. ing to dry skin, slow and poor wound healing,
and so on. These are detected as weak
responses to either deep breathing or Valsalva.
As discussed and referenced above, poor P&S
By the time this is associated with neuropathy,
results have been correlated with poor outcomes,
the disorder has progressed to deficits in pos-
and normal P&S results have been correlated
tural change. This often causes complaints of
with healthy outcomes. For example, OH is com-
lightheadedness or dizziness and is when most
mon in Diabetes. OH is a failure of the
patients present. Early detection and interven-
Sympathetics, known as Sympathetic Withdrawal
tion during these phases significantly slow the
(SWh). SW is not a structural deficit; it is a func-
progression of the disease.
tional deficit. Identifying this difference facilitates
diagnosis and therapy.
In addition to lightheadedness (or dizziness Small fiber disorder
often from orthostasis), the common comorbid-
The Sympathetics uniquely control the vascula-
ities that are associated with Diabetes Mellitus
ture, and may be affected by Small Fiber Disorder.
include:
However, Sympathetic deficit may come first,
• hypertension secondary to the diabetes causing poor peripheral circulation leading to
• upper and lower GI upset Small Fiber Disease, or it may be from Oxidative
• sleep disturbances Stress. The underlying, primary disorder may not
• genitourinary dysfunction be known, and depending on the individual, it
• cardiovascular diseases could be either or both. Since poor circulation will
• renal disorder weaken nerves (which require relatively large
h
Sympathetic Withdrawal (SW, which we will discuss later) is the neurological basis of Orthostatic disorders. It is the
abnormal Sympathetic response to postural change where the Sympathetics decrease rather than increase.

Diabetic Neuropathy
174 10. Diabetic autonomic neuropathy, measurement and management; part 1: measurement

FIGURE 10.4 Baseline test results from a 27-year-old male (BMI 5 25.8), diagnosed with Metabolic Syndrome.
Upon questioning, he reports dry skin, Hypohidrosis, slow wound healing, and occasional sensitivity to touch (mild,
generalized pain). These five graphs are the “Response Graphs” from P&S Monitoring, presenting the average P&S
responses to the four clinical challenges of the P&S test: resting Baseline, deep breathing, Valsalva, and Stand. The
gray areas represent the normal ranges. This patient’s responses are within normal limits at rest and while standing,
challenges that test the cooperative responses of the P&S nervous systems when resting and when active, respectively.
The responses are low in response to deep breathing and Valsalva, challenges that test the independent responses of
the P&S nervous systems, respectively. This patient’s results depict what is known as Peripheral Autonomic
Neuropathy (early P&S dysfunction) and possible Small Fiber Disease. These P&S results were corroborated by an
abnormal Sudomotor test showing moderate Small Fiber Disorder. Given the symptoms (dry skin, Hypohidrosis, and
slow wound healing), these results confirm poor peripheral circulation due to autonomic involvement in Small Fiber
Disorder and the generalized pain as a direct result of Small Fiber Disorder. P&S, Parasympathetic and Sympathetic.

amounts of oxygen), causing dry skin and poor sweating (aka, Hyperhidrosis) and Hypohidrosis
wound healing, the cause of poor circulation may may indicate P&S Dysfunction, but more infor-
be the primary disorder. This points to a mation is needed.
Sympathetic affect underlying the poor peripheral Small Fiber Disorder may be associated with
circulation leading to poor wound healing and Peripheral Autonomic Neuropathy (PAN), an ear-
dry skin, which are early signs of Diabetes, and lier Autonomic Dysfunction than DAN. The small
may be an early effect of Insulin or high blood (nerve) fibers include pain nerve fibers and auto-
sugar possibly causing sugar acidosis. nomic nerve fibers. Since PAN precedes DAN
Small Fiber Disorder is commonly diagnosed (and CAN), and since PAN involves Small Fiber
with a positive Sudomotor test. Other tests Disorder, often as indicated by Autonomic
include Skin Biopsy of sweat glands (the original Dysfunction, Small Fiber Disorder may help to
test) and Quantitative Sudomotor Axon Reflex indicate earlier the stages of Autonomic
(Q-SART) or Q-Sweat. This progression of sweat Dysfunction. These earlier conditions may be
gland testing (as listed above in order) decreases documented by P&S responses as the earliest defi-
the invasiveness and increases the ease of mea- cits in autonomic testing, specifically, abnormally
surement of sweat gland function. Sudomotor low responses to either or both deep breathing or
testing passes a very low voltage, high-frequency Valsalva (see Fig. 10.4 as an example).
current across the skin between two electrodes.
The perspiration (“sweat”) on the skin reduces
the impedance between the two electrodes. Dry
Postural change P&S dysfunction
skin (too little perspiration, aka, Hypohidrosis) Fig. 10.5 depicts the possible (average) P&S
will cause high impedance and low current, responses to postural change (i.e., standi [71])
indicating Small Fiber Disease. Both too much challenge. The abscissa is the measure of
i
The five minutes of rest (the initial baseline, section “A,” of the P&S Assessment) followed by five minutes of standing
(section “F,” of the P&S Assessment) is equivalent to a passive, head-up, tilt-table study [71].

Diabetic Neuropathy
 Earlier detection and earlier intervention 175

FIGURE 10.5 Sample postural change (i.e., stand) response plots demonstrating normal and abnormal states. See text
for details.

Sympathetic activity (LFa, Low Frequency area, indication of whether the P&S branches are coor-
bpm2) and the ordinate is the measure of dinating properly.
Parasympathetic activity (RFa, Respiratory
• Normally (Fig. 10.5, upper, left graph), the
Frequency area, bpm2). The gray area indicates a
Parasympathetics first decrease (in the first
normal response. The point “A” represents the
1 3 heart beats), and then the Sympathetics
average, initial resting baseline response. The
increase (in the following 2 5 heart beats,
point “F” represents the average stand response.
normally between 10% and 500% over
The red portion of the curve represents the
resting baseline).
Sympathetic response. The blue portion of the
curve represents the Parasympathetic response. This normal response is the proper coordi-
Point “A” is the patient’s own baseline from which nation between the P&S branches: first a
the patient’s individual stand response is refer- Parasympathetic decline to facilitate and
enced, enabling testing, diagnosis, and therapy potentiate (minimize) the Sympathetic
planning from the same visit. Again, from a response and then the Sympathetic response.
diagnostic point of view, the stand response pro- Consider the analogy of a car, with its brake
vides two important pieces of information: (1) and accelerator (like the P&S, respectively).
possible cause(s) of lightheadedness, and (2) an When stopped, the brakes are applied. To go,

Diabetic Neuropathy
176 10. Diabetic autonomic neuropathy, measurement and management; part 1: measurement

first the brakes are released, and then the accel- acceleration and more gas are needed to
erator is engaged. As soon as the brakes are reach the same speed. This “more gas”
released, the car starts to roll. It is already and “acceleration” artificially raises the
accelerating with minimal gas (read that adren- Sympathetic response. As PE is relieved,
aline or Sympathetic activation) and minimal the typical amount of “gas” and
additional stress on the engine. Then only a lit- “acceleration” is revealed, which may be
tle more gas is needed to accelerate up to insufficient, as in SW, especially if the
speed. P&S coordination is very similar. This stand BP or HR response is abnormal. In
analogy includes the “See-Saw” model taught these cases, PE is treated and the therapy
in Medical School, but the car analogy goes for SW is provided as needed. Often, we
further, as will be discussed. determine that SW is the primary
Autonomic Dysfunction.
1. Abnormal: Any decrease in Sympathetic
activity is abnormal and indicates Combining these stand P&S abnormalities
(preclinical) Orthostatic Dysfunction. with abnormal HR & BP responses to stand
Decreases in the Sympathetic response to completes the diagnostic picture:
stand are called Sympathetic Withdrawal
(SW, Fig. 10.5, middle, top and middle, • SW with a significant decrease in BP upon
bottom graphs). This is an abnormal alpha- standing is associated with (preclinical) NOH.
adrenergic response. • SW with an excessive increase in HR upon
2. Abnormal: Any increase in Parasympathetic standing is associated with (preclinical)
activity is abnormal and indicates PE. PE often POTS.
presents with SE (a beta-adrenergic response), • SW with an excessive increase in BP upon
which is associated with (preclinical) Vasovagal standing is associated with (preclinical)
Syncope. PE often masks SW. A masking of Orthostatic Hypertension. This is possible,
SW is most likely the case if there is also an but very rare.
abnormal BP or HR response to stand. • By default, SW with a change in BP upon
Examples of PE are in Fig. 10.5, top, left; standing that is within normal limits is
bottom, middle; and bottom, right graphs. associated with (preclinical) Orthostatic
Examples of SE are in Fig. 10.5, bottom, right Intolerance.
and bottom, left graphs. The bottom, right • Stand SE with a little or no HR response to
graph is an example of PE with SE, indicating stand is associated with Neurogenic
(preclinical) Vasovagal Syncope. The bottom, Syncope. Apparently, Sympathetic signaling
left graph is an example of SE without PE, is not reaching the heart.
indicating possible (preclinical) Neurogenic or • Stand SE with PE activity somewhere
Cardiogenic Syncope. Neurogenic Syncope is during the test is associated with Vasovagal
indicated if the HR response is abnormal. Syncope (Fig. 10.5, bottom, right graph).
Cardiogenic Syncope is a diagnosis by • Stand SE with a normal HR response to
omission. This helps to specify and improve the stand and no PE anywhere during the P&S
diagnostic yield of Neurocardiogenic Syncope. Assessment is not Neurogenic Syncope. It
• To understand how PE masks SW, may be associated with possible (preclinical)
consider the car analogy. If the brakes are Cardiogenic Syncope; however, additional
not released and the accelerator is testing is often required to fully diagnose.
engaged anyhow (like “riding the • Neurocardiogenic Syncope, the most
brakes”), the car still goes, but more common form of Syncope, is a combination

Diabetic Neuropathy
 Earlier detection and earlier intervention 177
of two of the above. The P&S help to that seem to not tolerate antihypertensive ther-
differentiate the neurologic component. apy due to lightheadedness.
A significant number of patients (perhaps as
With the additional information available many as 40%) demonstrating abnormal
from the P&S, preclinical, including presymp- Sympathetic responses to stand are simply
tomatic, conditions that lead to lightheaded- dehydrated. A first-line therapy is proper daily
ness and other symptoms of poor cerebral hydration: 6 8 glasses of water throughout the
perfusion are elucidated. This more informa- day with fewer caffeinated, sugar, and alcohol
tion helps to guide therapy to treat or prevent drinks and possibly with electrolytes; also con-
these symptoms, as well as symptoms of poor sider titrating lower any diuretic. If resting BP
cardiac and cerebral perfusion (while upright, is high, then potassium-based electrolytes are
sitting, or standing), including brain-fog, cogni- recommended to help lower resting BP. If rest-
tive and memory difficulties, sleep difficulties, ing BP is normal to low, then Sodium-based
difficult to control (resting) BP, blood glucose, electrolytes are recommended.
or hormone levels, headache and migraine, An example of the benefit of more informa-
“Coat-Hanger” pain, Fatigue, Depression, and tion is the ability to document comorbid
Anxiety. Poor cardiac and especially cerebral Orthostatic Dysfunction (e.g., POTS) and
or brain perfusion is caused by venous pooling Vasovagal Syncope. Current conventional
from SW, with or without inefficient heart thinking refutes the possibility of the two
function due to PE. occurring together, largely because it has never
The additional information from the P&S been measured (e.g., from tilt-table testing).
stand response is especially important in geriat- The point is that these two disorders are medi-
ric patients, or sicker patients, who are at risk for ated through two different nervous systems:
falling which may cause additional problems SW is an alpha-Sympathetic response and SE is
[72]. P&S responses may preidentify these a beta-Sympathetic response. Then there is the
patients to prevent falls. From an Autonomic fact that PE may mask SW, and PE is from yet
Dysfunction perspective, early identification of another nervous system (the Parasympathetic).
stand challenge abnormalities is coincident with Therefore SE and SW may occur together, and
early identification of the advanced stages of therefore, Orthostatic Dysfunction (including
Autonomic Dysfunction (DAN and CAN). POTS) and Vasovagal Syncope may occur
P&S stand dysfunction may contribute to simultaneously, especially if Vasovagal
hypertension [38]. In these cases, the high rest- Syncope is accompanied by a drop in BP upon
ing BP may be compensatory for the drop in standing. The more P&S information docu-
BP upon standing. Treating this condition as if ments when both are present, and thus they
the hypertension was the primary condition may be treated simultaneously.
may lead to further complications, including
difficult-to-manage or Labile Hypertension.
Remember, the brain wants its blood, and there
Autonomic neuropathy scoring
are at least seven pathways through which the Over the years, there have been several scor-
brain may call for more blood. Medicine may ing protocols. There is the protocol Dr. A.I.
only block three or four of those pathways. So Vinik developed based on the three Time
the body may defeat the therapy, especially if Domain Ratios that Ewing established: the E/I
the Hypertension is secondary to abnormal Ratio based on the deep breathing challenge,
P&S stand responses. Stand abnormalities may the Valsalva Ratio, and the 30:15 Ratio based
also be associated with hypertensive patients on the postural change challenge. If two of the

Diabetic Neuropathy
178 10. Diabetic autonomic neuropathy, measurement and management; part 1: measurement

three are abnormal then DAN is indicated. If Diagnosis of autonomic dysfunction

all three are abnormal then CAN is indicated.
However, over time, Vinik proved that proto- Autonomic Dysfunction should be suspected
col ineffective in identifying DAN earlier. based on a diagnosis of a chronic disease such as
A simplified version of the Composite Diabetes Mellitus or possible chronic diseases
Autonomic Scoring Scale was developed for that are often comorbid with Diabetes, such as
patients diagnosed with Diabetes as an Hypertension or Morbid Obesity. Questionnaire
attempt to document earlier onset of CAN. assessments help to determine the level of
This scale requires expertise in beat-to-beat Autonomic Dysfunction. If (in part) Autonomic
waveform reading and tilt-table testing Dysfunction is suspected, rule out other causes
which makes it difficult to implement in the of (common) signs and symptoms. The remain-
clinic. All autonomic measures are mixed der is Autonomic Dysfunction. Early stages of
measures of P&S activity and require Autonomic Dysfunction (see Table 10.1), includ-
assumption and approximation to theorize ing PAN, involve the peripheral vasculature and
the activity of each P&S branch indepen- the P&S nerve fibers that control peripheral cir-
dently, and symptoms are required to con- culation (directly or indirectly). This involvement
firm the measurement. By the time symptoms may lead to poor peripheral circulation reducing
present, the ANS has already been dysfunc- blood flow to peripheral nerves and may acceler-
tional, perhaps for years. This is part of the ate the progression of Autonomic Dysfunction,
reason why the ADA states that by the time including symptoms of numbness, cold extremi-
Diabetes is diagnosed, the patient has had ties, tingling and/or burning sensations, and
the disease for up to two years. If prediabetes pain. Table 10.1 summarizes and compares the
and Metabolic Syndrome are included in the HRV-based phases of the progression of
progression, the duration may be up to a Autonomic Neuropathy or Dysfunction and the
decade or more. These are the major weak- P&S-based phases of the progression of
nesses of these scoring systems. Autonomic Dysfunction.

TABLE 10.1 Stages of autonomic neuropathy based on HRV-alone (total autonomic measurement) computations
and autonomic dysfunction based on P&S monitoring computations.
HRV alone PAN DAN CAN

P&S Stage 1 Stage 2 Stage 3 Stage 4 Stage 5 Stage 6

Monitoring

Nomenclature P-weakness S-release S-decline SW and PE DAN CAN

Clinical k P-response to m S-response to k S- Abnormal P&S Abnormal P- Bx P-

feature deep breathing Valsalva, released response responses to or S-response response , 0.1 bpm2
(DB) as a reflex from to postural at baseline
DB Valsalva change (Bx)
(V)

Additional Peripheral Hypertension Poor Orthostatic Increased Increased mortality

characteristics numbness, tingling, secondary to wound Dysfunction morbidity risk risk
and burningDry, Autonomic healing
cold skin Dysfunction

CAN, cardiovascular autonomic neuropathy; DAN, diabetic autonomic neuropathy; HRV, heart rate variability; P, parasympathetic; PE,
P-excess; S, sympathetic; SW, S-withdrawal.

Diabetic Neuropathy
 Diagnosis of autonomic dysfunction 179
From P&S Monitoring, the earliest stage of The first three stages are associated with
autonomic decline (Stage 1) is characterized PAN as determined by the HRV-alone meth-
by a weak Parasympathetic response to deep ods. Stage 4 is characterized by abnormal
breathing (Ewing) challenge. It has been Parasympathetic or Sympathetic responses
argued that this is a result of the differential to postural change (stand): SW and often PE.
effects of sugar acidosis on the longer, more SW is typically demonstrated even before
exposed Vagus Nerve as compared to the symptoms or an abnormal BP response, indi-
shorter, Sympathetic nerves [73]. This is fol- cating (preclinical) NOH. PE is typically
lowed by a brief Stage 2, which is character- demonstrated in those who have difficult-to-
ized by an excessive Sympathetic response to manage blood sugars or BP. SW is typically
Valsalva. The SE is enabled by the weakened the first sign of DAN. Stage 5 is characterized
Parasympathetic response to challenge, as by weak P or S responses at rest: DAN, indi-
represented by the weak Parasympathetic cating increased morbidity risk, and Stage 6
response to deep breathing. The SE is often is characterized by a very weak parasympa-
the beginning of Hypertension, a common thetic response at rest, defined as
comorbidity of Diabetes. Stage 3 is character- RFa , 0.1 bpm 2 , which is CAN, indicating
ized by both a weak Parasympathetic increased mortality risk. This staging is for
response to deep breathing and a weak adults. In pediatric patients, orthostasis
Sympathetic response to Valsalva challenge. (abnormal Parasympathetic or Sympathetic

FIGURE 10.6 The natural history of autonomic balance, based on the effect of Diabetes on the P&S nervous system. IL-6,
Interleukin-6, an inflammatory marker; HMWA/L, High Molecular Weight Adiponectin-to-Leptin ratio, an inflammatory
marker; LFa, Low Frequency area, a pure measure of Sympathetic activity; RFa, Respiratory Frequency area, a pure measure of
Parasympathetic activity; PAI-1, Plasminogen Activator Inhibitor 1, an inflammatory marker; TA/L ratio, Total Adiponectin-to-
Leptin ratio, an inflammatory marker; SB, Sympathovagal Balance, the ratio of resting Sympathetic activity to resting
Parasympathetic activity. Low RFa or LFa is a definition of Diabetic Autonomic Neuropathy (DAN), indicating increased morbid-
ity risk. Very low RFa is a definition of Cardiovascular Autonomic Neuropathy (CAN), indicating increased mortality risk. CAN
with high SB is associated with high mortality risk. Source: Adapted from Vinik A, Ziegler D. Diabetic cardiovascular autonomic
neuropathy. Circulation 2007;115:387 97.

Diabetic Neuropathy
180 10. Diabetic autonomic neuropathy, measurement and management; part 1: measurement

responses to stand, including SW or PE) or • CAN with low SB, is “Broken Heart”
Syncope may occur independent of any stage Syndrome, or
of autonomic decline. A graphical timeline of • SW or PE; and
Autonomic Dysfunction and inflammation is • renal disorder, which may be a function of
depicted in Fig. 10.6 [10]. • high SB or SE with Valsalva.
With additional P&S information, let us
Independent, simultaneous measures of the
revisit the common comorbidities that are asso-
P&S branches provide more quantitative measures
ciated with Diabetes and their associated P&S
and predictors of autonomic neuropathy (see
dysfunction:
Figs. 10.2, 10.4, and 10.5). These measures help
• lightheadedness, due to NOH, which may document, monitor, and establish balance, as mea-
be a function of sured for the individual patient at that time in their
• SW or PE or both; life, including the effects of antioxidant oxidant
• hypertension secondary to the Diabetes, (including sugar and insulin) balance at the cellu-
which may also be a function of lar level, P&S balance at the system level, and the
• SW or PE, or high SB, or SE with effects of psychosocial and physiological stresses
Valsalva; and therapy balance at the whole-body level.
• upper and lower GI upset, which may be a Establishing and maintaining proper balance for
function of the individual patient, given their history, leads to
• PE or low SB, if Gastric Dumping or improved patient outcomes, reduced morbidity
irritable bowel syndrome (IBS)-diarrhea, and mortality risks, and improved QoL.
or
• CAN, if Gastroparesis or IBS-constipation;
• sleep disturbances, which may be a function of Acknowledgments
• high SB if sleep apnea or nighttime The years of collaboration, mutual learning, and shared data
sleeplessness, or with Drs. Aaron I. Vinik, Gary L. Murray, Nicholas L. DePace,
• low SB if daytime sleepiness, or SW or PE; and many more are greatly appreciated. No funding was
• genitourinary dysfunction, which may be a received for this work. Dr. Colombo is the inventor of P&S
Monitoring and an officer and part owner of Physio PS, Inc.
function of
• insufficient Parasympathetic activity (high
SB) associated with erectile dysfunction or References
poor vaginal lubrication, or
• insufficient Sympathetic activity (low SB) [1] Vinik AI, Maser RE, Ziegler D. Neuropathy. The crystal
ball for cardiovascular disease. Diabetes Care 2010;33
associated with abnormal or absent (7):1688 90.
ejaculation or orgasm; [2] Vinik AI, Murray GL. Autonomic neuropathy is treat-
• chronic Inflammation, which may be a able. US Endocrinol 2008;2:82 4.
function of [3] Vinik AI, Maser RE, Ziegler D. Autonomic imbalance:
• overactive or prolonged Sympathetic prophet of doom or scope for hope? Diabet Med
2011;28:643 51.
activity which helps to drive excessive or [4] Hayano J, Mukai S, Sakakibara M, Okada A, Takata K,
prolonged histaminergic responses; Fujinami T. Effects of respiratory interval on vagal
• cardiovascular diseases, which may be a modulation of heart rate. Am J Physiol Heart Circ
function of Physiol 1994;267(36):H33 40.
• CAN with high SB, if MACE risk [5] Colombo J, Arora RR, DePace NL, Vinik AI. Clinical
autonomic dysfunction: measurement, indications, ther-
(including MI, Stroke, Heart Failure, apies, and outcomes. New York: Springer Science 1
Ventricular arrhythmia, etc.), or Business Media; 2014.

Diabetic Neuropathy
 References 181
[6] Vinik AI. The conductor of the autonomic orchestra. artery disease have reduced heart rate variability
Front Endocrinol 2012;3(71):1 13. mainly affecting vagal tone. Heart 2000;83:290 4.
[7] Vinik AI, Erbas T, Casellini CM. Diabetic cardiac auto- [20] Latson TW, Ashmore TH, Reinhart DJ, Klein KW,
nomic neuropathy, inflammation and cardiovascular Giesecke AH. Autonomic reflex dysfunction in
disease. J. Diabetes Invest 2013;4(1):4 18. patients presenting for elective surgery is associated
[8] Curtis BM, O’Keefe JH. Autonomic tone as a cardio- with hypotension after anesthesia induction.
vascular risk factor: the dangers of chronic fight or Anesthesiology 1994;80:326 37.
flight. Mayo Clin Proc 2002;77:45 54. [21] Parmer RJ, Cervenka JH, Stone RA, O’Connor DT.
[9] Vinik AI, Maser RE, Nakave AA. Diabetic cardiovascular Autonomic function in hypertension. Are there racial
autonomic nerve dysfunction. US Endocr Dis 2007;2 9. differences? Circulation 1990;81:1305 11.
[10] Vinik A, Ziegler D. Diabetic cardiovascular autonomic [22] Guzzetti S, Piccaluga E, Casati R, Cerutti S, Lombardi F,
neuropathy. Circulation 2007;115:387 97. Pagani M, et al. Sympathetic predominance in essential
[11] Tsuji H, Venditti Jr FJ, Manders ES, Evans JC, Larson hypertension: a study employing spectral analysis of
MG, Feldman CL, et al. Reduced heart rate variability heart rate variability. J Hypertens 1988;6:711 17.
and mortality risk in an elderly cohort. The Framingham [23] Istenes I, Keresztes K, Hermanyi Z, Putz Z, Vargha P,
Heart Study. Circulation 1994;90(2):878 83. Gandhi R, et al. Relationship between autonomic neu-
[12] Hoshi RA, Santos IS, Dantas EM, Andreão RV, ropathy and hypertension are we underestimating
Schmidt MI, Duncan BB, et al. Decreased heart rate the problem? Diabetes Med 2008;25:863 6.
variability as a predictor for diabetes a prospective [24] Alter P, Grimm W, Vollrath A, Czerny F, Maisch B.
study of the Brazilian longitudinal study of adult Heart rate variability in patients with cardiac hyper-
health. Diabetes Metab Res Rev 2019;35(7):e3175. trophy relation to left ventricular mass and etiology.
Available from: https://doi.org/10.1002/dmrr.3175. Am Heart J 2006;151(4):829 36.
[13] Benichou T, Pereira B, Mermillod M, Tauveron I, Pfabigan [25] Packer M. Neurohormonal interactions and adaptations
D, Maqdasy S, et al. Heart rate variability in type 2 diabe- in congestive heart failure. Circulation 1988;77:721 30.
tes mellitus: a systematic review and meta-analysis. PLoS [26] Rea RF, Berg WJ. Abnormal baroreflex mechanisms in
One 2018;13(4):e0195166. Available from: https://doi.org/ congestive heart failure. Recent. insights. Circulation
10.1371/journal.pone.0195166. 29608603. 1990;81:2026 7.
[14] Liao D, Carnethon M, Evans GW, Cascio WE, Heiss G. [27] Saul JP, Arai Y, Berger RD, Lilly LS, Colucci WS,
Lower heart rate variability is associated with the Cohen RJ. Assessment of autonomic regulation in
development of coronary heart disease in individuals chronic congestive heart failure by heart rate spectral
with diabetes: Atherosclerosis Risk Communities analysis. Am J Cardiol 1988;61:1292 9.
(ARIC) study. Diabetes 2002;51(12):3524 31. Available [28] Trimarco B, Lembo G, De Luca N, Volpe M,
from: https://doi.org/10.2337/diabetes.51.12.3524. Ricciardelli B, Condorelli G, et al. Blunted sympathetic
[15] Monmeneu JV, Chorro FJ, Bodi V, Sanchis J, Llacer A, response to cardiopulmonary receptor unloading in
Garcia-Civera R, et al. Relationships between heart rate hypertensive patients with left ventricular hypertro-
variability, functional capacity, and left ventricular func- phy. A possible compensatory role of atrial natriuretic
tion following myocardial infarction: an evaluation after factor. Circulation 1989;80:883 92.
one week and six months. Clin Cardiol 2001;24:313 20. [29] Piccirillo G, Fimognari FL, Munizzi MR, Bucca C,
[16] Rich MW, Saini JS, Kleiger RE, Carney RM, teVelde A, Cacciafesta M, Marigliano V. Age-dependent influence
Freedland KE. Correlation of heart rate variability with on heart rate variability in salt-sensitive hypertensive
clinical and angiographic variables and late mortality after subjects. J Am Geriatr Soc 1996;44:530 8.
coronary angiography. Am J Cardiol 1988;62:714 17. [30] Lipsitz LA, Mietus J, Moody GB, Goldberger AL.
[17] Hayano J, Sakakibara Y, Yamada M, Ohte N, Fujinami T, Spectral characteristics of heart rate variability before
Yokoyama K, et al. Decreased magnitude of heart rate and during postural tilt. Relations to aging and risk of
spectral components in coronary artery disease. Its rela- syncope. Circulation 1990;81:1803 10.
tion to angiographic severity. Circulation 1990;81:1217 24. [31] Farrell TG, Paul V, Cripps TR, Malik M, Bennett ED,
[18] Bonnemeier H, Richardt G, Potratz J, Wiegand UK, Ward D, et al. Baroreflex sensitivity and electrophysio-
Brandes A, Kluge N, et al. Circadian profile of cardiac logical correlates in patients after acute myocardial
autonomic nervous modulation in healthy subjects: infarction. Circulation 1991;83:945 52.
differing effects of aging and gender on heart rate var- [32] Schwartz PJ, Vanoli E, Stramba-Badiale M, De
iability. J Cardiovasc Electrophysiol 2003;14:791 9. Ferrari GM, Billman GE, Foreman RD. Autonomic
[19] Wennerblom B, Lurje L, Tygesen H, Vahisalo R, mechanisms and sudden death. New insights from
Hjalmarson A. Patients with uncomplicated coronary analysis of baroreceptor reflexes in conscious dogs

Diabetic Neuropathy
182 10. Diabetic autonomic neuropathy, measurement and management; part 1: measurement

with and without a myocardial infarction. Circulation [44] Pfeifer MA, Weinberg CR, Cook DL, Reenan A, Halter
1988;78:969 79. JB, Ensinck JW. Autonomic neural dysfunction in
[33] Lombardi F, Sandrone G, Pernpruner S, Sala R, recently diagnosed diabetic subjects. Diabetes Care
Garimoldi M, Cerutti S, et al. Heart rate variability 1984;7:447 53.
as an index of sympathovagal interaction after acute [45] Ziegler D. Diabetic cardiovascular autonomic neurop-
myocardial infarction. Am J Cardiol 1987;60: athy: prognosis, diagnosis and treatment. Diabetes
1239 45. Metab Rev 1994;10:339 83.
[34] La Rovere MT, Bigger JT, Marcus Jr. FI, Mortara A, [46] Valensi P. Diabetic autonomic neuropathy: what are
Schwartz PJ. Baroreflex sensitivity and heart-rate vari- the risks? Diabetes Metab 1998;24:66 72.
ability in prediction of total cardiac mortality after [47] Murray GL, Colombo J. Maintenance (r)alpha lipoic
myocardial infarction. ATRAMI (Autonomic Tone and acid reduces sudden cardiac death 43% in geriatric
Reflexes After Myocardial Infarction) Investigators. diabetes mellitus II patients with predominantly mild,
Lancet 1998;351:478 84. asymptomatic dysautonomia. 2020, Submitted to
[35] Bauer A, Kantelhardt JW, Barthel P, Schneider R, Clinical Cardiology and Cardiovascular Medicine.
Makikallio T, Ulm K, et al. Deceleration capacity of heart [48] Levitt NS, Stansberry KB, Wychanck S, Vinik AI.
rate as a predictor of mortality after myocardial infarc- Natural progression of autonomic neuropathy and
tion: cohort study. Lancet 2006;367:1674 81. autonomic function tests in a cohort of IDDM.
[36] Ponikowski P, Anker SD, Chua TP, Szelemej R, Piepoli Diabetes Care 1996;19:751 4.
M, Adamopoulos S, et al. Depressed heart rate vari- [49] Vinik AI, Casellini C, Nakave A, Patel C. Diabetic neurop-
ability as an independent predictor of death in chronic athies. Endotext.com. http://diabetesmanager.pbworks.
congestive heart failure secondary to ischemic or idio- com/w/page/17680180/Diabetic%20Neuropathies#
pathic dilated cardiomyopathy. Am J Cardiol Oxidativestress; 2009 [Last edited by Rushakoff R.].
1997;79:1645 50. [50] Maser R, Mitchell B, Vinik AI, Freeman R. The associa-
[37] Tobias H, Vinitsky A, Bulgarelli RJ, Ghosh-Dastidar S, tion between cardiovascular autonomic neuropathy
Colombo J. Autonomic nervous system monitoring of and mortality in individuals with diabetes, a meta-
patients with excess Parasympathetic responses to analysis. Diabetes Care 2003;26(6):1895 901.
Sympathetic challenges clinical observations. US [51] Vinik AI, Maser RE, Mitchell BD, Freeman R. Diabetic
Neurol 2010;5(2):62 6. autonomic neuropathy. Diabetes Care 2003;26
[38] Arora RR, Bulgarelli RJ, Ghosh-Dastidar S, Colombo J. (5):1553 79.
Autonomic mechanisms and therapeutic implications [52] DePace NL, Mears JP, Yayac M, Colombo J. Cardiac
of postural diabetic cardiovascular abnormalities. J. autonomic testing and diagnosing heart disease. “A clin-
Diabetes Sci Technol 2008;2(4):568 71. ical perspective.”. Heart Int 2014;9(2):37 44. Available
[39] Nanavati SH, Bulgarelli RJ, Vazquez-Tanus J, Ghosh- from: https://doi.org/10.5301/heartint.5000218.
Dastidar S, Colombo J, Arora RR. Altered autonomic [53] DePace NL, Mears JP, Yayac M, Colombo J. Cardiac
activity with atrial fibrillation as demonstrated by autonomic testing and treating heart disease. “A clini-
non-invasive autonomic monitoring. US Cardiol cal perspective.”. Heart Int 2014;9(2):45 52. Available
2010;7(1):47 50. from: https://doi.org/10.5301/heartint.5000216.
[40] Stansberry KB, Shapiro SA, Hill MA, McNitt PM, [54] Bullinga JR, Alharethi R, Schram MS, Bristow MR,
Meyer MD, Vinik AI. Impaired peripheral vasomotion Gilbert EM. Changes in heart rate variability are corre-
in diabetes. Diabetes Care 1996;19:715 21. lated to hemodynamic improvement with chronic
[41] Stansberry KB, Hill MA, Shapiro SA, McNitt PM, CARVEDILOL therapy in heart failure. J Card Fail
Bhatt BA, Vinik AI. Impairment of peripheral blood 2005;11(9):693 9.
flow responses in diabetes resembles an enhanced [55] Fatoni C, Raffa S, Regoli F, Giraldi F, La Rovere MT,
aging effect. Diabetes Care 1997;20:1711 16. Prentice J, et al. Cardiac resynchronization therapy
[42] Stansberry KB. Primary nociceptive afferents mediate improves heart rate profile and heart rate variability
the blood flow dysfunction in non-glabrous (hairy) of patients with moderate to severe heart failure. J Am
skin of type 2 diabetes. Diabetes Care Coll Cardiol 2005;46(10):1875 82.
1999;22:1549 54. [56] Fathizadeh P, Shoemaker WC, Woo CCJ, Colombo J.
[43] Haak ES, Usadel KH, Kohleisen M, Yilmaz A, Autonomic activity in trauma patients based on vari-
Kusterer K, Haak T. The effect of alpha-lipoic on the ability of heart rate and respiratory rate. Crit Care
neurovascular reflex arc in patients with diabetic neu- Med 2004;32(5):1300 5.
ropathy assessed by capillary microscopy. Microvasc [57] Chen JY, Fung JW, Yu CM. The mechanisms of atrial fibril-
Res 1999;58:28 34. lation. J Cardiovasc Electrophysiol 2006;17(Suppl 3):S2 7.

Diabetic Neuropathy
 References 183
[58] Copie X, Lamaison D, Salvador M, Sadoul N, DaCosta A, and Electrophysiology. Heart rate variability, stan-
Faucher L, et al.VALID Investigators Heart rate variability dards of measurement, physiological interpretation,
before ventricular arrhythmias in patients with coronary and clinical use. Circulation 1996;93:1043 65.
artery disease and an implantable cardioverter defibrilla- [66] Malik Mthe Task Force of the European Society of
tor. Ann Noninvasive Electrocardiol 2003;8(3):179 84. Cardiology and the North American Society of Pacing
[59] Debono M, Cachia E. The impact of cardiovascular auto- and Electrophysiology. Heart rate variability, stan-
nomic neuropathy in diabetes: is it associated with left ven- dards of measurement, physiological interpretation,
tricular dysfunction? Auton Neurosci 2007;132(1 2):1 7. and clinical use. Eur Heart J 1996;17:354 81.
[60] Just H. Peripheral adaptations in congestive heart fail- [67] DePace NL, Colombo J. Autonomic and mitochondrial
ure: a review. Am J Med 1991;90:23S 6S. dysfunction in clinical diseases: diagnostic, preven-
[61] Nakamura K, Matsumura K, Kobayashi S, Kaneko T. tion, and therapy. New York: Springer Science 1
Sympathetic premotor neurons mediating thermoreg- Business Media; 2019.
ulatory functions. Neurosci Res 2005;51(1):1 8. [68] American Diabetes Association. Standards of medical
[62] Manfrini O, Morgagni G, Pizzi C, Fontana F, care in diabetes 2008. Diabetes Care 2008;31(Suppl
Bugiardini R. Changes in autonomic nervous system 1):S12 54.
activity: spontaneous vs balloon-induced myocardial [69] American Diabetes Association. Standards of medical care
ischemia. Eur Heart J 2004;25(17):1502 8. in diabetes 2013. Diabetes Care 2013;36(Suppl 1):S11 66.
[63] Clarke B, Ewing D, Campbell I. Diabetic autonomic [70] Umetani K, Singer DH, McCraty R, Atkinson M.
neuropathy. Diabetologia 1979;17:195 212. Twenty-four hour time domain heart rate variability
[64] Litchman JH, Bigger Jr JT, Blumenthal JA, Frasure-Smith and heart rate: relations to age and gender over nine
N, Kaufmann PG, Lespérance F, et al. Depression and cor- decades. J Am Coll Cardiol 1998;31(3):593 601.
onary heart disease recommendations for screening, refer- [71] Bloomfield DM, Kaufman ES, Bigger Jr JT, Fleiss J,
ral, and treatment: a science advisory from the American Rolnitzky L, Steinman R. Passive head-up tilt and
Heart Association Prevention Committee of the Council actively standing up produce similar overall changes
on Cardiovascular Nursing, Council on Clinical in autonomic balance. Am Heart J 1997;134(2 Pt
Cardiology, Council on Epidemiology and Prevention, 1):316 20.
and Interdisciplinary Council on Quality of Care and [72] Post RE, Dickerson LM. Dizziness: a diagnostic
Outcomes Research: Endorsed by the American approach. Am. Fam. Physician 2010;82(4):361 8 369.
Psychiatric Association. Circulation 2008;118:1768 75. [73] Low PA, editor. Clinical autonomic disorders: evalua-
[65] Malik Mthe Task Force of the European Society of tion and management. 2nd ed. Philadelphia, PA:
Cardiology and the North American Society of Pacing Lippincott-Raven; 1997.

Diabetic Neuropathy
 C H A P T E R

11
Diabetic neuropathy in children and youth
Gulcin Akinci, Masha G. Savelieff, Gary Gallagher, Brian C. Callaghan
and Eva L. Feldman
Department of Neurology, University of Michigan, Ann Arbor, MI, United States

Background to symptoms (e.g., pain, numbness, paresthesia)

that affect quality of life, and, in severe cases,
Diabetes is a major health concern in children contributes to fall risk, foot ulcerations, and ampu-
and youth, and has increased in incidence and tation. Diabetic foot ulcers are associated with a
prevalence in recent years [1]. Although type 1 high risk for long-term mortality. DN often devel-
diabetes (T1D) accounts for greater than 80% of ops years after a T1D diagnosis, but may be
the cases of diabetes in youth less than 20 years already present at the time of diagnosis of T2D.
old, recent large population-based studies reveal The SEARCH cohort showed that the rate of com-
the incidence of type 2 diabetes (T2D) plications of diabetes, including DN, is nearly three
is rapidly increasing [2]. The SEARCH for times higher in T2D compared to T1D youth [6].
Diabetes in Youth study highlighted an increased Although T2D in adults and youths share similar
prevalence of both T1D and T2D among children pathophysiology, youth T2D has unique features
and youth over a 7-year period; however, the characterized by rapid loss of beta-cell function
rate of increase was particularly striking for T2D. and early development of diabetic complications
The rise in T2D incidence based on SEARCH [2]. Longer lifetime exposure to hyperglycemia is a
data suggests that the number of youth with T2D major risk factor for developing DN; however,
may quadruple over the next several decades [3]. additional factors (e.g., genetic markers, obesity,
Diabetic complications emerge with disease lon- dyslipidemia, hypertension, inflammation) also
gevity, therefore the increasing diabetes rates in contribute to the risk [6 8].
children and youth is alarming since many risk DN is infrequently reported in pediatric
developing complications of diabetes in early practice, presumably due to insensitive initial
adulthood [4]. screening measures and subclinical presenta-
Diabetic neuropathy (DN) is one of the most tions in most cases [9,10]. Recent studies
prevalent diabetic complications. Up to 50% of showed that abnormal nerve function or other
adults with diabetes will eventually develop DN, DN signs can be detected in many children
defined as a symmetric, distal to proximal loss of and youth with diabetes, even in the absence
sensory greater than motor function [5]. DN leads of typical DN symptoms. It is likely that DN

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00002-5 185 © 2022 Elsevier Inc. All rights reserved.
186 11. Diabetic neuropathy in children and youth

signs are missed in routine practice, and that the of neuropathy in these patients [14]. Thus a
number of undiagnosed cases is likely higher detailed neurological examination of children
than currently appreciated. A careful clinical with diabetes is essential for discerning DN signs
approach requires a detailed neurological evalu- using sensitive techniques that detect nerve dys-
ation and the use of sensitive techniques to function [7].
detect DN in youths before reaching adulthood. In 2017 the ADA issued a position statement
More importantly, risk of progressive neuropa- which classified neuropathies that occur with dia-
thy can potentially be reversed by interventions betes into four main entities: (1) distal symmetric
such as intensive glycemic control and lifestyle polyneuropathy, (2) autonomic neuropathy, (3)
changes. Beyond hyperglycemia, emerging data one or more mononeuropathies (median, pero-
advocate components of the metabolic syndrome neal, ulnar, cranial nerve III), and (4) radiculopa-
(MetS) as additional potential therapeutic targets thy and polyradiculopathy (radiculoplexus
to prevent or improve DN. Obesity is a common neuropathy, thoracic radiculopathy) [14]. Distal
problem in the T2D population, and its preva- symmetric polyneuropathy is the most common
lence is also on the rise in T1D children and type of neuropathy seen with diabetes. In this
youth [11]. Identifying high-risk subjects would chapter, we will also refer to distal symmetric
help develop preventive strategies that can halt polyneuropathy as DN. DN is characterized by
or delay the development of DN and highly initial distally predominant, length-dependent
morbid conditions, such as foot ulcers and limb impairment and loss of sensory neuron function;
amputations. the later motor involvement is rarely seen in
youth. If a youth presents with acute or subacute
onset, proximal rather than distal involvement,
Definitions and classifications asymmetrical or nonlength-dependent neuropa-
thy, the diagnosis is not DN [14].
DN is described as the presence of symptoms Diabetic autonomic neuropathy (DAN) is a
and/or signs of neuropathy, either clinically evi- disorder of the autonomic nervous system. DAN
dent or subclinical, that occurs in people with dia- includes cardiac autonomic neuropathy (CAN),
betes in whom other neuropathy causes are diabetic gastroparesis, hypohidrosis/anhidrosis,
excluded. Excluding other causes of neuropathy is gustatory sweating, hypoglycemic unawareness,
crucial since some are treatable or reversible [12]. and urinary and sexual dysfunction. CAN has
B12 deficiency should be considered in youth on a been studied in children and youth, and will be
strict vegetarian diet or with pernicious anemia or discussed later in this chapter.
metformin use. A detailed family history, includ-
ing determining if family members have hammer-
toes or high-arched feet, is essential when Epidemiology
hereditary peripheral nerve disorders of childhood
are suspected. T1D youth are at increased risk for There is large uncertainty on DN prevalence
concomitant autoimmune neurological conditions, and incidence in childhood and youth due to
which can mimic DN symptoms [13]. limited population-based studies in the field,
An important DN characteristic in children wide range of clinical presentations, dissimilar
and youth is that it may be asymptomatic or methodologies used by different studies, varia-
only mildly symptomatic in many patients. The tion in study population selection, and a lack
current American Diabetes Association (ADA) of well-defined cohorts representative of the
position statement highlights the importance of general target population, with the exception
early recognition and appropriate management of a few recent studies [6,7,15].

Diabetic Neuropathy
 Epidemiology 187
The prevalence of DN in T1D youth varies subjects from the EURODIAB study, aged
widely, depending on the study population 15 29, exhibited a DN prevalence of 19%,
and methodology [16 18]. One of the earliest when diagnosing DN based on symptoms,
studies, the Pittsburgh Epidemiology of reflex loss, and abnormalities in vibration
Diabetes Complications (EDC) trial, reported perception threshold (VPT), and autonomic
DN in only 3% of T1D patients #18 years old, dysfunction. VPT abnormalities were reported
defining DN as the presence of at least two of in 62% of 339 T1D subjects aged 12 27 years
three abnormal signs, symptoms, and decreased in a nationwide, multicenter, cross-sectional
tendon reflexes [19]. DN prevalence was 2 Danish study [16]. A summary of major studies
8% overall in the multicenter EURODIAB reporting DN prevalence in T1D children and
IDDM Complications Study [20]. A subgroup of youth is presented in Table 11.1.

TABLE 11.1 Study populations of 100 or more examining the prevalence of DN in T1D and T2D youth.
Methods to
References Study population diagnose DN Prevalence/incidence Notes

Käär et al. [21] Finland, clinic-based 161 NCS (median and 30% peroneal and 5% NCS considered abnormal if
Cross-sectional T1D, .5 years peroneal) median nerve abnormality , 2 2SD

Maser et al. [19] United States, clinic- Clinical history and ,18 years: 3%18 29 DN diagnosed if 2 of 3
Pittsburgh EDC based 400 T1D, neurological exam years: 18% $ 30 years: 58% (symptoms, signs, reflex
Cross-sectional diagnosed , 17 years abnormalities) present

Hyllienmark et al. Sweden, clinic-based 75 NCS (median, 57% had abnormal NCS NCS considered abnormal if
[10] Cross- T1D, aged 7 20 years, peroneal, sural) , 2 2 SD
sectional duration of diabetes .3
years

Bognetti et al. [18] Italy, clinic-based 317 NCS (sural, median, 18.5% had alterations on Outcomes reported based on
Prospective T1D ulnar, peroneal), NCSAfter 5 7 years of last visit (cross-sectional);
observational performed in 292 diabetes in 17% of the NCS abnormal if , 6 2 SD
patients patients

Solders et al. [22] Sweden 144 T1D, newly NCS (median, 25% abnormal sural NCV Normal lower limit: mean
Prospective diagnosed, aged 4 16 peroneal, sural) at first visit; 36.2% 22.5 SD
longitudinal years abnormal peroneal NCV
after 10 years

Olsen et al. [16] Denmark, multicenter VPT 62.5% abnormal VPT abnormal if .6.5 V
Cross-sectional 339 T1D VPT10 15 years: 5%
15 20 years: 46.8% . 20
years: 76.6%

Nelson et al. [9] Canada 73 T1D, duration Neurological exam, Abnormal exam: 36% VPT abnormal if amplitude
Cross-sectional of diabetes $ 5 years NCS (median, Reduced NCV: 57% $ 0.5 microns; TPT abnormal
peroneal, sural) Abnormal VPT: 51% if the patient could not sense
VPT, TPT Abnormal TPT: 26% 1 g filament

Nordwall et al. Sweden, clinic-based 80 Clinical history and 39.2% at first examination DN diagnosed if $ 2 of 4
[17] Retrospective T1D, aged 7 22 years, neurological exam, 59% after 13 years of NCS measurements
cohort duration of diabetes NCS (sural and diabetes abnormal
.3 years peroneal)

(Continued)

Diabetic Neuropathy
188 11. Diabetic neuropathy in children and youth

TABLE 11.1 (Continued)

Methods to
References Study population diagnose DN Prevalence/incidence Notes

Dart et al. [23] Canada, clinic-based Health insurance T1D: 5%T2D: 7.6% Shorter Neuropathy prevalence
Clinical registry 1011 T1D, aged 1 18 records (ICD codes complication free survival lower than most other studies
years 342 T2D, aged to detect clinically in T2D likely due to methodology
1 18 years diagnosed (the study relies on ICD
neuropathy) coding)

Donaghue et al. Australia, clinic-based VPT, thermal 28.4% $ 1 abnormal test

[24] Prospective 102 T1D, aged 11 19 threshold for heat
longitudinal years, duration of and cold
diabetes .3 years

Young et al. [25] United Kingdom, clinic- Neurological exam, 32% at first examination Three standardized variables
Prospective based 75 T1D NCS (peroneal, 41% after 2.5 years summed to give a single
longitudinal sural) electrophysiological variable

Karanavaki et al. United Kingdom, VPT, limited joint 6.2% decreased VPT 6.2%
[26] Prospective population-based 129 mobility limited joint mobility
longitudinal T1D

Mohsin et al. [27] Australia, clinic-based VPT, thermal 12% $ 1 abnormal test 1990 94 (T1, n 5 257)
Cross-sectional 878 T1D, aged 12 20 threshold result T1: 12% T2: 19% T3: 1995 98 (T2, n 5 284)
years, duration of 24% 1999 2002 (T3, n 5 337)
diabetes .3 years Closely comparable groups

Eppens et al. [28] Australia, clinic-based VPT, thermal T1D: 27%, T2D: 21%,
Longitudinal 1433 T1D, ,18 years 68 threshold
T2D, ,18 years

Cho et al. [29] Australia, clinic-based VPT, thermal 1990 94: 14% 1995 98:
Observational 819 T1D, aged 11 17 threshold 19% 1999 2002: 28%
years with 2- to 5-years 2003 06: 23%
diabetes duration

Jaiswal et al. [7] Five US centers 1734 MNSI exam T1D: 7% T2D: 22% DN diagnosed if MNSI exam
SEARCH T1D, duration of diabetes Diabetes for 5 10 years score .2
Longitudinal $ 5 years; 258 T2D, versus .10 years T1D 5%
population-based duration of diabetes $ 5 versus 13% T2D 19%
observational years versus 36%

Amutha et al. [30] India, clinic-based 3252 VPT T1D: 8.8 /1000 person VPT abnormal if $ 20 V DN
Retrospective T1D, ,20 years 899 T2D, years T2D: 24/1000 assessed in a small subgroup
,20 years person years of subjects

TODAY2 [2] 517 T2D Adjudicated DN in 28% 33% of Presented at ADA 2019
Preliminary neuropathy events subjects by year 12 meeting
analysis

ADA, American Diabetes Association; DN, diabetic peripheral neuropathy; EDC, Epidemiology of Diabetes Complications; ICD, international
classification of diseases; MNSI, Michigan Neuropathy Screening Instrument; NCS, nerve conduction study; NCV, nerve conduction velocity; SD,
standard deviation; T, time period; TPT, tactile perception threshold; T1D, type 1 diabetes; T2D, type 2 diabetes; VPT, vibration perception
threshold.

Diabetic Neuropathy
 Clinical presentations 189
Early reports of T2D youth with DN included Large fiber damage produces tingling, loss of pro-
few subjects. However, systematic investigation tective sensation, and progressive numbness.
by the SEARCH study, the first US population- DN symptoms of small and large fibers are fre-
based study of an ethnically diverse diabetes quently mild in children and youth. Commonly,
youth cohort, reported a DN prevalence of 7% in the neurologic examination and nerve conduction
T1D and 22% in T2D subjects with similar diabe- studies (NCS), which measure only large fiber
tes duration, using the Michigan Neuropathy function, may uncover only subtle neuropathy
Screening Instrument (MNSI) [7]. Similarly, pre- signs in this age group [35]. Other symptoms of
liminary analysis of the TODAY2 study in youth late-stage DN include distal motor nerve involve-
revealed high rates of DN in 28% 33% of T2D ment causing toe extension weakness, imbalance
subjects after 12 years [2]. The list of major stud- and gait problems, and altered proprioception;
ies on DN prevalence in T2D is presented in however, late DN symptoms are rare in chil-
Table 11.1. dren and youth. A differential diagnosis is
CAN is an independent predictor of cardio- indicated if muscle weakness develops early,
vascular mortality in people with diabetes. The for example, Charcot Marie Tooth disease
SEARCH study found prevalence rates of 12% or treatable neuropathies [12].
and 17% in T1D and T2D participants, respec-
tively [31]. In the TODAY study, the preva-
lence of CAN was 8% in a T2D cohort slightly
younger and with lower Hemoglobin A1c
Diabetic autonomic neuropathy
(HbA1c) than SEARCH [15]. Table 11.2 sum- DAN is a serious diabetes complication
marizes major studies reporting DAN/CAN in associated with an increased morbidity due to
T1D and T2D children and youth. organ dysfunction, such as orthostatic hypo-
tension, gastroparesis, hypoglycemia, and high
risk of death, primarily from cardiac events.
Although overt DAN is rare in diabetic chil-
Clinical presentations dren and youth, signs of autonomic dysfunc-
tion can be detected early in the disease course
Diabetic neuropathy [24,36,37]. DAN can affect both sympathetic
and parasympathetic nervous systems; thus
DN is a sensory greater than motor neuropathy clinical presentation is heterogeneous and
[5,14]. Nerve damage usually begins in small results in a variety of symptoms.
fibers, defined as predominantly small diameter In early stages, CAN, the best studied DAN
somatic and autonomic nerves, unmyelinated C in children and youth, is usually asymptom-
fibers, and/or thinly myelinated A-delta sensory atic, but may still be detected by standard tests,
fibers. Small fibers carry afferent impulses that such as evaluation of heart rate variability
mediate pain, temperature, and unpleasant (HRV) [31,34,38]. As the disease progresses,
stimuli, which protect the individual from harm CAN causes symptoms ranging from ortho-
and also innervate the autonomic nervous system. static tachycardia syndrome and lightheaded-
Classically, small fiber damage causes the earliest ness caused by postural hypotension to silent
DN symptoms, including both spontaneous and myocardial infarctions and sudden cardiac
stimulus-evoked distal extremity pain. As DN death. Typical CAN symptoms are challenging
progresses, large nerve fibers are affected. Large to detect in children and youth, that may suffer
fibers are myelinated axons; these axons have dif- from nonspecific symptoms, such as general
ferent degrees of myelination and axon diameter. weakness, dizziness, and fatigue. In advanced

Diabetic Neuropathy
190 11. Diabetic neuropathy in children and youth

TABLE 11.2 Major studies reporting AN/CAN in T1D and T2D children and youth.
Methods to diagnose AN/
References Study population CAN Prevalence/incidence Notes

Young et al. [25] United Kingdom, clinic- HR response to DB and 19% at first examination
Prospective based 75 T1D Valsalva; BP change in 28% after 2.5 years
longitudinal response to standing and
handgrip

Ringel et al. [32] United States, clinic- HR response to DB and Abnormal response to Also, follow-up data
Cross-sectional based 248 T1D standing either HR-DB or standing: reported from 75 patients;
29%; both test abnormal: no change in 1 year follow-
3% up

Solders et al. [22] Sweden 144 T1D, newly HR response to DB 25% at first visit 56%
Prospective diagnosed, aged 4 16 abnormal HRV after 10
longitudinal years years

Verrotti et al. [33] Italy, clinic-based110 HR response to DB, At least one abnormal 43%
Cross-sectional T1D Valsalva, and standing;
and BP response to
standing and handgrip

Donaghue et al. Australia, clinic-based HR response to DB, 29.5% $ 1 abnormal test

[24] Prospective 102 T1D, aged 11 19 Valsalva, and standing;
longitudinal years, duration of and BP response to
diabetes .3 years standing

Massin et al. [34] Belgium, clinic-based HRV with 24-h Holter All indices within the Two patients excluded (73
Cross-sectional 75 T1D monitoring (five time normal range # 10 years subjects analyzed)
domain and three 54% $ 1 abnormal test
frequency domain result, $ 11 years
measures)

Karanavaki et al. United Kingdom, HRV response to DB, 15.5% 1 abnormal HRV
[26] Prospective population-based 129 Valsalva and standing; test 7.7% $ 2 abnormal
longitudinal T1D pupillometry tests 7.9% abnormal
pupillometry

Mohsin et al. [27] Australia, clinic-based HRV response to DB, $ 1 abnormal test resul 1990 94 (T1, n 5 257)
Cross-sectional 878 T1D, aged 12 20 standing and postural BP tT1: 18% T2: 21% T3: 18% 1995 98 (T2, n 5 284)
years, duration of 1999 2002 (T3, n 5 337)
diabetes .3 years Closely comparable groups

Eppens et al. [28] Australia, clinic-based Pupillometry T1D: 61% T2D: 57%
Longitudinal 1433 T1D, ,18 years 68
study T2D ,18 years

Jaiswal et al. [31] Five US centers HRV using five domains T1D: 12% T2D: 17% Dysfunction defined as the
Longitudinal (SEARCH) 1646 T1D presence of $ 3 of 5
population-based 252 T2D abnormal HRV indices
observational

Shah et al. [15] 397 T2D (TODAY HRV using five domains, 8.1% versus obese controls Dysfunction defined as the
Data from a cohort) arterial stiffness 12.1% versus lean controls presence of $ 3 of 5
randomized Greater pulse-wave abnormal HRV indices
clinical trial velocity in T2D

AN, autonomic neuropathy; BP, blood pressure; CAN, cardiac autonomic neuropathy; DB, deep breathing; HR, heart rate; HRV, heart rate variability;
T, time period; T1D, type 1 diabetes; T2D, type 2 diabetes.

Diabetic Neuropathy
 Risk factors 191
cases, CAN may be complicated by hypoglyce- axis [63]. Thus it is difficult to examine the iso-
mia awareness, a dangerous condition that lated effect of puberty on DN outcomes, since it
poses a significant risk of severe hypoglycemia is confounded with other factors, such as height,
[39,40]. CAN is also associated with arterial age, and diabetes duration.
stiffness in both T1D and T2D children and Despite these limitations, several studies
youth, a well-documented risk factor that pre- suggested a link between puberty and DN [64].
dicts future cardiovascular events [15,41]. The rarity of diabetic complications in prepu-
Gastrointestinal presentations of DAN bertal children suggests a potential role of
include gastroparesis, esophageal and/or gas- puberty in DN development [65], although
trointestinal dysmotility, but these symptoms some studies have not supported such a rela-
are rare in youth [42]. Sudomotor dysfunction tionship [66,67]. Sosenko et al. reported a sig-
occurs as a result of sympathetic denervation nificant relationship between HbA1c and DN
and can develop distally early in the course of in postpubertal, but not in younger prepuber-
DN. Clinically, sudomotor dysfunction is asso- tal, children [55]. Barkai et al. proposed the late
ciated with impaired sweating that can lead to stages of puberty as an independent DN risk
dry skin, a predisposing factor to skin cracks factor [56]. Riihimaa et al. reported a progres-
and foot ulceration. Again, this symptom is sive decrease in several measures of motor and
rarely reported in youth. sensory function in late-pubertal and postpu-
bertal T1D adolescents [58]. Massin et al.
showed that depressed HRV correlated with
metabolic control in T1D children aged $11 years,
Risk factors
but not in younger children [34], suggesting
the critical role of early puberty for developing
The increasing prevalence of diabetes in
CAN. Additional studies are needed to clarify
children and youth necessitates urgent devel-
the role of puberty in relation to the risk of
opment of early screening strategies and
developing DN.
improved risk factor modification to prevent
DN onset and/or progression. Beyond hyper-
glycemia, recent studies suggest a wide range Sex
of emerging risk factors such as obesity, MetS,
Earlier development of DN in males versus
and genetic markers [43], listed in Table 11.3
females was previously reported in a study of
along with major studies that reported them. A
T2D adults. The TODAY study recently
detailed review of DN pathogenesis can be
reported a significant association between male
found in two of our recent papers [5,43].
sex in youth and elevated low-to-high fre-
quency ratio or sympathetic overdrive, sup-
porting a higher CAN risk in males [15]. The
Role of puberty and height SEARCH study also found a higher DN preva-
As children enter puberty, hormonal changes lence in T2D males compared to females [7].
occur that contribute to physiological changes in The mechanism for sex dimorphism in DN
insulin dynamics, characterized by reduced insu- development remains unclear.
lin sensitivity, leading to impaired glycemic con-
trol, especially in those with limited beta-cell
Race and ethnicity
function [62]. Also, puberty is associated with a
growth spurt and hormonal changes in the The prevalence of DN in adults is dispropor-
growth hormone/insulin-like growth factor-I tionally high across some varying racial and

Diabetic Neuropathy
192 11. Diabetic neuropathy in children and youth

TABLE 11.3 Major studies proposing DN risk factors in T1D and T2D childhood and youth.
Risk factor Population References

Glycemic control T1D Käär et al. [21], Maser et al. [19], Hyllienmark et al. [10], Tesfaye et al. [20],
Bognetti et al. [18], Allen et al. [44], Solders et al. [22], Jaiswal et al. [7,31],
Young et al. [25], Massin et al. [34], Cho et al. [45,46]
T2D Dart et al. [23], Shah et al. [15]
Disease duration T1D Käär et al. [21], Maser et al. [19], Tesfaye et al. [20], Eng et al. [47], Bognetti
et al. [18], Solders et al. [22], Jaiswal et al. [7], Massin et al. [34]
T2D Jaiswal et al. [7], Shah et al. [15]
Older age at DN screening T1D Tesfaye et al. [20], Olsen et al. [16,48], Jaiswal et al. [7], Massin et al. [34]

T2D Jaiswal et al. [7], Shah et al. [15]

Presence of other T1D Maser et al. [19], Bjerg et al. [49], Tesfaye et al. [20], Olsen et al. [16], Young
microvascular complications et al. [25], Massin et al. [34], Maguire et al. [50], Karavanaki et al. [36]
T2D Jaiswal et al. [31]
Obesity/BMI T1D Jaiswal et al. [7], Cho et al. [46], Nagai et al. [51]
T2D/IFG/IR Dart et al. [23], Shah et al. [15], Ince et al. [52], Akin et al. [53]
Dyslipidemia T1D Maser et al. [19], Tesfaye et al. [20], Jaiswal et al. [7,31]
T2D Jaiswal et al. [7,31]

Blood pressure T1D Maser et al. [19], Tesfaye et al. [20], Olsen et al. [16], Jaiswal et al. [7,31]
Genetic markers T1D Thamotharampillai et al. [54]
T2D Dart et al. [23]
Ethnicity T2D Jaiswal et al. [31], Shah et al. [15]
Sex T1D Mohsin et al. [27]
T2D Jaiswal et al. [7], Shah et al. [15]

Smoking T1D Maser et al. [19], Tesfaye et al. [20], Jaiswal et al. [7]
T2D Jaiswal et al. [7], Shah et al. [15]
Role of puberty T1D Sosenko et al. [55], Barkai et al. [56], Hoffman et al. [57], Olsen et al. [48],
Massin et al. [34], Riihimaa et al. [58]
Height T1D Duck et al. [59], Hyllienmark et al. [10], Olsen et al. [16,48], Tesfaye et al.
[20], Mohsin et al. [27]
Eating disorders T1D Steel et al. [60], Scheuing et al. [61]

BMI, Body mass index; DN, diabetic neuropathy; IFG, impaired fasting glucose; IR, insulin resistance; T1D, type 1 diabetes; T2D, type 2
diabetes.

ethnic groups with T2D [5]. The impact of race was more prevalent among non-Hispanic Whites
and ethnicity on DN has not been systematically and Hispanics than non-Hispanic Blacks and
studied in children and youth. However, CAN other minority groups, including Asian/Pacific

Diabetic Neuropathy
 Risk factors 193
Islanders, in T2D youths from the SEARCH study Obesity and other metabolic syndrome
[31]. Similarly, non-Hispanic Black race-ethnicity components
was found to be associated with less DAN in the
TODAY study cohort of T2D youths [15]. It is MetS is a constellation of metabolic abnormal-
possible, but unproven, that racial or ethnic popu- ities frequently associated with insulin resistance.
lations share common genetic material or environ- The clinical definition of MetS is more controver-
mental exposures that contribute to DN risk. sial in children and youth than adults, and
Additionally, there are racial, ethnic, and systemic requires abdominal obesity plus at least two
disparities that affect access to appropriate health- additional risk factors. Abdominal obesity can be
care, including diet and exercise regimens, that defined by an increased waist circumference in
could prevent or delay the development of DN. children and youth by age group, either older
than 10 (aged 10 16 years, $ 90th percentile) or
aged .16 years (by Adult Treatment Panel III
criteria [71], which can also be used in adults).
Glycemic control The definition of most metabolic abnormalities
Long-term impairment of glycemic control is requires age-specific thresholds in childhood, but
the strongest predictor of DN development in a simple algorithm to diagnose these abnormali-
T1D, as found by several clinical studies, that ties associated with MetS has been proposed as
enrolled younger individuals. The Diabetes follows: triglycerides $150 mg/dL for both age
Control and Complications Study is a landmark groups; high-density lipoprotein cholesterol
study, which demonstrated the benefit of inten- (HDL-C) ,40 mg/dL for ages 10 16 years;
sive glycemic control for improving DN outcomes HDL-C ,40 mg/dL for males aged .16 years,
in T1D subjects. The association of DN with and ,50 mg/dL females aged .16 years; ele-
HbA1c was also shown in subjects with vated fasting glucose ($100 mg/dL or known
childhood-onset T1D from the Pittsburgh EDC T2D), and hypertension (systolic $ 130 or dia-
study [19]. Similarly, glycemic control was linked stolic $ 85 mmHg).
to DN risk in a subset of T1D youth in the A few small scale studies have shown the
EURODIAB study, as well as diabetes duration, association of obesity with DN in children with
and deterioration of glycemic control during the impaired fasting glucose or insulin resistance,
follow-up period led to further risk [68]. Further without clinical diabetes [52,53].
studies, including the recent SEARCH study, con- Higher body mass index (BMI), dyslipide-
firmed the significant correlation between DN mia, and hypertension have been identified as
and poor glycemic control in youth with T1D DN risk factors in the T1D youth EURODIAB
[7,25,44,69,70]. Concerning T2D, the SEARCH cohort [68]. Diastolic blood pressure (BP) sig-
study found a significant relationship between nificantly correlated with DN in T1D children
DN and diabetes duration in T2D children and and adolescents in a Danish population-based
adolescents [7], as also noted in an Asian Indian longitudinal study [16]. In addition to glycemic
study [30]. Finally, poorly controlled, long-term control over time, increased diastolic BP, obe-
diabetes is a significant risk factor for developing sity, increased LDL cholesterol, triglycerides,
CAN in children and youth with T1D [31,45,46] and low HDL cholesterol were identified as
and T2D. The TODAY study showed an inde- DN risk factors in T1D youth in the SEARCH
pendent effect of HbA1c on reduced HRV, para- study [7].
sympathetic loss, and sympathetic overdrive in a There are two large studies regarding the role
large cohort of adolescents and young adults of obesity and other MetS components for devel-
with youth-onset T2D [15]. oping DN in youth with T2D. Preliminary

Diabetic Neuropathy
194 11. Diabetic neuropathy in children and youth

analysis of TODAY2, a long-term follow-up of factor for CAN in the TODAY cohort, which
the TODAY cohort that included T2D youth with includes adolescents and young adults with
BMI $ 85th percentile at diagnosis [72,73], youth-onset T2D [15].
showed evidence of DN in 28% 33% of partici-
pants by year 12 [2]. The SEARCH study also
found a significant correlation between low HDL Eating disorders
cholesterol and higher DN risk in T2D youths,
Eating disorders (EDs) and disordered eat-
independent of glucose regulation [7].
ing behavior (DEB) are common clinical pro-
The impact of obesity on CAN is well docu-
blems in adolescents with T1D and T2D. DEB
mented in adult diabetes studies. A similar
was reported in 21.2% of T1D and 50.3% of
association has been examined in several stud-
T2D in the SEARCH study [78]. DEB was asso-
ies of obese children and youth, showing auto-
ciated with higher BMI, lower insulin sensitiv-
nomic imbalance and lower HRV in obese
ity, more depression symptoms, and reduced
participants [74,75]. Higher BMI and central
quality of life in both T1D and T2D. Similarly,
adiposity predicted CAN in children and
6% of T2D youths had clinical, and 20% had
youth, and young T1D adults [46]. Besides gly-
subclinical binge eating in the TODAY study,
cemic control, CAN correlated with high BP
which was associated with higher rates of
and elevated triglycerides in T1D subjects and
extreme obesity, depression symptoms, and
high triglycerides in T2D subjects from the
poor quality of life [79].
SEARCH cohort [31]. Recently, the TODAY
EDs are usually associated with poor glyce-
study showed that BMI was associated with
mic control [80], and increased risk for acute
reduced HRV, parasympathetic loss, and sym-
and chronic diabetes complications [81]. A pro-
pathetic overdrive [15].
spective multicenter German/Austrian study
showed the association of ED with poor meta-
bolic control and higher rates of diabetes com-
Smoking plications in a cohort of 52,215 T1D patients
aged 8 30 years [61]. Besides poor metabolic
It is unclear whether cigarette smoking is an
control, DEB has also been linked to sedentary
independent risk factor for developing DN;
lifestyle, low socioeconomic status, and skip-
however, previous data do demonstrate raised
ping insulin injections [82].
DN risk among smokers with diabetes. A
metaanalysis of prospective studies comparing
ever-smokers with never-smokers reported an
association between smoking and DN preva- Screening and diagnosis
lence and incidence in adult populations [76].
The SEARCH study identified smoking as a Diagnosing DN and CAN in children
DN risk factor in both T1D and T2D children requires a high index of suspicion and close
and youth [7]. It also reported relatively high neurologic examination. An early DN diagno-
rates of tobacco use among youths with diabe- sis is important since neuropathy may be stabi-
tes (for T1D: 10 14 years, 2.7%, and 15 19 lized or improved with early intervention [5].
years, 17.1%; for T2D: 10 14 years, 5.5%, and Current ADA recommendations include rou-
15 19 years, 16.4%) [77]. Cigarette smoking tine screening for DN in children and youth
was associated with higher odds of elevated with diabetes. An annual comprehensive foot
triglyceride levels and physical inactivity in exam is recommended in T1D youth beginning
T1D youths. Smoking also emerged as a risk 5 years after diagnosis, at the start of puberty,

Diabetic Neuropathy
 Managing diabetic neuropathy 195
or at age $10 years, whichever comes first. issue in T1D youth that raises the risk of DN as
CAN screening is not recommended, but well as other diabetes complications [11]. Thus
should be performed if there is any clinical it is reasonable to provide both T1D and T2D
suspicion. youths and their guardians with counseling on
nutrition and exercise and encourage weight
loss for overweight and obese diabetic children
Diabetic neuropathy, quality of life, and and youth.
mortality

Healthcare-related quality of life is impaired Painful neuropathy

in children with diabetes and neuropathy, Painful DN is a rare clinical presentation in
compared to their peers [83]. Unrecognized or children and youth, despite its high frequency
undertreated neuropathic pain impacts dia- in adult DN patients. Experience in children is
betic children’s quality of life. Youth with dia- limited, but calcium channel a2δ ligands (gaba-
betes are at higher risk of depression and pentin, pregabalin), serotonin norepinephrine
anxiety than children without diabetes. reuptake inhibitors, and tricyclic antidepres-
Deaths related to diabetes arise from hypogly- sants can be used in youth with painful neu-
cemia, diabetic ketoacidosis, and CAN occur pri- ropathy, as they are used in adults [85]. These
marily in T1D youth. Numerous studies have treatments are used in children for neuropathic
demonstrated an increased risk of early mortality pain from many etiologies, but clinical evi-
in T1D children around the world, ranging from dence is lacking to assess the efficacy and
slight increase up to 8 3 risk of mortality [84]. safety of these drugs in pediatric patients with
Children from disadvantaged socioeconomic back- painful DN. Of note, there are no Food and
grounds and children that face racial/ethnic dispa- Drug Administration-approved neuropathic
rities in healthcare access are at an even higher pain medications for painful DN in children.
risk for DN, CAN, and impaired quality of life.

Diabetic autonomic neuropathy

Managing diabetic neuropathy CAN patients have an increased cardiovascular
risk and thus must be carefully managed. Special
The management of DN focuses on the man- considerations may be required for CAN patients
agement of diabetes itself, coupled with modifi- embarking on exercise programs, or those under-
able risk factors and symptomatic improvement. going surgery due to higher arrhythmia risk and
Management in youth parallels that of adults, hemodynamic instability. Orthostatic hypotension
except in T1D, where separate glycemic criteria is rare in children and youth, but may require
are commonly used. special attention if it develops [14]. Maintaining
optimal hydration is crucial in these patients.
Patient should avoid rapid positional changes
Emerging risk factors causing orthostatic symptoms. Differential diag-
The effect of obesity and related risk factors nosis of endocrine abnormalities, such as thyroid
is reported in several pediatric studies that dysfunction and adrenal insufficiency, should be
enrolled participants with diabetes and predia- ruled out, especially in T1D patients. Refractory
betes [7,31,46,52,53]. Although this effect is less cases can be treated with sympathomimetics
clear in T1D, excess weight is an emerging (e.g., midodrine) or fludrocortisone. Gastroparesis

Diabetic Neuropathy
196 11. Diabetic neuropathy in children and youth

is also rare in pediatric clinical practice. Treatment References

should start by adapting dietary changes, such as
[1] Mayer-Davis EJ, Lawrence JM, Dabelea D, Divers J,
implementing multiple small meals and decreas- Isom S, Dolan L, et al. Incidence trends of type 1 and
ing dietary fat intake. type 2 diabetes among youths, 2002 2012. N Engl J
Med 2017;376(15):1419 29.
[2] Zeitler P. Progress in understanding youth-onset type
2 diabetes in the United States: recent lessons from
clinical trials. World J Pediatr 2019;15(4):315 21.
Conclusion [3] Jensen ET, Dabelea D. Type 2 diabetes in youth: new
lessons from the SEARCH study. Curr Diab Rep
The prevalence of both T1D and T2D is ris- 2018;18(6):36.
[4] Graves LE, Donaghue KC. Management of diabetes
ing in children and youth, placing them at complications in youth. Ther Adv Endocrinol Metab
significant risk of developing DN. Initial mani- 2019;10 2042018819863226.
festations of childhood DN are commonly [5] Feldman EL, Callaghan BC, Pop-Busui R, Zochodne
asymptomatic or mild, and may only be DW, Wright DE, Bennett DL, et al. Diabetic neuropa-
detected through sensitive screening and neu- thy. Nat Rev Dis Primers 2019;5(1):41.
[6] Dabelea D, Stafford JM, Mayer-Davis EJ, D’Agostino Jr.
rologic exam. The most common presentation R, Dolan L, Imperatore G, et al. Association of type 1
involves small fiber greater than large fiber diabetes vs type 2 diabetes diagnosed during childhood
sensory involvement in a distal and symmetric and adolescence with complications during teenage
fashion. Poorly controlled diabetes increases years and young adulthood. JAMA 2017;317(8):825 35.
the risk of DN. In parallel, emerging risk fac- [7] Jaiswal M, Divers J, Dabelea D, Isom S, Bell RA,
Martin CL, et al. Prevalence of and risk factors for dia-
tors for developing DN also involve MetS betic peripheral neuropathy in youth with type 1 and
components, especially obesity, though the precise type 2 diabetes: SEARCH for Diabetes in Youth Study.
pathophysiology remains elusive. Neuropathic Diabetes Care 2017;40(9):1226 32.
pain secondary to DN and CAN may erode [8] Lascar N, Brown J, Pattison H, Barnett AH, Bailey CJ,
quality of life for children and youth, placing Bellary S. Type 2 diabetes in adolescents and young
adults. Lancet Diabetes Endocrinol 2018;6(1):69 80.
them at substantial risk of high morbidity and [9] Nelson D, Mah JK, Adams C, Hui S, Crawford S,
early mortality. This is a mounting epidemic Darwish H, et al. Comparison of conventional and
that requires a concerted and urgent effort to non-invasive techniques for the early identification of
better understand DN pathophysiology to diabetic neuropathy in children and adolescents with
develop interventional strategies and therapies type 1 diabetes. Pediatr Diabetes 2006;7(6):305 10.
[10] Hyllienmark L, Brismar T, Ludvigsson J. Subclinical
that delay or prevent DN development in dia- nerve dysfunction in children and adolescents with
betic and obese youth. IDDM. Diabetologia 1995;38(6):685 92.
[11] DuBose SN, Hermann JM, Tamborlane WV, Beck RW,
Dost A, DiMeglio LA, et al. Obesity in youth with
type 1 diabetes in Germany, Austria, and the United
Acknowledgments States. J Pediatr 2015;167(3):627 32 e1 4.
[12] Callaghan BC, Price RS, Feldman EL. Distal symmetric
Funding support was provided by the National Institutes polyneuropathy: a review. JAMA 2015;314(20):2172 81.
of Health (R24 DK082841 to E.L.F. and R01 DK115687 to [13] Bechtold S, Blaschek A, Raile K, Dost A, Freiberg C,
B.C.C.), the Novo Nordisk Foundation Askenas M, et al. Higher relative risk for multiple sclero-
(NNF14OC0011633 to E.L.F. and B.C.C.), the Sinai sis in a pediatric and adolescent diabetic population: anal-
Medical Staff Foundation Neuroscience Scholars Fund, ysis from DPV database. Diabetes Care 2014;37(1):96 101.
the NeuroNetwork for Emerging Therapies, and the A. [14] Pop-Busui R, Boulton AJ, Feldman EL, Bril V,
Alfred Taubman Medical Research Institute. The authors Freeman R, Malik RA, et al. Diabetic neuropathy: a
would like to thank Dr. Stacey Sakowski Jacoby for position statement by the American Diabetes
expert editorial assistance. Association. Diabetes Care 2017;40(1):136 54.

Diabetic Neuropathy
 References 197
[15] Shah AS, El Ghormli L, Vajravelu ME, Bacha F, Farrell [26] Karavanaki K, Baum JD. Prevalence of microvascular
RM, Gidding SS, et al. Heart rate variability and car- and neurologic abnormalities in a population of diabetic
diac autonomic dysfunction: prevalence, risk factors, children. J Pediatr Endocrinol Metab 1999;12(3):411 22.
and relationship to arterial stiffness in the treatment [27] Mohsin F, Craig ME, Cusumano J, Chan AK, Hing S,
options for Type 2 Diabetes in Adolescents and Youth Lee JW, et al. Discordant trends in microvascular com-
(TODAY) Study. Diabetes Care 2019;42(11):2143 50. plications in adolescents with type 1 diabetes from
[16] Olsen BS, Johannesen J, Sjolie AK, Borch-Johnsen K, 1990 to 2002. Diabetes Care 2005;28(8):1974 80.
Hougarrdss P, Thorsteinsson B, et al. Metabolic con- [28] Eppens MC, Craig ME, Cusumano J, Hing S, Chan
trol and prevalence of microvascular complications in AK, Howard NJ, et al. Prevalence of diabetes compli-
young Danish patients with Type 1 diabetes mellitus. cations in adolescents with type 2 compared with type
Danish Study Group of Diabetes in Childhood. Diabet 1 diabetes. Diabetes Care 2006;29(6):1300 6.
Med 1999;16(1):79 85. [29] Cho YH, Craig ME, Hing S, Gallego PH, Poon M, Chan
[17] Nordwall M, Hyllienmark L, Ludvigsson J. Early dia- A, et al. Microvascular complications assessment in ado-
betic complications in a population of young patients lescents with 2- to 5-yr duration of type 1 diabetes from
with type 1 diabetes mellitus despite intensive treat- 1990 to 2006. Pediatr Diabetes 2011;12(8):682 9.
ment. J Pediatr Endocrinol Metab 2006;19(1):45 54. [30] Amutha A, Ranjit U, Anjana RM, Shanthi RC,
[18] Bognetti E, Calori G, Meschi F, Macellaro P, Bonfanti Rajalakshmi R, Venkatesan U, et al. Clinical profile and
R, Chiumello G. Prevalence and correlations of early incidence of microvascular complications of childhood
microvascular complications in young type I diabetic and adolescent onset type 1 and type 2 diabetes seen at
patients: role of puberty. J Pediatr Endocrinol Metab a tertiary diabetes center in India. Pediatr Diabetes 2020;.
1997;10(6):587 92. [31] Jaiswal M, Divers J, Urbina EM, Dabelea D, Bell RA,
[19] Maser RE, Steenkiste AR, Dorman JS, Nielsen VK, Bass EB, Pettitt DJ, et al. Cardiovascular autonomic neuropathy
Manjoo Q, et al. Epidemiological correlates of diabetic neu- in adolescents and young adults with type 1 and type
ropathy. Report from Pittsburgh Epidemiology of Diabetes 2 diabetes: The SEARCH for Diabetes in Youth Cohort
Complications Study. Diabetes 1989;38(11):1456 61. Study. Pediatr Diabetes 2018;19(4):680 9.
[20] Tesfaye S, Stevens LK, Stephenson JM, Fuller JH, [32] Ringel RE, Chalew SA, Armour KA, McLaughlin J,
Plater M, Ionescu-Tirgoviste C, et al. Prevalence of McCarter Jr. RJ, Kramer WE. Cardiovascular reflex
diabetic peripheral neuropathy and its relation to gly- abnormalities in children and adolescents with diabe-
caemic control and potential risk factors: the tes mellitus. Diabetes Care 1993;16(5):734 41.
EURODIAB IDDM Complications Study. Diabetologia [33] Verrotti A, Chiarelli F, Blasetti A, Morgese G.
1996;39(11):1377 84. Autonomic neuropathy in diabetic children. J Paediatr
[21] Kaar ML, Saukkonen AL, Pitkanen M, Akerblom HK. Child Health 1995;31(6):545 8.
Peripheral neuropathy in diabetic children and adoles- [34] Massin MM, Derkenne B, Tallsund M, Rocour-
cents. A cross-sectional study. Acta Paediatr Scand Brumioul D, Ernould C, Lebrethon MC, et al. Cardiac
1983;72(3):373 8. autonomic dysfunction in diabetic children. Diabetes
[22] Solders G, Thalme B, Aguirre-Aquino M, Brandt L, Care 1999;22(11):1845 50.
Berg U, Persson A. Nerve conduction and autonomic [35] Oaklander AL, Nolano M. Scientific advances in and
nerve function in diabetic children. A 10-year follow- clinical approaches to small-fiber polyneuropathy: a
up study. Acta Paediatr 1997;86(4):361 6. review. JAMA Neurol 2019;.
[23] Dart AB, Martens PJ, Rigatto C, Brownell MD, Dean [36] Karavanaki K, Baum JD. Coexistence of impaired indi-
HJ, Sellers EA. Earlier onset of complications in youth ces of autonomic neuropathy and diabetic nephropa-
with type 2 diabetes. Diabetes Care 2014;37(2):436 43. thy in a cohort of children with type 1 diabetes
[24] Donaghue KC, Fung AT, Fairchild JM, Howard NJ, mellitus. J Pediatr Endocrinol Metab 2003;16(1):79 90.
Silink M. Prospective assessment of autonomic and [37] Tang M, Donaghue KC, Cho YH, Craig ME.
peripheral nerve function in adolescents with diabetes. Autonomic neuropathy in young people with type 1
Diabet Med 1996;13(1):65 71. diabetes: a systematic review. Pediatr Diabetes 2013;14
[25] Young RJ, Macintyre CC, Martyn CN, Prescott RJ, (4):239 48.
Ewing DJ, Smith AF, et al. Progression of subclinical [38] Metwalley KA, Hamed SA, Farghaly HS. Cardiac
polyneuropathy in young patients with type 1 (insu- autonomic function in children with type 1 diabetes.
lin-dependent) diabetes: associations with glycaemic Eur J Pediatrics 2018;177(6):805 13.
control and microangiopathy (microvascular compli- [39] Barkai L, Vamosi I, Lukacs K. Prospective assessment
cations). Diabetologia 1986;29(3):156 61. of severe hypoglycaemia in diabetic children and

Diabetic Neuropathy
198 11. Diabetic neuropathy in children and youth

adolescents with impaired and normal awareness of [53] Akin O, Eker I, Arslan M, Tasdemir S, Tascilar ME,
hypoglycaemia. Diabetologia 1998;41(8):898 903. Ulas UH, et al. Association of nerve conduction
[40] Bober E, Buyukgebiz A, Verrotti A, Chiarelli F. impairment and insulin resistance in children with
Hypoglycemia, hypoglycemia unawareness and coun- obesity. Childs Nerv Syst 2016;32(11):2219 24.
terregulation in children and adolescents with type 1 [54] Thamotharampillai K, Chan AK, Bennetts B, Craig
diabetes mellitus. J Pediatr Endocrinol Metab 2005;18 ME, Cusumano J, Silink M, et al. Decline in neuro-
(9):831 41. physiological function after 7 years in an adolescent
[41] Sauder KA, Stafford JM, Mayer-Davis EJ, Jensen ET, diabetic cohort and the role of aldose reductase gene
Saydah S, Mottl A, et al. Co-occurrence of early diabetes- polymorphisms. Diabetes Care 2006;29(9):2053 7.
related complications in adolescents and young adults [55] Sosenko JM, Boulton AJ, Kubrusly DB, Weintraub JK,
with type 1 diabetes: an observational cohort study. Skyler JS. The vibratory perception threshold in young
Lancet Child Adolesc Health 2019;3(1):35 43. diabetic patients: associations with glycemia and
[42] Koch KL, Calles-Escandon J. Diabetic gastroparesis. puberty. Diabetes Care 1985;8(6):605 7.
Gastroenterol Clin North Am 2015;44(1):39 57. [56] Barkai L, Kempler P, Vamosi I, Lukacs K, Marton A,
[43] Callaghan BC, Gallagher G, Fridman V, Feldman EL. Keresztes K. Peripheral sensory nerve dysfunction in
Diabetic neuropathy: what does the future hold? children and adolescents with type 1 diabetes mellitus.
Diabetologia 2020;63(5):891 7. Diabet Med 1998;15(3):228 33.
[44] Allen C, Duck SC, Sufit RL, Swick HM, D’Alessio DJ. [57] Hoffman WH, Hart ZH, Frank RN. Correlates of delayed
Glycemic control and peripheral nerve conduction in chil- motor nerve conduction and retinopathy in juvenile-onset
dren and young adults after 5 6 mo of IDDM. Wisconsin diabetes mellitus. J Pediatr 1983;102(3):351 6.
Diabetes Registry. Diabetes Care 1992;15(4):502 7. [58] Riihimaa PH, Suominen K, Tolonen U, Jantti V, Knip
[45] Cho YH, Craig ME, Srinivasan S, Benitez-Aguirre P, M, Tapanainen P. Peripheral nerve function is increas-
Mitchell P, Jopling T, et al. Heart rate variability in ingly impaired during puberty in adolescents with
pubertal girls with type 1 diabetes: its relationship with type 1 diabetes. Diabetes Care 2001;24(6):1087 92.
glycaemic control, insulin resistance and hyperandro- [59] Duck SC, Wei FF, Parke J, Swick HM. Role of height
genism. Clin Endocrinol (Oxf) 2014;80(6):818 24. and glycosylated hemoglobin in abnormal nerve con-
[46] Cho YH, Craig ME, Jopling T, Chan A, Donaghue KC. duction in pediatric patients with type I diabetes mel-
Higher body mass index predicts cardiac autonomic litus after 4 9 yr of disease. Diabetes Care 1991;14
dysfunction: a longitudinal study in adolescent type 1 (5):386 92.
diabetes. Pediatr Diabetes 2018;19(4):794 800. [60] Steel JM, Young RJ, Lloyd GG, Clarke BF. Clinically
[47] Eng GD, Hung W, August GP, Smokvina MD. Nerve apparent eating disorders in young diabetic women:
conduction velocity determinations in juvenile diabe- associations with painful neuropathy and other compli-
tes: continuing study of 190 patients. Arch Phys Med cations. Br Med J (Clin Res Ed) 1987;294(6576):859 62.
Rehabil 1976;57(1):1 5. [61] Scheuing N, Bartus B, Berger G, Haberland H, Icks A,
[48] Olsen BS, Nir M, Kjaer I, Volund A, Mortensen HB. Knauth B, et al. Clinical characteristics and outcome of
Elevated vibration perception threshold in young-patients 467 patients with a clinically recognized eating disor-
with type-1 diabetes in comparison to nondiabetic chil- der identified among 52,215 patients with type 1 dia-
dren and adolescents. Diabet Med 1994;11(9):888 92. betes: a multicenter German/Austrian study. Diabetes
[49] Bjerg L, Hulman A, Charles M, Jorgensen ME, Witte DR. Care 2014;37(6):1581 9.
Clustering of microvascular complications in type 1 diabe- [62] Amiel SA, Sherwin RS, Simonson DC, Lauritano AA,
tes mellitus. J Diabetes Complications 2018;32(4):393 9. Tamborlane WV. Impaired insulin action in puberty. A
[50] Maguire AM, Craig ME, Craighead A, Chan AK, contributing factor to poor glycemic control in adoles-
Cusumano JM, Hing SJ, et al. Autonomic nerve testing cents with diabetes. N Engl J Med 1986;315(4):215 19.
predicts the development of complications: a 12-year [63] Moran A, Jacobs Jr. DR, Steinberger J, Cohen P, Hong
follow-up study. Diabetes Care 2007;30(1):77 82. CP, Prineas R, et al. Association between the insulin
[51] Nagai N, Matsumoto T, Kita H, Moritani T. resistance of puberty and the insulin-like growth fac-
Autonomic nervous system activity and the state and tor-I/growth hormone axis. J Clin Endocrinol Metab
development of obesity in Japanese school children. 2002;87(10):4817 20.
Obes Res 2003;11(1):25 32. [64] Cho YH, Craig ME, Donaghue KC. Puberty as an
[52] Ince H, Tasdemir HA, Aydin M, Ozyurek H, Tilki HE. accelerator for diabetes complications. Pediatr
Evaluation of nerve conduction studies in obese chil- Diabetes 2014;15(1):18 26.
dren with insulin resistance or impaired glucose toler- [65] Hietala K, Harjutsalo V, Forsblom C, Summanen P,
ance. J Child Neurol 2015;30(8):989 99. Groop PHFinnDiane Study Group. Age at onset and

Diabetic Neuropathy
 References 199
the risk of proliferative retinopathy in type 1 diabetes. [76] Clair C, Cohen MJ, Eichler F, Selby KJ, Rigotti NA.
Diabetes Care 2010;33(6):1315 19. The effect of cigarette smoking on diabetic peripheral
[66] Donaghue KC, Fung AT, Hing S, Fairchild J, King J, neuropathy: a systematic review and meta-analysis.
Chan A, et al. The effect of prepubertal diabetes dura- J Gen Intern Med 2015;30(8):1193 203.
tion on diabetes. Microvascular complications in early [77] Reynolds K, Liese AD, Anderson AM, Dabelea D,
and late adolescence. Diabetes Care 1997;20(1):77 80. Standiford D, Daniels SR, et al. Prevalence of tobacco
[67] Kordonouri O, Danne T, Enders I, Weber B. Does the use and association between cardiometabolic risk factors
long-term clinical course of type I diabetes mellitus and cigarette smoking in youth with type 1 or type 2 dia-
differ in patients with prepubertal and pubertal onset? betes mellitus. J Pediatr 2011;158(4):594 601 e1.
Results of the Berlin Retinopathy Study. Eur J Pediatr [78] Nip ASY, Reboussin BA, Dabelea D, Bellatorre A,
1998;157(3):202 7. Mayer-Davis EJ, Kahkoska AR, et al. Disordered eat-
[68] Tesfaye S, Chaturvedi N, Eaton SE, Ward JD, Manes ing behaviors in youth and young adults with type 1
C, Ionescu-Tirgoviste C, et al. Vascular risk factors or type 2 diabetes receiving insulin therapy: the
and diabetic neuropathy. N Engl J Med 2005;352 SEARCH for Diabetes in Youth Study. Diabetes Care
(4):341 50. 2019;42(5):859 66.
[69] Olsen BS, Sjolie A, Hougaard P, Johannesen J, Borch- [79] Wilfley D, Berkowitz R, Goebel-Fabbri A, Hirst K,
Johnsen K, Marinelli K, et al. A 6-year nationwide Ievers-Landis C, Lipman TH, et al. Binge eating,
cohort study of glycaemic control in young people mood, and quality of life in youth with type 2 diabe-
with type 1 diabetes. Risk markers for the develop- tes: baseline data from the today study. Diabetes Care
ment of retinopathy, nephropathy and neuropathy. 2011;34(4):858 60.
Danish Study Group of Diabetes in Childhood. J [80] Young V, Eiser C, Johnson B, Brierley S, Epton T,
Diabetes Complications 2000;14(6):295 300. Elliott J, et al. Eating problems in adolescents with
[70] Hajas G, Kissova V, Tirpakova A. A 10-yr follow-up Type 1 diabetes: a systematic review with meta-
study for the detection of peripheral neuropathy in analysis. Diabet Med 2013;30(2):189 98.
young patients with type 1 diabetes. Pediatr Diabetes [81] Pursey KM, Hart M, Jenkins L, McEvoy M, Smart CE.
2016;17(8):632 41. Screening and identification of disordered eating in
[71] National Cholesterol Education Program Expert Panel people with type 1 diabetes: a systematic review.
on Detection Evaluation, Treatment of High Blood J Diabetes Complications 2020;34(4):107522.
Cholesterol in Adults (Adult Treatment Panel III). [82] Cherubini V, Skrami E, Iannilli A, Cesaretti A,
Third Report of the National Cholesterol Education Paparusso AM, Alessandrelli MC, et al. Disordered
Program (NCEP) Expert Panel on Detection, eating behaviors in adolescents with type 1 diabetes: a
Evaluation, and Treatment of High Blood Cholesterol cross-sectional population-based study in Italy. Int J
in Adults (Adult Treatment Panel III) final report. Eat Disord 2018;51(8):890 8.
Circulation 2002;106(25):3143 421. [83] Naughton MJ, Ruggiero AM, Lawrence JM,
[72] Group TS, Zeitler P, Epstein L, Grey M, Hirst K, Imperatore G, Klingensmith GJ, Waitzfelder B, et al.
Kaufman F, et al. Treatment options for type 2 diabe- Health-related quality of life of children and adoles-
tes in adolescents and youth: a study of the compara- cents with type 1 or type 2 diabetes mellitus: SEARCH
tive efficacy of metformin alone or in combination for Diabetes in Youth Study. Arch Pediatr Adolesc
with rosiglitazone or lifestyle intervention in adoles- Med 2008;162(7):649 57.
cents with type 2 diabetes. Pediatr Diabetes 2007;8 [84] Morgan E, Cardwell CR, Black CJ, McCance DR,
(2):74 87. Patterson CC. Excess mortality in Type 1 diabetes
[73] Copeland KC, Zeitler P, Geffner M, Guandalini C, diagnosed in childhood and adolescence: a systematic
Higgins J, Hirst K, et al. Characteristics of adolescents review of population-based cohorts. Acta Diabetol
and youth with recent-onset type 2 diabetes: the 2015;52(4):801 7.
TODAY cohort at baseline. J Clin Endocrinol Metab [85] Bril V, England J, Franklin GM, Backonja M, Cohen J,
2011;96(1):159 67. Del Toro D, et al. Evidence-based guideline: treat-
[74] Baum P, Petroff D, Classen J, Kiess W, Bluher S. ment of painful diabetic neuropathy: report of the
Dysfunction of autonomic nervous system in child- American Academy of Neurology, the American
hood obesity: a cross-sectional study. PLoS One 2013;8 Association of Neuromuscular and Electrodiagnostic
(1):e54546. Medicine, and the American Academy of Physical
[75] Rodriguez-Colon SM, Bixler EO, Li X, Vgontzas AN, Medicine and Rehabilitation. Neurology 2011;76
Liao D. Obesity is associated with impaired cardiac (20):1758 65.
autonomic modulation in children. Int J Pediatr Obes
2011;6(2):128 34.

Diabetic Neuropathy
 C H A P T E R

12
Treatment-induced painful diabetic
neuropathy
Milla Rosengård-Bärlund1 and Hanna Harno2
1
Division of Endocrinology, Abdominal Center, Helsinki University Hospital and University of
Helsinki, Helsinki, Finland 2Clinical Neurosciences, Neurology, Helsinki University Hospital and
University of Helsinki, Helsinki, Finland

Introduction in some cases accompanied by worsening dia-

betic retinopathy (DR) and nephropathy.
“Insulin neuritis” is a historical term for Gibbons and Freeman retrospectively reviewed
acute painful neuropathy related to the initia- all individuals referred to a tertiary care diabetes
tion of insulin treatment and the subsequent neuropathy clinic over 5 years [10]. They identi-
improvement in glycemic control in patients fied a subset of patients with type 1 diabetes and
with diabetes. Caravati was the first to describe painful neuropathy, preceded by a significant
this condition in 1933 in a type 1 diabetic improvement in HbA1c. Accordingly, they pro-
patient who developed severe pain following posed the following three clinical criteria for the
rapid glycemic control with insulin [1]. Since diagnosis of TIND: (1) a decrease in HbA1c $ 2%
the initial report in 1933, several small case (22 mmol/mol) over 3 months; (2) the acute onset
series described an association between acute of neuropathic pain and/or autonomic dysfunc-
onset of painful neuropathy and a rapid correc- tion, developing acutely and severely enough to
tion of hyperglycemia [2 9]. The term “insulin cause the patient to seek medical attention; and
neuritis” does not correctly describe the condi- (3) the neuropathic pain or autonomic dysfunc-
tion since this form of painful neuropathy can tion beginning within 8 weeks of the decrease in
occur after any antidiabetic agent is adminis- HbA1c. In addition, they noticed that patients
tered. Thus it has been proposed that treatment- who developed TIND typically had very high
induced neuropathy of diabetes (TIND) would pretreatment HbA1c . 10% (86 mmol/mol). In
be a more descriptive name for it [9]. The main this cohort, none of the patients had newly diag-
clinical manifestations are severe treatment- nosed type 1 diabetes. Nevertheless, there are sev-
resistant pain and autonomic dysfunction sec- eral case reports of patients with a recent
ondary to a rapid decrease in hemoglobin A1c, diagnosis of type 1 diabetes and TIND, even in

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00011-6 201 © 2022 Elsevier Inc. All rights reserved.
202 12. Treatment-induced painful diabetic neuropathy

children and adolescent patients [5,11 14]. Data addition, pro-inflammatory cytokines have been
regarding the prevalence of TIND in type 2 diabe- proposed to have a role in the development of
tes are even more limited, but TIND appears hyperalgesia and autonomic dysfunction [18].
more frequently reported in patients with type 1 Moreover, fiber regeneration and spontaneous
diabetes. This chapter will review the data avail- firing of regenerating nerve fibers may play a
able on the epidemiology of TIND, as well as its role in the pain mechanisms [3].
mechanisms, risk factors, clinical presentation, One important predisposing factor for
diagnostic methods, and management. chronic neuropathic pain in diabetes patients is
an impaired function of the descending, inhi-
biting pain pathway of the dorsal horn of the
Potential mechanisms and pathogenesis spinal cord, which has been shown, for exam-
ple, in a brain imaging study that compared
Although several theories have been pro- patients with painful and nonpainful DPN
posed, the pathophysiology of TIND is not fully [19], and in a duloxetine intervention study of
understood. In diabetic polyneuropathy (DPN), painful diabetic neuropathy [20]. However, the
the most typical finding is a mixed fiber neurop- function of the descending, inhibiting pain
athy with both myelinated and unmyelinated pathway has not been studied in TIND patients
fibers affected. However, the typical clinical pre- thus far.
sentation of TIND is a length-dependent, sym-
metrical small-fiber neuropathy (SFN) associated
with sensory and/or autonomic symptoms. In Epidemiology
SFN, thinly myelinated Aδ-fibers and unmyelin-
ated C-fibers are affected [15]. Large fibers are The prevalence of TIND in the general pop-
not affected, and thus nerve conduction studies ulation is unknown. The available data are
(NCSs) remain normal. However, some TIND based on case reports [21] and small,
patients have shown mild sensory, axonal neu- nonpopulation-based studies [10,22]. The larg-
ropathy in NCSs [10]. est cohort thus far was studied by Gibbons and
One suggested pathogenetic mechanism of Freeman [10]. Over 5 years, 954 patients with
TIND is the formation of arteriovenous shunting type 1 diabetes were retrospectively evaluated
and endoneurial ischemia as a result of a steal for diabetic neuropathy at a tertiary center for
effect from the proliferating epineurial vessels neurologic disorders, and 10.9% of these indi-
[4]. The coincident changes in the retina and kid- viduals met the suggested criteria for TIND.
ney support this microvascular mechanism. The study design did not allow assessment of
Another proposed mechanism is relative the population-based prevalence of the disor-
hypoglycemia, since the previous long-standing der. However, the findings indicated that
hyperglycemia may have led to energy- TIND might be more common than previously
dependent failure of axonal transport [10,16]. believed [10]. There is also no precise informa-
Consequently, a rapid reduction in the blood tion on the incidence of TIND.
glucose level may result in energy deficiency.
Unmyelinated fibers with a large surface area-
to-size ratio, compared to myelinated fibers, Risk factors
could be particularly susceptible to a sudden
reduction in glucose supply. In a cell model of The factors that predispose patients to TIND
diabetic neuropathy, relative glucose depriva- are unknown. Gibbons and Freeman reported
tion resulted in cellular apoptosis [17]. In an association between the severity of TIND

Diabetic Neuropathy
 Clinical presentation 203
and the extent of the change in HbA1c; the risk go unnoticed for months [26]. This period of
of TIND depended on the reduction in HbA1c undiagnosed and untreated hyperglycemia
level. TIND had a 20% chance of occurring often results in high HbA1c at diagnosis and,
when HbA1c levels decreased during 3 months frequently, in ketoacidosis. Thus neural dys-
by two to three units and more than an 80% function may already be present when type 1
chance when HbA1c levels decreased by more diabetes is diagnosed.
than four units [10]. Young women with type 1
diabetes and a history of eating disorders
appear to be at increased risk of developing Clinical presentation
TIND [10,23]. Moreover, TIND has been
described in the context of rapid improvements The typical clinical presentation of TIND is
in glycemic control and weight loss, both after an acute or subacute onset of severe neuro-
bariatric surgery and as a result of a low- pathic pain and autonomic dysfunction within
calorie diet. Common denominators seem to be 2 8 weeks of the improvement in glucose con-
weight loss, parallel with the rapid correction trol (Table 12.1). Typically, patients experience
of glycemic control and nutritional deficiencies. a feeling of constant burning or lancinating
Painful neuropathy and autonomic dysfunc- pain in the feet. In addition, patients have
tion similar to TIND have been reported in reported tingling, numbness, chills, a feeling of
some patients after bariatric surgery, attributed tightness or pressure, and sometimes itching.
to malnutrition and deficiencies in micronutri- Patients frequently experience pain induced by
ents such as vitamin B12 [24,25]. touch (allodynia) where, for example, skin con-
Notably, there are no published neuropatho- tact with sheets or socks feels painful, or ther-
logical or neurophysiological data regarding mal dysesthesia and allodynia, where a hot
patients with TIND prior to the onset of neuro- shower may be perceived as unpleasant or
pathic symptoms, nor are there prospective painful [9,10].
studies that assessed the correlation between Moreover, weakness, painful muscle cramps
the development of TIND and the change in in the legs, and restless legs syndrome have
HbA1c level or other biomarkers. TIND has not been described as TIND symptoms. Usually,
been as well described in newly diagnosed such symptoms begin in the distal parts of the
patients with type 1 diabetes, although case lower limbs and slowly progress proximally in
reports do exist [5,12 14]. Notably, adult-onset a length-dependent fashion. In some cases, the
type 1 diabetes is usually a slowly progressive symptoms spread to the upper limbs or pres-
disease, where symptoms and weight loss can ent in patches in a nonlength-dependent

TABLE 12.1 Typical clinical presentation of treatment-induced neuropathy (TIND).

• Long-standing hyperglycemia HbA1c . 10% (86 mmol/mol)
• Type 1 diabetes, either long duration or newly diagnosed with slow development, often ketoacidosis and weight loss
• Type 2 diabetes, either remote or recent diagnosis
• Rapidly improved glycemic control with any antidiabetic pharmacotherapy or a combination (insulin, other injectable,
oral antidiabetic agents) or by bariatric surgery or diet
• A drop of HbA1c $ 2% (22 mmol/mol) within 3 months
• Onset of symptoms within 2 8 weeks of the change in glucose control
• Painful polyneuropathy and/or small-fiber neuropathy (SFN)
• Symptoms of autonomic dysfunction
• Occasionally history of eating disorder, weight loss, bariatric surgery

Diabetic Neuropathy
204 12. Treatment-induced painful diabetic neuropathy

pattern. They typically get worse when the per- hypoglycemia unawareness and sudomotor dys-
son is at rest and may interfere with sleep. function with either increased or decreased
In addition to neuropathic pain, a patient may sweating may be seen.
present with various symptoms of autonomic
dysfunction, which are probably underreported
or dismissed due to severe neuropathic pain. Diagnostic methods
Autonomic dysfunction may be completely
asymptomatic and detected only by thorough In general, the diagnosis of typical diabetic
clinical examination. Symptoms of dysautonomia peripheral neuropathy is clinical [27]. DPN usu-
are more likely to be identified through detailed ally involves a mix of large-fiber neuropathy
questioning or by using validated questionnaires and SFN, and of the patients with diabetic neu-
[9]. Occasionally, autonomic symptoms may be ropathy, 25% report painful neuropathy [27].
the presenting feature. Severe autonomic failure TIND is considered primarily an SFN, thus its
may also be seen in some cases. There appears to diagnosis may be challenging. Notably, electro-
be a relationship between the severity of auto- myoneurography (EMNG) remains normal in
nomic failure and the decrease in HbA1c [10]. SFN, and other tests are needed to confirm sus-
Common symptoms of cardiovascular autonomic pected SFN. However, DPN may be preexisting
dysfunction are arrhythmias or palpitations (rest- in TIND patients with a long history of diabetes
ing tachycardia), orthostatic intolerance, or ortho- and inadequate glycemic control. NCS (EMNG)
static hypotension with syncope. Gastrointestinal is the gold standard as a confirmatory test in
autonomic dysfunction may manifest with early painful DPN diagnostic work-up.
satiety, nausea, vomiting, or dyspepsia due to The diagnosis of TIND is based on the
oesophagic dysmotility or gastroparesis or with patient’s history and symptoms, as well as a
constipation, diarrhea, or fecal incontinence. clinical examination that includes a bedside
Possible genitourinary disturbances are erectile assessment of their sensory and motor activities
dysfunction and bladder dysfunction. Moreover, (Table 12.2). At the bedside, small-fiber function

TABLE 12.2 The diagnostic tests of treatment-induced neuropathy (TIND).

Clinical examination • Usually normal

• Painful distal paresthesia (sock-glove distribution) and symptoms of autonomic
dysfunction
Nerve conduction studies • Sensorimotor polyneuropathy
(EMNG)
Small-fiber testing • Quantitative Sensory Testing (QST)
• Skin biopsy (10 cm above lateral malleolus)
• Corneal Confocal Microscopy (CCM)
Autonomic testing • Parasympathetic and sympathetic dysfunction tests
• Quantitative Sudomotor Axon Reflex Test (QSART)
• Thermoregulatory Sweat Testing (TST)
Laboratory tests • For exclusion of other causes
• TSH, ANA, ESR, vitamin B12, vitamin B1, serum protein electrophoresis and light
chain analysis
Differential diagnosis • Exclusion of other rare forms of polyneuropathies in atypical or exceptionally severe
cases

Diabetic Neuropathy
 Diagnostic methods 205
is examined by testing the patient’s responses to biopsy (Fig. 12.1), quantitative sensory testing
different stimuli, including heat, cold, pinprick, (QST), autonomic function tests according to
pressure, and touch (with a metal roller or symptoms, and exclusion of other causes
object, sharp wooden cocktail stick, finger com- (Table 12.2).
pression, painter’s brush, and a cotton tuft, Skin punch biopsy and the assessment of
respectively). Pinprick and temperature sensa- intraepidermal nerve fiber density (IENFD) is
tions are mediated via small nerve fibers, the most sensitive tool and the gold standard
whereas large nerve fibers mediate monofila- for diagnosing SFN [28]. However, a skin
ment sensation, vibration sensation, and propri- biopsy is invasive and thus primarily used in
oception. For motor testing, heel-toe walking specialized clinics. In clinical practice, skin
and reflexes are important. Sensory responses biopsy should most likely be used more often
may include increased painful sensation (hyper- in the SFN diagnostic procedure. Fig. 12.1
algesia), a nonpainful stimulus (e.g., heat, cold, demonstrates skin biopsies from a healthy con-
painter’s brush, or light touch) causing pain trol and a patient with severe TIND.
(allodynia), a stimulus causing an unpleasant QST is a psychophysical measurement that
sensation (dysesthesia), or a stimulus causing a requires relevant patient cooperation in evaluat-
diminished sensation (hypoesthesia). ing responses to stimuli. In recent years, a more
The tests that can confirm TIND-related extensive QST has been developed to cover, not
polyneuropathy and/or SFN are EMNG, skin only thermal thresholds, but all the sensory
FIGURE 12.1 Images from
skin biopsies and corneal confo-
cal microscopy (CCM) from a
healthy control (A and B) and a
patient with newly diagnosed
type 1 diabetes and severe TIND
(C and D). Note the numerous
nerve fibers in the epidermis of
the skin biopsy of the healthy
control (A) versus the biopsy of
the patient with TIND (C). The
CCM shows a significant loss of
corneal nerve fibers in the
patient with TIND (D) versus a
normal nerve fiber density in
the healthy control (B).

Diabetic Neuropathy
206 12. Treatment-induced painful diabetic neuropathy

modalities concerned for SFN. This larger QST is Management of symptoms associated with
called DFNS (Deutscher Forschungsverbund treatment-induced neuropathy of diabetes
Neuropatischer Schmertz)-QST, according to the
German Research Network on Neuropathic Pain, Typically, patients with TIND experience
and it is currently in clinical and research use in severe, subacute neuropathic pain, and over-
several centers [29 31]. With DFNS-QST, a the-counter analgesics usually do not alleviate
patient-specific sensory profile can be developed it. In pain management, the same medications
[31]. This may be a helpful tool for improving are used as in other forms of painful diabetic
the accuracy of SFN diagnostics in the future. neuropathy [tricyclic antidepressants, gabapen-
Corneal confocal microscopy (CCM) is a tinoids, and serotonin-norepinephrine reup-
promising repeatable, noninvasive method for take inhibitors (SNRIs)]. However, due to its
assessing small-fiber damage in diabetic periph- rapid start, the neuropathic pain may be more
eral neuropathy [32 34]. In several studies, mea- severe than usual in painful diabetic neuropa-
sures of CCM have correlated with IENFD, the thy, and the pain medication may need rapid
severity of painful neuropathic symptoms and escalation and/or the addition of several neu-
quality of life, and autonomic neuropathy mea- ropathic pain medications. Correspondingly,
sures in patients with painful DPN [35 37]. the intake of this medication should be gradu-
However, a study attempting to detect DPN ally reduced as the pain dissipates.
from a cohort of confirmed and unconfirmed Treatment of neuropathic pain may not alter
DPN, and healthy controls by assessing CCM the natural course of TIND but is essential to
remained negative, although corneal nerve fiber improving the patient’s quality of life and pain
density was lower in patients with DM com- control. In some cases of TIND, maximizing treat-
pared with controls [38]. Fig. 12.1 demonstrates ment with a single neuropathic pain agent may
corneal confocal microscopy and nerve fiber den- help. Although tricyclic antidepressant medica-
sity in a healthy control and a patient with severe tions can be used, the anticholinergic side effects
TIND. Further studies are needed to validate may be worsen autonomic dysfunction. In those
CCM for routine clinical use and in the assess- cases, gabapentinoids or SNRIs may be a good
ment of TIND. alternative. Eventually, neuropathic pain may be
Methods that assess sweat gland density and resolved with stable glycemic control. It is essen-
distribution, and stimulated sweat production, tial to inform the patient about the disease course
have not yet been validated for routine clinical and its spontaneous tendency to heal, although
use. Other confirmatory tests of SFN include healing may take anywhere from 6 to 24 months.
laser Doppler flare imaging studies and posi- Particularly in patients with newly diagnosed
tron emission tomography scans. These have no type 1 diabetes, the burden of a recent diagnosis
exact role in routine clinical diagnosis. combined with severe neuropathic pain may be
If a patient under suspicion of having TIND challenging. Chronic pain is also associated with
shows atypical features, such as asymmetric pre- worse self-management of diabetes and higher
sentation of neuropathy or motor predominance, blood glucose levels [39]. Thus the patient may
a neurological consult, as in other forms of dia- need psychological support. Historically, it was
betic neuropathy, should be considered. believed that the pain related to TIND could be
Additional testing options include measuring relieved through “permissive hyperglycemia.”
the patient’s serum vitamin B12 levels, conduct- However, no data support that view. On the con-
ing thyroid function tests, serum protein electro- trary, it appears that allowing hyperglycemia
phoresis with immunofixation, and assessing for increases the risk of DPN and new episodes of
markers of autoimmune disorders. TIND [22], thus sustained glycemic control is

Diabetic Neuropathy
 Discussion 207
recommended. Pharmacologic treatment of auto- supported by this study, TIND is considered a
nomic dysfunction is challenging, although reversible form of diabetic neuropathy, and,
symptomatic treatment of tachycardia, gastroin- importantly, it has been observed in structural
testinal symptoms, and erectile dysfunction may neural outcomes in patients with a long history
improve a patient’s quality of life. Diabetic auto- of type 1 diabetes.
nomic neuropathy is considered an irreversible, Several studies have reported the develop-
progressive form of neuropathy. Importantly, ment or progression of DR and microalbumi-
autonomic dysfunction occurring with TIND is nuria in association with TIND; however, data
usually reversible and resolves along with the are limited [9 11,23]. In a follow-up study con-
other symptoms. ducted by Gibbons and Freeman, more than
half of those with TIND had no DR, and only
3% had proliferative retinopathy at baseline
Outcomes of treatment-induced [10]. After the development of TIND, 90% had
neuropathy of diabetes and association either severe nonproliferative or proliferative
with other microvascular complications retinopathy. The pathophysiology of this phe-
nomenon is not well understood but supports a
Data on long-term follow-up of patients common microvascular etiology. Early worsen-
with TIND are limited. Based on previous case ing of DR is a well-described phenomenon asso-
reports and small groups of studied cohorts, ciated with the rapid improvement in glycemic
manifestations of TIND are typically resolved control in patients with both type 1 and type 2
within 1 2 years. In a recent study that is, to diabetes [40 44]. However, longitudinal studies
our knowledge, the only follow-up study that have demonstrated that intensive glucose con-
has been conducted on this topic, 26 individuals trol is the cornerstone of the reduction of dia-
with type 1 diabetes and TIND were followed betic microvascular complications. Despite the
for 8 years [22]. The neuropathic pain scores initial worsening, intensively treated patients
improved within 18 36 months after the onset, have demonstrated a sustained protective effect
which supports previous data. Interestingly, of improved glycemic control during follow-up.
although the patients in both groups were simi- Therefore it is likely that the deterioration of DR
lar at baseline, the outcomes diverged substan- seen in patients with TIND can be attributed to
tially, depending on the glycemic control improved glycemic control and not neuropathy.
during follow-up. For the 19 patients who main- In a follow-up study of patients with TIND,
tained stable glycemic control during follow-up, the patients with inadequate glycemic control
their autonomic scores, NCSs, IENFDs, pain after experiencing TIND had a substantial
scores, and microvascular complications were decline in renal function, and some patients
observed to have improved. progressed to hemodialysis [10,22].
In contrast, a subgroup with unstable glycemic
control during follow-up experienced new TIND
episodes, persistent neuropathic pain, and dem- Discussion
onstrated deterioration in all measures. Notably,
despite severe TIND at baseline, stable glycemic TIND is a rare, underreported, and often
control allowed almost complete recovery in both unidentified form of diabetic neuropathy. Its
structural and functional neural parameters. The prevalence in the general population, though
neural damage in diabetic peripheral polyneuro- unknown, appears to be increasing [10].
pathy is considered irreversible, as intervention During the last decades, diabetes guidelines
studies have failed to reverse the process. As have emphasized the importance of achieving

Diabetic Neuropathy
208 12. Treatment-induced painful diabetic neuropathy

treatment goals to prevent microvascular com- a known phenomenon associated with rapid
plications. Thus increased attention to reaching improvement in glycemic control. Accordingly,
adequate glycemic control could be one expla- eye status should be assessed in patients at
nation for the suggested increase in TIND increased ocular risk before antidiabetic treat-
prevalence. Moreover, emerging potent antidi- ment is intensified. Moreover, screening for
abetic agents and new diabetes technology, both retinopathy and microalbuminuria could
including flash glucose monitoring, continuous be recommended in all TIND presentations,
glucose monitoring, and hybrid closed-loop irrespective of diabetes duration.
systems, allow better glycemic control. There are several pharmacological treatment
Nonetheless, there is a mismatch between options for painful diabetic neuropathy; how-
the number of patients whose glycemic control ever, disease-modifying therapies for TIND, and
rapidly improves and those who develop even diabetic neuropathy in general, are still
TIND. It is also not fully understood why only lacking. Importantly, painful neuropathy is both
some patients with neuropathy experience underreported and inadequately treated [47].
painful symptoms. Patients who develop TIND Severe TIND often has debilitating consequences
may also have an unidentified predisposing for the patient, including interference with daily
condition. The association between variants in activities, work disability, and reduced health-
the voltage-gated sodium channel NaV1.7 related quality of life. For patients at risk of
encoded by the SCN9A gene [45,46] and TIND developing TIND, regularly conducted question-
has not been studied. ing with a pain scale and patient pain drawings
No recommendations have been made to could allow for early recognition and pain man-
identify individuals at risk or prevent the agement. Whether patients with TIND are at risk
development of TIND. Based on the available of later diabetic neuropathy is not known.
data, a precaution in the intensification of treat- However, based on the available data, it appears
ment is essential in long-standing hyperglyce- that, at least in patients with type 1 diabetes,
mia, especially in young females who have stable glycemic control prevents the progression
experienced weight loss [22]. In a newly diag- of neuropathy and even improves neural mar-
nosed type 1 diabetic patient, hyperglycemia kers after a 9-year follow-up [22,23]. Regarding
may have lasted for months; thus symptoms diabetic peripheral neuropathy, the American
lasting for weeks or months, high HbA1c, or a Diabetes Association recommends assessing for
history of ketoacidosis may indicate an it in patients who have had type 1 diabetes for
increased risk for TIND. Therefore TIND could 5 years, and at diagnosis in patients with type
be even more common in patients with new- 2 diabetes or those with impaired glucose toler-
onset type 1 diabetes, as initiation of insulin ance or neuropathic symptoms [27]. Importantly,
treatment almost invariably results in rapid TIND should be considered in any diabetes
improvement of glycemic control. Gibbons patient experiencing acute neuropathic pain or
et al. propose limiting the reduction in HbA1c exhibiting signs of autonomic dysfunction
to less than 2% (22 mmol/mol) over 3 months regardless of the type or duration of diabetes or
[10], although prospective data on prevention the current treatment.
are lacking. Thus, based on the available data, Many questions remain unanswered. The
stable glycemic control is recommended in the optimal rate for reestablishing glycemic control
management of individuals with type 1 diabe- that prevents TIND but does not compromise
tes and TIND. No longitudinal follow-up stud- the future risk of microvascular complications
ies of people with type 2 diabetes and TIND is unknown. Thus far, treatment for TIND
have been conducted. Early worsening of DR is comprises pain management, symptomatic

Diabetic Neuropathy
 References 209
treatment of autonomic dysfunction, and opti- [10] Gibbons CH, Freeman R. Treatment-induced neuropa-
mization of glycemic control. Further research thy of diabetes: an acute, iatrogenic complication of
diabetes. Brain 2015;138(1):43 52.
into the pathophysiology, risk factors, treat- [11] Alexandrou EG, Corathers SD, Lahoti A, Redel J, Tellez S,
ment, and prevention of TIND is needed. Jones NHY, et al. Treatment-induced neuropathy of dia-
betes in youth: case series of a heterogeneous and chal-
lenging complication. J Endocr Soc 2020;4(12):1 6.
[12] Chandler E, Brown M, Wintergerst K, Doll E.
Declaration of competing interests Treatment-induced neuropathy of diabetes (TIND) in
pediatrics: a case report and review of the literature. J
None Clin Endocrinol Metab 2020;105(2) dgz067.
[13] Dayal D, Jayaraman D, Sankhyan N, Singhi P. Acute
painful neuropathy in a girl with type 1 diabetes: long
term follow-up. J Clin Diagnostic Res 2016;10(5):1 2.
Acknowledgments [14] Tran C, Philippe J, Ochsner F, Kuntzer T, Truffert A.
The authors want to thank Prof. Anders Paetau for the skin Acute painful diabetic neuropathy: an uncommon,
biopsy images. remittent type of acute distal small fibre neuropathy.
Swiss Med Wkly 2015;145:w14131.
[15] Terkelsen AJ, Karlsson P, Lauria G, Freeman R,
Finnerup NB, Jensen TS. The diagnostic challenge of
References small fibre neuropathy: clinical presentations, evalua-
[1] Caravati CM. Insulin neuritis. A case report. Va Med tions, and causes. Lancet Neurol 2017;16(11):934 44.
Mon 1933;59:745 6. [16] Low PA, Singer W. Treatment-induced neuropathy of
[2] Archer AG, Watkins PJ, Thomas PK, Sharma AK, Payan J. diabetes: an energy crisis? Brain 2015;138(Pt 1):2 3.
The natural history of acute painful neuropathy in diabetes [17] Honma H, Podratz JL, Windebank AJ. Acute glucose
mellitus. J Neurol Neurosurg Psychiatry 1983;46(6):491 9. deprivation leads to apoptosis in a cell model of acute
[3] Llewelyn JG, Thomas PK, Fonseca V, King RHM, diabetic neuropathy. J Peripher Nerv Syst 2003;2:65 74.
Dandona P. Acute painful diabetic neuropathy precip- [18] Gibbons CH, Adler GK, Bonyhay I, Freeman R.
itated by strict glycaemic control. Acta Neuropathol Experimental hypoglycemia is a human model of
1986;72(2):157 63. stress-induced hyperalgesia. Pain 2012;11:2204 9.
[4] Tesfaye S, Malik R, Harris N, Jakubowski JJ, Mody C, [19] Segerdahl AR, Themistocleous AC, Fido D, Bennett
Rennie IG, et al. Arterio-venous shunting and prolifer- DL, Tracey I. A brain-based pain facilitation mecha-
ating new vessels in acute painful neuropathy of rapid nism contributes to painful diabetic polyneuropathy.
glycaemic control (insulin neuritis). Diabetologia Brain 2018;141(2):357 64.
1996;39(3):329 35. [20] Yarnitsky D, Granot M, Nahman-Averbuch H,
[5] Wilson JL, Sokol DK, Smith LH, Snook RJ, Waguespack Khamaisi M, Granovsky Y. Conditioned pain modula-
SG, Kincaid JC. Acute painful neuropathy (insulin neuri- tion predicts duloxetine efficacy in painful diabetic
tis) in a boy following rapid glycemic control for type 1 neuropathy. Pain 2012;153(6):1193 8.
diabetes mellitus. J Child Neurol 2003;18(5):365 7. [21] Hwang YT, Davies G. ‘Insulin neuritis’ to ‘treatment-
[6] Leow MKS, Wyckoff J. Under-recognised paradox of induced neuropathy of diabetes’: new name, same
neuropathy from rapid glycaemic control. Postgrad mystery. Pract Neurol 2016;16(1):53 5.
Med J 2005;81(952):103 7. [22] Gibbons CH. Treatment induced neuropathy of diabe-
[7] Song KB, Cho SJ, Minn YK, Kwon KH, Park MK. A tes—long term implications in type 1 diabetes. J
case of insulin neuritis that developed in a patient Diabetes Complicat 2017;31(4):715 20.
under regular insulin treatment on increasing the [23] Gibbons CH, Goebel-Fabbri A. Microvascular compli-
insulin dose: insulin neuritis: is it a misnomer? J cations associated with rapid improvements in glyce-
Neurol 2009;256(2):274 5. mic control in diabetes. Curr Diab Rep 2017;17(7):48.
[8] Dabby R, Sadeh M, Lampl Y, Gilad R, Watemberg N. [24] Juhasz-Pocsine K, Rudnicki SA, Archer RL, Harik SI.
Acute painful neuropathy induced by rapid correction Neurologic complications of gastric bypass surgery for
of serum glucose levels in diabetic patients. Biomed morbid obesity. Neurology 2007;68(21):1843 50.
Pharmacother 2009;63(10):707 9. [25] Thaisetthawatkul P, Collazo-Clavell ML, Sarr MG,
[9] Gibbons CH, Freeman R. Treatment-induced diabetic Norell JE, Dyck PJB. A controlled study of peripheral
neuropathy: a reversible painful autonomic neuropa- neuropathy after bariatric surgery. Neurology 2004;63
thy. Ann Neurol 2010;67(4):534 41. (8):1462 70.

Diabetic Neuropathy
210 12. Treatment-induced painful diabetic neuropathy

[26] Sabbah E, Savola K, Ebeling T, Kulmala P, Vahasalo P, diabetic autonomic neuropathy. Muscle Nerve 2015;52
Ilonen J, et al. Genetic, autoimmune, and clinical char- (3):363 70.
acteristics of childhood- and adult-onset type 1 diabe- [38] Andersen S, Grosen K, Tankisi H, Charles M,
tes. Diabetes Care 2000;23(9):1326 32. Andersen N, Andersen H, et al. Corneal confocal
[27] Pop-Busui R, Boulton AJM, Feldman EL, Bril V, microscopy as a tool for detecting diabetic polyneuro-
Freeman R, Malik RA, et al. Diabetic neuropathy: a pathy in a cohort with screen-detected type 2 diabetes:
position statement by the American diabetes associa- ADDITION-Denmark. J Diabetes Complicat 2018;32
tion. Diabetes Care 2017;40(1):136 54. (12):1153 9.
[28] Tesfaye S, Boulton AJ, Dyck PJ, Freeman R, Horowitz [39] Krein SL, Heisler M, Piette JD, Makki F, Kerr EA. The
M, Kempler P, et al. Diabetic neuropathies: update on effect of chronic pain on diabetes patients’ self-
definitions, diagnostic criteria, estimation of severity, management. Diabetes Care 2005;28(1):65 70.
and treatments. Diabetes Care 2010;33(10):2285 93. [40] Bain SC, Klufas MA, Ho A, Matthews DR. Worsening
[29] Baron R, Maier C, Attal N, Binder A, Bouhassira D, of diabetic retinopathy with rapid improvement in
Cruccu G, et al. Peripheral neuropathic pain: a systemic glucose control: a review. Diabetes Obes
mechanism-related organizing principle based on sen- Metab 2019;21(3):454 66.
sory profiles. Pain 2017;158(2):261 72. [41] The Diabetes Control and Complications Trial
[30] Rolke R, Baron R, Maier C, Tölle TR, Treede RD, Beyer A, Research Group. Early worsening of diabetic retinopa-
et al. Quantitative sensory testing in the German Research thy in the diabetes control and complications trial.
Network on Neuropathic Pain (DFNS): standardized pro- Arch Ophthalmol 1998;116(7):874 86.
tocol and reference values. Pain 2006;123(3):231 43. [42] Lachin JM, White NH, Hainsworth DP, Sun W, Cleary
[31] Üçeyler N, Vollert J, Broll B, Riediger N, Langjahr M, PA, Nathan DM. Effect of intensive diabetes therapy
Saffer N, et al. Sensory profiles and skin innervation on the progression of diabetic retinopathy in patients
of patients with painful and painless neuropathies. with type 1 diabetes: 18 years of follow-up in the
Pain 2018;159(9):1867 76. DCCT/EDIC. Diabetes 2015;64(2):631 42.
[32] Malik RA, Kallinikos P, Abbott CA, Van Schie CHM, [43] Turner R. Intensive blood-glucose control with sul-
Morgan P, Efron N, et al. Corneal confocal microscopy: a phonylureas or insulin compared with conventional
non-invasive surrogate of nerve fibre damage and repair treatment and risk of complications in patients with
in diabetic patients. Diabetologia 2003;46(5):683 8. type 2 diabetes (UKPDS 33). Lancet 1998;352
[33] Quattrini C, Tavakoli M, Jeziorska M, Kallinikos P, (9131):837 53.
Tesfaye S, Finnigan J, et al. Surrogate markers of small [44] Vilsbøll T, Bain SC, Leiter LA, Lingvay I, Matthews D,
fiber damage in human diabetic neuropathy. Diabetes Simó R, et al. Semaglutide, reduction in glycated hae-
2007;56(8):2148 54. moglobin and the risk of diabetic retinopathy.
[34] Tavakoli M, Petropoulos IN, Malik RA. Corneal confo- Diabetes Obes Metab 2018;20(4):889 97.
cal microscopy to assess diabetic neuropathy: an eye [45] Blesneac I, Themistocleous AC, Fratter C, Conrad LJ,
on the foot. J Diabetes Sci Technol 2013;7(5):1179 89. Ramirez JD, Cox JJ, et al. Rare NaV1.7 variants associ-
[35] Chen X, Graham J, Dabbah MA, Petropoulos IN, ated with painful diabetic peripheral neuropathy. Pain
Ponirakis G, Asghar O, et al. Small nerve fiber quanti- 2018;159(3):469 80.
fication in the diagnosis of diabetic sensorimotor poly- [46] Li QS, Cheng P, Favis R, Wickenden A, Romano G,
neuropathy: comparing corneal confocal microscopy Wang H. SCN9A variants may be implicated in neuro-
with intraepidermal nerve fiber density. Diabetes Care pathic pain associated with diabetic peripheral neu-
2015;38(6):1138 44. ropathy and pain severity. Clin J Pain 2015;31
[36] Kalteniece A, Ferdousi M, Azmi S, Mubita WM, (11):976 82.
Marshall A, Lauria G, et al. Corneal confocal micros- [47] Daousi C, MacFarlane IA, Woodward A, Nurmikkot
copy detects small nerve fibre damage in patients with TJ, Bundred PE, Benbow SJ. Chronic painful periph-
painful diabetic neuropathy. Sci Rep 2020;10(1):3371. eral neuropathy in an urban community: a controlled
[37] Tavakoli M, Begum P, Mclaughlin J, Malik RA. comparison of people with and without diabetes.
Corneal confocal microscopy for the diagnosis of Diabet Med 2004;21(9):976 82.

Diabetic Neuropathy
 C H A P T E R

13
Health economics of diabetic foot
disease: costs of diabetic neuropathy and
diabetic foot
Beatriz Rodrı́guez-Sánchez1 and Alan Sinclair2
1
Department of Applied Economics, Public Economics and Political Economy, Faculty of Law,
University Complutense of Madrid, Madrid, Spain 2Foundation for Diabetes Research in Older People,
King’s College, London, United Kingdom

Background years [1]. These figures have their direct transla-

tion in terms of economic impact, since it is esti-
Chronic diseases are the main cause of mortal- mated that 80% of resources are devoted to the
ity and morbidity among older people. In fact, it treatment and prevention of chronic diseases,
is estimated that 80% of the older population has with the distribution of spending being concen-
at least one chronic disease. The most function- trated in a small percentage of the population,
ally limiting conditions in the population over 65 mostly older people (approximately 5% of the
years of age are, among others, type 2 diabetes population consumes about 30% of the
mellitus (T2DM), cardiovascular disease, or resources). Particularly to the case of diabetes,
chronic obstructive pulmonary disease (COPD), developing countries spend nearly 40% of their
and some degenerative diseases, such as demen- healthcare expenditure on people with diabetes,
tia. Due to population aging, longevity, and an whereas in developed countries, such disease-
increase in unhealthy behaviors, which might related expenditure accounts for approximately
lead to additional health problems (e.g., high 12% 15% of national healthcare spending [2].
cholesterol, obesity, or hypertension), the preva- One of those chronic diseases that pose a
lence of older people with some chronic disease great burden on its sufferers is T2DM, which is
has increased markedly in recent decades. In especially relevant given the high prevalence
fact, in 2016, noncommunicable diseases, includ- of T2DM among total diabetes cases and the
ing the main chronic diseases, were shown to be effect of aging on its prevalence. Older people
responsible for approximately 82% of the disease represent around half of the people with diabe-
burden, measured in disability-adjusted life tes and diabetes prevalence reaches one in

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00017-7 211 © 2022 Elsevier Inc. All rights reserved.
212 13. Health economics of diabetic foot disease: costs of diabetic neuropathy and diabetic foot

every four older adults 65 years old and above if at particularly high risk, referral to multidisci-
[3]. Furthermore, diabetes is one of the largest plinary foot care teams [14,15].
factors increasing the risk of mortality, morbid- Early detection might also play a relevant
ity, and disability over the world and its eco- role with respect to the health-related quality of
nomic burden demands new ways to curb life (HRQoL) on people with diabetes addition-
diabetes healthcare expenditure [4]. ally suffering of DPN. In fact, the negative asso-
Diabetic peripheral neuropathy (DPN) is a ciation between DPN and HRQoL has been
frequent long-term diabetes-related complica- supported by several studies [16,17], highlight-
tion, which consists of nerve damage in hands ing the relevance of such relationship in case of
and feet and it may also affect other internal painful DPN. Moreover, it has been supported
organs such as heart and bladder [2,5]. that quality of life in people with current ulcers
Similarly, DPN affects up to 50% of people and for those with major amputations is lower
with diabetes [6,7], with chronic painful neu- than for individuals with other long-term dis-
ropathy affecting up to 26%, leading to an eases such as COPD or renal disease [13].
increasing risk of foot ulceration and subse- In addition to causing pain and morbidity,
quent amputation [8]. More than one in two- foot lesions among people with diabetes have
foot ulcers is expected to become infected, substantial economic consequences. The cost of
raising the risk of hospitalization, amputation, diabetic foot lesions is, in fact, affected by sev-
and mortality [9]. A diabetic foot ulcer (DFU) eral interventions such as foot ulcers prevention
precedes 85% of nontraumatic lower limb programs, management strategies to heal ulcers,
amputations [10]. While diabetes itself doubles which shorten wound healing time and prevent
the rate of mortality by all causes as compared amputation, and by the appropriate manage-
to people without diabetes, foot ulceration ment and care for disability after amputations
imposes an even higher morbidity [11]. [18]. The following two sections aim to provide
Due to such issues related to diabetes’ compli- with a general picture that diabetic foot poses
cations, improving screening and earlier identifi- on national healthcare budgets, especially in
cation of patients at risk of developing the DPN European countries (section entitled “Cost-of-ill-
might offer an excellent opportunity for patients ness studies on diabetic foot disease”), but also
with diabetes. First of all, the majority of people to facilitate the readers with some interventions
with diabetes might not be aware of suffering that have been assessed in the existing literature
DPN since around 50% 85% of cases are asymp- and pursue to improve diabetic foot care, pre-
tomatic [12]. Moreover, the current practice vention, and diagnosis (section named “Cost-
barely focuses on early detection of DPN, maybe effectiveness and cost-utility of diabetic foot
due to complications related to the diagnosis. disease-related interventions”).
Furthermore, if, due to lack of early diagnosis,
people with diabetes and at high risk of develop-
ing neuropathy first present with an ulcer, the Cost-of-illness studies on diabetic foot
annual cost of treating these patients would be in disease
the range of additional d29 116 million in case
of the United Kingdom [13]. Furthermore, Having been widely used in the existing lit-
screening would also imply behavioral changes erature of health economics, the aim of cost-of-
to reduce the risk of unperceived trauma and illness studies is to provide with an estimation
identify those patients who should undergo of the economic burden that a certain disease
more intense intervention including improved poses on the society as a whole [19]. The
glycemic, blood pressure, and lipid control and, advantages of cost-of-illness studies could be

Diabetic Neuropathy
 Cost-of-illness studies on diabetic foot disease 213
summarized in three points: (1) they can serve with diabetes vs h35,858 if nondiabetic),
as an argument to inform policies regarding a whereas they are substantially different after 3
specific disease, as well as its related complica- years since amputation (h115,676 if the person
tions [19]; (2) they might be helpful in identify- had diabetes vs h92,862 if not) [25], pointing
ing target populations that may be subject to toward the relevance of the severity of the
specific health problems or policies [20]; and prognosis of DFUs.
(3) the results derived from this type of studies When such higher healthcare use is trans-
could be used to determine the efficacy of any lated into monetary terms, even though huge
health intervention aimed at reducing or elimi- heterogeneity of foot complications and geo-
nating, if possible, the economic effects of a graphic variations existed, substantial costs
disease or condition [21]. and further healthcare burden for people with
As it has been mentioned in the introduction diabetes were found, regardless of the coun-
section, in addition to the poorer quality of life, tries analyzed, being amputation procedures
mortality, and morbidity associated with the greatest component contributing to higher
diabetes-related complications, diabetic foot healthcare costs. A recent review showed that
also has adverse consequences on healthcare the cost of amputation ranged between $35,000
system and its sustainability, being projected to and $45,000, per year, in the developed coun-
reach $15 22 billion annually by the year 2024 tries, highlighting the fact that the cost of
in the United States in case of wound care pro- amputation in the United States is generally
ducts [22]. The cost of care for people with dia- higher than in European countries [26].
betes and a foot ulcer poses a greater economic Another more recent review showed that the
burden compared with the management of a burden of DFUs in European countries was
person with diabetes but with no foot ulcera- estimated to reach $13,561 annually per person
tion. Economic disparities (i.e., insurance cov- with diabetes, containing both direct and indi-
erage, type of healthcare provider) may play rect costs [27], which became even higher if
an escalating role because the reimbursement amputation procedures were introduced. For
of foot ulcer treatments, especially when example, in Sweden, the healthcare expendi-
chronic, might not be covered by third-party tures incurred per each healed DFU without
payers. Healthcare use might endorse not only amputation were nearly $24,965, whereas in
the immediate costs related to foot ulcer, case of minor and major amputation surgeries,
chronic wounds, and eventual amputation pro- those costs increased to $47,518 and $42,858,
cedures related costs, such as inpatient costs, respectively [28]. In case of the English
surgery, drugs, and dressings, but also long- National Health Service (NHS), Guest et al.
term foot ulcer care costs. Once an ulcer occurs, showed, using data from 130 patients with a
higher costs are projected to be incurred over newly diagnosed DFU, that the average DFU
the following 2 3 years, compared with people care costs significantly varied depending on
with diabetes but without foot ulcers [23], disease severity: for example, the average cost
although the literature has showed that there over a 12-month period of a healed ulcer,
seemed to be a tendency toward the nonulcer unhealed ulcer, and amputated wound was,
group after 2 years too, but still higher [24]. respectively, d2140, d8800, and d16,900 [29].
However, if amputation surgery had to be Our review of cost-of-illness studies per-
made, healthcare costs do not differ that much formed in a European setting related to dia-
between people with diabetes and without in betic foot revealed the relevant economic
the short-term (healthcare costs 24 weeks after burden that diabetic foot presents for several
lower extremity amputation: h36,686 in people national health services, with even higher

Diabetic Neuropathy
214 13. Health economics of diabetic foot disease: costs of diabetic neuropathy and diabetic foot

figures in terms of percentage attributed to dia- interventions, it should be important to high-

betic foot care within the NHS healthcare light what an economic evaluation is and its
expenditure for England. The study performed purpose. An economic evaluation consists of
by Kerr et al. in 2014 [30] showed that the comparing at least two available alternatives
healthcare costs due to diabetic foot care repre- (option A vs option B) in terms of their costs
sented 0.4% of the total NHS healthcare bud- and expected benefits [19]. Although there are
get; however, only 5 years later, in 2019, the many more pros, the advantages of carrying out
same authors concluded that diabetic foot care a full economic evaluation could be summarized
related expenses had already doubled, repre- as follows: (1) economic evaluations minimize
senting 0.8% 0.9% of the NHS budget [31]. In the risk of excluding an important alternative by
case of the Danish healthcare system, wound assuming a systematic analysis of the available
care represented an even higher proportion of options; (2) they take into account the perspec-
the national healthcare budget, between 1.5% tive of the analysis (healthcare provider or finan-
and 2.4% [32], depending on the hospital cer, patient or society as a whole), and the
assessed. conclusions achieved may vary depending on
When healthcare costs per person with dia- the perspective applied [39]; (3) the quantifica-
betic foot problems, the studies identified pro- tion and comparison of the costs and health out-
vided with a wide range of figures, which comes associated with two different alternatives
varied from $14,288 per year in Turkey [33] to facilitate the estimation of the real cost of a pro-
$1727 [34] in case of Wales. Differences might gram or intervention (not only its “budget” but
be due to methodology issues, as in case of the also its opportunity cost with respect to the
Turkish estimation, the authors simulated the available alternatives); and (4) offer an orga-
data, but included lots of healthcare use com- nized consideration of a range of possible alter-
ponents (outpatient care, imaging costs, labora- natives and evidence of their possible health
tory tests, drugs, medical materials, hospital effects. Furthermore, it may imply a clear dis-
admission, complication costs, intervention, tinction between questions of fact and the
rehabilitation, and trainings), while the Welsh inevitable questions of social value, assuming a
authors used real data from 78,090 subjects change not so much in the decisions that are
with chronic wounds, but only considered pri- finally made but in how they are made [19,40].
mary and secondary care consultations. Closer Within economic evaluations, there are dif-
estimations were observed between the ferent types according to the way of evaluating
Turkish [33] and Spanish study [35] with the health effects, which may vary substan-
respect to hospitalization costs due to diabetic tially [19,41]. Compared to cost minimization
foot since both works estimated an average analysis in which the health result or benefit is
cost per hospital admission episode of around identical between alternatives, both in the cost-
h7500. effectiveness and cost-utility analyses, these
A synopsis of the studies evaluating the effects differ between the evaluated interven-
costs of diabetic foot is reported in Table 13.1. tion and the comparator. The difference
between these two types of studies focuses on
the outcome measure, being in the former
Cost-effectiveness and cost-utility of (cost-effectiveness analysis) common clinical
diabetic foot disease-related interventions units while in the second case (cost-utility anal-
ysis) they are the so-called quality-adjusted life
Before moving on with some examples on the years (QALYs). Finally, in cost-benefit analysis,
cost-effectiveness of diabetic foot related health outcomes are also measured in

Diabetic Neuropathy
 Cost-effectiveness and cost-utility of diabetic foot disease-related interventions 215
TABLE 13.1 Summary of cost-of-illness studies in diabetic foot disease in Europe from 2010 to 2020.
First author
name and
year of
publication Country Study population Costs included Results

Kerr, 2019 [31] England Patient-level datasets at a Healthcare costs (ulceration care, outpatient Total cost to the NHS: d837 962
national and local level and primary care costs, inpatient admissions) million (0.8% 0.9%) of the NHS
budget for England

Guest, 2018 United 130 patients with a newly Healthcare costs (specialist physicians, Average NHS cost of wound care:
[29] Kingdom diagnosed DFU in THIN podiatrists, nurses’ visits, pressure-offloading d7800 per DFU over 12 months
dataset device, amputation procedures)
Average NHS cost per healed ulcer:
d2140

Average NHS cost per unhealed ulcer:

d8800

Average NHS cost per amputated

wound: d16,900

Nieto-Gil, 2018 Spain 2700 hospital admissions from Healthcare costs (hospital admission costs, Average cost per hospitalization:
[35] the year 2009 13 including amputation procedures) h7633

Oksuz, 2016 Turkey Simulated data Healthcare costs (outpatient care, imaging Average annual per patient cost with a
[33] costs, laboratory tests, drugs, medical materials, diabetic foot ulcer: $14,287.70
hospital admission, complication costs,
intervention, rehabilitation, and trainings) Mean intervention, rehabilitation, and
trainings costs: $2291.7

Mean drugs costs: $2545.8

Mean hospital admission costs:

$7357.4

Philipps, 2016 Wales 78,090 subjects with chronic Healthcare costs (primary and secondary care Total cost of chronic wounds care:
[34] wounds consultations, including postoperative care) d328.8 million (5% of total expenditure
on NHS in Wales)

Average cost per patient: d1727

Ignatyeva, Russia Medical records from the Healthcare costs (hospital admission, surgery, Mean treatment cost was h3051 per
2015 [36] Diabetic Foot Department of ulcer treatment) patient with diabetic foot ulcer (per
the Endocrinology Scientific hospital episode)
Center

Søndergaard, Denmark 48 patients with a diabetic foot Healthcare costs (hospital admission due to Median cost per case: 133,867 DKK
2015 [37] ulcer surgery and readmission costs)

Kerr, 2014 [30] England National datasets Healthcare costs (hospitalization costs, Total cost of diabetic foot care: d580
ulceration care, inpatient care) million (0.6% of NHS expenditure)

Ulceration care: d307 million

Inpatient care: d219 million

Amputation care: d55 million

Gottrup, 2013 Denmark 830 patients with a wound Healthcare costs only (inpatient and outpatient Total annual costs: h232,458 3.6
[32] visits, dressing changes costs) million, depending on the hospital,
accounting for 1.5% 2.4% of the total
annual budgets

(Continued)

Diabetic Neuropathy
216 13. Health economics of diabetic foot disease: costs of diabetic neuropathy and diabetic foot

TABLE 13.1 (Continued)

First author
name and
year of
publication Country Study population Costs included Results

Hoffmann, Germany 444 patients with LEA Healthcare costs (hospitalization, rehabilitation, Costs 24 weeks after LEA: h36,686 in
2013 [25] outpatient care, outpatient drug prescriptions, people with diabetes vs h35,858 if
nonphysician services, durable medical nondiabetic
equipment, and long-term care)
Costs 3 years after LEA: were h115,676
if the person had diabetes vs h92,862 if
not

McInnes, 2012 United National Diabetes Audit 2011 Healthcare costs (hospital admission costs and Foot ulcers and amputations related
[38] Kingdom data diabetic foot complications-related costs) costs: d985,600,282

DFU, Diabetic foot ulcer; LEA, lower extremity amputations; NHS, National Health Service; THIN, The Health Improvement Network.

FIGURE 13.1 Type of diabetic

foot intervention assessed in eco-
nomic evaluations (2010 20).

monetary units. In spite of these types of eco- PubMed. Within those 48 economic evalua-
nomic evaluations, in this section we will focus tions, different types of interventions can be
on cost-effectiveness and cost-utility analyses found. The main types of intervention assessed
only, as they represent the main bulk of eco- were pharmacological treatments, such as
nomic evaluations performed in any therapeu- wound dressing or plasma, which were evalu-
tic area. ated in 42% of the economic evaluations identi-
Fig. 13.1 shows that a total of 48 economic fied (n 5 20), as well as nonpharmacological
evaluations on any intervention related to dia- treatments (n 5 10, 21%). In addition, Fig. 13.2
betic foot disease from 2010 up to May 2020 displays the differences in terms of geographi-
have been identified through a search in cal location, with more than half of the

Diabetic Neuropathy
 Cost-effectiveness and cost-utility of diabetic foot disease-related interventions 217
FIGURE 13.2 Geographical
context of diabetic foot economic
evaluations (2010 20).

economic evaluations performed in a US set- treatment would only be implemented in

ting (n 5 25), followed by European countries France [53], but not in Turkey [50]. However,
(n 5 12). two studies are not representative enough to
In order to have a more in-depth analysis of establish implications of the results obtained.
the economic evaluations that have been Another care intervention that implied health-
recently published, we have summarized in care technologies, such as telemonitoring, is
Table 13.2 the results obtained in the economic proved that telemonitoring was the dominant
evaluations performed in Europe during the alternative compared with usual care in
last 10 years. As it has been illustrated in Denmark [46], as lower costs than standard
Fig. 13.1, also in Europe, pharmacological treat- monitoring and similar amputation rates were
ments have been the most popular interven- reported.
tions assessed in relation to diabetic foot. Moreover, DFU prevention is crucial to
Platelet-rich plasma was found to be domi- reduce diabetic foot disease burden. Taking
nated in a Spanish context [42,51], although its into account that each ulcer episode not pre-
cost-effectiveness and eventual implementation vented costs about h10,000 when only consid-
depended on some methodological issues [42]. ering medical costs [54,55], such prevention
Moreover, different dressing types were also investments (i.e., interdisciplinary teams for
found to be cost-effective [44] or even domi- ulcer prevention, implementation of ulcer pre-
nant (led to better health outcomes and lower vention, and social awareness campaigns) can
costs) [45,47], as well as healing therapies [49], be cost-effective. With respect to the cost-
regardless of the advanced stage of the effectiveness of specialized staff, the literature
wounds. Three additional studies were identi- shows that the introduction of a multidisciplin-
fied in the literature review to assess the cost- ary foot protection team in hospitals, which
effectiveness of healthcare technologies, includ- consisted of a bi-weekly consultant-led includ-
ing vacuum-assisted closure (VAC) therapy ing vascular surgery, endocrinology, orthope-
and telemonitoring. In terms of the former, the dic surgery, podiatry, orthotics, and tissue
literature led to inconclusive results, as VAC viability, led to lower costs (overall saving of

Diabetic Neuropathy
218 13. Health economics of diabetic foot disease: costs of diabetic neuropathy and diabetic foot

TABLE 13.2 Summary of economic evaluations found performed in diabetic foot disease within Europe during the
last 10 years (2010 20).
Country Economic
and time evaluation
Intervention assessed Intervention type horizon type Conclusions

PRP added to usual care vs usual Pharmacological Spain, 5 CUA Compared with usual care, the PRP treatment with the
care [42] treatment years manual method was more effective and less costly
(dominant option)

PRP added to usual care vs usual Pharmacological Spain, 5 CUA Compared with usual care, the PRP treatment with the
care [39] treatment years commercial kit was more effective but also more costly,
turning out to be noncost-effective

Neuropad 1 10-g SWME vs 10-g Screening/diagnostic United CUA Neuropad together with the SWME was the dominant
SWME [43] Kingdom, strategy (higher QALY gains and lower costs)
3 years

Neuropad vs 10-g SWME [43] Screening/diagnostic United CUA The solely use of Neuropad led to higher costs and lower
Kingdom, health gains than the 10-g SWME
3 years

TLC-NOSF dressing vs a neutral Pharmacological Germany, CEA TLC-NOSF dressing’s cost-effectiveness was superior
dressing [44] treatment 100 weeks compared with the neutral wound dressing

Collagen-containing dressing plus Pharmacological United CUA Collagen-containing dressing plus standard care is a
standard care compared with treatment Kingdom, dominant strategy, as it improved outcomes for lower costs
standard care alone [45] 4 months

Telemonitoring vs standard Medical device Denmark, CEA Telemonitoring was dominant, as it led to lower costs than
monitoring [46] 6 months standard monitoring and similar amputation rates

SBG gel vs standard care [47] Pharmacological United CEA SBG gel is the dominant alternative, as it steered to more
treatment Kingdom, wounds healed (and faster) and lower costs
1 year

Fiberglass heel cast in addition to Nonpharmacological United CUA/CEA Fiberglass heel cast in addition to standard of care is
standard of care vs standard of care treatment Kingdom, dominated by usual care, as it led to higher costs and poorer
[48] 24 weeks health outcomes

Heberprot-P as an add-on therapy to Pharmacological Slovakia, CUA Heberprot-P in addition to usual care is not a cost-effective
good wound care vs usual care [49] treatment 10 years option

VAC treatments vs connector [50] Medical device Turkey, 2 CEA The authors found no significant difference in the cost-
years effectiveness of VAC vs Y-connector

Platelet-rich plasma vs standard of Pharmacological Spain CEA Platelet-rich plasma is cost-effective, leading to faster
care [51] treatment healing ulcers

Multidisciplinary foot protection Specialized staff Ireland, 4 CEA A multidisciplinary foot protection clinic should be
clinic vs standard of care [52] years implemented, as it led to lower costs and better health
outcomes (lower number of major amputations)

VAC therapy vs advanced wound Medical device France, 1 CUA/CEA VAC therapy was found to be the dominant alternative
care [53] year (higher QALY gains, more healed ulcers, and lower costs)

CEA, cost-effectiveness analysis; CUA, cost-utility analysis; PRP, platelet-rich plasma; QALY, quality-adjusted life year; SWME, Semmes-Weinstein Monofilament Examination;
TLC-NOSF, Lipido-Colloid Technology-Nano-Oligo Saccharide factor; SBG, soluble beta-glucan; VAC, vacuum-assisted closure.

h114,063 per year) and better health outcomes Finally, although it is well-known that
(five major amputations less) within an Irish improved screening and earlier identification
context after 3 years [52]. of patients at risk of developing the diabetic

Diabetic Neuropathy
 References 219
foot syndrome offer an excellent opportunity References
for patients with diabetes to make behavioral
[1] Vos T, Abajobir AA, Abate KH, Abbafati C, Abbas KM,
changes to reduce the risk of unperceived Abd-Allah F, et al. Global, regional, and national inci-
trauma and identify those patients who should dence, prevalence, and years lived with disability for
undergo more intense intervention [14,15], the 328 diseases and injuries for 195 countries, 1990 2016: a
literature review performed has identified only systematic analysis for the Global Burden of Disease
Study 2016. Lancet 2017;390(10100):1211 59.
one study in which a screening tool was
[2] NIDDKD, https://www.niddk.nih.gov/health-informa-
assessed [43]. Neuropad, a simple test of sudo- tion/diabetes/overview/preventing-problems/nerve-
motor function, is a device that allows indivi- damage-diabetic-neuropathies; 2018 [accessed 07.06.18].
duals to self-test, facilitating early detection of [3] Soriguer F, Goday A, Bosch-Comas A, Bordiú E, Calle-
DPN. The authors concluded that the joint use Pascual A, Carmena R, et al. Prevalence of diabetes
mellitus and impaired glucose regulation in Spain: the
of Neuropad and the 10-g Semmes-Weinstein
[email protected] Study. Diabetologia 2012;55:88 93.
Monofilament Examination (SWME) compared [4] de Lagasnerie G, Aguadé AS, Denis P, Fagot-Campagna
to the individual use of the latter reported A, Gastaldi-Menager C. The economic burden of diabetes
better health results (0.04 QALYs more over 3 to French national health insurance: a new cost-of-illness
years) and savings in costs (d1049 per person), method based on a combined medicalized and incremen-
tal approach. Eur J Health Econ 2018;19(2):189 201.
being the dominant (and preferred) strategy.
[5] Stratton IM, Adler AI, Neil HAW, et al. Association of
Thus the authors point out that the joint use of glycaemia with macrovascular and microvascular
Neuropad, as a test of sudomotor function, complications of Type 2 diabetes (UKPDS 35): pro-
and the SWME as tests for screening and diag- spective observational study. BMJ 2000;321:405 12.
nosis of diabetic neuropathy, is an optimal [6] Strachan MWJ, et al. Cognitive function, dementia and
type 2 diabetes mellitus in the elderly. Nat Rev
strategy compared to normal practice, suggest-
Endocrinol 2011;7:108 14.
ing its use taking into account the high preva- [7] Boulton AJM. Management of diabetic peripheral neu-
lence of such diabetic complication and the ropathy. Clin Diabetes 2005;23:9 15.
limitations existing in its early detection. [8] Ziegler D. Diabetic peripheral neuropathy in holt RIG.
In: Cockram CS, Flyvbjerg A, et al., editors. Textbook
of diabetes. 4th ed. Oxford: Wiley-Blackwell; 2010.
[9] Hobizal KB, Wukich DK. Diabetic foot infections: current
Conclusion concept review. Diabet Foot Ankle 2012;3(1):18409.
Available from: https://doi.org/10.3402/dfa v3i0.18409.
The burden that diabetic foot poses on [10] Mavrogenis AF, Megaloikonomos PD, Antoniadou T,
healthcare systems is undeniable, as the data Igoumenou VG, Panagopoulos GN, Dimopoulos L, et al.
Current concepts for the evaluation and management of
reported in this chapter have supported. The
diabetic foot ulcers. EFORT Open Rev 2018;3(9):513 25.
available projections on population aging and [11] Adler AI, Boyko EJ, Ahroni JH, Smith DG. Lower-
the increasing prevalence of diabetes mellitus, extremity amputation in diabetes. The independent
particularly in older people [56], provide further effects of peripheral vascular disease, sensory neurop-
evidence of the likely future impact of diabetes- athy, and foot ulcers. Diabetes Care 1999;22:1029 35.
[12] Prashanth RJV, Edmonds ME. Early recognition of
related foot disease. The associated pain, dis-
diabetic peripheral neuropathy and the need for one-
comfort as well as infective, neuropathic, and stop microvascular assessment. Lancet Diabetes
vascular complications have a dramatic nega- Endocrinol 2016;4(9):723 5.
tive impact on quality of life of the affected indi- [13] Kerr M. Diabetic foot care in England: an economic
vidual. A modern diabetes healthcare service study. London: Insight Health Economics for Diabetes
UK; 2017.
requires a prompt and effective mechanism for [14] Gæde P, Vedel P, Parving HH, Pedersen O. Intensified
detecting foot disease early and instituting cost- multifactorial intervention in patients with type 2 dia-
effective interventions if it wants to avoid rising betes mellitus and microalbuminuria: the Steno type 2
and unsustainable health and social care costs. randomised study. Lancet 1999;353(9153):617 22.

Diabetic Neuropathy
220 13. Health economics of diabetic foot disease: costs of diabetic neuropathy and diabetic foot

[15] Gæde P, Lund-Andersen H, Parving HH, Pedersen O. [30] Kerr M, Rayman G, Jeffcoate WJ. Cost of diabetic foot
Effect of a multifactorial intervention on mortality in disease to the National Health Service in England.
type 2 diabetes. N Engl J Med 2008;358(6):580 91. Diabet Med 2014;31(12):1498 504.
[16] Kim CH, Jeong SJ, Mok JO, Lee KY, Kim SR. Impact [31] Kerr M, Barron E, Chadwick P, Evans T, Kong WM,
on quality of life in peoples with painful diabetic Rayman G, et al. The cost of diabetic foot ulcers and
peripheral neuropathy. Endocr Abstr 2015;37:EP443. amputations to the National Health Service in
[17] Van Acker K, Bouhassira D, De Bacquer D;, Weiss S;, England. Diabet Med 2019;36(8):995 1002.
Matthys K;, Raemen H;, et al. Prevalence and impact [32] Gottrup F, Henneberg E, Trangbæk R, Bækmark N,
on quality of life of peripheral neuropathy with or Zøllner K, Sørensen J. Point prevalence of wounds
without neuropathic pain in type 1 and type 2 diabetic and cost impact in the acute and community setting in
patients attending hospital outpatients clinics. Denmark. J Wound Care 2013;22(8):413 22.
Diabetes Metab 2009;35(3):206 13. [33] Oksuz E, Malhan S, Sonmez B, Tekin RN. Cost of ill-
[18] Boulton AJ, Vileikyte L, Ragnarson-Tennvall G, ness among patients with diabetic foot ulcer in
Apelqvist J. The global burden of diabetic foot disease. Turkey. World J Diabetes 2016;7(18):462.
Lancet 2005;366(9498):1719 24. [34] Phillips CJ, Humphreys I, Fletcher J, Harding K,
[19] Drummond MF, Schulper MJ, Claxton K, Stoddart GL, Chamberlain G, Macey S. Estimating the costs associ-
Torrance GW. Methods for the economic evaluation of ated with the management of patients with chronic
health care programmes. 4th ed. New York: Oxford wounds using linked routine data. Int Wound J
University Press; 2015. 2016;13(6):1193 7.
[20] Carey K. Cost function estimates. Encyclopaedia of [35] Nieto-Gil P, Ortega-Avila AB, Pardo-Rios M, Cobo-
Health Economics. Elsevier; 2014. Najar M, Blasco-Garcia C, Gijon-Nogueron G.
[21] Larg A, Moss JR. Cost-of-illness studies: a guide to Hospitalisation cost of patients with diabetic foot
critical evaluation. Pharmacoeconomics 2011;29 ulcers in Valencia (Spain) in the period 2009 2013: a
(8):653 71. retrospective descriptive analysis. Int J Environ Res
[22] Sen CK. Human wounds and its burden: an updated Public Health 2018;15(9):1831.
compendium of estimates. Adv Wound Care (New [36] Ignatyeva VI, Severens JL, Ramos IC, Galstyan GR,
Rochelle) 2019;8(2):39 48. Avxentyeva MV. Costs of hospital stay in specialized
[23] Driver VR, Fabbi M, Lavery LA, Gibbons G. The costs diabetic foot department in Russia. Value Health
of diabetic foot: the economic case for the limb salvage Regional Issues 2015;7:80 6.
team. J Vasc Surg 2010;52(3):17S 22S. [37] Søndergaard LN, Christensen AB, Vinding AL, Kjær
[24] Ramsey SD, Newton K, Blough D, McCulloch DK, IL, Larsen P. Elevated costs and high one-year mortal-
Sandhu N, Reiber GE, et al. Incidence, outcomes, and ity in patients with diabetic foot ulcers after surgery.
cost of foot ulcers in patients with diabetes. Diabetes Dan Med J 2015;62(4):A5050.
care 1999;22(3):382 7. [38] McInnes AD. Diabetic foot disease in the United
[25] Hoffmann F, Claessen H, Morbach S, Waldeyer R, Kingdom: about time to put feet first. J Foot Ankle Res
Glaeske G, Icks A. Impact of diabetes on costs before 2012;5(1):26.
and after major lower extremity amputations in [39] Peña-Longobardo LM, Rodrı́guez-Sánchez B, Oliva-
Germany. J Diabetes Complicat 2013;27(5):467 72. Moreno J, Aranda-Reneo I, López-Bastida J. How rele-
[26] Petrakis I, Kyriopoulos IJ, Ginis A, Athanasakis K. vant are social costs in economic evaluations? The
Losing a foot vs losing a dollar; a systematic review case of Alzheimer’s disease. Eur J Health Econ 2019;20
of cost studies in diabetic foot complications. (8):1207 36.
Expert Rev Pharmacoecon Outcomes Res 2017;17 [40] Culyer AJ. The dictionary of health economics. 3rd ed.
(2):165 80. Cheltenham: Edward Elgar; 2014.
[27] Raghav A, Khan ZA, Labala RK, Ahmad J, Noor S, [41] Oliva Moreno J, González López-Valcárcel B, Trapero
Mishra BK. Financial burden of diabetic foot ulcers to Bertrán M, Hidalgo Vega A, del Llano Señarı́s JE.
world: a progressive topic to discuss always. Therap Economı́a de la Salud. Madrid: Ediciones Pirámide; 2018.
Adv Endocrinol Metab 2018;9(1):29 31. [42] Linertová R, del Pino-Sedeño T, Pérez LG, Aragón-
[28] Tennvall GR, Apelqvist J, Eneroth M. Costs of deep Sánchez J, Andia-Ortiz I, Trujillo-Martı́n M, et al.
foot infections in patients with diabetes mellitus. Cost-effectiveness of platelet-rich plasma for diabetic
Pharmacoeconomics 2000;18(3):225 38. foot ulcer in Spain. Int J Low. Extremity Wounds 2020;
[29] Guest JF, Fuller GW, Vowden P. Diabetic foot ulcer 1534734620903239.
management in clinical practice in the UK: costs and [43] Rodrı́guez-Sánchez B, Peña-Longobardo LM, Sinclair
outcomes. Int Wound J 2018;15(1):43 52. AJ. Cost-effectiveness analysis of the Neuropad device

Diabetic Neuropathy
 References 221
as a screening tool for early diabetic peripheral neu- [50] Uğurlar M, Sönmez MM, Armağan R, Eren OT.
ropathy. Eur J Health Econ 2019;1 15. Comparison of two different vacuum-assisted closure
[44] Lobmann R, Augustin M, Lawall H, Tigges W, (VAC) treatments of multiple chronic diabetic foot wounds
Potempa C, Thiem H, et al. Cost-effectiveness of TLC- in the same extremity. Foot Ankle Surg 2017;23(3):173 8.
sucrose octasulfate vs control dressings in the treat- [51] Cobos R, Aizpuru F, Parraza N, Anitua E, Orive G.
ment of diabetic foot ulcers. J Wound Care 2019;28 Effectiveness and efficiency of platelet rich plasma in
(12):808 16. the treatment of diabetic ulcers. Curr Pharm
[45] Guest JF, Singh H, Vowden P. Potential cost- Biotechnol 2015;16(7):630 4.
effectiveness of using a collagen-containing dressing [52] Nason GJ, Strapp H, Kiernan C, Moore K, Gibney J,
in managing diabetic foot ulcers in the UK. J Wound Feeley TM, et al. The cost utility of a multi-
Care 2018;27(3):136 44. disciplinary foot protection clinic (MDFPC) in an Irish
[46] Fasterholdt I, Gerstrøm M, Rasmussen BSB, hospital setting. Ir J Med Sci 2013;182(1):41 5.
Yderstræde KB, Kidholm K, Pedersen KM. Cost- [53] Whitehead SJ, Forest-Bendien VL, Richard JL, Halimi
effectiveness of telemonitoring of diabetic foot ulcer S, Ha Van G, Trueman P. Economic evaluation of
patients. Health Inform J 2018;24(3):245 58. Vacuum Assisted Closures Therapy for the treatment
[47] Cutting KF. The cost-effectiveness of a novel soluble of diabetic foot ulcers in France. Int Wound J 2011;8
beta-glucan gel. J Wound Care 2017;26(5):228 34. (1):22 32.
[48] Jeffcoate W, Game F, Turtle-Savage V, Musgrove A, [54] Posnett J, Gottrup F, Lundgren H, Saal G. The
Price P, Tan W, et al. Evaluation of the effectiveness resource impact of wounds on health-care providers
and cost-effectiveness of lightweight fibreglass heel in Europe. J Wound Care 2009;18(4):154.
casts in the management of ulcers of the heel in diabe- [55] Lindholm C, Searle R. Wound management for the
tes: a randomised controlled trial. Health Technol 21st century: combining effectiveness and efficiency.
Assess (Winchester, Engl.) 2017;21(34):1. Int Wound J 2016;13:5 15.
[49] Tesar T, Szilberhorn L, Nemeth B, Nagy B, Wawruch [56] Sinclair A, Saeedi P, Kaundal A, Karuranga S, Malanda
M, Kalo Z. Cost-utility analysis of Heberprot-P as an B, Williams R. Diabetes and global ageing among
add-on therapy to good wound care for patients in 65 99-year-old adults: findings from the International
Slovakia with advanced diabetic foot ulcer. Front Diabetes Federation Diabetes Atlas, 9th edition. Diabetes
Pharmacol 2017;8:946. Res Clin Pract 2020;162:108078.

Diabetic Neuropathy
 C H A P T E R

14
Diabetic foot
Frank Lee Bowling1,2,3, Keeley Jane Foley3 and Andrew J.M. Boulton1,2,4
1
Division of Endocrinology, Diabetes and Gastroenterology, Faculty of Medicine, Biology and Health,
University of Manchester, Manchester, United Kingdom 2Manchester Royal Infirmary, Manchester,
United Kingdom 3Vascular Surgery and Diabetes, Central Manchester University Hospital NHS
Foundation Trust, Manchester, United Kingdom 4Diabetes Research Institute, University of Miami,
Miami, FL, United States

Introduction diagnosed diabetes mellitus, usually accompa-

nied by neuropathy and/or PAD in the lower
Diabetic foot disease (DFD) is a major global extremity.
burden for patients and health care systems While the prevalence of DFD varies in dif-
and is one of the most serious complications of ferent regions of the world, the pathways for
diabetes mellitus [1]. It encompasses infection, ulceration are similar in most patients. A holis-
ulceration, and osseous destruction of the foot tic patient approach is recommended to iden-
of a person with diabetes. Frequently accompa- tify the epidemiology of DFD to successfully
nied with peripheral arterial disease (PAD), treat and implement a suitable management
and neuropathy, which play a central role in plan [3].
the disease [2].
The International Working Group on the
Diabetic Foot (IWGDF) publishes international, Epidemiology of diabetic foot problems
multidisciplinary, evidence-based guidelines to
inform health professionals all over the world Every hour, in England, someone older than
on the prevention and management of DFD [2]. 50 years of age has a foot amputation (below
DFD encompasses diabetic peripheral neu- the ankle) due to DFD, and every 2 hours
ropathy (DPN), foot ulceration, Charcot neu- someone loses their whole leg (above or below
roarthropathy (CN), peripheral vascular disease the knee). The current UK annual cost of lower
(PVD), which is a major etiological factor in dia- limb ulcer care is d4bn and predicted to rise to
betic foot lesions as neuroischemic ulcers pres- d7bn by 2022 [4]. The burden on the NHS of
ent a major challenge in management. Infection, patients with foot and leg ulcers is greater than
ulceration, or destruction of tissues of the obesity. One in three individuals with neuro-
foot of a person with currently or previously pathic DFD have a lower extremity amputation,

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00022-0 223 © 2022 Elsevier Inc. All rights reserved.
224 14. Diabetic foot

and one in six have an early demise [4]. The The range of rates of foot LEA in the diabetic
incidence of neuropathic foot ulcers is associ- population ranges from 1.2 to 362.9 per 100,000.
ated with more than seven times increase risk The National Diabetes Foot Care Audit (NDFA)
of mortality, compared to those who do not reports that the number of ulcer episodes sub-
develop a foot ulcer, despite similar age and mitted to the audit increased by 57% between
duration of diabetes [4]. 2016 17 and 2017 18 [4] (Fig. 14.1).
Annual incidence and prevalence of foot
ulceration and amputation in the United
Kingdom is not certain and is heavily debated, Risk assessment
due to anomalies with how amputations are
categorized, and the methodology used in the Lack of symptoms and evidence of diabetic
studies that report on incidence and prevalence. foot ulceration (DFU) do not exclude DFD.

FIGURE 14.1 Trends in lower extremity amputation incidence in European Union 15 1 countries 1990 2017 sets out a
global pattern of LEA [5].

Diabetic Neuropathy
 Diabetic peripheral neuropathy 225
There is no evidence on the effect of screening approximately 2%, and the lifetime incidence
for preventing DFU or its frequency, it is of neuropathy has been estimated to be 37%
based on expert opinion, since there is no pub- 45% for patients with type 2 diabetes and
lished evidence to support the required inter- 54% 59% for patients with type 1 diabetes [7].
vals. Implementing community diabetic foot Studies of nerve conduction tests performed at
screening is reported to cause greater good the time of diabetes diagnosis demonstrate
than harm [2]. National Institute for Clinical that neuropathy is already present in 10%
Excellence Guidelines Ng19 [6] set out the 18% of patients and subclinical neuropathy
current UK guidelines for the prevention, man- is also present [7]. It is one of the most com-
agement, and treatment of DFD, however the mon diabetic complications and leading cause
guidelines are focused on infection with lim- of disability, foot ulceration, and ultimately
ited guidance on PAD and foot ischemia amputation [7]. It is estimated to be present
detection, or diagnostic assessments relating to in at least 50% of people diagnosed with dia-
DFD [6]. betes. The incidence of DPN in the diabetic
It is crucial to identify an “at risk foot” by population is associated with other potentially
carrying out an initial assessment and proceed- modifiable cardiovascular risk factors, includ-
ing to in-depth diagnostics based upon the ing a raised triglyceride level, hypertension,
findings. Thorough history taking including obesity, and smoking [7].
diabetes diagnosis, comorbidities, current med- It is defined as asymmetrical and length-
ication, previous ulcer/lower extremity ampu- dependent sensorimotor polyneuropathy, attrib-
tation, claudication, loss of protective sensation uted to metabolic and micro vessel alterations
(LOPS), foot deformity, vascular status, neuro- due to long-standing hyperglycemia and meta-
pathy status, should be identified. The IWGDF bolic derangement [8]. Sensorimotor peripheral
2019 Risk Stratification System [2] can be used neuropathy can be asymptomatic, painful,
to categorize risk (Table 14.1). painless, or both. The characteristic clinical
manifestation of distal symmetrical sensorimo-
tor peripheral neuropathy is a feeling of burn-
Diabetic peripheral neuropathy ing, tingling, electric, sharp, and shooting pain
or a numbness and heaviness. Patients can
The yearly incidence of distal symmetric also be unaware they have DPN as they may
polyneuropathy in patients with diabetes is be asymptomatic [9]. It’s reported that 80% of

TABLE 14.1 The IWGDF 2019 risk stratification system and corresponding foot screening frequency [2].
Category Ulcer risk Characteristics Frequency

0 Very low No LOPS and no PAD Once a year

1 Low LOPS or PAD Once every 6 12 months
2 Moderate LOPS 1 PAD, or LOPS 1 foot deformity, or PAD 1 foot deformity Once every 3 6 months
3 High LOPS or PAD, and one or more of the following: Once every 1 3 months
history of a foot ulcer
a lower extremity amputation (minor or major)
end-stage renal disease

IWGDF, International Working Group Diabetic Foot Guidelines; LOPS, loss of protective sensation; PAD, peripheral arterial disease.

Diabetic Neuropathy
226 14. Diabetic foot

the DPN patients have depression and anxiety screening scores are subjective, and reliant on
disorders, due to the ongoing symptoms of the clinician’s interpretation [10].
DPN [10]. Autonomic dysfunction in the dia- Semmes Weinstein Monofilament Examination
betic lower extremity can manifest as anhidro- (10-g monofilament) [11] is currently the gold
sis, and commonly callus formation will occur standard for first-line screening of DPN in the
on weight-bearing areas of the foot [11]. feet. Other neurological screening tools include
The presence of moderately and severely the Ipswich Touch Test, whereby the clinician
increased albuminuria, that is, UACR of uses their index finger to apply light touch
$ 30 mg/g, is a strong predictor of DPN [12]. to the tips of the first, third, and fifth toes.
Long-term glycemic variability is also associ- Neuropathy is identified when detection of
ated with diabetics that have DPN. Further sensation fails at two or more sites (out of
studies are required to conclude albuminuria the total six) [11]. Stimulation tests are also
or glycemic variability for identifying indivi- used such as biothesiometer and the vibration
duals at increased risk for DPN [12]. However, perception threshold testing with the Vibratip,
DPN is a culmination of a complex interaction a tuning fork, or automated devices such as
of several causatively linked pathophysiologi- the Vibration Sensory Analyser VSA-3000, or
cal processes, many of which are not fully Neurometor [10]. These assessments are reliant
understood. Hyperglycemia and duration of on patients’ subjective response. They are
diabetes have an important role in DPN [13]. mainly used to determine loss of protective foot
It is important to determine risk factors sensation and risk of ulceration [10].
and control them at an early stage to prevent Nerve condition studies only assess large
serious consequences, such as foot ulcers, gan- nerve fibers. DPN involves both small and
grene, and amputation. Current studies suggest large nerve fibers, with small nerve fiber
that the risk factors include the duration of dia- involvement occurring at the early stage of the
betes, age, increased HbA1c, diabetic retino- disease. Small nerve fibers constitute 80% 91%
pathy, smoking, and a high body mass index. of peripheral nerve fibers, which control
Appropriate interventions to address risk fac- pain perception and autonomic function. There
tors can reduce ulcers by 60% and amputations has been development for the early detection
by 85% in those with high-risk diabetic neuro- of DPN using devices such as DPNCheck
pathy [13]. that assesses the sural nerve in 3 minutes,
with 95% sensitivity and a 71% specificity
(DPNCheckTM, 2020), and Neuropad that is a
Neurological assessment 10-minute test that measures sweat production.
Nerve conduction studies are the gold stan- While Neuropad reported to be highly sensitiv-
dard for diagnosis of DPN, however are time ity for small fiber neuropathy, it has low speci-
consuming, costly, and impractical to imple- ficity [14]. These devices are in their infancy of
ment in routine clinical care screening pro- experimental use and have high heterogeneity
grams. Measures used to screen for DPN in and participant selection bias cannot be dis-
clinical practice are rudimentary and do not counted [14]. Intraepidermal nerve fiber density
detect the disease until the late stages of its measurement from lower limb skin biopsy is
development [10]. considered the gold standard for the diagnosis
Using standardized clinical assessment scor- of small fiber neuropathy but is invasive and
ing such as the Michigan Neuropathy Screening not suitable for DPN screening [15] (Table 14.2).
Instrument, the Toronto Clinical Neuropathy Five simple clinical tests are considered use-
Score, and the UK neuropathy disability ful in the diagnosis of LOPS in the diabetic

Diabetic Neuropathy
 Peripheral neuropathy and autoimmune disease 227
TABLE 14.2 Key components of the diabetic foot TABLE 14.3 Neuropathy assessment of the foot [16].
examination [16].
10-g monofilament at 4 sites on each foot 1 1 of the
Inspection following:

Dermatologic Vibration using 128 Hz tuning fork

• skin status: color, thickness, dryness, cracking Pinprick sensation

• sweating
Ankle reflexes
• infection: check between toes for fungal infection
• ulceration Vibration perception threshold
• calluses/blistering: hemorrhage into callus?

Musculoskeletal
• deformity (e.g., claw toes, prominent metatarsal heads, protein kinase C pathways [17]. These abnor-
Charcot joint) malities trigger the release of proinflammatory
• muscle wasting (guttering between metatarsals) cytokines, accumulation of advanced glycation
Neurological assessment end products, and generation of reactive oxy-
10-g monofilament 1 1 of the following 4
gen changes of the vasa nervorum result in
neuroischemia [18]. Major international clinical
• vibration using 128-Hz tuning fork guidelines for the management of DPN recom-
• pinprick sensation
• ankle reflexes
mend several treatments. First-line therapies
• VPT include tricyclic antidepressants, serotonin
noradrenaline reuptake inhibitors, and anticon-
Vascular assessment
vulsants that act on calcium channels. Other
• foot pulses therapies include opioids and topical agents
• ABI, if indicated such as capsaicin and lidocaine [7]. All current
guidelines advise a personalized approach
with a low-dose start that is tailored to the
foot [16]. Any of the five tests could be used by
maximum response having the least side
clinicians to identify LOPS. Two of these tests
effects or adverse event [7].
should be regularly performed during screen-
ing assessments which are usually the 10-g
monofilament and one other test [16].
Peripheral neuropathy and autoimmune
One or more abnormal tests would suggest
disease
LOPS, while at least two normal tests (and no
abnormal test) would rule out LOPS [16]. The
Peripheral neuropathies may present in the
vibration assessment using a biothesiometer or
context of systemic vasculitis and other primary
similar instrument is a useful tool in the assess-
or secondary to autoimmune diseases. Vasculitis
ment of LOPS (Table 14.3).
affecting small arteries or large arterioles is likely
to produce a peripheral neuropathy [19]. Among
the autoimmune diseases, rheumatoid arthritis,
Treatment hepatitis B virus (HBV)-associated PAN, and
DPN is often inadequately treated, and the hepatitis C virus (HCV)-related cryoglobulinemic
role of improving glycemic control specifically vasculitis are the most frequent causes of periph-
in type 2 diabetes remains unclear, however it eral neuropathy [20].
has been reported that hyperglycemia leads to While multiple mononeuropathy occurs in
abnormalities in nerve polyol, hexosamine, and 35% 65% of the patients with vasculitic

Diabetic Neuropathy
228 14. Diabetic foot

neuropathy, these patients may also present

with a distal asymmetric polyneuropathy or,
more rarely, a distal symmetric polyneuro-
pathy. It is therefore important to establish
a correct diagnosis so appropriate treatment
can commence and etiologic characterization of
peripheral neuropathies is found to prevent
potential complications and to optimize the
therapeutic approaches. Considering the broad
etiologic and phenotypic diversity of periph-
eral neuropathies, the management of these
patients must be individualized to identify the
underlying etiology and multidisciplinary team-
work with a rheumatologist is advocated [20].
FIGURE 14.2 Radiograph—Charcot foot.

Charcot neuroarthropathy of the tarsal bones with slight medial subluxa-

tion of the navicular and medical cuneiform.
Charcot neuroarthropathy is linked with There is dorsal subluxation/dislocation of the
distal symmetrical somatic and autonomic neu- tarsometatarsal joints with associated overlying
ropathy, which results in a nonreducible foot proliferative bone formation. Appearances are
deformity, risk of chronic neuropathic foot keeping with Charcot’s arthropathy with gener-
ulcers, and major lower extremity amputation. alized soft tissue swelling.
The presenting symptoms of the early stages
include a warm swollen foot with or without
pain. Osseous problems can occur such as
Classification of Charcot
spontaneous fractures with associated tendon
neuroarthropathy
and ligament damage. The pathogenetic
mechanisms are unknown [21]. The widely used CN classification system of
The condition is often misdiagnosed in the Eichenholtz [22] which describes three stages
early stages due to lack of knowledge in seen on radiographs:
the nonspecialist foot and ankle community.
• Stage 1—osseous debris, fragmentation,
Patients often present at the advanced stages of
disruption, and dislocation seen of involved
the condition, which is extra difficult to man-
joints;
age. Osseous and joint destructive changes lead
• Stage 2—the stage of coalescence, sclerosis,
to a gross alteration of the architecture of the
absorption of fine debris, and fusion of most
foot, which can occur in the forefoot, midfoot,
large osseous fragments; and
hindfoot, and ankle regions with avulsion frac-
• Stage 3—the reconstruction and
tures effecting the calcaneus [21]. Patients with
reconstitution stage, sclerosis becomes less,
Charcot foot continue to mobilize due to lack of
the major fragments are rounded and there
sensory awareness and in the late stages of the
is attempt at reformation of joint
disease the complete midfoot collapses and a
architecture.
rocker bottom foot shape occurs [21] (Fig. 14.2).
There is significant pathology pertaining to the This classification system does not describe
tarsus/TMT joints. There is early disorganization the prodromal phase and misses the important

Diabetic Neuropathy
 Management of diabetic foot ulceration 229
earliest inflammatory phase. A modern approach SINBAD, Meggitt-Wagner and PEDIS scoring
is to identify Stage 0 which is the prodromal systems are used internationally by health care
state of the disease for early detection and treat- professionals. They have high reliability and
ment. Stage 0 is when a foot demonstrates, can be used by different clinicians, even those
swelling, warmth, and pain, redness, signs with varied experience in wound care [24].
representing inflammation, in the neuropathic In 2014 the Society for Vascular Surgery
patient. These signs and symptoms are precur- published the (WIfI) classification system for
sor to foot architecture breakdown, seen in the the clinical staging of chronic limb threatening
later stages [23] and requires modern imaging ischemia, to review amputation risk and need
modalities such as MRI scans [23]. for revascularization [25].
Wifi Classification is especially useful in
predicting the possibility of amputation for
Charcot neuroarthropathy treatment 1 year and has benefits to assist with the man-
There are no proven medical or pharmaco- agement of patients presented with a foot
logical approaches other than immobilization ulcer [26].
and offloading. The treatment for acute CN is SINBAD wound classification is advocated
offloading, usually with an irremovable below- by IWGDF [2], and the National Institute for
the-knee cast walker. Timing of surgical inter- Clinical Excellence Guidelines Ng19 [6]. The
vention is controversial due to a lack of SINBAD system is simple and quick to use [27]
evidence regarding whether this should be per- (Table 14.4).
formed in the acute or later phase of the dis-
ease [21]. Tendon transfers in the early stages,
exostectomies in the latter stages can offer Neuropathic and neuroischemic ulcers
reduction in bony prominences, thus reducing
the risk of ulcerative episodes. There is little Peripheral ischemia in conjunction with a
evidence to show that a surgically recon- diabetic foot ulcer is an independent risk factor
structed Charcot deformity functions better for amputation. While some patients may only
than a nonsurgically corrected deformity, and have mild ischemia, others can have profound
the reoccurrence of plantar foot ulceration is vascular insufficiency which can significantly
higher in people with Charcot foot than those impair healing of diabetic foot [28].
without [21].
TABLE 14.4 SINBAD [27].
S: Site (Forefoot 5 0; mid- and hindfoot 5 1)
Management of diabetic foot ulceration I: Ischemia (pedal blood flow intact [at least one palpable
pulse] 5 0; clinical evidence of reduced blood flow 5 1)
Wound classification
N: Neuropathy (protective sensation intact 5 0; protective
The need to classify, score, and describe sensation lost 5 1)
DFU is required for accurate clinical documen- B: Bacterial infection (none 5 0; present 5 1)
tation and vital communication between health
A: Area (ulcer area ,1 cm2 5 0; ulcer area .1 cm2 5 1)
care professionals, as well as recording impor-
tant outcomes [24]. It can also establish if D: Depth (ulcer confined to skin and subcutaneous
urgent vascular referral is required for critical tissue 5 0; ulcer reaching muscle, tendon or deeper 5 1)
limb ischemia. The widely used classification Less severe ulcer 5 SINBAD score of less than 3; Severe
systems; namely the University of Texas, ulcer 5 SINBAD score of 3 or more.

Diabetic Neuropathy
230 14. Diabetic foot

amputation. The first choice for plantar fore-

foot or midfoot ulcers in a people with diabetes
should be a nonremovable knee-high offload-
ing device [28]. In the event of patient intoler-
ance, or contraindications such as ischemia
or infection a removable knee-high device
could be used, or an ankle-high offloading
device [28]. Appropriately, fitting footwear
combined with felted foam can be considered
as the fourth-choice offloading treatment [28].
The IWGF, 2019 Guidelines provides detailed
evidence-based advice for correct offloading at
all stages and classification of DFU [2].
For nonplantar foot ulcers, removable ankle-
high offloading device, footwear modifications,
therapeutic footwear, toe spacers, orthoses,
heel casts can be used depending on the type
and location of the foot ulcer [28].

Surgical offloading
FIGURE 14.3 Neuroischemic toe which led to amputa- Interventions for metatarsal head and digital
tion of the digit. ulcers can be considered such as Achilles ten-
don lengthening (ATL), metatarsal head (MTH)
Fig. 14.3 Neuroischemic second digit with resection, osteotomy, arthroplasty, ostectomy,
dorsal abrasion fourth digit. Full-thickness tis- exostectomy, external fixation, flexor tendon
sue necrosis with pus over the nail plate and transfer or tenotomy, and tissue fillers such
adjacent soft tissue, high index of suspicion of as silicone or fat. These procedures aim to pro-
osteomyelitis second digit, fourth digit superfi- vide a more permanent offloading of elevated
cial soft tissue loss. mechanical stress; however, surgery comes
with increased risk [28].

Neuropathic and neuroischemic ulcer

treatment Antibiotic-resistant organisms
Debridement Foot infections are the main complication
Conservative debridement may be necessary in patients DFD. Biofilm formation is also an
using minimal sharp debridement, however important pathophysiology consideration in
debriding agents such as honey or larval ther- the development of diabetic foot ulcers, due to
apy should be considered dependent on vascu- the disease progression and chronic ulceration
lar status [29]. caused by biofilm [30]. The development of
antibiotic resistance makes wound healing dif-
Offloading ficult and the difficulty lies in distinguishing
This is arguably the most important aspect between infection and colonization in DFU [30].
of the treatment of neuropathic foot ulcers Bacteria present can have close interactions
that are at risk of infection, hospitalization, and between bacteria and biofilm. Consequently,

Diabetic Neuropathy
 Treatment of infection in diabetic foot disease 231
certain bacterial species that would be consid- slow to clinically improve, or if the patient has
ered nonpathogenic, could coaggregate in path- severe PAD [31]. If the infection has not resolved
ogenic biofilm and cause a chronic infection after 4 weeks the patient should be reevaluated
[30]. Biofilms contribute to delayed healing and further diagnostics carried out [31].
and are characterized by a complex microbiome
and a polymicrobial organization on the surface
of the wound. Further studies are required to
validate biofilm and develop potential thera-
Osteomyelitis
peutic agents [30]. Surgical specialists should be consulted in
the case of severe infections involving
bone, or complicated gangrene, necrotizing
Treatment of infection in diabetic foot infection, and abscesses [31]. Uncomplicated
disease forefoot osteomyelitis may be treated with anti-
biotic therapy, without surgical intervention.
Antibiotic therapy Diabetic foot osteomyelitis with concomitant
soft tissue infection requires an urgent surgical
Antibiotic selection for treating diabetic foot opinion [31].
infections should be based on the likely or Diabetic foot osteomyelitis should not be
proven causative pathogen(s) and their antibiotic treated for longer than 6 weeks. If there is
susceptibilities [31]. It is important to consider no improvement within 2 4 weeks bone cul-
the clinical severity of the infection, efficacy of ture should be carried out and surgical resec-
the antibiotic of choice, risk of adverse events, tion may be required [31]. There have been
potential damage to the commensal flora as well recent prospective controlled studies which
as the likelihood of drug interactions in patients have demonstrated that in carefully selected
that are polypharmacy [31]. patients, antibiotic therapy alone has been
effective and surgical resection of the infected
bone was not required [31]. Diabetic foot osteo-
Soft tissue infection
myelitis, without surgical intervention requires
Effective antibiotic agents for soft tissue 6 weeks of antibiotic therapy. Patients that
infection include penicillins, cephalosporins, have had bone resection do not require more
carbapenems, metronidazole in combination than a week of antibiotic treatment. Parenteral
with clindamycin, linezolid, daptomycin, therapy should be the route of choice for the
fluoroquinolones, or vancomycin, but not tige- first 1 7 days followed by oral antibiotics of
cycline, First-line antibiotics ought to be well- an agent that elect specific antibiotic agents
established agents, and newer agents are often for treatment based on the likely or proven
held in reserve for antibiotic-resistant patho- causative pathogens, antibiotic susceptibilities,
gens [31]. severity of infection, and efficacy [31].
Parenteral route should be used until the Severe infection and diabetic foot complica-
infection clinically approves, and oral therapy tions may lead to pedal amputation as demon-
continued thereafter [31]. There are no pub- strated in Figs. 14.4 and 14.5.
lished studies to support topical therapy for Soft tissue balancing for hindfoot stabiliza-
diabetic foot infections [31]. tion. Plantar incision slight wound dehiscence
Patients with skin or soft tissue infection with clean granulation tissue, dorsal soft tissue
should be treated with antibiotics for 1 2 breakdown exposed lesion showing anterior
weeks, and up to 3 4 weeks if the infection is portion of talus.

Diabetic Neuropathy
232 14. Diabetic foot

diabetic foot reduces shear pressure, and

abnormally high plantar pressures. Wounds
with extensive bone and soft tissue involve-
ment require deeper and more aggressive
debridement. Drainage of purulent discharge
can significantly reduce healing time compared
with ulcers managed more conservatively [32].

Offloading
Total contact casting prevents prompt detec-
tion of infection and are contraindicated
with ischemic ulcers and osteomyelitis [28] The
IWGDF recommended various types of offload-
ing devices for infected diabetic foot ulcers [2].
Removal of pressure, restoration of perfu-
sion, eradication of infection, and local wound
care should be regularly addressed when treat-
ing diabetic foot ulcers [28].
IWGDF Guidelines 2019 for Interventions to
enhance healing of chronic foot ulcers in diabe-
tes are detailed and concise evidence-based
FIGURE 14.4 Destructive forefoot osteomyelitis with guidelines. The guidelines aim to assist with
extensive soft tissue loss leading to a midfoot amputation. the reduction of ulcer healing time and preven-
Dorsal to plantar flap with inline blood flow via the ante-
rior tibial artery.
tion of amputation [30].

References
[1] World Diabetes Foundation. Diabetes and foot care:
time to act; 2020. https://www.worlddiabetesfounda-
tion.org/files/diabetes-and-foot-care-time-act [accessed
15.11.20].
[2] IWGDF Guidelines - IWGDF guidelines; 2019. https://
iwgdfguidelines.org/guidelines/guidelines/ [accessed
15.11.20].
[3] Schaper N, Netten J, Apelqvist J, Bus S, Hinchliffe R,
Lipsky B. Practical Guidelines on the prevention and
management of diabetic foot disease (IWGDF 2019
update). Diabetes/Metabolism Res Rev 2020;36(S1).
Available from: https://doi.org/10.1002/dmrr.3266.
[4] NDFA. National Diabetes Foot Care Audit - NHS
Digital; 2020. https://digital.nhs.uk/data-and-informa-
FIGURE 14.5 Midfoot amputation is the same foot as tion/clinical-audits-and-registries/national-diabetes-
in Fig. 14.3 when healed. foot-care-audit [accessed 26.11.20].
[5] Hughes W, Goodall R, Salciccioli J, Marshall D, Davies A,
Debridement Shalhoub J, editors. Editor’s choice trends in lower
extremity amputation incidence in European Union
Regular debridement of hyperkeratotic (cal- 15 1 Countries 1990 2017. Eur J Vasc Endovasc Surg
lused) tissue on the plantar aspect of the 2020;60(4):602 12.

Diabetic Neuropathy
 References 233
[6] NICE NG19. Overview: diabetic foot problems: preven- targets. Nat Rev Neurol 2011;7(10):573 83. Available
tion and management. Guidance. NICE; 2020. https:// from: https://doi.org/10.1038/nrneurol.2011.137.
www.nice.org.uk/guidance/ng19 [accessed 17.11.20]. [18] Malik R, Newrick P, Sharma A, Jennings A, Ah-See A,
[7] Khdour M. Treatment of diabetic peripheral neuropa- Mayhew T, et al. Microangiopathy in human diabetic
thy: a review. J Pharm Pharmacol 2020;72(7):863 72. neuropathy: relationship between capillary abnormali-
Available from: https://doi.org/10.1111/jphp.13241. ties and the severity of neuropathy. Diabetologia
[8] Liu X, Xu Y, An M, Zeng Q. The risk factors for dia- 1989;32(2):92 102. Available from: https://doi.org/
betic peripheral neuropathy: a meta-analysis. PLOS 10.1007/bf00505180.
ONE 2019;14(2):e0212574. Available from: https://doi. [19] Rodrigues R, Branco M, Silva R, Ruano L, Fontão L,
org/10.1371/journal.pone.0212574. Lopes M, et al. Peripheral neuropathy in systemic vascu-
[9] Boulton A. The diabetic foot: a global view. Diabetes litis and other autoimmune diseases a report of five
Metab Res Rev 2000;16(S1):S2 5. Available from: cases emphasizing the importance of etiologic characteri-
https://doi.org/10.1002/1520-7560(200009/10)16:1 1 zation. Eneurologicalsci 2020;21:100272. Available from:
, ::aid-dmrr105 . 3.0.co;2-n. https://doi.org/10.1016/j.ensci.2020.100272.
[10] Selvarajah D, Kar D, Khunti K, Davies M, Scott A, [20] Blaes F. Diagnosis and therapeutic options for peri-
Walker J, et al. Diabetic peripheral neuropathy: pheral vasculitic neuropathy. Therapeutic Adv
advances in diagnosis and strategies for screening and Musculoskelet Dis 2015;7(2):45 55. Available from:
early intervention. Lancet Diabetes Endocrinol 2019;7 https://doi.org/10.1177/1759720x14566617.
(12):938 48. Available from: https://doi.org/10.1016/ [21] Schmidt B. Clinical insights into Charcot foot. Best
s2213-8587(19)30081-6. Pract Res Clin Rheumatol 2020;34(3):101563. Available
[11] Boulton A. The diabetic foot: grand overview, epide- from: https://doi.org/10.1016/j.berh.2020.101563.
miology and pathogenesis. Diabetes Metab Res Rev [22] Ehrlich G. Charcot joints. Arch Intern Med 1966;118
2008;24(S1):S3 6. Available from: https://doi.org/ (5):521.
10.1002/dmrr.833. [23] Ahluwalia R, Armstrong D, Petrova N, Papanas N,
[12] Pai Y, Lin C, Lin S, Lee I, Chang M. Reconfirmation of Edmonds M. Stage 0 Charcot neuroarthropathy in
newly discovered risk factors of diabetic peripheral neu- the diabetic foot: an emerging narrow window of
ropathy in patients with type 2 diabetes: a case-control opportunity? Int J Low Extremity Wounds 2021;
study. PLOS ONE 2019;14(7):e0220175. Available from: 153473462110118.
https://doi.org/10.1371/journal.pone.0220175. [24] Camilleri A, Gatt A, Formosa C. Inter-rater reliability
[13] Grisold A, Callaghan B, Feldman E. Mediators of dia- of four validated diabetic foot ulcer classification sys-
betic neuropathy. Curr Opin Endocrinol Diabetes tems. J Tissue Viability 2020;29(4):284 90. Available
Obes 2017;24(2):103 11. Available from: https://doi. from: https://doi.org/10.1016/j.jtv.2020.09.002.
org/10.1097/med.0000000000000320. [25] Conte M, Bradbury A, Kolh P, White J, Dick F,
[14] Manes C, Papanas N, Exiara T, Katsiki N, Fitridge R, et al. Global vascular guidelines on the
Papantoniou S, Kirlaki E, et al. The indicator test neu- management of chronic limb-threatening ischemia.
ropad in the Aassessment of small and overall nerve J Vasc Surg 2019;69(6):3S 125S. Available from:
fibre dysfunction in patients with type 2 diabetes: a https://doi.org/10.1016/j.jvs.2019.02.016 e40.
large multicentre study. Exp Clin Endocrinol Diabetes [26] Salib W, Desokey A, Wahba R, Zayed M. The ability
2014;122(03):195 9. Available from: https://doi.org/ of the Society for Vascular Surgery Wound, Ischemia
10.1055/s-0034-1367061. and Foot Infection (WIFI) Classification system to pre-
[15] Malik R, Veves A, Tesfaye S, Smith G, Cameron N, dict risk of amputation in patients with non-healing
Zochodne D, et al. Small fibre neuropathy: role in the ulcer during the first year atAin Shams University
diagnosis of diabetic sensorimotor polyneuropathy. Hospitals. QJM: An Int J Med 2020;113(Supplement_1).
Diabetes Metab Res Rev 2011;27(7):678 84. Available Available from: https://doi.org/10.1093/qjmed/
from: https://doi.org/10.1002/dmrr.1222. hcaa050.059.
[16] Boulton A, Armstrong D, Albert S, Frykberg R, [27] Ince P, Abbas Z, Lutale J, Basit A, Ali S, Chohan F,
Hellman R, Kirkman M, et al. Comprehensive foot et al. Use of the SINBAD classification system and
examination and risk assessment: a report of the Task score in comparing outcome of foot ulcer management
Force of the Foot Care Interest Group of the American on three continents. Diabetes Care 2008;31(5):964 7.
Diabetes Association, with endorsement by the [28] Bus S, Armstrong D, Gooday C, Jarl G, Caravaggi C,
American Association of Clinical Endocrinologists. Viswanathan V, et al. Guidelines on offloading foot
Diabetes Care 2008;31(8):1679 85. ulcers in persons with diabetes (IWGDF 2019 update).
[17] Vincent A, Callaghan B, Smith A, Feldman E. Diabetic Diabetes Metab Res Rev 2020;36(S1). Available from:
neuropathy: cellular mechanisms as therapeutic https://doi.org/10.1002/dmrr.3274.

Diabetic Neuropathy
234 14. Diabetic foot

[29] Conte M, Mills J, Bradbury A, White J. Implementing diabetes (IWGDF 2019 update). Diabetes Metab Res
global chronic limb-threatening ischemia guidelines in Rev 2020;36(S1). Available from: https://doi.org/
clinical practice: utility of the Society for Vascular 10.1002/dmrr.3280.
Surgery Threatened Limb Classification System (WIfI). [32] Rayman G, Vas P, Dhatariya K, Driver V, Hartemann A,
J Vasc Surg 2020;72(4):1451 2. Available from: Londahl M, et al. Guidelines on use of interventions
https://doi.org/10.1016/j.jvs.2020.06.049. to enhance healing of chronic foot ulcers in diabetes
[30] Pouget C, Dunyach-Remy C, Pantel A, Schuldiner S, (IWGDF 2019 update). Diabetes Metab Res Rev
Sotto A, Lavigne J. Biofilms in diabetic foot ulcers: 2020;36(S1). Available from: https://doi.org/10.1002/
significance and clinical relevance. Microorganisms dmrr.3283.
2020;8(10):1580. Available from: https://doi.org/10.
3390/microorganisms8101580.
[31] Lipsky B, Senneville É, Abbas Z, Aragón-Sánchez J,
Diggle M, Embil J, et al. Guidelines on the diagnosis
and treatment of foot infection in persons with

Diabetic Neuropathy
 C H A P T E R

15
Management of diabetic foot disease
Prashanth R.J. Vas1,2 and Venu Kavarthapu1,3
1
Diabetes Foot Clinic, King’s College Hospital, London, United Kingdom 2King’s Health Partners’
Institute of Diabetes, Endocrinology and Obesity, London, United Kingdom 3Department of
Orthopaedics, King’s College Hospital, London, United Kingdom

Introduction as loss of earnings effects of employee absentee-

ism and the impact on careers and family. The
The unrelenting increase in the global preva- reported rate of diabetes-related foot complica-
lence in diabetes mellitus has meant that the tions in low- and middle-income countries is
burden of its associated complications is higher, occurring at a relatively younger age [7].
expected to increase significantly over the next Nonetheless, many recent improvements and
few decades. Diabetic foot disease (DFD) is per- innovations have led to a significant improve-
haps the most feared of all such complications ment in the outlook for the individual with dia-
leading to a high degree of individual and soci- betic foot. In this chapter we discuss the
etal burden [1]. The rate of lower extremity multidisciplinary management of the DFD, focus-
amputation (LEA) continues to rise in many ing on diabetic foot ulceration (DFU) and Charcot
countries [2,3]—for example, in the United neuroarthropathy (CN). We explore the contem-
Kingdom, 169 of such procedures are per- porary principles of care including the evolving
formed every week. The mortality risk is sub- role of surgical specialties within the diabetic
stantial, with the 5-year mortality from DFD foot.
being worse than many cancers [4]. The eco-
nomic budern is also enormous—the estimated
direct costs of diabetic foot care in England was
equivalent to d1 in every d140 spent by the Management of diabetic foot ulceration
National Health Service in 2014 15 [5]. Incident
new DFD could add between $11,710 and DFU has a lifetime incidence between 19%
$16,833 incremental burden to an individual’s and 35% in individuals with diabetes. It is
annual healthcare expenditure in the United defined as skin breakdown of the foot (at the
States, doubling the cost of delivering diabetes level or below of the malleoli) penetrating the
care [6]. There are considerable invisible indirect dermal layer as a minimum. In the western
costs which are more difficult to estimate such hemisphere, the annual incidence of DFU is

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00004-9 235 © 2022 Elsevier Inc. All rights reserved.
236 15. Management of diabetic foot disease

estimated to be around 2%, although rates as with the role being often handled by physicians
high as 8% in other geographical locations have or nursing personnel with wound care expertise
been reported [8,9]. Prevalence rates of DFU are in many countries.
between 5% and 13% depending on the popula-
tions and health systems evaluated. The triad of
loss of protective sensation (LOPS), ischemia, Management from ulceration to healing
and foot deformities come together in DFU, and
Responsiveness of healthcare structure:
the clinical course is often complicated by infec-
time to first expert assessment
tion. Therefore, the key principles of DFU care
are centered around prompt recognition of Timing of referral is crucial. Longer time to
ulceration, timely recognition of ischemia and first expert assessment is associated with worse
the provision of high-quality standard care DFU prognosis [16,17]. DFU that have experi-
(SC) (to include wound care, timely treatment enced a delay in referral are often graded as
of infection, revascularization where needed severe during their first expert review [18]. The
and the provision of quality offloading) pref- challenge to ensure individuals with new DFU
erably in a multidisciplinary setting. Equally, are prioritized and referred early is a global one,
concurrent optimization of the many medical including for health systems that are deemed
comorbidities that individuals with DFU expe- advanced. In one recent exploratory study across
rience is essential [10]. Europe, 48% of patients were referred longer
than 1 month after first reporting then to their
primary care team [19]. During the COVID-19
pandemic which strained healthcare resources
Multidisciplinary diabetic foot service worldwide, suboptimal DFU outcomes alongside
Modern DFU care is provided through a col- a delay in first expert review have been noted,
lective of healthcare professionals working in col- underscoring the importance of ensuring a
laboration for the benefit of the patient—termed robust pathway [20,21]. Patient factors may also
a “multidisciplinary” diabetic foot service contribute to a delay in expert assessment—those
(MDfS). This allows for a systematic, consistent, with LOPS may neglect the presence of DFU and
and coordinated delivery of care, often under a infection as they do not experience the pain and
single auspice. Significant improvement in DFU discomfort that would typically raise alarm.
healing rates and reduction in major amputation Guidance and toolkits to improve time to referral
rates have been universally reported with the are available [22,23], some with adaptations to
development of such services [11 13]. In addi- cope with the inexorable service reorganization
tion, improved mortality rates (by up to 20%) that has been mandated by the COVID-19 pan-
and reduction in the duration of hospitalization demic [24].
(up to 40%) and medical treatment costs have
been reported [14,15]. The components of the Classification of diabetic foot ulceration
MDfS team varies globally but as a minimum Ulcer classification and grading systems facili-
includes podiatry, vascular surgery, physicians tate the communication of ulcer severity and
with diabetes experience and a orthotist but serial progression between various members of
larger units will have orthopedics, plastics, infec- the MDfS, homogenize DFUs recruited into stud-
tious diseases, specialist musculoskeletal radiolo- ies and audits, but importantly, influence treat-
gists, and physiotherapy teams providing close ment strategies [25]. Numerous classifications
support. It is notable that many countries pres- and scoring systems with varying degrees of val-
ently do not offer training programs in podiatry idation exist [25,26]. The University of Texas

Diabetic Neuropathy
 Management of diabetic foot ulceration 237
classification (UTC) and the Wagner classifica- glycemic control. Imaging choices will be guided
tion systems are widely used in clinical practice by the clinical picture. Table 15.1 details the vari-
and research. Site, Ischemia, Neuropathy, ous blood, vascular, imaging, and microbiological
Bacterial Infection, Area and Depth (SINBAD) is assessments that should be considered at the first
utilized in the United Kingdom National patient visit to an MDfS. Comment should also be
Diabetes Foot Audit, which by virtue of its sim- made of current footwear, ambulatory status, and
plicity, allows for evaluation of high-volume if any mobility aids are used by the individual.
data between multiple institutions [25]. The
Wound, Infection, Ischemia (WIfI) facilitates the Standard care in the management of
assessment of ischemia and the likely benefit of diabetic foot ulceration
revascularization and has been shown to predict
SC in DFU is the algorithmic and protocolized
risk of LEA and DFU healing in cohorts [27]; but
provision of the best possible treatment supported
its use to prognosticate outcomes of individual
by an evidence-based and cost-effectiveness data,
DFU is not recommended.
individualized to the patients’ needs, to achieve
DFU healing/remission. The key SC spokes
Key assessments at first visit include wound control through regular debride-
ment and the judicious use of dressings, identifi-
At first assessment, a detailed history to under-
cation of ischemia and revascularization where
stand the context of the DFU is needed. This
needed, infection control, offloading, and meta-
should include estimating the duration of ulcera-
bolic control. Ensuring pain relief and continued
tion. In addition, a detailed medical, drug and
surveillance after healing to maintain remission
personal history should be elicited. Examination
are also increasingly considered part of SC
of the foot should include assessments for LOPS
(Fig. 15.1).
and foot perfusion. Ulcer characteristics should
then be determined including DFU site, size,
depth (superficial or probe to the bone) and infec- Wound control: debridement and dressings
tion status and graded using an ulcer classifica- The TIME framework (tissue debridement,
tion system. Blood tests are recommended to inflammation control, moisture balance, and
assess overall health, estimate level of systemic epithelialization of wound edges) underscores
infection (if ulcer appears infected) and current the key elements of wound control [28].

TABLE 15.1 Investigations to be considered at first patient visit to a multidisciplinary diabetic foot service (MDfS).

Tests at first DFU assessment in an MDfS

Hematology Full blood count, erythrocyte sedimentation rate (ESR)

Biochemistry Urea and electrolytes, liver functions, CRP, HbA1C, lipid profile

Radiology X-ray foot and able with weight bearing views. Ensure clear views of calcaneal available for hind foot DFU
Where needed, magnetic resonance imaging (MRI) foot or ultrasound assessment

Vascular lab One or more of Duplex waveforms or Ankle Brachial Pressure Index (ABPI) or Toe pressures or toe
brachial index (TBI) as minimum. Transcutaneous oxygen (TcPO2) if accessible
Full arterial Duplex if waveforms suggestive of arterial disease
Microbiology Tissue specimen or deep swab for culture (when tissue specimen is not possible)
Bone or pus for culture where possible

DFU, Diabetic foot ulceration.

Diabetic Neuropathy
238 15. Management of diabetic foot disease

FIGURE 15.1 The spokes of high-quality stan-

dard care diabetic foot care. DFU, Diabetic foot
Wound Control ulceration.
• Dressings
• Debridement
Vascular/perfus
ion status
Pain
Management • Assessment
• Revascula-
risaon

Diabec
Foot Offloading &
Maintaining Ulceraon Mechanical
remission
Control
• Educaon
• Conservave
• Surveillance
• Surgical

Diabetes care Infecon

& medical control
comorbidies • So ssue
control • Osteomyelis

Regular debridement of devitalized and a tissue disruptor [30]. In enzymatic debridement,

necrotic tissue promotes wound healing by gradual chemical liquefaction of the devitalized
activating repair and reducing the bacterial tissues is observed after topical application of
load. Scalpel or “sharp” debridement is consid- agent. Water-soluble proteinases such as collage-
ered as the gold standard, as it triggers the nase and papain are typical examples. The use of
wound healing process by swiftly changing the larvae-based debridement (maggot debridement)
surface of a chronic wound into an acute wound has been popular for many decades [31], being
[29] but provides the added advantage of reliev- particularly helpful when there is significant
ing pressure created by callus (Fig. 15.2). It can be necrotic tissue that is too painful for sharp or
undertaken in the clinic, or if extensive ulceration mechanical debridement. Autolytic debridement
is present, may also be performed in a surgical using hydrogels is a very slow process best suited
theater facility. It is very unusual for patients with for dry necrotic wounds in care facilities where
DFU to require an anesthetic for sharp debride- limited expertise is available to offer other forms
ment. Limitations of the technique include that it of debridement and the individuals themselves
can on occasions be painful, requires careful prac- are too frail to visit a hospital clinic. The recent
tice and not all practitioners are allowed to per- consensus statement from the International
form them. Working Group on the Diabetic Foot (IWGDF)
Other available methods include mechanical, noted that while there is clear consensus support-
enzymatic, or autolytic debridement techniques. ing debridement for wound cleansing, the superi-
Examples of mechanical debridement include ority of a single technique over another could not
hydrosurgical, using pressurized irrigation, deep be established [29]. It is therefore important to
cleansing with moistened gauze, brushing with consider factors such as pain, operator confi-
monofilament pads, and the use of ultrasound as dence, need for swift control of necrosis, and

Diabetic Neuropathy
 Management of diabetic foot ulceration 239

FIGURE 15.2 Before (A) and after (B) sharp debridement. Note the devitalized and hyperkeratotic skin (callus) has
been removed.

costs of debridement when considering the opti- additional hydrocolloid or hydrofiber layers)
mal approach. However, in the presence of acute which are absorbent and effective for heavily
severe infection, sharp surgical debridement is exuding wounds [33]. Hydrogel-based dres-
the mainstay. sings can moisturize dry, eschar heavy wound
The application of the dressing to a DFU is and consequently facilitate autolytic debride-
intended to relieve symptoms, protect the ment but their use in dry gangrene needs to be
wound and support healing [32]. A wide array carefully monitored for signs of progression to
of dressings are now available with some mod- wet gangrene at the interface of dead and
ern dressings capable of targeting specific healthy tissue. Dressings containing iodine and
aspects of healing. The wound bed in DFU is silver are regarded for their antiseptic poten-
notoriously dynamic with a progressive tem- tial, used typically when local infection is pres-
poral impact from infection, undulations in ent. All dressings should be changed
peri-wound and leg edema, impact of patient frequently, and occlusive dressings ideally
mobilization and adherence to the offloading avoided in infected DFUs. However, presently,
provided. Therefore, it is unlikely that a single there is no evidence to support the use of one
dressing type or formulation can support the dressing type over another as SC in DFU, a fact
ulcer over the entirety of its clinical course. that has been reiterated in the most recent con-
Nonadherent dressings such as saline-soaked sensus statements from the IWGDF and the
gauze are simple to use, inexpensive, and for American Podiatric Medical Association/
many years, have been used as SC in control Society for Vascular Medicine [29,34]. The only
arms of dressing trials. Highly exuding wound exception to this is the recommendation to con-
require dressings that support moisture control sider sucrose-octasulfate potassium salt
such alginate or foam dressings (with impregnated dressing as an adjunct to SC in

Diabetic Neuropathy
240 15. Management of diabetic foot disease

difficult-to-heal neuroischemic DFU without dialysis have higher prevalence of neuroische-

infection [29,35]. mia resulting in lower wound healing and
greater amputation rates compared to those
Ischemia and vascular control not receiving dialysis [38].
It is understood that ischemia complicates The decision to revascularize an individual
45% 60% of all DFU with prevalence rates with DFU and ischemia will depend on multi-
becoming evident over the last few decades. It ple factors such as severity of perfusion abnor-
is a clinical spectrum ranging from subtle malities identified, size of ulceration, extent of
peripheral arterial disease (PAD) which may tissue loss, and infection [36] with restoration
have negligible impact on wound healing to of pulsatile in-line flow to the affected region
more established chronic limb threatening being the primary goal. Those with milder per-
ischemia (CLTI). In itself, CLTI represents a fusion abnormalities can be observed closely
clinical continuum of increasing degrees of within the MDfS with SC provision and a deci-
ischemia which can delay wound healing and sion to revascularization made if the DFU dete-
increase risk of amputation [36]. Early charac- riorates or when progress is unpredictably
terization of limb perfusion advisable, if possi- slow. Recent advances in interventional radiol-
ble, on the first visit itself as the outcome will ogy alongside ultradistal/pedal bypass surgery
help guide the DFU management options. within an integrated pathway have facilitated
Neuroischemic DFUs are usually seen on the delivery of complex revascularization [39].
nonweight-bearing sites of the foot such as the
medial and lateral aspects of the feet or the tips Offloading
of the toes. These regions are often in contact The primary aim of adequate offloading is to
with footwear and experience shear forces dur- reduce abnormal pressure over the ulcer and redis-
ing mobilization but may also occur spontane- tribute it to the area surrounding the DFU to pro-
ously [37]. The callus formation surrounding mote healing. A wide array of offloading choices is
the ulcers is limited, thin, and shiny. This dis- available, with varying degrees of clinical effective-
tinguishes neuroischemic DFU from classical ness to support their use such as the total contact
neuropathic DFU which typically occur on cast (TCC), removable prefabricated cast walkers,
weight-bearing plantar surfaces overlying custom splints and extra-depth shoes, surgical
bony prominences and often display thick cal- boots, and half-shoes. In addition, the use of
lus overgrowth. The loss of protective sensa- wheelchairs, crutches, and scooters can also be
tion also means that individuals with diabetes recommended to support with achieving the bal-
often do not report intermittent claudication ance between desired offloading and maintaining
but rather present directly with tissue loss. quality of life. Among these, the nonremovable
Thus a common presentation of CLTI in diabe- TCC is recommended as the first choice for neuro-
tes is a foot with features of critical limb ische- pathic forefoot or midfoot plantar ulcerations
mia and areas dry gangrene indicating regions [23,40]. However, in those with neuroischemic
of direct cutaneous infarction (Fig. 15.3). DFU, TCC is a relative contraindication. If a TCC
Individuals with diabetes, with or without is not tolerated or contraindicated, options include
ulceration, can also develop acute limb ische- a removable knee-high or ankle-high boot.
mia (from emboli, acute trauma, graft thrombo- Padding the peri-ulcer skin with semicompressed
sis, or nonatherosclerotic conditions) which has adhesive felt combined with proper fitting foot-
a distinct clinical phenotype and needs imme- wear may be adequate in small DFUs. Those with
diate support from the vascular surgical spe- chronic or recurrent ulcerations resulting from
cialists. Importantly, individuals undergoing high-pressure points related to bony deformities

Diabetic Neuropathy
 Management of diabetic foot ulceration 241

FIGURE 15.3 The various presentation of diabetic foot ulcers. (A) Classical neuropathic diabetic foot ulceration on the
right foot with peri ulcer callus overgrowth over the bony prominence of first MTPJ. Note the forefoot deformities on both
feet and preulcerative lesions on the left plantar foot. (B) Lateral neuroischemic diabetic foot ulceration. Noted limited foot
deformity and extensive lateral necrosis. (C) Ischemic features in a young individual with longstanding type 1 diabetes.
Note the ischemic rubor and areas of early dry gangrene over dorsal second toe. The individual did not report significant
rest pain and ignored early symptoms and signs. MTPJ, metatarsophalangeal joint.

should be considered for surgical offloading. to asess the extent of the infection and confirm
Advocating complete nonweight bearing to indivi- the presence of osteomyelitis. Microbiological
duals with DFU is impractical and is not recom- specimens are critical to identify the
mended, except in specific situations (e.g., pathogen—tissue sampling is encouraged
postmajor bony debridement or postorthopedic where such expertise exists; ideally bone speci-
reconstruction) and even then, for a limited period mens are encouraged for the confirmation of
only. Ensuring adherence to offloading and recom- ostemyelitis. In those with mild infection,
mendations on mobilization remains a challenge. debridement of surrounding callus, slough and
necrotic tissue may be sufficient, contributing
Infection control including osteomyelitis to reducing the microbial bioburden. The selec-
It is estimated that half of all DFUs become tion of an antibiotic agent will depend on the
infected during their lifetime. Diabetic foot likely pathogen (or proven by culture) as well
infection (DFI) can lead to delayed healing, as allergies, drug interactions, and the cost. In
greater odds of hospitalization [41,42]. The most European countries gram-positive organ-
severity of infection can range from mild with isms such as Staphylococcus aureus and
features localized around the ulcer, moderate Streptococcus spp. are the most commonly
with more locally extensive features including encountered pathogens [44]. However, polymi-
extension to subdermal structures including crobial isolates, including a wide array of
tendon or bone, and severe infection present- gram-negative organisms such as Klebsiella
ing with systemic inflammatory response spp., Escherichia coli, and Pseudomonas spp. and
alongside limb-threatening features [43]. anaerobes are common in more chronic DFU
Imaging with plain radiography and magnetic [45]. In Africa and Asian regions, gram-
resosnace imaging (MRI) should be undertaken negative organisms may be presenting

Diabetic Neuropathy
242 15. Management of diabetic foot disease

FIGURE 15.4 Ascending

soft infection from a forefoot
plantar diabetic foot ulceration.
Note the cellulitis tracking into the
medial arch which is extending
into the dorsum and distal leg.

pathogen, sometimes declaring multidrug lacking but in the presence of severe soft tissue
resistance at the onset [46,47]. infection (with or without DFO) broad-spectrum
For soft tissue infection (Fig. 15.4), empirical intravenous antibiotics such as Co-amoxiclav
antibiotic therapy should target gram-positive 1.2 g thrice daily or tazobactam-piperacillin 4.5 g
organisms (as a minimum), however, DFUs that thrice daily along with a glycopeptide such as
are chronic or isolating antibiotic-resistant organ- teicoplanin or vancomycin (in those with sus-
isms may require bespoke antibiotic treatment pected or known MRSA infection), are the
plans form the outset. Co-amoxiclav 625 mg thrice empirical agents of choice—the glycopeptide is
daily is our empiric antibiotic of choice for mild typically discontinued once absence of MRSA is
infections in individuals without penicillin allergy. confirmed. Targeted antibiotic regimes are
The high prevalence of antibiotic-resistant strains guided by the microbiology results.
such as methicillin-resistant Staphylococcus aureus Diabetic foot osteomyelitis (DFO) is common,
(MRSA), vancomycin-resistant Enterococci (VRE), occurring in nearly one in five of moderately
and carbapenem-resistant organisms is emerging infected and one in two of severely infected
as a major challenge in determining antimicrobial DFU [50]. Severe DFO (acute or chronic) typi-
therapy [48,49]. The recent IWGDF consensus cally displays active soft tissue infective compo-
statement has recommended a treatment duration nent, however chronic DFO can exist with only
of 1 or 2 weeks for mild soft tissue infection, with mild extrinsic soft tissue features. Presence of a
longer doses to be considered in the presence of swollen toe or foot, history of long-standing
slow healing or PAD [43]. Antibiotic treatment ulceration or exacerbations of foot infection
should be combined with other principles of SC every time antibiotics are stopped should raise
including debridement. concerns of DFO. Findings include presence of
Infections exhibiting moderate or severe fea- a deep ulcer (or a sinus) which probes to bone.
tures are considered as potentially limb threaten- Plain radiography can reveal evidence of corti-
ing. These may require hospitalization and cal destruction or fragmentation (Fig. 15.5) but
possibly surgical debridement. Indications for has poor sensitivity, especially in early DFO
urgent surgical intervention, among others, and over bony resection areas. In comparison,
include rapidly spreading necrosis, deep abscess MRI has a high sensitivity ( . 90%) and specific-
with systemic features and gas in the soft tissues ity ( . 90%) for the radiological detection of
on X-ray examination. A consensus on the best DFO and can also provide insight to the extent
choice of antimicrobial agent or combination is of soft tissue involvement [51].

Diabetic Neuropathy
 Management of diabetic foot ulceration 243

FIGURE 15.5 Diabetic foot osteomyelitis (A). Note the swollen “sausage”-shaped right hallux. The corresponding
X-ray (B) demonstrates destruction of the distal phalanx tip (yellow arrow). There is significant soft tissue swelling extending
beyond the visible bony changes.

Management of DFO consists of bone in forefoot predominant, stable (low-grade)

debridement along with removal of any osteomyelitis without evidence of necrotizing
sequestrum, targeted antibiotic therapy guided infection, managed through an MDfS with
by bone specimen culture and ongoing inten- close podiatric support [55,56]. There is limited
sive DFU care. Antibiotic therapy should be consensus on the optimal mode of antibiotic
offered for 6 weeks but if there is no clinical delivery [52,57,58] and recent evidence sug-
improvement within 4 weeks, further inquiry gests that short courses of antibiotics may suf-
into slow progress and antimicrobial confirma- fice if a high-quality source control has been
tion is required [52]. The recently published achieved [59,60]. Surgical bone debridement
OVIVA study, while not specifically designed should be considered, as part of SC, in DFO
to evaluate response in DFO, has demonstrated with recurrent soft tissue infections, extensive
the safety and clinical effectiveness of early midfoot and hindfoot osteomyelitis or when
switch from intravenous to oral antibiotics infection control is not achieved as predicted
[53,54] with no differences in outcomes or there is a strong suspicion of nonviable
between the early oral switch group and the infected bone [61].
group receiving only intravenous therapy for
the study duration [53]. There is evidence to Diabetic foot attack
support the notion that primary antibiotic ther- Not infrequently, DFUs can present with rap-
apy only may be adequate, but this is mainly idly spreading infection and ascending cellulitis,

Diabetic Neuropathy
244 15. Management of diabetic foot disease

tissue necrosis, and systemic inflammatory syn- Advanced therapies in diabetic foot
drome. Such a presentation can potentially esca- ulceration
late into a limb threatening scenario without
timely intervention and is labeled as a “Diabetic Despite the wide-ranging components of SC,
Foot Attack” [62]. Without urgent (at times, the healing of foot ulcers remains a challenge.
immediate) intervention, it carries a significant In the United Kingdom National Diabetic Foot
risk of a major amputation and threat to life. Audit, approximately half of new DFUs healed
Management of a diabetic foot attack (DFA) fre- within the first 12 weeks [68] and among those
quently involves admission to hospital, initial still alive at 12 months, a quarter were
broad-spectrum empirical antibiotics, medical removed unhealed [69,70]. Provision of adju-
stabilization, and urgent surgical debridement of vant advanced wound healing therapies may
all infected tissue [63] followed by culture- therefore be attractive in difficult-to-heal DFUs
guided antibiotic therapy, high-quality wound or when wound progress has simply stalled
care and where necessary, revascularization [5,6]. [71]. The effectiveness of these interventions
Soft tissue defects occurring following radical can be variable, and their role is supplemen-
debridement can be managed with negative tary to SC. Indeed, it is important to ensure
pressure would therapy (NPWT) and early skin that any significant ischemia is corrected
grafting when granulation has reached the sur- beforehand (or at times, alongside) and ade-
face of the wound. Fig. 15.6 describes a phased quate offloading and infection control interven-
approach to the management of a DFA, which is tions are in place. A few of the important
best undertaken in a multidisciplinary setting. interventions include negative pressure wound
Such a protocolized approach has demonstrated therapy (NPWT), hyperbaric oxygen therapy
high rates of healing and limb salvage [64]. (HBOT), artificial skin substitutes, growth fac-
tors including platelets and their derived pro-
ducts, acellular dermal matrices and physical
Evaluating response to standard care therapies such as laser and electromagnetic
treatment applications [30]. While a detailed overview is
The percentage area reduction (PAR) in DFU out of scope of this chapter, a brief description
can be considered a useful guide of DFU progress is provided.
during SC provision. Individuals with .50%
reduction in PAR by 4 weeks are up to six times Negative pressure wound therapy
more likely to progress toward healing by 12 20 NPWT is used to enhance the healing of
weeks [65,66]. However, the evidence is based acute, chronic (converted to acute through
around classically neuropathic DFU. It may not debridement) or closed incisional wounds.
transfer easily to the prediction of healing in neu- The device applies negative pressure to the
roischemic DFU and therefore requires cautious ulcer/wound bed thereby providing an air-
interpretation. Measurement of biomarkers such tight seal using an electrically powered vac-
as levels of wound fluid matrix metalloproteinases uum pump. This leads to enhancement in
(MMPs) or tissue inhibitors of MMP or inflam- local blood flow, removal of excess exudate
matory cytokines such as interleukin-6 (IL-6) and stimulation of granulation. A Cochrane
remain very much research tools and are presently meta-analysis reported that NPWT increased
not geared toward application into frontline proportion of healed wounds compared with
practice [67]. moist wound therapy, a faster healing time
The King’s College Hospital SC approach to and reduction in amputation risk [72]. The
healing a DFU is described in Flowchart 15.1. evidence for NPWT in DFU healing is

Diabetic Neuropathy
 Management of diabetic foot ulceration 245

Phase 1 : Inial steps

Urgent Hospital admission,

Urgent orthopaedic/vascular/general surgery
Follow local sepsis management protocol: fluid consultaon
resuscitaon, acid-base control
Imaging (X-rays, MRI if feasible)
Empirical broad-spectrum anbiocs including MRSA
Surgical planning to debride all necroc ssue
cover

Phase 2: Urgent Infecon Control (for limb salvage)

Idenfy the proximal aspect of infecon spread, rule Plan ahead:

out necrosing process, Consider re-look in theatre in 48-72 hours
Radical debridement of all infected ssue including Arterial duplex (if not already obtained) or as per local
tendon and muscle down to healthy bleeding ssue. vascular assessment protocols
Copious lavage with normal saline Consider MRI if unsure the proximal infecve focus is cleared.

Phase 3: Intensive Wound Care

Bed rest and non-weight bearing on affected limb Connued Radiology and vascular assessments as
necessary
Daily bedside wound assessment, wound toilet and scalpel
debridement of necroc ssue Targeted anbiocs and Medical /Metabolic care
Revascularisaon as a priority (when indicated). Plascs input /Skin gra when ready
Negave Pressure Wound Therapy (NPWT) Safe discharge planning

Close follow-up in foot protecon service

If significant post DFA foot deformity - plan deformity correcon where appropriate
NB: High risk for DFU recurrence
Cardiovascular risk modificaon and glycaemic control

FIGURE 15.6 Management of a diabetic foot attack.

stronger in postsurgical diabetic foot standard pressure of NPWT application is

wounds [30] than for chronic ulcers offered 125 mmHg; evidence for the efficacy and
NPWT after receiving clinic-based debride- safety of lower suction pressures is not
ment or wound cleansing only [73]. The available.

Diabetic Neuropathy
 Diabec Foot Ulcer

Is there peripheral arterial disease (PAD)?

No Yes

Neuropathic DFU Neuroischaemic DFU or Ischaemic DFU

Assess for ulcer depth and presence of infecon. Involve vascular surgery urgently

Ulcer, no infecon Ulcer with cellulis or Ulcer with underlying

Ischaemia, but not crical Crical Limb ischaemia
so ssue infecon osteomyelis
and not known to be or already known to
non/slow- healing be non/slow-healing

Sharp debridement Sharp debridement Sharp debridement and wound care

and wound care and wound care Bone culture targeted anbiocs
TCC or appropriate DFU management as in Organise angioplasty
TCC* or Appropriate Assess if bone is viable!
Off-loading, Early input from orthopaedic surgery, Neuropathic DFU pathway or surgical bypass to
Off-loading, Adequate
Supporve medical infecon control, Podiatric or surgical removal of facilitate wound
care Supporve medical sequestrum where appropriate, healing.
care TCC or appropriate Off-loading,
Healing as predicted
Supporve medical care
(4-6 weeks)
DFU management as in
Neuropathic DFU
pathway
43
Yes No

Further assessment in Healing as predicted

Healing as predicted (4-6 weeks) 4-6 weeks, (4-6 weeks)

Healing as predicted?

Yes No No Yes

Yes No
Ensure appropriate bespoke Revisit Ischaemia,
footwear / orthocs provided, Revisit offloading technique,
Ensure CVD risk opmised, Revisit infecon- consider
Supporve medical care, surgical debridement of infected Ensure appropriate bespoke
ssue + May need connuaon footwear / orthocs provided,
Close supervision for recurrence of anbiocs or change in Vascular surveillance as per
once healed (follow up in DFC or anbiocs to reflect current local protocol for those with
community foot health clinic), pathogens,
angioplasty/bypass,
Ensure CVD risk opmised,
Paent educaon on ongoing Consider adjuvant advanced
Supporve medical care
foot care. therapies.

Discuss home environment,

Close supervision for recurrence
Assess compliance. once healed follow up in DFC or
community foot health clinic,

Paent educaon
 Management of diabetic foot ulceration 247
Oxygen therapies low as it was based on moderate quality evi-
Oxygen is the essential element in wound dence [29].
healing as it modulates many biological pro- Topical oxygen therapy (TOT) involves the
cesses that progress the wound from the prolifer- process of direct wound oxygenation with
ative stage to the remodeling stage of healing. either continuous delivery or intermittent
Chronic hypoxia can lead to inadequate collagen (cyclical) pressurized systems. The equipment
synthesis, downregulates angiogenesis and used is portable or can be easily installed at
reduces antimicrobial defense, contributing to home. One study reported nearly twofold
formation of a chronic, static wound [74]. higher rate of complete DFU closure at 12
Adjunctive treatment with HBOT was the weeks with continuously diffused topical oxy-
first available technology. It consists of indivi- gen therapy [80]. More recently, a study using
duals either alone or in groups breathing 100% cyclical pressurized topical wound oxygen
oxygen at 2.2 2.5 atmospheric pressure (atm) (TWO2) technology reported a four times
of oxygen in pressurized chambers daily for increased likelihood of DFU healing at 12
several weeks. Early studies indicated HBOT weeks compared to SC alone [81]. A topical
benefitted amputation avoidance [75]. The spray containing purified hemoglobin, which
Hyperbaric Oxygen Therapy in Diabetics with can bind oxygen from the local atmosphere
Chronic Foot Ulcers (HODFU) trial reported and subsequently diffuse it into the wound
higher healing rates of ischemic and neurois- bed has shown early promise [82]. While the
chemic DFU with HBOT compared to placebo technology can be arguably viewed as in its
over a 12-month period [76]. The two recent infancy, the results from the studies are
studies did not demonstrate a benefit with encouraging. The future adjunctive use of TOT
HBOT but had significant methodological lim- may appeal to both individuals with difficult-
itations and/or subjective endpoints [77,78]. to-heal DFU and healthcare providers.
One challenge in interpreting results of the
many trials of this technology is the significant Placental membrane-derived products
variation in study methodologies and heteroge- Placental membranes contain components
neity of clinical end points used. A recent necessary for stimulating wound healing such
meta-analysis of 11 studies concluded that as collagen-rich matrices, growth factors, cyto-
adjuvant HBOT lowered the major amputation kines in addition to viable endogenous cells
rate but not improve DFU healing. [79] such as mesenchymal stem cells (MSCs), neo-
Individuals with significant frailty and cardiac natal fibroblasts, and epithelial cells which
failure are unable to tolerate HBOT [78]. induce angiogenesis and potentiate wound
Furthermore, access of the technology is lim- repair [83]. Formulations available include
ited in most countries and its cost-effectiveness cryopreserved membranes (containing live
in DFU remains unproven. The IWGDF has cells) and dehydrated chorionic amniotic mem-
recommended consideration of HBOT as an brane products where live components are
adjunct in nonhealing ischemic or neurois- removed by a purification process. The advan-
chemic DFU already receiving best standard of tage with dehydrated products is that they do
care but the strength of recommendation was not require refrigeration and have a longer
L

FLOWCHART 15.1 Algorithm for management of diabetic foot ulceration (DFU) without acute limb threatening fea-
tures in a multidisciplinary: Authors’ perspective. Schematic overview of the approach to a nonacutely limb threatening
DFU. CVD, Cardiovascular disease; DFC, diabetic foot clinic, TCC, total contact cast.

Diabetic Neuropathy
248 15. Management of diabetic foot disease

shelf life. After debridement and wound Bioengineered skin and acellular dermal
cleansing, the product is applied directly on products
top of the ulcer base and covered with non- A number of skin substitutes are available.
permeable dressing. In their 2019 update, the They can be either cellular, containing either
IWGDF concluded that placental-derived pro- allogeneic or autologous cells, or acellular
ducts may lead to a higher wound closure rate matrices comprised of collagen subtypes and
in difficult-to-heal DFU when compared to chondritin [90,91]. They provide an extracellu-
standard of care only [30]. The technology is lar matrix scaffold to support fibroblast migra-
expensive and cost-effectiveness data is limited tion and angiogenesis. Studies have reported
[84,85]. apparent increased rate of DFU healing when
compared to SC only, but the studies have gen-
erally been at moderate to high risk of bias
Growth factors
[30]. The exact role of these products in the
Stalling of tissue regeneration in the inflam- healing algorithm is still being debated and
matory phase can lead to chronic wounds. A current guidance does not fully recommend
lack of growth factors has been implicated in their use [29].
chronic wounds [86] with several trials have
explored the clinical effectiveness of products Physical therapies
that contain one or more of such growth factors
Several physical therapies have claimed effi-
for stimulating wound healing. Commercially
cacy in the healing of DFU. Evidence of appar-
applications are now available to use as adjuvant
ent superiority in DFU healing is available for
wound therapy contain factors such as platelet-
devices using laser [92], ultrasound (mainly
derived growth factor (PDGF), fibroblast growth
used a tool for debridement to stimulate granu-
factor (FGF), epidermal growth factor (EGF), or a
lation) [93], therapeutic magnetic resonance
combination of FGF and EGF [87]. The use
[94], electrical stimulation [95], and photother-
recombinant platelet-derived growth factor (r-
apy/radiation, but in general, the quality of
PDGF, Becaplermin) is approved by the Food
evidence supporting their use is low [30]. Their
and Drug Agency (FDA) for the treatment of
use is currently not recommended in any con-
difficult-to-heal DFU without ischemia or infec-
sensus documents on DFU healing.
tion and not penetrating into deeper structures
of the foot [34]. However, Becaplermin is expen-
sive and efficacy data has been inconsistent
Metabolic control and treatment of
between studies and therefore has not been
widely recommended in international guidelines
associated comorbidities
[29,34]. A Cochrane review which examined A brief overview is provided in Table 15.2.
trials of 11 different growth factor formulations The risk of diabetic micro- and macrovascular
reported a positive signal, albeit weak, to sup- complications is greater in those with suboptimal
port superior healing of DFU compared to SC glucose control [96,97] which in turn increases
only but no effect on amputations [88]. Another the risk of developing DFD. Lower HbA1C
recent innovation is a patch dressing of autolo- (,7% or 53 mmol/mol) has been shown to
gous leukocytes, platelets, and fibrin lower the risk of amputation [98] but the impact
(Leucopatch 3C Patch, Reapplix Inc.) applied of short-term improvement in glycated hemoglo-
directly on the DFU demonstrating a higher heal- bin (HbA1C) on DFU healing has been less obvi-
ing rate compared to SC in difficult-to-heal DFU ous [99]. However glycemic control should be
including those with mild neuroischemia [89]. encouraged given the pleiotropic benefits of

Diabetic Neuropathy
 Management of diabetic foot ulceration 249
TABLE 15.2 A brief overview of key medical optimization interventions in diabetic foot disease (DFD).
Intervention Impact Pitfalls

Glycemic control Lower HbA1C has been shown to reduce risk of Wound healing benefit from improved
amputation. control has not been demonstrated.
Need to ensure hypoglycemia is avoided in
frail individuals.
Antiplatelets and Reduction in major adverse cardiovascular events Needs to be balanced against risk of
anticoagulants (MACE) such as stroke and coronary syndromes. bleeding.
Reduction in major adverse limb events (MALE) and Those at highest benefit risk are also at risk of
need for minor amputations. serious adverse events.
Direct Antihyperglycemic agents such as GLP-1 and SGLT2i Adverse events from medications including
cardiovascular have shown reduction in MACE events including risk of pancreatitis with GLP-1 analogs (low).
risk modification reduction in heart failure episodes and hospitalization. Risk of diabetic ketoacidosis from SGLT2i
Statins have established benefit profile on therapy. Muscle aches with statins.
cardiovascular outcomes. Treatment compliance.
Treatment of Allows for control of limb edema and moisture Use of diuretics associated with a higher risk
heart failure and around the DFU. of developing acute kidney injury and
leg edema Better wound care, and quality of life postural hypotension.

DFU, Diabetic foot ulceration; GLP-1, glucagon-like peptide-1; SGLT2i, sodium glucose cotransporter 2 inhibitor.

improved glucose control—from reducing the risk of amputation from canagliflozin is lower
potential for DFU infection, around periopera- than was initially estimated and with the FDA
tive/periprocedural outcomes and cardiovascu- now recommending removal of the boxed warn-
lar health [100]. ing [105].
DFD is associated with high mortality, at Evidence for use of statins in PAD is mainly
times greater than many cancers, mainly related derived from studies on coronary artery dis-
to cardiovascular disease. Therefore, pharmaco- ease and stroke outcomes [106]; direct exami-
therapy designed to improve the vascular out- nation of the impact on statins on limb-related
look of such individuals is attractive [101]. outcomes is limited. In the Fenofibrate
Cardiovascular trials of antihyperglycemic Intervention and Event Lowering in Diabetes
agents such as glucagon-like peptide-1 analogs (FIELD) study, reduction in amputation risk
(liraglutide, semaglutide, dulaglutide, albiglu- emerged after 1.5 years in the arm receiving
tide), and sodium glucose cotransporter 2 inhi- fenofibrate therapy [107]. The use of single
bitors (SGLT2i, empagliflozin, dapagliflozin, antiplatelet agent, either aspirin or clopidogrel,
canagliflozin) have indicated improved outcomes is currently recommended in diabetes indivi-
[102]. One liraglutide trial reported a lowered duals with PAD including in those receiving
risk of DFU-related amputations compared with revascularization [108]. Tigeclor has not shown
placebo in their posthoc analysis [103]. A higher superiority compared to clopidogrel alone in
risk of lesser-toe amputation was reported in the reduction of acute limb events [109]. The
two early studies of the SGLT2i canagliflozin but debate surrounding the correct use of dual
no such association has been reported for empa- antiplatelet agents (DAPT) to treat PAD con-
gliflozin or dapagliflozin [104]. However, a more tinues [110] including when revascularization
recent literature review has suggested that the of crural and pedal arteries is being considered.

Diabetic Neuropathy
250 15. Management of diabetic foot disease

The role of fixed low-dose direct oral anticoa- foot”) [118]. Approximately half of individuals
gulants, specifically rivaroxaban, is receiving recollect a history of trauma. Uncontrolled CN
attention for the secondary prevention of PAD. can result in considerable foot deformity lead-
The COMAPSS study, undertaken in those ing to an increased risk of ulceration, and
with stable atherosclerotic disease with previ- potentially amputation. A low index of suspi-
ous history of revascularization [111] and cion in all individuals with diabetes along with
VOYAGER-PAD study, undertaken in those imaging is key to early detection of CN. It is
having active revascularization [112], the com- important to remember that in cases presenting
bination of rivaroxaban 2.5 mg twice daily with early, typical radiographic signs such as sub-
aspirin demonstrated a significantly lower risk chondral fractures, joint dislocations, and sub-
of lower limb vascular events and need for luxations may be lacking [119]. In such
repeat revascularization when compared to instances, MRI or nuclear imaging help in the
aspirin alone. In addition, there was a reduc- diagnosis [119,120]; serial MRI may also help in
tion in all-cause cardiovascular morbidity and the follow-up period to assess treatment
mortality. There was however an approxi- response.
mately 40% higher rate of bleeding risk com- Medical management of CN primarily
pared to aspirin alone [113]. involves provision of rigid offloading and ade-
Reduction of lower limb is important for the quate rest of the affected region. Casting ther-
control of ulcer moisture and to improve healing apy is commonly considered as the “gold
[114]. If lower limb edema is present, compres- standard” [116,121]. Early initiation of high-
sion therapy should be considered. Presence of quality rigid offloading improves time to CN
PAD, especially of greater severity is a contrain- remission and reduction in deformity develop-
dication. Presence of cardiac failure should be ment [122,123]. The superiority of TCC in com-
managed in close liaison with the heart failure/ parison to removable systems is acknowledged
cardiology teams. The use of diuretics may be of by most experts, although no definitive trial
benefit, but their introduction should be cau- data are available to support these views [123].
tiously supervised. The effectiveness of agents to attenuate the
inflammatory milieu and osteoblastic activity has
been explored the management of CN. Most
Charcot neuroarthropathy “Charcot foot” widely explored has been the use of bisphospho-
nates (oral and intravenous) [124 127]. However,
CN is an uncommon but potentially devas- none of studies have reported an unequivocal
tating condition affecting the bones, joints, and positive effect on clinically relevant endpoints
soft tissues of the foot (or elsewhere in the such as reduction in casting time, fracture healing
limbs) affected by neuropathy [115,116]. While rate, or deformity development [128]. Likewise,
it was initially described by Charcot himself in the use of intranasal calcitonin, while reducing
an individual with neurosyphilis, diabetes is bone turnover markers, has not demonstrated
the most common cause today. Existence of a improvement in bone edema score, or time to
proinflammatory state leading to the activation remission [129]. A recent study reported that time
of abnormal osteoclastic activity underpins the to remission in CN was paradoxically prolonged
development of characteristic clinical and with the use of methylprednisolone as an antiin-
radiological appearances [117]. Presentation flammatory agent [127]. Therefore, it can be pres-
with a hot swollen foot is typical, with ently surmised that there is insufficient evidence
notable overlying dorsal erythema in lighter to recommend the use of any therapeutic agents
skinned individuals (“the red, hot swollen as first line or as adjuvant therapy in the

Diabetic Neuropathy
 Surgical management of the diabetic foot 251
management of CN. Future studies would benefit battery of investigations prior to surgery
from the recruitment of individuals with early- [43,62]. However, nonemergency presenta-
stage Charcot, as this group has been grossly tions would benefit from detailed investiga-
underrepresented in putative treatment trials of tions to assess the microbiological status by
CN. obtaining deep tissue or bone specimens for
Therefore, the current management of CN is microbiological sensitivities, distal vascular
based on early recognition, offering rigid off- status by performing appropriate studies and
loading with a TCC (or equivalent if facilities cross-sectional imaging to assess the extent of
for casting are not available) and close moni- infection spread [62].
toring in an MDfS. In the event of CN progres- The aim of surgical resection of DFI is to
sion despite best supportive therapy, an remove all the infected and necrotic tissues to
orthopedic referral should be considered. reduce the infection burden. During surgical
debridement, the edge and floor the ulcer is
excised, and all macroscopically visible
infected and necrotic tissue is removed. The
Surgical management of the diabetic foot
same principles are applicable to osteomyelitic
bone. Together with administration of targeted
Surgery is an important component in the
antibiotic therapy (guided by cultures of
management of DFD and a surgical member
intraoperative specimens) will allow for the
should be an integral part of an MDfS. Surgical
eradication of any residual infection. The
management is often required for the manage-
resulting open wound and tissue defect
ment of infection, chronic ulceration, vascular
is managed with NPWT. Emphasis should be
compromise, and deformities that fail nonoper-
maintained on ensuring any PAD is closely
ative measures [130]. All diabetic foot surgeons
monitored and revascularization considered
providing any component of surgical care are
in those with PAD with large tissue defects or
expected to be competent in performing com-
significantly abnormal vascular perfusion
mon and generic surgical procedures such as
indices [108].
infection and ulcer debridement, tendon bal-
ancing procedures, and minor and major lower
limb amputations. In addition, the orthopedic
or podiatric surgical member of the MDfS is Surgical management of a deformed
should be able to offer minor deformity correc- diabetic foot
tions and functional limb salvage of “foot at Forefoot deformity correction
risk” major diabetic foot deformities [131].
Lesser toe claw deformities that are partially
flexible often respond well to percutaneous
flexor tenotomy procedure [132]. These with
Surgical management of infected diabetic
rigid bone deformities usually require bone
foot ulcer and infection
osteotomies for correction. Marked plantar
Infected diabetic foot, with or without ulcer- bone prominence of metatarsal heads due to
ation, requires a protocol-based approach for the associated deformity respond to excision of
successful eradication of infection and soft tis- the prominent metatarsal head. Such deformi-
sue healing. Presentation with a “diabetic foot ties in the hallux are often managed with
attack” requires emergency surgical interven- deformity correction and fusion at the first
tion with radical debridement, and it may not metatarsophalangeal joint (MTPJ) to maintain
always be possible to complete ideally desirable load transfer ability across the first ray.

Diabetic Neuropathy
252 15. Management of diabetic foot disease

Exostectomy necessary in established deformities with signifi-

Minor diabetic foot deformities with local- cant joint contractures and bone damage.
ized bone prominence leading to abnormal Diabetic foot infections, particularly when sur-
plantar pressures and ulceration can be consid- gery is undertaken for infection control, can also
ered for surgical removal of the bony exostosis trigger postsurgical deformity formation. This is
[133]. When an ulcer is present overlying the typically noted in the forefoot, as the infective
bony deformity, the exostectomy is preceded episode frequently results in tendon and joint
by a thorough ulcer debridement. If the contractions [137]. The forefoot deformities asso-
amount of bone resection needed to normalize ciated with ulceration can be considered for toe
plantar pressures is excessive and lead to insta- deformity corrections. Lesser toe deformities due
bility of the foot in this area, this is often com- to marked flexor tendon contractures but no joint
bined with a reconstruction. dislocation or bone deformities, can be consid-
ered for percutaneous flexor tenotomy proce-
dures. Hallux-fixed deformities and the lesser toe
Reconstruction of deformed diabetic foot deformities with bone changes often require bone
A deformed Charcot or neuropathic foot that procedures, including excision arthroplasty, in
develops ulceration or is impending to ulcerate conjunction with tendon realignment surgery.
despite adequate offloading measures may be
considered for a surgical procedure. In the pres-
ence of actively infected ulcer or osteomyelitis, Future outlook
this is often performed as a two-stage proce-
dure—the first stage involved aggressive surgical The ongoing advances medical and surgical
debridement and administration of targeted anti- concepts for the management of diabetic foot
biotics to eradicate the infection, followed a few presage an improved outlook for this common
weeks later by the second stage of surgical recon- but condition [134,138,139]. Working within an
struction [134]. The principles of reconstruction MDfS environment allows specialists to provide
include wedge bone resections to correction the close supervision, bounce ideas, and concepts
deformity, followed by stabilization of the bone among each other and provide intensive diabetic
in the corrected position using internal or exter- foot care of a high quality for the benefit of the
nal fixation methods. individual. Future work exploring functional
Motor neuropathy leads to progressive muscle outcomes, cost-effectiveness, and improvement
imbalance and often affects the ankle and hindfoot in quality of life after introduction of advanced
regions. Relative weakness of peroneus brevis and wound care therapies or surgical offloading tech-
tibialis anterior, along with contraction of tibialis niques is urgently required. Equally, understand-
posterior and Achilles tendons result in static and ing the key characteristics (patient, care delivery
dynamic equinovarus deformities in the ankle, process, social and psychological) which can
hindfoot, and midfoot region [135]. Most deformi- prognosticate outcomes is important.
ties can be managed with offloading of the foot an
appropriate custom-built brace, but, as the neurop-
athy if often progressive, the deformities tend to Conclusion
continue to worsen over a period of time despite
these measures. Tendon balancing surgical proce- DFD and its two major constituents—DFUs
dures can provide good outcomes if performed and CN—represent a major challenge to clini-
during early stages of the disease [136], but bone cians and healthcare systems. They occur typi-
and joint reconstructive procedures are often cally in frail individuals with long-standing

Diabetic Neuropathy
 References 253
diabetes and many comorbidities. This under- management of diabetic foot complications. Diabetes
scores the need for early specialist referral and 2018;2018(2):1 20.
[10] Vas PRJ, Vainieri E, Patel N. Pharmacological treat-
care provision within a multidisciplinary service. ments for diabetic foot disease: current state and
The management principles require the control of future perspectives. Curr Pharm Des 2020.
wounds, vascular system, infection, and pressure [11] Edmonds M. The multidisciplinary diabetic-foot clinic.
to be achieved in parallel and close follow-up Int J Low Extrem Wounds 2008;7(2):66 7.
once healing is achieved to maintain remission. [12] Apelqvist J, Ragnarson-Tennvall G, Persson U, Larsson
J. Diabetic foot ulcers in a multidisciplinary setting. An
Advanced wound acre techniques and cardiovas- economic analysis of primary healing and healing with
cular risk modification offer an opportunity for amputation. J Intern Med 1994;235(5):463 71.
us to truly improve the future outlook and [13] Rubio JA, Aragon-Sanchez J, Jimenez S, Guadalix G,
reduce amputation rates and achieve functional Albarracin A, Salido C, et al. Reducing major lower
limb salvage. extremity amputations after the introduction of a mul-
tidisciplinary team for the diabetic foot. Int J Low
Extrem Wounds 2014;13(1):22 6.
[14] Joret MO, Osman K, Dean A, Cao C, van der Werf B,
References Bhamidipaty V. Multidisciplinary clinics reduce treat-
ment costs and improve patient outcomes in diabetic
[1] Edmonds M, Manu C, Vas P. The current burden of foot disease. J Vasc Surg 2019;70(3):806 14.
diabetic foot disease. J Clin Orthop Trauma
[15] Buggy A, Moore Z. The impact of the multidisciplin-
2021;17:88 93.
ary team in the management of individuals with dia-
[2] Hughes W, Goodall R, Salciccioli JD, Marshall DC,
betic foot ulcers: a systematic review. J Wound Care
Davies AH, Shalhoub J, editors. Editor’s Choice -
2017;26(6):324 39.
Trends in lower extremity amputation incidence in
[16] Ince P, Abbas ZG, Lutale JK, Basit A, Ali SM, Chohan F,
European Union 15 1 Countries 1990 2017. Eur J
et al. Use of the SINBAD classification system and score
Vasc Endovasc Surg 2020;60:602 12.
in comparing outcome of foot ulcer management on
[3] Harding JL, Andes LJ, Rolka DB, Imperatore G, Gregg
three continents. Diabetes Care 2008;31(5):964 7.
EW, Li Y, et al. National and state-level trends in non-
[17] Jeffcoate W, Young B. National Diabetic Foot Audit of
traumatic lower-extremity amputation among U.S.
England and Wales yields its first dividends. Diabet
Medicare beneficiaries with diabetes, 2000 2017.
Med 2016;33(11):1464 5.
Diabetes Care 2020;43(10):2453 9.
[4] Armstrong DG, Swerdlow MA, Armstrong AA, Conte [18] Jeffcoate WJ, Vileikyte L, Boyko EJ, Armstrong DG,
MS, Padula WV, Bus SA. Five year mortality and Boulton AJM. Current challenges and opportunities in
direct costs of care for people with diabetic foot com- the prevention and management of diabetic foot
plications are comparable to cancer. J Foot Ankle Res ulcers. Diabetes Care 2018;41(4):645 52.
2020;13(1):16. [19] Manu C, Lacopi E, Bouillet B, Vouillarmet J,
[5] Kerr M. Improving footcare for people with diabetes Ahluwalia R, Ludemann C, et al. Delayed referral of
and saving money:an economic study in England; patients with diabetic foot ulcers across Europe: pat-
2017. terns between primary care and specialised units.
[6] Rice JB, Desai U, Cummings AK, Birnbaum HG, J Wound Care 2018;27(3):186 92.
Skornicki M, Parsons NB. Burden of diabetic foot [20] Meloni M, Izzo V, Giurato L, Gandini R, Uccioli L.
ulcers for medicare and private insurers. Diabetes Management of diabetic persons with foot ulceration
Care 2014;37(3):651 8. during COVID-19 health care emergency: effectiveness
[7] Harding JL, Pavkov ME, Magliano DJ, Shaw JE, Gregg of a new triage pathway. Diabetes Res Clin Pract
EW. Global trends in diabetes complications: a review 2020;165:108245.
of current evidence. Diabetologia. 2019;62(1):3 16. [21] Musajee M, Zayed H, Thulasidasan N, Sayed M,
[8] Zhang P, Lu J, Jing Y, Tang S, Zhu D, Bi Y. Global epi- Francia F, Green M, et al. Impact of COVID-19 pan-
demiology of diabetic foot ulceration: a systematic demic on the outcomes in patients with critical limb
review and meta-analysis (dagger). Ann Med 2017;49 threatening ischaemia and diabetic foot infection. Ann
(2):106 16. Surg 2020.
[9] Boulton AJ, Armstrong DG, Kirsner R, Attinger CE, [22] Meloni M, Izzo V, Manu C, Ahluwalia R, Pedro J,
Lavery LA, Lipsky BA, et al. Diagnosis and Sánchez-Rı́os CL, et al. Fast-track pathway: an easy-to-

Diabetic Neuropathy
254 15. Management of diabetic foot disease

use tool to reduce delayed referral and amputations in Society for Vascular Medicine. J Vasc Surg 2016;63
diabetic patients with foot ulceration. Diabet Foot (2 Suppl.):3S 21S.
2019;22(2):39. [35] Edmonds M, Lazaro-Martinez JL, Alfayate-Garcia JM,
[23] Schaper NC, van Netten JJ, Apelqvist J, Bus SA, Martini J, Petit JM, Rayman G, et al. Sucrose octasulfate
Hinchliffe RJ, Lipsky BA, et al. Practical guidelines on dressing vs control dressing in patients with neu-
the prevention and management of diabetic foot dis- roischaemic diabetic foot ulcers (Explorer): an interna-
ease (IWGDF 2019 update). Diabetes Metab Res Rev tional, multicentre, double-blind, randomised, controlled
2020;36(Suppl. 1):e3266. trial. Lancet Diabetes Endocrinol 2018;6(3):186 96.
[24] Meloni M, Bouillet B, Ahluwalia R, Ludemann C, [36] Conte MS, Bradbury AW, Kolh P, White JV, Dick F,
Sanchez-Rios JP, Iacopi E, et al. Fast-track pathway for Fitridge R, et al. Global vascular guidelines on the
diabetic foot ulceration during COVID-19 crisis: a doc- management of chronic limb-threatening ischemia.
ument from International Diabetic Foot Care Group J Vasc Surg 2019;69(6S):3S 125S e40.
and D-Foot International. Diabetes Metab Res Rev [37] Ndip A, Jude EB. Emerging evidence for neurois-
2020;e3396. chemic diabetic foot ulcers: model of care and how to
[25] Monteiro-Soares M, Russell D, Boyko EJ, Jeffcoate W, adapt practice. Int J Low Extrem Wounds 2009;8
Mills JL, Morbach S, et al. Guidelines on the classifica- (2):82 94.
tion of diabetic foot ulcers (IWGDF 2019). Diabetes [38] Lavery LA, Lavery DC, Hunt NA, La Fontaine J, Ndip
Metab Res Rev 2020;36(Suppl. 1):e3273. A, Boulton AJ. Amputations and foot-related hospita-
[26] Game F. Classification of diabetic foot ulcers. Diabetes lisations disproportionately affect dialysis patients. Int
Metab Res Rev 2016;32(Suppl. 1):186 94. Wound J 2015;12(5):523 6.
[27] Zhan LX, Branco BC, Armstrong DG, Mills Sr. JL. The [39] Huang DY, Wilkins CJ, Evans DR, Ammar T, Deane C,
Society for Vascular Surgery lower extremity threat- Vas PR, et al. The diabetic foot: the importance of coordi-
ened limb classification system based on Wound, nated care. Semin Intervent Radiol 2014;31(4):307 12.
Ischemia, and foot Infection (WIfI) correlates with risk [40] Bus SA, Armstrong DG, Gooday C, Jarl G, Caravaggi
of major amputation and time to wound healing. C, Viswanathan V, et al. Guidelines on offloading foot
J Vasc Surg 2015;61(4):939 44. ulcers in persons with diabetes (IWGDF 2019 update).
[28] Harries RL, Bosanquet DC, Harding KG. Wound bed Diabetes Metab Res Rev 2020;36(Suppl 1):e3274.
preparation: TIME for an update. Int Wound J 2016;13 [41] Skrepnek GH, Mills Sr. JL, Lavery LA, Armstrong DG.
(Suppl. 3(S3)):8 14. Health care service and outcomes among an estimated
[29] Rayman G, Vas P, Dhatariya K, Driver V, Hartemann 6.7 million ambulatory care diabetic foot cases in the
A, Londahl M, et al. Guidelines on use of interven- U.S. Diabetes Care 2017;40(7):936 42.
tions to enhance healing of chronic foot ulcers in dia- [42] Ndosi M, Wright-Hughes A, Brown S, Backhouse M,
betes (IWGDF 2019 update). Diabetes Metab Res Rev Lipsky BA, Bhogal M, et al. Prognosis of the infected
2020;36(Suppl. 1):e3283. diabetic foot ulcer: a 12-month prospective observa-
[30] Vas P, Rayman G, Dhatariya K, Driver V, Hartemann tional study. Diabet Med 2018;35(1):78 88.
A, Londahl M, et al. Effectiveness of interventions to [43] Lipsky BA, Senneville E, Abbas ZG, Aragon-Sanchez
enhance healing of chronic foot ulcers in diabetes: a J, Diggle M, Embil JM, et al. Guidelines on the diagno-
systematic review. Diabetes Metab Res Rev 2020;36 sis and treatment of foot infection in persons with dia-
(Suppl. 1):e3284. betes (IWGDF 2019 update). Diabetes Metab Res Rev
[31] Sherman RA. Mechanisms of maggot-induced wound 2020;36(Suppl. 1):e3280.
healing: what do we know, and where do we go from [44] Spichler A, Hurwitz BL, Armstrong DG, Lipsky BA.
here? Evid Based Complement Altern Med Microbiology of diabetic foot infections: from Louis
2014;2014:592419. Pasteur to ‘crime scene investigation’. BMC Med 2015;13:2.
[32] Jeffcoate WJ, Harding KG. Diabetic foot ulcers. Lancet [45] Manas AB, Taori S, Ahluwalia R, et al. Admission
2003;361(9368):1545 51. time deep swab specimens compared with surgical
[33] Hilton JR, Williams DT, Beuker B, Miller DR, Harding bone sampling in hospitalized individuals with dia-
KG. Wound dressings in diabetic foot disease. Clin betic foot osteomyelitis and soft tissue infection. Int J
Infect Dis 2004;39(Suppl. 2):S100 3. Low Extrem Wounds 2021;20:300 8.
[34] Hingorani A, LaMuraglia GM, Henke P, Meissner [46] Hatipoglu M, Mutluoglu M, Turhan V, Uzun G,
MH, Loretz L, Zinszer KM, et al. The management of Lipsky BA, Turk-Day Study G, et al. Causative patho-
diabetic foot: a clinical practice guideline by the gens and antibiotic resistance in diabetic foot infec-
Society for Vascular Surgery in collaboration with the tions: a prospective multi-center study. J Diabetes
American Podiatric Medical Association and the Complications 2016;30(5):910 16.

Diabetic Neuropathy
 References 255
[47] Rastogi A, Sukumar S, Hajela A, Mukherjee S, Dutta osteomyelitis: a multicenter open-label controlled ran-
P, Bhadada SK, et al. The microbiology of diabetic foot domized study. Diabetes Care 2015;38(2):302 7.
infections in patients recently treated with antibiotic [60] Gariani K, Pham TT, Kressmann B, Jornayvaz FR,
therapy: a prospective study from India. J Diabetes Gastaldi G, Stafylakis D, et al. Three vs six weeks of
Complications 2017;31(2):407 12. antibiotic therapy for diabetic foot osteomyelitis: a
[48] Bowling FL, Jude EB, Boulton AJ. MRSA and diabetic prospective, randomized, non-inferiority pilot trial.
foot wounds: contaminating or infecting organisms? Clin Infect Dis 2020.
Curr Diab Rep 2009;9(6):440 4. [61] Vas PRJ, Demetriou M, Papanas N. Oral antibiotic
[49] Lipsky BA. Diabetic foot infections: current treatment therapy in diabetic foot osteomyelitis: one small step
and delaying the ‘post-antibiotic era’. Diabetes Metab or a giant leap of faith? Ann Transl Med 2019;7(Suppl.
Res Rev 2016;32(Suppl. 1):246 53. 8):S266.
[50] Lazaro Martinez JL, Garcia Alvarez Y, Tardaguila- [62] Vas PRJ, Edmonds M, Kavarthapu V, Rashid H,
Garcia A, Garcia, Morales E. Optimal management of Ahluwalia R, Pankhurst C, et al. The diabetic foot
diabetic foot osteomyelitis: challenges and solutions. attack: “Tis too late to retreat!”. Int J Low Extrem
Diabetes Metab Syndr Obes 2019;12:947 59. Wounds 2018;17(1):7 13.
[51] Lauri C, Tamminga M, Glaudemans A, Juarez Orozco [63] Ahluwalia R, Vainieri E, Tam J, Sait S, Sinha A, Manu
LE, Erba PA, Jutte PC, et al. Detection of osteomyelitis CA, et al. Surgical diabetic foot debridement: improv-
in the diabetic foot by imaging techniques: a system- ing training and practice utilizing the traffic light prin-
atic review and meta-analysis comparing MRI, white ciple. Int J Low Extrem Wounds 2019;18(3):279 86.
blood cell scintigraphy, and FDG-PET. Diabetes Care [64] Vainieri E, Ahluwalia R, Slim H, Walton D, Manu C,
2017;40(8):1111 20. Taori S, et al. Outcomes after emergency admission
[52] Lipsky BA, Aragon-Sanchez J, Diggle M, Embil J, with a diabetic foot attack indicate a high rate of heal-
Kono S, Lavery L, et al. IWGDF guidance on the diag- ing and limb salvage but increased mortality: 18-
nosis and management of foot infections in persons month follow-up study. Exp Clin Endocrinol Diabetes
with diabetes. Diabetes Metab Res Rev 2016;32(Suppl. 2020.
1):45 74. [65] Sheehan P, Jones P, Giurini JM, Caselli A, Veves A.
[53] Li HK, Rombach I, Zambellas R, Walker AS, McNally Percent change in wound area of diabetic foot ulcers
MA, Atkins BL, et al. Oral vs intravenous antibiotics over a 4-week period is a robust predictor of complete
for bone and joint infection. N Engl J Med 2019;380 healing in a 12-week prospective trial. Plast Reconstr
(5):425 36. Surg 2006;117(7 Suppl.):239S 244SS.
[54] Scarborough M, Li HK, Rombach I, Zambellas R, [66] Sheehan P, Jones P, Caselli A, Giurini JM, Veves A.
Walker AS, McNally M, et al. Oral vs intravenous anti- Percent change in wound area of diabetic foot ulcers
biotics for bone and joint infections: the OVIVA non- over a 4-week period is a robust predictor of complete
inferiority RCT. Health Technol Assess 2019;23 healing in a 12-week prospective trial. Diabetes Care
(38):1 92. 2003;26(6):1879 82.
[55] Lazaro-Martinez JL, Aragon-Sanchez J, Garcia- [67] Liu T, Yang F, Li Z, Yi C, Bai X. A prospective pilot
Morales E. Antibiotics vs conservative surgery for study to evaluate wound outcomes and levels of
treating diabetic foot osteomyelitis: a randomized serum C-reactive protein and interleukin-6 in the
comparative trial. Diabetes Care 2014;37(3):789 95. wound fluid of patients with trauma-related chronic
[56] Yosuf MK, Mahadi SI, Mahmoud SM, Widatalla AH, wounds. Ostomy Wound Manage 2014;60(6):30 7.
Ahmed ME. Diabetic neuropathic forefoot and heel [68] Digital N. National Diabetes Foot Care Audit Report
ulcers: management, clinical presentation and out- 2014 2016; 2017.
comes. J Wound Care 2015;24(9):420 5. [69] Prompers L, Schaper N, Apelqvist J, Edmonds M,
[57] Peters EJ. Pitfalls in diagnosing diabetic foot infec- Jude E, Mauricio D, et al. Prediction of outcome in
tions. Diabetes Metab Res Rev 2016;32(Suppl. individuals with diabetic foot ulcers: focus on the dif-
1):254 60. ferences between individuals with and without
[58] Bateman AH, Bradford S, Hester TW, Kubelka I, peripheral arterial disease. The EURODIALE study.
Tremlett J, Morris V, et al. Modern orthopedic inpa- Diabetologia 2008;51(5):747 55.
tient care of the orthopedic patient with diabetic foot [70] Fife CE, Eckert KA, Carter MJ. Publicly reported
disease. Int J Low Extrem Wounds 2015;14(4):384 92. wound healing rates: the fantasy and the reality. Adv
[59] Tone A, Nguyen S, Devemy F, Topolinski H, Valette Wound Care (N Rochelle) 2018;7(3):77 94.
M, Cazaubiel M, et al. Six-week vs twelve-week antibi- [71] Uccioli L, Izzo V, Meloni M, Vainieri E, Ruotolo V,
otic therapy for nonsurgically treated diabetic foot Giurato L. Non-healing foot ulcers in diabetic

Diabetic Neuropathy
256 15. Management of diabetic foot disease

patients: general and local interfering conditions and [82] Elg F, Hunt S. Hemoglobin spray as adjunct therapy
management options with advanced wound dressings. J in complex wounds: meta-analysis vs standard care
Wound Care 2015;24(4 Suppl.):35 42. alone in pooled data by wound type across three ret-
[72] Liu Z, Dumville JC, Hinchliffe RJ, Cullum N, Game F, rospective cohort controlled evaluations. SAGE Open
Stubbs N, et al. Negative pressure wound therapy for Med 2018;6 2050312118784313.
treating foot wounds in people with diabetes mellitus. [83] Zelen CM. An evaluation of dehydrated human amni-
Cochrane Database Syst Rev 2018;10 CD010318. otic membrane allografts in patients with DFUs.
[73] Seidel D, Storck M, Lawall H, Wozniak G, Mauckner J Wound Care 2013;22(7):347 8 50-1.
P, Hochlenert D, et al. Negative pressure wound ther- [84] Lakmal K, Basnayake O, Hettiarachchi D. Systematic
apy compared with standard moist wound care on review on the rational use of amniotic membrane allo-
diabetic foot ulcers in real-life clinical practice: results grafts in diabetic foot ulcer treatment. BMC Surg
of the German DiaFu-RCT. BMJ Open 2020;10(3): 2021;21(1):87.
e026345. [85] Su YN, Zhao DY, Li YH, Yu TQ, Sun H, Wu XY, et al.
[74] Bishop A. Role of oxygen in wound healing. J Wound Human amniotic membrane allograft, a novel treat-
Care 2008;17(9):399 402. ment for chronic diabetic foot ulcers: a systematic
[75] Abidia A, Laden G, Kuhan G, Johnson BF, Wilkinson review and meta-analysis of randomised controlled
AR, Renwick PM, et al. The role of hyperbaric oxygen trials. Int Wound J 2020;17(3):753 64.
therapy in ischaemic diabetic lower extremity ulcers: a [86] Barrientos S, Stojadinovic O, Golinko MS, Brem H,
double-blind randomised-controlled trial. Eur J Vasc Tomic-Canic M. Growth factors and cytokines in wound
Endovasc Surg 2003;25(6):513 18. healing. Wound Repair Regen 2008;16(5):585 601.
[76] Londahl M, Katzman P, Nilsson A, Hammarlund C. [87] Werner S, Grose R. Regulation of wound healing by
Hyperbaric oxygen therapy facilitates healing of growth factors and cytokines. Physiol Rev 2003;83
chronic foot ulcers in patients with diabetes. Diabetes (3):835 70.
Care 2010;33(5):998 1003. [88] Marti-Carvajal AJ, Gluud C, Nicola S, Simancas-
[77] Fedorko L, Bowen JM, Jones W, Oreopoulos G, Goeree Racines D, Reveiz L, Oliva P, et al. Growth factors for
R, Hopkins RB, et al. Hyperbaric oxygen therapy does treating diabetic foot ulcers. Cochrane Database Syst
not reduce indications for amputation in patients with Rev 2015;10:CD008548.
diabetes with nonhealing ulcers of the lower limb: a [89] Game F, Jeffcoate W, Tarnow L, Jacobsen JL, Whitham
prospective, double-blind, randomized controlled clin- DJ, Harrison EF, et al. LeucoPatch system for the man-
ical trial. Diabetes Care 2016;39(3):392 9. agement of hard-to-heal diabetic foot ulcers in the UK,
[78] Santema KTB, Stoekenbroek RM, Koelemay MJW, Denmark, and Sweden: an observer-masked, rando-
Reekers JA, van Dortmont LMC, Oomen A, et al. mised controlled trial. Lancet Diabetes Endocrinol
Hyperbaric oxygen therapy in the treatment of 2018;6(11):870 8.
ischemic lower-extremity ulcers in patients with dia- [90] Kirsner RS, Sabolinski ML, Parsons NB, Skornicki M,
betes: results of the DAMO2CLES multicenter ran- Marston WA. Comparative effectiveness of a bioengi-
domized clinical trial. Diabetes Care 2018;41 neered living cellular construct vs. a dehydrated
(1):112 19. human amniotic membrane allograft for the treat-
[79] Brouwer RJ, Lalieu RC, Hoencamp R, van Hulst RA, ment of diabetic foot ulcers in a real world setting.
Ubbink DT. A systematic review and meta-analysis of Wound Repair Regen 2015;23(5):737 44.
hyperbaric oxygen therapy for diabetic foot ulcers [91] Dixon D, Edmonds M. Managing diabetic foot ulcers:
with arterial insufficiency. J Vasc Surg 2020;71 pharmacotherapy for wound healing. Drugs 2021;81
(2):682 92 e1. (1):29 56.
[80] Niederauer MQ, Michalek JE, Liu Q, Papas KK, [92] Maltese G, Karalliedde J, Rapley H, Amor T, Lakhani
Lavery LA, Armstrong DG. Continuous diffusion of A, Gnudi L. A pilot study to evaluate the efficacy of
oxygen improves diabetic foot ulcer healing when class IV lasers on nonhealing neuroischemic diabetic
compared with a placebo control: a randomised, foot ulcers in patients with type 2 diabetes. Diabetes
double-blind, multicentre study. J Wound Care Care 2015;38(10):e152 3.
2018;27(Suppl. 9):S30 45. [93] Kotronis G, Vas PRJ. Ultrasound devices to treat
[81] Frykberg RG, Franks PJ, Edmonds M, Brantley JN, chronic wounds: the current level of evidence. Int J
Teot L, Wild T, et al. A multinational, multicenter, ran- Low Extrem Wounds 2020;19(4):341 9.
domized, double-blinded, placebo-controlled trial to [94] Piaggesi A, Sambataro M, Nicoletti C, Goretti C,
evaluate the efficacy of cyclical topical wound oxygen Lacopi E, Coppelli A. Safety and effectiveness of ther-
(TWO2) therapy in the treatment of chronic diabetic apeutic magnetic resonance in diabetic foot ulcers: a
foot ulcers: the TWO2 study. Diabetes Care 2020;43 prospective randomised controlled trial. J Wound
(3):616 24. Care 2016;25(12):704 11.

Diabetic Neuropathy
 References 257
[95] Petrofsky JS, Lawson D, Berk L, Suh H. Enhanced [107] Rajamani K, Colman PG, Li LP, Best JD, Voysey M,
healing of diabetic foot ulcers using local heat and D’Emden MC, et al. Effect of fenofibrate on amputa-
electrical stimulation for 30 min three times per tion events in people with type 2 diabetes mellitus
week. J Diabetes 2010;2(1):41 6. (FIELD study): a prespecified analysis of a rando-
[96] Stratton IM, Adler AI, Neil HA, Matthews DR, mised controlled trial. Lancet. 2009;373(9677):1780 8.
Manley SE, Cull CA, et al. Association of glycaemia [108] Hinchliffe RJ, Forsythe RO, Apelqvist J, Boyko EJ,
with macrovascular and microvascular complications Fitridge R, Hong JP, et al. Guidelines on diagnosis,
of type 2 diabetes (UKPDS 35): prospective observa- prognosis, and management of peripheral artery dis-
tional study. BMJ 2000;321(7258):405 12. ease in patients with foot ulcers and diabetes
[97] Pop-Busui R, Herman WH, Feldman EL, Low PA, (IWGDF 2019 update). Diabetes Metab Res Rev
Martin CL, Cleary PA, et al. DCCT and EDIC studies 2020;36(Suppl. 1):e3276.
in type 1 diabetes: lessons for diabetic neuropathy [109] Jones WS, Baumgartner I, Hiatt WR, Heizer G, Conte
regarding metabolic memory and natural history. MS, White CJ, et al. Ticagrelor compared with clopi-
Curr Diab Rep 2010;10(4):276 82. dogrel in patients with prior lower extremity revas-
[98] Lane KL, Abusamaan MS, Voss BF, Thurber EG, Al- cularization for peripheral artery disease.
Hajri N, Gopakumar S, et al. Glycemic control and Circulation. 2017;135(3):241 50.
diabetic foot ulcer outcomes: a systematic review and [110] Aboyans V, Ricco JB, Bartelink MEL, et al., editors.
meta-analysis of observational studies. J Diabetes Editor’s Choice - 2017 ESC guidelines on the diagno-
Complications 2020;34(10):107638. sis and treatment of peripheral arterial diseases, in
[99] Vas PRJ, Whyte MB, Papanas N. Association between collaboration with the European Society for Vascular
glycaemic control and diabetic foot outcomes: dark Surgery (ESVS). Eur J Vasc Endovasc Surg
side of the moon? J Diabetes Complications 2020;34 2018;55:305 68.
(10):107650. [111] Eikelboom JW, Connolly SJ, Yusuf S. Rivaroxaban in
[100] Rollins KE, Varadhan KK, Dhatariya K, Lobo DN. stable cardiovascular disease. N Engl J Med 2018;378
Systematic review of the impact of HbA1c on out- (4):397 8.
comes following surgery in patients with diabetes [112] Bonaca MP, Bauersachs RM, Anand SS, Debus ES,
mellitus. Clin Nutr 2016;35(2):308 16. Nehler MR, Patel MR, et al. Rivaroxaban in periph-
[101] Mader JK, Haas W, Aberer F, Boulgaropoulos B, eral artery disease after revascularization. N Engl J
Baumann P, Pandis M, et al. Patients with healed dia- Med 2020;382(21):1994 2004.
betic foot ulcer represent a cohort at highest risk for [113] Zhang YS, Chen JH, Mei T, Li S, Lu YM. Efficacy and
future fatal events. Sci Rep 2019;9(1):10325. safety of DOACs in patients with peripheral arterial
[102] Giugliano D, Bellastella G, Longo M, Scappaticcio L, disease: a systematic review and meta-analysis.
Maiorino MI, Chiodini P, et al. Relationship between Vascular. 2021; 1708538120987935.
improvement of glycaemic control and reduction of [114] Atkin L, Tansley J, Stephenson J. Diabetic foot ulcera-
major cardiovascular events in 15 cardiovascular out- tion: the impact of oedema. Wounds UK 2018;14
come trials: a meta-analysis with meta-regression. (1):33 9.
Diabetes Obes Metab 2020;22(8):1397 405. [115] Metcalf L, Musgrove M, Bentley J, Berrington R,
[103] Dhatariya K, Bain SC, Buse JB, Simpson R, Tarnow L, Bunting D, Mousley M, et al. Prevalence of active
Kaltoft MS, et al. The impact of liraglutide on Charcot disease in the East Midlands of England.
diabetes-related foot ulceration and associated com- Diabet Med 2018;35(10):1371 4.
plications in patients with type 2 diabetes at high risk [116] Rogers LC, Frykberg RG, Armstrong DG, Boulton AJ,
for cardiovascular events: results from the LEADER Edmonds M, Van GH, et al. The Charcot foot in dia-
trial. Diabetes Care 2018;41(10):2229 35. betes. Diabetes Care 2011;34(9):2123 9.
[104] Papadokostaki E, Rizos E, Tigas S, Liberopoulos EN. [117] Mabilleau G, Petrova NL, Edmonds ME, Sabokbar A.
Canagliflozin and amputation risk: evidence so far. Increased osteoclastic activity in acute Charcot’s
Int J Low Extrem Wounds 2020;19(1):21 6. osteoarthropathy: the role of receptor activator of
[105] Food and Drug Administration. FDA removes Boxed nuclear factor-kappaB ligand. Diabetologia. 2008;51
Warning about risk of leg and foot amputations for the (6):1035 40.
diabetes medicine canagliflozin. https://www.fda.gov/ [118] Petrova NL. The Charcot foot revisited: how the new
media/141533/download; 2020. [accessed 31.08.20]. pathogenetic findings explain the clinical course of
[106] Authors/Task Force M, Guidelines ESCCfP, Societies the disease. The diabetic foot syndrome, vol. 26.
ESCNC. ESC/EAS guidelines for the management of Karger Publishers; 2018. p. 48 59.
dyslipidaemias: lipid modification to reduce cardio- [119] Meacock L, Petrova NL, Donaldson A, Isaac A,
vascular risk. Atherosclerosis. 2019;(290):140 205. Briody A, Ramnarine R, et al. Novel

Diabetic Neuropathy
258 15. Management of diabetic foot disease

semiquantitative bone marrow eedema score and [130] Frykberg RG, Wukich DK, Kavarthapu V, Zgonis T,
fracture score for the magnetic resonance imaging Dalla Paola L, Board of the Association of Diabetic
assessment of the active Charcot foot in diabetes. J Foot Surgeons. Surgery for the diabetic foot: a key
Diabetes Res 2017;2017:8504137. component of care. Diabetes Metab Res Rev 2020;36
[120] Ahluwalia R, Bilal A, Petrova N, Boddhu K, Manu C, (Suppl. 1):e3251.
Vas P, et al. The role of bone scintigraphy with [131] Kavarthapu V. Is there a need for diabetic foot ortho-
SPECT/CT in the characterization and early diagno- paedic surgery as a sub-specialty? J Clin Orthop
sis of stage 0 Charcot neuroarthropathy. J Clin Med Trauma 2021;17:72 3.
2020;9(12). [132] van Netten JJ, Bril A, van Baal JG. The effect of flexor
[121] Petrova NL, Edmonds ME. Conservative and phar- tenotomy on healing and prevention of neuropathic
macologic treatments for the diabetic Charcot foot. diabetic foot ulcers on the distal end of the toe. J Foot
Clin Podiatr Med Surg 2017;34(1):15 24. Ankle Res 2013;6(1):3.
[122] Chantelau E. The perils of procrastination: effects of [133] Frykberg RG, Bevilacqua NJ, Habershaw G. Surgical
early vs. delayed detection and treatment of incipient off-loading of the diabetic foot. J Am Podiatr Med
Charcot fracture. Diabet Med 2005;22(12):1707 12. Assoc 2010;100(5):369 84.
[123] Game FL, Catlow R, Jones GR, Edmonds ME, Jude [134] Vasukutty N, Jawalkar H, Anugraha A, Chekuri R,
EB, Rayman G, et al. Audit of acute Charcot’s disease Ahluwalia R, Kavarthapu V. Correction of ankle and
in the UK: the CDUK study. Diabetologia. 2012;55 hind foot deformity in Charcot neuroarthropathy
(1):32 5. using a retrograde hind foot nail—the Kings’ experi-
[124] Jude EB, Selby PL, Burgess J, Lilleystone P, Mawer EB, ence. Foot Ankle Surg 2018;24(5):406 10.
Page SR, et al. Bisphosphonates in the treatment of [135] Lavery LA, Armstrong DG, Boulton AJ, Diabetex
Charcot neuroarthropathy: a double-blind randomised Research G. Ankle equinus deformity and its rela-
controlled trial. Diabetologia. 2001;44(11):2032 7. tionship to high plantar pressure in a large popula-
[125] Pitocco D, Ruotolo V, Caputo S, Mancini L, Collina tion with diabetes mellitus. J Am Podiatr Med Assoc
CM, Manto A, et al. Six-month treatment with alen- 2002;92(9):479 82.
dronate in acute Charcot neuroarthropathy: a ran- [136] Dallimore SM, Kaminski MR. Tendon lengthening
domized controlled trial. Diabetes Care 2005;28 and fascia release for healing and preventing diabetic
(5):1214 15. foot ulcers: a systematic review and meta-analysis.
[126] Pakarinen TK, Laine HJ, Maenpaa H, Mattila P, J Foot Ankle Res 2015;8:33.
Lahtela J. The effect of zoledronic acid on the clinical [137] Molines-Barroso RJ, Lazaro-Martinez JL, Aragon-
resolution of Charcot neuroarthropathy: a pilot ran- Sanchez J, Garcia-Morales E, Beneit-Montesinos JV,
domized controlled trial. Diabetes Care 2011;34 Alvaro-Afonso FJ. Analysis of transfer lesions in
(7):1514 16. patients who underwent surgery for diabetic foot
[127] Das L, Bhansali A, Prakash M, Jude EB, Rastogi A. ulcers located on the plantar aspect of the metatarsal
Effect of methylprednisolone or zoledronic acid on heads. Diabet Med 2013;30(8):973 6.
resolution of active Charcot neuroarthropathy in dia- [138] Dalla Paola L, Volpe A, Varotto D, Postorino A,
betes: a randomized, double-blind, placebo- Brocco E, Senesi A, et al. Use of a retrograde nail for
controlled study. Diabetes Care 2019;42(12):e185 6. ankle arthrodesis in Charcot neuroarthropathy: a
[128] Rastogi A, Bhansali A, Jude EB. Efficacy of medical limb salvage procedure. Foot Ankle Int 2007;28
treatment for Charcot neuroarthropathy: a systematic (9):967 70.
review and meta-analysis of randomized controlled [139] Lowery NJ, Woods JB, Armstrong DG, Wukich DK.
trials. Acta Diabetol 2021. Surgical management of Charcot neuroarthropathy
[129] Bem R, Jirkovska A, Fejfarova V, Skibova J, Jude EB. of the foot and ankle: a systematic review. Foot
Intranasal calcitonin in the treatment of acute Ankle Int 2012;33(2):113 21.
Charcot neuroosteoarthropathy: a randomized con-
trolled trial. Diabetes Care 2006;29(6):1392 4.

Diabetic Neuropathy
 C H A P T E R

16
Strategies for the prevention or reversal
of neuropathy
Lindsay A. Zilliox, Krish Chandrasekaran and James W. Russell
Department of Neurology, University of Maryland and Maryland VA Healthcare System, Baltimore,
MD, United States

Introduction less. Third, the endpoints and the trial design

that have been used may lack adequate sensi-
Peripheral somatic and autonomic neuropa- tivity to detect small changes in the severity of
thy are major risk factors predicting future diabetic polyneuropathy. Fourth, and perhaps
mortality in diabetes [1,2]. Therefore preven- most importantly, many of the trials of diabetic
tion and treatment of neuropathy are impor- neuropathy have approached treatment when
tant not only in reducing the patient burden the patient’s disease has progressed to the
from this common complication but is also point where nerve regeneration and concomi-
vital in reducing overall morbidity and mortal- tant improvement in symptoms and signs of
ity in diabetes. At the present time, there are disease are limited. Treatment of diabetic neu-
no medications that have been proven to be ropathy early in the disease before there is sig-
effective in randomized, blinded (masked) clin- nificant neurodegeneration is critical if new
ical trials. Numerous clinical trials aimed at a therapies are to show efficacy in clinical trials
diverse array of biochemical pathways have and be introduced into practice.
failed, up to this point, to show efficacy for dia- Although there are no specific medications
betic polyneuropathy. This failure of clinical to prevent or treat the pathophysiology of dia-
trials is likely due to several causes. First, the betic neuropathy, there are approaches that
pathophysiology of diabetic neuropathy is can slow the progression of neuropathy and
extremely complex, and it is likely that more may impact mortality from the disease. These
than one treatment approach will be required approaches are discussed in the following sec-
to achieve clinical improvement. Second, stabi- tions and include improved glycemic control
lization or improvement in diabetic neuropa- for neuropathy associated with type 1 diabetes
thy may take years. A small treatment effect mellitus (T1DM), lifestyle interventions, nutra-
may be missed by trials, which because of ceuticals, and medications that require further
research funding limit treatment to 1 year or clinical testing.

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00016-5 259 © 2022 Elsevier Inc. All rights reserved.
260 16. Strategies for the prevention or reversal of neuropathy

Improved glycemic control over 1400 patients with T1DM found that
intensive diabetes therapy with either multiple
Intensive glucose control with the goal of nor- daily insulin injections or continuous subcuta-
moglycemia is one of the mainstays of treatment neous insulin infusion reduced the develop-
in diabetes [3,4]. Strict insulin therapy has been ment of clinical neuropathy by 64% (95% CI,
shown to be effective in the prevention or delay 45% 76%) compared to conventional therapy
of progression of diabetic polyneuropathy (DPN) over 5 years. The prevalence of abnormalities
in T1DM patients [5]. However, in type 2 diabetes on nerve conduction testing was reduced by
mellitus (T2DM), the evidence that improved gly- 44% (CI, 34% 53%) in the intensive treatment
cemic control can slow progression of neuropathy group compared to conventional treatment.
is limited [5,6]. Seven of the studies addressing In terms of neuropathy progression, the
the effect of improved glycemic control examined DCCT found that intensive glycemic control
DPN in T1DM; however, only two reported out- was effective in preventing decreases in nerve
comes related to clinical impairment. Overall, conduction velocities (NCV) observed in parti-
there is a large, significant reduction in neuropa- cipants in the conventional treatment cohort
thy with tighter glucose control in T1DM. There [8,9].
have been eight randomized clinical trials of The question of whether normoglycemia can
patients with T2DM, and the results have been prevent the development of neuropathy in
less definitive than those in T1DM. Half of these T1DM was examined by a prospective observa-
studies examined the effects of glucose control on tional study of 11 patients with well-controlled
clinical impairments from neuropathy. Overall, T1DM (mean HbA1c , 7.0%) compared to 21
there is evidence of a modest reduction in neu- patients with poor control (mean HbA1c . 7.0%)
ropathy in patients with T2DM receiving [12]. The participants were followed from the
enhanced glucose control, which is in contrast to time of diagnosis (duration of diabetes 2.7 1 0.3
the substantial effect seen in those with T1DM weeks) for over 24 years. None of the patients
[5,7 11]. Explanations for this difference may be with well-controlled diabetes developed clinical
because of the different outcome measures used polyneuropathy. Compared to participants with
in the trials, different treatment regimens utilized poorly controlled T1DM, there was significantly
for glycemic control, the higher incidence of DPN less decline in peripheral and autonomic nerve
in controls with T2DM, and differences in base- function as measured by motor and NCV, quan-
line glucose control among the clinical trials. titative sensory testing, and heart rate variability.
Importantly, this difference in the impact of glu- At the completion of the study intervention in
cose control on DPN in T1DM and T2DM high- the DCCT, the participants originally assigned to
lights the differences in the disease mechanisms conventional insulin treatment were encouraged
and complications between these two diseases. to begin intensive treatment. Despite similar
HbA1c levels 13 14 years later, the participants
who were originally assigned to the conventional
treatment group continued to develop neuropathy
Somatic neuropathy in type 1 diabetes
at a higher rate than those originally assigned to
mellitus the intensive insulin therapy [9]. The persistent
One of the largest trials that established that effect of early intensive insulin therapy is referred
tight glycemic control can effectively delay the to as “metabolic memory,” which has been
onset or slow the progression of neuropathy ascribed to a change in the individual’s epigenetic
in T1DM was the Diabetes Control and expression. Support for this concept was provided
Complications Trial (DCCT) [8]. This trial of by evidence of differences in epigenetic DNA

Diabetic Neuropathy
 Improved glycemic control 261
methylation during the DCCT that were found to often irreversible and supports the idea that
persist at certain loci associated with glycemia dur- early prevention of neuropathy is paramount.
ing the Epidemiology of Diabetes Interventions
and Complications (EDIC) study [13].
Based upon the beneficial effects of normo- Somatic neuropathy in type 2 diabetes
glycemia achieved by intensive improvement
mellitus
in glycemic control, several small trials have
examined the effects of simultaneous pancreas In contrast to T1DM, there is no convincing
and kidney transplants on neuropathy in evidence in T2DM patients that intensive glu-
T1DM. Overall, transplantation is associated cose control can prevent or delay the develop-
with both early and maintained small nerve ment or progression of DPN. There is evidence
fiber regeneration, followed by improvement to support an important role of glucose control
in nerve conduction indices. Improvements in in DPN associated with T2DM, but normoglyce-
corneal confocal microscopy assessments, a mia alone is not sufficient to prevent or reverse
measure of small fiber pathology, have been DPN. A Cochrane systemic review in 2012 con-
seen within 6 and 12 months of transplantation cluded that while enhanced glucose control pre-
[14,15]. A controlled, 3-year-long observational vents the development of neuropathy in
study found that, when compared to the patients with T1DM, there is a nonsignificant
patients with T1DM and severe DPN who did reduction in the incidence of clinical neuropathy
not undergo transplant, those in the transplant in patients with T2DM. However, there is a sig-
group saw improvements in the corneal nerve nificant reduction in nerve conduction and
fiber density and length that were maintained vibration threshold abnormalities [18].
over the course of the study. While the intrae- One of the largest trials of the effect of normo-
pidermal nerve fiber density (IENFD) did not glycemia on neuropathy in T2DM is the UK
change, the mean dendritic length improved Prospective Diabetes Study (UKPDS) [19]. The
significantly at 12 and 36 months. By 36 UKPDS randomized 3867 patients with newly
months, there were also significant improve- diagnosed T2DM to either intensive glucose ther-
ments in the neuropathy symptom profile and apy with a sulphonylurea, intensive glucose ther-
peroneal NCV. There was no change in quanti- apy with insulin, or conventional diet therapy. At
tative sensory testing or cardiac autonomic the end of 10 years, there was a small difference
function assessments [16]. The long-term in HgBA1c between the intensive (HgBA1c 7.0%)
effects of simultaneous pancreas and kidney and standard (HgBA1c 7.9%) control groups. An
transplantation on patients with T1DM and aggregate 25% risk reduction (P 5 .0099) in
severe neuropathy at baseline have also been microvascular endpoints was found in the inten-
examined. An observational study of 12 partici- sive blood-glucose groups after 10 years. The
pants with T1DM and severe neuropathy prior main neuropathy outcome measure, vibration
to transplant found that after 8 years, all the detection threshold, had a lower rate of
participants had excellent glycemic control impairment with intensive glucose therapy com-
(median HgBA1c 5.5%) but there was no pared to conventional therapy. However, this dif-
improvement seen in IENFD, vibration percep- ference was only significant after 15 years
tion threshold (VPT) or autonomic function (relative risk 0.60, 95% CI 0.39 0.94). More
testing results. However, a statistically signifi- importantly, despite the small improvement,
cant improvement in the median motor NCV there was a higher rate of hypoglycemic compli-
was seen [17]. The results from this long-term cations in the intensive glucose therapy groups,
study highlight the fact that advanced DPN is particularly with insulin treatment.

Diabetic Neuropathy
262 16. Strategies for the prevention or reversal of neuropathy

There have been other more recent large stud- T2DM and DPN at baseline and determined
ies of the effects of strict glucose control on the the effect of intensive glucose treatment on
development of T2DM complications. The VA measures of small fiber neuropathy [11].
Diabetes Trial randomized 1791 military veter- Patients were treated with either standard care
ans with T2DM to either intensive or standard or intensive HbA1c control, which included
glucose control. The primary outcome measure diet restriction and hypoglycemic agents, and
was time to a major cardiovascular event. The were then followed for an average of 4 years.
development of neuropathy was determined The participants who were treated with inten-
based upon self-report. While they did not find sive HbA1c control demonstrated normaliza-
any difference in the rate of new microvascular tion of HbA1c (9.6% 5.9%), which was
complications, including neuropathy, there was maintained for 2 years. At the end of the study,
a nonsignificant 5% reduction in the incidence of there were significant improvements in several
neuropathy [20]. The Action to Control measures of nerve conduction studies as well
Cardiovascular Risk in Diabetes trial was also as corneal nerve fiber measures seen in the
designed to determine if strict glycemic control group treated with extensive HbA1c control,
in T2DM was effective in reducing the rate of while the vibration detection threshold
cardiovascular events [21]. This trial randomized improved in the standard care group. These
10,251 patients to either intensive glucose control improvements were associated with a signifi-
(HgBA1c goal , 6.0%) or standard care (HgBA1c cant reduction in weight in the extensive
goal 7.0% 7.9%). The trial was stopped early HbA1c control group.
after 3.5 years due to higher mortality in the Not all studies have shown that aggressive
intensive therapy group; however, planned glucose control reduces the development or
follow-up continued for 5 years. The incidence progression of neuropathy in T2DM. The
of neuropathy, as measured by a Michigan Veterans Affairs Cooperative Study on Type II
Neuropathy Screening Instrument (MNSI) Diabetes Mellitus (VA-CSDM) was designed as
score . 2.0 was significantly reduced in the a feasibility study that compared standard
intensive glucose control group (Hazard Ratio insulin treatment to an intensive therapy group
(HR) 0.92, Confidence Interval (CI) 0.86 0.99, in men with T2DM who required insulin. The
P 5 .027, Number Needed to Treat 5 33). enrolled men had poorly controlled T2DM
Intensive treatment was also associated with sig- with an average duration of 7.8 6 4 years and
nificant reductions in the development of loss of about half had neuropathy at baseline.
ankle jerk and light touch by 10-g monofilament, Intensive insulin therapy did result in an
but the loss of vibratory sensation did not differ. improvement in HgBA1c after 6 months and
Additional evidence from smaller trials of was 2% lower than that in the standard treat-
patients with T2DM lend support for the role ment group. However, there was no difference
of glucose control in preventing the progres- found in the prevalence of DPN between the
sion of neuropathy. A Japanese trial enrolled intensive versus standard arms (64% intensive
110 patients with T2DM who were randomly vs 69% standard) after 2 years [23].
assigned to intense or conventional insulin Glucose control is probably only part of the
treatment. Intensive insulin therapy group optimal treatment approach to DPN, and the
showed significant improvement in NCV and type of glucose-lowering agent used to obtain
vibration threshold while those in the conven- normoglycemia may play a role in preventing
tional treatment group showed deterioration in the development of DPN in T2DM. The Bypass
these measures [22]. More recently, another Angioplasty Revascularization Investigation 2
study assessed patients with uncontrolled Diabetes trial enrolled 2159 participants with

Diabetic Neuropathy
 Improved glycemic control 263
T2DM and coronary artery disease to examine T1DM, the data for diabetic autonomic neurop-
the effect of insulin-sensitizing treatments athy (DAN) from the DCCT and the follow-up
(metformin, thiazolidinediones, or both) versus EDIC study was more conclusive for CAN
insulin-providing treatments (sulfonylureas/ than it was for DPN. The prevalence of CAN
meglitinides, insulin, or both) on cardiovascu- was low at the start of the DCCT. While the
lar outcomes [24]. DPN was a predetermined prevalence of CAN almost doubled in the con-
secondary outcome measure and was defined ventional group, it remained stable in the
as an MNSI clinical examination score . 2. intensive group and the risk reduction in inci-
After adjusting for HbA1c, which was signifi- dent CAN with intensive therapy was 45%. At
cantly lower in the insulin-sensitizing therapy the end of the DCCT, both groups were
arm, the cumulative incidence of new onset instructed to adhere to the intensive therapy
DPN in those participants without DPN at and were followed in the EDIC study, which
baseline (n 5 1075) over 4 years was significantly had more conclusive data for CAN than it did
reduced in the group of patients treated with for DPN. After 13 14 years of observation, the
insulin-sensitizing therapies compared to those prevalence of CAN increased in both groups
receiving insulin-providing therapies (66% vs but remained significantly lower in the former
72%, P 5 .02). There was no difference between intensive control group compared to the for-
the two treatment arms in the prevalence of mer conventional group (28.9% vs 35.2%,
DPN (51% vs 53%) and there was no difference P 5 .018). After adjusting for covariates, former
in the remission rates in those who had estab- participation in the intensive glucose control
lished DSP at baseline. There was a significant arm of the DCCT reduced the risk of incident
association between changes in the MNSI ques- CAN by 31% (OR 0.69, 95% CI 0.51 0.93) [25].
tionnaire and changes in body weight, It is possible that the more robust data for the
HgBA1c, and serum lipids. The finding that effect of early, strict glycemic control on CAN
insulin-sensitizing treatments are associated as opposed to DPN may be attributed to the
with a reduced development of DPN in T2DM, choice of outcome measures. The effect of met-
when compared to insulin-providing treat- abolic memory may be greater on the small
ments, indicates that the effect of glycemic con- unmyelinated nerve fibers that are measured
trol may be pathway specific. with the DCCT autonomic measures (R-R vari-
The methodological weakness in these trials ation with paced breathing, Valsalva ratio, and
is that most of the trials only enrolled a minor- postural blood pressure changes) as opposed
ity of patients with DPN, only included a few to the tests of large fiber function (nerve con-
clinical endpoints of DPN, different endpoints duction studies and VPT) that were chosen as
were used, only a few included quantitative measures of DPN.
measures of small fiber pathology, DPN and Similar to DPN, the data for the effect of inten-
cardiac autonomic neuropathy (CAN) were sive glycemic control on CAN in T2DM is not as
secondary and not primary outcomes, and the convincing as for T1DM. The lack of sensitive,
duration of intervention was relatively short. standardized outcome measures is problematic
for both DPN and CAN and there is even less
data on outcomes of autonomic function. The
results from the United Kingdom Prospective
Diabetic autonomic neuropathy Diabetes Study (UKPDS) and the Anglo-Danish-
Data regarding the effect of strict glucose Dutch Study of Intensive Treatment in People
control on CAN in T1DM and T2DM is similar with Screen Detected Diabetes in Primary Care
to that for somatic DPN. In patients with (ADDITION)-Denmark trials found that intensive

Diabetic Neuropathy
264 16. Strategies for the prevention or reversal of neuropathy

glucose control in participants with newly diag- with both impaired fasting glucose and IGT
nosed T2DM had no effect on CAN compared to (23.9%) has been found to be close to those
standard care. However, there were no measures with known T2DM (22.0%) [33]. There is less
of neuropathy included in the observational data regarding the association between predia-
follow-up of the UKPDS, and the ADDITION- betes and DAN, but the German KORA S4
Denmark study did not include any baseline eva- study found similar prevalence rates of CAN
luations for CAN. Similarly, the Veterans Affairs in participants with newly diagnosed T2DM
Diabetes Trial reported a nonsignificant increase and those with combined impaired fasting glu-
in CAN in the intensive therapy group and the cose and IGT [34]. However, no difference in
VA-CSDM found no difference in CAN after 2 the prevalence of DPN was found in patients
years of intensive insulin treatment. with impaired glycemia compared with
The Steno type 2 study also examined the matched controls with normoglycemia in a
effects of intensive treatment of T2DM and population-based study in Olmsted County,
included CAN as a secondary endpoint [26]. MN, USA [35]. However, this study has several
This study utilized an intensive multifactorial methodological problems: (1) The study used
treatment, which included not only intensive gly- relatively insensitive measures of neuropathy
cemic control but also treatment of hypertension, as endpoint measures, for example, composite
dyslipidemia, and microalbuminuria in T2DM. scores of nerve conduction (which are fre-
Over approximately 4 years of follow-up, inten- quently normal in prediabetes) and quantita-
sive treatment was found to be associated with a tive sensation tests, rather than using more
lower rate of progression of DAN compared to sensitive measures of early neuropathy such as
standard care (OR 0.32, CI 0.12 0.78). the IENFD; (2) rather than using prospective
data capture for baseline “glycemic groups,” a
reclassification was applied to the existing gly-
Lifestyle interventions cemic data; (3) if the “normative” comparison
data is obtained from a population group that
Traditionally, DPN was considered to be a is already known to be prone to metabolic syn-
late complication of diabetes with no effective drome, then the comparator data may be inad-
treatments other than symptomatic neuro- equate to detect a difference between groups;
pathic pain control. However, a growing body (4) the study did actually show decreased heart
of evidence indicates that small fiber neuropa- rate response to deep breathing between the
thy can occur early in the disease course at the impaired glycemia (nondiabetic) group and the
earliest stages of glucose dysregulation. This is nonimpaired glycemia (healthy subject) group
an important finding with significant clinical and there was no difference in decreased heart
implications because the unmyelinated small rate response to deep breathing between the
nerve fibers that are vulnerable to effects from impaired glycemia and diabetic groups. Heart
metabolic derangements are also more respon- rate responses to deep breathing, including the
sive to therapies compared to large myelinated inspiration:expiration ratio, have been shown
fibers [5,27,28]. Evidence for an association to be sensitive and reproducible measures of
between prediabetes and small fiber neuropa- CAN [4,31,36,37]. Another reason for the
thy comes from studies demonstrating an conflicting data on the relationship between pre-
increased prevalence of impaired glucose toler- diabetes and neuropathy is that impaired glyce-
ance (IGT) in patients with idiopathic DPN mia alone is not enough to cause neuropathy
and DAN compared to controls [4,28 32]. In and there must be another metabolic driver(s)
addition, the prevalence of DPN in individuals for the onset and progression of neuropathy.

Diabetic Neuropathy
 Lifestyle interventions 265
Metabolic syndrome refers to a collection of patients with prediabetes due to cost and poten-
abnormalities that occur together and include tial for serious side-effects such as hypoglycemia.
hypertension, hyperlipidemia, central obesity, A more suitable approach for these patients
and insulin resistance. Having metabolic syn- would be a lifestyle intervention that could arrest
drome has been found to increase an indivi- the underlying process that leads to neuropathy
dual’s risk for heart disease, stroke, and and its associated functional disability. Exercise
diabetes. While knowledge of how metabolic has been shown to improve neuropathy function,
syndrome and its individual components effect for example, an improvement in the 6-minute
peripheral nerves is growing, metabolic syn- walk test (Fig. 16.1) and neuropathy symptoms
drome has been linked to an increased risk of and measures of somatic and autonomic neuropa-
neuropathy. In patients with T2DM, those with thy (Fig. 16.2). In addition, exercise can promote
metabolic syndrome were found to be twice as regrowth of cutaneous small diameter nerve
likely as those with T2DM alone to have neu- fibers in both animal models and human trials,
ropathy [38]. In nondiabetic patients, metabolic but data from well-designed randomized control
syndrome has been found to be associated trials is needed [3].
with a small fiber neuropathy [39]. As There have been several studies that found
described above, strict glucose control alone is that lifestyle interventions are effective in pre-
not enough to halt the progression or reverse venting or delaying the onset of T2DM in peo-
DPN in T2DM. Components of metabolic syn- ple with IGT. A 6-year long lifestyle
drome, specifically obesity and dyslipidemia, intervention in people from Da Qing, China
have been identified as risk factors for DPN
independent of glucose control [40,41].
Dyslipidemia is thought to play a critical role
in the development and progression of DPN
through associated metabolic, endocrine, and
inflammatory effects and are attractive targets
for disease-modifying treatment of DPN [5,28].
One way to target several of the components
of metabolic syndrome as well as many of the
metabolic and inflammatory pathways involved
in the pathogenesis of DPN is through aerobic
exercise. The pathophysiology of DPN involves FIGURE 16.1 Improvement in the 6-minute walk
(6MW) in subjects with diabetic neuropathy receiving
persistent hyperglycemia as well as multiple bio- “standard of care” dietary and exercise counseling over a
chemical pathways that ultimately result in oxida- 12-month period [3]. Participants had mild, recently diag-
tive stress and inflammation. In contrast to nosed type 2 diabetes mellitus or IGT. (A) Change in the
targeted therapies, aerobic exercise has multiple 6MW between baseline and 6 months (n 5 74) and baseline
synergistic effects on many of the pathways that and 12 months (n 5 64). The 6MW was performed prospec-
tively, was masked, and completed in accordance with a
are adversely affected by diabetes. Exercise not standardized protocol. “Standard of care” advice included
only improves insulin sensitivity and glucose con- recommendations for 150 minutes/week of aerobic exer-
trol, but it increases end-organ perfusion, reduces cise, up to 30 minutes/session, and a goal of reducing base-
lipid and protein oxidation, inhibits adipocyte line weight by 7%. The change in the 6MW was statistically
production of free fatty acids and deleterious adi- significant at 6 months (P 5 .023) and 12 months (P 5 .011).
(B) Percent subjects showing an improvement in the 6MW
pokines, and reduces humeral inflammation at 12 months ( . 5% improvement in the 6MW), no change
[3,29,42 44]. Furthermore, glucose-modifying (,5% improvement), or worse 6MW with “standard of
drug therapy is typically not appropriate in care” dietary and exercise counseling.

Diabetic Neuropathy
266 16. Strategies for the prevention or reversal of neuropathy

FIGURE 16.2 Improvement in measures of autonomic function in subjects with neuropathy and impaired glucose
tolerance or type 2 diabetes mellitus receiving “standard of care” dietary and exercise counseling over a 12-month
period [3]. The protocol was performed prospectively, was masked, and standardized. “Standard of care” advice included
recommendations to undertake aerobic exercise for 150 minutes/week, up to 30 minutes/session, and a goal of reducing
baseline weight by 7%. However, none of the recommendations were enforced or monitored. The technician performing
the outcome measures was masked to the patient intervention. (A) Increased expiration:inspiration (E:I ratio, n 5 50) after 1
year (P 5 .042). Base 95% CI: 1.14, 1.22. 1 year 95% CI: 1.20 1.44. (B) Decreased Total Impact Score (TIS) on the Survey of
Autonomic Symptoms (SAS, n 5 71). (C) Decreased resting diastolic blood pressure (n 5 72).

with IGT found a 43% reduction in the inci- one case of diabetes in 3 years, 6.9 persons
dence of diabetes in a combined diet and exer- would have to participate in the lifestyle inter-
cise intervention group compared to the vention program, compared to 13.9 people
control group [45]. The participants were fol- who would have to receive metformin [46]. At
lowed for an additional 20 years and the com- the end of the DPP, all participants were
bined lifestyle intervention was associated offered the lifestyle intervention training. After
with a 47% lower incidence of severe retinopa- 15 years of follow-up, there was evidence of
thy after 20 years. However, no significant dif- the effects of metabolic memory [47]. The inci-
ference was reported in the prevalence of dence of diabetes was reduced by 27% in the
neuropathy, as assessed by monofilament test- original lifestyle intervention group compared
ing, in the combined intervention group com- to the placebo group. In terms of other out-
pared to the control group. However, come measures, the prevalence rates of the
monofilament testing is not the most sensitive aggregate microvascular outcome did not dif-
measure for neuropathy. fer between the treatment groups. However,
The Diabetes Prevention Program (DPP) when examining the cohort of participants
also demonstrated that a physical activity and who developed T2DM, the original lifestyle
dietary intervention can normalize features of intervention group showed reductions in neu-
metabolic syndrome and reduce the incidence ropathy, which was defined as loss of sensa-
of T2DM by 58% in people with impaired fast- tion to a 10-g monofilament, compared to the
ing glucose (IFG) or IGT. In fact, the DPP placebo and metformin groups [47]. This find-
found that a lifestyle intervention was almost ing suggests previous lifestyle intervention in
twice as effective as metformin in preventing people with IGT or IFG who convert to diabe-
conversion from IGT to diabetes. To prevent tes may reduce the prevalence of DPN. It is

Diabetic Neuropathy
 Lifestyle interventions 267
important to note that in addition to DPN measurements did not include assessments for
being defined by monofilament testing, there neuropathy. The primary outcome measure
were no baseline assessments of neuropathy was rate of cardiovascular events, which did
included in the DPP, so incident DPN could not show a difference after 9.6 years when the
not be evaluated. intervention was stopped. However, partici-
The Impaired Glucose Tolerance Causes pants were followed after the intervention was
Neuropathy (IGTN) Study was a 12-month nat- stopped and the prevalence of neuropathy was
ural history study of 32 patients with either assessed by the MNSI. The lifestyle interven-
IGT or IFG and mild neuropathy at the time of tion group was found to have a lower increase
enrollment [29]. It was designed to examine in neuropathic symptoms compared to the
the effects of a lifestyle intervention on DPN standard care group and this in turn was asso-
and specifically included measures of small ciated with the amount of weight loss.
fiber neuropathy. Participants were given gen- However, there was no effect of the lifestyle
eral dietary and physical activity advice that intervention on the MNSI physical examination
was similar to the goals of the DPP lifestyle score with the exception of light touch
intervention group. Those individuals who lost sensation.
weight and/or increased their physical activity In contrast to the Look AHEAD study,
were found to have significant improvements which was designed to assess cardiovascular
in the proximal thigh IENFD obtained from a events and did not assess for DPN at enroll-
3-mm skin punch biopsy (increase of 1.4 ment, an Italian study aimed to assess the
fibers/mm) and foot sweat volume measured effects of exercise training on the development
by quantitative sudomotor axon reflex. The of DPN [49]. This 4-year-long study enrolled 78
improvements in IENFD and sweat volume, participants with either T1DM or T2DM and
which are both small fiber nerve measure- no signs or symptoms of DPN. The partici-
ments, were associated with significant pants were randomized to either an intense
improvements in weight, glucose tolerance, aerobic exercise program (4 hours of treadmill
and lipid profile. The increase in the thigh walking weekly) or a control group with no
IENFD levels was clinically relevant in that it intervention. There was no dietary intervention
was significantly correlated with decreased included in the protocol. The exercise group
neuropathic pain scores. Although there was did not significantly change their BMI, waist
no control group for comparison, these find- circumference, or lipid profile. However, they
ings suggest that aggressive treatment of IGT did significantly improve their exercise capac-
with a lifestyle intervention might improve ity and there was improvement in the NCV for
clinically relevant measures of neuropathy. It both the peroneal and sural nerves (P , .001)
also highlights the importance of intervention and fewer participants in the exercise group
during the earliest stages of impaired glucose developed an increased VPT (P , .5). The per-
regulation before the development of frank centage of diabetic participants who developed
diabetes. DPN, based on clinical evaluation, at the end
The effect of lifestyle interventions has also of the study was significantly lower in the
been examined in T2DM. The Look AHEAD exercise group compared to the control group
(Action for Health in Diabetes) study enrolled (P , .05). These findings suggest that aerobic
over 5000 overweight or obese patients with exercise training may be an effective preven-
T2DM and randomized them to either an tion strategy for DPN in diabetic patients, but
intensive lifestyle intervention or diabetes edu- larger confirmatory studies are needed. It is
cation [48]. Similar to the DPP, baseline also important to note that assessments of

Diabetic Neuropathy
268 16. Strategies for the prevention or reversal of neuropathy

small fiber neuropathy were not included in aerobic exercise program (n 5 60) or general
this study. health counseling (n 5 40). After the 1-year inter-
Using measurements of small fiber neuropa- vention, participants in the exercise group had
thy obtained from distal and proximal lower a significantly increased IENFD at the distal
extremity skin biopsies (IENFD and branching) leg (increased by 1.5 fibers/mm) compared to
as one of the predetermined outcome mea- control subjects (decreased by 0.1 fibers/mm;
sures, a pilot study from the University of P 5 .03). This improvement was seen despite
Kansas examined the effect of an exercise pro- no significant changes in metabolic parameters,
gram on DPN progression in 17 patients with other than an improvement in HDL. Although
T2DM and neuropathy [50]. This small group it was not statistically significant, fewer sub-
of subjects participated in a 10-week aerobic jects in the exercise program developed neu-
and resistance exercise program. At the end of ropathy, defined as an abnormal score on the
the study, there was no change in weight, but Utah Early Neuropathy Scale, compared to
the participants significantly improved their those in the control group (5.7% vs 17%;
HbA1c as well as neuropathic symptoms mea- P 5 .08). These results provide further support
sured by the Michigan Neuropathy Symptom that IENFD is a sensitive and responsive bio-
Inventory and pain levels on a visual analog marker for future clinical trials in DPN. They
scale. There was also a significant improve- also suggest that exercise can prevent or even
ment seen on the epidermal nerve fiber branch- reverse nerve damage to unmyelinated axons
ing (0.11 branch nodes per fiber; P 5 .008) on in T2DM.
the proximal skin biopsy. There was also a The Utah group also used the same capsai-
nonsignificant improvement in the IENFD cin axotomy technique to examine the effects
(1.68 fibers/mm; P 5 .09) at the proximal of exercise on cutaneous regenerative capacity
biopsy site [50]. Considering that this pilot in metabolic syndrome [52]. Since nerve injury
study was only 10 weeks long, this finding was in DPN is thought to start prior to the onset of
consistent with greater improvements seen diabetes, successful prevention strategies may
over 1 year in the IGTN trial [29]. be needed during the earliest stages of glucose
Results from the IGTN natural history study dysregulation. This study enrolled 32 patients
and pilot data from the University of Kansas with metabolic syndrome and 35 patients
suggest that lifestyle interventions may exert with T2DM without signs or symptoms of
an effect on small unmyelinated epidermal neuropathy. Participants underwent thigh
axons. Probably this is because this subset of capsaicin axotomy and their baseline distal
neurons is not only more vulnerable to the leg IENFD and 30-day regeneration rate were
toxic effects of metabolic derangements, but assessed and compared. A subset of 36 patients
also has a higher regenerative capacity and is (17 with metabolic syndrome) was then
therefore more amenable to therapies. assigned to participate in a 6-month intensive
Researchers at the University of Utah exam- exercise and lifestyle counseling program and
ined this hypothesis by utilizing a chemical the regeneration rate was reassessed in month 4.
axotomy technique that utilizes topical capsai- At baseline, distal leg IENFD was significantly
cin to assess epidermal reinnervation by skin reduced in both metabolic syndrome and
biopsy and serial measurement of IENFD [51]. T2DM groups and the 30-day regeneration rate
To examine the effect of exercise on unmyelin- was comparable in both groups. After 4 months
ated axons in T2DM, researchers randomized of the lifestyle intervention, the 30-day reinner-
100 patients with T2DM and no signs or symp- vation rate increased to 0.051 fibers/mm/
toms of DPN to either a weekly resistance and day compared to the baseline rate of 0.072

Diabetic Neuropathy
 Bariatric surgery 269
fibers/mm/day (P 5 .002). Participants who intervention in DPN described above have sug-
achieved improvements in more components of gested that exercise without significant weight
metabolic syndrome had a greater increase in loss or change in metabolic parameters is suffi-
their reinnervation rate (P , .012). These results cient to improve DPN. While improved glyce-
provide further evidence that metabolic syn- mic control alone has not been convincingly
drome is associated with a similar reduction in shown to be an effective treatment for DPN, is
IENFD and cutaneous regenerative capacity normoglycemia together with weight loss
than is seen in T2DM. through bariatric surgery able to achieve this
In general, patients with diabetes are encour- goal? While there has not been a randomized
aged to participate in regular physical exercise control trial to address this question, several
and supervised exercise programs are well tol- small uncontrolled studies have found low-
erated [53]. However, patients with proliferative grade evidence that bariatric surgery may
nephropathy should avoid exercise that can improve measures of DPN.
cause hypertension [54] and weight-bearing Two prospective studies of patients with
exercises should be performed with caution in T2DM who underwent gastric bypass surgery
patients with more advanced neuropathy and found conflicting effects on measures of DPN.
insensate feet due to risk of developing a dia- The first study found that participants with
betic foot ulcer. In particular, patients with auto- existing preoperative neuropathy had an
nomic neuropathy can have an increased risk of improvement in their Neuropathy Symptom
exercise-induced injury due to decreased car- Score and Neuropathy Deficit Score 6 months
diac responsiveness to exercise, impaired ther- postoperatively [55]. Interestingly, this improve-
moregulation, orthostatic hypotension, and ment was independent of improved glucose
greater susceptibility to hypoglycemia. In addi- control. Significant limitations of this study
tion, because CAN is an independent risk factor include the lack of a control group and there
for the development of silent myocardial infarc- was no masking of the intervention during
tion and cardiovascular death, patients with assessment of outcome measures. Another small
CAN should be screened and possibly undergo prospective case control study of patient with
an exercise stress test before starting an exercise T2DM who underwent gastric bypass surgery
program. Patients with autonomic dysfunction found no change in nerve conduction studies
should also avoid exercising in hot or cold 1 year after surgery [56]. Neither of these two
environments because of the risk of dehydration studies focused on possible prevention of DPN
due to difficulty with thermoregulation. Finally, in patients with T2DM who underwent gastric
in patients with symptomatic orthostatic hypo- bypass surgery. A large retrospective observa-
tension, recumbent exercises or water aerobics tional cohort of over 4500 participants with
may be safer activities. T2DM who underwent bariatric surgery found
a 29% lower risk of incident microvascular dis-
ease in participants whose T2DM remitted com-
Bariatric surgery pared to those who never remitted [57].
Interestingly, retinopathy accounted for much
Similar to intensive lifestyle intervention of the incident microvascular disease and neu-
programs, bariatric surgery can also bring ropathy had a very low incidence. It remains
about sustained weight loss. While bariatric unknown if bariatric surgery can prevent or
surgery can result in the remission T2DM, it is treat existing DPN, and one must also consider
not known if it can be an effective treatment the risks of surgery that include subacute axonal
for DPN. The clinical trials of a lifestyle neuropathy due to micronutrient deficiencies.

Diabetic Neuropathy
270 16. Strategies for the prevention or reversal of neuropathy

Dietary interventions Multifactorial treatment

There is a lack of evidence from clinical Hypertension plays a role in the microvas-
trials in humans on the effect of diet on DPN. cular dysfunction seen in diabetes through an
Evidence from animal studies is also lacking, unclear mechanism or interplay among several
but mice fed with a high-fat diet provide an mechanisms. Overall, it is thought that hyper-
animal model of obesity and insulin resistance tension exerts an effect on pathways that con-
that can be used to study metabolic syndrome tribute to an overall state of oxidative stress,
and the development of DPN. In a dietary inflammation, and ultimately disruption of
study of the effects of a ketogenic diet, mice normal cellular functioning and alterations to
fed with a ketogenic diet had improvement in the endoneurial vasculature. One of the path-
neuropathy based on an increase in epidermal ways that is likely involved is the renin-
axon density compared to mice fed with a angiotensin system. Angiotensin-converting
high-fat diet or mice fed with a high-fat diet enzyme (ACE) inhibitors improve vascular
and subjected to exercise [58]. The ketogenic dysfunction and promote vasodilation by pre-
diet both prevented and reversed mechanical venting the production of angiotensin II, which
allodynia and preserved IENFD [58]. Despite a is a pressor, and preventing the breakdown of
high fat content in both the high-fat and keto- bradykinin, which is a vasodilator. ACE inhibi-
genic diets, the ketogenic diet appears to affect tors have been shown to delay progression of
peripheral nerves differently and promote both nephropathy and retinopathy in humans,
axon growth, but this requires further study. and animal studies have found an improve-
There is considerable interest in the effect of ment in measures of diabetic neuropathy
lipid metabolism and chronic inflammation on [62,63]. In humans, over 9000 individuals with
the pathogenesis of DPN. Animal studies indi- IGT and cardiovascular disease or cardiovascu-
cate that high-fat diets can cause neuropathy in lar risk factors were randomized to treatment
diabetes and that withdrawing or manipulating with either valsartan, an angiotensin-receptor
a high-fat diet can reduce neuropathy [58 60]. blocker, or placebo in addition to lifestyle mod-
In humans, a nutritional intervention study that ification [64]. The participants were followed
was targeted to improve essential fatty acid dys- for 5 years and the cumulative incidence of
metabolism in T1DM showed improvement in diabetes was 33.1% in the valsartan group com-
neuropathy [61]. In this uncontrolled open-label pared with 36.8% in the placebo group (HR
trial, patients with T1DM and DPN were given 0.86, CI 0.80 0.92; P , .001). There was no sig-
supplementation with seal oil omega-3 polyun- nificant change in the incidence of cardiovascu-
saturated fatty acids. After 12 months of supple- lar outcomes between the two groups.
mentation, the participants were found to have There have been several small studies that
a 29% increase in corneal nerve fiber length but have examined the effects of ACE inhibitors on
there was no improvement found in other mea- DPN. A small randomized, double-blind,
sures of neuropathy. Despite there being no placebo-controlled trial of diabetic patients
improvement in other neuropathy outcomes, with mild DPN found that treatment with tran-
there was no observed progression of clinical dolopril was associated with improvements in
disease symptoms over the year-long study and nerve conduction studies but not in the VPT,
there were no declines in small and large fiber neuropathic symptoms, or deficits [65]. There
sensory and functional measures. This lack of was also no improvement in autonomic func-
progression may be significant, but the study tion. Another ACE inhibitor, quinapril, has
lacked a control group. been shown to increase parasympathetic

Diabetic Neuropathy
 Treatment based on pathogenic concepts 271
activity in DAN [66]. A study of the effects of modifiable. The ADDITION-Denmark study
combination therapy with an ACE inhibitor enrolled over 1500 patients with newly diagnosed
and a third-generation dihydropyridine cal- T2DM and randomized them to intensive multi-
cium channel blocker found a positive effect on factorial treatment or routine care. After 6 years,
DPN. Therapy with an ACE inhibitor (delapril) there was no statistically significant difference in
alone or in combination with a calcium channel the prevalence of DPN. However, there was no
blocker (manidipine) for 3 years was found to baseline neuropathy assessments included at the
reduce progression of diabetic neuropathy in time of enrollment and the two groups had little
200 patients with T2DM and hypertension [67]. to no differences in glycemic or other metabolic
The odds ratio for combined therapy was 0.45 measurements. The multicenter ADDITION-
(0.24 0.87, P 5 .017) and for ACE inhibitors Europe study followed over 3000 patients from
was 0.52 (0.27 0.99, P 5 .048). Although these the original ADDITION trial for 5 years to deter-
results are encouraging, the true effects of ACE mine the effect of the multifactorial treatment on
inhibitors on DPN remain unknown. Many microvascular complications [71]. Compared to
diabetic patients require treatment of hyperten- routine care, the intense multifactorial treatment
sion and ACE inhibitors are often a first-line in early T2DM did not significantly reduce the
treatment in this patient population. frequency of microvascular events at 5 years.
Few studies have examined the effect of DPN, which was assessed by questionnaire, was
multifactorial cardiovascular risk intervention present in 4.9% of patients in the intensive care
on DPN. The Steno 2 study was an open, paral- group and 5.9% of patients receiving routine
lel trial of an intensive multifactorial interven- care. Similar to the original ADDITION trial,
tion versus standard care in patients with there was a lack of baseline assessment of neu-
T2DM [68,69]. The intervention was an aggres- ropathy and both the intervention and standard
sive, stepwise implementation of lifestyle and care groups achieved similar improvements in
pharmacologic therapies (including renin- cardiovascular disease risk factors, which might
angiotensin system blockers) that focused on account for the lack of differences in DPN preva-
hypertension and lipid abnormalities as well as lence between the two groups.
exercise, smoking cessation and glucose con-
trol. There was no effect from the intensive
multifactorial intervention on the progression Treatment based on pathogenic concepts
of relatively advanced somatic DPN as mea-
sured by vibration detection threshold, which The underlying pathogenesis of DPN involves
is not a sensitive measure of neuropathy. various pathologic pathways (glucose entry into
However, there was a reduced risk of auto- the polyol pathway, oxidative stress, advanced
nomic neuropathy in the intensive therapy glycation end-product (AGE) formation, micro-
group compared to those in the conventional vascular ischemia, or adipocyte derived toxicity)
therapy group. that are each a potential therapeutic target.
In contrast, the ADDITION study examined Several of these targeted treatments have been
the effect of an early multifactorial intervention in evaluated through randomized clinical trials
participants with newly diagnosed T2DM [70]. with varying levels of evidence, but in general
This was done to see if the beneficial effects of unequivocal evidence from phase 3 trials is lack-
the intensive multifactorial intervention utilized ing. Targeted therapies are not ideal due to the
in the Steno 2 study on cardiovascular mortality multitude of mechanisms underlying the devel-
might be amplified if implemented earlier in the opment and progression of DPN. However, they
course of T2DM when the risk factors are still do have the advantage of being able to exert

Diabetic Neuropathy
272 16. Strategies for the prevention or reversal of neuropathy

their effects despite patient noncompliance with someone with an improvement of at least 2
a lifestyle intervention or the inability to main- points on the NIS and NIS-LL. After 4 years of
tain normoglycemia. treatment, the rates of clinical responders were
higher and the rates of progressors were lower
with the ALA versus placebo (P 5 .013 and
Alpha-lipoic acid P 5 .025 for the NIS and NIS-LL, respectively)
Alpha-lipoic acid (ALA) is a naturally occur- [76]. However, there was no significant differ-
ring, dietary supplement with antioxidant ence in the primary endpoint and for this rea-
properties and it is an essential cofactor for son the study drug was not approved by the US
mitochondrial bioenergetic enzymes. It has FDA. Post hoc analysis of the NATHAN 1 trial
been studied in clinical trials of DPN because found that improvement and prevention of pro-
oxidative stress is thought to be a major con- gression of the NIS-LL with ALA versus pla-
tributor to the pathogenesis of DPN. Several cebo was predicted by higher age, lower BMI,
metaanalyses suggest that ALA is an effective male sex, normal blood pressure, history of car-
and safe drug for the treatment of symptomatic diovascular disease, insulin treatment, longer
DPN [72 74]. The SYDNEY 2 trial was a ran- duration of diabetes and neuropathy, and
domized, double-blind, placebo-controlled trial higher neuropathy stage [77]. The authors con-
of different doses of once daily, oral doses of cluded that better outcomes in neuropathic
either 600, 1200, or 1800 mg ALA in patients impairments with treatment of mild to moder-
with existing DPN [75]. A statistically signifi- ate DPN with ALA 600 mg daily was predicted
cant reduction in the Total Symptom Score by optimal control of cardiovascular risk factors
(TSS) compared to placebo was seen for each in patients with more severe neuropathy.
dose of ALA after 5 weeks. The TSS decreased The effects of ALA on CAN in patients with
by 51% in the 600 mg group, 48% in the T2DM were examined by the Deutsche
1200 mg group, and 52% in the 1800 mg group Kardiale Autonome Neuropathie (DEKAN)
compared to 32% in the placebo group (P , .05 Study [78]. This randomized, double-blind,
vs placebo). Significant improvements were placebo-controlled multicenter trial assigned
also found in the Neuropathy Symptoms and patients with T2DM and a reduced heart rate
Change score and a lower score was found for variability at baseline to treatment with 800 mg
the Neuropathy Impairment Score (NIS). All ALA orally or placebo. After 4 months of treat-
three ALA groups had improvement in neuro- ment, the ALA was well tolerated and there
pathic symptoms and deficits; however, side- were small, but not statistically significant
effects were greater with higher ALA doses, improvements in components of the cardiac
and thus the recommended dose of ALA is autonomic spectral analysis in the ALA-treated
600 mg once daily. patients compared to placebo.
The long-term efficacy and safety of ALA
600 mg once daily in patients with mild to mod- Sirtuin 1 (SIRT1) and diabetic
erate DPN was evaluated in the NATHAN 1
trial. This multicentered, randomized, double-
polyneuropathy
blind, parallel-group trial suggested that ALA Activators of SIRT1 have been shown to
may improve neuropathic deficits in the long prime the mitochondrion by increasing its
term with good tolerability. The primary end- reserve capacity to combat mitochondrial stress
point was a composite score of the NIS-lower under diabetic conditions [79 81]. Recent
limb (NIS-LL) and seven neurophysiological research has shown that SIRT1 is decreased in
tests and a clinical responder was defined as diabetic dorsal root ganglion neurons and

Diabetic Neuropathy
 Treatment based on pathogenic concepts 273
overexpression can increase axonal outgrowth Role of NAD1 analogs in diabetic
in the peripheral nerve [79]. SIRT1 promotion of polyneuropathy
axonal repair would be critical in repairing
degenerating axons in diabetic neuropathy. The precursors of NAD1 are nicotinamide
Neuronal overexpression of SIRT1 enhances the (NAM), nicotinic acid (NA), NR, and NMN.
SIRT1/PGC-1α axis in neurons and protects The reduction in NAD1 induced in diabetes is
against DPN. SIRT1 is upregulated by the due to increased DNA damage, stimulating
NAD1 precursors nicotinamide mononucleotide PARP1 activity [83]. Long-term treatment with
(NMN) or nicotinamide riboside (NR; Fig. 16.3) NAM is detrimental because they increase
and is able to prevent and reverse DPN through development of a fatty liver, due to reductions
improving Mt oxidative metabolism [79 82]. in available methyl groups [84]. NAD1 cannot

FIGURE 16.3 Potential mechanisms by which nicotinamide riboside (NR) or nicotinamide mononucleotide (NMN)
can prevent DPN. The nuclear factors that regulate the expression of nDNA-encoded genes that regulate Mt metabolism
and organelle biogenesis are SIRT1 and peroxisome proliferator-activated receptor γ coactivator-1 (PGC-1α). PGC-1α
serves as a central component of the transcriptional regulatory circuitry that coordinately controls the energy-generating
functions of mitochondria in accordance with the metabolic demands imposed by changing physiological conditions and
disease. NAD1 is depleted in diabetes. NR or NMN is a source of NAD1 that deacetylates SIRT1, activates PGC-1α, and
results in improvement in mitochondrial function in the diabetic nerve. Source: Adapted from Chandrasekaran K, Choi J, Arvas
MI, Salimian M, Singh S, Xu S, et al. Nicotinamide mononucleotide administration prevents experimental diabetes-induced cognitive
impairment and loss of hippocampal neurons. Int J Mol Sci 2020;21(11).

Diabetic Neuropathy
274 16. Strategies for the prevention or reversal of neuropathy

be given directly because of its toxic effects improvement in the primary outcome parame-
that include serious hyperglycemia lasting for ter, the Neuropathy Symptom Score, compared
hours. NA often causes severe flushing, medi- to placebo in the per-protocol group, but no sig-
ated by the binding of NA to the GPR109A nificant difference was seen in the intention to
receptor [85,86]. NAD1 precursors that do not treat population (P 5 .055). A 24-month trial of
activate GPR109A [87], for example NR and 300 mg/day benfotiamine compared to placebo
NMN, but still increase NAD1 levels are more in patients with T1DM found no significant dif-
likely to be clinically useful. The dietary sup- ference in peripheral nerve function or soluble
plement NR is a major generator of NAD1, markers of inflammation [93]. However, only a
which activates neuronal protection and regen- few of the enrolled participants had neuropathy
eration pathways [80]. NR is well tolerated and or inflammation at baseline.
orally bioavailable and protects against both
sensory and motor neuropathy in animals
[80,88]. NR and NMN function through
Actovegin
numerous pathways. However, one of the pri- Actovegin is a deproteinized hemoderivative
mary pathways is to replete NAD1, thereby that is extracted from calf blood by ultrafiltration.
deacetylating SIRT1 that in turn activates PGC- It is made up of several hundred bioactive sub-
1 and improves mitochondrial function in dia- stances that exert an effect on many intracellular
betic neuropathy (Fig. 16.3). processes and its neuroprotective effects are
thought to be due in part to its ability to stimu-
late oxygen uptake and consumption. In addi-
Benfotiamine
tion, actovegin has been shown to have an
Benfotiamine is a lipid-soluble derivative of insulin-like effect and results in increased glu-
vitamin B1 that blocks several hyperglycemia- cose oxidation [94]. In animal models of diabetes,
induced pathways that are implicated in the actovegin has been found to improve DPN
development of hyperglycemia-induced vascu- through suppression of poly(ADP-ribose) poly-
lar damage: the hexosamine pathway, the AGE merase activation [95]. A clinical trial in humans
formation pathway, the diacylglycerol-protein randomized 567 patients with DPN to receive 20
kinase C pathway, and NF-kappa B activation daily infusions (2000 mg/day) followed by three
by activating the pentose phosphate pathway tablets (1800 mg/day) a day of actovegin or pla-
enzyme transketolase [89]. Benfotiamine has cebo for 140 days [96]. After 160 days, there was
been shown to increase levels of intracellular a significant improvement seen in the two pri-
thiamine and reduce AGEs that induce experi- mary outcome measures, the TSS of the lower
mental diabetic neuropathy [90]. The BEDIP limbs and VPT in the actovegin group compared
study, a pilot study in early DPN of benfotia- to placebo. There was also a significant improve-
mine 100 mg four times a day for 3 weeks, ment in the sensory nerve function component
found a statistically significant improvement in of the NIS-LL. There were no differences in the
a neuropathy score that included evaluation of incidence of adverse events between the groups.
patients’ symptoms, VPT, and exam findings While these results are promising, long-term,
of neuropathy [91]. The follow-up double- confirmatory trials with inclusion of objective
blind, placebo-controlled, phase III BENDIP neuropathy outcome measures are needed. In
study randomized 165 patients with DPN to 6 addition, the mechanism of action of actovegin is
weeks of treatment with benfotiamine 600 mg still not elucidated and further investigations
a day, benfotiamine 300 mg/day, or placebo into the mode of action with in vitro and in vivo
[92]. The 600 mg/day dose group had a small studies are needed.

Diabetic Neuropathy
 Treatment based on pathogenic concepts 275

Epalrestat Antiinflammatory drugs

Epalrestat is an aldose reductase inhibitor Subclinical inflammation is a characteristic
that is marketed in Japan, China, and India for of T1DM and T2DM and there is a growing
the treatment of DPN. Aldose reductase inhibi- body of evidence that low-grade, chronic
tors were a promising therapeutic agent in DPN inflammation plays an important role in the
due to their effects on the polyol pathway, but pathogenesis of DPN and may be a potential
clinical data from randomized control trials in therapeutic target. Hyperglycemia, along with
DPN is lacking. Aldose reductase is a key insulin resistance and dyslipidemia, leads to
enzyme of the polyol pathway, which in systemic inflammation and a cycle of oxidative
enhanced in the development of DPN, and its and mitochondrial stress that causes cellular
inhibition was hypothesized to be a key compo- damage and perpetuates the cycle. Multiple
nent in the treatment of DPN. Numerous clini- factors are involved in this cascade that ulti-
cal trials with aldose reductase inhibitors have mately produces cytokines and chemokines
been carried out but none of them have been that enhance inflammation. The result of the
approved in the United States due to poor clini- inflammation is an increase in downstream
cal trial designs, lack of a control group, limited oxidative stress and further neuronal damage.
efficacy, or unacceptable adverse events. One A full review of inflammation in diabetic neu-
open-label, controlled trial showed a small but ropathy is outside the scope of this chapter,
significant change from baseline in motor NCV but the role of chronic inflammation as a
in patients with DPN with epalrestat compared potential therapeutic target in DPN will be
to control after 3 years [97]. discussed.
Another aldose reductase inhibitor, ranire- A link between diabetes and inflammation
stat, has also been examined for its effect on was first suggested by epidemiological studies
measures of DPN. A clinical trial of patients that found a correlation between T2DM and
with DPN randomized 549 participants with inflammatory markers such as fibrinogen,
mild to moderate DPN to placebo or 10, 20, or C-reactive protein, interleukin-6, plasminogen
40 mg/day of ranirestat for a year [98]. activator inhibitor-1, and sialic acid [100]. More
Treatment with ranirestat was associated with concrete evidence for the role of inflammation
an improvement in motor nerve function on in T2DM came from the finding that salsalate,
nerve conduction studies, but there was no sta- a salicylate prodrug, improved glycemic con-
tistically significant difference in sensory nerve trol and decreased inflammatory markers in
function compared to placebo. There was also T2DM [101,102]. Salicylate acts on the nuclear
no improvement seen in the modified Toronto factor-kB (NF-kB) pathway, which has been
Clinical Neuropathy Score (mTCNS) or quanti- found to be a key pathway in the inflammatory
tative sensory testing. A more recent double- response seen in diabetic nerves [103]. The NF-
blind, placebo-controlled, phase III study of 557 kB pathway is activated by hyperglycemia,
participants with DPN found a significant oxidative stress, and cytokines and goes on to
increase in tibial motor NCV (0.52 m/s, P 5 .021), trigger an inflammatory response that not only
the median motor nerves, proximal median sen- results in direct cellular injury, but it also regu-
sory nerves, and distal median sensory nerves lates the expression of many inflammatory
in those treated with ranirestat 40 mg/day com- genes (COX-2, NO-synthase, lipoxygenase, and
pared to placebo after 1 year [99]. No significant endothelin-1) [100]. The NF-kB pathway also
differences were found in the mTCNS between regulates the expression of cytokines, including
the two groups. Ranirestat was well tolerated. tumor necrosis factor-α (TNF-α). Activation of

Diabetic Neuropathy
276 16. Strategies for the prevention or reversal of neuropathy

TNF-α activates mitogen-activated protein Conclusion

kinase [104,105], which is found in association
with DPN. Furthermore, inactivation of the For DPN, the most effective interventions
COX-2 gene has been found to prevent periph- are improved glycemic control in T1DM and
eral nerve dysfunction [104]. lifestyle interventions in T2DM. Several medi-
Observations in human DPN mirror those in cations show promise in modifying DPN but
animals. In DPN, compared to diabetics without require confirmation with large phase 3 trials.
neuropathy and control subjects, there is an In future trials, the endpoints need to be care-
increase in serum levels of inflammatory cyto- fully selected to ensure adequate sensitivity to
kines, including TNF-α [106]. In addition to the change; therapies should be introduced earlier
increase in TNF-α in DPN, levels of the antiin- in the course of DPN to avoid expensive,
flammatory cytokine IL-10 were reduced [107]. extended duration studies; drugs need to tar-
Similarly, a population-based study found that get pathogenetic mechanisms; and therapies
serum levels of several inflammatory cytokines, may be more effective in combination because
including IL-1β and IL-6, were associated with they target different pathogenic mechanisms.
DPN [108]. Further evidence that cytokines may
be a potential target for the treatment of DPN is
the finding that higher levels of IL-1 receptor Acknowledgments
antagonist (IL-1RA) are associated with incident This work was supported in part by the National Institute
DPN and the progression of DPN [109]. of Diabetes and Digestive and Kidney Diseases, National
Dysregulation of heat shock proteins (HSPs), Institutes of Health 1R01DK107007 01A1, Office of
which chaperone cells and protect them from Research Development, Department of Veterans Affairs
(Biomedical and Laboratory Research Service and
environmental stress, has also been found to be Rehabilitation Research and Development, 101RX001030),
associated with DPN. In humans, diabetes has Diabetes Action Research and Education Foundation,
been shown to decrease the expression of HSP70 University of Maryland Institute for Clinical &
and HSP 27, and increasing the expression of Translational Research (ICTR), and the Baltimore GRECC
HSP70 can improve DPN [110]. In animal mod- (JWR), 1K2RX001651 (LAZ).
els, mice overexpressing HSP27 were less likely
to develop DPN. This reduced incidence of neu-
ropathy was correlated with a decrease of NF-kB
References
activation [111]. In humans, however, a cross- [1] Soedamah-Muthu SS, Chaturvedi N, Witte DR, Stevens
sectional study of patients with T1DM found LK, Porta M, Fuller JH, et al. Relationship between risk
factors and mortality in type 1 diabetic patients in
that higher levels of HSP27 correlated with the Europe: the EURODIAB Prospective Complications
presence of DPN [112]. An explanation for this Study (PCS). Diabetes Care 2008;31(7):1360 6.
finding may be that the increased levels of [2] Baltzis D, Roustit M, Grammatikopoulou MG,
HSP27 are a result of a secondary neuroprotec- Katsaboukas D, Athanasiou V, Iakovou I, et al. Diabetic
tive response to the developing DSP [110]. peripheral neuropathy as a predictor of asymptomatic
myocardial ischemia in type 2 diabetes mellitus: a
There is strong support for the role of cross-sectional study. Adv Ther 2016;33(10):1840 7.
inflammation early in the development of both [3] Zilliox LA, Russell JW. Physical activity and dietary
diabetes and DPN. Together with insulin resis- interventions in diabetic neuropathy: a systematic
tance and adiposity, inflammation contributes review. Clin Auton Res 2019;4:443 55.
to ongoing metabolic dysfunction and cytotox- [4] Zilliox LA, Russell JW. Is there cardiac autonomic neurop-
athy in prediabetes? Auton Neurosci 2020;229:102722.
icity that results in peripheral nerve dysfunc- [5] Feldman EL, Callaghan BC, Pop-Busui R, Zochodne
tion. Thus targeting inflammation is a potential DW, Wright DE, Bennett DL, et al. Diabetic neuropathy.
therapy for DPN. Nat Rev Dis Prim 2019;5(1):41.

Diabetic Neuropathy
 References 277
[6] Ang L, Jaiswal M, Martin C, Pop-Busui R. Glucose nerve fiber loss in diabetic neuropathy is not reversed
control and diabetic neuropathy: lessons from recent by long-term normoglycemia after simultaneous pan-
large clinical trials. Curr Diab Rep 2014;14(9):528. creas and kidney transplantation. Am J Transpl
[7] DCCT. Effect of intensive diabetes treatment on nerve 2016;16(7):2196 201.
conduction in the Diabetes Control and Complications [18] Callaghan BC, Little AA, Feldman EL, Hughes RA.
Trial. Ann Neurol 1995;38(6):869 80. Enhanced glucose control for preventing and treating
[8] DCCT. The effect of intensive treatment of diabetes on diabetic neuropathy. Cochrane Database Syst Rev
the development and progression of long-term com- 2012;6:CD007543. Available from: https://doi.org/
plications in insulin-dependent diabetes mellitus. 10.1002/14651858.CD007543.pub2.:CD007543.
Diabetes Control Complications Trial Res Group N [19] UK Prospective Diabetes Study (UKPDS) Group.
Engl J Med 1993;329:977 86. Intensive blood-glucose control with sulphonylureas
[9] Albers JW, Herman WH, Pop-Busui R, Feldman EL, or insulin compared with conventional treatment and
Martin CL, Cleary PA, et al. Effect of prior intensive risk of complications in patients with type 2 diabetes
insulin treatment during the Diabetes Control and (UKPDS 33). Lancet. 1998;352(9131):837 53.
Complications Trial (DCCT) on peripheral neuropathy [20] Duckworth W, Abraira C, Moritz T, Reda D,
in type 1 diabetes during the Epidemiology of Emanuele N, Reaven PD, et al. Glucose control and
Diabetes Interventions and Complications (EDIC) vascular complications in veterans with type 2 diabe-
Study. Diabetes Care 2010;33(5):1090 6. tes. N Engl J Med 2009;360(2):129 39.
[10] Albers JW, Brown MB, Sima AAF, Greene DA. Nerve [21] Ismail-Beigi F, Craven T, Banerji MA, Basile J, Calles J,
conduction measures in mild diabetic neuropathy in Cohen RM, et al. Effect of intensive treatment of
the Early Diabetes Intervention Trial: the effects of hyperglycaemia on microvascular outcomes in type 2
age, sex, type of diabetes, disease duration, and diabetes: an analysis of the ACCORD randomised
anthropometric factors. Neurology. 1996;46:85 91. trial. Lancet. 2010;376(9739):419 30.
[11] Ishibashi F, Taniguchi M, Kosaka A, Uetake H, [22] Ohkubo Y, Kishikawa H, Araki E, Miyata T, Isami S,
Tavakoli M. Improvement in neuropathy outcomes Motoyoshi S, et al. Intensive insulin therapy prevents
with normalizing HbA1c in patients with type 2 diabe- the progression of diabetic microvascular complica-
tes. Diabetes Care 2019;42(1):110 18. tions in Japanese patients with non-insulin-dependent
[12] Ziegler D, Behler M, Schroers-Teuber M, Roden M. diabetes mellitus: a randomized prospective 6-year
Near-normoglycaemia and development of neuropa- study. Diabetes Res Clin Pract 1995;28:103 17.
thy: a 24-year prospective study from diagnosis of [23] Azad N, Emanuele NV, Abraira C, Henderson WG,
type 1 diabetes. BMJ Open 2015;5(6):e006559. Colwell J, Levin SR, et al. The effects of intensive gly-
[13] Chen Z, Miao F, Paterson AD, Lachin JM, Zhang L, cemic control on neuropathy in the VA cooperative
Schones DE, et al. Epigenomic profiling reveals an asso- study on type II diabetes mellitus (VA CSDM). J
ciation between persistence of DNA methylation and Diabetes Complications 1999;13:307 13.
metabolic memory in the DCCT/EDIC type 1 diabetes [24] Pop-Busui R, Lu J, Brooks MM, Albert S, Althouse
cohort. Proc Natl Acad Sci U S A 2016;113(21):E3002 11. AD, Escobedo J, et al. Impact of glycemic control strat-
[14] Tavakoli M, Mitu-Pretorian M, Petropoulos IN, Fadavi egies on the progression of diabetic peripheral neu-
H, Asghar O, Alam U, et al. Corneal confocal micros- ropathy in the Bypass Angioplasty Revascularization
copy detects early nerve regeneration in diabetic neu- Investigation 2 Diabetes (BARI 2D) Cohort. Diabetes
ropathy after simultaneous pancreas and kidney Care 2013;36(10):3208 15.
transplantation. Diabetes. 2013;62(1):254 60. [25] Pop-Busui R, Low PA, Waberski BH, Martin CL,
[15] Mehra S, Tavakoli M, Kallinikos PA, Efron N, Boulton Albers JW, Feldman EL, et al. Effects of prior intensive
AJ, Augustine T, et al. Corneal confocal microscopy insulin therapy on cardiac autonomic nervous system
detects early nerve regeneration after pancreas trans- function in type 1 diabetes mellitus: the Diabetes
plantation in patients with type 1 diabetes. Diabetes Control and Complications Trial/Epidemiology of
Care 2007;30(10):2608 12. Diabetes Interventions and Complications study
[16] Azmi S, Jeziorska M, Ferdousi M, Petropoulos IN, (DCCT/EDIC). Circulation. 2009;119(22):2886 93.
Ponirakis G, Marshall A, et al. Early nerve fibre regen- [26] Gaede P, Vedel P, Parving HH, Pedersen O. Intensified
eration in individuals with type 1 diabetes after simul- multifactorial intervention in patients with type 2 diabe-
taneous pancreas and kidney transplantation. tes mellitus and microalbuminuria: the Steno type 2 ran-
Diabetologia. 2019;62(8):1478 87. domised study. Lancet. 1999;353(9153):617 22.
[17] Havrdova T, Boucek P, Saudek F, Voska L, [27] Polydefkis M, Hauer P, Sheth S, Sirdofsky M, Griffin
Lodererova A, Uceyler N, et al. Severe epidermal JW, McArthur JC. The time course of epidermal nerve

Diabetic Neuropathy
278 16. Strategies for the prevention or reversal of neuropathy

fibre regeneration: studies in normal controls and in [41] Van Acker K, Bouhassira D, De Bacquer D, Weiss S,
people with diabetes, with and without neuropathy. Matthys K, Raemen H, et al. Prevalence and impact on
Brain. 2004;127:1606 15. quality of life of peripheral neuropathy with or with-
[28] Zilliox L, Russell JW. Treatment of diabetic sensory out neuropathic pain in type 1 and type 2 diabetic
polyneuropathy. Curr Treat Options Neurol 2011;13 patients attending hospital outpatients clinics.
(2):143 59. Diabetes Metab 2009;35(3):206 13.
[29] Smith AG, Russell J, Feldman EL, Goldstein J, Peltier [42] Russell JW, Kaminsky AJ. Oxidative injury in diabetic
A, Smith S, et al. Lifestyle intervention for pre-diabetic neuropathy. In: Opara E, editor. Nutrition and diabe-
neuropathy. Diabetes Care 2006;29(6):1294 9. tes: pathophysiology and management. Boca Raton:
[30] Zilliox LA, Ruby SK, Singh S, Zhan M, Russell JW. Taylor & Francis; 2005, p. 381 97.
Clinical neuropathy scales in neuropathy associated [43] Vincent AM, Olzmann JA, Brownlee M, Sivitz WI,
with impaired glucose tolerance. J Diabetes Russell JW. Uncoupling proteins prevent glucose-
Complications 2015;15:10. induced neuronal oxidative stress and programmed
[31] Zilliox L, Peltier AC, Wren PA, Anderson A, Smith AG, cell death. Diabetes. 2004;53:726 34.
Singleton JR, et al. Assessing autonomic dysfunction in [44] Russell JW, Golovoy D, Vincent AM, Mahendru P,
early diabetic neuropathy: the Survey of Autonomic Olzmann JA, Mentzer A, et al. High glucose-induced
Symptoms. Neurology. 2011;76(12):1099 105. oxidative stress and mitochondrial dysfunction in neu-
[32] Singleton JR, Smith AG, Bromberg MB. Increased rons. FASEB 2002;16(13):1738 48.
prevalence of impaired glucose tolerance in patients [45] Gong Q, Gregg EW, Wang J, An Y, Zhang P, Yang W,
with painful sensory neuropathy. Diabetes Care et al. Long-term effects of a randomised trial of a 6-
2001;24:1448 53. year lifestyle intervention in impaired glucose toler-
[33] Bongaerts BW, Rathmann W, Kowall B, Herder C, ance on diabetes-related microvascular complications:
Stockl D, Meisinger C, et al. Postchallenge hyperglyce- the China Da Qing Diabetes Prevention Outcome
mia is positively associated with diabetic polyneuro- Study. Diabetologia. 2011;54(2):300 7.
pathy: the KORA F4 study. Diabetes Care 2012;35 [46] Knowler WC, Barrett-Connor E, Fowler SE, Hamman
(9):1891 3. RF, Lachin JM, Walker EA, et al. Reduction in the inci-
[34] Ziegler D, Voss A, Rathmann W, Strom A, Perz S, dence of type 2 diabetes with lifestyle intervention or
Roden M, et al. Increased prevalence of cardiac auto- metformin. N Engl J Med 2002;346(6):393 403.
nomic dysfunction at different degrees of glucose [47] Diabetes Prevention Program Research G. Long-term
intolerance in the general population: the KORA S4 effects of lifestyle intervention or metformin on diabe-
survey. Diabetologia. 2015;58(5):1118 28. tes development and microvascular complications
[35] Dyck PJ, Clark VM, Overland CJ, Davies JL, Pach JM, over 15-year follow-up: the Diabetes Prevention
Dyck PJ, et al. Impaired glycemia and diabetic poly- Program Outcomes Study. Lancet Diabetes Endocrinol
neuropathy: the OC IG Survey. Diabetes Care 2012;35 2015;3(11):866 75.
(3):584 91. [48] Look ARG. Effects of a long-term lifestyle modifica-
[36] Low PA, Denq JC, Opfer-Gehrking TL, Dyck PJ, tion programme on peripheral neuropathy in over-
O’Brien PC, Slezak JM. Effect of age and gender on weight or obese adults with type 2 diabetes: the Look
sudomotor and cardiovagal function and blood pres- AHEAD study. Diabetologia. 2017;60(6):980 8.
sure response to tilt in normal subjects. Muscle Nerve [49] Balducci S, Iacobellis G, Parisi L, Di Biase N,
1997;20(12):1561 8. Calandriello E, Leonetti F, et al. Exercise training can
[37] Russell JW, Zilliox LA. Diabetic neuropathies. modify the natural history of diabetic peripheral neu-
Continuum (Minneap Minn) 2014;20(5 Peripheral ropathy. J Diabetes Complications 2006;20(4):216 23.
Nerv Syst Disord):1226 40. [50] Kluding PM, Pasnoor M, Singh R, Jernigan S, Farmer
[38] Costa LA, Canani LH, Lisbôa HR, Tres GS, Gross JL. K, Rucker J, et al. The effect of exercise on neuropathic
Aggregation of features of the metabolic syndrome is symptoms, nerve function, and cutaneous innervation
associated with increased prevalence of chronic complica- in people with diabetic peripheral neuropathy. J
tions in Type 2 diabetes. Diabet Med 2004;21(3):252 5. Diabetes Complications 2012;26(5):424 9.
[39] Smith AG, Rose K, Singleton JR. Idiopathic neuropa- [51] Singleton JR, Marcus RL, Jackson JE, Lessard K,
thy patients are at high risk for metabolic syndrome. J Graham TE, Smith AG. Exercise increases cutaneous
Neurol Sci 2008;273(1 2):25 8. nerve density in diabetic patients without neuropathy.
[40] Smith AG, Singleton JR. Obesity and hyperlipidemia Ann Clin Transl Neurol 2014;1(10):844 9.
are risk factors for early diabetic neuropathy. J [52] Singleton JR, Marcus RL, Lessard MK, Jackson JE,
Diabetes Complications 2013;27(5):436 42. Smith AG. Supervised exercise improves cutaneous

Diabetic Neuropathy
 References 279
reinnervation capacity in metabolic syndrome [64] McMurray JJ, Holman RR, Haffner SM, Bethel MA,
patients. Ann Neurol 2015;77(1):146 53. Holzhauer B, Hua TA, et al. Effect of valsartan on the
[53] Kluding PM, Pasnoor M, Singh R, D’Silva LJ, Yoo M, incidence of diabetes and cardiovascular events. N
Billinger SA, et al. Safety of aerobic exercise in people Engl J Med 2010;362(16):1477 90.
with diabetic peripheral neuropathy: single-group [65] Malik RA, Williamson S, Abbott C, Carrington AL,
clinical trial. Phys Ther 2015;95(2):223 34. Iqbal J, Schady W, et al. Effect of angiotensin-
[54] Burr JF, Shephard RJ, Riddell MC. Physical activity in converting-enzyme (ACE) inhibitor trandolapril on
type 1 diabetes mellitus: assessing risks for physical human diabetic neuropathy: randomised double-blind
activity clearance and prescription. Can Fam controlled trial. Lancet. 1998;352(9145):1978 81.
Physician 2012;58(5):533 5. [66] Kontopoulos AG, Athyros VG, Didangelos TP,
[55] Müller-Stich BP, Fischer L, Kenngott HG, Gondan M, Papageorgiou AA, Avramidis MJ, Mayroudi MC,
Senft J, Clemens G, et al. Gastric bypass leads to et al. Effect of chronic quinapril administration on
improvement of diabetic neuropathy independent of heart rate variability in patients with diabetic auto-
glucose normalization results of a prospective cohort nomic neuropathy. Diabetes Care 1997;20(3):355 61.
study (DiaSurg 1 study). Ann Surg 2013;258(5):760 5 [67] Ruggenenti P, Lauria G, Iliev IP, Fassi A, Ilieva AP,
discussion 5 6. Rota S, et al. Effects of manidipine and delapril in
[56] Miras AD, Chuah LL, Khalil N, Nicotra A, Vusirikala hypertensive patients with type 2 diabetes mellitus:
A, Baqai N, et al. Type 2 diabetes mellitus and micro- the delapril and manidipine for nephroprotection in
vascular complications 1 year after Roux-en-Y gastric diabetes (DEMAND) randomized clinical trial.
bypass: a case-control study. Diabetologia. 2015;58 Hypertension. 2011;58(5):776 83.
(7):1443 7. [68] Gaede P, Vedel P, Larsen N, Jensen GV, Parving HH,
[57] Coleman KJ, Haneuse S, Johnson E, Bogart A, Fisher Pedersen O. Multifactorial intervention and cardiovas-
D, O’Connor PJ, et al. Long-term microvascular dis- cular disease in patients with type 2 diabetes. N Engl J
ease outcomes in patients with type 2 diabetes after Med 2003;348(5):383 93.
bariatric surgery: evidence for the legacy effect of sur- [69] Gaede P, Lund-Andersen H, Parving HH, Pedersen O.
gery. Diabetes Care 2016;39(8):1400 7. Effect of a multifactorial intervention on mortality in
[58] Cooper MA, Menta BW, Perez-Sanchez C, Jack MM, type 2 diabetes. N Engl J Med 2008;358(6):580 91.
Khan ZW, Ryals JM, et al. A ketogenic diet reduces [70] Charles M, Ejskjaer N, Witte DR, Borch-Johnsen K,
metabolic syndrome-induced allodynia and promotes Lauritzen T, Sandbaek A. Prevalence of neuropathy
peripheral nerve growth in mice. Exp Neurol and peripheral arterial disease and the impact of treat-
2018;306:149 57. ment in people with screen-detected type 2 diabetes:
[59] Hinder LM, O’Brien PD, Hayes JM, Backus C, Solway the ADDITION-Denmark study. Diabetes Care
AP, Sims-Robinson C, et al. Dietary reversal of neu- 2011;34(10):2244 9.
ropathy in a murine model of prediabetes and meta- [71] Sandbaek A, Griffin SJ, Sharp SJ, Simmons RK, Borch-
bolic syndrome. Dis Model Mech 2017;10(6):717 25. Johnsen K, Rutten GE, et al. Effect of early multifacto-
[60] Coppey L, Davidson E, Shevalye H, Torres ME, Yorek rial therapy compared with routine care on microvas-
MA. Effect of dietary oils on peripheral neuropathy- cular outcomes at 5 years in people with screen-
related endpoints in dietary obese rats. Diabetes detected diabetes: a randomized controlled trial: the
Metab Syndr Obes 2018;11:117 27. ADDITION-Europe Study. Diabetes Care 2014;37
[61] Lewis EJH, Perkins BA, Lovblom LE, Bazinet RP, (7):2015 23.
Wolever TMS, Bril V. Effect of omega-3 supplementa- [72] Mijnhout GS, Kollen BJ, Alkhalaf A, Kleefstra N, Bilo
tion on neuropathy in type 1 diabetes: a 12-month HJ. Alpha lipoic acid for symptomatic peripheral neu-
pilot trial. Neurology. 2017;88(24):2294 301. ropathy in patients with diabetes: a meta-analysis of
[62] Davidson EP, Coppey LJ, Holmes A, Yorek MA. Effect randomized controlled trials. Int J Endocrinol
of inhibition of angiotensin converting enzyme and/or 2012;2012:456279.
neutral endopeptidase on vascular and neural compli- [73] Ziegler D, Nowak H, Kempler P, Vargha P, Low PA.
cations in high fat fed/low dose streptozotocin- Treatment of symptomatic diabetic polyneuropathy
diabetic rats. Eur J Pharmacol 2012;677(1 3):180 7. with the antioxidant alpha-lipoic acid: a meta-analysis.
[63] Oltman CL, Davidson EP, Coppey LJ, Kleinschmidt Diabet Med 2004;21(2):114 21.
TL, Dake B, Yorek MA. Role of the effect of inhibition [74] McIlduff CE, Rutkove SB. Critical appraisal of the use
of neutral endopeptidase on vascular and neural com- of alpha lipoic acid (thioctic acid) in the treatment of
plications in streptozotocin-induced diabetic rats. Eur symptomatic diabetic polyneuropathy. Ther Clin Risk
J Pharmacol 2011;650:556 62. Manag 2011;7:377 85.

Diabetic Neuropathy
280 16. Strategies for the prevention or reversal of neuropathy

[75] Ziegler D, Ametov A, Barinov A, Dyck PJ, Gurieva I, [87] Tunaru S, Kero J, Schaub A, Wufka C, Blaukat A,
Low PA, et al. Oral treatment with alpha-lipoic acid Pfeffer K, et al. PUMA-G and HM74 are receptors for
improves symptomatic diabetic polyneuropathy: the nicotinic acid and mediate its anti-lipolytic effect. Nat
SYDNEY 2 trial. Diabetes Care 2006;29(11):2365 70. Med 2003;9(3):352 5.
[76] Ziegler D, Low PA, Litchy WJ, Boulton AJ, Vinik AI, [88] Trammell SA, Weidemann BJ, Chadda A, Yorek MS,
Freeman R, et al. Efficacy and safety of antioxidant Holmes A, Coppey LJ, et al. Nicotinamide riboside
treatment with alpha-lipoic acid over 4 years in dia- opposes type 2 diabetes and neuropathy in mice. Sci
betic polyneuropathy: the NATHAN 1 trial. Diabetes Rep 2016;6:26933. Available from: https://doi.org/
Care 2011;34(9):2054 60. 10.1038/srep26933.:26933.
[77] Ziegler D, Low PA, Freeman R, Tritschler H, Vinik AI. [89] Hammes HP, Du X, Edelstein D, Taguchi T,
Predictors of improvement and progression of diabetic Matsumura T, Ju Q, et al. Benfotiamine blocks three
polyneuropathy following treatment with alpha-lipoic major pathways of hyperglycemic damage and pre-
acid for 4 years in the NATHAN 1 trial. J Diabetes vents experimental diabetic retinopathy. Nat Med
Complications 2016;30(2):350 6. 2003;9(3):294 9.
[78] Ziegler D, Schatz H, Conrad F, Gries FA, Ulrich H, [90] Balakumar P, Rohilla A, Krishan P, Solairaj P,
Reichel G. Effects of treatment with the antioxidant Thangathirupathi A. The multifaceted therapeutic
alpha-lipoic acid on cardiac autonomic neuropathy in potential of benfotiamine. Pharmacol Res 2010;61
NIDDM patients. A 4-month randomized controlled mul- (6):482 8.
ticenter trial (DEKAN Study). Deutsche Kardiale [91] Haupt E, Ledermann H, Kopcke W. Benfotiamine in
Autonome Neuropathie. Diabetes Care 1997;20(3):369 73. the treatment of diabetic polyneuropathy a three-
[79] Chandrasekaran K, Salimian M, Konduru SR, Choi J, week randomized, controlled pilot study (BEDIP
Kumar P, Long A, et al. Overexpression of Sirtuin 1 study). Int J Clin Pharmacol Ther 2005;43(2):71 7.
protein in neurons prevents and reverses experimental [92] Stracke H, Gaus W, Achenbach U, Federlin K, Bretzel
diabetic neuropathy. Brain. 2019;142(12):3737 52. RG. Benfotiamine in diabetic polyneuropathy
[80] Chandrasekaran K, Anjaneyulu M, Choi J, Kumar P, (BENDIP): results of a randomised, double blind,
Salimian M, Ho CY, et al. Role of mitochondria in dia- placebo-controlled clinical study. Exp Clin Endocrinol
betic peripheral neuropathy: influencing the NAD Diabetes 2008;116(10):600 5.
( 1 )-dependent SIRT1-PGC-1alpha-TFAM pathway. [93] Fraser DA, Diep LM, Hovden IA, Nilsen KB, Sveen KA,
Int Rev Neurobiol 2019;145:177 209. Seljeflot I, et al. The effects of long-term oral benfotiamine
[81] Chandrasekaran K, Choi J, Arvas MI, Salimian M, supplementation on peripheral nerve function and
Singh S, Xu S, et al. Nicotinamide mononucleotide inflammatory markers in patients with type 1 diabetes: a
administration prevents experimental diabetes- 24-month, double-blind, randomized, placebo-controlled
induced cognitive impairment and loss of hippocam- trial. Diabetes Care 2012;35(5):1095 7.
pal neurons. Int J Mol Sci 2020;21(11). [94] Jacob S, Dietze GJ, Machicao F, Kuntz G, Augustin HJ.
[82] Chandrasekaran K, Muragundla A, Demarest TG, Choi J, Improvement of glucose metabolism in patients with
Sagi AR, Najimi N, et al. mGluR2/3 activation of the type II diabetes after treatment with a hemodialysate.
SIRT1 axis preserves mitochondrial function in diabetic Arzneimittelforschung 1996;46(3):269 72.
neuropathy. Ann Clin Transl Neurol 2017;4(12):844 58. [95] Dieckmann A, Kriebel M, Andriambeloson E, Ziegler
[83] Yamamoto H, Uchigata Y, Okamoto H. Streptozotocin D, Elmlinger M. Treatment with Actovegin(R)
and alloxan induce DNA strand breaks and poly improves sensory nerve function and pathology in
(ADP-ribose) synthetase in pancreatic islets. Nature. streptozotocin-diabetic rats via mechanisms involving
1981;294(5838):284 6. inhibition of PARP activation. Exp Clin Endocrinol
[84] Kang-Lee YA, McKee RW, Wright SM, Swendseid ME, Diabetes 2012;120(3):132 8.
Jenden DJ, Jope RS. Metabolic effects of nicotinamide [96] Ziegler D, Movsesyan L, Mankovsky B, Gurieva I,
administration in rats. J Nutr 1983;113(2):215 21. Abylaiuly Z, Strokov I. Treatment of symptomatic
[85] Benyo Z, Gille A, Kero J, Csiky M, Suchankova MC, polyneuropathy with actovegin in type 2 diabetic
Nusing RM, et al. GPR109A (PUMA-G/HM74A) med- patients. Diabetes Care 2009;32(8):1479 84.
iates nicotinic acid-induced flushing. J Clin Invest [97] Hotta N, Akanuma Y, Kawamori R, Matsuoka K, Oka Y,
2005;115(12):3634 40. Shichiri M, et al. Long-term clinical effects of epalrestat,
[86] Bogan KL, Brenner C. Nicotinic acid, nicotinamide, an aldose reductase inhibitor, on diabetic peripheral
and nicotinamide riboside: a molecular evaluation of neuropathy: the 3-year, multicenter, comparative Aldose
NAD 1 precursor vitamins in human nutrition. Annu Reductase Inhibitor-Diabetes Complications Trial.
Rev Nutr 2008;28:115 30. Diabetes Care 2006;29(7):1538 44.

Diabetic Neuropathy
 References 281
[98] Bril V, Hirose T, Tomioka S, Buchanan R. Ranirestat activated protein kinases in the etiology of diabetic
for the management of diabetic sensorimotor poly- neuropathy. FASEB J 2001;15(13):2508 14.
neuropathy. Diabetes Care 2009;32(7):1256 60. [106] Doupis J, Lyons TE, Wu S, Gnardellis C, Dinh T, Veves
[99] Sekiguchi K, Kohara N, Baba M, Komori T, Naito Y, A. Microvascular reactivity and inflammatory cytokines
Imai T, et al. Aldose reductase inhibitor ranirestat in painful and painless peripheral diabetic neuropathy.
significantly improves nerve conduction velocity in J Clin Endocrinol Metab 2009;94(6):2157 63.
diabetic polyneuropathy: a randomized double-blind [107] Duksal T, Tiftikcioglu BI, Bilgin S, Kose S, Zorlu Y.
placebo-controlled study in Japan. J Diabetes Investig Role of inflammation in sensory neuropathy in predi-
2019;10(2):466 74. abetes or diabetes. Acta Neurol Scand 2016;133(5):
[100] Pop-Busui R, Ang L, Holmes C, Gallagher K, 384 90.
Feldman EL. Inflammation as a therapeutic target for [108] Herder C, Bongaerts BW, Rathmann W, Heier M,
diabetic neuropathies. Curr Diab Rep 2016;16(3):29. Kowall B, Koenig W, et al. Association of subclinical
[101] Goldfine AB, Silver R, Aldhahi W, Cai D, Tatro E, inflammation with polyneuropathy in the older pop-
Lee J, et al. Use of salsalate to target inflammation in ulation: KORA F4 study. Diabetes Care 2013;36(11):
the treatment of insulin resistance and type 2 diabe- 3663 70.
tes. Clin Transl Sci 2008;1(1):36 43. [109] Herder C, Kannenberg JM, Huth C, Carstensen-
[102] Goldfine AB, Buck JS, Desouza C, Fonseca V, Chen Kirberg M, Rathmann W, Koenig W, et al.
YD, Shoelson SE, et al. Targeting inflammation using Proinflammatory cytokines predict the incidence and
salsalate in patients with type 2 diabetes: effects on progression of distal sensorimotor polyneuropathy:
flow-mediated dilation (TINSAL-FMD). Diabetes KORA F4/FF4 study. Diabetes Care 2017;40(4):
Care 2013;36(12):4132 9. 569 76.
[103] Wang Y, Schmeichel AM, Iida H, Schmelzer JD, Low [110] Dobrowsky RT. Targeting the diabetic chaperome to
PA. Enhanced inflammatory response via activation improve peripheral neuropathy. Curr Diab Rep
of NF-kappaB in acute experimental diabetic neurop- 2016;16(8):71.
athy subjected to ischemia-reperfusion injury. J [111] Korngut L, Ma CH, Martinez JA, Toth CC, Guo GF,
Neurol Sci 2006;247(1):47 52. Singh V, et al. Overexpression of human HSP27 pro-
[104] Kellogg AP, Wiggin TD, Larkin DD, Hayes JM, tects sensory neurons from diabetes. Neurobiol Dis
Stevens MJ, Pop-Busui R. Protective effects of 2012;47(3):436 43.
cyclooxygenase-2 gene inactivation against periph- [112] Gruden G, Bruno G, Chaturvedi N, Burt D,
eral nerve dysfunction and intraepidermal nerve Schalkwijk C, Pinach S, et al. Serum heat shock pro-
fiber loss in experimental diabetes. Diabetes. 2007;56 tein 27 and diabetes complications in the EURODIAB
(12):2997 3005. prospective complications study: a novel circulating
[105] Purves T, Middlemas A, Agthong S, Jude EB, marker for diabetic neuropathy. Diabetes. 2008;57(7):
Boulton AJ, Fernyhough P, et al. A role for mitogen- 1966 70.

Diabetic Neuropathy
 C H A P T E R

17
Treatment of diabetic peripheral
neuropathy: technologies, exercise, and
alternative treatments
Kalliopi Pafili and Nikolaos Papanas
Diabetes Centre-Diabetic Foot Clinic, Second Department of Internal Medicine, Democritus University
of Thrace, University Hospital of Alexandroupolis, Alexandroupolis, Greece

Introduction favorably combined with technology, but the

efficacy of this combination remains to be proven
Diabetic peripheral neuropathy (DPN) remains [10]. Finally, alternative treatments are now
a devastating microvascular complication of dia- being increasingly appreciated, especially due to
betes [1] with current guidelines pointing to lim- their more holistic, individual, and preventative
ited pharmacologic treatment options [2], mainly nature [11].
targeting symptomatic alleviation, optimization of
glycemic control and improvement of cardiovascu-
Technologies
lar risk factors [1,3]. Although disease-modifying
agents have been examined in clinical trials, clini-
Technologies targeting improvement in
cally meaningful benefits were few or applied to
restricted populations with special baseline charac-
quality of life and medical adherence
teristics [4,5]. Hence, technology seems a modern The contribution of new technologies to the
and appealing strategy targeting alleviation of neu- treatment of DPN remains an appealing, albeit
ropathic symptoms, improved medical adherence, questionable issue. The recently published
personalized offloading, and monitoring of physi- Diabetes Telephone Study has tested whether
cal activity [6,7]. Furthermore, exercise has recently automated verbal communication between 1179
emerged as an additional potential therapeutic humans with symptomatic DPN at the point of
option [8], especially in the context of prior evi- beginning treatment might improve quality of
dence pointing to association of the other diabetic life and contribute to alleviation of neuropathic
microvascular complications, but not DPN, with symptoms [12]. Interactive voice response tech-
exercise capacity [9]. Intriguingly, exercise can be nology (a total of three calls) over 6 months

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00008-6 283 © 2022 Elsevier Inc. All rights reserved.
284 17. Treatment of diabetic peripheral neuropathy: technologies, exercise, and alternative treatments

failed to prove beneficial [12]. During a more phone-based thermal camera [20] and the wire-
intensified approach, individuals with painful less, sensor-embedded continuous temperature
DPN were randomized to receive either standard monitoring socks for use in the home environ-
care or the latter plus twice-daily diabetes self- ment [21]. In a similar context [22], 129 diabetic
management text messages for 6 months [13]. persons with previously healed DFU were
The latter aimed to provide education on diabe- instructed to simply step on a wireless thermo-
tes, address lifestyle modifications, glucose moni- metric smart mat based on the telehealth con-
toring and medical adherence, whereas 20% of cept, which could address plantar temperature
them emphasized individual foot care. There variations. An asymmetry of .2.2 C (.4 F)
were no significant differences in terms of pain between corresponding sites of the 2 feet soles
and improvement of glycated hemoglobin correctly predicted pending DFU with 97% sensi-
(HbA1c), but text messaging was related to tivity but a modest specificity (,45%) [22]. Of
increased self-management activities and amelio- note, evidence supporting the cost-effectiveness
rated diabetes-related health beliefs [13]. of approaches implementing at home tempera-
ture monitoring is currently missing [23].

Technologies to identify high-risk for

impendent diabetic foot ulceration Technologies targeting personalized
offloading
Prevention is better than cure in diabetes melli-
tus [14]. Aiming to reduce recurrence of diabetic High plantar pressure has been traditionally
foot, artificial intelligence technology on mobile recognized as a risk factor for the development
devices has recently been applied to automatically of DFUs and, indeed, humans with history of
detect and localize diabetic foot ulceration (DFU) DFU frequently exhibit high foot pressure [24].
with a precision approaching 92% [15]. In the Hence, studies focusing on barefoot strain anal-
same context, thermography is increasingly being ysis have been conducted, the majority focus-
appreciated as a further modality contributing to ing on the forefoot and the assessment of
timely detection of incipient foot damage [16]. Via ventrical plantar stress [25 28]. Both barefoot
implementation of thermography, high tissue tem- peak pressure and peak plantar pressure above
perature reflecting inflammatory response can be .200 kPa have been reported as independent
detected and monitored [17]. Lavery et al. con- predictors of future neuropathic DFU [29].
ducted one of the leading studies including 173 Nonetheless, it seems plausible that site-
individuals with a history of DFU [18]. Daily use specific plantar pressure may be more relevant
of infrared skin thermometry on 6 foot sites for 15 in the prediction of DFUs. Indeed, in the study
months was compared with standard therapy by Ledoux et al., increased baseline peak meta-
alone. In the case of temperature differences tarsal plantar pressure was associated with a
.2.2 C ( . 4 F) between left and right corre- significant risk of subsequent metatarsal ulcer-
sponding sites, participants were instructed to ation over a mean period of 2.4 years, but peak
immediately contact the study nurse and reduce plantar pressure in distinct foot sites including
ambulation [18]. An almost fivefold lower risk for the heel, midfoot, and the hallux was unable to
development of DFU was reported in the temper- predict DFU [30].
ature monitoring group. These results were not More recently, in-shoe analysis has been
replicated in a subsequent study [19]. implemented to detect elevated plantar pres-
To simplify thermometric evaluation, new sure. In a proof-of-concept study [31], dynamic
devices have been developed, such as the smart- in-shoe plantar pressure evaluation was

Diabetic Neuropathy
 Technologies 285
proven an effective tool to guide footwear therapeutic footwear. Indeed, the latter does not
modifications leading to substantial reduction seem to exceed 30% of total daily activity [39].
of the peak plantar pressure and an almost Even most compliant humans wear the offloading
30% pressure reduction among persons with device no more than 60% of their daily activity
DPN wearing custom-designed footwear. [39]. The home is probably further regarded as a
Subsequent studies have highlighted the clini- comfort zone, with mean adherence approxi-
cal significance of feedback approaches in the mately 60% [40]. Therefore, technology has been
reduction of plantar pressure for the prevention implemented to enable objective monitoring of
of DFU [32 36]. Feedback strategies implement adherence. The temperature-based monitoring
mainly in-shoe technologies to warn of excessive system attached to the inner surface of the device
plantar pressure (usually though visual or acous- (footwear, upper limb or lower limb orthosis) was
tic signs). They aim to make up for the loss of used to assess how long the device was used [41].
pain and to help humans with DPN toward Alternatively, it was combined with a step activity
suitable foot offloading during daily activities monitor to determine use during ambulatory
[34,37]. The first outcomes of this method were activity [41]. Adequate validity of this technology
controversial [33,35]. Nevertheless, these prelimi- has been reported, but almost one-third of partici-
nary studies included a very limited number of pants found the step activity monitor unpleasant
participants and applied a relatively short-term [41]. Similar technology has reported validity
analysis of the intervention effects. In a subse- among healthy participants [42].
quent study implementing wireless, pressure- To further improve adherence, Najafi et al.
sensitive insoles, an improved foot-off loading implemented a 3-month smart insole system
was reported at 6 weeks following training on a designed to prompt offloading in order to avert
new walking strategy to reduce maximal peak high plantar pressures [43]. They attempted to pre-
pressure [36]. This was based on pressure map vent DFU among participants with prior neuro-
analysis as obtained from insoles. Again, only pathic DFU [43]. The device alerted the user
half of steps were below peak pressure threshold through smartwatch auditory, vibratory, and
[36]. More recently, a prospective, randomized, visual feedback when “safe” pressure and time
multidisciplinary study assessed the efficacy of thresholds were exceeded. The procedure
continuous, during daily activities, audiovisual involved the user whenever there was an alert
alerts via a smartwatch linked to the insole sys- [43]. High-pressure areas were identified via
tem and offloading instructions among 58 DPN smartphone technology [43]. Whenever partici-
participants with a recent history of DFU [32]. pants applied offloading within 30 minutes of an
This intervention did not reduce the incidence of alarm, this was defined as a successful response to
DFU overall. However, a subanalysis demon- the alert [43]. A minimum number of alerts (one
strated almost 90% decreased incidence of DFU every 2 hours) was required to enhance adherence
with the intervention among the 40 with ade- and to improve successful offloading [43].
quate compliance [32].

Technologies to monitor physical activity

Technologies to increase adherence to
Early guidelines of the American Diabetes
therapeutic footwear Association (ADA) discouraged walking and
Effective offloading of the diabetic foot is a pre- weight-bearing activities in DPN [44]. The main
requisite toward DFU healing [38]. Nonetheless, concern was the susceptibility of the insensate
evidence points to suboptimal adherence to foot to injuries and, thus the increased risk of

Diabetic Neuropathy
286 17. Treatment of diabetic peripheral neuropathy: technologies, exercise, and alternative treatments

DFUs [44]. This approach changed with the (Table 17.1) point to increased weight-bearing
study by LeMaster et al. in 2008 [45] showing activity among humans who did not previously
that weight-bearing activity does not increase ulcerate [47,49]. However, whether external ambu-
neuropathic DFUs if proper footwear is worn. lation exceeds internal is still controversial [48,51].
Accordingly, the latest ADA guidelines have Importantly, increased activity ($7.5 hours/day)
been rephrased [46]. Individuals with DPN are was reported to reduce the risk of recurrent
no more precluded from physical weight- DFU [50].
bearing activity [46]. However, several issues
remain controversial, including optimal duration
Exercise
of exercise. Technology has now been employed
to provide physical activity feedback (Table 17.1).
Quality of life, neuropathic symptoms,
Following validation of the monitoring devices to
detect weight-bearing activities [52 54], no asso-
and deficits
ciation between physical activity and walking Almost one out of 10 persons with diabetes
capacity was shown [54]. Overall, studies mellitus experience debilitating neuropathic

TABLE 17.1 Characteristics of studies reporting physical activity feedback among humans with diabetic peripheral
neuropathy.
Participants First author
(N) Feedback method (study duration) Amount of activity reported (year)

100 High-capacity computerized Significantly lower average daily activity among Armstrong
accelerometer/pedometer (minimum of 25 individuals who ulcerated as compared to those (2004) [47]
weeks or until ulceration) who did not
20 Activity monitoring worn on the Restricted ambulation indoors, more external Armstrong
waistband, measuring number of steps activity among humans with diabetes, (2001) [48]
taken over a period of time (1 week) neuropathy, deformity, or a history of lower-
extremity ulceration or partial foot amputation
30 Two-dimensional accelerometer (15 days) Humans with diabetes and a previous history of Maluf (2003)
DFU were almost 50% less active as compared [49]
to nondiabetic humans but no difference was
shown in comparison with diabetic humans
without a previous history of plantar ulcer
400 24-hour physical activity questionnaire Following adjustment for confounders, most Lemaster
assessing daily weight-bearing activity (2 active participants presented a significantly (2003) [50]
years) reduced risk of recurrent DFU as compared to
the least active
10 Tri-axial accelerometer combined with DFU participants’ weight-bearing activity was Crews (2017)
GPS monitoring (3 days) almost double indoors than away from home. [51]
At-home activity was similar between at-risk
and active DFU participants. At-risk
participants had 132% more external weight-
bearing activity as compared to humans with
active DFU

DFU, diabetic foot ulceration; GPS, global positioning system.

Diabetic Neuropathy
 Exercise 287
symptoms [55]. Loss of sensation and increased interfere with physical activity [54], improvement
vibration perception threshold represent important of the gait pattern is expected to optimize the
deficits of DPN, increasing the risk of DFUs [56]. functional outcomes of exercise without increas-
There is no consensus on improving neuropathic ing DFUs [65]. Thus a number of studies have
deficits [57]. Thus lifestyle interventions might addressed the association between exercise and
prove beneficial. A randomized controlled trial improvement in physical status in DPN. In a ran-
examined the effects of 8-week aerobic exercise of domized controlled trial [67,68], 12-week exercise
modest intensity on quality of life among 29 for foot-ankle and gait training (40 60 minutes,
humans with DPN and 37 controls [58]. Exercise twice a week) was associated with a significant
significantly improved quality of life, pain scores, improvement of the foot rollover process during
sensory symptoms, restriction in activity of daily gait. An improved eccentric control of forefoot
life and disruptions in social relationships [58]. contact, increased participation of hallux and toes,
The same research group then examined the and increase in overall foot and ankle function
ability of this training intervention to improve were observed [67] in comparison with the con-
vibration perception [59]. Neuropathic symptoms trol group. Nevertheless, these effects were not
were further assessed in the context of a random- sustainable during the following 12 months (24
ized controlled trial of 31 DPN humans [60]. months from baseline), highlighting the need for
These were assigned to either perform cycle train- continuous training.
ing for 12 weeks (50% 70% of heart rate reserve, Similarly, simple home-based exercises were
30 45 minutes, 3 sessions/week) or to continue evaluated in 51 DPN humans, as compared
without training [60]. A significant improvement with 53 controls [69]. The intervention group
of neuropathic symptoms (assessed with the was instructed to perform 10-minute exercises
Michigan Diabetic Neuropathy Score) was including eight exercises for both hands and
reported with exercise intervention [60]. Moreover, four exercises for both feet, 3 times daily for 8
exercise was associated with reduced HbA1c [60]. weeks [69]. The exercise group significantly
Finally, in a clinical trial among persons .60 improved motor function (assessed with a spe-
years of age with type 2 diabetes mellitus cial 5-point Likert scale) and specific activities
(T2DM), 12 weeks of plantiflexor and dorsiflexor of daily living, such as climbing stairs and per-
muscle strengthening exercises with resistance forming household chores [69]. In a more
bands, proprioceptive exercises on balance boards intensified approach, a 12-week group-specific
and a buoy combined with plantar sensory stimu- progressive balance, flexibility, strengthening,
lation with bristle brushes and cloths significantly and aerobic weight-bearing exercise (60 min-
improved plantar cutaneous sensibility of both utes, 3 times/week) was compared with the
feet (as evaluated with Semmes-Weinstein mono- same nonweight-bearing training in a group of
filament) [61]. Of note, limitations of this study 29 participants with a mean age of 64.5 years
include the lack of assessment of body weight [70]. In this group of older DPN humans,
parameters and glycemic control. Importantly, weight-bearing activity significantly improved
evidence form randomized controlled trials is the 6-minute walk distance (between group
only scarce [62 65], pointing to the need for difference reached 29 m) and daily step counts
future well-designed studies [66]. (between group difference was almost 1200
steps) [70]. The clinical relevance of extra 1200
steps daily remains to be identified, but it
Physical capacity appears important granted that persons aged
Although previous evidence has implied between 55 and 95 years have a mean daily
that the level of physical capacity does not physical activity of almost 4000 steps/day [71].

Diabetic Neuropathy
288 17. Treatment of diabetic peripheral neuropathy: technologies, exercise, and alternative treatments

Balance 73 years) [75]. Postural sway, trunk reposition-

ing errors, dynamic and static balance signifi-
Impaired balance represents an important risk cantly improved with the intervention [75].
factor for falls, especially among older DPN The risk of falls was not studied.
populations [72]. Therefore a number of studies More recently, wearable technology has
have examined the effects of exercise interven- been combined with balance exercise (45 min-
tions on balance parameters and risk for falls in utes, twice a week for 4 weeks) to improve pos-
DPN and some systematic analyses have been tural stability in DPN humans (mean age 64
conducted (Table 17.2). The effects of a lower- years) [76]. Body-worn sensors were imple-
extremity exercise and walking intervention (8 mented to acquire kinematic data and provide
individual sessions with a physical therapist, real-time joint visual feedback during training
plus exercises at home, 60 minutes, 3 times/ [76]. Significant improvement of postural bal-
week) were examined among 79 DPN partici- ance was shown [76].
pants with a mean age of approximately 65 years
in randomized controlled trial [73]. At the end of
12 months, intervention resulted in a small
increase in the amount of time that participants
Nerve conduction parameters
could perform a one leg stand with their eyes Whether exercise is able to modify the natu-
closed [73]. Conversely, there were no differ- ral history of DPN has been a matter of discus-
ences in other balance parameters or fall rates sion. A 4-year supervised brisk walking on
[73]. However, this study was limited by a signif- treadmill (heart-rate reserve 50% 85%, 4
icant decline of lower-extremity muscle strength, times/week, 60 minutes) in DPN participants
pointing to an insufficient intensity of the inter- (mean age 49 years), led to a significant
vention to provide benefit [73]. improvement of nerve conduction velocity of
In another work, 8-week sensorimotor train- both peroneal and sural nerves and to a
ing, including wall slides, core exercises, balance decreased risk of DPN development [77].
exercises on unstable surface and gait training Further evidence pointed to the favorable
(65 80 minutes, 3 times/week), improved static effects of aerobic exercise mainly on sural
and dynamic balance as well as proprioception nerve conduction velocity of DPN humans but
measures among middle-aged and older humans less so on peroneal nerve conduction [78,79].
[74]. The risk of falls was not studied [74]. More recently, aerobic exercise, isokinetic
Morrison et al. examined the ability of a 12-week strength and their combination 3 times/week
moderate (heart-rate reserve 50%, 3 times/week, for 12 weeks, were compared in a randomized
45 minutes), or intense (heart-rate reserve 70%, 3 controlled design among 45 DPN persons with
times/week, 30 minutes) supervised aerobic a mean age .60 years [80]. Regardless of the
training to improve standing balance and risk of type of exercise, sensory (sural, median, ulnar)
falls among 16 DPN humans and 21 humans and motor (tibial and peroneal) nerve conduc-
with diabetes mellitus but no DPN (age range tion did not improve [80]. Finally, Hung et al.
42 70 years) [72]. Following interventions, both assessed the effects of exercise on nerve func-
groups significantly improved balance metrics, tion in the diabetic hand exclusively [81], a
but the risk of falls was not reduced [72]. rather forgotten diabetes complication [82]. A
In a more focused approach, the effects of 12-week Tai-chi Chuan exercise program sig-
an 8-week balance exercise program (60 min- nificantly increased nerve conduction velocities
utes, twice a week) were investigated among but not amplitudes of the median, tibial, and
38 DPN participants (mean age approximately ulnar nerves [81].

Diabetic Neuropathy
 Exercise 289
TABLE 17.2 Systematic reviews and meta-analyses addressing the effects of exercise interventions on the progression
of diabetic peripheral neuropathy and risk of falls among diabetic humans.

Number of pooled First

trials (number of Eligibility Outcomes of Author
Type(s) of exercise participants) criteria interest Results (year)

Systematic review and meta-analysis

Aerobic exercise, 20 RCTs (1357 RCTs reporting DPN The effect of exercise Liao
resistance exercise, participants) effects of on DPN was not (2019) [62]
combination of both exercise on risk assessed—only one
factors for DFU RCT
Mostly balance and 10 RCTs (889 Exercise Falls-related Exercise interventions Chapman
lower limb strength participants) interventions outcomes (static were more effective (2017) [63]
training assessing falls- and dynamic than the control
related balance, lower limb condition for static
outcomes strength, gait, falls balance, lower-limb
among older rates, and falls strength and gait; no
diabetic risks) RCTs assessed falls-
humans risk; one RCT reported
12-month falls-rate,
with no differential
treatment effect
Systematic review
Moderate- or higher- 12 studies: 4 RCTs, 2 Effect of Nerve function Eleven studies Gu (2019)
intensity exercise, quasiexperimental exercise on the (self-reported reported that exercise [64]
resistance exercise, studies, 1 one pre-post progression and neuropathy scores, training had a positive
balance training, or a parallel group study, 5 development of clinical nerve influence on nerve
combination of pre-post single-group DPN function tests, NCS, function or
exercise modes studies (5743 IENFD) neuropathy-related
participants) symptoms; one study
reported mild adverse
events (pain in the
legs, knees, back of the
hands with exercise,
and pain/itchiness at
the skin biopsy site)
Lower limb 10 studies: 7 RCTs, 1 Effect of Falls risks People with T2DM and Gu (2017)
strengthening, quasiexperimental exercise DPN can improve their [66]
balance practice, study, 1 noncontrolled training on falls balance and walking
aerobic exercise, trial, 1 case-report (352 risk among after a targeted
walking programs, participants) T2DM DPN multicomponent
and Tai Chi humans program without risk
of serious adverse
events

DPN, Diabetic peripheral neuropathy; IENFD, intraepidermal nerve fiber density; NCS, nerve conduction study; RCT, randomized
controlled trial; T2DM, type 2 diabetes mellitus.

Diabetic Neuropathy
290 17. Treatment of diabetic peripheral neuropathy: technologies, exercise, and alternative treatments

Intraepidermal nerve fiber density In a randomized crossover study in T2DM

persons with DPN, a single session of self-
Loss of small unmyelinated nerve fibers is applied Thai foot massage for 25 minutes was
traditionally deemed an early manifestation of equally effective in terms of increasing the range
DPN [83]. Improvement through exercise has of motion of foot and ankle to the same method
hitherto been unanimously supported by vari- applied by a massage therapist [90]. Whole body
ous training interventions [80,84,85]. Indeed, vibration is another alternative intervention [92].
the first study [84] reported increased intraepi- A systematic review has shown its slight positive
dermal nerve fiber branching ( 1 0.11 6 0.15 effect on both neuropathic pain and balance [93].
branch nodes/fiber) after 8 weeks of aerobic However, it included very few studies and had a
and strengthening exercises. These favorable high risk of bias [93].
changes were sustained at 12 months [85].
However, it is now known that DPN may even
occur in prediabetes [86,87]. Smith et al.
showed that 12-month lifestyle intervention Targeting neuropathic pain
(exercise plus diet) in humans with impaired Neuropathic pain remains a major clinical
glucose tolerance resulted in cutaneous rein- concern, negatively affecting the quality of life
nervation of both distal leg and proximal thigh, [1]. Nonpharmacological approaches, including
as well as reduced neuropathic pain [88]. electrotherapy, have been investigated for the
treatment of painful DPN [94]. Although the
underlying mechanisms remain obscure, poten-
tial improvement of microcirculation [95] and
Alternative treatments interference with pain gait control [96] have been
suggested. An initial study of frequency-
Targeting balance modulated electromagnetic neural stimulation
Despite promising molecules [2,57], there is cur- for a maximum of 3 weeks [97] suggested a sig-
rently an unmet need for improvement of treat- nificant improvement of daytime and nocturnal
ment efficacy, for instance in balance. Foot plantar visual analog pain score and quality of life, as
pressure massage (classic massage and deep fric- compared with placebo. In this study, four elec-
tion massage at the right and left foot dorsum, trodes administering monophase-compensated
medial, lateral regions of the foot, the toes, and the negative potential electrical pulses were applied
plantar region) for 10 minutes by a trained physio- to the lower extremities for 30 minutes and parti-
therapist has been shown to improve balance cipants were required to discontinue all analge-
immediately after the intervention [89]. However, sics 3 weeks prior to study entry [97].
this study had no control group [89]. Thai foot Importantly, these beneficial effects persisted
massage is another type of deep massage using during a 4-month follow-up [97]. Conversely, the
thumb, finger, palm, or elbow pressure [90,91]. It is same therapeutic intervention applied for 3
applied along the meridian lines of the foot and weeks in a double-blind, randomized, placebo-
leg and is combined with toe distraction [90,91]. controlled clinical trial [96], significantly reduced
Chatchawan et al. examined the effects of daytime and nocturnal pain, but beneficial effects
30-minute Thai foot massage, 3 days weekly in a were not sustained at 3 months following the
randomized, parallel-controlled trial among T2DM intervention. These two studies selected very dia-
participants with DPN [91]. There was a significant betic populations: the former included partici-
improvement of balance after 2 weeks of Thai mas- pants with severe DPN [97], whereas the latter
sage [91]. relatively mild symptomatic DPN [96].

Diabetic Neuropathy
 Discussion 291
It has long been known that hyperglycemia- Following 25-minute sessions over a 12-week
related nonenzymatic collagen glycation may period (twice weekly for the first week and
increase diameter and rigidity of ligaments weekly thereafter), no significant pain improve-
and fascias [98,99]. Thus surgical decompres- ment was shown [107]. However, all participants
sion of entrapped nerves in the lower extremi- were permitted to continue analgesics and low
ties has been attempted to reduce neuropathic pain scores were reported at baseline [107].
pain. Initial attempts that microsurgical In addition, herbal products seem attractive,
decompression of the common peroneal nerve, given that they have fewer untoward effects
deep peroneal nerve, and posterior tibial nerve [108]. A recent Cochrane database systematic
resulted in .50% pain relief at 2 weeks postop- review has included two studies reporting out-
eratively, with sustainable results even after 2 comes on both DPN and nondiabetic neuro-
years [100]. Favorable changes were also seen pathic pain [108]. These assessed two herbal
in physical capacity, social function, and men- medicinal products, namely nutmeg nasal-
tal health [100]. Of note, a limitation of this spray (125 mL applied topically for 4 weeks)
study and previous reports [101] was the fail- and St John’s wort capsules (900 mg total
ure to control for the nonoperated limb. hypericin, ingested 3 times daily for 5 weeks)
A further randomized controlled trial with a [108]. The authors concluded that this evidence
12-month follow-up assessed the benefit of uni- was insufficient to determine the efficacy of
lateral decompression of the tibial common these herbal products for neuropathic pain
peroneal, deep peroneal, and superficial pero- alleviation [108]. Finally, acupuncture therapy
neal nerves in painful DPN [102]. Significant may improve pain of DPN, but relevant studies
pain relief was achieved by approximately two are characterized by varying methodologies
out of five humans in the operated leg [102]. and outcomes [109].
Noninvasive brain neurostimulation techni-
ques have been used to inhibit pain perception
in DPN [103 105]. Daily application of this
technique (deep H-coil repetitive transcranial Discussion
magnetic stimulation applied via a helmet
placed on the participant’s scalp at the point at Technologies have been used to identify per-
which a minimum magnetic field was able to sons at high risk of DFU. Evidence supports the
cause a motor response of the tibialis anterior benefits of home foot temperature monitoring
muscle) for 5 consecutive days was well toler- [18], but this technology appears to provide the
ated and significantly reduced neuropathic maximal benefit only in humans consistently
pain, as well as the threshold of RIII nocicep- wearing therapeutic footwear [19] and its cost-
tive flexion reflex [103], a measure of pain per- effectiveness is unknown [23]. Conversely,
ception [104]. Nevertheless, changes were not conflicting results have been achieved with tech-
sustainable 3 weeks later, in line with a more nology in the improvement of quality of life in
recent study [105]. DPN [12,13]. Furthermore, dynamic in-shoe plan-
Furthermore, Reiki therapy was examined tar pressure evaluation has emerged as a promis-
in a randomized, semidouble-blind, placebo- ing approach to guide footwear modification,
controlled 12-week trial [106]. Reiki therapy is significantly reducing mechanical stress [31].
classed as a bioenergy field therapy [106]. It is Nevertheless, implementation of feedback visual
characterized as a spiritual practice without or acoustic approaches targeting personalized off-
any religion references and involves a practi- loading was not very consistent [32 37] and was
tioner guiding energy to a receiver [107]. most beneficial in compliant persons [32]. Of

Diabetic Neuropathy
292 17. Treatment of diabetic peripheral neuropathy: technologies, exercise, and alternative treatments

note, improvement of adherence to therapeutic offloading. Importantly, adequate compliance is a

footwear through technology is promising [43]. prerequisite. Furthermore, exercise emerges as an
Exercise has also been reported to improve additional therapeutic option with beneficial
the quality of life and mitigate neuropathic effects even in prediabetes. Finally, alternative
symptoms [58,59,61]. However, the role of treatments provide additional option but we
improved glycemic control [59] or use of addi- need long-term, well-designed studies.
tional modalities [61] needs to be clarified, while
evidence from randomized controlled trials is
limited [62 64,66]. Furthermore, the effect of References
exercise on physical status was overall beneficial [1] Bonhof GJ, Herder C, Strom A, Papanas N, Roden M,
[67 70], but continuous training was needed for Ziegler D. Emerging biomarkers, tools, and treatments for
sustainable benefits [67]. Importantly, balance diabetic polyneuropathy. Endocr Rev 2019;40(1):153 92.
[2] Papanas N, Ziegler D. Emerging drugs for diabetic
was improved with different exercise protocols peripheral neuropathy and neuropathic pain. Expert
among populations with different ages [72 76]. Opin Emerg Drugs 2016;21(4):393 407.
Nevertheless, future research needs to examine [3] Pop-Busui R, Boulton AJM, Feldman EL, Bril V,
whether falls may also be reduced. Surprisingly, Freeman R, Malik RA, et al. Diabetic neuropathy: a
exercise seems to also reverse small nerve fiber position statement by the American Diabetes
Association. Diabetes Care 2017;40(1):136 54.
damage [84,85], even in prediabetes [88], while [4] Ziegler D, Edmundson S, Gurieva I, Mankovsky B,
its effect on improvement of nerve conduction Papanas N, Strokov I. Predictors of response to treat-
parameters (large nerve fiber function) is less cer- ment with actovegin for 6 months in patients with
tain [77 80]. type 2 diabetes and symptomatic polyneuropathy. J
Among alternative treatments, foot plantar Diabetes Complications 2017;31(7):1181 7.
[5] Ziegler D, Low PA, Freeman R, Tritschler H, Vinik AI.
pressure massage [89] and Thai foot massage Predictors of improvement and progression of diabetic
[90,91] have improved balance, but studies were polyneuropathy following treatment with alpha-lipoic
limited by short follow-up [89 91]. Electrotherapy acid for 4 years in the NATHAN 1 trial. J Diabetes
has yielded contradictory findings in terms of neu- Complications 2016;30(2):350 6.
ropathic pain reduction [96,97]. It may be more [6] Lung CW, Wu F-L, Liao F, Pu F, Fan Y, Jan Y-K.
Emerging technologies for the prevention and man-
effective in severe DPN [97]. Surgical decompres- agement of diabetic foot ulcers. J Tissue Viability
sion of entrapped nerves has shown promising 2020;29(2):61 8.
and more durable results [100,102]. Noninvasive [7] Chatwin KE, Abbott CA, Boulton AJM, Bowling FL,
brain neurostimulation appears to offer only tem- Reeves ND. The role of foot pressure measurement in
porary benefit [103,105]. Finally, Reiki therapy the prediction and prevention of diabetic foot ulcera-
tion—a comprehensive review. Diabetes Metab Res
[107], herbal products [108] and acupuncture [109] Rev 2020;36(4):e3258.
have been investigated, but sufficient evidence of [8] Vas PRJ, Pafili K, Papanas N. Exercise to improve dia-
their efficacy has not yet accumulated. betic peripheral neuropathy: an additional option?
Neurophysiol Clin 2018;48(4):191 3.
[9] Estacio RO, Regensteiner JG, Wolfel EE, Jeffers B,
Dickenson M, Schrier RW. The association between
Conclusion diabetic complications and exercise capacity in
NIDDM patients. Diabetes Care 1998;21(2):291 5.
DPN remains a frequently overlooked diabetes [10] Ferreira J, Cruvinel RH, Junior, Silva EQ, Verı́ssimo
complication [110,111]. In this context, technology JL, Monteiro RL, et al. Study protocol for a random-
provides modern possibilities, for instance foot ized controlled trial on the effect of the diabetic foot
guidance system (SOPeD) for the prevention and
temperature monitoring, dynamic in-shoe plantar treatment of foot musculoskeletal dysfunctions in peo-
pressure evaluation and implementation of feed- ple with diabetic neuropathy: the FOotCAre (FOCA)
back approaches targeting personalized trial. I Trials 2020;21(1):73.

Diabetic Neuropathy
 References 293
[11] Baute V, Zelnik D, Curtis J, Sadeghifar F. temperature monitoring in reducing the incidence of
Complementary and alternative medicine for painful foot ulcer recurrence in patients with diabetes
peripheral neuropathy. Curr Treat Options Neurol (DIATEMP): study protocol for a randomized con-
2019;21(9):44. trolled trial. Trials 2018;19(1):520.
[12] Adams AS, Schmittdiel JA, Altschuler A, Bayliss EA, [24] Lazzarini PA, Crews RT, van Netten JJ, Bus SA,
Neugebauer R, Ma L, et al. Automated symptom and Fernando ME, Chadwick PJ, et al. Measuring plantar
treatment side effect monitoring for improved quality tissue stress in people with diabetic peripheral neu-
of life among adults with diabetic peripheral neuropa- ropathy: a critical concept in diabetic foot manage-
thy in primary care: a pragmatic, cluster, randomized, ment. J Diabetes Sci Technol 2019;13(5):869 80.
controlled trial. Diabet Med 2019;36(1):52 61. [25] Caselli A, Pham H, Giurini JM, Armstrong DG, Veves
[13] Bauer V, Goodman N, Lapin B, Cooley C, Wang E, A. The forefoot-to-rearfoot plantar pressure ratio is
Craig TL, et al. Text messaging to improve disease increased in severe diabetic neuropathy and can pre-
management in patients with painful diabetic periph- dict foot ulceration. Diabetes Care 2002;25(6):1066 71.
eral neuropathy. Diabetes Educ 2018;44(3):237 48. [26] Fernando M, Crowther R, Lazzarini P, Sangla K,
[14] Smith J. Diabetes: prevention is better than cure. Br J Cunningham M, Buttner P, et al. Biomechanical charac-
Community Nurs 2010;15:5. teristics of peripheral diabetic neuropathy: a systematic
[15] Goyal M, Reeves ND, Rajbhandari S, Yap MH. Robust review and meta-analysis of findings from the gait cycle,
methods for real-time diabetic foot ulcer detection and muscle activity and dynamic barefoot plantar pressure.
localization on mobile devices. IEEE J Biomed Health Clin Biomech (Bristol, Avon) 2013;28(8):831 45.
Inform 2019;23(4):1730 41. [27] Lavery LA, Armstrong DG, Wunderlich RP, Tredwell
[16] Pafili K, Papanas N. Thermography in the follow up J, Boulton AJM. Predictive value of foot pressure
of the diabetic foot: best to weigh the enemy more assessment as part of a population-based diabetes dis-
mighty than he seems. Expert Rev Med Devices ease management program. Diabetes Care 2003;26(4):
2015;12(2):131 3. 1069 73.
[17] Bergtholdt HT, Brand PW. Thermography: an aid in [28] Veves A, Murray HJ, Young MJ, Boulton AJ. The risk
the management of insensitive feet and stumps. Arch of foot ulceration in diabetic patients with high foot
Phys Med Rehabil 1975;56(5):205 9. pressure: a prospective study. Diabetologia 1992;35
[18] Lavery LA, Higgins KR, Lanctot DR, Constantinides (7):660 3.
GP, Zamorano RZ, Athanasiou KA, et al. Preventing [29] Waaijman R, de Haart M, Arts MLJ, Wever D.
diabetic foot ulcer recurrence in high-risk patients: use Verlouw AJWE, Nollet F, et al. Risk factors for plantar
of temperature monitoring as a self-assessment tool. foot ulcer recurrence in neuropathic diabetic patients.
Diabetes Care 2007;30(1):14 20. Diabetes Care 2014;37(6):1697 705.
[19] Skafjeld A, Iversen MM, Holme I, Ribu L, Hvaal K, [30] Ledoux WR, Shofer JB, Cowley MS, Ahroni JH, Cohen
Kilhovd BK. A pilot study testing the feasibility of V, Boyko EJ. Diabetic foot ulcer incidence in relation
skin temperature monitoring to reduce recurrent foot to plantar pressure magnitude and measurement loca-
ulcers in patients with diabetes a randomized con- tion. J Diabetes Complications 2013;27(6):621 6.
trolled trial. BMC Endocr Disord 2015;15:55. [31] Bus SA, Haspels R, Busch-Westbroek TE. Evaluation
[20] van Doremalen RFM, van Netten JJ, van Baal JG, and optimization of therapeutic footwear for neuro-
Vollenbroek-Hutten MMR, van der Heijden F. pathic diabetic foot patients using in-shoe plantar
Validation of low-cost smartphone-based thermal pressure analysis. Diabetes Care 2011;34(7):1595 600.
camera for diabetic foot assessment. Diabetes Res Clin [32] Abbott CA, Chatwin KE, Foden P, Hasan AN, Sange
Pract 2019;149:132 9. C, Rajbhandari SM, et al. Innovative intelligent insole
[21] Reyzelman AM, Koelewyn K, Murphy M, Shen X, Yu system reduces diabetic foot ulcer recurrence at plan-
E, Pillai R, et al. Continuous temperature-monitoring tar sites: a prospective, randomised, proof-of-concept
socks for home use in patients with diabetes: observa- study. Lancet Dig Health 2019;1(6):e308 18.
tional study. J Med Int Res 2018;20(12):e12460. [33] De León Rodriguez D, Allet L, Golay A, Philippe J,
[22] Frykberg RG, Gordon IL, Reyzelman AM, Cazzell SM, Assal J-Ph, Hauert C-A, Pataky Z. . Biofeedback can
Fitzgerald RH, Rothenberg GM, et al. Feasibility and effi- reduce foot pressure to a safe level and without caus-
cacy of a smart mat technology to predict development of ing new at-risk zones in patients with diabetes and
diabetic plantar ulcers. Diabetes Care 2017;40(7):973 80. peripheral neuropathy. Diabetes Metab Res Rev
[23] An de Stegg WB, Mejaiti N, van Netten JJ, Dijkgraaf 2013;29(2):139 44.
MGW, van Baal JG, Busch-Westbroek TE, et al. The [34] Pataky Z, Faravel L, Da Silva J, Assal J. A new ambu-
cost-effectiveness and cost-utility of at-home infrared latory foot pressure device for patients with sensory

Diabetic Neuropathy
294 17. Treatment of diabetic peripheral neuropathy: technologies, exercise, and alternative treatments

impairment. A system for continuous measurement of [46] American Diabetes Association. 2. Classification and
plantar pressure and a feed-back alarm. J Biomech diagnosis of diabetes: standards of medical care in
2000;33(9):1135 8. diabetes-2019. Diabetes Care 2019;42(Suppl 1):S13 28.
[35] York RM, Perell-Gerson KL, Barr M, Durham J, Roper [47] Armstrong DG, Lavery LA, Holtz-Neiderer K, Mohler
JM. Motor learning of a gait pattern to reduce forefoot MJ, Wendel CS, Nixon BP, et al. Variability in activity
plantar pressures in individuals with diabetic periph- may precede diabetic foot ulceration. Diabetes Care
eral neuropathy. PMR 2009;1(5):434 41. 2004;27(8):1980 4.
[36] Van GH, Haddad J, Bensimon Y, Dillard L, Jacobs D. [48] Armstrong DG, Abu-Rumman PL, Nixon BP, Boulton
Biofeedback strategy and reduction of foot pressure in AJ. Continuous activity monitoring in persons at high
diabetic patient with neuropathy: benefit of connected risk for diabetes-related lower-extremity amputation. J
insoles FEETME: a preliminary study. Ann Phys Am Pod Assoc 2001;91(9):451 5.
Rehab Med 2017;60:e40. [49] Maluf KS, Mueller MJ. Comparison of physical activ-
[37] Pataky Z, de León Rodriguez D, Allet L, Golay A, ity and cumulative plantar tissue stress among sub-
Assal M, Assal J-P, et al. Biofeedback for foot offload- jects with and without diabetes mellitus and a history
ing in diabetic patients with peripheral neuropathy. of recurrent plantar ulcers. Clin Biomech 2003;18(7):
Diabet Med 2010;27(1):61 4. 567 75.
[38] Armstrong DG, Nguyen HC, Lavery LA, van Schie [50] Lemaster JW, Reiber GE, Smith DG, Heagerty PJ,
CH, Boulton AJ, Harkless LB. Off-loading the diabetic Wallace C. Daily weight-bearing activity does not
foot wound: a randomized clinical trial. Diabetes Care increase the risk of diabetic foot ulcers. Med Sci Sports
2001;24(6):1019 22. Ex 2003;35(7):1093 9.
[39] Armstrong DG, Lavery LA, Kimbriel HR, Nixon BP, [51] Crews RT, Yalla SV, Dhatt N, Burdi D, Hwang S.
Boulton AJM. Activity patterns of patients with dia- Monitoring location-specific physical activity via inte-
betic foot ulceration: patients with active ulceration gration of accelerometry and geotechnology within
may not adhere to a standard pressure off-loading patients with or at risk of diabetic foot ulcers: a tech-
regimen. Diabetes Care 2003;26(9):2595 7. nological report. J Diabetes Sci Techol 2017;11(5):
[40] Waaijman R, Keukenkamp R, de Haart M, Polomski 899 903.
WP, Nollet F, Bus SA. Adherence to wearing prescrip- [52] Jao Y-L, Gardner SE, Carr LJ. Measuring weight-
tion custom-made footwear in patients with diabetes bearing activities in patients with previous diabetic
at high risk for plantar foot ulceration. Diabetes Care foot ulcers. J Wound Ost Cont Nurs 2017;44(1):34 40.
2013;36(6):1613 18. [53] Dasanayake IS, Bevier WC, Castorino K, Pinsker JE,
[41] Bus SA, Waaijman R, Nollet F. New monitoring tech- Seborg DE, Doyle 3rd FJ, et al. Early detection of phys-
nology to objectively assess adherence to prescribed ical activity for people with type 1 diabetes mellitus. J
footwear and assistive devices during ambulatory Diabetes Sci Technol 2015;9(6):1236 45.
activity. Arch Phys Med Rehab 2012;93(11):2075 9. [54] van Schie CHM, Noordhof ELJ, Busch-Westbroek TE,
[42] Lutjeboer T, van Netten JJ, Postema K, Hijmans JM. Beelen A, Nollet F. Assessment of physical activity in
Validity and feasibility of a temperature sensor for people with diabetes and peripheral neuropathy.
measuring use and non-use of orthopaedic footwear. J Diabetes Res Clin Pract 2011;92(1):e9 e11.
Rehabil Med 2018;50(10):920 6. [55] Argoff C, Cole BE, Fishbain DA, Irving GA. Diabetic
[43] Najafi B, Ron E, Enriquez A, Marin I, Razjouyan J, peripheral neuropathic pain: clinical and quality-of-
Armstrong DG. Smarter sole survival: will neuro- life issues. Mayo Clin Proc 2006;81(4 Suppl):S3 11.
pathic patients at high risk for ulceration use a smart [56] Abbott CA, Vileikyte L, Williamson S, Carrington AL,
insole-based foot protection system? J Diabetes Sci Boulton AJ. Multicenter study of the incidence of and
Technol 2017;11(4):702 13. predictive risk factors for diabetic neuropathic foot
[44] Sigal RJ, Kenny GP, Wasserman DH, Castaneda- ulceration. Diabetes Care 1998;21(7):1071 5.
Sceppa C, White RD. Physical activity/exercise and [57] Ardeleanu V, Toma A, Pafili K, Papanas N, Motofei I,
type 2 diabetes: a consensus statement from the Diaconu CC, et al. Current pharmacological treatment
American Diabetes Association. Diabetes Care 2006;29 of painful diabetic neuropathy: a narrative review.
(6):1433 8. Medicina (Kaunas.) 2020;56(1):25.
[45] Lemaster JW, Mueller MJ, Reiber GE, Mehr DR, [58] Dixit S, Maiya A, Shastry B. Effect of aerobic exercise
Madsen RW, Conn VS. Effect of weight-bearing activ- on quality of life in population with diabetic periph-
ity on foot ulcer incidence in people with diabetic eral neuropathy in type 2 diabetes: a single blind, ran-
peripheral neuropathy: feet first randomized con- domized controlled trial. Qual Life Res 2014;23(5):
trolled trial. Phys Ther 2008;88(11):1385 98. 1629 40.

Diabetic Neuropathy
 References 295
[59] Dixit S, Maiya A, Shastry B. Effects of aerobic exercise [70] Mueller MJ, Tuttle LJ, Lemaster JW, Strube MJ, McGill
on vibration perception threshold in type 2 diabetic JB, Hastings MK, et al. Weight-bearing versus
peripheral neuropathy population using 3-sites nonweight-bearing exercise for persons with diabetes
method: single-blind randomized controlled trial. and peripheral neuropathy: a randomized controlled
Alter Ther Health Med 2019;25(2):36 41. trial. Arch Phys Med Rehab 2013;94(5):829 38.
[60] Gholami F, Nazari H, Alimi M. Cycle Training [71] Tudor-Locke C, Craig CL, Aoyagi Y, Bell RC, Croteau
improves vascular function and neuropathic symp- KA, De Bourdeaudhuij I, et al. How many steps/day
toms in patients with type 2 diabetes and peripheral are enough? For older adults and special populations.
neuropathy: a randomized controlled trial. Exper Int J Behav Nutr Phys Act. 2011;8:80.
Gerontol. 2020;131:110799. [72] Morrison S, Colberg SR, Parson HK, Vinik AI.
[61] Navarro-Peternella FM, Teston EF, Dos Santos Exercise improves gait, reaction time and postural sta-
Santiago Ribeiro BM, Marcon SS. Plantar cutaneous bility in older adults with type 2 diabetes and neurop-
sensory stimulation improves foot sensibility and gait athy. J Diabetes Complications 2014;28(5):715 22.
speed in older adults with diabetes: a clinical trial. [73] Kruse RL, Lemaster JW, Madsen RW. Fall and balance
Adv Skin Wound Care 2019;32(12):568 73. outcomes after an intervention to promote leg
[62] Liao F, An R, Pu F, Burns S, Shen S, Jan Y-K. Effect of strength, balance, and walking in people with diabetic
exercise on risk factors of diabetic foot ulcers: a sys- peripheral neuropathy: “feet first” randomized con-
tematic review and meta-analysis. Am J Phys Med trolled trial. Phys Ther 2010;90(11):1568 79.
Rehab 2019;98(2):103 16. [74] Ahmad I, Noohu MM, Verma S, Singla D, Hussain
[63] Chapman A, Meyer C, Renehan E, Hill KD, Browning ME. Effect of sensorimotor training on balance mea-
CJ. Exercise interventions for the improvement of sures and proprioception among middle and older
falls-related outcomes among older adults with diabe- age adults with diabetic peripheral neuropathy. Gait
tes mellitus: a systematic review and meta-analyses. J Posture 2019;74:114 20.
Diabetes Complications 2017;31(3):631 45. [75] Song CH, Petrofsky JS, Lee SW, Lee KJ, Yim JE. Effects
[64] Gu Y, Dennis SM, Kiernan MC, Harmer AR. Aerobic of an exercise program on balance and trunk proprio-
exercise training may improve nerve function in type ception in older adults with diabetic neuropathies.
2 diabetes and pre-diabetes: a systematic review. Diabetes Technol Ther 2001;13(8):803 11.
Diabetes Metab Res Rev 2019;35(2):e3099. [76] Grewal GS, Schwenk M, Lee-Eng J, Parvaneh S,
[65] Kanade RV, van Deursen RWM, Harding K, Price P. Bharara M, Menzies RA, et al. Sensor-based interactive
Walking performance in people with diabetic neurop- balance training with visual joint movement feedback
athy: benefits and threats. Diabetologia 2006;49(8): for improving postural stability in diabetics with
1747 54. peripheral neuropathy: a randomized controlled trial.
[66] Gu Y, Dennis SM. Are falls prevention programs effec- Gerontology 2015;61(6):567 74.
tive at reducing the risk factors for falls in people with [77] Balducci S, Iacobellis G, Parisi L, Di Biase N,
type-2 diabetes mellitus and peripheral neuropathy: a Calandriello E, Leonetti F, et al. Exercise training
systematic review with narrative synthesis. J Diabetes can modify the natural history of diabetic peripheral
Complications 2017;31(2):504 16. neuropathy. J Diabetes Complications 2006;20(4):
[67] Sartor CD, Hasue RH, Cacciari LP, Butugan MK, 216 23.
Watari R, Pássaro AC, et al. Effects of strengthening, [78] Dixit S, Maiya A, Shastry B. Effect of aerobic exercise
stretching and functional training on foot function in on peripheral nerve functions of population with dia-
patients with diabetic neuropathy: results of a ran- betic peripheral neuropathy in type 2 diabetes: a sin-
domized controlled trial. BMC Musculoskelet Disord gle blind, parallel group randomized controlled trial. J
2014;15:137. Diabetes Complications 2014;28(3):332 9.
[68] Sartor CD, Watari R, Pássaro AC, Picon AP, Hasue [79] Gholami F, Nikookheslat S, Salekzamani Y, Boule N,
RH, Sacco ICN. Effects of a combined strengthening, Jafari A. Effect of aerobic training on nerve conduction
stretching and functional training program vs usual- in men with type 2 diabetes and peripheral neuropa-
care on gait biomechanics and foot function for dia- thy: a randomized controlled trial. Neurophysiol Clin
betic neuropathy: a randomized controlled trial. BMC 2018;48(4):195 202.
Musculoskelet Disord 2012;13:36. [80] Stubbs EB, Fisher MA, Miller CM, Jelinek C, Butler J,
[69] Win M, Fukai K, Nyunt HH, Linn KZ. Hand and foot McBurney C, et al. Randomized controlled trial of
exercises for diabetic peripheral neuropathy: A ran- physical exercise in diabetic veterans with length-
domized controlled trial. Nurs Health Sci 2020;22(2): dependent distal symmetric polyneuropathy. Front
416 26. Neurosci 2019;13:51.

Diabetic Neuropathy
296 17. Treatment of diabetic peripheral neuropathy: technologies, exercise, and alternative treatments

[81] Hung J-W, Liou C-W, Wang P-W, Yeh S-H, Lin L-W, [94] Pieber K, Herceg M, Paternostro-Sluga T. Electrotherapy
Lo S-K, et al. Effect of 12-week tai chi chuan for the treatment of painful diabetic peripheral neuropa-
exercise on peripheral nerve modulation in patients thy: a review. J Rehabil Med 2010;42(4):289 95.
with type 2 diabetes mellitus. J Rehabil Med 2009;41 [95] Jacobs MJ, Jorning PJ, Joshi SR, Kitslaar PJ, Slaaf DW,
(11):924 9. Reneman RS. Epidural spinal cord electrical stimula-
[82] Papanas N, Maltezos E. The diabetic hand: a forgot- tion improves microvascular blood flow in severe
ten complication? J Diabetes Complications 2010;24 limb ischemia. Ann Surg 1988;207(2):179 83.
(3):154 62. [96] Bosi E, Bax G, Scionti L, Spallone V, Tesfaye S,
[83] Pafili K, Trypsianis G, Papazoglou D, Maltezos E, Valensi P, et al. Frequency-modulated electromag-
Papanas N. Correlation of cardiac autonomic neurop- netic neural stimulation (FREMS) as a treatment for
athy with small and large peripheral nerve function symptomatic diabetic neuropathy: results from a
in type 2 diabetes mellitus. Diabetes Res Clin Pract double-blind, randomised, multicentre, long-term,
2019;156:107844. placebo-controlled clinical trial. Diabetologia 2013;56
[84] Kluding PM, Pasnoor M, Singh R, Jernigan S, (3):467 75.
Farmer K, Rucker J, et al. The effect of exercise on [97] Bosi E, Conti M, Vermigli C, Cazzetta G, Peretti E,
neuropathic symptoms, nerve function, and cutane- Cordoni MC, et al. Effectiveness of frequency-
ous innervation in people with diabetic peripheral modulated electromagnetic neural stimulation in the
neuropathy. J Diabetes Complications 2012;26 treatment of painful diabetic neuropathy. Diabetologia
(5):424 9. 2005;48(5):817 23.
[85] Singleton JR, Marcus RL, Jackson JE, Lessard MK, [98] Chao CY, Zheng YP, Cheing GL. Epidermal thickness
Graham TE, Smith AG. Exercise increases cutaneous and biomechanical properties of plantar tissues in dia-
nerve density in diabetic patients without neuropa- betic foot. Ultrasound Med Biol 2011;37(7):1029 38.
thy. Ann Clin Trans Neurol 2014;1(10):844 9. [99] Sun JH, Cheng BK, Zheng Y-P, Huang Y-P, Leung JY,
[86] Papanas N, Vinik AI, Ziegler D. Neuropathy in pre- Cheing GL. Changes in the thickness and stiffness of
diabetes: does the clock start ticking early? Nat Rev plantar soft tissues in people with diabetic peripheral
Endocrinol 2011;7(11):682 90. neuropathy. Arch Phys Med Rehabil 2011;92
[87] Pafili K, Papanas N, Ziegler D. Neuropathy in diabe- (9):1484 9.
tes: “one cannot begin it too soon.”. Angiology [100] Yang W, Guo Z, Yu Y, Xu J, Zhang L. Pain relief and
2018;69(9):752 4. health-related quality-of-life improvement after
[88] Smith AG, Russell J, Feldman EL, Goldstein J, Peltier A, microsurgical decompression of entrapped periph-
Smith S, et al. Lifestyle intervention for pre-diabetic eral nerves in patients with painful diabetic periph-
neuropathy. Diabetes Care 2006;29(6):1294 9. eral neuropathy. J Foot Ankle Surg 2016;55
[89] Tütün Yümin E, Şimşek TT, Sertel M, Ankarali H, (6):1185 9.
Yumin M. The effect of foot plantar massage on bal- [101] Chaudhry V, Russell J, Belzberg A. Decompressive
ance and functional reach in patients with type II dia- surgery of lower limbs for symmetrical diabetic
betes. Physiother Theory Pract 2017;33(2):115 23. peripheral neuropathy. Cochrane Database Syst Rev
[90] Chatchawan U, Jarasrungsichol K, Yamauchi J. 2008;3 CD006152.
Immediate effects of self-Thai foot massage on skin [102] van Macaré Maurik JFM, Oomen RTW, van Hal M,
blood flow, skin temperature, and range of motion of Kon M, Peters EJG. The effect of lower extremity
the foot and ankle in type 2 diabetic patients. J Altern nerve decompression on health-related quality of life
Compl Med 2020;26(6):491 500. and perception of pain in patients with painful dia-
[91] Chatchawan U, Eungpinichpong W, Plandee P, betic polyneuropathy: a prospective randomized
Yamauchi J. Effects of Thai foot massage on balance trial. Diabet Med 2015;32(6):803 9.
performance in diabetic patients with peripheral neu- [103] Onesti E, Gabriele M, Cambieri C, Ceccanti M,
ropathy: a randomized parallel-controlled trial. Med Raccah R, Di Stefano G, et al. H-coil repetitive tran-
Sci Monit Basic Res 2015;21:68 75. scranial magnetic stimulation for pain relief in
[92] Kessler NJ, Hong J. Whole body vibration therapy for patients with diabetic neuropathy. Eur J Pain 2013;17
painful diabetic peripheral neuropathy: a pilot study. (9):1347 56.
J Bodyw Mov Ther 2013;17(4):518 22. [104] Willer JC. Comparative study of perceived pain and
[93] Robinson CC, Barreto RPG, Della, Plentz RM. Effects nociceptive flexion reflex in man. Pain 1977;3
of whole body vibration in individuals with diabetic (1):69 80.
peripheral neuropathy: a systematic review. J [105] Shimizu T, Hosomi K, Maruo T, Goto Y, Yokoe M,
Musculoskel Neur Interact 2018;18(3):382 8. Kageyama Y, et al. Efficacy of deep rTMS for

Diabetic Neuropathy
 References 297
neuropathic pain in the lower limb: a randomized, or preparations for neuropathic pain. Cochrane
double-blind crossover trial of an H-coil and figure-8 Database Syst Rev 2019;4(4) CD010528.
coil. J Neurosurg 2017;127(5):1172 80. [109] Nash J, Armour M, Penkala S. Acupuncture for the
[106] Gillespie EA, Gillespie BW, Stevens MJ. Painful dia- treatment of lower limb diabetic peripheral neuropathy:
betic neuropathy: impact of an alternative approach. a systematic review. Acupunct Med 2019;37(1):3 15.
Diabetes Care 2007;30(4):999 1001. [110] Pafili K, Papanas N, Maltezos E. Treatment of dia-
[107] Billot M, Daycard M, Wood C, Tchalla A. Reiki ther- betic complications: how can we learn by seeking
apy for pain, anxiety and quality of life. BMJ Support and blundering? Angiology 2015;66(4):301 3.
Palliat Care 2019;9(4):434 8. [111] Herman WH, Kennedy L. Underdiagnosis of periph-
[108] Boyd A, Bleakley C, Hurley DA, Gill C, Hannon- eral neuropathy in type 2 diabetes. Diabetes Care
Fletcher M, Bell P, et al. Herbal medicinal products 2005;28(6):1480 1.

Diabetic Neuropathy
 C H A P T E R

18
Treatment of diabetic polyneuropathy
Amanda C. Peltier
Neuromuscular Division, Department of Neurology and Medicine, Vanderbilt University Medical
Center, Nashville, TN, United States

Epidemiology and historical trials in reductase inhibitors, and other agents have
diabetic polyneuropathy been tested without success in preventing fur-
ther progression of DSP [7]. It has been theo-
Diabetic neuropathy is the most common rized that focusing on nerve conduction
cause of distal symmetric polyneuropathy velocity (NCV) and sural amplitude as the out-
(DSP) worldwide [1]. It generates the most come measure of most treatment trials may
morbidity and is linked to significant mortality have been a limiting factor in determining effi-
rates of all diabetic complications [2,3]. DSP is cacy [8]. The evaluation showed the change of
becoming more common with the global pan- sural nerve amplitude in well-controlled T1
demic of diabetes and is expected to affect over and T2DM patients was less than 1 μV per year
2.3 million in the United States alone [4]. DSP [9]. Evaluation of intraepidermal nerve fiber
is the leading contributor to nontraumatic density utilizing minimally invasive punch
amputations in the United States [5]. DSP is biopsies may be more sensitive for the assess-
also highly linked to diabetic autonomic neu- ment of change in neuropathy over time. In
ropathy (DAN), which is a significant predictor addition, utilizing a model of capsaicin-
of increased mortality in patients with both induced denervation to evaluate nerve regener-
types of diabetes [2]. DSP is a large source of ation may be a more sensitive measure [10].
loss productivity and disability in the United Multiple studies were performed on therapy
States. The cost of DSP is 11 14 billion in the based on pathogenetically linked treatments.
United States alone [5,6]. The first wave of these studies was in aldose
Despite the high number of patients affected reductase inhibitors, based on the polyol path-
with DSP and its morbidity, treatment of DSP way in which glucose is converted to sorbitol
has been limited to management of symptoms via aldose reductase, leading to cellular injury
and glycemic control, which benefits type 1 [11]. Aldose reductase inhibitors also pre-
diabetes mellitus (T1DM) patients more than vented change in sodium channel displace-
type 2 diabetes mellitus (T2DM) patients. ment, which may explain its effect on
Multiple trials of growth factors, aldose conduction velocity slowing [12].

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00021-9 299 © 2022 Elsevier Inc. All rights reserved.
300 18. Treatment of diabetic polyneuropathy

The finding of improvement in NCV with Glucose control and diabetic

aldose reductase inhibitors led to 13 large clinical polyneuropathy
trials of aldose reductase inhibitors sorbinul,
ponalrestat, and tolrestat which were negative [7]. The most successful treatment of DPN has
The endpoint for all the trials was NCV which been found to be glycemic control, and this was
did not meet clinical significance in individual much more effective in T1DM patients. The
trials. However, there was a trend, but a pooled Diabetes Control and Complications Trial (DCCT)
meta-analysis showed improvement in peroneal study was the largest study to evaluate the effect
motor conduction velocity and median motor con- of intensive glycemic control on the development
duction velocity but not in sural sensory conduc- of complications in T1DM. A total of 1441 patients
tion velocity [13]. were randomized to conventional versus inten-
Thiocytic acid, an antioxidant also called sive control and followed in the Epidemiology of
α-lipoic acid (ALA), was found to be effective
Diabetes Interventions and Complications (EDIC)
in several studies to help with pain but did not
[19]. The DCCT was stopped early when the par-
significantly improve NCV in the trials. The
ticipants in the intensive group had significantly
Neurological Assessment of Thiocytic Acid in
fewer complications and rates of progression. The
Diabetic Neuropathy (NATHAN) study, which
DCCT/EDIC studies also demonstrated the con-
was a large, multicenter trial of ALA in the
United States, found no significant change in cept of “metabolic memory,” in which patients
NCV and NCS data in the placebo or treated who had been randomized to conventional con-
groups, resulting in a negative result. trol had persistently worse outcomes over time
However, there was a clinical improvement in despite later adherence to intensive glycemic con-
the ALA treated group in the NIS-LL trol. Early treatment was shown to be most effec-
(Neuropathy Impairment Score-lower limb) tive in reducing the rate of complications. The
score [14]. There was also a significant DCCT/EDIC study has also shown that in T1DM
improvement in pain control, for which ALA is after glucose and age, hypertension, elevated tri-
still recommended. glycerides, and elevated LDL cholesterol were
In addition to aldose reductase inhibitors also significant risk factors for the development of
and ALA, growth factors were also studied in DPN and DAN [20].
diabetic polyneuropathy (DPN) with limited However, glucose control for T2DM is not as
success. Recombinant human nerve growth helpful in preventing DPN, DAN, and was actu-
factor (rhNGF) was studied in a large 84 center ally in the ACCORD study linked to higher mor-
randomized controlled study which revealed tality in the intensive control causing the study
no significant benefit from rhNGF despite sig- to be stopped early [21]. The UK Prospective
nificant improvement in animal preclinical Diabetes Study did not use a strict definition of
data [15]. Increased pain, myalgias, and edema DPN and had narrow windows between the two
were seen in the treatment group. An groups which might have decreased the ability
angiotensin-converting enzyme inhibitor, tran- to detect improvement in DPN in this study [22].
dolapril, was also studied in diabetic neuropa- However, the Veteran Affairs Cooperative Study
thy but did not show significant improvement on Glycemic Control and Complications in Type
[16]. Angiotensin-converting enzyme inhibitors 2 Diabetes Study and the Veterans Affairs
have been shown to improve microangiopathy Diabetes Trial did not show improvements in
in animal models [17]. A smaller study had DPN with intensive glucose control [23].
been previously positive for quinapril in DAN It has been theorized that the secondary meta-
and peripheral neuropathy [18]. bolic factors in T2DM drive the lack of

Diabetic Neuropathy
 Serotonin norepinephrine reuptake inhibitors 301
improvement from intensive glycemic control. painful DPN: pregabalin, duloxetine, tapentadol,
Type 2 patients are typically older at the onset of and the new 8% capsaicin patch. All approach
diabetes and have years of prediabetes prior to pain through different mechanisms. See
diagnosis. Prediabetes in itself has been linked to Table 18.1 for the list of first-line medications for
polyneuropathy in itself [24,25]. Obesity and painful DPN [39,40].
chronic inflammation related to obesity is theo-
rized to be yet another mechanism contributing
to DPN in the T2DM population [26]. DPN is Gabapentanoids
typically detected much earlier in the course of
Pregabalin and gabapentin are both alpha2
T2DM and often at the time of diagnosis of dia-
delta calcium channel blockers. By blocking
betes mellitus. DPN in T2DM is also more often
hyperexcitable dorsal root ganglion cells, gaba-
painful [27]. Focus on factors beyond glucose in
pentanoids decrease neurotransmitter release of
T2DM is crucial to improvement in DPN treat-
norepinephrine, glutamate, and aspartate [41,42].
ments. Future directions have included a focus
Gabapentin was originally designed as a GABA
on triglycerides, diet and exercise, and chronic
analog to treat epilepsy and was serendipitously
inflammation [28].
found to treat neuropathic pain. Gabapentin is
Bariatric surgery has some promise in treat-
typically started between 300 and 900 mg daily in
ment of DPN. Several prospective studies have
divided doses. The half-life is short, with peak
shown the benefit of bariatric surgery in reduc-
concentration at 3 hours, requiring 3 4 times a
ing neuropathy scores after surgery [29 33].
day dosing. It also has limited bioavailability at
Care must be taken as bariatric surgery, espe-
higher doses, requiring more frequent dosing to
cially roux-en-y bypass has a risk of peripheral
maintain pain control. The most common side
neuropathy after surgery especially associated
effects are sedation, weight gain, and peripheral
with vitamin B deficiency [34], and rapid
edema. It is limited by renal dysfunction as it is
weight loss can lead to foot drop rarely [35].
renally cleared. It has the significant benefit of
Meta-analyses suggest an overall benefit of bar-
having minimal interaction with other drugs and
iatric surgery in the treatment of DPN [31].
can also treat restless leg syndrome (RLS), com-
This may be due to the reversal of multiple
monly seen in polyneuropathy patients [41,43].
risk factors such as hypertension, elevated cho-
Pregabalin has a longer half-life, can be dosed
lesterol and triglycerides in addition to glucose
twice to three times a day, and has a more linear
control [36]. Bariatric surgery may also reduce
bioavailability. It has the same side effects and
inflammation which is associated with obesity
limitations by renal function [44]. It has a level A
and type 2 diabetes [37].
recommendation from the American Academy of
Neurology’s (AAN’s) evidence-based guidelines
for the treatment of painful DPN [40]. Both drugs
Treatment of pain in diabetic
are Schedule V controlled substances limiting
polyneuropathy
refills and requiring greater monitoring of abuse.
DPN is one of the most common causes of
neuropathic pain and costs $11 14 billion US
Dollars annually in lost wages, direct costs of Serotonin norepinephrine reuptake
health care, and disability. Painful DPN also con- inhibitors
tributes to increased incidence of mood disorders
such as diabetes and anxiety [38]. Currently, there Duloxetine was the first medication approved
are only four FDA approved treatments for for DPN in 2004. It acts to inhibit the reuptake of

Diabetic Neuropathy
302 18. Treatment of diabetic polyneuropathy

TABLE 18.1 Medications for the treatment of painful diabetic neuropathy [37].
American Academy of
Neurology Level of
Medication Typical dose range Common side effects Recommendation Notes

Gabapentin 300 1200 mg three times Sedation, weight gain, B Adjust dose in patients
a day peripheral edema with renal dysfunction
Pregabalin 150 600 mg/d in two to Sedation, weight gain, A Adjust dose in patients
three divided doses peripheral edema with renal dysfunction;
Schedule V controlled
substance
Tricyclic 10 25 mg at bedtime; Anticholinergic: dry B Avoid in patients with a
antidepressants titrate up to a maximum mouth, constipation, history of a prior suicide
(e.g., of 150 mg/d orthostatic attempts; obtain ECG if
amitriptyline, hypotension, urinary titrating to high doses or
nortriptyline) retention sedation, if on concomitant QTc
weight gain prolonging drugs. Also
avoid in individuals with
specific cardiac
conductance
abnormalities
Duloxetine 30 mg/d to 60 mg two Nausea, dizziness, B Avoid in patients with
times a day increased blood hepatic dysfunction, and
pressure, in patients with renal
hyperhidrosis function with GFR below
30 mL/min. 60 mg two
times a day has not been
shown to be more
effective than 60 mg/d
Venlafaxine Immediate release: Nausea, dizziness, B Extended release
37.5 225 mg/d in two to increased blood formulation may be better
three divided doses once pressure, tolerated and less likely to
higher than 37.5 mg/d. hyperhidrosis be associated with a
Extended release: withdrawal syndrome on
37.5 mg/d; increase in discontinuation
37.5 mg increments to
total daily dose of
225 mg/d

ECG, electrocardiogram.
Data from Bril V, England J, Franklin GM, Backonja M, Cohen J, Del Toro D, et al. Evidence-based guideline: treatment of painful diabetic neuropathy:
report of the American Academy of Neurology, the American Association of Neuromuscular and Electrodiagnostic Medicine, and the American Academy
of Physical Medicine and Rehabilitation. Neurology. 2011;76(20):1758 65 (r 2011 American Academy of Neurology); Peltier AC, Wood D.
Management of neuropathic pain in polyneuropathy. Continuum 2020;26(5):1299 1322.

serotonin and norepinephrine. By upregulating patients with coexistent depression. It can be

norepinephrine, it upregulates descending cen- dosed daily, and the typical efficacious dose is
tral pain control pathways using norepinephrine 60 mg daily, although it can be increased to
as the neurotransmitter [45]. Duloxetine is also 120 mg daily. Typical side effects include nausea,
FDA approved for depression and is useful in increased sweating, hypertension, and decreased

Diabetic Neuropathy
 Opioids 303
libido. It should be used with caution with other are thought to be hyperexcitable in painful DPN
serotoninergic medications to avoid serotonin [52]. Genetic mutations in sodium channel 1.7
syndrome. It should also not be stopped gene SCN9A is known to cause erythromelalgia
abruptly to avoid withdrawal symptoms. and insensitivity to pain [53]. However, trials of
Venlafaxine has also been shown to be benefi- epileptic drugs known to target sodium channels
cial in neuropathic pain, although it is thought to have led to minimally positive results. Valproic
be less effective than duloxetine. It has an acid (VPA) has the most efficacy of all antiepilep-
extended-release formulation helpful in painful tic drugs (AEDs) tested in painful DPN [54]. Not
DPN with 32% reduction in pain at the 75 mg only does it affect sodium channels, but also
dose and 50% reduction at the 150 225 mg dose potassium and calcium channels. It also has a
compared to placebo. It has similar side effects to number of significant side effects minimizing its
duloxetine and the risk of withdrawal syndrome use, including weight gain, thrombocytopenia,
[46,47]. pancreatitis, tremor, sedation, hair loss, and signif-
icant teratogenicity. VPA is, therefore, typically a
second-line treatment for DPN. Other AEDs used
Tricyclic antidepressants are carbamazepine and oxcarbamazepine, both
sodium channel blockers and have been used for
Tricyclic antidepressants were among the earli- neuropathic pain in the past. A study of oxcarba-
est medications used for neuropathic pain. mazepine initiated at 300 mg/d and titrated to a
Cochran reviews show amitriptyline has the low- maximum of 1800 mg showed efficacy using a
est number needed to treat of all medications visual analog pain scale, but other studies were
studied, 67% of patients experiencing benefit over negative [55]. Both oxcarbamazepine and carba-
placebo [48,49]. However, the benefit is severely mazepine can cause hyponatremia, and carbamaz-
limited by side effects. In addition to targeting the epine can cause bone marrow depression [56].
same mechanisms as serotonin norepinephrine Another non-AED sodium channel blocker
reuptake inhibitors (SNRIs), they also antagonize used is mexiletine, an antiarrhythmic agent, and
histamine and muscarinic anticholinergic recep- used typically for myotonia. Mexiletine was not
tors, increasing the side effect profile with seda- significantly better than placebo [57,58]. Topical
tion, dry mouth, constipation, urinary retention, lidocaine has been studied for postherpetic neu-
and orthostatic hypotension. At high doses, they ralgia but has been used as an adjunctive agent
can also prolong the QT interval requiring ECG [46,57].
monitoring [50]. This led to a level B recommen-
dation from the AAN because of the high side
effect and concern for treatment in elderly Opioids
patients [40]. Nortriptyline is also frequently used
and has a similar side-effect profile. Both are Opioids have also been studied in painful
dosed daily, typically at night, due to sedating DPN. Tapentadol has been approved for painful
side effects starting at 10 25 mg and titrating up DPN [59]. It has a weak μ-opioid receptor antag-
to 150 mg/d maximum dose [51]. onism and norepinephrine reuptake inhibitor
activity. The extended-release formulation
showed a greater than 30% efficacy over placebo,
Antiepileptic drugs which led to its approval [60,61]. It has relatively
weak efficacy and, like other opioids, can lead to
Voltage-gated sodium channels have been opioid-induced hyperalgesia [62]. Oxycontin and
implicated in neuropathic pain mechanisms and morphine have been shown to be effective in

Diabetic Neuropathy
304 18. Treatment of diabetic polyneuropathy

placebo-controlled studies, especially in combi- therapies and have few side effects. The most
nation with other medications [63,64]. However, common adverse effects are acid reflux and nau-
their abuse potential and development of toler- sea. Doses of 600, 1200, and 1800 mg/d were all
ance makes them a last-step option and typically, found to be superior to placebo in pain relief and
management is performed in pain clinics. neuropathy symptoms and change score [14].
Opioid-induced hyperalgesia is a phenomenon The NATHAN study suggested patients with
thought to be secondary to neuroplasticity and fewer comorbidities of hypertension, cardiovas-
sensitization to μ-opioid receptors and nociceptor cular disease, or obesity were more likely to
toll-like receptor 4, which may arise from opioid improve in NIS-LL [14,71].
activation of mast cells, microglia, and other
cells. Opioid-induced hyperalgesia causes
increased pain and allodynia. It typically occurs Benfotiamine
with higher potency analgesics [65]. The treat-
Benfotiamine, a synthetic derivative of thia-
ment is to recognize it, taper opioid treatment,
mine, has been studied to treat diabetic neuropa-
and determine whether to restart after a holiday.
thy with conflicting evidence. Thiamine is a
Patients with refractory neuropathic pain
cofactor for several enzymes important in glu-
should be screened for depression, as overlap-
cose metabolism in the Krebs cycle, such as
ping depression can increase catastrophization
transketolase, pyruvate dehydrogenase, and
and minimize benefits from medications
α-ketoglutarate dehydrogenase. Their actions are
[38,66,67]. A multimodal pain clinic may be
important for the prevention of toxic metabolites
helpful for these patients. Although typical
[72]. In animal studies, benfotiamine reversed
dogma is to raise an existing medication to
structural changes in the nerve [73] and reversed
higher doses before the trial of a second medi-
the accumulation of advanced glycation endpro-
cation, in neuropathic pain, patients often ben-
ducts [74]. Several studies showed possible
efit from the utilization of two different
improvement in pain or in neuropathy symptom
medications with different modes of action to
score [75,76] but did not improve overall nerve
improve pain control and minimize side
function compared to placebo in a study of type
effects. Some initial work on phenotyping pain
1 diabetic patients [77].
patients to identify the most likely successful
agent to individualize therapy has been per-
formed using quantitative sensory testing,
which may also lead to more successful pain Cannabidiol and cannabis-derived
treatment [68]. treatments

Cannabinoids bind to cannabinoid receptors,

CB1 in the brain and TRPV1, peroxisome
Supplements and alternative treatments
proliferator-activated receptor γ (PPARγ) GABA
for painful diabetic polyneuropathy
receptors, and calcium channels [78,79].
Cannabinoid products studied include cannabi-
α-Lipoic acid diol oil, tetrahydrocannabinol (THC) and canna-
ALA is an antioxidant studied initially for the bidiol nasal spray, nabilone (synthetic mimic of
treatment of DPN but was shown symptomati- THC), inhaler herbal cannabis and plant-
cally to improve pain [69,70]. It is often used as derived THC but showed a modest benefit of
an adjunctive pain reliever. It may be attractive 28% improvement compared to 22% in the pla-
to patients who are reluctant to use conventional cebo group and were typically of small numbers

Diabetic Neuropathy
 Emerging treatments and recent trials 305
[80]. More data are needed before cannabinoids damage to nerve roots or cord during surgical
can be recommended for painful DPN. placement, and hardware failure.
Other supplements studied have included A typical algorithm for pain management
acetyl-L-carnitine, thought to enhance mito- includes using a gabapentinoid or SNRI as a
chondrial activity, curcumin, nutmeg, St. Johns first-line agent, and the choice depends on
wort, B vitamins (Metanx), and Lion’s mane comorbidities such as RLS, depression. Many
(Hericium erinaceus), fish oil (containing omega- patients require more than one agent, and
3 fatty acids) in animal models, but clinical trial unlike standard pharmacotherapy where the
data are lacking [81 87]. first agent is pushed to maximum dose before
the trial of a second, in pain control, usage of
two agents may synergistically improve pain
by acting on multiple pathways. All patients
Exercise and nonpharmacologic should be counseled on the potential benefit of
treatments exercise. If first and second-line agents are not
helpful, referral to a multidisciplinary pain
Exercise is a promising therapeutic option for clinic may be necessary. Screening for depres-
neuropathic pain and may effectively treat DPN sion is also key, as depression may lead to pain
itself [88]. Animal exercise studies demonstrate catastrophization and impaired sleep, increas-
increased nerve growth factor production in the ing pain and minimizing response to pain
dorsal root ganglion, including brain-derived neu- management [67] (Fig. 18.1).
rotrophic factor and decreased pro-inflammatory
cytokines such as tumor necrosis factor-α and IL-
1β [89,90]. Moderate aerobic exercise, in addition Emerging treatments and recent trials
to strength training, has been studied, showing
benefits in pain and intraepidermal nerve fiber Nav1.7 channels have been well documen-
density [88,89,91]. Exercise may also be helpful in ted to be involved in the development of neu-
the treatment of DAN [92]. Yoga has been studied ropathic pain [57]. Diabetic neuropathy models
in chronic pain but not specifically in painful have shown an upregulation of Na1.7 channel
DPN [93]. Tai chi has been studied to help subtype in painful diabetic neuropathy, possi-
improve balance and decrease fall risk [94,95]. bly from TNF-alpha induced upregulation [98].
Acupuncture has been studied in several A trial of a Na1.7 sodium channel blocker in
small studies but it has been challenging to painful DPN did not lead to superior pain
develop a rigorous clinical methodology to relief over pregabalin, but further studies may
prove efficacy [96]. Limited insurance coverage be helpful [99].
and limited availability in some areas minimize Trial of stromal cell-like mesenchymal products
its potential benefit. It is not clear whether acu- was also studied in diabetic neuropathy. Human
puncture activates descending pain control placental-derived adherent cells are a mesenchy-
pathways or decrease neuroinflammation. mal stromal cell-like population found in animal
Spinal cord stimulation has been studied in a models to decrease neuroinflammation pain and
few studies with patients having at least short- promote angiogenesis [100,101]. However, a
term pain benefits [97]. Surgically laced electrodes phase 2a study of injection of placental-derived
in the epidural space produce electrical stimula- adherent cells did not significantly improve or
tion of nerve roots or dorsal columns theoretically prevent the decline of intraepidermal nerve fiber
block pain perception without ablation. Short- density. Angiotensin 2 receptor antagonists were
term cost is high, and risks include infection, also recently studied with negative results [102].

Diabetic Neuropathy
306 18. Treatment of diabetic polyneuropathy

FIGURE 18.1 Algorithm of approach to pain management. Source: From Peltier AC, Wood D. Management of neuropathic
pain in polyneuropathy. Continuum 2020;26(5):1299 1322.

Nicotinamide adenine dinucleotide phosphate control. The development of improved animal

(NADPH) oxidase, specifically the NADPH oxi- models and therapeutic endpoints would be
dase, type 5 isoform restricted to humans, has helpful for future therapeutic trials [107,108].
been recently studied as a potential target for dia- Further efforts to better identify clinical trial
betic complications of retinopathy and nephropa- participants and refine the definition of dia-
thy and may also be a potential new target for betic neuropathy will hopefully lead to more
treatments of diabetic neuropathy [103 106]. positive treatment outcomes [8,109].
Further work on mechanisms of pain in DPN
will hopefully lead to improved treatment of neu-
Conclusion ropathic pain in DPN. Evaluation of potential
genetic variants in ion channels may also lead to
More work is needed in the treatment of improved pain treatments. The development of
DPN, which is currently limited to mainly pain multimodal therapies involving multiple pain

Diabetic Neuropathy
 References 307
pathways may reduce the morbidity and loss of [11] Oka M, Kato N. Aldose reductase inhibitors. J Enzyme
function in patients with refractory neuropathic Inhib 2001;16(6):465 73.
[12] Cherian PV, Kamijo M, Angelides KJ, Sima AA. Nodal
pain. Na(1)-channel displacement is associated with nerve-
Utilization of nonpharmacologic therapies such conduction slowing in the chronically diabetic BB/W
as exercise needs more study and may act on mul- rat: prevention by aldose reductase inhibition.
tiple DPN mechanism. Treatments such as bariat- J Diabetes Complications 1996;10(4):192 200.
ric surgery affect multiple pathways in the [13] Nicolucci A, Carinci F, Cavaliere D, Scorpiglione N,
Belfiglio M, Labbrozzi D, et al. A meta-analysis of trials
development of DPN and may be more effective on aldose reductase inhibitors in diabetic peripheral
in the treatment of DPN, especially in type 2 dia- neuropathy. The Italian Study Group. The St. Vincent
betes where glucose control is less effective in the Declaration. Diabet Med 1996;13(12):1017 26.
prevention of DPN. [14] Ziegler D, Low PA, Litchy WJ, Boulton AJ, Vinik AI,
Freeman R, et al. Efficacy and safety of antioxidant
treatment with alpha-lipoic acid over 4 years in dia-
betic polyneuropathy: the NATHAN 1 trial. Diabetes
References Care 2011;34(9):2054 60.
[15] Apfel SC, Schwartz S, Adornato BT, Freeman R, Biton V,
[1] England JD, Asbury AK. Peripheral neuropathy. Rendell M, et al. Efficacy and safety of recombinant
Lancet 2004;363(9427):2151 61. human nerve growth factor in patients with diabetic poly-
[2] Coppini DV, Bowtell PA, Weng C, Young PJ, Sonksen neuropathy: a randomized controlled trial. rhNGF
PH. Showing neuropathy is related to increased mor- Clinical Investigator Group. JAMA 2000;284(17):2215 21.
tality in diabetic patients a survival analysis using [16] Malik RA, Williamson S, Abbott C, Carrington AL,
an accelerated failure time model. J Clin Epidemiol Iqbal J, Schady W, et al. Effect of angiotensin-
2000;53(5):519 23. converting-enzyme (ACE) inhibitor trandolapril on
[3] Boyko EJ, Ahroni JH, Smith DG, Davignon D. human diabetic neuropathy: randomised double-blind
Increased mortality associated with diabetic foot ulcer. controlled trial. Lancet 1998;352(9145):1978 81.
Diabet Med 1996;13(11):967 72. [17] Malik RA. Can diabetic neuropathy be prevented by
[4] Organization WH. Factsheets diabetes [web page]. World angiotensin-converting enzyme inhibitors? Ann Med
Health Organization; 2020Available from. Available from: 2000;32(1):1 5.
https://www.who.int/diabetes/global-report/. [18] Didangelos T, Tziomalos K, Margaritidis C, Kontoninas
[5] Gordois A, Scuffham P, Shearer A, Oglesby A, Z, Stergiou I, Tsotoulidis S, et al. Efficacy of administra-
Tobian JA. The health care costs of diabetic periph- tion of an angiotensin converting enzyme inhibitor for
eral neuropathy in the US. Diabetes Care 2003;26 two years on autonomic and peripheral neuropathy in
(6):1790 5. patients with diabetes mellitus. J Diabetes Res
[6] Kiyani M, Yang Z, Charalambous LT, Adil SM, Lee 2017;2017:6719239.
HJ, Yang S, et al. Painful diabetic peripheral neuropa- [19] Diabetes Control Complications Trial Research Group,
thy: health care costs and complications from 2010 to Nathan DM, Genuth S, Lachin J, Cleary P, Crofford O,
2015. Neurol Clin Pract 2020;10(1):47 57. Davis M, Rand L, Siebert C. The effect of intensive treat-
[7] Calcutt NA, Fernyhough P. A brief introduction to the ment of diabetes on the development and progression of
history and controversies of clinical trials in diabetic long-term complications in insulin-dependent diabetes
neuropathy. Int Rev Neurobiol 2016;127:3 8. mellitus. N Engl J Med 1993;329(14):977 86.
[8] Bril V. The perfect clinical trial. Int Rev Neurobiol [20] Braffett BH, Gubitosi-Klug RA, Albers JW, Feldman EL,
2016;127:27 41. Martin CL, White NH, et al. Risk factors for diabetic
[9] Gibbons CH, Freeman R, Tecilazich F, Dinh T, Lyons TE, peripheral neuropathy and cardiovascular autonomic neu-
Gnardellis C, et al. The evolving natural history of neu- ropathy in the diabetes control and complications trial/
rophysiologic function in patients with well-controlled epidemiology of diabetes interventions and complications
diabetes. J Peripher Nerv Syst 2013;18(2):153 61. (DCCT/EDIC) study. Diabetes 2020;69(5):1000 10.
[10] Polydefkis M, Hauer P, Sheth S, Sirdofsky M, Griffin [21] Action to Control Cardiovascular Risk in Diabetes
JW, McArthur JC. The time course of epidermal nerve Study G, Gerstein HC, Miller ME, Byington RP, Goff
fibre regeneration: studies in normal controls and in Jr. DC, Bigger JT, et al. Effects of intensive glucose
people with diabetes, with and without neuropathy. lowering in type 2 diabetes. N Engl J Med 2008;358
Brain 2004;127(Pt 7):1606 15. (24):2545 59.

Diabetic Neuropathy
308 18. Treatment of diabetic polyneuropathy

[22] Adler AI, Stevens RJ, Neil A, Stratton IM, Boulton AJ, [36] Buchwald H, Avidor Y, Braunwald E, Jensen MD, Pories
Holman RR. UKPDS 59: hyperglycemia and other W, Fahrbach K, et al. Bariatric surgery: a systematic
potentially modifiable risk factors for peripheral vas- review and meta-analysis. JAMA 2004;292(14):1724 37.
cular disease in type 2 diabetes. Diabetes Care 2002;25 [37] Villarreal-Calderon JR, Cuellar-Tamez R, Castillo EC,
(5):894 9. Luna-Ceron E, Garcia-Rivas G, Elizondo-Montemayor
[23] Ang L, Jaiswal M, Martin C, Pop-Busui R. Glucose L. Metabolic shift precedes the resolution of inflamma-
control and diabetic neuropathy: lessons from tion in a cohort of patients undergoing bariatric and
recent large clinical trials. Curr Diabetes Rep metabolic surgery. Sci Rep 2021;11(1):12127.
2014;14(9):528. [38] Selvarajah D, Cash T, Sankar A, Thomas L, Davies J,
[24] Smith AG, Russell J, Feldman EL, Goldstein J, Peltier Cachia E, et al. The contributors of emotional distress
A, Smith S, et al. Lifestyle intervention for pre-diabetic in painful diabetic neuropathy. Diab Vasc Dis Res
neuropathy. Diabetes Care 2006;29(6):1294 9. 2014;11(4):218 25.
[25] Singleton JR, Smith AG, Bromberg MB. Increased preva- [39] Peltier AC, Wood D. Management of neuropathic pain
lence of impaired glucose tolerance in patients with pain- in polyneuropathy. Continuum. 2020;26(5):1299 322.
ful sensory neuropathy. Diabetes Care 2001;24(8):1448 53. [40] Bril V, England J, Franklin GM, Backonja M, Cohen J,
[26] Dyck PJ, Giannini C. Pathologic alterations in the dia- Del Toro D, et al. Evidence-based guideline: treatment
betic neuropathies of humans: a review. J Neuropathol of painful diabetic neuropathy: report of the American
Exp Neurol 1996;55(12):1181 93. Academy of Neurology, the American Association of
[27] Callaghan BC, Hur J, Feldman EL. Diabetic neuropa- Neuromuscular and Electrodiagnostic Medicine, and
thy: one disease or two? Curr Opin Neurol 2012;25 the American Academy of Physical Medicine and
(5):536 41. Rehabilitation. Neurology 2011;76(20):1758 65.
[28] Feldman EL, Nave KA, Jensen TS, Bennett DLH. New [41] McLean MJ. Clinical pharmacokinetics of gabapentin.
horizons in diabetic neuropathy: mechanisms, bioen- Neurology. 1994;44(Suppl 5):S17 22 discussion S31-2.
ergetics, and pain. Neuron 2017;93(6):1296 313. [42] Fink K, Meder W, Dooley DJ, Gothert M. Inhibition of
[29] O’Brien R, Johnson E, Haneuse S, Coleman KJ, O’Connor neuronal Ca(2 1 ) influx by gabapentin and subse-
PJ, Fisher DP, et al. Microvascular outcomes in patients quent reduction of neurotransmitter release from rat
with diabetes after bariatric surgery vs usual care: a neocortical slices. Br J Pharmacol 2000;130(4):900 6.
matched cohort study. Ann Intern Med 2018;169 [43] Sills GJ. The mechanisms of action of gabapentin and
(5):300 10. pregabalin. Curr Opin Pharmacol 2006;6(1):108 13.
[30] Adam S, Azmi S, Ho JH, Liu Y, Ferdousi M, [44] Bockbrader HN, Wesche D, Miller R, Chapel S,
Siahmansur T, et al. Improvements in diabetic neurop- Janiczek N, Burger P. A comparison of the pharmaco-
athy and nephropathy after bariatric surgery: a pro- kinetics and pharmacodynamics of pregabalin and
spective cohort study. Obes Surg 2021;31(2):554 63. gabapentin. Clin Pharmacokinet 2010;49(10):661 9.
[31] Chen X, Zhang J, Zhou Z. The effects of metabolic sur- [45] Pergolizzi Jr. JV, Raffa RB, Taylor Jr. R, Rodriguez G,
gery on microvascular complications in obese patients Nalamachu S, Langley P. A review of duloxetine 60 mg
with type 2 diabetes: a meta-analysis. Surg Obes Relat once-daily dosing for the management of diabetic
Dis 2021;17(2):434 43. peripheral neuropathic pain, fibromyalgia, and chronic
[32] Chang YC, Chao SH, Chen CC, Ser KH, Chong K, Lu musculoskeletal pain due to chronic osteoarthritis pain
CH, et al. The effects of bariatric surgery on renal, neuro- and low back pain. Pain Pract 2013;13(3):239 52.
logical, and ophthalmic complications in patients with [46] Attal N, Cruccu G, Baron R, Haanpaa M, Hansson P,
type 2 diabetes: the Taiwan diabesity study. Obes Surg Jensen TS, et al. EFNS guidelines on the pharmacologi-
2021;31(1):117 26. cal treatment of neuropathic pain: 2010 revision. Eur J
[33] Azmi S, Ferdousi M, Liu Y, Adam S, Iqbal Z, Dhage S, Neurol 2010;17(9):1113 e88.
et al. Bariatric surgery leads to an improvement in [47] Rowbotham MC, Goli V, Kunz NR, Lei D. Venlafaxine
small nerve fibre damage in subjects with obesity. Int J extended release in the treatment of painful diabetic
Obes 2021;45(3):631 8. neuropathy: a double-blind, placebo-controlled study.
[34] Punchai S, Hanipah ZN, Meister KM, Schauer PR, Pain 2004;110(3):697 706.
Brethauer SA, Aminian A. Neurologic manifestations [48] Moore RA, Derry S, Aldington D, Cole P, Wiffen PJ.
of vitamin B deficiency after bariatric surgery. Obes Amitriptyline for neuropathic pain and fibromyalgia in
Surg 2017;27(8):2079 82. adults. Cochrane Database Syst Rev 2012;12:CD008242.
[35] Sen O, Karaca FC, Turkcapar A. Neurological compli- [49] Moore RA, Derry S, Aldington D, Cole P, Wiffen PJ.
cation after laparoscopic sleeve gastrectomy: foot Amitriptyline for neuropathic pain in adults.
drop. Obes Surg 2020;30(3):957 60. Cochrane Database Syst Rev 2015;7:CD008242.

Diabetic Neuropathy
 References 309
[50] Boulton AJ. Is duloxetine more effective than amitrip- [65] Chu LF, Angst MS, Clark D. Opioid-induced hyperal-
tyline for painful diabetic neuropathy? Curr Diabetes gesia in humans: molecular mechanisms and clinical
Rep 2011;11(4):230 2. considerations. Clin J Pain 2008;24(6):479 96.
[51] Derry S, Wiffen PJ, Aldington D, Moore RA. [66] Vileikyte L, Peyrot M, Gonzalez JS, Rubin RR, Garrow
Nortriptyline for neuropathic pain in adults. Cochrane AP, Stickings D, et al. Predictors of depressive symp-
Database Syst Rev 2015;1:CD011209. toms in persons with diabetic peripheral neuropathy:
[52] Brismar T, Sima AA, Greene DA. Reversible and irre- a longitudinal study. Diabetologia 2009;52(7):1265 73.
versible nodal dysfunction in diabetic neuropathy. [67] McMillan SC, Tofthagen C, Morgan MA. Relationships
Ann Neurol 1987;21(5):504 7. among pain, sleep disturbances, and depressive symp-
[53] Han C, Dib-Hajj SD, Lin Z, Li Y, Eastman EM, Tyrrell toms in outpatients from a comprehensive cancer center.
L, et al. Early- and late-onset inherited erythromelal- Oncol Nurs Forum 2008;35(4):603 11.
gia: genotype-phenotype correlation. Brain 2009;132 [68] Reimer M, Helfert SM, Baron R. Phenotyping neuro-
(Pt 7):1711 22. pathic pain patients: implications for individual ther-
[54] Gill D, Derry S, Wiffen PJ, Moore RA. Valproic acid apy and clinical trials. Curr Opin Support Palliat Care
and sodium valproate for neuropathic pain and 2014;8(2):124 9.
fibromyalgia in adults. Cochrane Database Syst Rev [69] Vallianou N, Evangelopoulos A, Koutalas P. Alpha-
2011;10:CD009183. lipoic acid and diabetic neuropathy. Rev Diabet Stud
[55] Zhou M, Chen N, He L, Yang M, Zhu C, Wu F. 2009;6(4):230 6.
Oxcarbazepine for neuropathic pain. Cochrane Database [70] Han T, Bai J, Liu W, Hu Y. A systematic review and
Syst Rev 2013;3:CD007963. meta-analysis of alpha-lipoic acid in the treatment of
[56] Eisenberg E, River Y, Shifrin A, Krivoy N. Antiepileptic diabetic peripheral neuropathy. Eur J Endocrinol
drugs in the treatment of neuropathic pain. Drugs 2012;167(4):465 71.
2007;67(9):1265 89. [71] Ziegler D, Low PA, Freeman R, Tritschler H, Vinik AI.
[57] Kalso E. Sodium channel blockers in neuropathic pain. Predictors of improvement and progression of diabetic
Curr Pharm Des 2005;11(23):3005 11. polyneuropathy following treatment with alpha-lipoic
[58] Romman A, Salama-Hanna J, Dwivedi S. Mexiletine acid for 4 years in the NATHAN 1 trial. J Diabetes
usage in a chronic pain clinic: indications, tolerabil- Complications 2016;30(2):350 6.
ity, and side effects. Pain Physician 2018;21(5): [72] Beltramo E, Mazzeo A, Porta M. Thiamine and diabe-
E573 9. tes: back to the future? Acta Diabetol 2021;.
[59] Afilalo M, Morlion B. Efficacy of tapentadol ER for [73] Vafadar Ghasemi L, Behnam Rassouli M, Matin MM,
managing moderate to severe chronic pain. Pain Mahdavi-Shahri N. Benfotiamine reduced collagen IV
Physician 2013;16(1):27 40. contents of sciatic nerve in hyperglycemic rats.
[60] Hartrick CT, Rodriguez Hernandez JR. Tapentadol for J Diabetes Metab Disord 2021;20(1):21 30.
pain: a treatment evaluation. Expert Opin Pharmacother [74] Karachalias N, Babaei-Jadidi R, Ahmed N, Thornalley
2012;13(2):283 6. PJ. Accumulation of fructosyl-lysine and advanced
[61] Schwartz S, Etropolski MS, Shapiro DY, Rauschkolb C, glycation end products in the kidney, retina and
Vinik AI, Lange B, et al. A pooled analysis evaluating peripheral nerve of streptozotocin-induced diabetic
the efficacy and tolerability of tapentadol extended rats. Biochem Soc Trans 2003;31(Pt 6):1423 5.
release for chronic, painful diabetic peripheral neurop- [75] Haupt E, Ledermann H, Kopcke W. Benfotiamine in
athy. Clin Drug Investig 2015;35(2):95 108. the treatment of diabetic polyneuropathy a three-
[62] Mao XF, Ahsan MZ, Apryani E, Tang XQ, Zhao MJ, Li week randomized, controlled pilot study (BEDIP
XY, et al. Dual mu-opioid receptor and norepineph- study). Int J Clin Pharmacol Ther 2005;43(2):71 7.
rine reuptake mechanisms contribute to dezocine- and [76] Stracke H, Gaus W, Achenbach U, Federlin K, Bretzel
tapentadol-induced mechanical antiallodynia in can- RG. Benfotiamine in diabetic polyneuropathy
cer pain. Eur J Pharmacol 2020;876:173062. (BENDIP): results of a randomised, double blind,
[63] Chaparro LE, Wiffen PJ, Moore RA, Gilron I. placebo-controlled clinical study. Exp Clin Endocrinol
Combination pharmacotherapy for the treatment of Diabetes 2008;116(10):600 5.
neuropathic pain in adults. Cochrane Database Syst [77] Fraser DA, Diep LM, Hovden IA, Nilsen KB, Sveen KA,
Rev 2012;7:CD008943. Seljeflot I, et al. The effects of long-term oral benfotia-
[64] Gilron I, Bailey JM, Tu D, Holden RR, Weaver DF, mine supplementation on peripheral nerve function and
Houlden RL. Morphine, gabapentin, or their combina- inflammatory markers in patients with type 1 diabetes: a
tion for neuropathic pain. N Engl J Med 2005;352 24-month, double-blind, randomized, placebo-controlled
(13):1324 34. trial. Diabetes Care 2012;35(5):1095 7.

Diabetic Neuropathy
310 18. Treatment of diabetic polyneuropathy

[78] Cristino L, Bisogno T, Di Marzo V. Cannabinoids and [90] Cooper MA, Kluding PM, Wright DE. Emerging rela-
the expanded endocannabinoid system in neurologi- tionships between exercise, sensory nerves, and neu-
cal disorders. Nat Rev Neurol 2020;16(1):9 29. ropathic pain. Front Neurosci 2016;10:372.
[79] O’Hearn S, Diaz P, Wan BA, DeAngelis C, Lao N, [91] Singleton JR, Marcus RL, Jackson JE, M.K.L., Graham
Malek L, et al. Modulating the endocannabinoid TE, Smith AG. Exercise increases cutaneous nerve
pathway as treatment for peripheral neuropathic density in diabetic patients without neuropathy. Ann
pain: a selected review of preclinical studies. Ann Clin Transl Neurol 2014;1(10):844 9.
Palliat Med 2017;6(Suppl 2):S209 14. [92] Nakos I, Kadoglou NPE, Gkeka P, Tzallas AT,
[80] Mucke M, Phillips T, Radbruch L, Petzke F, Hauser Giannakeas N, Tsalikakis DG, et al. Exercise training
W. Cannabis-based medicines for chronic neuro- attenuates the development of cardiac autonomic
pathic pain in adults. Cochrane Database Syst Rev dysfunction in diabetic rats. In Vivo 2018;32(6):
2018;3:CD012182. 1433 41.
[81] Shevalye H, Watcho P, Stavniichuk R, Dyukova E, [93] Mooventhan A, Nivethitha L. Evidence based effects
Lupachyk S, Obrosova IG. Metanx alleviates multiple of yoga in neurological disorders. J Clin Neurosci
manifestations of peripheral neuropathy and 2017;43:61 7.
increases intraepidermal nerve fiber density in [94] Ahn S, Song R. Effects of Tai Chi exercise on glucose
Zucker diabetic fatty rats. Diabetes 2012;61(8): control, neuropathy scores, balance, and quality of
2126 33. life in patients with type 2 diabetes and neuropathy.
[82] Fonseca VA, Lavery LA, Thethi TK, Daoud Y, J Altern Complement Med 2012;18(12):1172 8.
DeSouza C, Ovalle F, et al. Metanx in type 2 diabetes [95] Gu Y, Dennis SM. Are falls prevention programs
with peripheral neuropathy: a randomized trial. Am effective at reducing the risk factors for falls in peo-
J Med 2013;126(2):141 9. ple with type-2 diabetes mellitus and peripheral neu-
[83] Rolim LC, da Silva EM, Flumignan RL, Abreu MM, ropathy: a systematic review with narrative
Dib SA. Acetyl-L-carnitine for the treatment of dia- synthesis. J Diabetes Complications 2017;31(2):
betic peripheral neuropathy. Cochrane Database Syst 504 16.
Rev 2019;6:CD011265. [96] Ju ZY, Wang K, Cui HS, Yao Y, Liu SM, Zhou J, et al.
[84] Sima AA, Calvani M, Mehra M, Amato A, Acetyl Acupuncture for neuropathic pain in adults.
LCSG. Acetyl-L-carnitine improves pain, nerve Cochrane Database Syst Rev 2017;12.
regeneration, and vibratory perception in patients [97] Mekhail N, Visnjevac O, Azer G, Mehanny DS,
with chronic diabetic neuropathy: an analysis of Agrawal P, Foorsov V. Spinal cord stimulation 50
two randomized placebo-controlled trials. years later: clinical outcomes of spinal cord stimula-
Diabetes Care 2005;28(1):89 94. tion based on randomized clinical trials a system-
[85] Rowin J. Integrative neuromuscular medicine: neu- atic review. Reg Anesth Pain Med 2018;43(4):
ropathy and neuropathic pain: consider the alterna- 391 406.
tives. Muscle Nerve 2019;60(2):124 36. [98] Huang Y, Zang Y, Zhou L, Gui W, Liu X, Zhong Y.
[86] Liu PS, Chueh SH, Chen CC, Lee LY, Shiu LY. Lion’s The role of TNF-alpha/NF-kappa B pathway on the
mane medicinal mushroom, Hericium erinaceus up-regulation of voltage-gated sodium channel
(Agaricomycetes), modulates purinoceptor-coupled Nav1.7 in DRG neurons of rats with diabetic neurop-
calcium signaling and murine nociceptive behavior. athy. Neurochem Int 2014;75:112 19.
Int J Med Mushrooms 2017;19(6):499 507. [99] McDonnell A, Collins S, Ali Z, Iavarone L, Surujbally
[87] Unda SR, Villegas EA, Toledo ME, Asis Onell G, Laino R, Kirby S, et al. Efficacy of the Nav1.7 blocker PF-
CH. Beneficial effects of fish oil enriched in omega-3 05089771 in a randomised, placebo-controlled,
fatty acids on the development and maintenance of double-blind clinical study in subjects with painful
neuropathic pain. J Pharm Pharmacol 2020;72(3): diabetic peripheral neuropathy. Pain 2018;159(8):
437 47. 1465 76.
[88] Smith AG, Marcus R. Exercise for diabetic neuropa- [100] He S, Khan J, Gleason J, Eliav E, Fik-Rymarkiewicz E,
thy: a toe in the therapeutic door. J Diabetes Herzberg U, et al. Placenta-derived adherent cells
Complications 2012;26(5):361 2. attenuate hyperalgesia and neuroinflammatory
[89] Kluding PM, Pasnoor M, Singh R, Jernigan S, Farmer response associated with perineural inflammation in
K, Rucker J, et al. The effect of exercise on neuro- rats. Brain Behav Immun 2013;27(1):185 92.
pathic symptoms, nerve function, and cutaneous [101] He S, Gleason J, Fik-Rymarkiewicz E, DiFiglia A,
innervation in people with diabetic peripheral neuropa- Bharathan M, Morschauser A, et al. Human placenta-
thy. J Diabetes Complications 2012;26(5):424 9. derived mesenchymal stromal-like cells enhance

Diabetic Neuropathy
 References 311
angiogenesis via T cell-dependent reprogramming of NOX5 exacerbates renal inflammation, fibrosis and
macrophage differentiation. Stem Cell 2017;35 albuminuria in the Akita mouse. Diabetologia
(6):1603 13. 2019;62(9):1712 26.
[102] Rice ASC, Dworkin RH, Finnerup NB, Attal N, [106] Jha JC, Watson AMD, Mathew G, de Vos LC,
Anand P, Freeman R, et al. Efficacy and safety of Jandeleit-Dahm K. The emerging role of NADPH oxi-
EMA401 in peripheral neuropathic pain: results of 2 dase NOX5 in vascular disease. Clin Sci 2017;131
randomised, double-blind, phase 2 studies in patients (10):981 90.
with postherpetic neuralgia and painful diabetic neu- [107] Calcutt NA, Fernyhough P. An introduction to the
ropathy. Pain 2021;. history and controversies of animal models of dia-
[103] Wu J, Lu K, Zhu M, Xie X, Ding Y, Shao X, et al. betic neuropathy. Int Rev Neurobiol 2016;127:
miR-485 suppresses inflammation and proliferation 45 52.
of mesangial cells in an in vitro model of diabetic [108] Biessels GJ, Bril V, Calcutt NA, Cameron NE, Cotter
nephropathy by targeting NOX5. Biochem Biophys MA, Dobrowsky R, et al. Phenotyping animal mod-
Res Commun 2020;521(4):984 90. els of diabetic neuropathy: a consensus statement of
[104] Deliyanti D, Alrashdi SF, Touyz RM, Kennedy CR, the diabetic neuropathy study group of the EASD
Jha JC, Cooper ME, et al. Nox (NADPH oxidase) 1, (Neurodiab). J Peripheral Nerv Syst 2014;19(2):
Nox4, and Nox5 promote vascular permeability and 77 87.
neovascularization in retinopathy. Hypertension [109] Gewandter JS, Burke L, Cavaletti G, Dworkin RH,
2020;75(4):1091 101. Gibbons C, Gover TD, et al. Content validity of
[105] Jha JC, Dai A, Holterman CE, Cooper ME, Touyz symptom-based measures for diabetic, chemother-
RM, Kennedy CR, et al. Endothelial or vascular apy, and HIV peripheral neuropathy. Muscle Nerve
smooth muscle cell-specific expression of human 2017;55(3):366 72.

Diabetic Neuropathy
 C H A P T E R

19
Diabetic autonomic neuropathy,
measurement and management;
part II: management
Joseph Colombo1,2,3
1
Franklin Cardiovascular Associates, PA and Autonomic Dysfunction and POTS Center, Sewell, NJ,
United States 2Physio PS, Inc., Atlanta, GA, United States 3Neurocardiology Research Center,
Sicklerville, NJ, United States

Autonomic neuropathy poor outcomes, and normal P&S results have

been correlated with healthy outcomes. For
example, orthostatic hypotension (OH) is com-
Diabetes, like aging, precipitates a cascade of
mon in diabetes. OH is a failure of the sympa-
secondary symptoms: sleep and GI disturbance,
thetics, known as sympathetic withdrawal
dizziness or lightheadedness, secondary hyper-
tension, urogenital dysfunction, higher risk of (SW).1 SW is not a structural deficit; it is a
functional deficit. Identifying this difference
cardiovascular disease, and more. This is a
facilitates diagnosis and therapy.
recurring theme and may help to lend more
In addition to lightheadedness (or dizziness
insight to pathogenesis, identification, diagno-
often from orthostasis), the common comorbidities
sis, and treatment of the effects of diabetes on
that are associated with diabetes mellitus include:
the P&S. P&S decline, regardless of rate, seems
to follow the same progression whether it is • Hypertension secondary to the diabetes, and
caused by healthy aging or chronic disease. possibly secondary to SW;
As discussed and referenced in Chapter 10, • Upper and lower GI upset, possibly due to
poor P&S results have been correlated with parasympathetic dysfunction;

1
Sympathetic Withdrawal (SW, which we will discuss later) is the neurological basis of orthostatic disorders. It is the
abnormal Sympathetic response to postural change where the sympathetics decrease rather than increase.

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00020-7 313 © 2022 Elsevier Inc. All rights reserved.
314 19. Diabetic autonomic neuropathy measurement and management, part II: management

• Sleep disturbances, possibly due to either or resting P&S dysfunction (i.e., DAN or CAN).
both P&S dysfunction; SW and other postural-change P&S dysfunc-
• Genitourinary dysfunction, possibly due to tions are detailed in Chapter 10.
either or both P&S dysfunction; SW and the associated orthostatic dysfunction
• Cardiovascular diseases, possibly due to may be treated with high dose r-alpha-lipoic acid
either or both P&S dysfunction; and (600 mg, tid, time release recommended to mini-
• Renal disorder, possibly due to either or mize acid effect on stomach) [3] or low-dose oral
both P&S dysfunction. vasoactive therapy (i.e., midodrine, 2.5 mg, tid,
titrated slowly, assuming no supine hypertension
The effect of autonomically active therapy
or resting BP above 160/90 mmHg) [4,5].
(e.g., beta-blockers, antihypertensives, antide-
As discussed in Chapter 10, small fiber neu-
pressants) is to help establish and maintain nor-
ropathy is also a precursor to autonomic neu-
mal SB. Therapy should be titrated to
ropathy. The small nerve fibers include
normalizing SB for the individual patient to
autonomic nerve fibers, and both are often
relieve the comorbidities. Then the patient
comorbid. Small fiber disorder is mostly associ-
becomes more stable, and the physician may be
ated with generalized pain disorders, and it
more aggressive toward the primary disease.
may be an even earlier indicator of autonomic
dysfunction. The small sympathetic fibers,
which uniquely control the vasculature, may
Earlier detection and earlier intervention be affected by small fiber disorder. This
Sympathetic affect may underlie poor periph-
The above short list of common comorbid- eral circulation leading to poor wound healing
ities may be substantiated by the resting P&S and dry skin, which are also early signs of dia-
imbalances indicated in the curves presented betes. Small fiber disease may be treated with
in Fig. 10.2 of Chapter 10; however, they may OTC methyl-folate or, if prescription grade is
have origins long before these more traditional needed, metanx [6].
definitions of autonomic neuropathy. As men-
tioned earlier, orthostatic hypotension (OH),
or more generally, orthostatic dysfunction Disease initiation and progression of
(including OH, orthostatic intolerance, or OI, neuropathy
and postural orthostatic tachycardia syndrome,
or POTS), is arguably the most debilitating It is well known that in diabetes, causative fac-
symptom of diabetes [1,2], and possibly P&S tors include persistent hyperglycemia, microvas-
dysfunction in general. Again, the autonomic cular insufficiency, oxidative and nitrosative
basis underlying orthostatic dysfunction is SW. stress, defective neurotropism, and autoimmune-
This dysfunction, documented during a mediated nerve destruction [7]. DAN is a hetero-
head-up, postural change (e.g., standing), is a geneous group of conditions with widely varying
dysfunctional challenge response (rather than pathology, suggesting differences in pathogenic
dysfunctional resting response). While ortho- mechanisms for the different clinical syndromes
static dysfunction (e.g., OH) may mark the [8,9]. Recognition of the clinical homolog of these
beginning of DAN or CAN, SW often occurs pathologic processes is the first step in achieving
long before DAN or CAN. Recent clinical expe- the appropriate form of intervention.
rience suggests that correcting SW, and any The link between autonomic and inflamma-
accompanying challenge autonomic dysfunc- tory pathways and their response to endoge-
tion, may help to slow the progression of nous or exogenous stressors is the sympathetic

Diabetic Neuropathy
 Disease initiation and progression of neuropathy 315
nervous system. Prolonged or excessive sym- balance [10] which may reverse the metabolic
pathetic activation is abnormal and leads to abnormalities simply by enhancing parasym-
increased morbidity and mortality risk. pathetic dominance to normalize SB. There are
Psychosocial and physical stresses, as well as simple treatments and therapies to improve SB
the stresses of disease, disorder, or injury, [10]. In contrast, failure to identify loss of para-
may all lead to prolonged or excessive sympa- sympathetic integrity is accompanied by poten-
thetic activation. Prolonged or excessive sym- tially dire consequences as witnessed by the
pathetic activation results in increased hepatic 22% increase in sudden death in the ACCORD
gluconeogenesis, a rise in circulating free fatty study with intensification of therapy [11]. The
acids, resistance to the action of insulin, and association between inflammation (with activa-
increases in markers of inflammation and tion of inflammatory cytokines such as IL-6
hypercoagulation. All of these may be restored and leptin), oxidative stress, and abnormalities
toward normal by treatment to restore P&S in SB is depicted in Fig. 19.1 [12]. Autonomic

FIGURE 19.1 Pathogenesis of DAN: the initiation and progression of diabetes. Multiple etiologies are involved, including
metabolic, vascular, autoimmune, oxidative stress, and neurohormonal growth-factor deficiency. Thus DAN is a heterogeneous
disease, requiring a highly individualistic approach to intervention. Abbreviations: 12LO/5LO: eosinophil types involved in
inflammation; 262 T . C: a promoter polymorphism of the catalase gene associated with DAN; ACE: angiotensin converting
enzyme; AGEs: advanced glycation end-products are proteins or lipids that become harmful due to glycation as a result of expo-
sure to sugars promoting atherosclerosis, kidney disease, and other complications of diabetes; APoE4: apolipoprotein E protein
involved in the metabolism of fats; AR Z2: androgen receptor gene enabling responses to hormones; CHT1: high-affinity choline
transporter involved in acetylcholine synthesis; HETE: hydroxyeicosatetraenoic acid, a derivative of the polyunsaturated fatty
acid, arachidonic acid; IL6: interleukin 6 acts as both a proinflammatory cytokine and an anti-inflammatory myokine; NFκβ:
nuclear factor κβ, overactivation from persistent hyperglycemia causes key role cytokine overproduction leading to the patho-
genesis of vascular complications of diabetes; PARPs: poly (ADP-ribose) polymerase proteins involved in a number of cellular
processes such as DNA repair, genomic stability, and programmed cell death; PKC: protein kinase C which regulates a number
of cellular responses including inflammation; ROS, reactive oxygen species promote oxidative stress and damage to cell mem-
branes, especially mitochondria; TNFα: tumor necrosis factor-α, a multifunctional cytokine; toll rec, toll-like receptor poly-
morphisms that recognize pathogens and signal immune responses including inflammation; VCAMS: vascular cell adhesion
molecule, part of the immunoglobulin superfamily important in inflammation, immune responses, and intracellular signaling.
Source: Adapted from Vinik AI, Mehrabyan A. Diabetic neuropathies. Medical Clinics of North America 2004; 88: 947 999.

Diabetic Neuropathy
316 19. Diabetic autonomic neuropathy measurement and management, part II: management

FIGURE 19.2 Afferent sensory signals are transmitted by the vagus nerve to the nucleus tractus solitary (NTS) and
polysynaptic receptors relay to the sympathetic through the rostral ventromedullary nucleus (RVLS), the parasympathetic
nucleus ambiguous (NA), and the dorsal vagal nucleus. There is both P and S output to the celiac ganglion. The splenic
nerve activates the inflammatory cascade in macrophages, which might be in the spleen, but occur diffusely throughout
the body. Stimulation of the vagus nerve inhibits this activation by acetylcholine binding to the nicotinic acetylcholine
receptor subunit α7 receptor, which curtails the response. Also shown is the coactivation of the hypothalamic-pituitary
axis (HPA) with release of glucocorticoids, which also modulates the inflammatory response. It is unclear how this set-
point is determined in the body, but what is clear is that autonomic modulation of inflammation may be achieved by alter-
ing SB. Source: Adapted from Tracey KJ. Reflex control of immunity. Nat Rev Immunol 2009; 9: 418 428.

dysfunction has been shown to be a predictor Oxidative stress

of cardiovascular risk and sudden death.
Activation of the efferent arm of the reflex arc All-cause mortality is greater than double
(through the administration of an acetylcholine for those diagnosed with type 2 diabetes and
receptor agonist, see Fig. 19.2 [13]) causes a DAN [7,8,12,16 24]. Oxidative stress is
decrease in proinflammatory cytokine produc- described as the mechanism related to auto-
tion, and a reduction in disease severity nomic injury found with DAN [25], which
[14,15]. often deteriorates to CAN which, untreated,

Diabetic Neuropathy
 Disease initiation and progression of neuropathy 317
leads to sudden cardiac death [26]. Oxidative
stress is stress at the cellular level. The oxida-
tion of fruit, like the apples in the insert below,
demonstrates how oxidative stress breaks
down cells, causing premature aging, fatigue,
and disease. Oxidative stress is mediated
through free radicals and reactive oxygen spe-
cies (ROS) and reactive nitrogen species (RNS,
ROS, and RNS collectively are known as
RONS). Oxidative stress targets cellular mem-
branes, DNA, and especially mitochondria. In
fact, mitochondria are the largest producers of
ROS in the body. While most ROS produced
by the mitochondria are collected by the
immune system and used as a first-line defense
against invading pathogens, increased oxida-
tive stress overwhelms the system and
damages mitochondria, reducing energy (ATP)
production and leading to fatigue, especially
when attempting to be active, mentally as well
as physically. Over the short term, as in
Oxidative damage affects proteins, nucleic
responding to a cold or the flu, fatigue is help-
acids, and lipids; its effects on proteins and
ful in that it promotes rest and sleep to help
nucleic acids lead to increased inflammation
the healing process. However, with long-term
and its effects on lipids exacerbate to nerve
oxidative stress, mitochondria (among other
injury. Hyperglycemia reduces antioxidant
organelles) are severely damaged, and persis-
activity, increasing inflammation, lipid peroxi-
tent fatigue or chronic fatigue is one of the
dation, the production of AGEs, glycated
resulting symptoms [27]. Fig. 19.3 presents the
proteins, and ketones, altering genetic tran-
systemic effects of oxidative stress, highlight-
scription activation, damaging energy produc-
ing the many complications of diabetes.
tion, and increasing apoptosis and cell death.
All of these affects are also involved in aging.
Oxidative stress can feed on itself, for example,
through lipid peroxidation and increased
inflammatory cytokines [28].
Cardiovascular risk factors and vascular dis-
ease are associated with increased levels of oxi-
dative stress. Carvedilol possesses antioxidant
effects, which may explain its preference in dia-
betes [29]. Traditional risk factors for circulatory
diseases have been identified, and excellent

Diabetic Neuropathy
FIGURE 19.3 The effects of oxidative stress at the system level and on the human body.
 Disease initiation and progression of neuropathy 319
pharmacological and lifestyle measures to Chronic, low-grade inflammation of white adi-
improve these risk factors have been developed. pose tissue is a hallmark of obesity and a major
Traditional risk factors include diabetes, contributor to oxidative stress and lipid peroxi-
high LDL cholesterol, low HDL cholesterol, dation [36]. In the expanding adipose tissue,
family history of premature cardiovascular dis- hypertrophied adipocytes contribute to the
ease,2 male gender, hypertension, being seden- inflammation by upregulating the expression
tary, aging, obesity, and tobacco [30,31]. and release of proinflammatory cytokines.
Nontraditional risk factors include psychosocial Aldehydes also decrease the expression of the
stress, inflammation, endothelial dysfunction, antiinflammatory, insulin-sensitizing hormone,
kidney disease, and poor brachial index mea- adiponectin [37,38], therefore linking lipid per-
surements [30,31]. Atherosclerosis is a common oxidation by-products and chronic inflamma-
circulatory disease and is highly associated with tion [28].
endothelial dysfunction. A vicious cycle between With diabetes, aldehyde concentration is
oxidative stress and oxidative stress-induced ath- increased in several tissues, including the pan-
erosclerosis leads to the development and pro- creas, liver, brain, and heart. Oxidative stress,
gression of atherosclerosis [32]. through ROS, can positively and negatively
Sleep apnea is a cause of oxidative stress, regulate insulin signaling, depending on time,
compounding the stress, leading to or exacerbat- dose, mode, and free radical used [39].
ing associated conditions and comorbidities, However, prolonged oxidative stress impairs
such as sympathetic overactivation, persistent insulin signaling and glucose uptake.
obesity, hypertension, hyperlipidemia, and dia- Aldehyde inhibition of mitochondrial aconitase
betes. Of course, these conditions also cause oxi- (a hydroxyl radical) enables the generation of
dative stress, exacerbating the condition and precursors of fatty acid synthesis [28,40].
perpetuating this vicious cycle, eventually However, data suggest that accumulation of
involving inflammatory and immune cell activa- lipid aldehydes and fat deposition could be
tion and cardiovascular disease. Sleep apnea is mutually inductive, leading to a vicious cycle
associated with elevated sympathetic activity promoting fat accretion [28]. Supplemental glu-
that may be, partially, ameliorated by CPAP treat- tathione helps this, as glutathione levels in dia-
ment [33,34]. It is important to know, however, betic patients may be depleted. An increase in
that CPAP is not a sympatholytic. It only has an local glutathione may prevent the deleterious
indirect effect on the sympathetics. It relives the effects of aldehyde on ROS production [41],
stress of not sleeping well, but does not actively, and that dysfunction of glutathione S-
directly, reduce sympathetic excess. If there are transferase, a major enzyme for aldehyde
other reasons for sympathetic excess (e.g., diabe- detoxification, leads to oxidative stress [42,43].
tes, hypertension, or obesity), CPAP alone is typi- In diabetes, a contributor to oxidative stress
cally not sufficient to relieve sympathetic excess. is an excess of advanced glycation end-
Considering metabolic syndrome as a pre- products (AGEs). AGEs are the result of
cursor to type 2 diabetes, significant aldehyde nonenzymatic additions of glucose or other
production through lipid peroxidation is saccharides to proteins, lipids, and nucleotides.
tightly linked to high-fat diet and obesity. AGE generation leads to intra- and extracellu-
Increased oxidative stress in accumulated fat is lar protein cross-linking and protein aggrega-
an important pathogenic mechanism of tion, which alters intracellular signaling and
obesity-associated metabolic syndrome [35]. gene expression, releases proinflammatory

2
This is considered a nonmodifiable risk factor.

Diabetic Neuropathy
320 19. Diabetic autonomic neuropathy measurement and management, part II: management

molecules, and results in an increased produc- [62]. Heart failure is categorized by activation
tion of ROS [44]. Compounds including vita- of the sympathetic and renin-angiotensin sys-
mins B6 and B12 with the antioxidant r-ALA tems. These neuroendocrine activations are
have helped reduce the production of ROS and associated with oxidative stress in the myocar-
the subsequent effects of AGE-induced oxida- dium and vasculature. In heart failure, oxida-
tive stress. tive stress occurs in the myocardium [63,64]
Protein kinase C (PKC) activation is a critical and plasma and correlates with left ventricular
step in the pathway to diabetic microvascular com- dysfunction [65]. Excessive levels of ROS (due
plications leading to cardiovascular disease [45]. It to diabetes or heart failure) adversely affect the
is activated by both hyperglycemia and disordered disposition of myocardial calcium, causing
fatty-acid metabolism resulting in increased pro- arrhythmia and abnormal cardiac remodeling
duction of vasoconstrictive, angiogenic, and che- and left ventricular hypertrophy, apoptosis,
motactic cytokines including transforming growth and necrosis, and they cause uncoupled nitric
factor β (TGF-β), VEGF, endothelin-1, and intercel- oxide synthase (NOS). Angiotensin-converting
lular adhesion molecules [46]. enzymes and angiotensin receptor blockers
There is evidence that oxidative stress leads exert their beneficial effects in part through
to a deficiency of growth factors in diabetes antioxidative mechanisms by blocking ROS
and contributes to clinical perturbations in production at the enzymatic source, the
small-fiber function [47] and, in the pancreas, a mitochondria.
pancreatic cytokine [48]. As shown in Fig. 19.4 The omega-3 fatty acid, eicosapentaenoic
[49], increased oxidative stress effects four acid (EPA), is widely used to fill the “Statin
pathways that contribute to glucose toxicity Gap” in diabetics and others with cardiovascu-
[50 53]. These different pathways can in lar disease due to endothelial dysfunction. This
return exacerbate oxidative stress and can is due to EPA’s healthful effects on lipoprotein
induce changes in gene expression, transcrip- metabolism, platelet function, cytokine produc-
tion factors, diverse cellular products disrupt- tion, clotting, fibrinolysis, and inflammatory
ing several cellular functions and the factors. Diabetes and obesity have similarities
communication between the cell and the sur- in that fatty acid uptake and fatty acid oxida-
rounding matrix, all of which leads to neuronal tion are increased as levels of intramyocardial
dysfunction and death [54 56]. These path- circulating triglycerides, circulating free fatty
ways also result in impaired microvascular acids, glucose uptake, and glucose oxidation
regulation and endothelial dysfunction by dif- are reduced [66,67]. The ability of obese or dia-
ferent pathways [57 59]. Diabetes affects the betic hearts to use glucose results in reliance
enteric nervous system (considered as a part of on fatty acid oxidation and reduces metabolic
the parasympathetic nervous system) and flexibility. Free fatty acids increase ROS pro-
thereby GI manifestations of diabetes [60,61]. duction in mitochondria, which increases
Patients who have type 2 diabetes are two to mitochondrial dysfunction leading to (among
four times more likely to develop heart failure other dysfunctions) increased myocardial ROS
than those without diabetes. However, heart emission [68]. Hyperglycemia induces the
failure is also a risk factor for diabetes. At mitochondria to produce excess superoxide
healthy levels, ROS are used to regulate multi- (O2-, a type of ROS) from increased flux in the
ple vascular cell functions, including endothe- electron transport chain. This is a major
lial cell and smooth muscle cell growth cause of the clinical consequences associated
proliferation, angiogenesis, and vascular tone with diabetes and obesity [69]. In this way,

Diabetic Neuropathy
 Treating diabetic autonomic dysfunction and neuropathy 321

FIGURE 19.4 Hyperglycemia-induced neuronal ischemia, dysfunction, and death. Abbreviations: PKC: protein kinase
C; AGE: advanced glycation end-products; PARP: poly ADP-ribose polymerase; GAPDH: glyceraldehyde-3 phosphate dehydro-
genase; GSH: glutathione; NADH: nicotinamide adenine dinucleotide; TGF-β: transforming growth factor; VEGF: vascular
endothelial growth factor; PAI-1: plasminogen activator inhibitor-1; eNOS: endothelial nitric oxide synthase; IL-1: interleukin 1;
TNF-α: tumor necrosis factor-α; VCAM-1: vascular cell adhesion molecule 1. Source: Baishideng Publishing Group Inc.

hyperglycemia activates four major pathways Treating diabetic autonomic dysfunction

in the pathogenesis of cardiovascular complica- and neuropathy
tions of diabetes and obesity including (1) a
path that increases production of advanced Pharmacology and “Functional Medicine”
glycation end products (AGEs), (2) the polyol are complimentary. Prescription medicines are
pathway, (3) the hexosamine pathway, and (4) needed to ensure the disease or disorder is sta-
the protein kinase C-dependent signal trans- bilized and at least partially relieved, and then
duction [69]. functional medicine (i.e., supplements and

Diabetic Neuropathy
322 19. Diabetic autonomic neuropathy measurement and management, part II: management

lifestyle modifications, including diet and exer- approved specifically for NOH and help to
cise) completes the therapy and helps to estab- relieve SW. Patient reactions to disease,
lish and maintain wellness. Traditional therapy, medication, and dosing may be
approaches should not be discounted but directly monitored and documented by P&S
should incorporate complimentary (nontradi- responses and used to further guide
tional) approaches as well. To this end, for individualized therapy.
example, two pharmaceuticals often prescribed Of course, first treat diabetes according to
to diabetics are metformin and carvedilol; both standards to relieve the adverse effects of
have antioxidant effects [70 73]. abnormal sugar and insulin levels on the
The P&S have been shown to be reflective of nervous system and the rest of the body.
medication effects (both direct and indirect), as Standard multifactorial treatment of
well as disease and aging effects. Many com- hyperglycemia, lipids, BP, and other causal
mon medications are known autonomic agents factors of diabetes may reduce abnormalities
and may be used off-label, typically in very in autonomic function by up to 68%,
low dose, to balance P&S activity both at rest especially when detected early. Autonomic
and in response to challenge, for example: dysfunction will still progress, albeit slower
(perhaps), and dysautonomia, DAN, and
• Sympatholytics, including: beta-1 then CAN will ultimately present.
(adrenergic) blockers, beta-2 adrenergic Additional P&S therapy for dysautonomia
agonists (bronchodilators), angiotensin (including DAN and CAN) helps to
converting enzyme inhibitors (ACE-Is), further slow the progression of P&S
angiotensin receptor blockers, etc.: dysfunction and continue to reduce
• Carvedilol is both a beta-1 blocker and an morbidity and mortality risks. Frequent
alpha-adrenergic blocker; monitoring and earlier treatment of
• Antihypertensives (including alpha- dysautonomia helps to delay even further
adrenergic antagonists); the onset of DAN and CAN and helps to
• Anticholinergics with antidepressant affects, maintain a more normal quality of life.
including very low dose antidepressants Treating DAN, CAN, or any dysautonomia
(e.g., tricyclic antidepressants, serotonin- first involves establishing and maintaining
norepinephrine reuptake inhibitors (SNRIs), proper balance: normalized SB and
and selective serotonin reuptake inhibitors normalized challenge responses [74].
(SSRIs)); and Disorders or symptoms that remain are then
• Oral vasoactive medications, including: a result of end-organ dysfunction, including
alpha-1 (adrenergic) agonists (vasopressors, oxidative stress and inflammation [75].
e.g., midodrine), droxidopa (northera), Restoring proper SB3 and P&S4 balance is
pyridostigmine (mestinon), and (rarely) possible [74], for example:
pseudoephedrine. • Resting SE(measured as high SB) may be
Only midodrine and droxidopa are FDA treated with beta-blockers and
approved for dysautonomia. They are antihypertensives, typically low doses

3
SB is the measure of resting P&S balance.
4
P&S balance refers to a challenge P&S balance.

Diabetic Neuropathy
 Treating diabetic autonomic dysfunction and neuropathy 323

FIGURE 19.5 Pre- and postdosing P&S and hemodynamic responses. Left panel depicts changes in sympathetic activity
and HR in response to beta-blocker (e.g., carvedilol). As expected, beta-blockers reduce both sympathetic activity and HR signif-
icantly. Middle panel depicts changes in sympathetic activity and systolic BP in response to vasopressor (e.g., low-dose mido-
drine or droxidopa). As expected, vasopressor increases both sympathetic activity and systolic BP significantly. Right panel
depicts changes in parasympathetic activity and systolic and diastolic BP in response to low-dose anticholinergic therapy (e.g.,
very low-dose nortriptyline or duloxetine). As expected, anticholinergic therapy reduces parasympathetic activity and increases
both BPs significantly. B-blocker; beta-adrenergic blocker; dBP: diastolic blood pressure; d( ): denotes a change in the parameter
that is in the parentheses; d(d( )): denotes a change in the change in the parameter that is in the parentheses (for example, hypo-
tension is a drop in BP increase in response to stand. That would be a negative d(BP)). Treating this should reverse the drop
and cause an increase in BP with stand. That would be a positive change in the drop, denoted as an increase d(d(BP))); HR:
heart rate; P: parasympathetic activity; PE: parasympathetic excess; sBP: systolic blood pressure; S: sympathetic activity; SE:
sympathetic excess; SW: sympathetic withdrawal; Test n: the first of two consecutive tests, but not always their first test in a
series; Test n 1 1: the second of two consecutive tests. Source: Adapted from Vinik AI, Erbas T, Casellini CM. Diabetic cardiac auto-
nomic neuropathy, inflammation and cardiovascular disease. J Diabet Investig 2013; 4(1): 4 18.

unless complicated by known SW may be treated with proper daily hydration

cardiovascular disease. (including fewer alcoholic, sugary, and caffeinated
• Challenge SE (documented during either or drinks and perhaps fewer diuretics) typically with
both the Valsalva or stand challenge) is electrolytes, oral vasopressors, or r-ALA. SB may
typically secondary and is often relieved by be adjusted with agents that act on one or another
treating the underlying PE or SW. arm of the ANS [20] (see Fig. 19.5 [24]).
• Resting PE (measured as low SB) or Pharmacological, supplement, or lifestyle mod-
challenge PE (documented during any of the ification manipulates P&S balance and leads to
P&S challenges: deep breathing, Valsalva, or changes in the symptom complex and quality of
stand) may be treated with low dose life [20]. This may include overmedication which
anticholinergics [23] (e.g., very low-dose will shift balance too far in the other direction
antidepressants), or low-and-slow exercise: (e.g., too much beta-blocker or antihypertensive
• Challenge PE in response to deep breathing may drive the sympathetics too low, enabling
challenge may be confounded by parasympathetic dominance, which may induce
pulmonary or upper respiratory disorder. exercise intolerance and fatigue). Intervention

Diabetic Neuropathy
324 19. Diabetic autonomic neuropathy measurement and management, part II: management

studies have documented the protective and posi- due to oxidative stress), cardiogenic syncope,
tive P&S effects of glycemic control in type 1 dia- upper or lower nonneurogenic gastrointestinal
betes (e.g., DCCT [76]) and of multifactorial symptoms, persistent inflammation, generalized
strategy aimed at lifestyle change with pharmaco- pain, sudomotor dysfunction (including anhydro-
logical correction of hyperglycemia, hypertension, sis, heat intolerance, dry skin, or hyperhidrosis),
dyslipidemia, and microalbuminuria [77]. R-ALA vision affects, urogenital dysfunction (both male,
is an example of functional medicine. Ziegler [78] including erectile dysfunction, or female, including
and Murray [3] reported that r-ALA improves vaginal dryness), other hormonal disorders, and
P&S function and reduces morbidity and mortal- impaired visceral sensations.
ity risk in patients diagnosed with diabetes. Note, Continued low-and-slow or graded, super-
several diabetes medications are listed as “bad vised exercise often helps with many of these
drugs.” Remember, FDA drug approval is based remaining symptoms. Exercise is arguably the
on an approximate 80% effectiveness. The “bad most powerful of antioxidants, among its other
drugs” are labeled so, based on the other 20%. beneficial properties. CoQ10 and other antioxi-
P&S responses help to identify earlier when a dants in addition to r-ALA help to address any
patient is not responding to these (or any other) remaining effects of oxidative stress, athero-
medications as expected and provides the docu- sclerosis, and some of the other cardiovascular
mentation to switch to another agent, thereby, diseases. Supplemental oral vasoactives (e.g.,
preventing the effects of “bad drugs.” fludrocortisone5 or pyridostigmine6) may help
After establishing P&S balance [79], therapy to support midodrine or droxidopa. Carvedilol
may include treating any symptoms that remain. (with its additional antioxidant effects, rather
Take care to not overtreat and re-establish P&S than metoprolol7 [81,82]) may be considered
imbalance. Also, it may take 3 6 months for for tachycardia and some cardiovascular dis-
remaining symptoms to be relieved organically. eases. Methyl-folate (supplement) or metanx
For example, should high BP remain once SW or (prescription) [83] may be considered for sudo-
PE are relieved, it may take approximately 3 motor dysfunction to help relieve the inflam-
months for the secondary SE to be relieved, then mation of small fiber disease and oftentimes
approximately 3 more months later, associated heal the small fibers. Metanx has been shown
high BP will normalize [80]. Of course, this to increase the number of nerve fibers, possibly
assumes no atherosclerosis, or hardening of arter- reversing the damage done by diabetes, and is
ies, or other nonneurogenic causes of high BP. generally well tolerated. Desmopressin may be
In addition to high BP, remaining symptoms considered as adjunctive to fluids and electro-
may include resting tachycardia or exercise intoler- lytes if the fluids (with electrolytes) are not
ance, atherosclerosis, or other cardiovascular staying in the veins, but rather is being spilled
diseases, nonneurogenic postural hypotension, through the kidneys. The algorithm below pro-
nonneurogenic lightheadedness, dizziness (vestibu- vides an example of the therapy options
lar in origin), weakness, fatigue (perhaps largely for SW.

5
A short course of Fludrocortisone (Florinef), prior to Midodrine, is recommended if resting BP is more than 160/90.
Take care, over the long term, as fludrocortisone may induce fibrosis.
6
Pyridostigmine (Mestinon) often causes diarrhea.
7
Carvedilol demonstrates a small but significant rebound in parasympathetic activity; metoprolol does not [].

Diabetic Neuropathy
 Treating diabetic autonomic dysfunction and neuropathy 325

Antioxidants (as well as A and E and glutathione). r-ALA is

more selective for nerves and has been found
r-Alpha-lipoic acid effective in slowing the progression of P&S
The body has a natural network of antioxi- dysfunction [78]. r-ALA has also been found to
dants to combat oxidative stress [84]. Two of reduce sudden cardiac death in patients diag-
the most powerful antioxidants that the body nosed with diabetes [3,85], relieve orthostatic
makes are alpha-lipoic acid (ALA,8 aka., thioc- hypotension [18,22,86 89], and reduce resting
tic acid) and co-enzyme Q10 (CoQ10). In and BP by enhancing nitric oxide production. r-
of themselves, they are powerful, and they are ALA improves glucose tolerance and insulin
made more powerful by the fact that they help sensitivity in both adipocytes and muscle cells
to recycle other antioxidants, such as vitamin C [90 92] and increases immune responses and

8
There are two forms of ALA: r-ALA and s-ALA. Only r-ALA is effective. S-ALA is not known to have any effect in
humans and is often used as a filler.

Diabetic Neuropathy
326 19. Diabetic autonomic neuropathy measurement and management, part II: management

ATP (energy) production. By recycling vitamin occur naturally with aging and are accelerated
E, r-ALA reduces fasting glycemia and insulin with chronic diseases or disorders such as dia-
resistance both in diabetes and metabolic syn- betes. Acquired mitochondrial dysfunctions
drome [93]. r-ALA has been called the univer- are those that may be slowed or reversed with
sal antioxidant [18,94 98]. It is unique because a proper antioxidant oxidant balance, a
it is both lipid and water soluble. r-ALA is one proper P&S balance, proper glucose metabo-
of the medications that may improve parasym- lism through treating metabolic syndrome or
pathetic function and is endorsed by the diabetes, and a proper nitric oxide environ-
Toronto consensus [99]. An example of the ment to promote proper oxygenation through
effects of lifestyle on SB has been demonstrated proper blood flow.
[100].
Nitric oxide
Coenzyme Q10 Nitric oxide is also an antioxidant, among
CoQ10 is more selective for the heart (car- other functions, including as a parasympa-
diovascular muscle) and is well recommended thetic neurotransmitter increasing parasympa-
by cardiologists for patients, including patients thetic tone. Fig. 19.6 presents an example of
diagnosed with diabetes with heart disease how antioxidant balance and antioxidant
[101 104]. Nerve and cardiac cells contain the imbalance affect multiple systems of the
most mitochondria of any cells in the body. body. The affected systems include those
Together, r-ALA and CoQ10 help to keep these that are also affected by diabetes. Note, a slight
two cell types healthy, as well as mitochondria balance advantage to the oxidants is normal.
in general. Decreases in antioxidant production Remember, the immune system utilizes

FIGURE 19.6 An example of the effects of balance and imbalance between antioxidants and oxidants. Imbalance
between nitric oxide (NO) and reactive oxygen species (ROS) signaling has been linked to renal dysfunction, metabolic
syndrome, cardiovascular disease, and type 2 diabetes.

Diabetic Neuropathy
 Treating diabetic autonomic dysfunction and neuropathy 327
oxidants to help keep us healthy. Too much is all about balance. Antioxidant oxidant bal-
oxidant leads to excessive inflammation and cell ance is a dynamic balance established under
death, and sympathetic overactivation. Proper conditions of homeostasis between free radicals
antioxidant balance promotes proper P&S bal- that are created (e.g., from the Mitochondria
ance, helping to delay the onset of autonomic due to the need for more energy to support the
neuropathy by slowing its progression. exercise) and those consumed (scavenged by
antioxidants) [110,111]. Physical exercise of
Exercise and diet great intensity (too much for the individual)
As stated above, exercise is potentially the induces oxidative damage [112 115] (again, it
most powerful antioxidant. Of course, too intense is all about balance). While some patients
or the wrong type of exercise may create oxida- may tolerate intense exercise, and even feel
tive stress and more damage. This is a reason better during it, the oxidative stress effects may
why exercise should be planned with physician still be too great. Their hearts may be decondi-
guidance. The review, “Psychopharmacology tioned as a result of oxidative stress or prior
and Cardiovascular Disease” [105] documents oxidative stress, and more damage is the long-
the many physiologic and lifestyle risk factors for term result. True, they feel better and think
the development of cardiovascular disease, with more clearly during exercise because their
diabetes near the top of the list, much of which skeletal muscles are conditioned and help
may be prevented or relieved (especially in the the heart to circulate the blood. However, once
early stages) with lifestyle modifications, includ- they stop exercising, because the heart is decon-
ing diet and exercise [106]. ditioned, these patients tend to “crash” for up
Diets, like the Mediterranean and Japo- to 3 days afterward because the heart could not
Mediterranean diets of course, are replete with support the exercise or the body’s recovery
antioxidants, antiinflammatories, and many from the exercise. Mild exercise (low-and-slow
other nutrients that support health and well- exercise) helps to recondition the heart so
ness. Exercise (a healthy level for the individ- that more intense exercises may be added back
ual) is also well known to have a significant in once the heart is reconditioned. This takes
antioxidant effect, and perhaps the greatest time and discipline, up to 6 months, regardless
[107]. The benefit of lifestyle change was con- of a patient feeling better. Follow-up P&S
firmed in the Diabetes Prevention Program assessment helps to encourage and discipline
[108]. Lifestyle modification demonstrated patients.
superior results in the improvement of auto- An important factor which perturbs antioxidant
nomic dysfunction as compared to the use of balance is an increased core body temperature
metformin or placebo. Moderate endurance [107]. After mild exercise, the parasympathetics
and aerobic exercise in both type 1 and type 2 engage in recovery. Increased parasympathetic
diabetes mellitus improve HRV and cardiac activity in the brain, together with endorphin
autonomic function significantly in favor of release, helps dilate blood vessels and
parasympathetic dominance (reducing DAN/ improves blood flow and endothelial function,
CAN risk), independent of BMI, glycemic or reducing hypertension and risk for atheroscle-
BP control, and duration of diabetes [109]. A rosis and cardiovascular disease. Outside the
minimum of 150 minutes of mild to moderate brain, the parasympathetic nervous system
exercise per week is strongly recommended, controls the immune system. Increased para-
with physician supervision. sympathetic activity, together with increased
Exercise improves blood flow and nerve core body temperature (simulating a fever),
function and reduces atherosclerosis. Overall, it improves immune responses, “burning-out”

Diabetic Neuropathy
328 19. Diabetic autonomic neuropathy measurement and management, part II: management

invading pathogens and otherwise inducing (TNF-α), and plasminogen-activator inhibitor

antioxidant effects [116]. (PAI). Patients with metabolic syndrome have
Exercise is important in treating type 2 diabetes higher risk of atherosclerosis, type 2 diabetes,
and metabolic syndrome and preventing worsen- non-alcoholic fatty liver disease, various cancers,
ing or even emerging syndromes [117 121]. and Alzheimer’s or Parkinson’s disease. These
Patients with diabetes, as well as myocardial disorders are not seen as frequently in people
infarction, heart failure, valve replacement, hyper- who exercise regularly. Interestingly, treatment
tension, pulmonary hypertension, and COPD, can with diabetic drugs such as metformin or thiazo-
increase their oxygen consumption with exercise. lidinediones diminishes systemic insulin resis-
It causes a reduction in HR and BP, an improve- tance similar to exercise. (Exercise is still more
ment in blood lipid profile, and a sense of well- efficient and effective and provides additional
being. A common goal of exercise is to maintain benefits.)
or lose weight and strengthen the functional
capacity of the heart and lungs. In most people,
exercise makes the heart function more efficient.9 Antioxidant effects of psychosocial stress
Increasing heart and lung capacity is well known reduction
to decrease the risk of heart disease [122 126], Psychosocial stress reduction is another pow-
including heart disease due to diabetes, and is erful antioxidant. Stress reduction involves exer-
now also associated with reduced risk of demen- cise and therapies that typically increase
tia, including Alzheimer’s disease [127 130]. parasympathetic activity, thereby reducing per-
The Surgeon General and Medical organiza- sistent sympathetic activity (including inflam-
tions, including the ADA and ACCE, all state mation), ultimately reducing the production of
that every American adult should participate free radicals and the deleterious effects of oxida-
in 30 minutes or more of moderate-intensity tive stress. Psychosocial stress is one example of
activity on most and probably all days of the the mind-body connection gone awry. While
week. Exercise reduces CAN and mortality there is no physical insult to the body, an abnor-
risk of diabetes possibly even before age 45. As mal mental state is causing physical disorder in
one ages, autonomic and mitochondrial dys- the body and the brain, including suppressing
function becomes more evident, and this is the immune system (which is controlled by the
linked to many disease processes. Exercise is parasympathetic nervous system). Heavy work-
one of the few proven methods of improving loads, job insecurity, marital or family strife, or
autonomic balance and mitochondrial function. living in poverty are circumstances that can
Metabolic syndrome is a constellation of find- result in chronically increased stress, which in
ings which involves central obesity, especially turn can lead to chronic psychological disorders
around the abdomen, high blood sugar, hyper- such as depression or anxiety. Stress and
tension, and high triglycerides along with low depression lead to persistent inflammation.
HDL levels. These patients have insulin Persistent sympathetic activation is associated.
resistance, which is proinflammatory. It is also a In general, the symptoms of diabetes are
procoagulant state. Patients with metabolic syn- involved, and eventually psychosocial stress
drome present with increases in inflammatory may induce symptoms of DAN and CAN.
markers such as IL-6, tumor necrosis factor alpha Psychosocial stress is well known to increase

9
The most notable exception are Hypertrophic Cardiomyopathy patients who express the recessive gene, in whom the
heart muscle thickens with exercise, reducing stroke volume and ejection fraction, leading to heart attack or sudden death.
Normally, the heart muscle does not thicken with exercise or stress, like skeletal muscles, but still gets stronger.

Diabetic Neuropathy
 References 329
morbidity and mortality risks, effecting quality mitochondrial efficiency), improve cognitive
of life and longevity. function, improve sex function, promote more
There are three stages to the acute stress restful sleep, improve social interactions,
response [131]. First, the stressor is detected reduce sweating, promote more patience and
with sensory organs. The second stage is the less irritability, and promote a happier mood.
stress response: defense or escape. The third Autonomic dysfunction, including DAN
stage is when the acute becomes chronic. If the and CAN, is treatable! More information that
stressor cannot be avoided and it persists, then has been traditionally available is needed,
the third stage is entered, a state of exhaustion however, and that information is available
[132]. Stress responses are coordinated by the through blood and urine tests for oxidative
P&S nervous systems and the hypothalamus- stress and P&S assessment. Balance is the key.
pituitary-adrenal (HPA) axis [133]. The neural Oxidative balance, P&S balance, balance
and hormonal signals interact and complement between exercise and rest, balanced diet, men-
each other through the regulatory action of the tal balance (including techniques to reduce
HPA axis. Its interaction with the enteric ner- psychosocial stress, like meditation, yoga,
vous system and the digestive system, to mobi- prayer, etc.), balance between work and time
lize energy reserves and protect the body with family and friends, etc. Medication may
during the stress response, involves the pan- help to accelerate the establishment and main-
creas and insulin production and thereby tenance of proper balance, and in advanced
affects diabetes. The HPA axis underlies the cases help to compensate for end-organ dam-
fight-or-flight response via a hormonal cascade age; but in the long-term, lifestyle modification
of catecholamines and corticoid hormones. is also necessary to complete the therapy and
Repetitive HPA activation resulting in an enable not only health, but wellness as well.
excess of glucocorticoids in the blood can lead
to metabolic diseases such as metabolic syn-
drome or type 2 diabetes [134 136]. For Acknowledgements and disclaimers
instance, early childhood abuse in girls The years of collaboration, mutual learning, and shared
increases the risk of diabetes and cardiovascu- data with Drs. Aaron I. Vinik, Gary L. Murray, Nicholas L
lar diseases, possibly through immune and DePace, and many more, are greatly appreciation. Thank
chronic inflammatory dysregulation associated you.
with SE [137], often secondary to PE. This work is not funded.
Dr. Colombo is the inventor of P&S monitoring and an offi-
Again, psychosocial stress reduction cer and part owner of Physio PS, Inc.
increases parasympathetic activity which
decreases sympathetic activity, thereby indi-
rectly slowing HR, dilating blood vessels to References
reduce BP, reducing insulin levels, etc. This
organically helps to relieve or prevent diabetes, [1] Vinik AI, Maser RE, Nakave AA. Diabetic cardiovascu-
lar autonomic nerve dysfunction. US Endocr Dis
hypertension, heart failure, certain types of 2007;2 9.
arrhythmias, and more. At the same time, the [2] Vinik A, Ziegler D. Diabetic cardiovascular autonomic
parasympathetics directly increase immune neuropathy. Circulation. 2007;115:387 97.
responses, decrease inflammation, oxidation, [3] Murray GL, Colombo J Maintenance (r)alpha lipoic
and oxidative stress, increase detoxification, acid reduces sudden cardiac death 43% in geriatric dia-
betes mellitus II patients with predominantly milld,
increase the motility of the gastrointestinal sys- asymptomatic dysautonomia; 2020, Submitted.
tem, etc. Overall, the net result is to improve [4] Arora RR, Bulgarelli RJ, Ghosh-Dastidar S, Colombo J.
energy (including through improved Autonomic mechanisms and therapeutic implications

Diabetic Neuropathy
330 19. Diabetic autonomic neuropathy measurement and management, part II: management

of postural diabetic cardiovascular abnormalities. [18] Ziegler D, Low PA, Litchy WJ, Boulton AJM, Vinik AI,
J Diabetes Sci Technol 2008;2(4):568 71. Freeman Rthe NATHAN 1 Trial Group. Efficacy and
[5] DePace NL, Vinik AI, Acosta C, Colombo J. Oral vaso- safety of antioxidant treatment with α-lipoic acid over
active medications: a review of midodrine, droxidopa, 4 years in diabetic polyneuropathy. Diabetes Care
and pseudoephedrine as applied to orthostatic dys- 2011;34:2054 60.
function. NEJM; 2021. Submitted. [19] Vinik AI, Maser RE, Ziegler D. Neuropathy. The crys-
[6] Fonseca VA, Lavery LA, Thethi TK, Daoud Y, tal ball for cardiovascular disease. Diabetes Care
DeSouza C, Ovalle F, et al. Metanx in type 2 diabetes 2010;33(7):1688 90.
with peripheral neuropathy: a randomized trial. Am J [20] Vinik AI, Maser RE, Ziegler D. Autonomic imbalance:
Med 2013;126(2):141 9. Available from: https://doi. prophet of doom or scope for hope? Diabet Med
org/10.1016/j.amjmed.2012.06.022. Epub 2012 Dec 5. 2011;28:643 51.
PMID: 23218892. [21] Vinik AI. The conductor of the autonomic orchestra.
[7] Vinik AI, Nevoret ML, Casellini CM, Parson H. Front Endocrinol 2012;3(71):1 13.
Diabetic neuropathy. Endocrinol Metab Clin North [22] Maser RE, Mitchell BD, Vinik AI, et al. The association
Am 2013;42(4):747 87. Available from: https://doi. between cardiovascular autonomic neuropathy and
org/10.1016/j.ecl.2013.06.001 Review. mortality in individuals with diabetes: a meta-
[8] Vinik AI, Strotmeyer ES, Nakave AA, Patel CV. analysis. Diabetes Care 2003;26:1895 901.
Diabetic neuropathy in older adults. Clin Ger Med [23] Vinik AI, Erbas T. Cardiovascular autonomic neuropa-
2008;24:407 35. thy: diagnosis and management. Curr Diab Rep
[9] Wu JS, Yang YC, Lin TS, et al. Epidemiological evidence 2006;6:424 30.
of altered cardiac autonomic function in subjects with [24] Vinik AI, Erbas T, Casellini CM. Diabetic cardiac auto-
impaired glucose tolerance but not isolated impaired fast- nomic neuropathy, inflammation and cardiovascular
ing glucose. J. Clin Endocrinol Metab 2007;92:3885 9. disease. J Diabetes Invest 2013;4(1):4 18.
[10] Scranton RE, Gaziano JM, Rutty D, Ezrokhi M, [25] Feldman, EL. Oxidative stress and diabetic neuropa-
Cincotta A. A randomized, double-blind, placebo- thy: a new understanding of an old problem. J Clin
controlled trial to assess safety and tolerability during Investig 111(4):431 433.
treatment of type 2 diabetes with usual diabetes ther- [26] Tsuji H, Venditti Jr FJ, Manders ES, Evans JC,
apy and either cycloset or placebo. BMC Endocr Larson MG, Feldman CL, et al. Reduced heart rate
Disord 2007;7(3). Available from: https://doi.org/ variability and mortality risk in an elderly cohort.
10.1186/1472-6823-7-3. The framingham heart study. Circulation 1994;90
[11] Calles-Escandon J, Lovato L, Simons-Morton D, (2):878 83.
Kendell D, Pop-Busui R, Cohen R, et al. Effect of inten- [27] DePace NL, Colombo J. Fatigue & dyasutonomia
sive compared with standard glycemia treatment chronic or persistent what’s the difference?: the mind-
strategies on mortality by baseline subgroup charac- body wellness program. NY, NY: Skyhorse
teristics. Diabetes Care 2010;33:721 7. Publishing, Co.; 2020.
[12] Vinik AI, Mehrabyan A. Diabetic neuropathies. Med [28] Halliwell B, Gutteridge J. Free radicals in biology and
Clin N Am 2004;88:947 99. medicine. Oxford, UK: Oxford Univ. Press; 2007.
[13] Tracey KJ. Reflex control of immunity. Nat Rev [29] Dandona P, Ghanim H, Brooks DP. Antioxidant
Immunol 2009;9:418 28. activity of carvedilol in cardiovascular disease. J
[14] Bernik TR, Friedman SG, Ochani M, DiRaimo R, Ulloa Hyperten 2007;25(4):731 41. Available from:
L, Yang H, et al. Pharmacological stimulation of the https://doi.org/10.1097/HJH.0b013e3280127948
cholinergic anti-inflammatory pathway. J Exp Med PMID: 17351362.
2002;195:781 8. [30] DePace NL, Mears JP, Yayac M, Colombo J. Cardiac
[15] Van Maanen MA, Vervoordeldonk MJ, Tak PP. The autonomic testing and diagnosing heart disease. A
cholinergic anti-inflammatory pathway: towards inno- clinical perspective. Heart Int 2014;9(2):37 44.
vative treatment of rheumatoid arthritis. Nat Rev Available from: https://doi.org/10.5301/heartint.5000218
Rheumatol 2009;5:229 32. published online 12/5/2014 12:29:58 PM.
[16] Vinik AI, Freeman R, Erbas T. Diabetic autonomic [31] DePace NL, Mears JP, Yayac M, Colombo J. Cardiac
neuropathy. Semin Neurol 2003;23:365 72. autonomic testing and treating heart disease. A clini-
[17] Vinik A, Erbas T, Pfeifer M, Feldman E, Stevens M, cal perspective. Heart Int 2014;9(2):45 52. Available
Russell J. Diabetic autonomic neuropathy, 2004. from: https://doi.org/10.5301/heartint.5000216.
In: Zucchi SE, editor. The diabetes mellitus manual: [32] Förstermann U, Münzel T. Endothelial nitric oxide
a primary care companion to Ellenberg and Rifkin’s. synthase in vascular disease: from marvel to menace.
6th Edition New York: McGraw Hill; 2004. p. 351. Circulation 2006;113:1708 14.

Diabetic Neuropathy
 References 331
[33] Carlson JT, Hedner J, Elam M, et al. Augmented rest- Med Cell Longev 2013;2013:564961. Available from:
ing sympathetic activity in awake patients with https://doi.org/10.1155/2013/564961 Epub 2013 Mar
obstructive sleep apnea. Chest 1993;103:1763 8. 27. Review.
[34] Hedner J, Darpo B, Ejnell H, et al. Reduction in [46] Rivard A, Silver M, Chen D, Kearney M, Magner M,
Sympathetic activity after long-term CPAP treatment Annex B, et al. Rescue of diabetes-related impairment
in sleep apnoea: cardiovascular implications. Eur of angiogenesis by intramuscular gene therapy with
Respir J 1995;8:222 9. adeno-VEGF. Am J Pathol 1999;154:355 63.
[35] Furukawa S, Fujita T, Shimabukuro M, Iwaki M, [47] Pittenger G, Vinik A. Nerve growth factor and diabetic
Yamada Y, Nakajima Y, et al. Increased oxidative neuropathy. Exp Diabesity Res 2003;4:271 85.
stress in obesity and its impact on metabolic syn- [48] Tam J, Rosenberg L. Maysinger D: INGAP peptide
drome. J Clin Invest 2004;114:1752 61. improves nerve function and enhances regeneration in
[36] Wellen KE, Hotamisligil GS. Obesity-induced inflam- streptozotocin-induced diabetic C57BL/6 mice.
matory changes in adipose tissue. J Clin Invest FASEBJ 2004;18:1767 9.
2003;112:1785 8. [49] Dimitropoulos G, Tahrani AA, Stevens MJ. Cardiac
[37] Soares AF, Guichardant M, Cozzone D, Bernoud-Hubac autonomic neuropathy in patients with diabetes melli-
N, Bouzaidi-Tiali N, Lagarde M, et al. Effects of oxidative tus. World J Diabetes 2014;5(1):17 39. Available from:
stress on adiponectin secretion and lactate production in https://doi.org/10.4239/wjd.v5.i1.17 Published online
3T3-L1 adipocytes. Free Radic Biol Med 2005;38:882 9. Feb 15, 2014.
[38] Wang Z, Dou X, Gu D, Shen C, Yao T, Nguyen V, [50] Soriano FG, Virág L, Szabó C. Diabetic endothelial
et al. 4-Hydroxynonenal differentially regulates adipo- dysfunction: role of reactive oxygen and nitrogen spe-
nectin gene expression and secretion via activating cies production and poly(ADP-ribose) polymerase
PPARγ and accelerating ubiquitin-proteasome degra- activation. J Mol Med (Berl.) 2001;79:437 48.
dation. Mol Cell Endocrinol 2012;349:222 31. [51] Obrosova IG, Julius UA. Role for poly(ADP-ribose) poly-
[39] Bashan N, Kovsan J, Kachko I, Ovadia H, Rudich A. merase activation in diabetic nephropathy, neuropathy
Positive and negative regulation of insulin signaling and retinopathy. Curr.Vasc.Pharmacol. 2005;3:267 83.
by reactive oxygen and nitrogen species. Physiol Rev [52] Yamagishi S, Uehara K, Otsuki S, Yagihashi S.
2009;89:27 71. Differential influence of increased polyol pathway on
[40] Zimniak P. 4-Hydroxynonenal and fat storage: a protein kinase C expressions between endoneurial
paradoxical pro-obesity mechanism? Cell Cycle and epineurial tissues in diabetic mice. J Neurochem
2010;9:3393 4. 2003;87:497 507.
[41] Pillon NJ, Croze ML, Vella RE, Soulère L, Lagarde M, [53] Giacco F, Brownlee M. Oxidative stress and diabetic
Soulage CO. The lipid peroxidation by-product complications. Circ Res 2010;107:1058 70.
4-Hydroxy-2-nonenal (4-HNE) induces insulin resis- [54] Williams B, Gallacher B, Patel H, Orme C. Glucose-
tance in skeletal muscle through both carbonyl and induced protein kinase C activation regulates vascular
oxidative stress. Endocrinology 2012;153:2099 111. permeability factor mRNA expression and peptide
[42] Kostyuk V, Potapovich A, Cesareo E, Brescia S, production by human vascular smooth muscle cells
Guerra L, Valacchi G, et al. Dysfunction of in vitro. Diabetes 1997;46:1497 503.
Glutathione S-transferase leads to excess 4-hydroxy-2- [55] Russell JW, Sullivan KA, Windebank AJ, Herrmann
nonenal and H2O2 and impaired cytokine pattern in DN, Feldman EL. Neurons undergo apoptosis in ani-
cultured keratinocytes and blood of vitiligo patients. mal and cell culture models of diabetes. Neurobiol Dis
Antioxid Redox Signal 2010; Available from: https:// 1999;6:347 63.
doi.org/10.1089/ars.2009.2976. [56] Ramasamy R, Vannucci SJ, Yan SS, Herold K, Yan SF,
[43] Curtis JM, Grimsrud PA, Wright WS, Xu X, Foncea Schmidt AM. Advanced glycation end products and
RE, Graham DW, et al. Downregulation of adipose RAGE: a common thread in aging, diabetes, neurodegen-
glutathione S-transferase A4 leads to increased protein eration, and inflammation. Glycobiology 2005;15:16R 28R.
carbonylation, oxidative stress, and mitochondrial [57] Kolm-Litty V, Sauer U, Nerlich A, Lehmann R,
dysfunction. Diabetes 2010;59:1132 42. Schleicher (ED). High glucose-induced transforming
[44] Uribarri J, Tuttle KR. Advanced glycation end pro- growth factor beta1 production is mediated by the
ducts and nephrotoxicity of high-protein diets. CJASN hexosamine pathway in porcine glomerular mesangial
2006;1(6):1293 9 Epub 2006 Sep 27. cells. J Clin Invest 1998;101:160 9.
[45] De Marchi E, Baldassari F, Bononi A, Wieckowski MR, [58] Wada R, Yagihashi S. Role of advanced glycation end
Pinton P. Oxidative stress in cardiovascular diseases products and their receptors in development of diabetic
and obesity: role of p66Shc and protein kinase C. Oxid neuropathy. Ann NY Acad Sci 2005;1043:598 604.

Diabetic Neuropathy
332 19. Diabetic autonomic neuropathy measurement and management, part II: management

[59] Sayeski PP, Kudlow JE. Glucose metabolism to glucos- Prev 2019;8(1):54 61. Available from: https://doi.
amine is necessary for glucose stimulation of trans- org/10.15171/jrip.2019.11.
forming growth factor-alpha gene transcription. J Biol [72] Book WM. Carvedilol: a nonselective β blocking agent
Chem 1996;271:15237 43. with antioxidant properties. Congestive Heart Fail
[60] Yarandi S, Srinivasan S. Diabetic gastrointestinal motility 2002;8:173 90. Available from: https://doi.org/
disorders and the role of enteric nervous system: current 10.1111/j.1527-5299.2002.00718.x.
status and future directions. Neurogastroenterol Motil [73] Ayashi S, Assareh AR, Jalali MT, Olapour S, Yaghooti
2014;26(5):611 24. Available from: https://doi.org/ H. Role of antioxidant property of carvedilol in mild
10.1111/nmo.12330 Published online 2014 Mar 24. to moderate hypertensive patients: A prospective
[61] Rhee SH, Pothoulakis C, Mayer EA. Principles and open-label study. Ind J Pharmacol 2016;48:372 6.
clinical implications of the brain-gut-enteric microbio- [74] Vinik AI, Murray GL. Autonomic neuropathy is treat-
ta axis. Nat Rev Gastroenterol Hepatol 2009;6:306 14. able. US Endocrinol 2008;2:82 4.
[62] Münzel T, Camici GG, Maack C, Bonetti NR, Fuster V, [75] Lieb D, Parson H, Mamikunian G, Vinik A. Cardiac auto-
Kovacic JC. Impact of oxidative stress on the heart and nomic imbalance in newly diagnosed and established
vasculature: part 2 of a 3-part series. J Am Coll diabetes is associated with markers of adipose tissue
Cardiol 2017;70(2):212 29. Available from: https:// inflammation. Exp Diabetes Res 2011;2012:1 8.
doi.org/10.1016/j.jacc.2017.05.035. [76] Diabetes Control and Complications Trial (DCCT). The
[63] Maack C, Kartes T, Kilter H, et al. Oxygen free radical effect of intensive diabetes therapy on measures of auto-
release in human failing myocardium is associated with nomic nervous system function in the diabetes control
increased activity of rac1-GTPase and represents a target and complications trial (DCCT). Diabetologia. 1998; 41:
for statin treatment. Circulation 2003;108:1567 74. 416 423. DCCT EDIC research group. intensive diabetes
[64] Mollnau H, Oelze M, August M, et al. Mechanisms of treatment and cardiovascular disease in patients with
increased vascular superoxide production in an exper- type 1 diabetes. N Engl J Med 2005; 353: 2643 2653.
imental model of idiopathic dilated cardiomyopathy. [77] Gaede P, Vedel P, Larsen N, Jensen G, Parving H,
Arterioscler Thromb Vasc Biol 2005;25:2554 9. Pedersen O. Multi-factorial intervention and cardio-
[65] Belch JJ, Bridges AB, Scott N, Chopra M. Oxygen free vascular disease in patients with type2 diabetes. N
radicals and congestive heart failure. Brit Heart J Engl J Med 2003;348:383 93.
1991;65:245 8. [78] Ziegler D, Schatz H, Conrad F, Gries FA, Ulrich H,
[66] Lopaschuk GD, Ussher JR, Folmes CD, Jaswal JS, Reichel G. Effects of treatment with the antioxidant
Stanley WC. Myocardial fatty acid metabolism in alpha-lipoic acid on cardiac autonomic neuropathy in
health and disease. Physiol Rev 2010;90:207 58. NIDDM patients. A 4-month randomized controlled
[67] Niemann B, Rohrbach S, Miller MR, Newby DE, Fuster V, multicenter trial (DEKAN study). Deutsche Kardiale
Kovacic JC. Oxidative stress and cardiovascular risk: Autonome Neuropathie. Diabetes Care
Obesity, diabetes, smoking, and pollution: Part 3 of a 3- 1997;20:369 73.
part series. J Am Coll Cardiol 2017;70(2):230 51. [79] Vinik AI, Ullal J, Parson HK, Casellini CM. Diabetic
Available from: https://doi.org/10.1016/j.jacc.2017.05.043. neuropathies: clinical manifestations and current treat-
[68] Anderson EJ, Kypson AP, Rodriguez E, Anderson CA, ment options. Nat. Clin.Pract. Endocrinol. Metab.
Lehr EJ, Neufer PD. Substratespecific derangements in 2006;2(4):1 13.
mitochondrial metabolism and redox balance in the [80] Colombo, J, Arora, RR, DePace, NL, Vinik, AI. Clinical
atrium of the type 2 diabetic human heart. J Am Coll autonomic dysfunction: measurement, indications,
Cardiol 2009;54:1891 8. therapies, and outcomes. New York, NY: Springer
[69] Brownlee M. Biochemistry and molecular cell biology Science 1 Business Media; 2014.
of diabetic complications. Nature 2001;414:813 20. [81] Vinik AI, Bloom HL, Colombo J. Differential effects
[70] de Kreutzenberg SV, Ceolotto G, Cattelan A, Pagnin E, of adrenergic antagonists (carvedilol vs. metoprolol)
Mazzucato M, Garagnani P, et al. Metformin improves on parasympathetic and sympathetic activity: a com-
putative longevity effectors in peripheral mononuclear parison of measures. Heart Int Heart Int 2014;9
cells from subjects with prediabetes. A randomized (1):7 14. Available from: https://doi.org/10.5301/
controlled trial. Nutr Metab Cardiovasc Dis 2015;25 HEART.2014.12495.
(7):686 93. Available from: https://doi.org/10.1016/j. [82] Bloom HL, Vinik AI, Colombo J. Differential effects of
numecd.2015.03.007 Epub 2015 Mar 24. adrenergic antagonists (carvedilol vs. metoprolol) on para-
[71] Hasanpour Dehkordi A, Abbaszadeh A, Mir S, sympathetic and sympathetic activity: a comparison of
Hasanvand A. Metformin and its anti-inflammatory clinical results. Heart Int 2014;9(1):15 21. Available from:
and anti-oxidative effects; new concepts. J Ren Inj https://doi.org/10.5301/HEART.2014.12496.

Diabetic Neuropathy
 References 333
[83] Vinik AI. A medicinal food provides food for thought symptomatic diabetic peripheral neuropathy with the
in managing diabetic neuropathy. Am J Med anti-oxidant alpha-lipoic acid. A 3-week multicentre
2013;126(2):95 6. Available from: https://doi.org/ randomized controlled trial (ALADIN study).
10.1016/j.amjmed.2012.08.008. Diabetologia 1995;38:1425 33.
[84] Adly, AAM. Oxidative stress and disease: an [97] Reljanovic M, Reichel G, Rett K, Lobisch M, Schuette
updated review. Res J Immunol 3: 129-145. doi: K, Moller W, et al. Treatment of diabetic polyneuro-
10.3923/rji.2010.129.145. pathy with the antioxidant thioctic acid (alpha-lipoic
[85] Murray G. Oxidative cardiac autonomic neuropathy: acid): a two year multicenter randomized double-
a unifying paradigm in adult sudden death with nor- blind placebo-controlled trial (ALADIN II). Alpha
mal left ventricular fraction; early diagnosis and lipoic acid in diabetic neuropathy. Free Radic Res
treatment reduces sudden cardiac death at least 43%. 1999;31(3):171 9.
EC Cardiol 2020; in press. [98] Ziegler D, Nowak H, Kempler P, Vargha P, Low PA.
[86] Ziegler D, Gries F. Alpha-lipoic acid and the treat- Treatment of symptomatic diabetic polyneuropathy
ment of diabetic peripheral autonomic cardiac neu- with the antioxidant alpha-lipoic acid: a meta-analy-
ropathy. Diabetes 1997;46(Suppl 2):S62 6. sis. Diabet Med 2004;21:114 21.
[87] Prendergast JJ. Diabetic autonomic neuropathy: part [99] Tesfaye S, Boulton AJ, Dyck PJ, Freeman R, Horowitz
1. Early detection. Pract Diabetol 2001;7 14. M, Kempler P, et al.Toronto diabetic neuropathy
[88] Prendergast JJ. Diabetic autonomic neuropathy: part expert group Diabetic neuropathies update on defini-
2. Treat Pract Diabetol 2001;30 6. tions diagnostic criteria estimation of severity and
[89] Ziegler D, Ametov A, Barinov A, Dyck PJ, Gurieva I, treatments. Diabetes Care 2010;33:2285 93.
Low PA, et al. Oral treatment with alpha-lipoic acid [100] Motooka M, Koike H, Yokoyama T, Kennedy NL.
improves symptomatic diabetic polyneuropathy: The Effect of dog-walking on autonomic nervous activity
SYDNEY 2 trial. Diabetes Care 2006;29(11):2365 70. in senior citizens. Med J Aust 2006;184:60 3.
[90] Rudich A, Tirosh A, Potashnik R, Khamaisi M, [101] Mortensen SA, Rosenfeldt F, Kumar A, Dolliner P,
Bashan N. Lipoic acid protects against oxidative Filipiak KJ, Pella D, et al. Q-SYMBIO Study inves-
stress induced impairment in insulin stimulation of tigators. The effect of coenzyme Q10 on morbidity
protein kinase B and glucose transport in 3T3-L1 adi- and mortality in chronic heart failure: Results
pocytes. Diabetologia 1999;42:949 57. from Q-SYMBIO: a randomized double-blind trial.
[91] Maddux BA, See W, Lawrence Jr JC, Goldfine AL, JACC Heart Fail 2014;2(6):641 9. Available from:
Goldfine ID, Evans JL. Protection against oxidative https://doi.org/10.1016/j.jchf.2014.06.008 Epub
stress-induced insulin resistance in rat L6 muscle 2014 Oct 1.
cells by micromolar concentrations of alpha-lipoic [102] Rosenfeldt FL, Haas SJ, Krum H, Hadj A, Ng K,
acid. Diabetes 2001;50:404 10. Leong JY, et al. Coenzyme Q10 in the treatment
[92] Jacob S, Streeper RS, Fogt DL, Hokama JY, Tritschler of hypertension: a meta-analysis of the clinical trials. J
HJ, Dietze GJ, et al. The antioxidant alpha-lipoic acid Hum Hypertens 2007;21(4):297 306 Epub 2007 Feb 8.
enhances insulin-stimulated glucose metabolism in [103] Rosenfeldt F, Hilton D, Pepe S, Krum H. Systematic
insulin-resistant rat skeletal muscle. Diabetes review of effect of coenzyme Q10 in physical exer-
1996;45:1024 9. cise, hypertension and heart failure. Biofactors
[93] Manning PJ, Sutherland WHF, Walker RJ, Williams SM, 2003;18(1-4):91 100 Review.
De Jong SA, Ryalls AR, et al. Effect of high-dose Vitamin [104] Deichmann R, Lavie C, Andrews S. Coenzyme q10
E on insulin resistance and associated parameters in and statin-induced mitochondrial dysfunction.
overweight subjects. Diabetes Care 2004;27:2166 71. Ochsner J 2010;10(1):16 21 Spring.
[94] Ametov AS, Barinov A, Dyck PJ, Hermann R, Kozlova [105] Piña IL, Di Palo KE, Ventura HO. Psychopharma-
N, Litchy WJ, et al. The sensory symptoms of diabetic cology and cardiovascular disease. J. ACC 2018;71
polyneuropathy are improved with alpha-lipoic acid: (20):2346 59.
The Sydney trial. Diabetes Care 2003;26:770 6. [106] American Heart Association Nutrition
[95] Ruhnau KJ, Meissner HP, Finn R, Reljanovic M, CommitteeLichtenstein AH, Appel LJ, Brands M,
Lobisch M, Schutte K, et al. Effects of 3-week oral Carnethon M, Daniels S, Franch HA, et al. Diet and
treatment with the antioxidant thioctic acid (alpha- lifestyle recommendations revision 2006: a scientific
lipoic acid) in symptomatic diabetic polyneuropathy. statement from the american heart association nutri-
Diabet Med 1999;16(12):1040 3. tion committee. Circulation 2006;114(1):82 96 Epub
[96] Ziegler D, Hanefeld M, Ruhnau KJ, Meissner HP, 2006 Jun 19. Erratum in: Circulation. 2006 Dec 5;114
Lobisch M, Schutte K, et al. Treatment of (23):e629. Circulation. 2006 Jul 4;114(1):e27.

Diabetic Neuropathy
334 19. Diabetic autonomic neuropathy measurement and management, part II: management

[107] Matsuo M, Kaneko T. The chemistry of reactive oxy- exercise and antioxidant supplementation: effects on
gen species and related free radicals. In: Radak Z, lipid profile and insulin sensitivity. J.Sports Sci. Med.
editor. Free radicals in exercise and aging. Hum Kin. 2017;16(3):375 82 eCollection 2017 Sep.
Leeds; 2000. p. 1 34. [119] Villafaina S, Collado-Mateo D, Fuentes JP, Merellano-
[108] Carnethon MR, Prineas RJ, Temprosa M, Zhang ZM, Navarro E, Gusi N. Physical exercise improves heart rate
Uwaifo G, Molitch ME. The association among auto- variability in patients with type 2 diabetes: a systematic
nomic nervous system function, incident diabetes, review. Curr Diab Rep 2017;17(11):110. Available from:
and intervention arm in the diabetes prevention pro- https://doi.org/10.1007/s11892-017-0941-9 Review.
gram. Diabetes Care 2006;29:914 19. [120] Ostman C, Smart NA, Morcos D, Duller A, Ridley W,
[109] Voulgari C, Pagoni S, Vinik A, Poirier P. Exercise Jewiss D. The effect of exercise training on clinical
improves cardiac autonomic function in obesity and outcomes in patients with the metabolic syndrome: a
diabetes. Metabolism 2013;62:609 21. systematic review and meta-analysis. Cardiovasc
[110] Powers SK, Jackson MJ. Exercise-induced oxidative Diabetol 2017;16(1):110. Available from: https://doi.
stress: cellular mechanisms and impact on org/10.1186/s12933-017-0590-y PMID: 28854979.
muscle force production. Physiol Rev 2008;88: [121] Myers J, Atwood JE, Froelicher V. Active lifestyle
1243 76. and diabetes. Circulation. 2003;107(19):2392 4.
[111] Zembroń-Łacny A, Szyszka K, Hübner-Woźniak E. [122] Nichols S, Gleadall-Siddall DO, Antony R, Clark AL,
The pro-antioxidant-oxidant balance in the blood of Cleland JGF, Carroll S, et al. Estimated peak func-
middle—distance runners. Biol Sport 1998;16:51 9. tional capacity: an accurate method for assessing
[112] Alessio HM, Hagerman AE, Fulkerson BK, Ambrose change in peak oxygen consumption after cardiac
J, Rice RE, et al. Generation of reactive oxygen spe- rehabilitation? Clin Physiol Funct Imag 2017;
cies after exhaustive aerobic and isometric exercise. Available from: https://doi.org/10.1111/cpf.12468.
Med Sci Sports Exerc 2000;32(9):1576 81. [123] Palermo P, Corrà U. Exercise prescriptions for train-
[113] Goto C, Higashi Y, Kimura M, Noma K, Hara K, ing and rehabilitation in patients with heart and lung
et al. Effect of different intensities of exercise on disease. Ann Am Thorac Soc 2017;14(Supplement_1):
endothelium-dependent vasodilation in humans: role S59 66. Available from: https://doi.org/10.1513/
of endothelium-dependent nitric oxide and oxidative AnnalsATS.201702-160FR.
stress. Circulation 2003;108(5):530 5. [124] Abdul-Jawad Altisent O, Puri R, Regueiro A,
[114] Knez WL, Jenkins DG, Coombes JS. Oxidative stress Chamandi C, Rodriguez-Gabella T, Del Trigo M, et al.
in half and full Ironman triathletes. Med Sci Sports Predictors and association with clinical outcomes of
Exerc 2007;39(2):283 8. the changes in exercise capacity after transcatheter
[115] Pilch W, Szygula Z, Tyka AK, Palka T, Tyka A, Cison aortic valve replacement. Circulation 2017;136
T, et al. Disturbances in pro-antioxidant-oxidant bal- (7):632 43. Available from: https://doi.org/10.1161/
ance after passive body overheating and after exer- CIRCULATIONAHA.116.026349 Epub 2017 Jun 6.
cise in elevated ambient temperatures in athletes and [125] Talwar A, Sahni S, Verma S, Khan SZ, Dhar S, Kohn
untrained men. PLoS One 2014;9(1):e85320. Available N. Exercise tolerance improves after pulmonary reha-
from: https://doi.org/10.1371/journal.pone.0085320 bilitation in pulmonary hypertension patients. J
eCollection 2014. Exerc Rehabil 2017;13(2):214 17. Available from:
[116] Ackland GL, Minto G, Clark M, Whittle J, Stephens https://doi.org/10.12965/jer.1732872.436 eCollection
RCM, Owen T, et al. Autonomic regulation of sys- 2017 Apr PMID:28503536.
temic inflammation in humans: a multi-center, [126] Cheng ST, Wu YK, Yang MC, Huang CY, Huang HC,
blinded observational cohort study. Brain Behav Chu WH, et al. Pulmonary rehabilitation improves
Immun 2017; Available from: https://doi.org/ heart rate variability at peak exercise, exercise capac-
10.1016/j.bbi.2017.08.010 pii: S0889-1591(17)30398-7. ity and health-related quality of life in chronic
[117] S Delevatti R, Schuch FB, Kanitz AC, Alberton CL, obstructive pulmonary disease. Heart & Lung
Marson EC, Lisboa SC, et al. Quality of life and sleep 2014;43(3):249 55. Available from: https://doi.org/
quality are similarly improved after aquatic or 10.1016/j.hrtlng.2014.03.002 Epub 2014 Mar 29.
dry-land aerobic training in patients with type 2 dia- PMID:24685394.
betes: a randomized clinical trial. J Sci Med Sport [127] Karssemeijer EGAE, Aaronson JAJ, Bossers WJW, Smits
2017; Available from: https://doi.org/10.1016/j. TT, Rikkert MGMMO, Kessels RPCR. Positive effects of
jsams.2017.08.024 pii: S1440-2440(17)31029-0. combined cognitive and physical exercise training on
[118] Yfanti C, Tsiokanos A, Fatouros IG, Theodorou AA, cognitive function in older adults with mild cognitive
Deli CK, Koutedakis Y, et al. Chronic eccentric impairment or dementia: a meta-analysis. Ageing Res

Diabetic Neuropathy
 References 335
Rev 2017; Available from: https://doi.org/10.1016/j. [132] Selye H. The stress of life. 2nd (ed.) New York, NY,
arr.2017.09.003 [Epub ahead of print] Review. USA: McGraw-Hill; 1956.
[128] Ebrahimi K, Majdi A, Baghaiee B, Hosseini SH, Sadigh- [133] Ulrich-Lai YM, Herman JP. Neural regulation of
Eteghad S. Physical activity and beta-amyloid pathol- endocrine and autonomic stress responses. Nat Rev
ogy in Alzheimer’s disease: a sound mind in a sound Neurosci 2009;10:397 409.
body. EXCLI J 2017;16:959 72. Available from: [134] McEwen BS. The neurobiology of stress: From
https://doi.org/10.17179/excli2017-475 eCollection 2017. serendipity to clinical relevance. Brain Res.
Review. PMID: 28900376. 2000;886:172 89.
[129] Cheng ST. Cognitive reserve and the prevention of [135] Chrousos GP. Stress and disorders of the stress sys-
dementia: the role of physical and cognitive activities. tem. Nat Rev Endocrinol 2009;5:374 81.
Curr Psych Rep 2016;18(9):85. Available from: [136] Stratakis CA, Chrousos GP. Neuroendocrinology and
https://doi.org/10.1007/s11920-016-0721-2 Review. pathophysiology of the stress system. Ann NY Acad
[130] Richter EA, Ruderman NB. AMPK and the biochem- Sci 1995;771:1 18.
istry of exercise: Implications for human health and [137] Bertone-Johnson ER, Whitcomb BW, Missmer SA,
disease. Biochem J 2009;418(2):261 75. Available Karlson EW, Rich-Edwards JW. Inflammation and early-
from: https://doi.org/10.1042/BJ20082055 Review. life abuse in women. Am J Prev Med 2012;43:611 20.
[131] Cardinali DP. Autonomic nervous system: basic and clin-
ical aspects. Springer International Publishing AG; 2018.

Diabetic Neuropathy
 C H A P T E R

20
Translating a treatment for diabetic
peripheral neuropathy from rodents to
humans: can a case be made for fish oil
and salsalate?
Mark Yorek1,2,3
1
Department of Internal Medicine, University of Iowa, Iowa City, IA, United States 2Department of
Veterans Affairs Iowa City Health Care System, Iowa City, IA, United States 3Fraternal Order of
Eagles Diabetes Research Center, University of Iowa, Iowa City, IA, United States

Diabetic peripheral neuropathy type 2 diabetes [1,2,5]. This scenario is further

complicated by the realization that both animal
Diabetic peripheral neuropathy (DPN) affects and human subjects with impaired glucose tol-
about 50% of the diabetes population and an erance and insulin resistance develop forms of
effective treatment has eluded investigators for sensory neuropathy [6 8]. There are many pos-
decades in spite of repeated successful discover- sible reasons for the lack of success of transla-
ies in animal models. Human and animal stud- tion of effective treatments in animal models for
ies have widely demonstrated that the etiology DPN to human subjects. This includes the possi-
of DPN is complex with multiple mechanisms bility that diabetic animals do not adequately
contributing to vascular and nerve dysfunction, model human DPN, poor design of clinical
which are believed to be the primary factors trials such as duration of disease, selected pri-
responsible for sensory and motor nerve mary endpoints and length of treatment com-
impairment [1 5]. Hyperglycemia has long promises the overall outcome of the study, large
been thought to be the primary cause of periph- placebo effect commonly seen in human trials
eral neuropathy in patients with type 1 diabetes for DPN for some endpoints masks any true
but numerous studies have shown that good benefit, and singular interventions targeting
glycemic control is not sufficient to significantly only one potential mechanism for a disease
alter the course of this disease in patients with with multiple etiologies. Whatever the reasons

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00006-2 337 © 2022 Elsevier Inc. All rights reserved.
338 20. Translating a treatment for diabetic peripheral neuropathy from rodents to humans: can a case be made for fish oil and salsalate?

for the failure of past clinical trials there mammals, EPA and DHA can also be produced
remains a dire need to find a treatment that will in small amounts from α-linolenic acid. This
reduce the impact of DPN on patients, family, synthesis is thought to occur through a series of
and society. In this chapter, we will examine desaturation (involving Δ6, Δ5, and Δ4 desa-
evidence that the combination of fish oil turases) and elongation steps (Fig. 20.1) [13].
[enriched in the omega-3 polyunsaturated fatty However, it is now considered more likely that
acids (PUFAs); eicosapentaenoic acid (20:5, DHA is biosynthesized via a 24 carbon interme-
EPA), and docosahexaenoic acid (22:6, DHA)] diate fatty acid followed by beta-oxidation in
and salsalate may be an effective treatment for peroxisomes. In this pathway, EPA is twice
DPN and worthy of advancing to a clinical trial. elongated, yielding 24:5 ω-3, then desaturated to
24:6 ω-3, then shortened to DHA via beta-
oxidation. This pathway is known as Sprecher’s
Omega-3 polyunsaturated fatty acid; shunt [13,14].
eicosapentaenoic acid and In subjects with diabetes, the formation of EPA
docosahexaenoic acid and DHA via α-linolenic acid is impaired [11,15].
Furthermore, diabetes-induced changes in Δ-5
Omega-6 and omega-3 PUFA are essential desaturase activity causes an increase in the
fatty acids with multiple double bonds, the omega-6 to omega-3 fatty acid ratio, further pro-
first of which is located at the sixth or third moting an inflammatory state that contributes to
carbon from the methyl end of the molecule, the pathologies associated with obesity and diabe-
respectively [9]. Mammals lack enzymes that tes as well exacerbation of DPN [16 20]. Thus,
insert a double bond in the omega-6 and increasing the circulatory levels of EPA and
omega-3 position of fatty acids; therefore lino- DHA, and favorably altering the omega-6 to
leic acid (18:2, omega-6 PUFA), γ-linolenic acid omega-3 fatty acid ratio could be an effective ther-
(18:3, omega-6 PUFA), and α-linolenic acid apy for treating DPN. The most effective
(18:3, omega-3 PUFA), as well as some of their approach to increasing circulating levels of EPA
elongation products including EPA and DHA, and DHA is through the diet and consuming
are deemed to be essential dietary nutrients foods enriched in these lipids such as oily fish or
[10,11]. The omega-6 fatty acids are derived by taking supplements such as fish oil capsules.
from the diet primarily as linoleic acid and are Different forms of algae have also been isolated
elongated in vivo to γ-linolenic acid and ara- and commercially grown that produce primarily
chidonic acid (AA) [9]. Most eicosanoids EPA or DHA but studies of the impact of these
derived from AA are proinflammatory, selectively enriched omega-3 PUFA oils on health
whereas metabolites derived from omega-3 outcomes are lacking. However, there is great
PUFA have antiinflammatory properties. The interest in pursuing studies of these oils for
balance of omega-6 to omega-3 fatty acid con-
sumption can affect systemic inflammatory
processes and immune activity [9]. The ratio of
consumed omega-6 to omega-3 fatty acids in
the Western diet is approximately 15 to 1,
whereas a ratio of 4 to 1 is considered optimal
[12]. Plants, seeds, nuts and their oils are a
good source for α-linolenic acid, but fish oils
are the primary source for the long-chain FIGURE 20.1 Synthesis in vivo of eicosapentaenoic and
omega-3 PUFA such as EPA and DHA. In docosahexaenoic acids.

Diabetic Neuropathy
 Omega-3 polyunsaturated fatty acid and cardiovascular health 339
environmental reasons. Few side effects have have been conducted examining the effect of
been associated with taking the recommended omega-3 PUFA on cardiovascular disease did
doses of fish oil capsules containing omega-3 not include blood levels of fatty acids in their
PUFA. The minor side effects that have been outcome measures. This flaw in study design
reported include loose stools, stomach upset, and creates a doubt in the outcome of these studies
belching (that includes a “fish stick” like the taste). when negative since variability in subjects
However, consumption of omega-3 PUFA in iso- enrolled in these trials achieving a therapeutic
lated cases may raise the risk of bleeding and this blood level of omega-3 PUFA can be influ-
can be potentially serious for subjects with bleed- enced by many factors including, genetic vari-
ing disorders or taking blood-thinning medica- ability, noncompliance or low dosing as well as
tions. Thus, subjects with a known risk for by gender, age, body weight, and physical
bleeding disorders are consulted to seek the activity [39]. For example, in the recent
advice of their physician before taking fish oil- ASCEND study, 15,480 subjects with diabetes
containing capsules. received 1 g capsules daily containing omega-3
fatty acids or placebo. In this study, the pri-
mary endpoint was a first serious vascular
Omega-3 polyunsaturated fatty acid and event. After a mean follow-up period of 7.4
cardiovascular health years, the study found no significant difference
in the risk of serious vascular events between
There is significant rationale to pursue fish oil those assigned to receive omega-3 fatty acids
and salsalate as a potential treatment of DPN. or placebo [40]. In contrast, the REDUCE-IT
Cardiovascular disease has been the primary dis- trial enrolled 8179 subjects that included sub-
order to be evaluated for treatment by omega-3 jects with diabetes to receive a 4 g daily dose of
PUFA [21 26]. However; additional studies sug- icosapent ethyl, a highly enriched form of an
gest potential benefits of omega-3 PUFA con- ester derivative of EPA, or placebo. After a
sumption on other chronic diseases including follow-up period of 4.9 years, icosapent ethyl was
chronic inflammatory and autoimmune diseases, found to have a statistical benefit on reducing
stroke, muscle atrophy, and neurodegenerative ischemic events in subjects with hypertriglyceride-
disease [27 34]. Nonetheless, the consumption mia [41]. In August of 2019, the American Heart
of omega-3 PUFA remains low in the Western Association released a Science Advisory based on
diet due to historically increased consumption of examination of the REDUCE-IT trial and other
omega-6 PUFA. studies using “pharmacological doses” of omega-
There has been a mixed response and some- 3 fatty acids ( . 3 g/day total EPA 1 DHA) stat-
times controversial regarding the potential ing that omega-3 fatty acids (EPA 1 DHA or
benefits of omega-3 PUFA on cardiovascular EPA-only) are an effective and safe option for
disease [35,36]. However, recent literature is reducing triglycerides as monotherapy or as an
describing that beneficial effects of omega-3 adjunct to other lipid-lowering agents [42].
PUFA supplementation on cardiovascular dis- Was the apparent success of the REDUCE-IT
ease risk are apparent when subjects achieve trial due to increased dosing of the omega-3
higher blood levels of omega-3 fatty acids and PUFA derivative, study design, or other fac-
when the omega-3 index is in a healthy range tors? This question is difficult to answer
[37 39]. The omega-3 index is defined as the because neither trial analyzed circulation levels
sum of EPA and DHA as a percentage of total of omega-3 PUFA or their metabolites before
fatty acids in red blood cells [39]. This raises an and after treatment. The failure to measure
important issue because most of the trials that levels of omega-3 fatty acids in the blood of

Diabetic Neuropathy
340 20. Translating a treatment for diabetic peripheral neuropathy from rodents to humans: can a case be made for fish oil and salsalate?

FIGURE 20.2 Lipid structures.

subjects enrolled in trials examining the poten-

tial benefits of fish oils is a serious error.
Individual variability could be a limiting factor
in the success or failure of omega-3 PUFA ther-
apy. This is highlighted by a study of a
Japanese population that reported 4 g of fish
oil per day failed to achieve the target EPA/
FIGURE 20.3 Structure of aspirin and salsalate.
AA ratio in the blood in 16% of the enrolled
subjects [43]. In our animal studies, to be sum-
marized in this chapter, fish oil treatment vari- human population and disease being studied.
ability in the accumulation of omega-3 fatty For instance, it is unknown if diabetes in the
acids in blood and tissues is not an issue. absence or presence of DPN will affect the
Animals used for preclinical studies are geneti- omega-3 index or profile of production of the anti-
cally similar and gender, age, body weight, inflammatory mediators of omega-3 PUFA. Since
and activity are all usually well matched. the antiinflammatory metabolites of omega-3
However, this is not the case in studies of PUFAs have neuroprotective properties, we
human subjects and it is important that each hypothesize that a therapy leading to increasing
subject enrolled in future studies be evaluated their production in vivo may be a potential
for changes in omega-3 fatty acid levels as an disease-modifying therapeutic approach for DPN.
outcome measure. Moreover, since the metabo- Aspirin has been reported to indirectly
lites of omega-3 PUFA such as resolvins and increase the production of resolvins [45],
neuroprotectin-1 (NPD-1) are the potential although the long-term use of an effective dose
mediators of the beneficial effects of fish oil the of aspirin is prohibitive due to multiple unfa-
circulating levels of these compounds should vorable side effects. However, salsalate, a
also be determined (Fig. 20.2) [44]. However, prodrug of salicylate, is well tolerated and con-
there is no information from clinical trials sidered safe for long-term clinical use
using omega-3 PUFA on the levels of these (Fig. 20.3) [46,47]. Both aspirin and salsalate are
antiinflammatory metabolites. Furthermore, it rapidly broken down to salicylate in vivo [48].
is unknown what variability may exist for the In high-fat-fed mice salsalate activates brown
accumulation of omega-3 PUFA in circulation adipose tissue and improves glucose tolerance
(omega-3 index) or in the formation of their [49]. In GK rats salsalate increased levels of
metabolites due to diabetes. Therefore future adiponectin and decreased adipose-tissue-
studies of the potential benefits of omega-3 derived inflammation [50]. In patients with
PUFA as disease modifying therapy should diabetes, salsalate alleviated insulin resistance,
also include matched control subjects in order improved metabolic control, and improved the
to determine what variability may exist in the lipid, lipoprotein, and apoprotein profile of
metabolism of these fatty acids within the insulin-resistant individuals who were

Diabetic Neuropathy
 Omega-3 polyunsaturated fatty acid and peripheral neuropathy related to obesity and insulin resistance 341
overweight or obese [46,47,51,52]. In our pre- reactive oxygen species. We have also demon-
clinical studies salsalate alone provided some strated that blocking the angiotensin system
improvement in DPN (to be discussed further with an angiotensin-converting enzyme inhibi-
below); however, salsalate in combination with tor or angiotensin receptor blocker improved
fish oil increased the production of resolvins vascular relaxation by epineurial arterioles as
and had greater benefit toward vascular and well as peripheral neuropathy in both type 1
neural endpoints associated with DPN than fish and type 2 diabetic rats [64 66]. Working with
oil alone [53,54]. cultured vascular endothelial cells, Sakai et al.
demonstrated that fish oil-derived omega-3
PUFA prevented oxidative stress-induced DNA
Omega-3 polyunsaturated fatty acid and damage [67]. In ob/ob mice a diet enriched in
vascular dysfunction omega-3 PUFA significantly improved insulin
resistance and vascular reactivity of the aorta
Our studies as well as others have shown [68]. In KKAy mice, a model of type 2 diabetes,
that vascular dysfunction is a contributing fac- EPA ethyl ester was shown to improve endothe-
tor to the development and progression of DPN lial dysfunction and vascular reactivity to ace-
[55 57]. Thus, any potential treatment for DPN tylcholine in aortic rings [69]. In human
should also improve and/or protect vascular subjects, it has been demonstrated that omega-3
function. A number of studies in animal models PUFA consumption improved small peripheral
of obesity and diabetes including our own have artery function in patients at intermediate-high
demonstrated that treating rodents with omega- cardiovascular risk [70]. In addition, it has been
3 PUFA improves vascular function. Our stud- shown that increasing the EPA to AA ratio is
ies with type 2 diabetic rats treated with a diet associated with improved arterial stiffness in
enriched with menhaden oil have demonstrated obese patients with dyslipidemia [71]. In sum-
significant improvement in vascular reactivity mary, multiple studies in different animal mod-
of epineurial arterioles, blood vessels that pro- els of diabetes as well as human subjects have
vide circulation to the sciatic nerve, to acetyl- demonstrated that omega-3 PUFA derived from
choline and calcitonin gene-related peptide fish oil protects endothelial cells from oxidative
[58 60]. Other animal studies have reported stress and vascular reactivity by different vascu-
that in type 2 diabetic rats a diet enriched in lar beds.
DHA prevented diabetic retinopathy by inhibit-
ing retinal vascular damage [61]. In type 2 dia-
betic Otsuka Long-Evans Tokushima fatty rats Omega-3 polyunsaturated fatty acid and
EPA ameliorated endothelial dysfunction in peripheral neuropathy related to obesity
mesenteric arteries by correcting the imbalance and insulin resistance
between endothelium-derived factors in part by
inhibiting extracellular signal-regulated kinase, Studies in animal models of prediabetes and
decreasing nuclear factor-κB and reducing humans with impaired glucose tolerance and
cyclooxygenase-2 expression [62]. Farooq et al. insulin resistance have been shown to develop
in studies using “old rats” showed that treating forms of sensory neuropathy [6 8]. We have
these rats with an omega-3 formulation of 6:1 shown that rats fed a high-fat diet progres-
ratio of EPA and DHA improved endothelium- sively develop impaired glucose utilization
dependent relaxations [63]. An effect they attrib- and slowing of sensory nerve conduction
uted to prevention of the upregulation of the velocity, loss of sensory nerves in the skin and
local angiotensin system and reduction of cornea and decreased sensitivity of

Diabetic Neuropathy
342 20. Translating a treatment for diabetic peripheral neuropathy from rodents to humans: can a case be made for fish oil and salsalate?

intraepidermal nerve fibers to a thermal stimu- consume lean fish are at a reduced risk of
lus and corneal nerves to Cochet Bonnet fila- developing type 2 diabetes [86,87]. However,
ment esthesiometer [60]. All these deficits were this remains controversial since an earlier study
reversed by enriching their high-fat diet with found no evidence that higher consumption of
menhaden oil [60]. Obesity as described in our long-chain omega-3 PUFA and fish reduces the
high-fat-fed rat model is often associated with risk of type 2 diabetes [88]. In summary, pre-
low-grade inflammation due to overproduction clinical studies and studies with human subjects
of inflammatory mediators by adipose tissue have demonstrated that increasing the con-
[72]. Increased consumption of omega-3 PUFA sumption of omega-3 PUFA can improve many
or exogenous administration of resolvins, antiin- of the pathogenic characteristics associated with
flammatory and proresolving mediators endog- obesity.
enously generated by EPA (E series resolvins)
and DHA (D series resolvins), reduce the
inflammatory condition (Fig. 20.4) [72]. Omega-3 polyunsaturated fatty acids and
In obesity, the increased inflammatory state diabetic peripheral neuropathy
and increased levels of leptin and reduced
levels of adiponectin are corrected with Omega-3 PUFA have long been implicated in
increased consumption of omega-3 PUFA or the development and continued maintenance of
exogenous treatment with resolvins [73 77]. In healthy nerves and may be beneficial for pro-
adipose tissue, it appears that increased con- moting peripheral nerve restoration [9]. Our
sumption of EPA and DHA lowers lipogenesis, body of work over the last 8 years have demon-
increases lipolysis, and inhibits the nuclear fac- strated that enriching the diet of chronic diabetic
tor-κB pathway thereby decreasing inflamma- mice and rats with menhaden (fish) oil, enriched
tion and oxidative stress [73,78,79]. Preclinical in EPA and DHA, can not only slow progression
studies have shown that increased consump- of DPN as determined by protection of motor
tion of omega-3 PUFA or exogenous treatment and sensory nerve conduction velocity and pre-
with resolvins improves insulin resistance in vention of thermal hypoalgesia but can stimulate
animal models of obesity and diabetes reversal of neurovascular and neuropathic defi-
[75,77,80 82]. Studies conducted with obese cits and repopulation of sensory nerves in the
and type 2 diabetic human subjects have also skin and cornea [54,58,59,89 91]. These latter
shown that increased consumption of long- data is of potential interest from a clinical and
chain omega-3 PUFA improves insulin sensi- translational point. Historically intervention
tivity [83 85]. Several studies have even trials for treatment of DPN have depended on
shown that individuals with higher serum measurement of large fiber function as the pri-
levels of omega-3 PUFA and those that mary endpoint and outcomes from these trials
have been disappointing. More recent studies
have suggested that damage to small sensory
nerve fibers in the skin and cornea precede large
nerve fiber damage and may be a better early
marker for DPN [92]. Clinical trials focusing on
structural and functional changes of these nerves
as a primary endpoint for the treatment of DPN
are needed. In this regard a proof-of-concept
FIGURE 20.4 Metabolites of eicosapentaenoic acid and study focusing on the effect of 12 months of
docosahexaenoic acid. omega-3 PUFA supplementation sought to

Diabetic Neuropathy
 Omega-3 polyunsaturated fatty acids and diabetic peripheral neuropathy 343
determine whether omega-3 PUFA could stop DPN or other diabetes complications and did
the known progression of diabetic sensorimotor not measure circulating levels of omega-3 PUFA
polyneuropathy in type 2 diabetes subjects by metabolites. In a much older study subjects with
measuring changes in corneal nerves and mea- type 2 diabetes were treated for 49 weeks with
sures of small and large fiber function [93]. In 1.8 g/day of ethyl ester of EPA [95]. A similar
this study subjects were treated for 12 months EPA formulation was used in the REDUCE-IT
with a 10 mL dose of seal oil containing 2.33 g of trial that was found to improve after 4.9 years of
omega-3 PUFA (0.75 g EPA, 0.56 g docosapen- treatment cardiovascular disease and reduce
taenoic acid, and 1.02 g DHA) daily. The pri- serum triglyceride levels [41]. In this study, the
mary outcome measure was the 1-year change authors also noted a decrease in serum triglycer-
in corneal fiber length measured by in vivo cor- ide levels in their treated subjects as well as an
neal confocal microscopy, with sensory and improvement in vascular function and clinical
nerve conduction measures as secondary out- symptoms related to sensory nerve function and
comes. Unfortunately the authors of this study vibration perception threshold of the lower
did not examine the change in omega-3 PUFA extremities [95]. Clearly more clinical studies are
in serum following the study or the effect of needed to determine whether omega-3 PUFA
treatment on the omega-3 index or change in may be an effective therapy for DPN. These
omega-6 to omega-3 PUFA ratio in serum. The trials need to be of sufficient duration and use a
primary outcome measure showed a 29% signif- therapeutic dose of omega-3 PUFA that not only
icant increase in corneal nerve fiber length. increases blood levels of omega-3 PUFA but also
There was no significant change in the second- their metabolites. The importance of measuring
ary outcome measures. The change in corneal circulating levels of omega-3 PUFA metabolites
fiber length in these type 1 diabetic subjects was is supported by our studies that have shown
comparative to the increase in corneal fiber that treating type 1 and type 2 diabetic mice
length we have shown to occur in type 1 dia- with daily injections of resolvin D1 improved
betic rats supplemented with menhaden oil in DPN [54,89,96].
their diet [59]. In another recent study omega-3 Since the antiinflammatory metabolites of
PUFA supplementation of type 1 diabetes sub- omega-3 PUFAs have neuroprotective proper-
jects was used to determine the effect on vascu- ties, it seems reasonable to hypothesize that a
lar health, glycemic control and metabolic therapy leading to increasing their production
parameters [94]. In this placebo controlled study in vivo may be a potential disease-modifying
type 1 diabetes subjects were treated for 6 therapeutic approach for DPN. Aspirin has
months with 3.3 g/day encapsulated omega-3 been reported to indirectly increase the pro-
PUFA. Following treatment for 6 months the duction of resolvins [45], although the long-
omega-3 PUFA index significantly increased term use of an effective dose of aspirin is
from 4.9% 6 0.9% to 8.3% 6 1.5%. However, in prohibitive due to multiple unfavorable side
this study enrichment in omega-3 PUFA did not effects (see Fig. 20.4). However, salsalate, a pro-
improve vascular health, glucose homeostasis, drug of salicylate, is well tolerated and consid-
or metabolic parameters. The authors concluded ered safe for long-term clinical use [46]. In
that their preliminary randomized controlled phase 3 clinical trial performed in patients
trial does not support the use of therapeutic with type 2 diabetes salsalate was shown to
omega-3 PUFA supplementation in the treat- decrease inflammatory mediators, lower fast-
ment and management of type 1 diabetes. It ing blood glucose and HbA1C levels, and
should be noted that this trial was of short dura- improve lipid profiles [47]. In our preclinical
tion, did not examine any endpoints related to studies salsalate alone provided minimal

Diabetic Neuropathy
344 20. Translating a treatment for diabetic peripheral neuropathy from rodents to humans: can a case be made for fish oil and salsalate?

improvement in DPN; however, salsalate in provided evidence from our studies that the
combination with fish oil increased the produc- metabolites of omega-3 PUFA, resolvins and
tion of resolvins and had greater benefit neuroprotectin, are likely responsible for the
toward DPN than fish oil alone (see Fig. 20.4) reported benefits of these lipids. Biologically
[53,54]. active concentrations of these antiinflammatory
In our studies, we have attributed the bene- lipid mediators in serum can be achieved fol-
ficial effects of menhaden oil on diabetic lowing oral omega-3 PUFA supplementation
peripheral neuropathy to the reduction of oxi- [105]. Moreover, preclinical studies from my
dative and inflammatory stress and the neuro- laboratory have demonstrated that salsalate can
protective properties of resolvins and substitute for aspirin and when combined with
neuroprotectin [97,98]. Other laboratories have fish oil increases the circulatory levels of resol-
reported similar results. Gerbi et al. reported vin and the efficacy of fish oil toward diabetic
that fish oil supplementation prevents peripheral neuropathy. Given the availability of
diabetes-induced slowing of nerve conduction omega-3 PUFA as an over the counter supple-
velocity, neuroanatomical changes in rats and ment and their safety profile, after multiple
improves sciatic nerve Na,K-ATPase activity studies of improvement in cardiovascular
[99]. Julu reported that essential fatty acids that health, enriching the daily intake of these lipids,
included EPA prevented slowing of nerve con- unless the individual has some form of a bleed-
duction velocity in streptozotocin-induced dia- ing disorder, seems like a reasonable approach
betic rats [100]. Some recent studies have to improve the health of subjects with diabetes.
suggested that treatment with fish oils may Given the body of evidence of the many poten-
also be beneficial in preventing painful diabetic tial benefits of increased consumption of
neuropathy. Heng et al. reported that dietary omega-3 PUFA it seems that it is time to seri-
supplementation of DHA inhibited mechanical ously consider human clinical trials using fish
allodynia and thermal hyperalgesia by decreas- oil and salsalate for the treatment of diabetic
ing the excitability of dorsal root ganglions peripheral neuropathy.
[101]. Li et al. reported that supplementing the
diet of diabetic rats with fish oil prevented
mechanical allodynia and thermal hyperalgesia
by blocking nuclear factor-κB-mediated inflam-
References
matory pathways [102]. Wagner et al. have [1] Feldman EL, Callaghan BC, Pop-Busui R, et al. Diabetic
shown in a murine diabetic model that DHA is neuropathy. Nat Rev Dis Prim 2019;5:41.
[2] Callaghan BC, Junguk H, Feldman EL. Diabetic neuropa-
metabolized into epoxydocosapentanoic acids thy: one disease or two? Curr Opin Neurol 2012;25:536 41.
via cytochrome P450 enzymes and these meta- [3] Jolivalt CG, Frizzi KE, Guernsey L, et al. Peripheral
bolites are analgesic against chronic pain and neuropathy in mouse models of diabetes. Curr Protoc
more efficacious than opioids [103]. Another Mouse Biol 2016;6:223 55.
study demonstrated that local administration [4] Yorek MA. Alternatives to the streptozotocin-diabetic
rodent. Int Rev Neurobiol 2016;127:89 112.
of DHA reverted hyperalgesia and allodynia in [5] Pop-Busui R, Boulton AJ, Feldman EL, et al. Diabetic
diabetic rats with the results indicating that the neuropathy: a position statement by the american dia-
antihyperalgesic effect of DHA involves indi- betes association. Diabetes Care 2017;40:136 54.
rect activation of μ and δ receptors [104]. [6] Cortez M, Singleton JR, Smith AG. Glucose intolerance,
In conclusion, I have provided a condensed metabolic syndrome, and neuropathy. Handb Clin
Neurol 2014;126:109 22.
overview of some of the beneficial effects [7] Papanas N, Ziegler D. Polyneuropathy in impaired glu-
attributed to omega-3 PUFA with the primary cose tolerance: is postprandial hyperglycemia the main
focus on diabetic peripheral neuropathy. I have culprit? A mini-review. Gerontology 2013;59:193 8.

Diabetic Neuropathy
 References 345
[8] Davidson EP, Coppey LJ, Kardon RH, et al. membranes: insight from NMR studies of model sys-
Differences and similarities in development of corneal tems. Biochim Biophy Acta 2015;1848:211 19.
nerve damage and peripheral neuropathy and in diet- [24] Simopoulos AP. The importance of the omega-6/
induced obesity and type 2 diabetic rats. Invest omega-3 fatty acid ratio in cardiovascular disease and
Ophthalmol Vis Sci 2014;55:1222 30. other chronic diseases. Exp Biol Med 2008;233:674 88.
[9] Zhang AC, MacIsaac RJ, Roberts L, et al. Omega-3 [25] Gil A, Serra-Majem L, Calder PC, et al. Systematic
polyunsaturated fatty acid supplementation for reviews of the role of omega-3 fatty acids in the preven-
improving peripheral nerve health: protocol for a sys- tion and treatment of disease. Br J Untr 2012;107:S1 2.
tematic review. BMJ Open 2018;8:e020804. [26] Kaliannan K, Li X-Y, Wang B, et al. Multi-omic analy-
[10] Harris WS, Dayspring TD, Moran TJ. Omega-3 fatty sis in transgenic mice implicates omega-6/omega-3
acids and cardiovascular disease: new developments fatty acid imbalance as a risk factor for chronic dis-
and applications. Postgrad Med 2013;125:100 13. ease. Commun Biol 2019;2:276.
[11] Coste TC, Gerbi A, Vague P, et al. Peripheral diabetic [27] Lanchais K, Capel F, Tournadre A. Could omega-3
neuropathy and polyunsaturated fatty acids supple- fatty acids preserve muscle health in rheumatoid
mentations: natural sources or biotechnological needs? arthritis? Nutrients 2020;12:223.
Cell Mol Biol 2004;50:845 53. [28] Ueno Y, Miyamoto N, Yamashiro K, et al. Omega-3
[12] Simopoulos AP. The importance of the ratio of omega- polyunsaturated fatty acids and stroke burden. Int J
6/omega-3 essential fatty acids. Biomed Pharmacother Mol Sci 2019;20:5549.
2002;56:365 79. [29] Li X, Bi X, Wang S, et al. Therapeutic potential of ω-3
[13] Bazan N, Molina M, Gordon W. Docosahexaenoic acid polyunsaturated fatty acids in human autoimmune
signalolipidomics in nutrition: significance in aging, diseases. Front Immunol 2019;10:2241.
neuroinflammation, macular degeneration, Alzheimer’s [30] Marton LT, Goulart RA, Carvalho ACA, et al. Omega
and other neurodegenerative diseases. Annu Rev Nutr fatty acids and inflammatory bowel diseases: an over-
2011;31:321 51. view. Int J Mol Sci 2019;20:4851.
[14] Voss A, Reinhart M, Sankarappa S, et al. The metabolism [31] McGlory C, Calder PC, Nunes EA. The influence of
of 7,10,13,16,19-docosapentaenoic acid to 4,7,10,13,16,19- omega-3 fatty acids on skeletal muscle protein turn-
docosahexaenoic acid in rat liver is independent of a 4- over in health, disuse, and disease. Front Nutr
desaturase. J Biol Chem 1991;266:19995 20000. 2019;6:144.
[15] Mozaffarian D, Wu JHY. Omega-3 fatty acids and car- [32] Avallone R, Vitale G, Bertolotti M. Omega-3 fatty acids
diovascular disease. J Am Coll Cardiol 2011;58:2047 67. and neurodegenerative diseases: new evidence in clin-
[16] Agrawal NK, Kant S. Targeting inflammation in dia- ical trials. Int J Mol Sci 2019;20:4526.
betes: newer therapeutic options. World J Diabetes [33] Abdolmaleki F, Kovanen PT, Mardani R, et al. Resolvins:
2014;5:697 710. emerging players in autoimmune and inflammatory dis-
[17] Sandireddy R, Yerra VG, Areti A, et al. eases. Clin Rev Allergy Immunol 2019;58:82 91.
Neuroinflammation and oxidative stress in diabetic [34] Baum SJ, Scholz KP. Rounding the corner on residual
neuropathy: futuristic strategies based on these tar- risk: implications of REDUCE-IT for omega-3 polyun-
gets. Int J Endocrinol 2014;674987. saturated fatty acids treatment in secondary preven-
[18] Zhou J, Zhou S. Inflammation: therapeutic targets for tion of atherosclerotic cardiovascular disease. Clin
diabetic neuropathy. Mol Neurobiol 2014;49:536 46. Cardiol 2019;42:829 38.
[19] Vincent AM, Calabek B, Roberts L, et al. Biology of [35] Tummala R, Ghosh RK, Jain V, et al. Fish oil and car-
diabetic neuropathy. Handb Clin Neurol diometabolic diseases: recent updates and controver-
2013;115:591 606. sies. Am J Med 2019;132:1153 9.
[20] Sytze Van Dam P, Cotter MA, Bravenboer B, et al. [36] Wang DD, Hu FB. Dietary fat and risk of cardiovascu-
Pathogenesis of diabetic neuropathy: focus on neuro- lar disease: recent controversies and advances. Annu
vascular mechanisms. Eur J Pharmacol 2013;719:180 6. Rev Nutr 2017;37:423 46.
[21] Bittner V. Implications for reduce it in clinical practice. [37] Harris WS, Zotor FB. N-3 fatty acids and risk for fatal
Prog Cardiovasc Dis 2019;62:395 400. coronary disease. Proc Nutr Soc 2019;. Available from:
[22] Kris-Etherton PM, Richter CK, Bowen KJ, et al. Recent https://doi.org/10.1017/S0029665118002902.
clinical trials shed new light on the cardiovascular [38] Superko HR, Superko SM, Nasir K, et al. Omega-3
benefits of omega-3 fatty acids. Methodist Debakey fatty acid blood levels. Circulation 2013;128:2154 61.
Cardiovasc J 2019;15:171 8. [39] Superko HR, Superko AR, Lundberg GP, et al.
[23] Shaikh SR, Kinnun JJ, Leng X, et al. How polyunsatu- Omega-3 fatty acid blood levels clinical significance
rated fatty acids modify molecular organization in update. Curr Cardiovasc Risk Rep 2014;8:407 14.

Diabetic Neuropathy
346 20. Translating a treatment for diabetic peripheral neuropathy from rodents to humans: can a case be made for fish oil and salsalate?

[40] Bowman L, Mafham M, Wallendszus Kthe ASCEND type 1 diabetic mouse on neuropathic endpoints.
study collaborative group. Effects of n-3 fatty acids J Nutr Metab 2016;2016:5905891.
supplements in diabetes mellitus. N Engl J Med [55] Cameron NE, Eaton SE, Cotter MA, et al. Vascular fac-
2018;379:1540 50. tors and metabolic interactions in the pathogenesis of
[41] Bhatt DL, Steg PG, Miller M, et al. Effects of icosapent diabetic neuropathy. Diabetologia 2001;44:1973 88.
ethyl on total ischemic events: from REDUCE-IT. J Am [56] Yorek MA. Vascular impairment of epineurial arterioles
Coll Cardiol 2019;73:2791 802. of the sciatic nerve: implications for diabetic peripheral
[42] Skulas-Ray AC, Wilson PWF, Harris WS, et al. neuropathy. Rev Diabet Stud 2015;12:13 28.
Omega-3 fatty acids for the management of hypertri- [57] Kles KA, Vinik AI. Pathophysiology and treatment of
glyceridemia: a science advisory from the American diabetic peripheral neuropathy: the case for diabetic
Heart Association. Circulation 2019;140:E673 91. neurovascular function as an essential component.
[43] Laidlaw M, Holub BJ. Effects of supplementation with Curr Diabet Rev 2006;2:131 45.
fish oil-derived n-3 fatty acids and alpha-linolenic acid [58] Coppey L, Holmes A, Davidson E, et al. Partial
on circulating plasma lipids and fatty acid profiles in replacement with menhaden oil improves peripheral
women. Am J Clin Nutr 2003;77:37 42. neuropathy in high-fat-fed low-dose streptozotocin
[44] Serhan CN, Chiang N, Dalli J. New pro-resolving n-3 type 2 diabetic rat. J Nutr Metab 2012;2012:950517.
mediators bridge resolution of infectious inflammation [59] Coppey L, Davidson E, Obrosov A, et al. Enriching
to tissue regeneration. Mol Asp Med 2018;64:1 17. the diet with menhaden oil improves peripheral neu-
[45] Chen J, Shetty S, Zhang P, et al. Aspirin-triggered ropathy in streptozotocin-induced type 1 diabetic rats.
resolvin D1 down-regulates inflammatory responses J Neurophysiol 2015;113:701 8.
and protects against endotoxin-induced acute kidney [60] Coppey L, Davidson E, Shevalye H, et al. Progressive
injury. Tox Appl Pharma 2014;277:118 23. loss of corneal nerve fibers and sensitivity in rats
[46] Goldfine AB, Fonseca V, Jablonski KA, et al. Salicylate modeling obesity and type 2 diabetes is reversible with
(salsalate) in patients with type 2 diabetes: a random- omega-3 fatty acid intervention: supporting cornea
ized trial. Ann Intern Med 2013;159:1 12. analyses as a marker for peripheral neuropathy and
[47] Goldfine AB, Fonseca V, Jablonski KA, et al. The treatment. Diabet Metab Syndr Obes 2020;13:1367 84.
effects of salsalate on glycemic control in patients with [61] Tikhonenko M, Lydic T, Opreanu M, et al. N-3
type 2 diabetes: a randomized trial. Ann Intern Med Polyunsaturated fatty acids prevent diabetic retinopa-
2010;152:346 57. thy by inhibition of retinal vascular damage and
[48] Hawley S, Fullerton M, Ross F, et al. The ancient drug enhanced endothelial progenitor cell reparative func-
salicylate directly activates AMP-activated protein tion. PLoS One 2013;8:e55177.
kinase. Science 2012;336:918 22. [62] Matsumoto T, Nakayama N, Ishida K, et al.
[49] Van Dam A, Nahon K, Kooijman S, et al. Salsalate Eicosapentaenoic acid improves imbalance between
activates brown adipose tissue in mice. Diabetes vasodilator and vasoconstrictor actions of endothelium-
2015;64:1544 54. derived factors in mesenteric arteries from rats at
[50] Wang X, DuBois D, Cao Y, et al. Diabetes disease pro- chronic stage of type 2 diabetes. J Pharmacol Exp Ther
gression in Goto-Kakizaki rats: effect of salsalate treat- 2009;329:324 34.
ment. Diabetes Metab Syndr Obes 2014;7:381 9. [63] Farooq MA, Gaertner S, Amoura L, et al. Intake of
[51] Chai W, Liu J, Fowler D, et al. Salsalate attenuates free omega-3 formulation EPA:DHA 6:1 by old rats for 2
fatty acid-induced microvascular and metabolic insulin weeks improved endothelium-dependent relaxations
resistance in humans. Diabetes Care 2011;34:1634 8. and normalized the expression level of ACE/AT1R/
[52] Ariel D, Kim S, Liu A, et al. Salsalate-induced changes NADPH oxidase and the formation of ROS in the
in lipid, lipoprotein, and apoprotein concentrations in mesenteric artery. Biochem Pharmacol 2020;173
overweight or obese, insulin-resistant, nondiabetic 113749.
individuals. J Clin Lipidol 2015;9:658 63. [64] Davidson EP, Coppey LJ, Holmes A, et al. Effect of
[53] Davidson EP, Coppey LJ, Shevalye H, et al. Effect of inhibition of angiotensin converting enzyme and/or
dietary content of menhaden oil with or without salsa- neutral endopeptidase on vascular and neural compli-
late on neuropathic endpoints in high-fat-fed/low- cations in high fat fed/low dose streptozotocin-
dose streptozotocin-treated Sprague Dawley rats. diabetic rats. Eur J Pharmacol 2012;677:180 7.
J Diabetes Res 2018;2018:2967127. [65] Coppey LJ, Davidson EP, Rinehart TW, et al. ACE
[54] Yorek M, Coppey L, Shevalye H, et al. Effect of treat- inhibitor or angiotensin II receptor antagonist attenu-
ment with salsalate, menhaden oil, combination of sal- ates diabetic neuropathy in streptozotocin-induced
salate and menhaden oil, or resolvin D1 of C57Bl/6J diabetic rats. Diabetes 2006;55:341 8348.

Diabetic Neuropathy
 References 347
[66] Oltman CL, Davidson EP, Coppey LJ, et al. Attenuation [80] Jelinek D, Castillo J, Arora S, et al. A high-fat diet supple-
of vascular/neural dysfunction in Zucker rats treated mented with fish oil improves metabolic features associ-
with enalapril or rosuvastatin. Obesity 2008;16:82 9. ated with type 2 diabetes. Nutrition 2013;29:1159 65.
[67] Sakai C, Ishida M, Ohba H, et al. Fish oil omega-3 [81] Gonzalez-Periz A, Horrillo R, Ferre N, et al. Obesity-
polyunsaturated fatty acids attenuate oxidative stress- induced insulin resistance and hepatic steatosis are
induced DNA damage in vascular endothelial cells. alleviated by ω-3 fatty acids: a role for resolvins and
PLoS One 2017;12:e0187934. protectins. FASEB J 2009;23:1946 57.
[68] Mustad V, Demichele S, Huang Y, et al. Differential [82] Li J, Li F, Wei D, et al. Endogenous ω-3 PUFAs produc-
effects of n-3 polyunsaturated fatty acids on metabolic tion confers resistance to obesity, dyslipidemia, and
control and vascular reactivity in the type 2 diabetic diabetes in mice. Mol Endocrinol 2014;28:1316 28.
ob/ob mouse. Metabolism 2006;55:1365 74. [83] Dangardt F, Chen Y, Gronowitz E, et al. High physiolog-
[69] Takenouchi Y, Ohtake K, Nobe K, et al. Eicosapentaenoic ical omega-3 fatty acid supplementation affects muscle
acid ethyl ester improves endothelial dysfunction in type fatty acid composition and glucose and insulin homeo-
2 diabetic mice. Lipids Health Dis 2018;17:118. stasis in obese adolescents. J Nutr Met 2012;2012:395757.
[70] Merino J, Sala-Vila A, Kones R, et al. Increasing long- [84] Sarbolouki S, Javanbakht M, Derakhshanian H, et al.
chain n-3 PUFA consumption improves small periph- Eicosapentaenoic acid improves insulin sensitivity
eral artery function in patients at intermediate-high and blood sugar in overweight type 2 diabetic mellitus
cardiovascular risk. J Nutr Biochem 2014;25:642 6. patients: a double-blind randomised clinical trial.
[71] Ito R, Satoh-Asahara N, Yamakage H, et al. An Singap Med 2013;54:387 90.
increase in the EPA/AA ratio is associated [85] Kim YS, Xun P, He K. Fish consumption, long-chain
with improved arterial stiffness in obese patients with omega-3 polyunsaturated fatty acid intake and risk of
dyslipidemia. J Atheroscler Thromb 2014;21:248 60. metabolic syndrome: a meta-analysis. Nutrients
[72] Titos E, Claria J. Omega-3-derived mediators counter- 2015;7:2085 100.
act obesity-induced adipose tissue inflammation. [86] Steffan B, Steffan L, Zhou X, et al. n-3 Fatty acids
Prostaglandins Other Lipid Mediat 2013;107:77 84. attenuate the risk of diabetes associated with elevated
[73] Fan C, Zirpoli H, Qi K. n-3 Fatty acids modulate adi- serum nonesterified fatty acids: the multi-ethnic study
pose tissue inflammation and oxidative stress. Curr of atherosclerosis. Diabet Care 2015;38:575 80.
Opin Clin Nutr Metab Care 2013;16:124 32. [87] Rylander C, Sandanger T, Engeset D, et al.
[74] Gray B, Steyn F, Davies P, Vitetta L. Omega-3 fatty Consumption of lean fish reduces the risk of type 2 dia-
acids: a review of the effects on adiponectin and leptin betes mellitus: a prospective population based cohort
and potential implications for obesity management. study of Norwegian women. PLoS One 2014;9:E89845.
Eur J Clin Nutr 2013;67 1234-124. [88] Kaushik M, Mozaffarian D, Spiegelman D, et al. Long-
[75] Figueras M, Olivan M, Busquets S, Lopez-Soriano F, chain omega-3 fatty acids, fish intake, and the risk of
Argiles J. Effects of eicosapentaenoic acid (EPA) treat- type 2 diabetes mellitus. Am J Clin Nutr 2009;90:613 20.
ment on insulin sensitivity in an animal model of dia- [89] Shevalye H, Yorek M, Coppey L, et al. Effect of enrich-
betes: improvement of the inflammatory status. ing the diet with menhaden oil or daily treatment
Obesity, 19. 2011. p. 362 9. with resolvin D1 on neuropathy in a mouse model of
[76] Bargut T, Mandarim-de-Lacerda C, Aguila M. A high- type 2 diabetes. J Neurophysiol 2015;114:199 208.
fish-oil diet prevents adiposity and modulates white [90] Holmes A, Coppey L, Davidson E, et al. Rat models of
adipose tissue inflammation pathways in mice. J Nutr diet-induced obesity and high fat/low dose streptozo-
Biochem 2015;26:960 9. tocin type 2 diabetes: effect of reversal of high fat diet
[77] Hellman J, Tang Y, Kosuri M, Bhatnagar A, Spite M. compared to treatment with enalapril or menhaden oil
Resolvin D1 decreases adipose tissue macrophage on glucose utilization and neuropathic endpoints. J
accumulation and improves insulin sensitivity in Diabet Res 2015;2015:307285.
obese-diabetic mice. FASEB J 2001;25:2399 407. [91] Davidson E, Holmes A, Coppey L, Yorek M. Effect of
[78] Todorcevic M, Hodson L. The effect of marine derived combination therapy consisting of enalapril, α-lipoic
n-3 fatty acids on adipose tissue metabolism and func- acid, and menhaden oil on diabetic neuropathy in a
tion. J Clin Med 2016;5(10):3390. high fat/low dose streptozotocin treated rat. Eur J
[79] Siriwardhana N, Kalupahana N, Fletcher S, et al. n-3 Pharmacol 2015;765:258 67.
and n-6 polyunsaturated fatty acids differentially reg- [92] Alam U, Jeziorska M, Petropoulos IN, et al. Diagnostic
ulate adipose angiotensinogen and other inflamma- utility of corneal confocal microscopy and intra-
tory adipokines in part via NF-κB-dependent epidermal nerve fibre density in diabetic neuropathy.
mechanisms. J Nutr Biochem 2012;23:1661 7. PLoS One 2017;12:e0180175.

Diabetic Neuropathy
348 20. Translating a treatment for diabetic peripheral neuropathy from rodents to humans: can a case be made for fish oil and salsalate?

[93] Lewis EJH, Perkins BA, Lovblom LE, et al. Effect of velocity and neuroanatomical changes in rats. J Nutr
omega-3 supplementation on neuropathy in type 1 1999;129:207 13.
diabetes. Neurology 2017;88:2294 301. [100] Julu P. Essential fatty acids prevent slowed nerve
[94] O’Mahoney LL, Dunseath G, Churm R, et al. Omega- conduction in streptozotocin diabetic rats. J Diabet
3 polyunsaturated fatty acid supplementation vs Complicat 1998;2:185 8.
placebo on vascular health, glycaemic control, and [101] Heng L, Qi R, Yang R, et al. Docosahexaenoic acid
metabolic parameters in people with type 1 diabetes: inhibits mechanical allodynia and thermal hyperalge-
a randomized controlled preliminary trial. sia in diabetic rats by decreasing the excitability of
Cardiovasc Diabetol 2020;19:127. DRG neurons. Exp Neurol 2015;271:291 300.
[95] Okuda Y, Mizutani M, Ogawa M, et al. Long-term [102] Li M, Wang Y, Cao R, et al. Dietary fish oil inhibits
effects of eicosapentanoic acid on diabetic periph- mechanical allodynia and thermal hyperalgesia in dia-
eral neuropathy and serum lipids in patients with betic rats by blocking nuclear factor-κB-mediated inflam-
type II diabetes mellitus. J. Diabet Complicat matory pathways. J. Nutr Biochem 2015;26:1147 55.
1996;10:280 7. [103] Wagner K, Lee K, Yang J, et al. Epoxy fatty acids
[96] Obrosov A, Coppey LJ, Shevalye H, et al. Effect of mediate analgesia in murine diabetic neuropathy.
fish oil vs. resolvin D1, E1, methyl esters of resolvins Eur J Pain 2017;21:456 65.
D1 or D2 on diabetic peripheral neuropathy. J Neurol [104] Landa-Juarez AY, Perez-Severiano F, Castaneda-
Neurophysiol 2017;8:453. Hernandez G, et al. The antihyperalgesic effect of
[97] Yorek MA. Is fish oil a potential treatment for dia- docosahexaenoic acid in streptozotocin-induced neu-
betic peripheral neuropathy? Curr Diabet Rev ropathic pain in the rat involves the opioidergic sys-
2018;14:339 49. tem. Euro J Pharmacol 2019;845:32 9.
[98] Yorek MA. The potential role of fatty acids in treating [105] Mas E, Croft K, Zahra P, et al. Resolvins D1, D2 and
diabetic neuropathy. Curr Diab Rep 2018;18:86. other mediators of self-limited resolution of inflam-
[99] Gerbi A, Maixent J, Ansaldi J, et al. Fish oil supple- mation in human blood following n-3 fatty acid sup-
mentation prevents diabetes-induec nerve conduction plementation. Clin Chem 2012;58:1476 84.

Diabetic Neuropathy
 C H A P T E R

21
Bedside clinical features and translational
snapshots of diabetic polyneuropathy
Brendan N. Putko and Douglas W. Zochodne
Division of Neurology, Department of Medicine and the Neuroscience and Mental Health Institute,
University of Alberta, Edmonton, Canada

Introduction early stages of many diseases, the classical syn-

drome may not be readily apparent. The informa-
Our understanding of diabetes mellitus (DM) tion contained in this chapter is intended to equip
had advanced substantially by the end of the 19th the reader with an understanding of the clinical
century. Observations by Minkowski and von features of DPN to facilitate recognition of the
Mering around that time made a key conceptual component signs and symptoms even if a specific
linkage of DM to the pancreas [1]. In 1864, de syndromic diagnosis is not initially apparent. We
Calvi published the first recognized version of the will describe clinical features as they pertain to
hypothesis that neurological dysfunction could components of the peripheral nervous system
result from DM, and Charcot provided an and we will frame our discussion of clinical fea-
expanded clinical description in 1890 [2]. tures with contextual data to emphasize the
Following the seminal reports of diabetic poly- importance of a refined clinical approach.
neuropathy (DPN), Williamson published a suc- Neurological dysfunction, occurring most com-
cinct description of testing vibration sense at monly as neuropathy, is the most common of the
subcutaneous bony prominences and observed complications of diabetes [5]. Many sources refer
that it was absent in some patients with diabetes to these complications in aggregate as “micro- or
[3]. Although descriptions of the clinical manifes- macrovascular” complications. However, in the
tations of DPN were published more than a cen- case of DPN, the evidence for a primary micro-
tury ago, current estimates suggest only one-third vascular etiology is not established, suggesting its
of physicians recognize the manifestations of dia- more apt description might simply be as a com-
betic neuropathy in patients presenting with a plication, with other published reviews examining
compatible clinical syndrome [4]. The approach to etiology in more detail than here [6,7]. Diabetic
neurological disease relies on syndromic diagno- neuropathies encompass a group of disorders of
sis in parallel with a first-principles approach the peripheral nervous system in the setting of
founded in neuroanatomic localization. In the diabetes. They occur on a temporal continuum

Diabetic Neuropathy
DOI: https://doi.org/10.1016/B978-0-12-820669-0.00015-3 349 © 2022 Elsevier Inc. All rights reserved.
350 21. Bedside clinical features and translational snapshots of diabetic polyneuropathy

from acute to chronic and may be focal to diffuse. is the most common subtype of neuropathy and
Of the diabetic neuropathies, approximately 75% DM is the most common cause of this subtype of
are chronic distal symmetric polyneuropathy, polyneuropathy [11]. Individuals with an as-yet
which is commonly called “diabetic neuropathy” undiagnosed polyneuropathy are one of the
[5]. Not all diabetic neuropathies are the same, main populations of interest, in addition to indi-
however, as the relative involvement of small and viduals with known diabetes, when considering
large nerve fibers can differ substantially between DPN.
two otherwise similar patients. Small nerve fibers While persons with overt diabetes are the
have unmyelinated and thinly myelinated axons most obvious population of interest covered in
that transmit pain and temperature sensation this chapter, it is important to recognize that
as well as autonomic signals. Large nerve fibers, neuropathy caused by type 1 and type 2 DM
on the other hand, have heavily myelinated axons are distinct entities that converge on the syn-
that transmit proprioception and vibration as well drome of DPN. In type 1 diabetes, there is a
as motor signals. Small nerve fibers conduct rela- reliable dose response relationship between
tively slowly (0.5 2.0 m/s), while large nerve glycemic control and neuropathy progression
fibers conduct quickly (40 80 m/s) and are pref- that is not as robust in type 2 diabetes [12]. For
erentially studied by typical electrophysiological more on the treatment of diabetic neuropathy
tests. More recent recognition has been given to please see other chapters in this textbook of
small fiber predominant diabetic neuropathy, Diabetic Neuropathy. Differences in neuropathy
which may represent early DPN that will later attributable to type 1 and type 2 diabetes are
evolve to have more evident large fiber features, expected in view of the differences in patho-
although small fiber symptoms and signs may genesis of the two entities. While type 1 diabe-
always predominate. For the purposes of this tes is due to loss of insulin production
review, we will refer to the group as DPN. capacity, type 2 diabetes results from a com-
Simply diagnosing a neuropathic syndrome plex interplay of insulin resistance, insulin
in the context of diabetes is insufficient to reli- secretion, and disordered hepatic glucose pro-
ably call a disorder diabetic neuropathy. To duction. Interestingly a growing body of evi-
make a diagnosis of diabetic neuropathy, other dence has linked substrate conditions that
causes must be excluded. This is an important cascade into type 2 diabetes, namely prediabe-
exercise in the setting of a common condition. tes and metabolic syndrome, with neuropathy
Anchoring bias and premature closure, pro- [12]. Therefore individuals with risk factors for
cesses by which clinicians assume the most type 2 diabetes make up the third population
common explanation in the face of contradic- of interest in addition to individuals with poly-
tory evidence, are potential pitfalls in evaluat- neuropathy of unknown etiology and indivi-
ing the patient with DM presenting with duals with known diabetes.
symptoms localizable to the peripheral nerves.
The reported prevalence of secondary or alter-
native causes of peripheral neuropathy in Symptoms
patients with diabetes is 10% 50% [8 10],
which should serve to inform a search for The clinical encounter should begin with the
treatable causes—this begins with the clinical patient being allowed to relay their experience
evaluation. On the other hand, considering dia- with symptoms in their own narrative. The
betes is paramount in the evaluation of patients healthcare provider may clarify details, although
presenting with neuropathic symptoms and it is advisable to avoid asking directed questions
signs. Indeed, distal symmetric polyneuropathy early in the interview. Care should also be taken

Diabetic Neuropathy
 Symptoms 351
to avoid paraphrasing the history into terms extremity pain, such as osteoarthritis or plantar
more familiar in the medical lexicon. Doing so fasciitis. Neuropathic pain was defined by Treede
removes the patient voice from the recorded his- et al. as “pain arising as a direct consequence of a
tory, which is essential in tracking symptoms lesion or disease affecting the somatosensory sys-
over time, otherwise error due to misinterpreta- tem” [15]. There are four criteria for neuropathic
tion is possible [13]. Indeed, a common cause of pain: first, it should follow a known neuroana-
cognitive error in medical practice is the failure tomic distribution, such as the innervation terri-
to properly listen to the patient’s story [14]. tory of peripheral nerves. Second, it should have
Throughout the history, the clinician must a congruent spatial and temporal relationship
broadly organize the patient narrative into a with the proposed lesion, that is, the lesion should
neuroanatomical localization. A key point about be able to affect that area of the somatosensory
DPN is that it targets sensory more than motor system. Third, there should be supportive features
fibers. Significant, symptomatic weakness, espe- on clinical examination (see Examination section
cially if progressive, is a red flag for an alterna- below). Fourth, there should be compatible evi-
tive cause of neuropathy. dence on established diagnostic tests. When all
The sensory symptoms relayed to the clinician criteria are satisfied, then neuropathic pain may
may be positive or negative. Positive symptoms be considered definite, while if the first, second,
reflect aberrant impulses from damaged nerve and either the third or fourth are present, then
fibers or ganglia, while negative symptoms reflect neuropathic pain may be considered probable.
the absence of neural function. Another axis of The latter are minimum criteria to confirm neuro-
symptoms is spontaneous versus stimulus pathic pain, otherwise the diagnosis should be
induced. It is in describing positive symptoms reconsidered, and further history, examination,
that the patient’s own narrative of their experi- and diagnostic testing performed.
ence come to life. Spontaneous, abnormal sensa- Negative symptoms, on the other hand, may
tions are broadly called paresthesias. Descriptions engender descriptions such as the numb or
commonly used by patients include “tingling,” “dead” extremity. Numbness deserves specific
“prickling,” and “pins and needles.” Spontaneous mention, as clinicians often take it to mean
unpleasant or painful sensations are, on the other insensate, whereas patients may use numb to
hand, broadly called dysesthesias. Descriptions describe a nonspecific sensory or motor distur-
commonly used by patients include burning, bance—this is one instance in which clarifica-
walking on “hot sand,” tightness, aching, and jab- tion of what numbness means to the patient
bing. Stimulus-induced positive sensory phenom- using other descriptive terminology is helpful.
ena include allodynia and hyperpathia. Allodynia The patient may remark upon the contradiction
is the perception of nonpainful stimuli as painful. of an insensate limb that is simultaneously
One common example that patients will provide remarkably painful. On the other hand, loss of
is that they find the touch of bedsheets intolerable. sensation may be hard for the patient to detect.
This example is also representative of how sen- Therefore it is important to look for other his-
sory symptoms in neuropathy, including diabetic torical clues, such as the appearance of injuries
neuropathy, tend to be worse at night. In contrast to the feet for which an incident event cannot
to allodynia, hyperpathia, also called hyperalge- be recalled. For more discussion on the damage
sia, is enhanced pain sensation in response to a that follows neuropathy, see Complications
normally painful stimulus. See Examination sec- section below. Another clue is a change in gait
tion below for more on hyperpathia. Care should and balance. Particularly telling are imbalance in
be taken to elucidate a history consistent with darkness or when closing the eyes while washing
neuropathic pain, rather than other causes of the face or the hair, as these give important

Diabetic Neuropathy
352 21. Bedside clinical features and translational snapshots of diabetic polyneuropathy

clues to sensory dysfunction that is manifested commonly considered as negative. Subtle weak-
with visual fixation removed, but also alert the ness may manifest in a dragging or frankly
clinician to the potential for falls. Beyond the dropped foot, while clumsiness and incoordina-
sensory experience itself, patients may report tion of the hands may be seen in severe DPN or
the behavioral responses to positive and nega- DPN with concurrent carpal tunnel syndrome.
tive sensory symptoms. Many patients report Mild foot drop commonly contributes to falls for-
restless legs, and indeed polyneuropathy is a ward from catching the foot on obstacles. When
common element on the differential diagnosis eliciting subtle weakness, it may be helpful to ask
for restless legs syndrome. Adaptive strategies whether the patient has experienced any difficul-
used by patients include wearing soft-soled ties with common activities, such opening jars
shoes, open toed shoes out of season, and and climbing stairs. As stated above, however,
other departures from their normal footwear directed questioning should be used sparingly
(Fig. 21.1). and should occur after the patient has had the
Motor symptoms, likewise, may be considered opportunity to relay their narrative.
in the positive or negative dichotomy. Cramps The discussion of symptoms attributable to
are a frequently encountered symptom in diabetic the small fibers has so far focused on somatic
patients and tend to be more severe than in those small fibers. Dysfunction of autonomic small
without diabetes [16]. Fasciculations, a hallmark fibers is common and disabling in DM, although
of motor axon loss and often described by formal estimates of the prevalence of diabetic
patients as muscle jumps or twitches, are not typ- autonomic neuropathy vary widely depending
ical of DPN. Neither is tremor, and its presence on the method of diagnosis [17]. Patients may
would suggest an acquired or inherited demye- report feeling light-headed, “dizzy,” or “woozy”
linating neuropathy. Motor symptoms are more when experiencing orthostatic hypotension, which
occurs in approximately 30% of patients
with DM overall and is the most disabling symp-
tom of cardiovascular autonomic involvement
[18]. Other symptoms of cardiovascular involve-
ment include tachycardia and loss of heart rate
variability, which may manifest as exercise intol-
erance due to inability to mount a physiologic
tachycardic response [17]. DM patients may
experience a loss of hypoglycemic awareness.
While not a symptom but rather a sequela, car-
diovascular autonomic involvement increases
the risk of overall mortality and sudden death in
patients with DM. Early satiety, intolerance of
oral intake, nausea, vomiting, and belching are
symptoms of gastroparesis, which occurs in 50%
FIGURE 21.1 Clinical clues: Preferred footware of a of patients with DM. Constipation occurs in 60%
patient with longstanding diabetes mellitus and polyneuropa- of patients with DM, while other lower gastroin-
thy. Despite ambient outside temperatures as low as 220 C testinal symptoms, such as diarrhea and abdomi-
in Edmonton, Canada, this patient wore sandals year round nal discomfort are also common. Urogenital
to alleviate symptoms of burning discomfort in his feet.
Although insensate, he was careful to take precautions to
symptoms include urinary retention, overflow
avoid frostbite. Note the healing skin lesion on the dorsum of incontinence, and sexual dysfunction. Erectile
the left foot, probably related to the sandal straps. dysfunction is a common symptom that occurs

Diabetic Neuropathy
 Symptoms 353
early; however, patients may not disclose it in Regarding the temporal profile of symptoms,
view of its sensitive nature, and thus it should be there is no single natural history that encom-
specifically queried in male patients. Changes in passes all patients with diabetic neuropathy.
sweating patterns, when present, occur in a Early in the course, however, patients often
length-dependent pattern wherein there is loss of report predominant burning pain, implicating
distal sweating with compensatory excessive early involvement of small, unmyelinated sen-
proximal sweating. However, as the reader will sory nerve fibers. Pathological studies using
observe from the discussion above, many auto- intraepidermal nerve fiber density measurements
nomic symptoms do not follow such a clearly that indicate fiber loss, retraction or endbulbs are
defined length-dependent pattern, so patients in keeping with these clinical findings, and have
may not conceptually link them to their sensory shown loss of somatic small fibers in individuals
symptoms. It is therefore important to undertake with prediabetes and those with early diabetes
an autonomic review of systems as part of the (Fig. 21.2) [19]. Progression of symptoms should
history in the assessment of neuropathy. follow a length-dependent gradient in DPN.

FIGURE 21.2 Immunohistochemical

images of lower limb (ankle level) 3 mm
punch skin biopsies at lower (left column
Bar 5 15 µm) and higher (right column
Bar 5 15 µm) power. The axon images are
labeled with an antibody to PGP 9.5, a panax-
onal marker (black, white arrows). The control
images (A, B) shows plentiful dermal and epi-
dermal (dermal epidermal junction shown
by white arrowhead) axons, shown branching
under higher power. The images from a
patient with type 2 diabetes mellitus and
polyneuropathy are from the ankle level (C,
D), and more proximally, at the hip (E, F).
Note the severe loss of both dermal and epi-
dermal axon profiles. (Images provided with
thanks from Lucie Guernsey, Katie Frizzi and
Dr. Nigel Calcutt at UCSD).

Diabetic Neuropathy
354 21. Bedside clinical features and translational snapshots of diabetic polyneuropathy

Importantly some symptoms, such as pain, may cerebral small vessel disease, or nutritional defi-
attenuate as epidermal and dermal axons pro- ciency, and thus need to be taken in context of the
gressively disappear [20]. patient’s known past medical history. Similarly,
cranial nerve deficits are not typical of DPN,
although pupillomotor dysfunction in response to
Examination low or high light stimuli may reflect underlying
autonomic dysfunction. Cranial mononeuropa-
While sensory symptoms predominate and thies occur in the setting of diabetes, particularly
may be the sole reason for seeking assessment, affecting cranial nerves III, IV, and VI. Again, the
this section is organized to match the flow of a history will likely suggest cranial nerve deficits
typical neurological examination that is preferred that predate or punctuate the course of DPN.
by the authors. As such, many of the findings Examination of the motor systems begins with
mentioned are not typical of DPN per se, but are a careful inspection. Inspection of the integument
commonly encountered, and therefore deserve is detailed below, although many individuals
specific mention. Following a careful history, a may prefer to combine the inspection portion of
detailed neurological examination is sufficient to the motor examination with an integumentary
confirm a diagnosis of polyneuropathy in most exam. Atrophy of the thenar or hypothenar emi-
patients. The techniques used have not changed nences may be seen in severe DPN, or carpal
substantially over many decades and remain tunnel syndrome and ulnar neuropathy at the
indispensable. elbow, respectively—entrapment neuropathies
Routine assessment of vital signs is recom- are more common in individuals with diabetes
mended for all clinical encounters. While than the general population. On the other hand,
abnormalities in vital signs are not specific for atrophy of the extensor digitorum brevis is com-
DPN, they are important to document and may monly seen in DPN. As the disorder progresses,
provide an indication of needed primary and atrophy of the anterior compartment of the lower
secondary prevention, such as treating hyper- leg may be seen or felt, giving a sharply con-
tension. Assessment of orthostatic blood pres- toured edge to the lateral margin of the tibia.
sure and heart rate requires no specialized Marked proximal muscle atrophy is not expected
equipment and does not add significant time in DPN. Radiculoplexus neuropathy, typically
to the encounter. It is recommended that blood involving the lumbosacral segments, is another
pressure be measured after at least one minute neuropathic syndrome observed in individuals
of standing to allow the normal transient fluc- with diabetes that can result in marked atrophy
tuation of blood pressure that occurs in healthy of muscles of the affected limbs. We are assum-
individuals to pass [21]. Indeed, pathological ing this is the case that diabetic radiculoplexus
orthostatic hypotension requires a sustained neuropathy is described elsewhere in this text.
drop in blood pressure on standing. The cut- As mentioned above, fasciculations and tremor
off values are a drop in the systolic blood are not typical findings.
pressure of 20 mmHg or the diastolic blood Manual muscle testing should demonstrate
pressure of 10 mmHg. The blood pressure largely preserved strength, even in advanced
should otherwise be measured in the standard DPN. Proper technique is essential to confirm
seated position, as well as in the supine posi- preserved strength. The region proximal to the
tion after at least 2 minutes of supine rest. joint movement being evaluated should be sta-
Alterations in sensorium and cognitive pro- bilized by the examiner to isolate the move-
cesses are not expected in any form of diabetic ment of interest. For the upper extremities, we
neuropathy, but may reflect sequelae of strokes, suggest testing with the same muscles or

Diabetic Neuropathy
 Examination 355
similarly sized muscles, such as applying the 25% of patients in the Rochester Diabetic
dorsal surface of extended fingers to the Neuropathy Study cohort [22]. An appropri-
patient’s extended fingers to test the power of ately weighted reflex hammer is important to
digit extensors. In the lower extremities, the ensure adequate force to elicit a reflex, espe-
longest lever arm possible should be generated cially in the case of pathologically depressed
by applying force to the most distal portion of reflexes. Plantar responses may be equivocal in
the appendage in question. Weakness proximal patients with advanced neuropathy and severe
to the elbow is not expected. Weakness in the sensory loss, but are otherwise expected to be
hands may observed in the presence of a con- flexor unless a concurrent disorder of the cen-
comitant carpal tunnel syndrome or ulnar neu- tral nervous system is present.
ropathy, as mentioned above. Weakness of the Although the authors have included sensory
deep finger flexors suggesting involvement of examination after motor examination, it may
anterior interosseus nerves should prompt an be worthwhile to consider testing sensation
expanded search for causes of mononeuropa- earlier or following a short break to avoid
thy, such as vasculitis. Proximal weakness of patient fatigue if an extensive mental status,
the legs is rare unless there has been lumbosa- cranial nerve, or motor system examination is
cral radiculoplexus neuropathy. Distal lower planned [23]. In general, the sensory examina-
extremity weakness is expected to progress tion should proceed with brief instructions and
according to a length-dependent gradient that avoid suggesting abnormal responses if possi-
involves the anterior compartment to a greater ble. Sensory examination should interrogate
degree than the posterior compartment. the large myelinated and small unmyelinated
Weakness of dorsiflexion of the great toe may sensory nerve fibers. A 128-Hz tuning fork
be identified first, followed by the other toes, with weights on the prongs is the standard tool
and then the ankle, thereby producing foot for assessing vibratory sense, although a 256-
drop. The relative strength of an adult patient’s Hz fork has been advocated by some. The
ankle dorsiflexors compared to an examiner’s prongs may be drawn together or the end
upper extremity muscles means that it is advis- forcefully tapped against the palm to generate
able to evaluate heel-walking to demonstrate vibration. It is helpful first to establish that the
subtle weakness of the ankle dorsiflexors if none patient perceives vibration by applying the end
is overt on manual muscle testing. The MRC of the vibrating fork to a proximal bony promi-
(Medical Research Council) grading scale, long nence. Complete loss of vibratory sense at the
established and convenient, remains the simplest great toes or more proximal bony prominences
and most widely utilized approach to grading is clearly abnormal. However, only 8% of
muscle power. MRC grades are recognized to be patients in the Rochester Diabetic Neuropathy
nonlinear. Study cohort had loss of vibratory sense at the
In the case of prominent small fiber great toes [22]. A gradient of attenuation from
involvement, deep tendon reflexes may be pre- proximal to distal may be demonstrated when
served. With large fiber involvement, reflexes the patient still has some residual vibration
attenuate and are subsequently lost in a length- sensation at the great toes, keeping in mind
dependent manner. Tendon reflex abnormali- that bone vibratory transmission can be exten-
ties are attributable to dysfunction of the sive, making discrete changes problematic to
afferent (sensory) limb of the reflex arc. Ankle measure. While demonstrating a gradient of
jerk abnormalities are expected first, followed vibration sense can illustrate a pattern of length-
by knee jerks, and subsequently upper extrem- dependency, quantitation of vibratory loss is nor-
ity reflexes. Loss of ankle jerks was observed in mally not possible. The Rydel Seiffer tuning

Diabetic Neuropathy
356 21. Bedside clinical features and translational snapshots of diabetic polyneuropathy

and laboratory measures of peripheral nerve dys-

function in DM [24 26].
Evaluation of proprioception is the second
modality involved in assessment of large fiber
function. In general, proprioceptive deficits are
less evident than vibratory deficits, which may
reflect the comparative sensitivity of bedside
tests and the actual level of the sensory organ
activated. The standard method for evaluating
proprioception involves asking the patient to
note with their eyes closed the direction of
movement of an appendage attached to a joint.
We suggest stabilizing the joint proximal to the
area of interest and manipulating the append-
age with pressure applied from the sides to
remove pressure sensory information from the
patient. As with vibratory testing, the distal
phalanx of the great toe can be tested, followed
by more proximal joints if joint position sense
is abnormal. Deficits in proprioception at the
toes are a later finding in DPN and typically
accompany gait imbalance. They should raise
the possibility of another etiology of poly-
FIGURE 21.3 Image of a Rydel Seiffer semiquantita- neuropathy. The Romberg sign, a loss of bal-
tive tuning fork. The tool vibrates at 128 Hz without the ance elicited by having the patient stand
dampers (shown by arrow) and at 64 Hz with the dampers. straight and unsupported with their feet
A scale based on a visual image of vibrating triangles pro- apposed side-by-side then closing their eyes is
vides a numerical readout of the intensity at which the an expected concurrent finding when proprio-
patient notes cessation of vibratory sensation (insert shows
readout during vibration). ceptive loss is evident at the bedside. The
assertion that the Romberg sign is specific for
proprioceptive dysfunction has recently been
fork is a semiquantitative instrument developed questioned suggesting that vestibular or cere-
by Rydel and Seiffer in 1903 that addresses this bellar pathways may be associated with abnor-
issue (Fig. 21.3) [24]. The tuning fork vibrates at mal testing, although this differs from classical
128 Hz without dampers and 64 Hz with dam- teaching [23].
pers. Two inverted triangles are present on the Light touch is classically considered a large-
dampers, and each creates two illusions as the fiber mediated sensation, although delivering
tuning fork vibrates. As the amplitude of vibration identical light touch near threshold may be diffi-
attenuates, the point of intersection of the two illu- cult. Nonetheless, light touch testing is included
sions nears the top of the fork. An arbitrary scale as part of some instruments described in Clinical
of 0 8 is used to quantitate the point at which Scales below and thus deserves specific mention.
vibration is lost, with 8 denoting the longest time A piece of cotton wool is the best tool for asses-
to loss and thus highest sensitivity of vibration. sing light touch, as brushes are not as easily dis-
Studies showed that abnormalities with the infected. Light touch can be assessed on the skin
Rydel Seiffer tuning fork correlated with clinical of the dorsum of the great toe just proximal to

Diabetic Neuropathy
 Examination 357
the nail as part of a comprehensive clinical highly conducting wall that is used to hold warm
assessment of neuropathy. Marked abnormalities and cool water is likely to offer the most consistent
of proprioception, which can be evident in pseu- temperature as it is moved between limbs. Further
doathetosis and sensory ataxia would suggest an complicating matters, cool distal extremities may
alternative cause of sensory neuropathy or neu- confound comparisons with proximal temperature
ronopathy. Otherwise, examination of coordina- sensation.
tion and cerebellar function are not expected to The Semmes Weinstein 5.07/10 g filament
be abnormal in diabetic neuropathy. provides 10 g of pressure to the skin when
The somatic small fibers are assessed by applied perpendicular until the filament buckles.
evaluating pain and temperature sensation. The proper technique involves applying a consis-
Allodynia may be evident throughout the tent force until buckling occurs. The monofila-
examination, particularly when testing light ment examination is commonly included in
touch, and should be noted. The standard eval- toolkits for screening for DPN. Studies have
uation of pain involves testing pinprick sensa- examined its use in screening by applying the fil-
tion. A clean, unused safety pin, used only ament to dorsal and plantar areas of the feet.
once on a patient then discarded in a desig- Four applications to the distal dorsum of each
nated biohazard sharp waste container, is a great toe in a nonrhythmic manner is one possi-
standard tool. Safety pins do not typically pen- ble method [27]. Another possible method is
etrate beyond most superficial layers of skin three applications on the plantar surface of the
and are not expected to draw blood. The foot: over the great toe, first metatarsal, and fifth
authors recommend unclasping the pin to 90 metatarsal [28]. From a screening perspective,
degrees to provide the most consistent stimu- dorsal surface has fewer false positives, while
lus with each application. Analgesia may be the plantar surface has a higher positive predic-
demonstrated by asking the patient to discrimi- tive value [29]. Clinicians may therefore choose
nate between sharp and dull ends of the pin. which surface to test depending on the goal of
The territory of sensory loss or abnormality is their examination. A recent meta-analysis
important to document by applying the sharp reported low sensitivity of 53%, leading the
end of the pin to sequentially more proximal authors to conclude that the monofilament can-
patches of skin in increments of a few centi- not be recommended as a screening test in rou-
meters, starting in the distal extremities. In the tine clinical practice [30]. It is important to
authors’ experience sensation in the palmar distinguish between screening for DPN and
surface of the hands is recognized before the identifying the foot at high risk for ulceration. In
dorsal surface. Hyperpathia may also be evi- the latter context, the 10 g monofilament exami-
dent on examination, although it is advisable nation is an important component. Identification
to compare within, not between patients, such of the at-risk foot is detailed elsewhere [31].
as evaluating the relative perception of pain Quantitative sensory testing (QST) is an adjunct
attributed to pinprick testing between the prox- test of sensation that provides helpful data for
imal and distal portions of a limb. following patients. QST is not reviewed in fur-
Evaluation of temperature involves applying ther depth in this chapter.
warm and cool stimuli to the patient’s extremities, Evaluation of gait begins with observing the
first to determine if temperature sensation is pres- transfer from sitting to standing. If patients are
ent and second to evaluate for a gradient in tem- unable to walk independently, then they should
perature sensation. Methods vary and typically be allowed to use their normal gait aid. As stated
there is an inverse relationship between practical- above, evaluation of the feet may accompany
ity and reliability. A container with a thin or motor inspection, gait examination, or with the

Diabetic Neuropathy
358 21. Bedside clinical features and translational snapshots of diabetic polyneuropathy

patient standing alone. Evaluation of the foot- numbness lose the protective sensation of pain,
wear is an important component of the foot eval- which allows pressure to continuously be
uation. It is important to assess for a wide toe applied to regions of skin breakdown [35].
box, soft-cushioned soles, and sufficient depth to Infections occurs in more than half and defini-
allow modifications, such as orthotics [32]. tive management with amputation is needed in
Abnormal wear of the anterior sole may be an 20% of moderate to severe ulcers. Greater than
indicator of subtle foot drop. In the case of an 70% of patients with amputation related to dia-
ulcer, open footwear with a specialized sole for betes will die within 5 years. Beyond the grave
off-loading is important. prognosis for patients, the prevalence of diabe-
A focused integumentary and vascular exami- tes means that the societal cost of managing
nation of the lower limbs should accompany or diabetic foot ulcers is immense.
follow examination of gait and footwear. Skin Falls can also be a major source of morbidity.
should be inspected for areas of trauma, includ- In a recent cohort study of patients with type 2
ing induration and erythema. Changes in sweat- diabetes, those with symptoms suggestive of
ing patterns may cause visible or palpable DPN had a significantly increased risk of falling,
dryness. Dry skin due to anhidrosis has a while risk of fractures was not significantly dif-
rougher texture than normal skin. On the other ferent, although the incidence of fractures was
hand, hyperhidrosis may be observed in proxi- low [36]. That said, bone fragility is well recog-
mal areas where sudomotor axons are preserved. nized in diabetes and occurs due to complex
In cases of diffuse anhidrosis, flushing will occur interactions between factors, such as oxidative
as a compensatory mechanism to remove excess stress, accumulation of glycation products, and
heat. Care should be taken to inspect between medications [37]. In view of emerging evidence
the toes for cracks and fissures. Ulcers and defor- that strength and balance programs can generate
mities should be noted and potentially photo- lasting improvements in gait, balance, and confi-
graphed for the electronic medical record. Of dence [38,39], recognizing contributions to falls,
particular importance is the presence of neuro- including neuropathy-related gait and balance
pathic arthropathy or “Charcot joint,” which impairments, is very important.
most commonly involves the joints of the feet
and ankles. The posterior tibial and dorsalis
pedis arterial pulses should be palpated. If they Neuropathy scales
cannot be felt, then a hand-held continuous
wave Doppler device can be used to demonstrate Having devoted significant attention to
pulsatile arterial flow [33]. symptoms, it is important to note that a strik-
ing 50% of individuals with DM who have
demonstrable peripheral nerve dysfunction on
Complications clinical examination or paraclinical testing are
asymptomatic [5]. This underscores the impor-
Onset and progression of diabetic neuropa- tance of a complete clinical assessment that
thy are associated with important complica- encompasses both history and examination.
tions. Ulceration of the foot is the most feared Numerous instruments have been developed
complication of diabetes, particularly diabetic that incorporate elements of clinical assess-
neuropathy. Large published studies have ment. These are standardized and have found
implicated neuropathy as a major contributor favor in assessing patients for clinical trials,
in 62% 87% of patients with diabetic foot both in terms of recruitment and outcomes.
ulcers [34]. Patients who have developed Please see the Clinical Trials in DPN, a Recent

Diabetic Neuropathy
 Clinical trials in DPN, a recent snapshot 359
Snapshot section below for more on the impor- studies [44]. The value of the MNSI and MDNS
tance of clinical assessment in contemporary are that they are designed to be used sequen-
clinical trials in diabetic neuropathy. tially as a screening and diagnostic instrument.
The neuropathy impairment scale (NIS) is a The importance of telemedicine has been
scale that assigns points on a linear scale to underscored by the COVID-19 pandemic. The
quantify abnormality in muscle weakness, degree to which telemedicine has been adopted
reflexes, and multimodality sensation bilaterally depends in part on the subdiscipline of medicine,
[40]. Weakness is graded from 0 (normal) to 4 the geographical constraints of the patient popu-
(paralysis), while reflexes and sensation are lation, and the availability of reliable telemedicine
graded from 0 (normal) to 2 (absent). Subscores systems. The conceptual barrier in neuromuscu-
can be calculated and the NIS lower limbs is lar medicine has been physical examination. A
well-suited to evaluating DPN on the basis of its recent report from the Veterans Affairs Los
length-dependent nature [41]. The NIS and its Angeles group has shown promising sensitivity
subscores are openly available in published and specificity results for a clinical scale that can
forms and are available for use [40]. The modi- be administered by a trained technician with a
fied Toronto Clinical Neuropathy Score neurologist observer [45]. A subsequent study by
(mTCNS) similarly uses increasing scores to this group to pilot neuromuscular telemedicine
delineate more significant deficits, but focuses on showed that visits could be performed in a time-
sensory and motor symptoms as well as sensory efficient manner while maintaining high patient
tests to improve sensitivity over clinical tests that satisfaction [46]. Taking all of this together, clini-
capture later stage findings, such as loss of cal assessment and a strong familiarity with the
reflexes [42]. In a cohort of patients with clinical features of diabetic neuropathy will have
impaired glucose tolerance and early neuropathy an enduring importance in future care and
compared to controls, the mTCNS had the high- research.
est sensitivity and specificity of seven scales
tested [42]. The Utah Early Neuropathy Scale
(UENS) was developed to provide a standard- Clinical trials in DPN, a recent snapshot
ized examination that reflects the common
symptoms of neuropathy [43]; however, does not Clinical assessment remains the most impor-
rely on reported symptoms and could be applied tant method by which participants may be
in an asymptomatic patient. The UENS scores selected for clinical trials. Although paraclini-
great toe extension, pinprick sensation in six seg- cal tests provide objective evidence of pathol-
ments of the leg, allodynia in the toes or feet, ogy, they have numerous limitations. Nerve
vibration in the great toe, joint position in the conduction studies, which have been described
great toe, and ankle jerks. The UENS correlates as the gold standard for confirming neuropa-
well with intraepidermal nerve fiber density, thy, only interrogate the function of large
quantitative sudomotor axon reflex testing, and diameter, myelinated nerve fibers. As
nerve conduction studies [43]. The two-step sys- described above, the small diameter, unmyelin-
tem encompassed by the Michigan Neuropathy ated nerve fibers, which are not studied by
Screening Instrument (MNSI) followed by the nerve conduction studies, are frequently
Michigan Diabetic Neuropathy Score (MDNS) is involved first in DPN. Meanwhile, the most
designed to identify potential patients through supported test to evaluate the small fibers is
symptoms and a brief clinical examination, then skin punch biopsy with assessment of intraepi-
confirm a diagnosis with more detailed neuro- dermal nerve fiber density [47], which
logical examination and nerve conduction although minimally so, remains an invasive

Diabetic Neuropathy
360 21. Bedside clinical features and translational snapshots of diabetic polyneuropathy

test that is not widely available, is costly, and methods by which to determine participant eli-
requires laboratory expertise to reliably inter- gibility for clinical trials.
pret. Corneal confocal microscopy is emerging As a snapshot of recent investigations, we
as a noninvasive tool to quantitate nerve fiber have included a table of highly selected ran-
loss in DPN and correlates with other clinical domized controlled clinical trials published in
indices, such as heart rate variability, but is the previous 5 years (2016 21). We selected
currently limited in its use by its restriction to only trials in which 50 or more participants
specialized centers with appropriate equip- completed the study protocol and for which
ment [48]. The scales described in the section the duration was 26 or more weeks (see
Neuropathy Scales provide standardized Table 21.1). We did not include trials in

TABLE 21.1 Selected clinical trials in diabetic neuropathy from 2016 to 2021.
Publication
Reference date N Eligibility Notes

Didangelos et al. 2020 90 Mixed: Michigan Neuropathy Methylcobalamin treatment improved

[49] Screening Instrument and nerve multiple clinical and laboratory indices. 52
conduction studies. weeks.
Didangelos et al. 2020 85 Mixed: Michigan Neuropathy Superoxide dismutase, alpha-lipoic acid,
[50] Screening Instrument and nerve vitamin B12, and carnitine improved clinical
conduction studies. and laboratory measures of peripheral
neuropathy. 52 weeks .
Sekiguchi et al. [51] 2019 492 Mixed, San Antonio criteria: Ranirestat significantly improves nerve
Symptoms, signs, and nerve conduction velocity, but not signs and
conduction studies. symptoms. 52 weeks .
Venkataraman 2019 134 Mixed: Michigan Neuropathy Short-term strength and balance training
et al. [39] Screening Instrument, did not improve quality of life, but did
monofilament test, and improve functional status and balance
neurothesiometer. confidence. 26 weeks.
Vitamin E 2018 229 Clinical: Neuropathy Impairment Negative trial of oral mixed tocotrienols for
Neuroprotection Scale of 2 or higher. treating neuropathic pain. 52 weeks.
Study Investigators
[52]
Vinik et al. [53] 2016 388 Clinical: Michigan Neuropathy 8% capsaicin patch was well-tolerated and
Screening Instrument and ongoing safe. 52 weeks.
neuropathic pain.
Wahren et al. [54] 2016 239 Laboratory: Nerve conduction Weekly C-peptide injections improved
studies. vibration perception but not other clinical or
laboratory indices. 52 weeks.
Satoh et al. [55] 2016 73 Mixed, San Antonio criteria: Ranirestat did not improve the total
Symptoms, signs, and nerve modified Toronto Clinical Neuropathy
conduction studies. Score, but did improve sensory domain. 26
weeks.

N shows the number of participants who completed the trial protocol. See text body for selection criteria.

Diabetic Neuropathy
 Translational opportunities 361
diabetic foot ulcers. The reason for these selec- Translational opportunities
tion criteria is that this chapter is not intended
to provide a comprehensive overview of recent Several recent reviews, including our own,
clinical trial activity in diabetic neuropathy. have provided thoughts over new translational
Rather, this section is intended to examine the opportunities and the reader is referred to
selected larger, time-intensive trials with regard these [57 60]. While the focus from aldose
to inclusion criteria and endpoints. Eight trials reductase inhibitors (ARIs) as a treatment for
met these criteria. Of these, seven used clinical elevated sorbitol and depleted myo-inositol
assessments of neuropathy in the inclusion crite- metabolic changes has declined in recent years,
ria, while one used solely electrodiagnostic fea- other therapeutic approaches and trials are
tures. Two trials used the complete Michigan slowly emerging [61,62]. Benefits from trophic
Neuropathy score consisting of the screening factors applied systemically to ligate neuronal
questionnaire followed by confirmation [49,50], receptors have been limited in previous trials
while one used the clinical instrument without [63]. Our own laboratory is exploring intrinsic
confirmation [53], which still performs well as a molecular pathways that support regenerative
tool for identifying DPN, especially with an growth in adult sensory neurons as a pathway
altered cut point [56]. forward. The concept is that adult neurons tar-
All included trials examined clinical indices
geted by a range of insults, such as oxidative
as part of their selected endpoints. Two trials
and nitrative stress, free radical damage,
did not specify the primary endpoint [49,50],
advanced glycation end-products and others,
which potentially allows for reporting first the
can be supported [64]. The intrinsic neuronal
positive findings followed by the negative find-
mechanisms that support regeneration of neu-
ings. This unfortunately contributes to selec-
rons can similarly shore up targeted neurons.
tion bias toward positive studies, which is a
For example, maintenance of ongoing sensory
practice that devalues the scientific process.
innervation within the intact epidermis during
Four trials explicitly included clinical measures
as primary or coprimary endpoints [39,51 53]. normal skin shedding requires a plasticity in
Two trials explicitly reported laboratory nerve sensory terminals that is shared with regenera-
conduction indices as primary endpoints tive behavior [65]. Other groups have also
[54,55]. Electrophysiological tests are generally developed new ideas along these lines.
considered insensitive to placebo arm treat- Fernyhough and Calcutt have initiated work
ment, although one negative study recently on an intervention to support epidermal axons
raised the issue [54]. Composite clinical indices and facilitate their regrowth through musca-
take into account symptoms and common clini- rinic agonism [66,67]. Other approaches
cal signs and may yield higher quality informa- addressing mitochondrial function and lipo-
tion of importance to patients and toxicity are under exploration in a mix of pre-
understanding the biology of DPN. Some over- clinical and clinical studies. New larger and
all comments might add that while there is rigorous clinical trials to exploit these new
activity in the clinical trial space, most of the opportunities are urgently needed. For neuro-
recent trials listed have not had a comprehen- pathic pain, recent work exploring new targets,
sive battery of assessments included. Also particularly sodium and calcium channel func-
worth pointing out is that only limited num- tion have yielded insights into painful poly-
bers of the agents studied have had rigorous neuropathies, not yet trialed extensively, or for
preclinical evidence or plausible molecular longer durations in human DPN [68 70]. Most
based approaches. of the recent trials listed have been short term

Diabetic Neuropathy
362 21. Bedside clinical features and translational snapshots of diabetic polyneuropathy

FIGURE 21.4 A simplified schematic

summarizing some key bedside features of
diabetic polyneuropathy and their regional
distribution.

with a focus on neuropathic pain and signifi- involve small and large fibers as reflected in its
cant attention toward either newer ARIs (e.g., signs and symptoms (Fig. 21.4). The in-depth
ranirestat), vitamins or their combinations (lim- knowledge of the clinical features of this condi-
ited preclinical evidence) as discussed above. tion remains an important part of caring for
Most clinicians are now avoiding opioids for patients with diabetes, so that neuropathy is
the treatment of painful DPN given the associ- recognized early and devastating complica-
ated healthcare crisis and potential for abuse. tions can be prevented. While abbreviated
assessments are useful and many standardized
scales selectively use elements of the overall
Conclusions clinical assessment, the initial diagnosis of dia-
betic neuropathy is founded on a detailed clini-
Distal symmetric polyneuropathy is the cal characterization that proceeds from history
most common manifestation of diabetic neu- through physical examination. Furthermore,
ropathy and is a very common complication modern clinical trials continue to make use of
overall in diabetes. It often begins with primar- clinical parameters over paraclinical or labora-
ily small fiber symptoms, then progresses to tory measures to enroll patients.

Diabetic Neuropathy
 References 363

References predominantly involves sensory small fibres. J

Peripher Nerv Syst 2012;17:341 50.
[1] Porter R. The greatest benefit to mankind: a medical [20] Veves A, Backonja M, Malik RA. Painful diabetic neu-
history of humanity. New York: W. W. Norton & ropathy: epidemiology, natural history, early diagno-
Company; 1997. sis, and treatment options. Pain Med 2008;9:660 74.
[2] Charcot JM. Sur un cas de paraplégie diabétique. Arch [21] Cheshire Jr. WP. Autonomic history, examination, and
Neurol 1890;19:305 35. laboratory evaluation. Contin. (Minneap. Minn.)
[3] Williamson RT. The vibrating sensation affection of 2020;26:25 43.
the nervous system in diabetes. Lancet [22] Dyck JB, Dyck PJ. Diabetic neuropathy. Toronto: WB
1905;165:856 7. Saunders; 1999.
[4] International Diabetes Federation. IDF Diabetes Atlas [23] Ross RT. How to examine the nervous system.
2019; 2019. Totowa, New Jersey: Humana Press; 2006.
[5] Pop-Busui R, Boulton AJ, Feldman EL, et al. Diabetic [24] Panosyan FB, Mountain JM, Reilly MM, Shy ME,
neuropathy: a position statement by the american dia- Herrmann DN. Rydel-Seiffer fork revisited: beyond a sim-
betes association. Diabetes Care 2017;40:136 54. ple case of black and white. Neurology 2016;87:738 40.
[6] Zochodne DW. Nerve and ganglion blood flow in diabe- [25] Thivolet C, el Farkh J, Petiot A, Simonet C, Tourniaire
tes: an appraisal. Int Rev Neurobiol 2002;50:161 202. J. Measuring vibration sensations with graduated tun-
[7] Zochodne DW. Mechanisms of diabetic neuron damage: ing fork. Simple and reliable means to detect diabetic
molecular pathways. Handb Clin Neurol 2014;126:379 99. patients at risk of neuropathic foot ulceration.
[8] Dyck PJ, Kratz KM, Karnes JL, et al. The prevalence Diabetes Care 1990;13:1077 80.
by staged severity of various types of diabetic neurop- [26] Liniger C, Albeanu A, Bloise D, Assal JP. The tuning
athy, retinopathy, and nephropathy in a population- fork revisited. Diabet Med 1990;7:859 64.
based cohort: the Rochester Diabetic Neuropathy [27] Perkins BA, Olaleye D, Zinman B, Bril V. Simple
Study. Neurology 1993;43:817 24. screening tests for peripheral neuropathy in the diabe-
[9] Lozeron P, Nahum L, Lacroix C, et al. Symptomatic tes clinic. Diabetes Care 2001;24:250 6.
diabetic and nondiabetic neuropathies in a series of [28] Kamei N, Yamane K, Nakanishi S, et al. Effectiveness
100 diabetic patients. J Neurol 2002;249:569 75. of Semmes-Weinstein monofilament examination for
[10] Gorson KC, Ropper AH. Additional causes for distal diabetic peripheral neuropathy screening. J Diabetes
sensory polyneuropathy in diabetic patients. J Neurol Complications 2005;19:47 53.
Neurosurg Psychiatry 2006;77:354 8. [29] Dimitrakoudis D, Bril V. Comparison of sensory test-
[11] Callaghan BC, Price RS, Feldman EL. Distal symmetric ing on different toe surfaces: implications for neuropa-
polyneuropathy: a review. JAMA 2015;314:2172 81. thy screening. Neurology 2002;59:611 13.
[12] Kazamel M, Stino AM, Smith AG. Metabolic syn- [30] Wang F, Zhang J, Yu J, et al. Diagnostic accuracy of
drome and peripheral neuropathy. Muscle Nerve monofilament tests for detecting diabetic peripheral
2021;63:285 93. neuropathy: a systematic review and meta-analysis. J
[13] Dyck PJ. Detection, characterization, and staging of Diabetes Res 2017;2017:8787261.
polyneuropathy: assessed in diabetics. Muscle Nerve [31] Boulton AJ, Armstrong DG, Albert SF, et al.
1988;11:21 32. Comprehensive foot examination and risk assessment:
[14] Scott IA. Errors in clinical reasoning: causes and reme- a report of the Task Force of the Foot Care Interest
dial strategies. BMJ 2009;338:b1860. Group of the American Diabetes Association, with
[15] Treede RD, Jensen TS, Campbell JN, et al. Neuropathic endorsement by the American Association of Clinical
pain: redefinition and a grading system for clinical and Endocrinologists. Diabetes Care 2008;31:1679 85.
research purposes. Neurology 2008;70:1630 5. [32] Mishra SC, Chhatbar KC, Kashikar A, Mehndiratta A.
[16] Katzberg H, Kokokyi S, Halpern E, et al. Prevalence of Diabetic foot. BMJ 2017;359:j5064.
muscle cramps in patients with diabetes. Diabetes [33] Bandyk DF. The diabetic foot: pathophysiology, evalu-
Care 2014;37:e17 18. ation, and treatment. Semin Vasc Surg 2018;31:43 8.
[17] Freeman R. Diabetic autonomic neuropathy. Handb [34] Boulton AJ. Diabetic neuropathy and foot complica-
Clin Neurol 2014;126:63 79. tions. Handb Clin Neurol 2014;126:97 107.
[18] Freeman R, Abuzinadah AR, Gibbons C, et al. [35] Armstrong DG, Boulton AJM, Bus SA. Diabetic foot
Orthostatic hypotension: JACC state-of-the-art review. ulcers and their recurrence. N Engl J Med
J Am Coll Cardiol 2018;72:1294 309. 2017;376:2367 75.
[19] Divisova S, Vlckova E, Hnojcikova M, et al. [36] Khan KS, Christensen DH, Nicolaisen SK, et al. Falls
Prediabetes/early diabetes-associated neuropathy and fractures associated with type 2 diabetic

Diabetic Neuropathy
364 21. Bedside clinical features and translational snapshots of diabetic polyneuropathy

polyneuropathy: a cross-sectional nationwide ques- carnitine for 12 months in patients with diabetic neu-
tionnaire study. J Diabetes Investig 202112:1827 34. ropathy. Nutrients 2020;12:3254.
[37] Napoli N, Chandran M, Pierroz DD, et al. [51] Sekiguchi K, Kohara N, Baba M, et al. Aldose reduc-
Mechanisms of diabetes mellitus-induced bone fragil- tase inhibitor ranirestat significantly improves nerve
ity. Nat Rev Endocrinol 2017;13:208 19. conduction velocity in diabetic polyneuropathy: a ran-
[38] Gu Y, Dennis SM. Are falls prevention programs effec- domized double-blind placebo-controlled study in
tive at reducing the risk factors for falls in people with Japan. J Diabetes Investig 2019;10:466 74.
type-2 diabetes mellitus and peripheral neuropathy: a [52] Vitamin E in Neuroprotection Study (Venus)
systematic review with narrative synthesis. J Diabetes Investigators, Hor CP, Fung WY, et al. Efficacy of oral
Complications 2017;31:504 16. mixed tocotrienols in diabetic peripheral neuropathy: a
[39] Venkataraman K, Tai BC, Khoo EYH, et al. Short-term randomized clinical trial. JAMA Neurol 2018;75:444 52.
strength and balance training does not improve qual- [53] Vinik AI, Perrot S, Vinik EJ, et al. Capsaicin 8% patch
ity of life but improves functional status in individuals repeat treatment plus standard of care (SOC) vs SOC
with diabetic peripheral neuropathy: a randomised alone in painful diabetic peripheral neuropathy: a ran-
controlled trial. Diabetologia 2019;62:2200 10. domised, 52-week, open-label, safety study. BMC
[40] Dyck PJ, Boes CJ, Mulder D, et al. History of standard Neurol 2016;16:251.
scoring, notation, and summation of neuromuscular [54] Wahren J, Foyt H, Daniels M, Arezzo JC. Long-acting
signs. A current survey and recommendation. J c-peptide and neuropathy in type 1 diabetes: a 12-
Peripher Nerv Syst 2005;10:158 73. month clinical trial. Diabetes Care 2016;39:596 602.
[41] Bril V. NIS-LL: the primary measurement scale for [55] Satoh J, Kohara N, Sekiguchi K, Yamaguchi Y. Effect
clinical trial endpoints in diabetic peripheral neuropa- of ranirestat on sensory and motor nerve function in
thy. Eur Neurol 1999;41(Suppl 1):8 13. japanese patients with diabetic polyneuropathy: a ran-
[42] Bril V, Tomioka S, Buchanan RA, Perkins BA. m TSG. domized double-blind placebo-controlled study. J
Reliability and validity of the modified Toronto clini- Diabetes Res 2016;2016:5383797.
cal neuropathy score in diabetic sensorimotor poly- [56] Herman WH, Pop-Busui R, Braffett BH, et al. Use of
neuropathy. Diabet Med 2009;26:240 6. the michigan neuropathy screening instrument as a
[43] Singleton JR, Bixby B, Russell JW, et al. The utah early measure of distal symmetrical peripheral neuropathy
neuropathy scale: a sensitive clinical scale for early in type 1 diabetes: results from the diabetes control
sensory predominant neuropathy. J Peripher Nerv and complications trial/epidemiology of diabetes
Syst 2008;13:218 27. interventions and complications. Diabet Med
[44] Feldman EL, Stevens MJ, Thomas PK, et al. A practical 2012;29:937 44.
two-step quantitative clinical and electrophysiological [57] Kobayashi M, Zochodne DW. Diabetic polyneuropa-
assessment for the diagnosis and staging of diabetic thy: bridging the translational gap. J Peripher Nerv
neuropathy. Diabetes Care 1994;17:1281 9. Syst 2020;25:66 75.
[45] Wilson AM, Ong MK, Saliba D, Jamal NI. The veter- [58] Stino AM, Rumora AE, Kim B, Feldman EL. Evolving
ans affairs neuropathy scale: a reliable, remote poly- concepts on the role of dyslipidemia, bioenergetics,
neuropathy exam. Front Neurol 2019;10:1050. and inflammation in the pathogenesis and treatment
[46] Wilson AM, Jamal NI, Cheng EM, et al. Teleneurology of diabetic peripheral neuropathy. J Peripher Nerv
clinics for polyneuropathy: a pilot study. J Neurol Syst 2020;25:76 84.
2020;267:479 90. [59] Feldman EL, Callaghan BC, Pop-Busui R, et al.
[47] Oaklander AL, Nolano M. Scientific advances in and Diabetic neuropathy. Nat Rev Dis Prim 2019;5:41.
clinical approaches to small-fiber polyneuropathy: a [60] Zochodne DW. The challenges of diabetic polyneuropa-
review. JAMA Neurol 2019;76:1240 51. thy: a brief update. Curr Opin Neurol 2019;32:666 75.
[48] Papanas N, Ziegler D. Corneal confocal microscopy: [61] Oates PJ. Polyol pathway and diabetic peripheral neu-
recent progress in the evaluation of diabetic neuropa- ropathy. Int Rev Neurobiol 2002;50:325 92.
thy. J Diabetes Investig 2015;6:381 9. [62] Chalk C, Benstead TJ, Moore F. Aldose reductase inhi-
[49] Didangelos T, Karlafti E, Kotzakioulafi E, et al. bitors for the treatment of diabetic polyneuropathy.
Vitamin B12 supplementation in diabetic neuropathy: Cochrane Database Syst Rev 2007;(CD004572).
a 1-year, randomized, double-blind, placebo- [63] Apfel SC, Schwartz S, Adornato BT, et al. Efficacy and
controlled trial. Nutrients 2021;13:395. safety of recombinant human nerve growth factor in
[50] Didangelos T, Karlafti E, Kotzakioulafi E, et al. patients with diabetic polyneuropathy: a randomized
Efficacy and safety of the combination of superoxide controlled trial. rhNGF clinical investigator group.
dismutase, alpha lipoic acid, vitamin b12, and JAMA 2000;284:2215 21.

Diabetic Neuropathy
 References 365
[64] Kobayashi M, Zochodne DW. Diabetic neuropathy function and prevent neurodegeneration. Mol Neurobiol
and the sensory neuron: new aspects of pathogenesis 2020;57:2521 38.
and their treatment implications. J Diabetes Investig [68] Lauria G, Ziegler D, Malik R, et al. The role of sodium
2018;9:1239 54. channels in painful diabetic and idiopathic neuropa-
[65] Cheng C, Guo GF, Martinez JA, Singh V, Zochodne thy. Curr Diab Rep 2014;14:538.
DW. Dynamic plasticity of axons within a cutaneous [69] Alsaloum M, Estacion M, Almomani R, et al. A gain-
milieu. J Neurosci 2010;30:14735 44. of-function sodium channel beta2-subunit mutation in
[66] Jolivalt CG, Frizzi KE, Han MM, et al. Topical delivery painful diabetic neuropathy. Mol Pain 2019;15
of muscarinic receptor antagonists prevents and 1744806919849802.
reverses peripheral neuropathy in female diabetic [70] Persson AK, Hoeijmakers JGJ, Estacion M, Black JA,
mice. J. Pharmacol. Exp. Ther. 2020;374:44 51. Waxman SG. Sodium channels, mitochondria, and
[67] Saleh A, Sabbir MG, Aghanoori MR, et al. Muscarinic axonal degeneration in peripheral neuropathy. Trends
toxin 7 signals via ca(2 1 )/calmodulin-dependent pro- Mol Med 2016;22:377 90.
tein kinase kinase beta to augment mitochondrial

Diabetic Neuropathy