Full download ebook at ebookname.

com

Osteoarthritis 2nd Edition Kenneth D. Brandt

https://ebookname.com/product/osteoarthritis-2nd-edition-
kenneth-d-brandt/

OR CLICK BUTTON

DOWLOAD NOW

Download more ebook from https://ebookname.com

More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Effective Teacher Hiring Kenneth D. Peterson

https://ebookname.com/product/effective-teacher-hiring-kenneth-d-
peterson/

Effective Instructional Strategies 4th Edition Kenneth

D. Moore

https://ebookname.com/product/effective-instructional-
strategies-4th-edition-kenneth-d-moore/

Progress in Inorganic Chemistry 1st Edition Kenneth D.

Karlin

https://ebookname.com/product/progress-in-inorganic-
chemistry-1st-edition-kenneth-d-karlin/

Noise and Vibration Analysis Signal Analysis and

Experimental Procedures 2nd Edition Anders Brandt

https://ebookname.com/product/noise-and-vibration-analysis-
signal-analysis-and-experimental-procedures-2nd-edition-anders-
brandt/
Fat Facts Osteoarthritis second edition Fast Facts
Philip Conaghan

https://ebookname.com/product/fat-facts-osteoarthritis-second-
edition-fast-facts-philip-conaghan/

Advances in Cancer Research Volume 127 1st Edition

Kenneth D Tew

https://ebookname.com/product/advances-in-cancer-research-
volume-127-1st-edition-kenneth-d-tew/

Osteoarthritis A Companion to Rheumatology 1st Edition

Leena Sharma Md

https://ebookname.com/product/osteoarthritis-a-companion-to-
rheumatology-1st-edition-leena-sharma-md/

Brittle matrix composites 10 1st Edition Andrzej Brandt

https://ebookname.com/product/brittle-matrix-composites-10-1st-
edition-andrzej-brandt/

Social Psychology 2nd Edition Kenneth S. Bordens

https://ebookname.com/product/social-psychology-2nd-edition-
kenneth-s-bordens/
Osteoarthritis
Second edition

Oxford Medical Publications

Oxford University Press makes no representation, express or implied,
that the drug dosages in this book are correct. Readers must therefore
always check the product information and clinical procedures with the
most up to date published product information and data sheets
provided by the manufacturers and the most recent codes of conduct
and safety regulations. The authors and the publishers do not accept
responsibility or legal liability for any errors in the text or for the
misuse or misapplication of material in this work.
Osteoarthritis
Second edition

Edited by
Kenneth D. Brandt
Rheumatology Division, Department of Medicine, and Department of Orthopaedic Surgery
Indiana University School of Medicine
USA

Michael Doherty
Nottingham University Medical School
Nottingham
UK

L. Stefan Lohmander
University Hospital
Lund
Sweden

1
1
Great Clarendon Street, Oxford  
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide in
Oxford New York
Auckland Bangkok Buenos Aires Cape Town Chennai
Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata
Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi
São Paulo Shanghai Taipei Tokyo Toronto
Oxford is a registered trade mark of Oxford University Press
in the UK and in certain other countries
Published in the United States
by Oxford University Press Inc., New York
© Oxford University Press, 2003
The moral rights of the author have been asserted
Database right Oxford University Press (maker)
First edition Published 1998
Reprinted 2000
Second edition published 2003
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, or under terms agreed with the appropriate
reprographics rights organization. Enquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above
You must not circulate this book in any other binding or cover
and you must impose this same condition on any acquirer
British Library Cataloguing in Publication Data
Data available
Library of Congress Cataloging in Publication Data
ISBN 0 19 850967 7 (Hbk.)
10 9 8 7 6 5 4 3 2 1
Typeset in Minion
by Newgen Imaging Systems (P) Ltd., Chennai, India
Printed in Hong Kong
 Acknowledgements

The efficient and highly competent secretarial support that the editors editors and representatives of the Oxford University Press and, particularly,
received from Joanna Ramowski in Nottingham and Viveca Wiklund in Lund her help in proofing and editing, make this book much better than it would
is hugely appreciated; without their efforts this book would not have seen the have been otherwise. Furthermore, without her diplomacy in reeling in chap-
light of day. In Indianapolis, as usual, Deborah Jenkins provided excellent sec- ters that seemed to be in limbo, we could not have met our deadline.
retarial support. In addition, as was the case when we put together the 1st edi- Finally, we are grateful to Pharmacia for their generous support, which
tion of this book, the initiative, enthusiasm, and dedication to the task shown helped defray production costs for the 2nd edition and, in particular, per-
by Kathie Lane was exemplary. Her professionalism in working with the mitted the use of colour throughout the book.
This page intentionally left blank
 Preface

A number of exciting new developments make it highly desirable at this the pathogenesis of OA. The importance of bone in the pathogenesis of OA
time to summarize the large body of knowledge related to the etiopatho- is reflected by the fact that three chapters are now devoted to this topic.
genesis and management of osteoarthritis. For years, many primary care Whereas discussion of hereditary OA was covered in a single chapter in the
physicians and, indeed, many rheumatologists have considered osteoarthri- 1st edition, three chapters are now devoted to genetic aspects of OA. The
tis to be a boring condition for which they had little to offer patients. chapter on synovial physiology has also been wholly rewritten.
Osteoarthritis has been viewed by physicians and surgeons, erroneously, as The section of the book dealing with therapy continues to reflect the
an inevitable consequence of aging or repetitive usage of a joint and a con- recognition that optimal management of OA requires a comprehensive
dition which, once it becomes symptomatic, progresses inexorably. Doctors program involving both pharmacologic agents and nonmedicinal measures
and allied health professionals have conveyed these misconceptions and and, in some cases, surgery. It is the belief of the editors that it remains
oversimplifications to millions of patients. Consider the nihilism, pes- necessary to dispel the widely held notion that medical management is
simism and futility engendered in those patients by such remarks! ‘conservative’ while surgery is ‘radical’ therapy, whereas the opposite is
But things are changing—and changing rapidly. Our understanding of often true—i.e., withholding surgery from patients with advanced disease
mechanisms underlying the breakdown of articular cartilage in osteoarthri- while they become increasingly deconditioned is, in fact, the radical
tis has grown greatly in the past several years—perhaps to an even greater approach; in such instances it may be much more conservative to operate.
extent than our awareness that osteoarthritis is not a disease simply of car- A recent analysis of outcomes of patients undergoing total hip and total
tilage, but of an organ, the diarthrodial joint. Osteoarthritis represents fail- knee arthroplasty supports that view.
ure of the joint. It may be due in some instances to a primary abnormality As indicated in the 1st edition, studies of animal models have shown that
in the articular cartilage, but in other cases the initial problem resides in the the development and progression of OA may be prevented or retarded with
underlying bone, the synovium, the supporting ligaments or the neuro- pharmacologic or biologic agents. The past few years have witnessed
muscular system. further progress in the development of disease-modifying drugs for
The above two paragraphs are quotations from the Preface to the 1st edi- osteoarthritis (DMOADs). Clinical trials of such therapy have already been
tion of Osteoarthritis. And how rapidly things continue to change! For undertaken in humans. A new chapter is devoted to the possible role of
example, the 4 year interval between the publication of the 1st edition and nutraceuticals, such as glucosamine and chondroitin sulfate, in modifica-
the 2nd has witnessed the important development of cyclooxygenase-2 tion of structural damage in the OA joint. The outcome measures available
selective inhibitors, such as celecoxib and rofecoxib, which are now used for assessment of a DMOAD effect in clinical trials, and limitations thereof,
widely in management of osteoarthritis (OA). The new chapter on non- are also discussed.
steroidal anti-inflammatory drugs (NSAIDs) in the 2nd edition discusses The numerous changes that have been made in the 2nd edition of
this topic in considerable detail. The juxtaposed chapter on economic con- Osteoarthritis reflect the rapid, broad, exciting progress which has occurred
siderations in pharmacologic management of OA has also been entirely in the past few years. The new topics and changes in authorship, providing
rewritten. Both chapters have new authors. expertise in the new areas of discussion, reflect those changes. For the edi-
Another important therapeutic development in the interval between the tors, these changes have assured that their work on this book has been both
1st and 2nd editions is intra-articular hyaluronan therapy. The true efficacy educational and fun.
of this treatment with respect to relief of joint pain and modification of The primary care physician, clinical rheumatologist, orthopaedic sur-
structural damage in the OA joint has generated sufficient heat and con- geon, allied health professional, basic researcher, those in the pharmaceuti-
troversy that an entire chapter is now devoted to this topic. Given that cal industry who are involved in development of drugs and biologicals for
results of sham-controlled clinical trials of joint irrigation are now avail- osteoarthritis, and regulatory agency staff should all find this book useful.
able, a chapter also on this form of intra-articular therapy is now included. The high prevalence of OA—which will continue to increase because of the
The 2nd edition differs from the 1st also in a number of other significant aging of the population—guarantees that health professionals will see a
respects: The chapters on physical therapy and occupational therapy have growing number of patients with OA. The need for a timely review of the
been rewritten, with new authorship for both. Chapters have also been evidence that drives our current understanding of optimal management of
added on exercise in management of OA; weight loss and patient adher- OA and of the basic mechanisms involved in its etiopathogenesis serves as
ence. The section of the book dealing with the pathogenesis of OA has been the rationale for the 2nd edition, whose authors continue to represent
updated considerably, in the light of new knowledge: chapters have been experts on OA from both sides of the Atlantic.
added which discuss protective muscular reflexes; proprioception; the
importance of local mechanical factors, such as joint laxity and malalign- Kenneth D. Brandt March 2003
ment; peripheral and central mechanisms relating to the pathogenesis of Michael Doherty
OA pain; and the possible role of vitamin deficiencies and antioxidants in L. Stefan Lohmander
This page intentionally left blank
 Contents

List of contributors 7.2.2 Bone 125

7.2.2.1 Subchondral bone in the pathogenesis of
1 Definition and classification of osteoarthritis. Mechanical aspects 125
7.2.2.2 Subchondral bone in the pathogenesis of
osteoarthritis 1 osteoarthritis. Biological effects 133
7.2.2.3 Systemic changes in bone in osteoarthritis 142
2 Epidemiology of osteoarthritis 9 7.2.3 Synovium 147
7.2.3.1 Synovial mediators of cartilage damage and
repair in osteoarthritis 147
3 The economics of osteoarthritis 17
7.2.3.2 Synovial physiology in the context of
osteoarthritis 155
4 Genetic aspects of osteoarthritis 23 7.2.4 Neuromuscular system 161
7.2.4.1 Innervation of the joint and its role in
4.1 Evidence for the inheritance of osteoarthritis 25
osteoarthritis 161
4.2 Specific gene defects associated with 7.2.4.2 Neuromuscular protective mechanisms 167
osteoarthritis 33 7.2.4.3 Proprioception in osteoarthritis 172
4.3 Laboratory approaches to the identification of 7.2.5 Local mechanical factors in the natural history of
genetic association in osteoarthritis 41 knee osteoarthritis. Malalignment and joint
laxity 177
5 Pathology of osteoarthritis 49 7.3 Pathogenesis of joint pain in osteoarthritis 185
7.3.1 Peripheral and central pain mechanisms in
6 Paleopathology of osteoarthritis 59 osteoarthritis 185
7.3.2 Why does the patient with osteoarthritis
hurt? 189
7 Pathogenesis of osteoarthritis 67
7.1 Introduction: the concept of osteoarthritis as 8 Clinical features of osteoarthritis
failure of the diarthrodial joint 69 and standard approaches to the
7.2 Pathogenesis of structural changes in the diagnosis 195
osteoarthritic joint 73
7.2.1 Articular cartilage 73 8.1 Signs, symptoms, and laboratory tests 197
7.2.1.1 Biochemistry and metabolism of normal and 8.2 Plain radiographic features of osteoarthritis 211
osteoarthritic cartilage 73
8.3 The natural history and prognosis of
7.2.1.2 Proteolytic degradation of normal and
osteoarthritic cartilage matrix 82
osteoarthritis 227
7.2.1.3 Articular cartilage repair 93
7.2.1.4 Mechanical properties of normal and
9 Management of osteoarthritis 235
osteoarthritic articular cartilage, and the
mechanobiology of chondrocytes 102
9.1 Introduction: the comprehensive approach 237
7.2.1.5 Response of the chondrocyte to mechanical 9.2 Systemic analgesics 243
stimuli 112
9.3 The role of NSAIDs in the treatment of
7.2.1.6 Crystals and osteoarthritis 120
osteoarthritis 251
x 

9.4 Economic considerations in pharmacologic 11.2 Disease modifying osteoarthritis drugs

management of osteoarthritis 259 (DMOADs) 395
9.5 Topical NSAIDs 267 11.2.1 The biological perspective 395

9.6 Topical Capsaicin Cream 273 11.2.2 The clinical perspective 401

9.7 Intra-articular glucocorticoids and other injection 11.3 Advantages and limitations of animal models
therapies 277 in the discovery and evaluation of novel disease-
modifying osteoarthritis drugs (DMOADs) 411
9.8 Intra-artiuclar hyaluronan injection 283
11.4 Assessment of changes in joint tissues in patients
9.9 Nutritional therapies 289 treated with disease modifying osteoarthritis
9.10 Mechanisms by which micronutrients may drugs (DMOADs): monitoring outcomes 417
influence osteoarthritis 293 11.4.1 Radiographic grading systems 417
9.11 Physical therapy 299 11.4.2 Quantitation of radiographic changes 426
9.11.1 Exercise for the patient with osteoarthritis 299 11.4.3 Magnetic resonance imaging 433
9.11.2 Other physical therapies 305 11.4.4 Arthroscopic evaluation of knee articular
9.12 Occupational therapy for the patient with cartilage 451
osteoarthritis 311 11.4.5 Bone scintigraphy 456
9.13 Patient education 321 11.4.6 Ultrasonography 462
11.4.7 Defining and validating the clinical role of
9.14 Social support 327
molecular markers in osteoarthritis 468
9.15 Depression in osteoarthritis 333
9.16 Coping strategies for the patient with 12 Design of clinical trials for
osteoarthritis 339 evaluation of structure modifying
9.17 Irrigation of the osteoarthritic joint 347 drugs and new agents for
9.18 Surgical approaches to preserving and restoring symptomatic treatment of
articular cartilage 353 osteoarthritis 479
9.19 Arthroplasty and its complications 361
9.20 Weight loss and osteoarthritis 371 Appendices 489
9.21 Patient adherence 375 A1 The American College of Rheumatology (ACR)
criteria for the classification and reporting of
10 Outcome assessment in osteoarthritis 491
osteoarthritis: a guide for research A2 Lequesne’s algofunctional lower limb indices 493
and clinical practice 381 A3 WOMAC osteoarthritis Index 495
A4 Protocols for radiography 497
11 Prospects for pharmacological
modification of joint breakdown in Index 501
osteoarthritis 391
11.1 Pharmacological modification of joint breakdown
in OA: Do we need it, can we do it, can we prove
it, is it good? 393
 List of Contributors

Abramson, Steven B.; New York University School of Medicine, Hospital for Joint Griffiths, R.J.; Inflammation Biology, Pfizer Global Research and Development,
Diseases, New York, NY, USA Groton, CT, USA
Anis, Aslam H.; Centre for Health Evaluation and Outcome Sciences, St. Paul’s Grodzinsky, Alan J.; Departments of Electrical and Mechanical Engineering and
Hospital, Vancouver, BC, Canada Biological Engineering Division, Massachusetts Institute of Technology,
Ang, Dennis C.; Rheumatology Division, Indiana University School of Medicine Cambridge, MA, USA
Indianopolis, IN, USA Hadler, Nortin M.; Department of Medicine, University of North Carolina at
Aspnes, Ann; Department of Psychiatry and Behavioral Sciences, Duke Chapel Hill School of Medicine, Chapel Hill, NC, USA
University Medical Center, Durham, NC, USA Hart, Deborah J.; Department of Rheumatology, St. Thomas’ Hospital, London,
Ayral, Xavier; Department of Rheumatology, Cochin Hospital, René Descartes UK
University, Paris, France Hassett, Geraldine; Department of Rheumatology, St. Thomas’ Hospital,
Barlow, Julie; The Interdisciplinary Research Centre in Health, Coventry London, UK
University, Coventry, UK Heinegard, Dick; Department of Cell and Molecular Biology, University of Lund,
Bayliss, Michael; Department of Veterinary Basic Sciences, The Royal Veterinary Lund, Sweden
College, London, UK Hochberg, Marc; Departments of Medicine and Epidemiology and Preventive
Bellamy, Nicholas; CONROD, Faculty of Health Sciences, University of Medicine, University of Maryland School of Medicine, Baltimore, MD, USA
Queensland, Queensland, Australia Hurley, Michael V.; Rehabilitation Research Unit, Physiotherapy Division, King’s
Bischoff, Heike A.; Department of Medicine, Harvard Medical School, Brigham College London, London, UK
& Women’s Hospital, Boston, MA, USA Hung, Clark T.; Department of Biomedical Engineering, Columbia University,
Bradley, John D.; Rheumatology Division, Indiana University School of New York, NY, USA
Medicine, and Wishard Memorial Hospital, Indianapolis, IN, USA Hunziker, Ernst B.; ITI Research Institute, University of Bern, Switzerland
Brandt, Kenneth D.; Rheumatology Division, Department of Medicine, and Ingvarsson, Thorvaldur; Akureyri University Hospital, Akureyri, Iceland
Department of Orthopaedic Surgery, Indiana University School of Jones, Adrian; Rheumatology Unit, Nottingham City Hospital, Nottingham, UK
Medicine, Indianapolis, IN, USA
Kashikar-Zuck, Susmita; Department of Psychiatry and Behavioral Sciences,
Buckland-Wright, J. Christopher; University of London, Applied Clinical Duke University Medical Center, Durham, NC, USA
Anatomy Research Centre, King’s College London, School of Biomedical
Keefe, Francis J.; Department of Psychiatry and Behavioral Sciences, Duke
Sciences, London, UK
University Medical Center, Durham, NC, USA
Buckwalter, Joseph A.; Department of Orthopaedic Surgery, University of Iowa
Kidd, Bruce; Bone and Joint Research Unit, St. Barts and the Royal London
Hospitals, Iowa city, IA, USA
School of Medicine and Dentistry, London, UK
Burr, David B.; Department of Anatomy and Cell Biology, Indiana University
Kim, Young-Jo; Children’s Hospital, Harvard Medical School, Boston, MA, USA
School of Medicine, Indianapolis, IN, USA
Knutson, Kaj; Department of Orthopaedics, University Hospital, Lund, Sweden
Caldwell, David S.; Division of Rheumatology and Immunology, Department of
Medicine, Duke University Medical Center, Durham, NC, USA Kraus, Virginia Byers; Department of Medicine, Division of Rheumatology,
Carr, Alison; Academic Rheumatology, University of Nottingham, Nottingham, UK Duke University Medical Center, Durham, NC, USA
Cibere, Jolanda; Arthritis Research Centre of Canada, Vancouver, BC, Canada Kroenke, Kurt; Regenstrief Institute for Health Care, Indianapolis, IN, USA
Cooper, Cyrus; MRC Environmental Epidemiology Unit, Southampton General Liang, H. Mathew; Department of Medicine, Harvard Medical School, Brigham
Hospital, University of Southampton, Southampton, UK & Women’s Hospital, Boston, MA, USA
Dennison, Elaine; MRC Environmental Epidemiology Unit, Southampton Lohmander, L. Stefan; Department of Orthopedics, University Hospital, Lund,
General Hospital, University of Southampton, Southampton, UK Sweden
Dieppe, Paul MRC Health Services Research Collaboration, University of Lorenzo, Pilar; Department of Cell and Molecular Biology, University of Lund,
Bristol, Bristol, UK Lund, Sweden
DiMicco, Michael A.; Centre for Biomedical Engineering, Massachusetts institute Lorig, Kate; Stanford University, Palo Alto, CA, USA
of Technology, and Children’s Hospital, Harvard Medical School, Cambridge, Manek, Nisha; Twin Research and Genetic Epidemiology Unit, St. Thomas’
MA, USA Hospital, London, UK
Doherty, Michael; Academic Rheumatology, Nottingham City Hospital, Marra, Carlo A.; Centre for Health Evaluation and Outcome Sciences, St. Paul’s
Nottingham, UK Hospital, Vancouver, BC, Canada
Felson, David; Boston University Arthritis Center and the Department of Martin, Eden R.; Department of Medicine, Section of Medical Genetics, Duke
Medicine at Boston City and Boston University Medical Center Hospital, University Medical Center, Durham, NC, USA
Boston, MA, USA Mazucca, Steven A.; Rheumatology Division, Indiana University School of
Flores, Raymond; Department of Medicine, University of Maryland School of Medicine, Indianapolis, IN, USA
Medicine, Baltimore, MD, USA McAlindon, Timothy E.; Boston University School of Medicine, Boston, MA, USA
Ghosh, Peter; Institute of bone and joint research, Royal North Shore Hospital, Melvin, Jeanne L.; Chronic Pain and Fibromyalgia Management Programs,
St Leonards, NSW, Australia Cedars-Sinai Medical Center, Los Angeles, CA, USA
xii   

Minor, Marian A.; Department of Physical Therapy, School of Health Schrier, D.J.; Inflammation Pharmacology, Pfizer Global Research and
Professions, University of Missouri, Columbia, MO, USA Development, Ann Arbor, MI, USA
Mow, Van C.; Department of Biomedical Engineering and New York Orthopaedic Schauwecker, Donald S.; Department of Radiology, Indiana University School of
Hospital Research Laboratory, Columbia University, New York, NY, USA Medicine, Indianapolis, IN, USA
Myers, Stephen L.; Eli Lilly and Company, Indianapolis, IN, USA Sharma, Leena; Division of Rheumatology, Northwestern University Medical
Muir, Kenneth R.; Department of Public Health Medicine and Epidemiology, School, Chicago, IL, USA
Queen’s Medical Center, Nottingham, UK Simkin, Peter A.; Division of Rheumatology, University of Washington, Seattle,
O’Connor, Brian L.; Indiana University School of Medicine, Indianapolis, IN, WA, USA
USA Spector, Tim D.; Twin Research and Genetic Epidemiology Unit, St Thomas’
O’Reilly, Sheila; Rheumatology Unit, Nottingham City Hospital, Nottingham, UK Hospital, London, UK
Peterfy, Charles G.; Synarc, San Francisco, CA, USA Stefansson, Stefan Einar; deCODE, Reykjavik, Iceland
Plaas, Anna H.K.; Biochemistry and Molecular Biology, University of South Tyler, Jenny; Strangeways Research Laboratory, Cambridge, UK
Florida, and Center for Research in Paediatric Orthopaedics, Shriners van Beuningen, Henk M.; University Medical Centre Nijmegen, Nijmegen, the
Hospital, Tampa, FL, USA Netherlands
Poole, A. Robin; Joint Diseases Laboratory and Department of Surgery, Shriner’s van der Kraan, Peter M.; University Medical Centre Nijmegen, Nijmegen, the
Hospital for Crippled Children, Montreal, Quebec, Canada Netherlands
Pritzker, Kenneth P.H.; Pathology and Laboratory Medicine, Mount Sinai van den Berg, Wim B.; University Medical Centre Nijmegen, Nijmegen, the
Hospital, Toronto, Ontario, Canada and University of Toronto, Canada Netherlands
Reid, David M.; Department of Medicine and Therapeutics, University of Vilensky, Joel A.; Indiana University School of Medicine, Fort Wayne, IN, USA
Aberdeen, Aberdeen, UK Walsh, Nicola; Rehabilitation Research Unit, Physiotherapy Division, King’s
Rogers, Juliet; Bristol Royal Infirmary, Bristol, UK (deceased) College London, London, UK
Roos, Ewa M.; Department of Orthopedics, University Hospital, Lund, Sweden Watt, Iain; Directorate of Clinical Radiology, Bristol Royal Infirmary, Bristol, UK
Rosenthal, Ann K.; Division of Rheumatology, Department of Medicine, Medical Webber, Jonathan; Clinical Nutrition Unit, Queen’s Medical Center, Nottingham,
College of Wisconsin, and the Zablocki Veterans Administration Medical UK
Center, Milwaukee, WI, USA Westacott, C.I.; Department of Pathology and Microbiology, University of
Ryan, Lawrence M.; Division of Rheumatology, Department of Medicine, Bristol, School of Medical Sciences, University Walk, Bristol, UK
Medical College of Wisconsin, Milwaukee, WI, USA Weinberger, Morris; Department of Health Policy and Administration,
Sandy, John D.; Department of Pharmacology and Therapeutics, University of University of North Carolina at Chapel Hill, Chapel Hill, NC, USA
South Florida, and Senior Scientist, Center for Research in Skeletal Yelin, Edward; Arthritis Research Group, University of California, San Francisco,
Development and Pediatric Orthopedics, Shriners Hospital, Tampa, FL, USA CA, USA
 1 Definition and classification of
osteoarthritis
Raymond H. Flores and Marc C. Hochberg

Osteoarthritis (OA), formerly referred to as osteoarthrosis and degenerative capsule, synovial membrane, and periarticular muscles. Ultimately, the
joint disease, is the most common form of arthritis.1,2 Prior to 1986, no articular cartilage degenerates with fibrillation, fissures, ulceration, and
standard definition of OA existed; most authors described OA as a disorder full thickness loss of the joint surface. … OA diseases are a result of both
of unknown etiology that primarily affects the articular cartilage and sub- mechanical and biologic events that destabilize the normal coupling of
chondral bone in contrast to rheumatoid arthritis, a disorder that primar- degradation and synthesis of articular cartilage chondrocytes and extra-
ily affects the synovial membrane. In that year, the Subcommittee on cellular matrix, and subchondral bone. Although they may be initiated
Osteoarthritis of the American College of Rheumatology Diagnostic and by multiple factors, including genetic, developmental, metabolic, and
Therapeutic Criteria Committee, proposed the following definition of OA: traumatic, OA diseases involve all of the tissues of the diarthrodial joint.
a heterogeneous group of conditions that lead to joint symptoms and signs Ultimately, OA diseases are manifested by morphologic, biochemical,
which are associated with the defective integrity of articular cartilage, in molecular, and biomechanical changes of both cells and matrix which
addition to related changes in the underlying bone at the joint margins.3 lead to a softening, fibrillation, ulceration, loss of articular cartilage, scler-
A more comprehensive definition of OA was developed at the ‘Workshop osis and eburnation of subchondral bone, osteophytes, and subchondral
on Etiopathogenesis of Osteoarthritis’ sponsored by the National Institute cysts. When clinically evident, OA diseases are characterized by joint
of Arthritis, Diabetes, Digestive, and Kidney Diseases, the National Institute pain, tenderness, limitation of movement, crepitus, occasional effusion,
on Aging, the American Academy of Orthopedic Surgeons, the National and variable degrees of inflammation without systemic effects.
Arthritis Advisory Board, and the Arthritis Foundation.4 This definition
summarized the clinical, pathophysiologic, biochemical, and biomechanical
changes that characterize OA:
Classification of osteoarthritis
Clinically, the disease is characterized by joint pain, tenderness, limitation
of movement, crepitus, occasional effusion, and variable degrees of local OA, as noted above, is a disorder of diverse etiologies, which affects both
inflammation, but without systemic effects. Pathologically, the disease is the small and large joints, either singly or in combination. Table 1.1 con-
characterized by irregularly distributed loss of cartilage more frequently tains a classification schema for OA developed at the ‘Workshop on
in areas of increased load, sclerosis of subchondral bone, subchondral Etiopathogenesis of Osteoarthritis’4 in which idiopathic OA is divided into
cysts, marginal osteophytes, increased metaphyseal blood flow, and vari- two forms: localized or generalized; the latter represents the form of OA
able synovial inflammation. Histologically, the disease is characterized described by Kellgren and Moore involving three or more joint groups.6
early by fragmentation of the cartilage surface, cloning of chondrocytes, Furthermore, generalized OA may occur with or without Heberden’s and
vertical clefts in the cartilage, variable crystal deposition, remodeling, and Bouchard’s nodes, that is as either a nodal or non-nodal form.
eventual violation of the tidemark by blood vessels. It is also characterized The classification of OA into idiopathic (primary) and secondary forms
by evidence of repair, particularly in osteophytes, and later by total loss of was based on the knowledge that OA could result from some recognized
cartilage, sclerosis, and focal osteonecrosis of the subchondral bone. causative factors. These factors operate largely through two mechanisms:
Biomechanically, the disease is characterized by alteration of the tensile, abnormalities of the biomaterials of the joint, usually the articular cartilage;
compressive, and shear properties and hydraulic permeability of the and abnormalities of the biomechanics of the joint, usually due to the
cartilage, increased water, and excessive swelling. These cartilage changes abnormal joint structure, resulting in abnormalities in the distribution of
are accompanied by increased stiffness of the subchondral bone. loading forces across the joint. Thus, patients with an underlying disease
Biochemically, the disease is characterized by reduction in the proteo- that appears to have caused their OA are classified as having secondary OA.
glycan concentration, possible alterations in the size and aggregation of A detailed discussion of several forms of secondary OA can be found else-
proteoglycans, alteration in collagen fibril size and weave, and increased where.7 While this classification is helpful for teaching and research pur-
synthesis and degradation of matrix macromolecules. poses, it has obvious deficiencies as some risk factors for idiopathic OA, for
example obesity, may be considered, alternatively, as causes of secondary
The current definition was developed in 1994 at a workshop entitled OA (see Chapter 2).
‘New Horizons in Osteoarthritis’ sponsored by the American Academy of
Orthopedic Surgeons, the National Institute of Arthritis, Musculoskeletal
and Skin Diseases, the National Institute on Aging, the Arthritis
Foundation, and the Orthopaedic Research and Education Foundation.5
Diagnostic criteria for osteoarthritis
This definition underscores the concept that OA may not represent a single Radiographic criteria
disease entity:
Classically, the diagnosis of OA in epidemiological studies has relied on the
Osteoarthritis is a group of overlapping distinct diseases, which may have characteristic radiographic changes described by Kellgren and Lawrence in
different etiologies but with similar biologic, morphologic, and clinical 19578 and illustrated in the Atlas of Standard Radiographs.9 The cardinal
outcomes. The disease processes not only affect the articular cartilage, radiographic features of OA include: (1) the formation of osteophytes
but involve the entire joint, including the subchondral bone, ligaments, on the joint margins or in ligamentous attachments, as on the tibial spines;
2 1     

Table 1.1 Classification of osteoarthritis* (2) the periarticular ossicles, chiefly in relation to distal and proximal inter-
phalangeal joints; (3) the narrowing of the joint space associated with scler-
I Idiopathic osis of subchondral bone; (4) the cystic areas with sclerotic walls situated in
A Localized the subchondral bone; and (5) the altered shape of the bone ends, particu-
1 Hands: e.g. Heberden’s and Bouchard’s nodes (nodal), erosive larly the head of the femur. Combinations of these changes considered
interphalangeal arthritis (non-nodal), carpal-1st metacarpal together led the authors to the development of an ordinal grading scheme
2 Feet: e.g. hallux valgus, hallux rigidus, contracted toes for severity of radiographic features of OA: 0  normal; 1  doubtful;
(hammer/cock-up toes), talonavicular 2  minimal; 3  moderate; and 4  severe. Different joints are graded
using different characteristics. For the small joints of the hands, knees, and
3 Knee: (a) medial compartment; (b) lateral compartment;
(c) patello-femoral compartment
hips these differences are summarized in Tables 1.2–1.4 and illustrated in
Figs. 1.1–1.5, respectively.
4 Hip: (a) eccentric (superior); (b) concentric (axial, medial);
A number of potential limitations of the use of the Kellgren–Lawrence
(c) diffuse (coxae senilis)
grading scale, as illustrated in the Atlas on Standard Radiographs, have been
5 Spine: (a) apophyseal joints; (b) intervertebral joints (discs); noted.10–13 Foremost among these is the fact that the radiographs of nei-
(c) spondylosis (osteophytes); (d) ligamentous (hyperostosis, ther the hips nor knees were taken in the weight-bearing position, limiting
Forestier’s disease, DISH) the ability of the reader to accurately assess joint-space narrowing as a meas-
6 Other single sites: e.g. glenohumeral, acromioclavicular, tibiotalar, ure of cartilage loss. In an attempt to address the limitations of a global
sacroiliac, temporomandibular grading scale, several groups developed radiographic grading schema that
B Generalized (GOA) includes three or more areas above (6) focus on the individual radiographic features of OA at specific joint groups;
II Secondary reliable grading scales have been published for the hand,14 hip,15,16
knee,17–20 and for all three of these peripheral joint groups.21,22 The atlas
A Trauma
1 Acute
Table 1.2 Grades of severity of osteoarthritis in the small joints of
2 Chronic (occupational, sports)
the hands*
B Congenital or developmental diseases
1 Localized diseases: e.g. Legg–Calve–Perthes, congenital hip Distal interphalangeal joints:
dislocation, slipped epiphysis Grade 1 Normal joint except for one minimal osteophyte
2 Mechanical factors: e.g. unequal lower extremity length, Grade 2 Definite osteophytes at two points with minimal subchondral
valgus/varus deformity, hypermobility syndromes sclerosis and doubtful subchondral cysts, but good joint space
3 Bone dysplasias: e.g. epiphyseal dysplasia, spondyloapophyseal and no deformity
dysplasia, osteonychondystrophy Grade 3 Moderate osteophytes, some deformity of bone ends and
C Metabolic diseases narrowing of joint space
1 Ochronosis (alkaptonuria) Grade 4 Large osteophytes and deformity of bone ends with loss of joint
2 Hemochromatosis space, sclerosis, and cysts

3 Wilson’s disease Proximal interphalangeal joints:

4 Gaucher’s disease Grade 1 Minimal osteophytosis at one point and possible cyst
D Endocrine diseases Grade 2 Definite osteophytes at two points and possible narrowing of
joint space at one point
1 Acromegaly
Grade 3 Moderate osteophytes at many points, deformity of bone ends
2 Hyperparathyroidism
Grade 4 Large osteophytes, marked narrowing of joint space,
3 Diabetes mellitus
subchondral sclerosis, and slight deformity
4 Obesity
First carpometacarpal joint:
5 Hypothyroidism
Grade 1 Minimal osteophytosis and possible cyst formation
E Calcium deposition diseases
Grade 2 Definite osteophytes and possible cysts
1 Calcium pyrophosphate dihydrate deposition disease
Grade 3 Moderate osteophytes, narrowing of joint space, and
2 Apatite arthropathy subchondral sclerosis and deformity of bone ends
F Other bone and joint diseases Grade 4 Large osteophytes, severe sclerosis, and narrowing of joint space
1 Localized: e.g. fracture, avascular necrosis, infection, gout
* Modified from Ref. 9. Reproduced from Silman, A.J., Hochberg, M.C. (1993).
2 Diffuse: rheumatoid (inflammatory) arthritis, Paget’s disease, Epidemiology of the Rheumatic Diseases. Oxford: Oxford University Press.
osteopetrosis, osteochondritis
G Neuropathic (Charcot) arthropathy
H Endemic disorders Table 1.3 Grades of severity of osteoarthritis of the knee*
1 Kashin–Beck Grade 1 Doubtful narrowing of joint space and possible osteophytic lipping
2 Mseleni Grade 2 Definite osteophytes and possible narrowing of joint space
I Miscellaneous conditions
Grade 3 Moderate multiple osteophytes, definite narrowing of joint space
1 Frostbite and some sclerosis and possible deformity of bone ends
2 Caisson’s disease Grade 4 Large osteophytes, marked narrowing of joint space,
3 Hemoglobinopathies severe sclerosis, and definite deformity of bone ends

* Reproduced from Ref. 4 with permission. * Taken from Ref. 9. Reproduced from Silman, A.J., Hochberg, M.C. (1993). Epidemiology
of the Rheumatic Diseases. Oxford: Oxford University Press.
 1      3

Table 1.4 Grades of severity of osteoarthritis of the hip* (a) (b)

Grade 1 Possible narrowing of joint space medially and possible

osteophytes around femoral head
Grade 2 Definite narrowing of joint space inferiorly, definite osteophytes,
and slight sclerosis
Grade 3 Marked narrowing of joint space, slight osteophytes, some sclerosis
and cyst formation, and deformity of femoral head and acetabulum
Grade 4 Gross loss of joint space with sclerosis and cysts, marked deformity
of femoral head and acetabulum, and large osteophytes

* Taken from Ref. 9. Reproduced from Silman, A.J., Hochberg, M.C. (1993). Epidemiology
of the Rheumatic Diseases. Oxford: Oxford University Press.

(a) (b)

(c) (d)

(c) (d)

Fig. 1.2 Grades of severity of OA of the proximal interphalangeal joints:

(a) Grade 1; (b) Grade 2; (c) Grade 3; (d) Grade 4.
Source: Reproduced from Silman, A.J. and Hochberg, M.C. (1993). Epidemiology of the Rheumatic
Diseases. Oxford: Oxford University Press.

inter-reader reliability in measuring the presence and severity of OA of the

hand, hip, and knee; the results of reliability studies have been reviewed by
Lane and Kremer23 and Sun and colleagues.24 In a recent study using radio-
graphs from subjects registered in the Australian Twin Registry, inter-rater
and intra-rater agreement varied by different anatomic sites and different
radiographic features.25 The authors concluded that a single experienced
assessor could reliably classify subjects as having radiographic OA at the
Fig. 1.1 Grades of severity of OA of the distal interphalangeal joints: hand, hip, and knee; when less experienced assessors were involved, how-
(a) Grade 1; (b) Grade 2; (c) Grade 3; (d) Grade 4. ever, independent examinations should be made by at least two persons
Source: Reproduced from Silman, A.J. and Hochberg, M.C. (1993). Epidemiology of the Rheumatic with either adjudication or a consensus reached on disparate examinations.
Diseases. Oxford: Oxford University Press. The validity of using individual radiographic features and a revised com-
posite grading scale has been demonstrated in several studies. Croft et al.
published by the Osteoarthritis Research Society International contains examined the association of individual radiographic features of OA of the
152 black-and-white photos illustrating 39 different radiographic features hip with reported hip pain in 759 men, 60–75 years of age, who had under-
of OA (Table 1.5).18 Using this and other published atlases, trained readers gone intravenous urograms.15 The radiographic feature most strongly asso-
have been shown to have excellent intra-reader and very good-to-excellent ciated with reported hip pain was joint-space width, at the narrowest point,
4 1     

(a) (b)

(c) (d)

Fig. 1.3 Grades of severity of OA of the first carpometacarpal joint: (a) Grade 1; (b) Grade 2; (c) Grade 3; (d) Grade 4.
Source: Reproduced from Silman, A.J. and Hochberg, M.C. (1993). Epidemiology of the Rheumatic Diseases. Oxford: Oxford University Press.

(a) (b)

(c) (d)

Fig. 1.4 Grades of severity of OA of the knee: (a) Grade 1; (b) Grade 2; (c) Grade 3; (d) Grade 4.
Source: Reproduced from Silman, A.J. and Hochberg, M.C. (1993). Epidemiology of the Rheumatic Diseases. Oxford: Oxford University Press.
 1      5

Table 1.5 Radiographic features of osteoarthritis illustrated in the measured in millimeters; in addition, an overall qualitative grade of 3 or
Atlas of the Osteoarthritis Research Society International* higher (Table 1.6) was strongly associated with reported hip pain. These
findings were subsequently confirmed by Scott et al. in an analysis of
Joint group and feature Range of grades data from women aged 65 and older, who had pelvic radiographs obtained
Hand at entry into the ‘Study of Osteoporotic Fractures’, a longitudinal epidemi-
Marginal osteophytes 0–3
ologic study of risk factors for osteoporotic fractures.26 In a more recent
study, Ingvarsson and colleagues compared the reliability of measuring
Joint-space narrowing 0–3
minimum joint space with the Kellgren–Lawrence global scale for assessing
Subchondral sclerosis 0–3 the prevalence of hip OA using colon radiographs in Iceland.27 They noted
Subchondral erosions 0–3 that the measurement of the minimum joint space had better intra-observer
Malalignment 0–3 and inter-observer reliability than the global grade, although prevalence
estimates were similar between the two methods. Based on the validity of
Hip
the association of individual radiographic features with hip pain, and the
Marginal osteophytes 0–3 greater reliability of scoring individual features, as compared to the global
Joint-space narrowing 0–3 Kellgren–Lawrence score,15,16,27 future population-based epidemiological
Subchondral sclerosis 0–3 studies of OA of the hip should rely on the presence of individual radio-
Subchondral lucencies 0–3 graphic features and the Croft or modified Croft global scales, rather
than on the Kellgren–Lawrence grading scale, for classifying cases of OA of
Femoral buttressing 0–3
the hip.28
Knee (Tibiofemoral joint) Spector et al. examined the association of the individual radiographic
Marginal osteophytes 0–3 features of knee OA with reported knee pain in 977 women aged 45–
Joint-space narrowing 0–3 64 years who were participants in the Chingford Study, a longitudinal study
Subchondral sclerosis 0–3
Table 1.6 Croft’s overall qualitative grading of hip OA*
Subchondral erosions 0–3
Malalignment 0–3 Grade Definition
Attrition 0–3 0 No changes of osteoarthritis
Tibial spine hypertrophy 0–1 1 Osteophytosis only
Knee (Patellofemoral joint) 2 Joint-space narrowing only
Marginal osteophytes 0–3 3 Two of osteophytosis, joint-space narrowing,
Joint-space narrowing 0–3 subchondral sclerosis, and cyst formation
Subchondral sclerosis 0–1 4 Three of osteophytosis, joint-space narrowing,
Medial subluxation 0–1 subchondral sclerosis, and cyst formation
Lateral subluxation 0–3 5 As in grade 4, but with deformity of the femoral head

* Modified from Ref. 18. * Modified from Ref. 13.

(a) (b)

(c) (d)

Fig. 1.5 Grades of severity of OS of the hip: (a) Grade 1; (b) Grade 2; (c) Grade 3; (d) Grade 4.
Source: Reproduced from Silman, A.J. and Hochberg, M.C. (1993). Epidemiology of the Rheumatic Diseases. Oxford: Oxford University Press.
6 1     

of musculoskeletal disease in women recruited from a single general prac- In 1981, the Subcommittee on Osteoarthritis of the American College of
tice in Chingford, East London, England.29 They noted that, among the Rheumatology Diagnostic and Therapeutic Criteria Committee was estab-
individual radiographic features of OA, a definite osteophyte in the medial lished to develop clinical criteria for the classification of OA.37 Over the last
compartment was most strongly associated with reported knee pain. The decade, the Subcommittee has developed and published sets of classifica-
odds ratio for the association of grade 1–3 osteophytes with knee pain and tion criteria for OA of the knee,3 hand,38 and hip.39 Altman modified the
the proportion of subjects with grade 1–3 osteophytes who had knee pain criteria sets into algorithms, facilitating their use in clinical research and
were both similar to those for the Kellgren–Lawrence grade 2–4 changes. population-based studies.40 The algorithms for classification of OA of the
In an analysis of data from the Baltimore Longitudinal Study on Aging, knee (Table 1.7), hand (Table 1.8), and hip (Table 1.9), were all developed
Lethbridge-Cejku et al. examined the association of individual radio- using patients with site-specific joint pain due to other types of arthritis or
graphic features and the global Kellgren–Lawrence grade with reported musculoskeletal diseases as the comparison groups. For OA of the knee,
knee pain among 452 men and 223 women, aged 18 and older.30 In support data on 85 historical, physical, laboratory, and radiographic features were
of the findings of Spector et al.,29 they found that the strength of the asso- collected from 130 patients with symptomatic OA of the knee and 105 con-
ciation of definite grade 1–3 osteophytes with current knee pain was simi- trol patients with knee pain due to other etiologies; 55 of the controls had
lar to that for grade 2 or higher OA, using the Kellgren–Lawrence scale; the rheumatoid arthritis.3 For OA of the hand, the Subcommittee collected
odds ratios were 4.4 (95 per cent confidence intervals: 2.6, 7.5) and 4.8 (2.5, data on 51 historical, physical, laboratory, and radiographic features from
8.5), respectively. This relationship was stronger, and remained consistent, 100 patients with symptomatic hand OA and 99 control patients with hand
among the more severe grades of OA: grade 2–3 osteophytes were associ-
ated with current knee pain with an odds ratio of 17.1 (7.5, 38.7), while a
Kellgren–Lawrence grade of 3–4 was associated with current knee pain with Table 1.7 Algorithm for classification of osteoarthritis of the knee,
an odds ratio of 20.8 (8.6, 50.4). Both these studies provided data derived Subcommittee on Osteoarthritis, American College of Rheumatology
from the standing, extended knee radiographs of the tibiofemoral joint. Diagnostic and Therapeutic Criteria Committee*
Felson and colleagues examined the association between clinically
Clinical
diagnosed knee OA, defined as frequent knee symptoms plus crepitus on
physical examination, with radiographic features present on either antero- 1 Knee pain for most days of prior month
posterior weight-bearing or lateral semi-flexed weight-bearing knee radio- 2 Crepitus on active joint motion
graphs.31 The radiographic definitions that best identified knees with 3 Morning stiffness 30 minutes in duration
clinical OA were the presence of a moderate or larger osteophyte, or the
4 Age 38 years
presence of moderate or greater joint-space narrowing plus at least one
5 Bony enlargement of the knee on examination
bony feature (cyst, sclerosis, or osteophyte). Adding information from
the lateral views of the patellofemoral joint enhanced the ability to more Osteoarthritis is present if items 1, 2, 3, 4 or items 1, 2, 5 or items 1 and 5,
efficiently distinguish clinical knee OA from those without clinical knee OA. are present. Sensitivity and specificity are 89 and 88%, respectively.
This study, however, did not have skyline views of the patellofemoral joint, Clinical, laboratory, and radiographic
as illustrated in the atlas published by the Osteoarthritis Research Society 1 Knee pain for most days of prior month
International.22 A study by Lanyon and colleagues, however, did include
2 Osteophytes at joint margins (X-ray spurs)
both an anteroposterior standing and midflexion skyline radiograph of the
knee.32 These authors noted that a case definition based on the presence of 3 Synovial fluid typical of osteoarthritis (laboratory)
a definite osteophyte was more efficient at predicting pain than definitions 4 Age 40 years
based on joint-space width. Furthermore, the addition of the patellofemoral 5 Morning stiffness 30 minutes
joint improved sensitivity for the presence of knee pain. Thus, the use of 6 Crepitus on active joint motion
skyline views of the patellofemoral joint combined with standing views
Osteoarthritis is present if items 1 and 2 or items 1, 3, 5, 6 or items 1, 4, 5,
of the tibiofemoral joint is preferable for epidemiological studies. Based on
6 are present. Sensitivity and specificity are 94 and 88%, respectively.
the validity of the association of individual radiographic features with knee
pain, future population-based epidemiologic studies of OA of the knee * Modified from Refs 3 and 27. Reproduced from Silman, A.J., Hochberg, M.C. (1993).
should rely on the use of individual features alone or in combination to clas- Epidemiology of the Rheumatic Diseases. Oxford: Oxford University Press.
sify cases.33 Protocols for the precise positioning of the tibiofemoral and
patellofemoral compartments of the knee joint have been published.34
Table 1.8 Algorithm for classification of OA of the hand,
Subcommittee on Osteoarthritis, American College of Rheumatology
Clinical criteria
Diagnostic and Therapeutic Criteria Committee*
As noted above, there are potential limitations to the use of only radio-
graphic criteria for a case definition, especially in clinical research studies of Clinical
patients with OA. In particular, although a statistical association exists 1 Hand pain, aching, or stiffness for most days of prior month.
between the radiographic changes of OA and reported pain at both the hip
2 Hard tissue enlargement of 2 of 10 selected hand joints†
and knee, in the individual patient there is often a poor correlation between
the severity of radiographic changes and clinical symptomatology.35 3 Fewer than 3 swollen MCP joints.
At the Third International Symposium on Population Studies of the 4 Hard tissue enlargement of 2 or more DIP joints.
Rheumatic Diseases in 1966, the Subcommittee on Diagnostic Criteria for 5 Deformity of 2 or more of 10 selected hand joints†
Osteoarthrosis recommended that population-based studies should invest- Osteoarthritis is present if items 1, 2, 3, 4 or items 1, 2, 3, 5 are present.
igate the validity of certain historical, physical, and laboratory findings in Sensitivity and specificity are 92 and 98%, respectively.
predicting the typical radiographic features of OA on a joint-by-joint
basis.36 Such historical features include pain on motion, pain at rest, noc- DIP, distal interphalangeal; PIP, proximal interphalangeal; MCP, metacarpophalangeal;
CMC, carpo-metacarpal.
turnal joint pain, and morning stiffness. Features present on physical exam-
ination include bony enlargement and expansion, limitation of motion, * Modified from Refs 25 and 27. Reproduced from Silman, A.J., Hochberg, M.C. (1993).
Epidemiology of the Rheumatic Diseases. Oxford: Oxford University Press.
and crepitus. Laboratory features include the erythrocyte sedimentation † The 10 selected hand joints include bilateral 2nd and 3rd DIP joints, 2nd and 3rd PIP
rate, tests for rheumatoid factor, serum uric acid concentration, and appro-
joints, and 1st CMC joints.
priate analyses of synovial fluid.
 1      7

Table 1.9 Algorithm for classification of hip OA, Subcommittee on inter-rater agreement varied by different anatomic sites, it was excellent at
Osteoarthritis, American College of Rheumatology Diagnostic and all sites. The authors concluded that a single experienced assessor could
Therapeutic Criteria Committee* reliably classify subjects as having clinical OA at the hand, hip, and knee.48
In summary, OA is a complex disorder that may result from many
Clinical, laboratory, and radiographic potential etiologies. Definitions of OA developed at multidisciplinary con-
1 Hip pain for most days of the prior month ferences, with international representation of experts, reflect this complex-
2 Femoral and/or acetabular osteophytes on radiograph
ity. It is difficult, however, to apply these definitions to case definition and
diagnosis in the community or clinic setting.
3 Erythrocyte sedimentation rate 20 mm/h
In the community setting, the criteria for case definitions have tradition-
4 Axial joint-space narrowing on radiograph ally relied on the presence of radiographic features of OA, codified using
Osteoarthritis is present if items 1 and 2 or items 1, 3, 4 are present. the Kellgren–Lawrence grading schema as illustrated in the Atlas of Standard
Sensitivity and specificity are 91 and 89%, respectively. Radiographs. Recently, however, the use of reliable atlases for grading the
severity of the individual radiographic features of OA and of modified
* Modified from Refs 26 and 27. Reproduced from Silman, A.J., Hochberg, M.C. (1993).
Epidemiology of the Rheumatic Diseases. Oxford: Oxford University Press. global scales has received an enthusiastic acceptance among rheumatic dis-
ease epidemiologists. In the clinic setting, however, a case definition based
on radiographic features alone have limitations. For purposes of clinical
research, including therapeutic trials, classification criteria have been devel-
pain of other etiologies: 74 had rheumatoid arthritis.38 For OA of the hip,
oped that use combinations of symptoms, physical findings, laboratory data,
data on 76 historical, physical, laboratory, and radiographic features were
and radiographic features, and have high levels of sensitivity and specificity.
collected from 114 patients with symptomatic OA of the hip and 87 control
These classification criteria have been endorsed by the Osteoarthritis
patients with hip pain of other etiologies: 37 had rheumatoid arthritis.39 At
Research Society International, for use as inclusion criteria for patients in
all joint sites, the sensitivity, specificity, and accuracy of these algorithms
clinical research studies, including clinical trials.49,50
approached or exceeded 90 per cent. Misclassification bias, therefore,
would not likely be a major problem in clinical research studies that
employed these criteria.
Because the major inclusion parameter is joint pain on most days of the Key points
prior month, these criteria sets identify patients with clinically important
OA. This contrasts with the identification of cases of OA based on radio- 1. The definition of OA has evolved over the past two decades and now
graphic features alone, insofar as many, if not most, subjects with radio- recognizes OA as a syndrome with many complex etiologies rather
graphic evidence of OA do not report joint pain.35,41 Therefore, estimates than as a single disease entity.
of the prevalence of OA will be lower when based on the American College 2. In epidemiological studies, radiographic criteria remain the basis
of Rheumatology criteria as compared to the traditional radiographic cri- for classifying subjects as having OA.
teria. In population-based studies, misclassification, particularly of false- 3. Clinical criteria identify persons with symptomatic disease and
negative cases, may be considerable because of the high proportion of should be used for the entry of subjects into clinical trials.
subjects with radiographic evidence of OA who do not have joint pain. For
example, in one study of 400 women aged 45–65 years, Hart and colleagues
noted that the prevalence of symptomatic knee OA was only 2.3 per cent References
compared to a prevalence of radiographic knee OA of 17 per cent.42 In
another study of an elderly population in Iceland, Aspelund and colleagues (An asterisk denotes recommended reading.)
noted that the majority of persons who fulfilled the classification criteria 1. Scott, J.C., Lethbridge-Cejku, M., and Hochberg, M.C. (1999).
for hand OA lacked symptoms on most days of the prior month; further- Epidemiology and economic consequences of osteoarthritis: the American
more, symptoms, when present, were episodic in almost one-third of the viewpoint. In: J-Y. Reginster, J-P. Pelletier, J. Martel-Pelletier, and
Y. Henroitin (eds). Osteoarthritis: Clinical and Experimental Aspects. Berlin:
group.43 They concluded that the clinical criteria were of limited use in
Springer, pp. 20–38.
population surveys. Readers need to be aware of the difference between
2. Scott, J.C. and Hochberg, M.C. (1998). Arthritis and other musculoskeletal
prevalence estimates based on radiographic or clinical criteria for case def-
diseases. In: R.C. Brownson, P.L. Remington, and J.R. Davis (eds). Chronic
initions when reviewing published studies.
Disease Epidemiology and Control, 2nd edition. Washington, DC: American
McAlindon and Dieppe reviewed both the process and development of Public Health Association, pp. 465–89.
the American College of Rheumatology criteria for OA of the knee.44 They
3. *Altman, R., Asch, E., Bloch, D., Bole, G., Borenstein, D., Brandt, K., et al.
noted several potential limitations, including the lack of age- and gender- (1986). Development of criteria for the classification and reporting of
matched controls, inclusion of controls with rheumatoid arthritis, use of osteoarthritis: classification of osteoarthritis of the knee. Arthritis Rheum
criterion items that were largely subjective and not validated, and the 29:1039–49.
absence of a definition or test for OA. Their comments were echoed by The American College of Rheumatology classification schema for OA as well
Balint and Szebenyi.45 The comments were addressed by Altman et al. who as the preliminary criteria for classification of OA of the knee are provided in
noted that the methodology used to construct the criteria adjusted for dif- this article.
ferences between cases and controls, and that the final items included in the 4. Brandt, K.D., Mankin, H.J., and Shulman, L.E. (1986). Workshop on
criteria sets could be reliably and objectively measured.46 etiopathogenesis of osteoarthritis. J Rheumatol 13:1126–60.
The validity of the criteria for classifying patients with hip OA was exam- 5. Keuttner, K. and Goldberg, V.M. (ed.) (1995). Osteoarthritic Disorders.
ined in a primary care setting.47 The authors noted an excellent agreement Rosemont: American Academy of Orthopedic Surgeons, pp. xxi–v.
between the use of the classification tree approach and the criteria set 6. Kellgren, J.H. and Moore, R. (1952). Generalised osteoarthritis and
including radiographic findings, but poor agreement when only clinical Heberden’s nodes. Br Med J 1:181–7.
variables were included. They suggested that radiographic information was 7. Schumacher, H.R. Jr. (2001). Secondary osteoarthritis. In: R.W. Moskowitz,
necessary to apply the criteria for hip OA in clinical practice. D.S. Howell, R.D. Altman, J.A. Buckwalter, and V.M. Goldberg (eds).
In a more recent study, 85 pairs of twins registered in the Australian Twin Osteoarthritis: Diagnosis and Medical/Surgical Management, 3rd edition.
Registry were examined as part of a study to determine the reliability of Philadelphia: WB Saunders, pp. 327–58.
the ACR clinical criteria.48 Two rheumatologists performed independent 8. *Kellgren, J.H. and Lawrence, J.S. (1957). Radiologic assessment of
clinical assessments blinded to laboratory or radiographic data. While osteoarthrosis. Ann Rheum Dis 16:494–501.
8 1     

This landmark paper presents the radiographic features of OA for each of the 32. Lanyon, P., O’Reilly, S., Jones, A., and Doherty, M. (1998). Radiographic
most common sites affected. assessment of symptomatic knee osteoarthritis in the community: defini-
9. The Department of Rheumatology and Medical Illustration, University of tions and normal joint space. Ann Rheum Dis 57:595–601.
Manchester (1973). The Epidemiology of Chronic Rheumatism, Vol. 2, Atlas of 33. Jacobson LT. (1996). Definitions of osteoarthritis in the knee and hand. Ann
Standard Radiographs of Arthritis. Philadelphia: FA Davis, pp. 1–15. Rheum Dis 55:656–8.
10. Spector, T.D. and Cooper, C. (1993). Radiographic assessment of 34. *Buckland-Wright, C. (1995). Protocols for precise radio-anatomical posi-
osteoarthritis in population studies: whither Kellgren and Lawrence? tioning of the tibiofemoral and patellofemoral compartments of the knee.
Osteoarthritis Cart 1:203–6. Osteoarthritis Cart 3(Suppl A):71–80.
11. Spector, T.D. and Hochberg, M.C. (1994). Methodological problems in the This article provides standardized protocols for positioning patients to obtain
epidemiological study of osteoarthritis. Ann Rheum Dis 53:143–6. optimal radiographic views for determining the presence and progression of
12. Hart, D.J. and Spector, T.D. (1995). Radiographic criteria for epidemiologic osteoarthritis.
studies of osteoarthritis. J Rheumatol 22(Suppl 43):46–8. 35. Creamer, P. and Hochberg, M.C. (1997). Why does osteoarthritis of the knee
13. Hart, D.J. and Spector, T.D. (1995). The classification and assessment of hurt—sometimes? Br J Rheumatol 37:726–8.
osteoarthritis. Bailliéres Clin Rheumatol 9:407–32. 36. Bennett, P.H. and Wood, P.H.N. (ed.) (1968). Population Studies of the
14. Kallman, D.A., Wigley, F.M., Scott, W.W. Jr., Hochberg, M.C., and Tobin, J.D. Rheumatic Diseases. International Congress Series No. 148. Excerpta Medica
(1989). New radiographic grading scales for osteoarthritis of the hand. Foundation: Amsterdam, pp. 417–19.
Arthritis Rheum 32:1548–91. 37. Altman, R.D., Meenan, R.F., Hochberg, M.C., Bole, G.G. Jr, Brandt, K.,
15. Croft, P., Cooper, C., Wickham, C., and Coggon, D. (1990). Defining Cooke, T.D.V., et al. (1983). An approach to developing criteria for the
osteoarthritis of the hip for epidemiologic studies. Am J Epidemiol 132:514–22. clinical diagnosis and classification of osteoarthritis: a status report of
16. Lane, N.E., Nevitt, M.C., Genant, H.K., Hochberg, M.C. (1993). Reliability the American Rheumatism Association Diagnostic Subcommittee on
of new indices of radiographic osteoarthritis of the hand and hip and lum- Osteoarthritis. J Rheumatol 10:180–3.
bar disc degeneration. J Rheumatol 20:1911–8. 38. Altman, R., Alarcon, G., Appelrough, D., Bloch, D., Borenstein, D.,
17. Spector, T.D., Cooper, C., Cushnaghan, J., Hart, D.J., and Dieppe, P.A. Brandt, K., et al. (1990). The American College of Rheumatology criteria for
(1992). A Radiographic Atlas of Knee Osteoarthritis. London: Springer. the classification and reporting of osteoarthritis of the hand. Arthritis Rheum
18. Scott, W.W. Jr., Lethbridge-Cejku, M., Reichle, R., Wigley, F.M., Tobin, J.D., 33:1601–10.
and Hochberg, M.C. (1993). Reliability of grading scales for individual radio- 39. Altman R, Alarcon G, Appelrough D, Bloch D, Borenstein D, Brandt K,
graphic features of osteoarthritis of the knee: the Baltimore Longitudinal et al. (1991). The American College of Rheumatology criteria for the classi-
Study of Aging atlas of knee osteoarthritis. Invest Radiol 28:497–501. fication and reporting of osteoarthritis of the hip. Arthritis Rheum 34:
19. Cooke, T.D.V., Kelly, B.P., Harrison, L., Mohamed, G., and Khan, B. (1999). 505–14.
Radiographic grading for knee osteoarthritis: a revised scheme that relates to 40. Altman, R. (1991). Classification of disease: osteoarthritis. Sem Arthritis
alignment and deformity. J Rheumatol 26:641–4. Rheum 20(6,Suppl 2):40–7.
20. Nagaosa, Y., Mateus, M., Hassan, B., Lanyon, P., and Doherty, M. (2000). 41. Lawrence, J.S., Bremner, J.M., and Bier, F. (1966). Osteoarthritis: prevalence
Development of a logically devised line drawing atlas for grading of knee in the population and relationship between symptoms and X-ray changes.
osteoarthritis. Ann Rheum Dis 59:587–95. Ann Rheum Dis 25:1–25.
21. Burnett, S., Hart, D.J., Cooper, C., and Spector, T.D. (1994). A Radiographic 42. Hart, D.J., Leedham-Green, M., and Spector, T.D. (1991). The prevalence of
Atlas of Osteoarthritis. London: Springer. knee osteoarthritis in the general population using different clinical criteria:
22. *Altman, R.D., Hochberg, M.C., Murphy, W.A. Jr., Wolfe, F., and Lequesne, M. the Chingford Study. Br J Rheumatol 30(Suppl. 2):72.
(1995). Atlas of individual radiographic features in osteoarthritis. 43. Alpelund, G., Gunnarsdottir, S., Jonsson, P., and Jonsson, H. (1996).
Osteoarthritis Cart 3(Suppl. A):3–70. Hand osteoarthritis in the elderly: application of clinical criteria. Scand J
The photographs reproduced in this paper are the standards developed by Rheumatol 25:34–6.
the Osteoarthritis Research Society International (OARSI) for grading the indi- 44. McAlindon, T. and Dieppe, P. (1989). Osteoarthritis: definitions and
vidual radiographic features of OA at the most common sites affected. criteria. Ann Rheum Dis 48:531–2.
23. Lane, N.E. and Kremer, L.B. (1995). Radiographic indices for osteoarthritis. 45. Balint, G. and Szebenyi, B. (1996). Diagnosis of osteoarthritis: guidelines
Rheum Dis Clin North Am 21:379–94. and current pitfalls. Drugs 52(Suppl. 3):1–13.
24. Sun, Y., Gunther, K.P., and Brenner, H. (1997). Reliability of radiographic 46. Altman, R.D., Bloch, D.A., Brandt, K.D., Cooke, D.V., Greenwald, R.A.,
grading of osteoarthritis of the hip and knee. Scand J Rheumatol 36:155–65. Hochberg, M.C., et al. (1990). Osteoarthritis: definitions and criteria. Ann
25. Bellamy, N., Tesar, P., Walker, D., Klestov, A., Muirden, K., Kuhnert P, et al. Rheum Dis 49:201.
(1999). Perceptual variation in grading hand, hip and knee radiographs: 47. Bierma-Zeinstra, S., Bohnen, A., Ginai, A., Prins, A., and Verhaar, J. (1999).
observations based on an Australian twin registry study of osteoarthritis. Validity of American College of Rheumatology criteria for diagnosing hip
Ann Rheum Dis 58:766–9. osteoarthritis in primary care research. J Rheumatol 26:1129–33.
26. Scott, J.C., Nevitt, M.C., Lane, N.E., Genant, H.K., and Hochberg, M.C. 48. Bellamy, N., Klestov, A., Muriden, K., Kuhnert, P., Do, K.A., O’Gorman, L.,
(1992). Association of individual radiographic features of hip osteoarthritis and Martin, N. (1999). Perceptual variation in categorizing individuals
with pain. Arthritis Rheum 35(Suppl. 9):S81. according to American College of Rheumatology classification criteria for
27. Ingvarsson, T., Hagglund, G., Lindberg, H., Lohmander, L.S. (2000). hand, knee, and hip osteoarthritis (OA): observations based on an
Assessment of primary hip osteoarthritis: comparison of radiographic Australian Twin Registry study of OA. J Rheumatol 26:2654–8.
methods using colon radiographs. Ann Rheum Dis 59:650–3. 49. Altman, R., Brandt, K., Hochberg, M., and Moskowitz, R. (for the Task
28. Nevitt, M.C. (1996). Definition of hip osteoarthritis for epidemiological Force) (1996). Design and conduct of clinical trials in patients with
studies. Ann Rheum Dis 55:652–5. osteoarthritis: recommendations from a task force of the Osteoarthritis
Research Society. Osteoarthritis Cart 4:217–43.
29. Spector, T.D., Hart, D.J., Byrne, J., Harris, P.A., Dacre, J.E., and Doyle, D.V.
(1993). Defining the presence of osteoarthritis of the knee in epidemiologic This article provides the OARSI guidelines for designing and conducting
studies. Ann Rheum Dis 52:790–4. trials in patients with osteoarthritis of the hip or knee. A companion article
on guidelines for designing and conducting trials in patients with osteoarthritis
30. Lethbridge-Cejku, M., Scott, W.W. Jr, Reichle, R., Ettinger, W.H.,
of the hand is in preparation and will be published in Osteoarthritis Cart
Zonderman, A., Costa, P., et al. (1995). Association of radiographic features
in 2002.
of osteoarthritis of the knee with knee pain: Data from the Baltimore
Longitudinal Study of Aging. Arthritis Care Res 9:182–8. 50. Altman, R. and Maheu, E. (for the Task Force) (2002). Design and conduct
of clinical trials in patients with osteoarthritis of the hand: recommenda-
31. Felson, D.T., McAlindon, T.E., Anderson, J.J., Naimark, A., Weissman, B.W.,
tions from a task force of the Osteoarthritis Research Society. Osteoarthritis
Aliabadi, P., et al. (1997). Defining radiographic osteoarthritis for the whole
Cart 10:in press.
knee. Osteoarthritis Cart 5:241–50.
 2 Epidemiology of osteoarthritis
David T. Felson

studies from the United States of America and Europe suggest that radio-
Impact of OA graphic hip OA occurs in roughly 3–4 per cent of elders.5
Osteoarthritis is the most common form of arthritis. Its high prevalence, Symptomatic OA is generally defined as frequent joint pain plus radio-
especially in the elderly, and the frequency of OA-related physical disability graphic change; its prevalence in the knee has ranged in different studies
make OA one of the leading causes of disability in the elderly, especially from 1.6–9.4 per cent of adults and in 10–15 per cent of elders.
with respect to weight-bearing functional tasks.1 According to a report on Symptomatic hip OA occurs in anywhere from 0.7–4.4 per cent of adults,
the prevalence of arthritis: ‘By 2020 the estimated number of persons with and, using British data from the 1960s, symptomatic hand OA occurs in
arthritis is projected to increase by 57 per cent and activity limitations asso- about 2.6 per cent of adults.
ciated with arthritis by 66 per cent’. These projected increases are largely The prevalence of knee pain and OA by radiograph represents a good
attributable to the high prevalence of OA among older persons and the example of how the prevalence of OA differs depending on how the disease
increasing average age of the US population.2 is defined. Approximately 25 per cent of adults aged 55 and over experience
Epidemiology is the study of disease in populations and its association knee pain for a prolonged period every year (lasting at least a month).
with characteristics of people and their environments. Epidemiological Of these, roughly half have evidence of radiographic OA6 (see Fig. 2.1).
studies document the burden of disease in society and evaluate risk factors Even though they have frequent knee pain, many of these persons do not
for disease that, if modified, might lead to disease prevention and a lessen- necessarily experience limited function from this pain, so that when one
ing of the burden of disability associated with disease. Identifying modifi- addresses the prevalence of knee pain with limitation, it is roughly half the
able risk factors for OA is the first step to prevent this disease and lower its previous prevalence. Severely disabling knee pain is relatively uncommon.
formidable burden in our society. A similar relationship between disabling joint pain and radiographic
change exists for most regions of the body affected by OA.
The prevalence of OA in all joints is strikingly correlated with age.
Prevalence and incidence of OA Regardless of how OA is defined, it is uncommon in adults aged under 40
Osteoarthritis is an extremely common joint disorder in all populations. It and extremely prevalent in those aged above 60. Radiographic hand OA, for
often affects certain joints, yet spares others. For example, in the hands, the example, was present in only about 5 per cent of adults aged under 35, but
distal interphalangeal (DIP), proximal interphalangeal (PIP) joints, and the was seen in over 70 per cent of those who were aged 65 or older.7
carpometacarpal (CMC) joint of the thumb are frequently involved. Other Osteoarthritis has a higher prevalence, and more often exhibits a gen-
joints commonly affected include the cervical spine, lumbosacral spine, hip, eralized distribution, in women than in men. Before the age of 50, men
knee, and first metatarsophalangeal (MTP) joint. The ankle, wrist, elbow, have a higher prevalence than women, but after the age of 50 women have
and shoulder are usually spared. Our joints were designed and shaped, in an a higher prevalence, and this sex difference in prevalence further increases
evolutionary sense, when humans were brachiating apes. Only later did with age.8,9 These gender and age-related prevalence patterns are consistent
humans develop a pincer grip capability3 and full weight-bearing on their with a role of post-menopausal hormone deficiency in increasing the risk
legs. These evolutionary differences in joint function and, possibly, differ- of OA.
ences in the composition of articular cartilage among the different joints, Cross-national and cross-racial studies often produce insights about dis-
predispose some joints to cartilage breakdown, leading to OA. ease etiology. With respect to OA, black women have been reported to have
Osteoarthritis can be defined in many ways. On the one hand, there is higher rates of knee OA than white women10 even after adjustment for age
structural change in a joint often assessed by radiograph. Radiographic and weight. The results in black men have been inconsistent. However,
changes of OA include osteophytes and joint-space narrowing, the latter studies suggest very low rates of hip OA among the black populations in
reflecting cartilage loss. Many persons with radiographic OA do not have Jamaica, South Africa, Nigeria, and Liberia (1–4 per cent for radiographic
joint symptoms. Secondly, one can assess the occurrence of joint symp- OA), in comparison with European populations (7–25 per cent). In
toms. While this is an appealing definition from a clinical and public health Blackfeet and Pima Native Americans, the rates of hip OA are intermediate
standpoint, many with joint symptoms do not have radiographic changes between those for blacks and whites,11 even though the Pimas weigh more,
of disease and may not have OA. Prevalence estimates across studies vary on average, than Caucasians.
because of the inconsistent definitions of symptoms and radiographic The rates of hip OA are much lower in Asians than in Caucasians. The
change. The only exception to this is the Kellgren and Lawrence scale, rate of hip arthroplasty for OA in people of Asian extraction is lower
which has been widely used to evaluate the prevalence of radiographic OA than that in Caucasians in the United States of America.12 Using standard
in most joints. methods across population-based groups, Nevitt et al.13 report that symp-
Radiographic disease is highly prevalent with radiographic hand OA tomatic and radiographic hip OA are far less prevalent in Chinese than in
occurring in approximately 32.5 per cent of adults aged 30 and over.4 The Caucasians from the United States of America. The low rate of hip OA in
prevalence of radiographic knee and hip OA has been best studied in the Asians is not matched by a low rate of knee OA, as Zhang et al.14 using stand-
population surveys of elders. The Framingham Study suggests that radio- ardized methods across populations, found that knee OA prevalence was
graphic knee OA occurs in 33 per cent of people aged 63 and over, and actually more prevalent in Chinese than in Caucasian women and equally
10 2   

Knee pain and

severe disability 160*

Knee pain and some

1250
disability

4 Weeks of knee
2500
pain in past year

People aged 55+

10 000
years

0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10 000

Population of adults aged 55+ years

Fig. 2.1 Prevalence of knee pain and osteoarthritis in persons aged 55 and over. *The proportion with radiographic evidence in this category is not known, though it
seems likely to be high. The shading represents the proportion in each category with radiographic evidence of knee osteoarthritis.
Source: Taken from Peat et al.6

prevalent among men. Although it was initially felt that these racial 1200
differences in hip OA prevalence might be attributed to an absence of Hand males
developmental hip abnormalities in Asians, studies of the Chinese in Incidence rate per 100 000 person-years 1000
Hand females
Hong Kong15,16 fail to confirm a lower prevalence of developmental hip Hip male
× × Hip female
abnormalities. Thus, the explanation for these racial differences in preval-
Knee male
ence is unclear. 800
Knee female
Generalized OA involves hand joints, including the DIP, PIP, and the first
CMC joints; cervical spine; lumbosacral spine and knees; and may include
the hips. There are two types of generalized OA: nodal OA (Heberden’s 600 ×
nodes) and non-nodal OA.17 There is little question that this entity exists;
OA in one joint is associated with the presence of OA in other joints, even ×
400
after adjusting for age and sex. Generalized OA is most common in older
women and may be inherited in a polygenic pattern.18 ×
Because the risk of mortality may be increased in those with OA19,20 200
prevalence estimates may give erroneous estimates of actual disease incid-
ence, that is the new occurrence of disease. A large-scale study from a ×
0× × ×
Massachusetts health maintenance organization21 reported that the age-
20–29 30–39 40–49 50–59 60–69 70–79 80–89
and sex-standardized incidence rate for symptomatic hand OA was 100 per
Age group (years)
100 000 person years. For hip OA the rate was 88 per 100 000 person years
and for knee OA 240 per 100 000 person years. The incidence of hand, knee,
Fig. 2.2 Incidence of OA of the hand, hip, and knee, in members of the Fallon
and hip OA all increased with age (Fig. 2.2) and for each joint was higher in
Community Health Plan, 1991–2, by age and sex.
women than in men after the age of 50. At the age of 70–89, the knee OA
Source: Taken from Oliveria et al.21
rates among women reached a maximum incidence of 1 per cent per year.
Interestingly, a leveling off or decline in the incidence of symptomatic OA
occurred in both sexes around the age of 80. In studies in which serial knee
radiographs were obtained,22 rates of incident symptomatic OA in women
were 1 per cent per year. Radiographic OA of the knee (often asymptomatic) Risk factors for OA
was more frequent with an incidence of 2 per cent per year in women.23 Individual joints become especially susceptible to OA when local factors in
the joint combine with systemic vulnerability. Such local factors might
The descriptive epidemiology of OA include joint deformity or malalignment, but by far the most common local
factor we recognize is previous major injury to a joint, which has left the
◆ The most common form of arthritis
joint vulnerable. Local factors may be powerful determinants of OA by
◆ Occurs most frequently in knees, hands (DIPs, PIPs, MCPs, thumb themselves such that a joint may be sufficiently deformed or affected by a
base), hips, back, neck and spares wrists, ankles major injury as to inevitably lead to OA. But more commonly, this injury
◆ Incidence and prevalence higher in women than in men, especially
acts on a joint within a person whose own systemic vulnerability to OA
varies depending on their age and other factors. Thus, in many cases there
after the age of 50
is interplay of systemic and local vulnerability factors.
◆ Many have joint symptoms without X-ray change and vice versa Within that local and systemic environment, a variety of loading-related
◆ Emerging information on racial differences in occurrence may factors exert influence over whether a joint develops OA. The primary
provide etiologic clues to disease factors are obesity and particular physical activities, which we shall call
extrinsic factors (see Fig. 2.3). Thus, a person who participates in activities
 2    11

developing new radiographic knee OA, but, paradoxically, those with OA

Intrinsic joint vulnerabilities: who had high bone mineral density had a much lower risk of disease pro-
Previous damage (e.g. meniscectomy) gression by radiograph (joint space loss) than those with low bone density.32
Bridging muscle weakness Estrogen deficiency in women. In addition to the high incidence of OA in
Malalignment
women after the age of 50—the approximate age of menopause—some
Proprioceptive deficiencies
Laxity women develop ‘menopausal arthritis’ (rapidly progressive hand OA) at the
time of menopause.33 These gender and age-related prevalence patterns are
Systemic factors: Extrinsic factors
consistent with a role for post-menopausal hormone deficiency in increasing
acting on joints: the risk of OA. In coronary artery disease, gout, and osteoporosis—diseases
Age
Gender Obesity for which the risk in women rises dramatically after menopause as it does in
Genetic susceptibility Specific injurious activities OA—estrogen loss has been strongly implicated as a risk factor.
Nutritional factors Most, but not all, epidemiologic studies34–36 provide evidence that estro-
Susceptibility Osteoarthritis or its gen replacement therapy (ERT) is associated with a reduction in the risk of
to OA or to progression knee and hip OA. For example, both the Study of Osteoporotic Fractures
its progression (SOF)28 and the Framingham Study35 have reported a lower prevalence
of OA among those reporting long-term use of ERT, than among non-users
or among women who used ERT for shorter periods (OR for more than
Fig. 2.3 Risk factors and how they interact to cause OA. 10 years of ERT use, for both SOF and the Framingham Study  0.6). In
both studies the inverse association was stronger when analysis was
that repeatedly injure joints is more likely to develop OA in that joint if they restricted to more severe OA or to bilateral radiographic OA. Follow-up of
have either local or systemic vulnerabilities. Genetic susceptibilities may be the Framingham37 and Chingford Study38 subjects suggested that women
a prime example of the mechanism by which OA occurs in this fashion. in ERT were less likely than those not on ERT to have incident or progres-
Persons who have known mutations predisposing to OA do not get OA in sive OA. Further, ERT users may have more cartilage in their knees than
all joints; rather in isolated joints, presumably those in which there is either non-users.39 Clinical OA, however, may not be less prevalent among ERT
some local vulnerability or extrinsic injury produces OA. The sections users,40 possibly because ERT users see doctors more often and are there-
below detail the evidence for the contribution for specific systemic, local, fore more likely to get their OA diagnosed, a form of detection bias. Lastly,
and extrinsic factors and their relation to the development of OA. while ERT may prevent structural changes of OA such as cartilage loss,
clinical trial evidence suggests it does not have a measurable effect on knee
symptoms.41
Systemic risk factors
Nutritional factors. Vitamin C or ascorbic acid has a multitude of functions
The most potent systemic vulnerabilities are increasing age and female gen-
within cartilage. Among others, it protects against damage by reactive oxy-
der. While getting older does not cause OA per se, disease incidence and
gen species and it serves as a cofactor for enzymes contributing to type II
prevalence increase dramatically with age. Further, given this interaction of
collagen synthesis. In a longitudinal study based in Framingham, low
systemic vulnerabilities and either local or extrinsic factors, it is not sur-
vitamin C intake was associated with accelerated radiographic progression
prising that those who sustain major knee injuries after the age of 30 are at
of OA in those with prevalent disease at baseline. It was unassociated with
a much higher risk of rapidly developing OA48 than persons under the age
incident disease.
of 30 who sustain similar injuries.
Using the same study subjects, a similar protective effect on disease pro-
Also, as noted above, racial factors provide another systemic vulnerabil-
gression of high levels and intakes of vitamin D was seen.42 The rationale
ity with those of Asian extraction having very low rates of hip OA. Other
behind vitamin D is on the surface more farfetched, although chondrocytes
racial differences in disease occurrence are yet to be identified.
at the basal level of cartilage redevelop vitamin D receptors in OA, and
Heritability and genetics. Like many other chronic diseases of onset in mid- vitamin D sufficiency is necessary for active bone turnover, which may be
dle and later years, a large proportion of the occurrence of OA is due, critical in OA. The vitamin D results have been reexamined in a longitud-
in part, to inheritance. The proportion of OA cases attributable to inherit- inal hip study and similar results have been reported.43
ance varies according to the joint. For example, the proportion of hand and
hip OA due, in part, to inheritance is over 50 per cent, yet the proportion Local risk factors: intrinsic (in local joint
of knee OA ranges from 10–30 per cent.24,25 Also, heritability of disease is
higher in early onset of disease than it is among those who have onset of dis- environment)
ease later in life. The details of the heritability of OA and identified genetic Congenital and developmental deformities and hip OA. Three uncommon
factors that may predispose to disease are discussed in later chapters. developmental abnormalities, congenital dislocation, Legg-Perthes disease,
Bone density and OA. Persons with osteoporosis exhibit a lower than and slipped femoral capital epiphysis, lead invariably to hip OA in later life.
expected rate of OA.26 Further, bone density is greater in patients with OA Milder developmental abnormalities, including sub-clinical variants of
than in age-matched controls, even at sites distant from the OA joint.27,28 these developmental diseases may also presage hip OA. Circumstantial evid-
Some, but not all, of the increase in bone mass may be explained by the ence for this link includes the unusual predominance of hip OA in men,44
association of OA with obesity, which protects against osteoporosis. Also, the striking racial disparities in the prevalence of hip OA, and the weak
osteophyte formation, not cartilage loss, may be linked to high bone mass,29 association with hip OA of such traditional risk factors as obesity, suggest-
and a circulating bone growth factor may enhance the growth of osteo- ing the importance of other risk factors. Acetabular dysplasia, a mild variant
phytes and enthesophytes, extra bone formation at the sites of ligament and of congenital dislocation in which the acetabulum is shallow, increases the
tendon insertions.30 The relation of that circulating or local factor and sys- risk of incident hip OA in women and may account for a substantial pro-
temic bone mineral density is unknown. portion of hip OA in women, although probably not in men.45,46
Emerging longitudinal studies are beginning to suggest a complex rela- Major joint injury. With a major joint injury, a person can sustain perma-
tionship between bone density and OA. In one study,31 while women with nent damage of many of the structures within a joint. This damage alters the
incident hand OA had higher bone density than women without it, bone biomechanics of the joint, increases stress across particular areas of the joint
turnover assessed by osteocalcin levels was actually lower in those with incid- and often dramatically increases the risk of OA. Joint cartilage and other
ent OA. In the Framingham Study where subjects were followed longitud- joint structures are often damaged by sudden injuries such as fractures
inally to evaluate a change in OA, high bone density increased the risk of or ligamentous tears. Acute major knee injuries including cruciate
12 2   

ligament and meniscal tears are common causes of knee OA especially Does losing weight lower the risk of OA? For the Framingham women
among men.47,48 Such injuries are prevalent49 in young women athletes and whose baseline BMI values were greater than 25, that is, greater than the
increase their risk of developing OA at a young age. Epidemiologic studies median, weight loss significantly lowered the rate of incident symptomatic
have documented that those with a history of major knee injury are a high knee OA.63 The adjusted odds ratio per two units of BMI (approximately
risk of later, usually ipsilateral, knee OA. In the Framingham study,50 men 5 kg for a woman of normal height) was 0.41, a reduction of more than
with a history of major knee injuries had a relative risk of 3.5 for subsequent 50 per cent in the risk of developing knee OA. Weight gain was associated
knee OA; for women the relative risk was 2.2 (both p  0.05). At sites in with a slightly increased rate of subsequent knee OA (odds ratio 1.28 for a
which the disease is generally uncommon, for example, the ankle and two-unit weight gain). For women whose baseline weight was lower than
shoulder joints, major joint injury may account for a large proportion of all the median, neither weight gain nor weight loss significantly affected their
cases of OA. risk of later disease.
There is clearly interplay between systemic and local injury factors. For Substantial weight loss (dropping from obese to overweight, or from
example, those sustaining major knee injuries of their knees develop OA in overweight to normal weight range) would prevent about 21 per cent of
the knees rapidly if they are older but are less likely to develop OA with such knee OA in men; in women (in whom the association of obesity and knee
rapidity if they are younger.51 Either partial or complete meniscus removal, OA is stronger), 33 per cent of knee OA would be prevented.64 For women,
done when the meniscus is injured increases local stresses and laxity within weight accounts for more OA than any other known factor; for men, over-
the knee52 and is itself a prominent cause of OA.53 Hip injuries may also weight is second to knee injury as a preventable cause of knee OA.
cause OA there, but longitudinal studies evaluating this issue are lacking. Not only does being overweight increase the risk for OA, but for those
Other local risk factors. Those with joint deformities from fractures, with established knee OA it increases the odds of disease progression.65–67
osteonecrosis, or other reasons are at high risk for OA also because of the Furthermore, those with OA in one knee are at higher risk of developing
biomechanical modifications that occur with joint deformity. This consti- OA in the other knee if they are overweight.68 The effect of obesity on OA
tutes one of the reasons why those with destructive inflammatory arthritis progression in joints other than the knee has not been studied.
are at a high risk for getting secondary OA. Lastly, malalignment especially In those who already have OA, weight change is likely to affect symp-
in the knees has recently been linked to rapid disease progression for those toms. McGoey et al.69 studied morbidly obese women at the time of their
who already have knee OA.54 By altering loading across joints, malalign- gastric stapling operation and one year later, at which time their mean
ment may contribute to OA incidence in the knees and other joints. weight loss was approximately 45 kg. The proportion with knee symptoms
dropped from 57 to 14 per cent during this interval and the prevalence of
other regional symptoms, for example, hip pain and back pain, fell com-
Extrinsic risk factors (factors acting on mensurately. Unfortunately, the authors did not definitively ascertain
the joint) whether these subjects had OA.
In another study, 30 obese women (48.7 per cent above ideal weight)
Obesity. Population-based studies consistently show that overweight per-
with knee or hip OA were randomized to phentermine, an appetite sup-
sons are at higher risk of knee OA than non-overweight controls. Estimates
pressant, or placebo, with all subjects instructed to eat a low-energy diet.70
of risk vary from population to population and depend, to some degree, on
Six months later, among 22 patients who remained in the trial, those taking
the criteria for overweight and the definition of OA. In NHANES I, which
phentermine had lost 6.3 kg and those in the placebo group 4.5 kg (p  NS).
was conducted throughout the United States of America from 1971–5,55 the
The amount of weight loss correlated significantly with improvement in a
risk of radiographic OA in obese women (body mass index, BMI, greater
clinical score, which combined symptoms and physical findings. For knee
than 30 but less than 35) was almost four times that in women whose BMI
OA, the correlation between weight loss and improvement was especially
was lower than 25. For men in the same overweight category the risk was 4.8
strong (r  0.66, p  0.01).
times greater than that for men of normal weight. Three to six times body
weight is exerted across the knee during single leg stance in walking;56 Muscle weakness. It has long been recognized that persons with knee OA
therefore, any increase in weight may be multiplied by this factor to reveal have weaker quadriceps than those without the disease.71,72 In a longitud-
the excess force across the knee of an overweight person during walking. In inal follow-up study, women with incident knee OA were weaker at baseline
addition to increasing the risk of tibiofemoral OA, obesity augments the than women who did not develop knee OA.73 The same was not necessarily
risk of disease in the patellofemoral joint. Not only does obesity occur with true for men, but the numbers were small. Quadriceps contraction may
knee OA more than expected, it actually57,58 precedes the development decelerate impulse loading during gait and therefore protect the knee from
of knee OA, suggesting it causes disease, probably through excess loading. damage.74 Since muscle contraction across a joint may be the main source
The relation of obesity to hip OA is weaker than with knee OA, a difference of joint loading, strength may not protect all joints from getting OA.
that is possibly due to the different multiplier effects of body weight across Increased grip strength has been found to be associated with high rates of
the two joint sites or differences in distribution of load across the hip and OA in proximal hand joints like the MCP and base of the thumb, suggesting
knee during weight-bearing.59 that increased loading on the basis of powerful muscle contraction can
Weight could act to cause OA via two mechanisms: first, and most logi- actually damage joints.75
cally, because it increases the amount of load across a joint, obesity could Repeated use of joint. There are two types of activities that involve the
induce cartilage breakdown simply on the basis of excess load. This could repeated use of a joint and have been the focus of epidemiologic studies,
account for the apparent causal relationship between weight and knee and occupational and athletic activities.
hip OA. Indeed, the association of obesity and knee OA is so strong that it Occupational activities and OA. Farmers are at high risk for hip OA. Among
is not likely to be explained by confounding factors. Surprisingly, people farmers, standing, bending, walking long distances over rough ground, lift-
who are overweight may be at slightly higher risk of hand OA.60 The load ing, moving heavy objects, and tractor driving all appear to pose high risks
theory does not explain the relationship between overweight and hand OA, for hip OA, while climbing was not implicated as a risk factor.76
which suggests the involvement of a systemic factor. Following this line of Also, jackhammer operators have a high rate of OA in the upper extrem-
reasoning, it has been speculated that a circulating factor that acts to accel- ity joints that are otherwise rarely affected by disease. Jackhammers impart
erate cartilage breakdown may be present in overweight persons, serving as high impulse loads across joints that are not designed to sustain them.
a second mechanism leading to OA. Vibratory tools may subvert the effectiveness of joint shock absorbers, such
In addition, bone mineral density is increased in overweight persons and as muscles, leading to the transmission of excess stress across the joint and
this (or the absence of osteoporosis) may be a risk factor for OA.61 Additional leading to joint injury.
evidence in favor of a systemic factor is the possibility that the relationship Miners have high rates of OA in the knees and spine,77 while shipyard
between overweight and OA is stronger in women than in men.62 and dockyard workers have a higher prevalence of OA in the knees and
 2    13

fingers than office workers. Cotton mill workers may have high rates of Like occupational activities, OA in athletes seems to occur in especially
hand OA compared to controls.78 over-used joints so that weight lifters who squat and then lift heavy weights
While there are jobs that clearly predispose to OA through joint overuse, have a high risk of patellofemoral OA, whereas soccer players have a higher
there are stereotyped activities that themselves cause OA. Women whose risk of tibiofemoral knee OA.87
jobs require a fine pincer grip (increasing the stress across DIP joints) in a Given the widespread prescription to adopt a healthier more exercise-
Virginia textile mill had significantly more DIP joint OA than women filled lifestyle, large-scale longitudinal epidemiologic studies of the conse-
whose jobs required repeated power grip, a motion that does not stress the quences of exercise on OA contain cautionary notes. In a large-scale study
DIP joints.79 Workers whose jobs required regular knee bending and lifting of women followed for hip OA, those with increased levels of physical activ-
or carrying heavy loads80–83 had a higher rate of knee OA than workers ity either as a teenager or at age 50 were at a higher risk of developing symp-
whose jobs did not entail these activities. tomatic OA later in life88 than women who were sedentary. In a prospective
With respect to the prevention of OA, occupational activities may be follow-up of elders in Framingham, those who reported heavy physical
extremely important. In a study of older male workers, the proportion activity were at a three-fold increased risk of developing both radiographic
of OA potentially attributable to occupational knee bending was higher and symptomatic knee OA at follow-up compared to those who were inact-
(32 per cent) than that attributable to obesity (24 per cent).84 If specific ive. Not all literature is consistent on this matter, however.89 Since physical
job-related tasks, which increase the risk of OA, were changed or avoided activity in general has so many salutatory effects, identifying the particular
many cases of OA would be prevented. Unfortunately, occupation-related activities that pose risks of long-term joint damage (such as those listed
OA is not an easy target for disease prevention. Jobs contain complex above) is preferred to stigmatizing physical activity because it might, in
ergonomic activities and isolating the injurious ones may be difficult. Also, some cases, be risky. Because of the systemic vulnerability to OA among
workers often change jobs or rotate to other tasks. Elimination of all jobs those who are aged, this group may be at a high risk of OA when they
involving substantial physical labor is not realistic, but minor alterations in become physically active later in life.90
ergonomic activities might lower job-related OA risk if specific tasks are
shown as being linked to a high rate of OA.
Leisure time physical activities and OA. With exercise already assuming a
major role in OA treatment, it appears paradoxical to consider the idea that
exercise per se may actually cause OA. But the evidence suggests that certain Risk factors for symptoms
types of exercise may indeed do so. The association of running with OA is of Only approximately half of those with radiographic OA have frequent joint
special interest (see Table 2.1). Most studies evaluating recreational running symptoms. From a clinical and public health perspective, understanding the
have not shown that runners have an increased risk of knee OA, although etiology of joint symptoms is critical. In a recent MRI-based study in which
studies of hip OA have not been consistent. What has emerged, however, is those with knee symptoms were compared to others of a comparable age
an intriguing relation between elite running85 and OA of both the knee and and radiographic disease, MRI features that distinguished those with knee
hip. Professional runners and those on Olympic teams followed over time symptoms from those without were bone marrow edema lesions, synovial
have substantially higher rates of OA in weight-bearing joints than age- thickening, and effusions.91 Large bone marrow edema lesions were present
matched controls. Studies of recreational runners have two important almost exclusively in those with knee pain and were absent in those without
methodological limitations: first, they often miss persons who stop running it, even if they had radiographic disease. Symptom severity was strongly
because they develop knee or hip pain because of early disease and secondly, correlated with an MRI finding of synovitis.92 Findings from this study
they often fail to have information on and adjust for a history of major knee suggest that synovial inflammation and its product, joint effusion, and
injury. One study suggests that runners who get OA are those with the his- bone marrow lesions might be uniquely correlated with the development of
tory of major knee injury.86 knee pain.

Table 2.1 Controlled studies on running and OA

Author (year) Joint No. of subjects Time elapsed since OA in runners compared to controls
running (years)
Sohn and Micheli (1985) Knee 504 25 No increase in knee complaints over
controls (swimmers)
Lane et al. (1993) Knee 33 5 No increase in OA
Panush (1994) Knee 12 8 No increase in OA
Harris et al. (1995) Knee 73 0–40 Increase in osteophytes in runners
Kujala et al. (1995)* Knee 28 35 Non-significant increase (OR  4.8)
of knee OA in runners vs. shooters
Spector et al. (1996)* Knee 73 0–40 Increase in osteophytes in runners
Puranen (1975)* Hip 60 20 No increase in OA
Marti et al. (1989)* Hip 27 15 Runners had more OA than controls
Vingard et al. (1993)* Hip, Knee 233 20–50 Runners had more OA (RR  2.1, p  NS)
Kujala et al. (1994) Hip, 100 30 Runners had more hospitalization for knee
and Ankle and hip OA (RR  1.8; 95% CI 0.9–3.6)
Cheng et al. (2000) Hip and Knee 16 0–25 Male but not female runners ( 20 miles/week)
had more OA than non-runners
(OR  1.6; 95% CI 1.1–2.3)

* Studies of elite runners.

14 2   

Table 2.2 Do biomechanical factors affect the knee differently at 7. Engel, A. (1968). Osteoarthritis and body measurements. Rockville MD:
different stages of OA? National Center for Health Statistics, Series 11, no. 29, PHS publication
no. 1999.
Odds ratio for 8. Felson, D.T., Naimark, A., Anderson, J., Kazis, L., Castelli, W., and
Incident (new) disease Progressive disease
Meenan, R.F. (1987). The prevalence of knee osteoarthritis in the elderly.
Arthritis Rheum 30(8):914–18.
High BMI* 9.1 2.6 9. Van Saase, J.L.C.M., Van Romunde, L.K.J., Cats, A., Vandenbroucke, J.P.,
Previous knee injury* 4.8 1.2 and Valkenburg, H.A. (1989). Epidemiology of osteoarthritis: Zoetermeer
survey. Comparison of radiological osteoarthritis in a Dutch population
Regular sports* 3.2 0.7
with that in 10 other populations. Ann Rheum Dis 48:271–80.
High bone density† 2.3 0.1 10. Anderson, J. and Felson, D.T. (1988). Factors associated with osteoarthritis
* Taken from Cooper et al., 2000.94 of the knee in the First National Health and Nutrition Examination Survey
(NHANES I). Am J Epidemiol 128:179–89.
† Taken from Zhang et al., 2000.33
11. Lawrence, J.S. and Sebo, M. (1980). ‘The geography of osteoarthritis’. In
The Aetiopathogenesis of Osteoarthritis, Baltimore: University Park Press,
pp. 155–83.
12. Hoaglund, F.T., Oishi, C.S., and Gialamas, G.G. (1995). Extreme variations
in racial rates of total hip arthroplasty for primary coxarthrosis: A
population-based study in San Francisco. Ann Rheum Dis 54:107–10.
Risk factors for incidence vs. progression 13. Nevitt, M., Xu, L., Zhang, Y., et al. (2000). Chinese in Beijing have a very low
prevalence of hip OA compared to U.S. Caucasians. Arthritis Rheum
Many of the risk factors discussed above may operate differently on joints at 43:S171.
different stages of disease. Studies have suggested, for example, that vitamin 14. Zhang, Y., Xu, L., Nevitt, M.C., et al. (2001). Comparison of the prevalence
D deficiency may accelerate the risk of OA progression but does not appear of knee osteoarthritis between the elderly Chinese population in Beijing and
to affect the risk of incident disease. Vitamin C deficiency may have similar Whites in the U.S.: The Beijing Osteoarthritis Study. Arthritis Rheum
effects.89 As noted above, bone density may actually have opposite effects at 44:2065–71.
different disease stages with high bone density increasing the risk of inci- 15. Hoaglund, F.T., Yau, A.C.M.C., and Wong, W.I. (1973). Osteoarthritis of the
dent disease yet delaying progression. Some factors such as obesity, pre- hip and other joints in Southern Chinese in Hong Kong. J Bone Joint Surg
vious knee injury, and regular sports prescriptions, have been strongly 55(A):545–57.
linked to incidence yet in the same populations appear unrelated to the risk 16. Lau, E. and Lin, F. (1994). Low prevalence of osteoarthritis of the hip in
of progression90 (see Table 2.2). Chinese men. Arthritis Rheum 37(Suppl.):S239.
It would be unusual if risk factors varied at different stages of the disease, 17. Kellgren, J.H. and Moore, R. (1952). Generalized osteoarthritis and
but when disease is present the joint becomes more vulnerable and one Heberden’s nodes. Br Med J 1:181–7.
might expect risk factors to act in a more potent fashion, although the data 18. Lawrence, J.S. (1977). Osteoarthrosis. In J.S. Lawrence (ed.). Rheumatism in
do not necessarily suggest they do. Other explanations for the differences Populations. London: Heinemann.
reported between risk factors for incidence and progression include: the 19. Cerhan, J.R., Wallace, R.B., El-Khoury, G.Y., Moore, T.E., and Long, C.R.
numbers are small in each of these studies and differences reported may not (1995). Decreased survival with increasing prevalence of full-body, radi-
reflect real ones; risk factors for incidence are really risk factors for osteo- ographically defined osteoarthritis in women. Am J Epidemiol 141:225–34.
phytes and those for progression are risk factors for joint space or cartilage 20. Monson, R.R. and Hall, A.P. (1976). Mortality among arthritics. J Chronic
loss; and lastly that risk factors really do differ at different stages of the Dis 29:359–467.
disease perhaps because of the different metabolic or restorative powers of 21. Oliveria, S.A., Felson, D.T., Reed, J.I., Cirillo, P.A., and Walker, A.M.
the joint. (1995). Incidence of symptomatic hand, hip, and knee osteoarthritis
among patients in a health maintenance organization. Arthritis Rheum
38:1134–41.
22. Felson, D.T., Zhang, Y., Hannan, M.T., et al. (1995). The Incidence and
References natural history of knee osteoarthritis in the elderly: The Framingham
Osteoarthritis Study. Arthritis Rheum 38:1500–5.
(An asterisk denotes recommended reading.) 23. Hart, D.J., Doyle, D.V., and Spector, T. (1999). Incidence and risk factors for
1. Guccione, A.A., Felson, D.T., Anderson, J.J., et al. (1994). The effects of radiographic knee osteoarthritis in middle-aged women. Arthritis Rheum
specific medical conditions on the functional limitations of elders in the 42:17–24.
Framingham Study. Am J Public Health 84:351–8. 24. Demissie, S., Cupples, L.A., Myers, R., Aliabadi, P., Levy, D., and Felson, D.T.
2. US Department of Health and Human Services (1995). CDC: Prevalence (2002). Genome scan for quantity of hand osteoarthritis: the Framingham
and impact of arthritis among women—United States, 1989–1991. MMWR study. Arthritis Rheum 46:946–52.
44:329–34. 25. *Spector, T.D., Cicuttini, F., Baker, J., Loughlin, J., and Hart, D. (1996).
3. Hutton, C.W. (1987). Hypothesis. Generalised osteoarthritis: an evolution- Genetic influences on osteoarthritis in women: a twin study. Br Med J
ary problem? Lancet 1:1463–5. 312:940–4.
4. Lawrence, R.C., Helmick, C.G., Arnett, F.C., et al. (1998). Estimates of the In data from a large twin study, these investigators reported that radiographic
prevalence of arthritis and selected musculoskeletal disorders in the United OA was highly heritable, especially OA in the hand joints.
States. Arthritis Rheum 41:778–99. 26. Hart, D.J., Mootoosamy, I., Doyle, D.V., and Spector, T.D. (1994). The rela-
5. Nevitt, M.C., Lane, N.C., Scott, J.C., et al. (1995). Radiographic osteoarthri- tionship between osteoarthritis and osteoporosis in the general population:
tis of the hip and bone mineral density. Arthritis Rheum 38:907–16. the Chingford Study. Ann Rheum Dis 53:158–62.
6. *Peat, G., McCarney, R., and Croft, P. (2001). Knee pain and osteoarthritis 27. Gevers, G., Dequeker, J., Martens, M., et al. (1989). Biomechanical char-
in older adults: a review of community burden and current use of primary acteristics of iliac crest bone in elderly women according to osteoarthritis
health care. Ann Rheum Dis 60:91–7. grade at the hand joints. J Rheumatol 16:660–3.
This article provides a clear explanation and supporting data on the preva- 28. Dequeker, J., Boonen, N., Aerssens, J., Westhovens, R. (1996). Inverse
lence of knee pain and OA in the community and how these vary depending on relationship osteoarthritis–osteoporosis: what is the evidence? What are the
the definition. consequences? (Editorial). Br J Rheumatol 35:813–20.
 2    15

29. Hannan, M.T., Anderson, J.J., Zhang, Y., Levy, D., and Felson, D.T. (1993). 51. *Roos, H.P., Adalberth, T., Dahlberg, L., Lohmander, L.S. (1995).
Bone mineral density and knee osteoarthritis in elderly men and women. Osteoarthritis of the knee after injury to the anterior cruciate ligament or
Arthritis Rheum 36:1671–80. meniscus: the influence of time and age. Osteoarthritis Cart 3:261–7.
30. Rogers, J., Shepstone, L., Dieppe, P. (1997). Bone formers: osteophyte An important article suggesting that knee injury causes accelerated develop-
and enthesophyte formation are positively associated. Ann Rheum Dis ment of OA if it occurs after age 30.
56:85–90. 52. Walker, P.S., Erkman, M.J. (1975). The role of the menisci in force transmis-
31. Sowers, M., Lachance, L., Jamadar, D., et al. (1999). The associations of sion across the knee. Clin Orthop Rel Res 109:184–192.
bone mineral density and bone turnover markers with osteoarthritis of the 53. Englund, M., Roos, E.M., Roos, H.P., Lohmander, L.S. (2001). Patient-
hand and knee in pre- and perimenopausal women. Arthritis Rheum relevant outcomes fourteen years after meniscectomy: influence of types of
42:483–9. meniscal tear and size of resection. Rheumatology 40:631–639.
32. Zhang, Y., Hannan, M.T., Chaisson, C.E., et al. (2000). Bone mineral density 54. Sharma, L., Song, J., Felson, D.T., Cahue, S., Shamiyeh, E., Dunlop, D.D.
and risk of incident and progressive radiographic knee osteoarthritis in (2001). The role of knee alignment in disease progression and functional
women: the Framingham study. J Rheumatol 27:1032–7. decline in knee osteoarthritis. JAMA 286:188–195.
33. Kellgren, J.H. and Moore, R. (1952). Generalized osteoarthritis and 55. Anderson, J. and Felson, D.T. (1988). Factors associated with osteoarthritis
Heberden’s nodes. Br Med J 1:181–7. of the knee in the First National Health and Nutrition Examination Survey
34. Nevitt, M.C., Cummings, S.R., Lane, N.E., Genant, H.K., and (NHANES I). Am J Epidemiol 128:179–89.
Pressman, A.R. (1994). Current use of oral estrogen is associated with a 56. Maquet, P. (1976). Biomechanics of the Knee. Springer-Verlag, New York.
decreased prevalence of radiographic hip OA in elderly white women. 57. Felson, D.T., Anderson, J.J., Naimark, A., Walker, A.M. and Meenan, R.F.
Arthritis Rheum 37(Suppl.):S212. (1988). Obesity and knee osteoarthritis. Ann Intern Med 109:18–24.
35. Hannan, M.T., Felson, D.T., Anderson, J.J., Naimark, A., and Kannel, W.B. 58. Schouten, J. (1991). A 12-year follow-up study of osteoarthritis of the knee
(1990). Estrogen use and radiographic osteoarthritis of the knee in women. in the general population, Thesis, Erasmus University, pp. 149–64.
Arthritis Rheum 33:525–32. 59. Heliovaara, M., Mkel, M., and Impivaara, O., et al. (1993). Association of
36. Wolfe, F., Altman, R., Hochberg, M., Lane, N., Luggan, M., and Sharp, J. overweight, trauma and workload with coxarthrosis: a health survey of 7,217
(1994). Postmenopausal estrogen therapy is associated with improved radi- persons. Acta Orthop Scand 64:513–18.
ographic scores in OA and RA. Arthritis Rheum 37(Suppl.):S231. 60. Carman, W.J., Sowers, M.F., Hawthorne, V.M., and Weissfeld, L.A. (1994).
37. Zhang, Y, McAlindon, T.E., Hannan, M.T., et al. (1998). Estrogen replace- Obesity as a risk factor for osteoarthritis of the hand and wrist: a prospective
ment therapy and worsening of radiographic knee osteoarthritis. Arthritis study. Am J Epidemiol 39:119–29.
Rheum 41:1867–73. 61. Hannan, M.T., Anderson, J.J., Zhang, Y., Levy, D., and Felson, D.T. (1993).
38. Hart, D.J., Doyle, D.V., and Spector, T.D. (1999). Incidence and risk factors Bone mineral density and knee osteoarthritis in elderly men and women: the
for radiographic knee osteoarthritis in middle-aged women: the Chingford Framingham Study. Arthritis Rheum 36:1671–80.
Study. Arthritis Rheum 42:17–24. 62. *Felson, D.T., Zhang, Y., Anthony, J.M., Naimark, A., and Anderson, J.J.
39. Wluka, A.E., Davis, S.R., Bailey, M., Stuckey, S.L., and Cicuttini, F.M. (1992). Weight loss reduces the risk for symptomatic knee osteoarthritis in
(2001). Users of oestrogen replacement therapy have more knee cartilage women. Ann Intern Med 116:535–9.
than non-users. Ann Rheum Dis 60:332–6. Longitudinal study of women suggesting that weight 12 years prior to onset
40. Sandmark, H., Hogstedt, C., Lewold, S., and Vingard, E. (1999). Osteoarthrosis of symptomatic OA increases risk, and that modest weight loss (approximately
of the knee in men and women in association with overweight, smoking, and 12 lbs) can dramatically lower the risk of developing disease, especially among
hormone therapy. Ann Rheum Dis 58:151–5. those who are overweight at baseline.
41. Nevitt, M.C., Felson, D.T., Williams, E.N., and Grady, D. (for the Heart and 63. Schouten, J.S.A.G., van den Ouweland, and Valkenburg, H.A. (1992).
Estrogen/Progestin Replacement Study Research Group) (2001). The effect A 12-year follow-up study in the general population on prognostic factors of
of estrogen plus progestin on knee symptoms and related disability in post- cartilage loss in osteoarthritis of the knee. Ann Rheum Dis 51:932–7.
menopausal women. Arthritis Rheum 44:811–8. 64. *Felson, D.T. and Zhang, Y.Q. (1998). An update on the epidemiology of
42. *McAlindon, T.E., Felson, D.T., Zhang, Y., et al. (1996). Relation of dietary knee and hip osteoarthritis with a view to prevention. Arthritis Rheum
intake and serum levels of Vitamin D to progression of osteoarthritis of the 41:1343–55.
knee among participants in the Framingham Study. Ann Intern Med A comprehensive recent review of disease epidemiology including data on
125:353–9. prevalence of disease in knee and hip and a critical review of current thinking
A longitudinal study suggesting that low 25-OH vitamin D levels increase about risk factors.
the risk of progressive radiographic knee OA. 65. Schouten, J.S.A.G., van den Ouweland, and Valkenburg, H.A. (1992).
43. Lane, N.E., Gore, L.R., Cummings, S.R., et al. (1999). Serum vitamin D A 12-year follow-up study in the general population on prognostic factors of
levels and incident changes of radiographic hip osteoarthritis. Arthritis cartilage loss in osteoarthritis of the knee. Ann Rheum Dis 51:932–7.
Rheum 42:854–60. 66. Dougados, M., Gueguen, A., Nguyen, M., et al. (1992). Longitudinal radio-
44. Kellgren, J.H. (1961). Osteoarthritis in patients and populations. Br Med J logic evaluation of osteoarthritis of the knee. J Rheumatol 19:378–83.
243:1–6. 67. Altman, R.D., Fried, J.F., Bloch, D.A., et al. (1987). Radiographic assessment
45. Solomon, L. (1976). Patterns of osteoarthritis of the hip. J Bone Joint Surg of progression in osteoarthritis. Arthritis Rheum 30:1214–25.
58(B):176–83. 68. Spector, T.D., Hart, D.J., and Doyle, D.V. (1994). Incidence and progression
46. Lane, N.E., Lin, P., Christiansen, L., et al. (2000). Association of mild acetab- of osteoarthritis in women with unilateral knee disease in the general popu-
ular dysplasia with an increased risk of incident hip osteoarthritis in elderly lation: the effect of obesity. Ann Rheum Dis 53:565–8.
white women. Arthritis Rheum 43:400–4. 69. McGoey, B.V., Deitel, M., Saplys, R.J.F., and Kliman, M.E. (1990). Effect of
47. Jacobsen, K. (1977). Osteoarthrosis following insufficiency of the cruciate weight loss on musculoskeletal pain in the morbidly obese. J Bone Joint Surg
ligaments in man. Acta Orthop Scand 48:520–6. 72(B):322–3.
48. Roos, H.P., Lauren, M., Adalberth, T., Roos, E.M., Jonsson, K., and 70. Williams, R.A. and Foulsham, B.M. (1981). Weight reduction in osteoarthri-
Lohmander, L.S. (1998). Knee osteoarthritis after meniscectomy. Arthritis tis using phentermine. The Practitioner 225:231–2.
Rheum 41:687–93. 71. Slemenda, C., Brandt, K.D., Heilman, et al. (1997). Quadriceps weakness
49. Roos, H.P., Ostenberg, A. (2000). A prospective study of female soccer and osteoarthritis of the knee. Ann Intern Med 127:97–104.
players with anterior cruciate ligament tear. Radiographic findings and 72. Madsen, O.R., Bliddal, H., Egsmose, C., and Sylvest, J. (1995). Isometric and
symptoms 12 years after injury. Arthritis Rheum 43:913. isokinetic quadriceps strength in gonarthrosis; inter-relations between
50. Felson, D.T. (1990). The epidemiology of knee osteoarthritis: results quadriceps strength, walking ability, radiology, subchondral bone density
from the Framingham osteoarthritis study. Sem Arthritis Rheum 20: 42–50. and pain. Clin Rheumatol 14:308–14.
16 2   

73. *Slemenda, C., Heilman, D.K., Brandt, K.D., et al. (1998). Reduced quadri- 85. Spector, T.D., Harris, P.A., Hart, D.J., et al. (1996). Risk of osteoarthritis asso-
ceps strength relative to body weight. A risk factor for knee osteoarthritis in ciated with long-term weight-bearing sports. Arthritis Rheum 39:988–95.
women? Arthritis Rheum 41:1951–9. 86. McDermott, M. and Freyne, P. (1983). Osteoarthrosis in runners with knee
Compared to women who did not get OA, those who developed incident pain. Br J Sports Med 17:84–7.
radiographic OA had weaker quadriceps 3 years earlier. No association of 87. Kujala, U.M., Kettunen, J., Paananen, H., et al. (1995). Knee osteoarthritis in
quadriceps weakness with incident OA was found in men. former runners, soccer players, weight lifters, and shooters. Arthritis Rheum
74. Jefferson, R.J., Collins, J.J., Whittle, M.W., Radin, E.L., and O’Connor, J.J. 38:539–46.
(1990). The role of the quadriceps in controlling impulsive forces around 88. Lane, N.E., Hochberg, M.C., Pressman, A., Scott, J.C., and Nevitt, M.C.
heel strike. Proc Instn Mech Engrs 204:21–8. (1999). Recreational physical activity and the risk of osteoarthritis of the hip
75. Chaisson, C.E., Zhang, Y., Sharma, L., Kannel, W., Felson, D.T. (1999). Grip in elderly women. J Rheumatol 26:849–540.
strength and the risk of developing radiographic hand osteoarthritis: results 89. Manninen, P., Riihimaki, H., Heliovaara, M., Suomalainen, O. (2001).
from the Framingham study. Arthritis Rheum 42:33–8. Physical exercise and risk of severe knee osteoarthritis requiring arthro-
76. Croft, P., Coggon, D., Cruddas, M., and Cooper, C. (1992). Osteoarthritis of plasty. Rheumatology 40:432–7.
the hip: an occupational disease in farmers. Br Med J 304:1269–72. 90. McAlindon, T.E., Wilson, W.F., Aliabadi, P., Weissman, B., Felson, D.T.
77. Kellgren, J.H. and Lawrence, J.S. (1952). Rheumatism in miners. II: X-ray (1999). Level of physical activity and the risk of radiographic and sympto-
study. Br J Ind Med 9:197–207. matic knee osteoarthritis in the elderly: the Framingham study. Am J Med
78. Lawrence, J.S. (1961). Rheumatism in cotton operatives. Br J Ind Med 106:151–7.
18:270–6. 91. Felson, D.T., Chaisson, C.E., Hill, C.L., et al. (2001). The association of
79. Hadler, N.M., Gillings, D.B., Imbus, R. et al. (1978). Hand structure and bone marrow lesions with pain in knee osteoarthritis. Ann Intern Med
function in an industrial setting. Arthritis Rheum 21:210–20. 134:541–9.
80. Anderson, J. and Felson, D.T. (1988). Factors associated with osteoarthritis 92. *Hill, C.L., Gale, D.G., Chaisson, C.E., et al. (2001) Knee effusions, popliteal
of the knee in the First National Health and Nutrition Examination Survey cysts and synovial thickening: association with knee pain in those with and
(NHANES I). Am J Epidemiol 128:179–89. without osteoarthritis. J Rheumatol 28:1330–7.
81. Kujala, U.M., Kettunen, J., Paananen, H., et al. (1995). Knee osteoarthritis in Compared to persons without knee pain, those with knee pain more often
former runners, soccer players, weight lifters and shooters. Arthritis Rheum had synovitis and effusions on their knee MRI’s. Prior reference (Felson et al.)
38:539–46. suggested that also those with knee pain more often had bone marrow edema
82. Felson, D.T., Hannan, M.T., Naimark, A., et al. (1991). Occupational phys- lesions. These two studies provide evidence that bone marrow edema, synovitis,
ical demands, knee bending, and knee osteoarthritis: results from the and effusions are the structural findings correlated with knee pain in OA.
Framingham Study. J Rheumatol 18:1587–92. 93. Oliveria, S.A., Felson, D.T., Reed, J.I., Cirillo, P.A., and Walker, A.M. (1995).
83. Cooper, C., McAlindon, T., Coggon, D., Egger, P., and Dieppe, P. (1994). Incidence of symptomatic hand, hip, and knee osteoarthritis among patients
Occupational activity and osteoarthritis of the knee. Ann Rheum Dis 53:90–3. in a health maintenance organization. Arthritis Rheum 38:1134–41.
84. Anderson, J., and Felson, D.T. (1988). Factors associated with osteoarthritis 94. Cooper, C., McAlindon, T.E., Goggon, D., Egger, P., and Dieppe, P. (1994).
of the knee in the First National Health and Nutrition Examination Survey Occupational activity and osteoarthritis of the knee. Ann Rheum Dis
(NHANES I). Am J Epidemiol 128:179–89. 53:90–3.
 3 The economics of osteoarthritis
Edward Yelin

Demographers and economists traditionally divide chronic diseases into Kingdom have recently been completed. Table 3.1 summarizes the results of
two groups: those of high prevalence and low-average impact, for example, these studies. In the first systematic study of the cost of illness, Rice10
chronic sinusitis; and those of low prevalence and high-average impact, for reported that in the United States of America in 1963, musculoskeletal con-
example, systemic lupus erythematosus.1 OA sits astride these two major ditions accounted for $4 billion in total costs, half of which was attributed
classifications: it is clearly a high prevalence condition, but it must also be to the direct costs of medical care and the other half to wage losses and
classified as one with at least a moderate, if not a high level of impact. More the imputed value of homemaker losses. This sum was equivalent to 0.7 per
importantly, because OA is associated with age,2 the aging of the population cent of the gross national product (GNP).
puts a higher proportion of the total population at risk for OA. Moreover, The relative magnitude of the total costs of musculoskeletal disease in the
the aging of the population puts a higher proportion of persons with OA in United States of America remained fairly stable through 1980, though the
the age groups in which the severity of the disease is greatest. This combin- proportion due to medical care and wage losses changed substantially.14,15
ation of high and growing prevalence, and moderate to severe impact, In the 1960s, real wages were rising rapidly, while medical care inflation had
makes OA an important condition in health policy concerns. not yet become a significant problem. Accordingly, the proportion of total
Interestingly, the condition has received relatively scant attention in costs due to lost wages rose between the 1963 and 1972 studies. In contrast,
the cost-of-illness literature. Instead, health services researchers from the after the oil shock of the early 1970s, real wages stagnated while medical
rheumatology community have focussed on the cost of rheumatoid arthri- care prices rose quickly. By 1980, direct costs of musculoskeletal disease
tis (RA) in clinical samples, probably because it is a more prominent con- accounted for more than 60 per cent of the total costs of this group of
dition in the practices of rheumatologists.3–9 On the other hand, in their illnesses. However, by 1988, total costs had increased to $124 billion, with
own studies, demographers and economists tend to focus on the more more than half associated with indirect costs.16
encompassing rubrics of musculoskeletal disease or all forms of arthritis, The most recent national study of the costs of musculoskeletal conditions
probably because the data sets they analyse do not include discrete diag- concerned 1995.17 In that year, the total costs of these illnesses amounted to
noses, such as OA. However, persons with OA constitute a large fraction $215 billion, or about 2.9 per cent of the GNP. Less than half of the total was
of all persons with musculoskeletal conditions and, therefore, of the costs due to direct costs of medical care; the balance was due to wage losses. In the
of these diseases. same study, the cost of arthritis alone was estimated to be $83 billion, with
This chapter reviews the studies of the economic impact of muscu- about three-quarters of that total due to lost wages.
loskeletal disease and all forms of arthritis, in general, conducted on random These studies of the national economic impact of musculoskeletal condi-
samples of the population; summarizes the small literature on the cost of tions in the United States of America suggest a large increase in costs
OA, in particular, based on clinical samples, including those conducted in occurred between 1980 and 1988, differences among the studies in methods
managed care environments in the United States of America; compares notwithstanding.18
results from clinical studies of OA and RA; and then uses the 1979–81 The relative magnitude of the costs of musculoskeletal conditions in
National Health Interview Survey to make preliminary estimates of the Canada, Australia, and the United Kingdom would appear to be similar to
absolute national cost of arthritis and of the increment in costs experienced that in the United States of America.19–22 For example, in 1986 the total
by persons with arthritis compared to those of similar age, sex, and race costs of musculoskeletal conditions in Canada was approximately 1.7 per
without arthritis. The pharmacoeconomics of treatment of OA are discussed cent of the GNP, or about half-way between the estimates for the United
specifically in Chapter 8. States of America for 1988 of the costs of all forms of arthritis alone and all
forms of musculoskeletal disease. In addition to the studies of all forms of
musculoskeletal conditions in various nations, a study has been published
Cost-of-illness studies concerning the costs of OA, in particular, in France; total costs of OA were
approximately 6 billion francs per year.22
The costs of illness are divided into two distinct spheres, those due to direct
expenditures for medical care services and those due to the indirect impact
of illness on function, principally measured by lost wages due to reduced Costs of osteoarthritis from studies using
work effort or total cessation of work activities.10–13
clinical samples
National studies of the cost of Only four studies of the costs of osteoarthritis have been derived from clin-
ical samples and in only one of these was indirect cost due to lost wages
musculoskeletal diseases assessed. Table 3.2 summarizes the result of these studies, expressing all costs
Five studies of the cost of musculoskeletal diseases in the United States of in 1999 terms. The study by Liang et al.25 derives from a random sample of
America have been published, the two most recent of which presented those who had ever attended a tertiary care facility. They reported that physi-
the costs of arthritis separately. In addition, studies of the economic impact cian visits accounted for only a small portion of direct costs of medical care;
of musculoskeletal disease in Canada, Australia, France, and the United the majority of direct costs were due to a handful of hospital admissions and
18 3    

Table 3.1 Cost of musculoskeletal diseases in the current US, Canadian, and Australian dollars, French francs, and British pounds and as a % of
Gross National Product (GNP)

Year Direct ($ billion) Indirect ($ billion) Total ($ billion) % of GNP

All Musculoskeletal conditions-US
1963 2 2 4 0.7
1972 4 5 9 0.7
1980 13 8 21 0.8
1988 60 64 124 2.5
1995 89 126 215 2.9
All Forms of Arthritis-US
1988 13 42 55 1.2
1995 22 61 83 1.1

Country Year Currency Direct Indirect Total % of GNP

All Forms of Musculoskeletal Condtions: Non-US
Canada 1986 dollars-billions 2 6 8 1.7
Australia 1994 dollars-billions 1 4 5 1.3
United Kingdom 1986 pounds-billions 1 3 4 1.1
Osteoarthritis: Non-US
France 1991 francs-billions 4 2 6 n/a

Source: US, Refs 10, 14–17; Canada, Ref. 19; Australia, Ref. 20; United Kingdom, Ref. 21; France, Ref. 22. Adopted from Ref. 23.

Table 3.2 Costs of osteoarthritis, in 1999 dollars, from four studies using clinical samples

Study Year Direct costs Indirect Total

Costs Costs
Physician Hospital Other Total
Liang et al. 1984 183 628 958 1769 12 032 13 801
Holman et al. 1988 1294 846 2157
Gabriel et al. 1995 2427
Lanes et al. 1997 135 280 196 611

Notes: All costs have been expressed in 1999 terms by inflating study year costs by the change in the Consumer Price Index (24, p. 487). In the study by Liang et al.,25 indirect costs were estimated
jointly for persons with OA and RA; this amount was applied here for the persons with OA. In the study by Holman et al.,26 neither ‘other’ costs nor indirect costs were reported. The study by
Gabriel et al.27 reported the frequency of changes in employment, but not the associated indirect costs. Lanes et al.28 reported costs incurred by a member of a health maintenance organization.

to the category ‘other’, which include prescription and non-prescription a true population-based study, with the caveat that those who have OA
drugs and medical devices. Indirect costs of OA dwarfed medical costs, aver- but have not received a diagnosis would not appear in the database. In
aging $12 032 in 1999 terms among all persons with OA. this study, the authors did not report direct medical costs by category.
The study by Holman et al.26 was limited to costs of physician visits and Moreover, they accounted only for costs that generated medical bills.
hospital admissions. It was conducted in Northern California and included Nonetheless, their estimate of overall annual direct medical care costs for
respondents receiving care in the fee-for-service sector, in a pre-paid group OA—$2427 per person in 1999 terms—exceeds the estimates from the two
practice, and in an experimental center designed to lower utilization rates. previous studies. Although Gabriel et al. did not estimate the wage losses
The authors reported that the magnitude of the costs of hospitalization in associated with OA, they noted that about 11 per cent of persons with OA
their study was similar to that in the study by Liang et al., but they also reduced their hours of work, 9 per cent reported being unable to get a job,
reported much higher costs due to physician visits, and overall direct med- and 14 per cent retired early due to this condition.
ical care costs more than 20 per cent higher, even though the costs of drugs In another study using a clinical sample, Lanes et al.28 analysed costs
and devices were not estimated. Interestingly, costs were lower in the experi- among members of a single health maintenance organization, and found
mental setting than in the pre-paid group practice, and lower in the pre- costs to be considerably lower than other studies (direct costs totalled only
paid group practice than under fee-for-service. This suggests that medical $611 in 1999 terms), perhaps indicating lower costs in that form of man-
care costs of OA can be reduced through both financial incentives to aged-care organization than in other systems of care.
providers and patients and through non-monetary incentives in a practice
designed to reduce utilization through more interactive relationships
between patients and physicians. Comparison of the costs of OA and RA
Gabriel et al.27 assembled a database of all persons with a diagnosis of OA Table 3.3 compares the per-case costs of OA and RA by averaging values
among residents of Olmstead County, Minnesota. This sample approximates from each of the studies using clinical samples (two values are provided for
 3     19

Table 3.3 Costs, in 1999 dollars, of rheumatoid arthritis and osteoarthritis from average of clinical studies

Study Physician Hospital Other Total Direct Costs Indirect Costs Total
OA-including Lanes et al. 537 585 577 1699 12 032 13 731
OA-excluding Lanes et al. 738 737 958 2433 12 032 14 465
RA 1082 3361 1397 5840 18 907 24 747

Notes: All costs have been expressed in 1999 terms by inflating study year costs by the change in the Consumer Price Index (32, p. 487). The first estimate of the costs of OA was obtained by
averaging the four studies summarized in Table 3.2, or, where only one study provided as estimate, using that one value. The second estimate of the costs of OA excludes the value from the
study of Lanes et al.28 because the latter study is an outlier among the four studies. The costs of RA were estimated in a fashion similar to that for the OA estimates, using the studies of Meenan
et al.3 and Lubeck et al.4

OA, including and excluding the study by Lanes et al. because the costs in Table 3.4 Per capita and total utilization of persons reporting arthritis,
that study appear to be systematically lower than those in the other three US, 1989–91
studies). Not surprisingly, the per-case costs of RA exceed the costs of OA
in every category, with the costs of hospital admissions accounting for Type of utilization Mean Total
the largest relative difference. Overall, the per-case cost of RA is more than (millions)
1.7 times as great as the per-case cost of OA, even if the results of Lanes et al. Visits to physicians 7.34 226.3
are excluded. Nevertheless, because the prevalence of OA is so great, OA has
Hospital admissions 0.22 6.9
a far greater overall impact on the economy than RA. Assuming a liberal
estimate of the prevalence of RA of 1.0 per cent of the total population, a Hospital days 1.58 48.8
conservative estimate of the prevalence of OA of 4.2 per cent of males and Length of admissions (in days) 7.07
9.0 per cent of females 20 years of age or over;2 estimates of the per-case cost
Source and note: Author’s analysis of 1989–91 National Health Interview Surveys.
of OA, including those from Lanes et al.; and the US population in 1999,24 Estimates based on averaging across the three surveys.
there are 2.73 million persons with RA and 13.03 million with OA in the
nation. Accordingly, the total costs of OA, at $178.9 billion, are about
2.65 times as large as the total costs of RA, at $67.5 billion. Even using the
highest published estimates for the prevalence of RA—2 per cent,2 the These hospital admissions totaled 48.8 million days; the average length of
national cost of OA would still exceed the national cost of RA by about stay was 7.07 days per admission.
50 per cent. It should be noted that the combined cost of OA and RA— Table 3.5 provides estimates of the economic impact of this medical care
about $246 billion—is far larger than the most recent estimate of the utilization and of the wage losses associated with arthritis. These estimates
cost of all forms of arthritis from a US national study in 1995,18 $83 billion, were made by multiplying the number of units of physician visits used
or about $91 billion in 1999 terms (Table 3.1). The higher estimate from the by $90 (the approximate average of the costs of a physician visit in
clinical samples may be due to the greater severity of the disease among per- the nation in 1999), and the number of hospital admissions by $8220 (the
sons sampled in clinical environments. Alternatively, the difference may be average cost of all hospital stays in the United States of America in that
due to the greater level of detail in the enumeration of costs in the clinical year). Using these unit prices, all persons self-reporting arthritis incurred
studies. $2.0 billion in costs due to physician visits and $56.7 billion in costs due to
hospital admissions in the years 1989–91. Because many of the hospital
National community-based estimates of admissions for arthritis include surgical procedures, including total joint
replacement surgery, and surgical admissions are more expensive than this
the impact of arthritis average, the ‘true’ cost of the admissions probably exceeds the $56.7 billion
In studies using clinical samples, researchers are able to customize data col- estimate. Even without adjusting the unit price of hospital admissions to
lection to ensure relatively complete enumeration of the impacts of illness. take surgery into account, the annual cost of medical care for arthritis is at
However, the sample of persons with the illness is biased, either because least $58.7 billion. Including the costs of drugs and devices, which are not
those who receive a diagnosis may have inherently better access to care, or enumerated in the HIS, would substantially increase this total.
because they may have more severe illness than those who do not. In addition, 14.1 per cent fewer of those with arthritis are in the labor
In part, to avoid the bias from sampling in clinical environments, the force than persons without arthritis. Thus, 2.34 million persons with
federal government in the United States of America instituted an annual arthritis were not working who would have been working if they had the
community-based survey of the health of the population—the National same labor force participation rates as persons without arthritis. If a unit
Health Interview (HIS).29 In the HIS, individuals self-report their symp- price for wage losses, based on average earnings among all US workers in
toms and, unless a physician has told the individual a specific diagnosis 1999, is applied, the indirect costs of arthritis total at least $66.8 billion.
associated with the symptoms, a general diagnostic code, such as arthritis, This figure omits all losses due to reduction in hours worked or the lesser
is given. The analysis reported below provides estimates of the impact of degree of career advancement experienced by persons with arthritis in
all forms of self-reported arthritis for the years 1989–91 (using three years comparison with those without arthritis.
reduces the sampling variability associated with any one year’s data). OA Even though the estimate of the direct costs omits common expenses
accounts for most of the subjects in this diagnostic classification.30 for arthritis, and the estimate of indirect costs omits partial work disability,
Over the period 1989–91, an average of 30.8 million individuals report- the estimate of the total cost of arthritis—$125.5 billion—represents
ed symptoms consistent with arthritis, providing an overall prevalence rate 1.4 per cent of the gross domestic product for the United States of America
of 12.3 per cent, that is, roughly twice the rate that has been estimated from for the year 1999.24
epidemiological studies of OA based on physician examination.2 Table 3.4 Table 3.5 provides estimates of the total costs incurred by persons with
summarizes the average and total health care utilization of these individu- arthritis. Table 3.6, in contrast, shows the increment in costs experienced by
als self-reporting arthritis. Overall, they made a mean of 7.34 visits to persons with arthritis relative to the remainder of the population and
physicians in the year prior to the interview, or 226.3 million visits overall. assuming persons with arthritis have the same age, race, and sex distribu-
In addition, they experienced an average of 0.22 hospital admissions per tion as those without arthritis. Persons with arthritis make slightly less than
person, amounting to 6.9 million admissions in the nation as a whole. one more visit per physician year than similar persons without arthritis.
20 3    

Table 3.5 Estimates of the absolute national costs of arthritis in 1999, US, 1989–91

Direct costs
Physician visits Hospital admissions Total
Number Unit price Total Number Unit price* Total
226.3 mil $90 $2.0 bil 6.9 mil $8220 $56.7 bil $58.7 bil
Indirect costs
Number leaving work† Unit price‡ Total Total costs
2.34 mil $28,548 $66.8 bil $125.5 bil

Source and notes: Author’s analysis of 1989–91 National Health Interview Surveys.
* Unit price of hospital admission based on average cost of patient stay in US short-term hospitals in 1992 (33, p. 127), with the value inflated to 1999 terms using the Medical Care
Component of the Consumer Price Index (24, p. 487).
† Number of persons who left work based on difference in age, sex, and race matched labor force participation rates of persons without arthritis and actual labor force participation rate of
persons with arthritis (14.1%).
‡ Unit price of cessation of work based on the US median weekly wage times 52 weeks (24, p. 437).

Table 3.6 Estimates of the incremental national costs of arthritis in 1994, US, 1989–91

Direct costs
Physician visits Hospital admissions Total
Incremental# Unit price Total Incremental* Unit price# Total
0.98 mil $90 $0.1 bil 0.06 mil $8220 $0.5 bil $0.6 bil

Indirect costs
Incremental# Leaving work† Unit price‡ Total Total costs
1.00 mil $28,548 $28.6 bil $29.2 bil

Source and notes: Author’s analysis of 1989–91 National Health Interview Surveys.
* Incremental number of physician visits and hospital admissions per person based on regressions estimating the utilization among persons with OA assuming the age, sex, and race distribu-
tion in the remainder of the population.
# Unit price of hospital admission based on average cost of patient stay in US short-term hospitals in 1992 (31, p. 127), with this value inflated to 1999 terms using the Medical Care
Component of the Consumer Price Index (24, p. 487).
† Number of persons who left work based on difference in labor force participation of persons without arthritis, and the estimated labor force participation rate of persons with arthritis and
the same distribution of age, sex, and race as persons without arthritis (6.0%).
‡ Unit price of cessation of work based on the US median weekly wage times 52 weeks (24, p. 437).

Accordingly, the incremental costs of their physician visits are relatively account for a large proportion of the increment in costs attributable to
small—$0.1 billion. The incremental cost of their hospitalizations is much hospitalization.32 To an even greater extent, policymakers would do well
larger: based on an excess of 0.06 admissions per capita and a unit cost of to emphasize the impact of arthritis on work, since wage losses account for
$8220 per admission, the increment in the costs of the hospital admissions almost two-thirds of the incremental cost of arthritis.
of persons with arthritis is $0.5 billion. Thus, of the total incremental med-
ical care costs of $0.6 billion incurred by persons with arthritis, more than
80 per cent is due to excess hospital admissions.
However, the incremental costs of wage losses dwarf even those due to Summary and conclusions
hospitalization. In the period 1989–91, an extra 1.0 million persons with In contrast to RA, the cost of OA has been the subject of few studies. From
arthritis were out of work, relative to the expected labor force participation these studies, the annual cost of OA would appear to be in the range of from
rate of persons with the same age, sex, and race characteristics but without $13 000–15 000 per case, with most of the costs due to lost wages. Though its
arthritis. After multiplying this number by the average wages of US work- per-case cost is lower than that of RA, because of the much greater preval-
ers, the increment in wage losses totals $28.6 billion. All told, the increment ence of OA, the overall economic impact of OA is much greater. However, it
in wage losses of persons with arthritis relative to those without is far larg- is very difficult to provide a good estimate of the exact magnitude of the eco-
er than the increment in direct medical care costs. nomic cost of OA for the nation. In self-report studies, such as the HIS, the
Policymakers concerned about the medical care costs of arthritis would prevalence of OA is under-reported; most OA is classified under the more
do well to focus on reducing hospital admissions, since excess admissions are encompassing rubric of ‘all forms of arthritis’. Indeed, a diagnosis of OA
responsible for the bulk of the increment in medical care costs. Although occurs relatively infrequently in the clinical environment, because physicians
one cannot easily attribute surgical procedures to particular conditions often do not differentiate among musculoskeletal complaints in their treat-
in the HIS, we know from other sources of data that surgical admissions ment plans. As shown above, it can be estimated that the economic impact of
 3     21

this more encompassing rubric is $125.5 billion, and that people with arthri- 11. Walker, K. and Gauger, W. (1970, revised 1980). The dollar value of house-
tis incur incremental costs of about $29.2 billion, mostly due to lost wages. hold work. In Information Bulletin, No. 60. Ithaca, New York: New York State
The estimates of the per-case cost and of the national impact must both be College of Human Ecology.
viewed as preliminary, however, in the former case because the clinical stud- 12. Lubeck, D. and Yelin, E. (1988). A question of value: measuring the impact
ies are few and not as systematic as studies of the cost of RA and in the of chronic disease. Milbank Quart 66:445–64.
latter case because of the lack of differentiation of OA from other forms of 13. Reisine, S., Goodenow, C., and Grady, K. (1987). The impact of rheumatoid
musculoskeletal disease in the relevant databases. arthritis on the Homemaker. Soc Sci Med 25:89–95.
However, to a certain extent these shortcomings in the literature reflect a 14. Cooper, B. and Rice, D. (1976). The economic cost of illness revisited. Soc
shortcoming of concern about OA itself. Even these preliminary findings Sec Bull 39:21–35.
should indicate to policymakers that the combination of high prevalence, 15. Rice, D., Hodgson, T., and Kopstein, A. (1985). The economic costs of ill-
and moderate to high impact, combine to make OA a costly illness. It is ness: a replication and update. Health Care Finance Rev 7:61–80.
time to conduct more systematic studies on the economic impact of OA, so 16. *Rice, D. (1992). Cost of musculoskeletal conditions. In A. Pramer,
that we can plan adequately for the pandemic of OA we face with the aging S. Furner, and D. Rice (eds.) Musculoskeletal Conditions in the US. Chicago:
of the population. American Academy of Orthopedics, pp. 143–70.
This article documents the high toll that all forms of musculoskeletal disease,
but especially arthritis, played in the national economy in the early 1990s.
Key points Using methods from prior studies, the author shows that the impact of these
conditions has grown dramatically because of the aging of the population.
1. Musculoskeletal conditions account for as much as 1.1–2.9 per cent 17. Rice, D. (1999). Cost of musculoskeletal conditions. In A. Pramer, S. Furner,
of the GNP of nations with advanced economies, including the and D. Rice (eds.) Musculoskeletal conditions in the US (2nd ed.), Chigaco:
direct costs associated with medical care and the indirect costs due American Academy of Orthopedics, pp. 141–62.
to lost function and work disability. 18. *Yelin, E. and Callahan, L. (1995). The economic cost and social and psy-
chological impact of musculoskeletal conditions. Arthritis Rheum
2. All forms of arthritis account for as much as 1.2 per cent of the GDP
38:1351–62.
of the United States of America.
This article, written by members of the National Arthritis Data Task Force, is
3. The direct cost of OA averages between $1699 and 2433, while indi- a comprehensive review of the literature on the economic costs and psychologi-
rect costs are $12 032, in 1999 terms. cal impact of musculoskeletal conditions. It can be used to access the key articles
4. Even though the per-case cost of RA is higher than that of OA because in the cost of arthritis literature.
of the much higher prevalence of OA, the total costs of OA—$178.9 19. Badley, E. (1995). The economic burden of musculoskeletal disorders in
billion in 1999 terms—exceed those of RA by at least 50 per cent. Canada is similar to that for cancer, and may be higher. J Rheumatol 22:204–6.
20. Arthritis Foundation of Australia. (1994). In Industry Commission Report.
5. Most of the increment in costs of arthritis beyond that expected of Cost of Arthritis to the Australian Community, pp. 14–22.
persons of similar age and gender is due to work loss, not to expen- 21. Freedman, D. (1989). Arthritis: the painful challenge, Searle Social Research
ditures for medical care. Fellowship Report.
22. Levy, E., Ferme, A., Perocheau, D., and Bono, I. (1993). Les couts socio-
economiques de l’arthrose en France. Rev Rheum 60:63s–67s.
References 23. March, L. and Bachmeier, C. (1997). Economics of osteoarthritis: a global
(An asterisk denotes recommended reading.) perspective. Bailliere’s Clin Rheum 11:817–34.
1. *Verbrugge, L. and Patrick, D. (1995). Seven chronic conditions: their 24. USGPO. (2000). Statistical Abstract of the US, 2000, p. 13, 127, 437, 487.
impact on US adults’ activity levels and use of medical services. Am J Public 25. Liang, M., Larson, M., Thompson, M., Eaton, H., McNamera, E., and
Health 85:173–82. Katz, R., et al. (1984). Costs and outcomes in rheumatoid and osteoarthritis.
This article shows the differential impact of various chronic conditions on Arthritis Rheum 27:522–9.
disability and access to medical care. It demonstrates that arthritis is among the
26. Holman, H., Lubeck, D., Dutton, D., and Brown, B. (1988). Improving
leading causes of disability, starting in middle age, but is not among the leading
health service performance by modifying medical practices. Trans Assoc Am
reasons for physician visits or hospital admissions in any age group.
Phys 101:173–9.
2. Silman, A. and Hochberg, M. (1993). In Epidemiology of the rheumatic
27. *Gabriel, S., Crowson, C., and O’Fallon, W. (1995). Costs of osteoarthritis:
diseases. Oxford: Oxford University Press, pp. 257–88.
estimates from a geographically defined population. J Rheumatol 22(Suppl.
3. Meenan, R., Yelin, E., Henke, C., Curtis, D., and Epstein, W. (1978). The 43):23–5.
costs of rheumatoid arthritis. Arthritis Rheum 21:827–33. This article is unique among studies of the costs of arthritis because of the
4. Lubeck, D., Spitz, P., Fries, J., Wolfe, F., Mitchell, D., and Roth, S. (1986). combination of the community-based sampling frame and certainty of diagnosis
A multicenter study of annual health services utilization and costs in rheuma- provided by a database of all medical care encounters in the area surrounding
toid arthritis. Arthritis Rheum 29:488–93. Rochester, Minnesota.
5. Thompson, M., Read, J., and Liang, M. (1984). Feasibility of willingness-to- 28. *Lanes, S., Lanza, L., Randensky, P., Yood, R., Meenan, R., Walker, A., and
pay measurements in chronic arthritis. Med Decis Making 4:195–212. Dreyer, N. (1997). Resource utilization and cost of care for rheumatoid
6. Stone, C. (1984). The lifetime costs of rheumatoid arthritis. J Rheumatol arthritis and osteoarthritis in a managed care setting. Arthritis Rheum
11:819–27. 40:1475–81.
7. Pugner, K., Scott, D., Holmes, J., and Hieke, K. (2000). The costs of rheuma- This article provided the first indication of the extent of the medical care
toid arthritis: an international long-term view. Sem Arthritis Rheum utilization and economic costs for RA in a closed-system sample, that is, in a
29:305–20. managed care organization.
8. Albers, J., Kuper, H., van Riel, P., Prevoo, M., Van’t Hof, M., van Gestel, A., 29. Kovar, M. and Poe, G. (1985). National Health Interview Survey design,
and Severens, J. (1999). Socio-economic consequences of rheumatoid 1973–84 and procedures, 1975–83. Vital Health Stat 18 (Series 1):1–130.
arthritis in the first years of the disease. Rheumatology 38:423–30. 30. LaPlante, M. (1988). Data on disability from the National Health Interview
9. Merkesdal, S., Ruof, J., Schoofski, O., Bernitt, K., Zeidler, H., and Mau, W. Survey, National Institute on Disability and Rehabilitation Research,
(2001). Indirect costs in early rheumatoid arthritis. Arthritis Rheum Washington, DC.
44:528–34. 31. USGPO (1994). Statistical Abstract of the US, 1994, p. 127.
10. Rice, D. (1966). Estimating the cost of illness. Health Econ Ser, No. 6, 32. Felts, W. and Yelin, E. (1989). The economic impact of the rheumatic
National Center for Health Statistics. diseases in the United States. J Rheumatol 16:867–84.
This page intentionally left blank
 4
Genetic aspects of
osteoarthritis

4.1 Evidence for the inheritance of osteoarthritis 25

Nisha J. Manek and Tim D. Spector
4.2 Specific gene defects associated with osteoarthritis 33
Thorvaldur Ingvarsson and Stefan Einar Stefansson
4.3 Laboratory approaches to the identification of genetic association
in osteoarthritis 41
Virginia B. Kraus and Eden R. Martin
This page intentionally left blank
 4.1 Evidence for the inheritance of
osteoarthritis
Nisha J. Manek and Tim D. Spector

Genetic studies are under way for many complex traits, spurred by the hereditary factors. Other earlier clinical studies by Kellgren and associates6
improvement in genetic tools and resources during the past decade. The and Lawrence and Moore7 on generalized OA and other OA subtypes, sug-
mechanism by which DNA sequence variations or alleles in genes con- gested that specific forms of common OA cluster within families. A retro-
tribute to differences within the human population with respect to the spective radiographic analysis of hip OA by Lindberg provided further
occurrence of OA is unknown. support for this observation.8 In this study, siblings of patients who had
Epidemiological studies have provided evidence to support the long- undergone hip replacement were noted to have twice the incidence of hip
standing clinical perspective that environmental factors predominate in the OA than that of age- and sex-matched control subjects. Chitnavis et al.9
etiology of OA. Important risk factors affecting disease status include age, have estimated the relative risks of advanced, symptomatic OA of the hip
occupation, body mass index, trauma, and joint deformity.1 Conversely, and knee in siblings of probands having total joint replacement for OA.
genetic liability is traditionally seen to predominate in only a small minor- Spouse-based controls were used to correct for environmental influences. A
ity of the OA population. Families that transmit OA as a simple Mendelian sibling relative risk of 1.9 for total hip replacement and 4.8 for total knee
genetic trait have been described but are rare. Partitioning OA into these replacement was calculated. The overall combined relative risk for joint
two conceptual groups, that is, strongly genetic or strongly environmental, replacement was 2.3. A sibling study by Lanyon and colleagues10 showed
neglects the majority of the population in which the interplay between that siblings of patients with hip OA of sufficient severity to warrant total
genes and environment in the occurrence of ‘common’ OA remains hip replacement have a high risk of radiographic hip OA. Each study,
significant. despite using different approaches, demonstrates that relatives of affected
To study sources of individual differences (i.e., the variance) in a pheno- individuals have higher rates of OA than the general population. Table 4.1
type such as OA, genetically related subjects are required and methods of summarizes these trials. Clustering of OA within families is suggestive of a
assessing genetic influence in common OA include estimations of relative genetic contribution to common OA, but they do not provide conclusive
risk such as sibling recurrence risk (s). This is an estimate of familiality or evidence. Clearly, familial aggregation of disease might also be explained by
the risk of disease in first-degree relatives of the patients affected with OA non-genetic factors owing to siblings sharing important environmental
(see practice point 1) and depends on the choice of controls and prevalence influences. Therefore, an alternative explanation for familial aggregation of
in the general population. Estimation of genetic relative risk in family stud- disease is that members within a family and members within a racial group
ies, however, does not provide information about the number of genes that share similar environmental influences.
contribute to this risk or the strength of each gene’s contribution.
A second method of assessing the genetic influence in OA is the total
heritability (h2) estimate. This is usually defined as the proportion of
population variance explained by genetic factors.2 It can be calculated by Twin pair studies:
comparing the liability to OA in first-degree relatives to that in more dis-
tant relatives, spouses, or unrelated individuals. Twins are a useful group heritability estimates for OA
matched for upbringing and age, where identical or monozygotic (MZ) The publication in 1996 of the first large-scale OA twin study by Spector
twins share 100 per cent of their genes and non-identical or dizygotic (DZ) et al.11 gave compelling support that the familial clustering was due to a
twins share 50 per cent of their genes. It is assumed that both twins have major genetic contribution to common OA. Using a cohort of 130 MZ and
roughly the same family environment, and that therefore any greater 120 DZ female twin pairs in the age group 48–70 years, a higher intra-class
similarity between the MZ and DZ twins is due to genetic influences.3 correlation among MZ twins compared with DZ twins was observed for
As well as estimating traditional heritability without many of the problems several clinical and radiographic features of OA. Significant differences in
of family studies, comparison of the covariance (or correlation) in OA OA concordance between the MZ and DZ twin pairs remained after control
between the MZ and DZ twins allows separation of the observed pheno- of environmental contributors such as weight, smoking, and oestrogen
typic variance into additive and dominant genetic components and into replacement. The study revealed an h2 between 39–65 per cent, with a con-
common and unique environmental components. This is reviewed in detail cordance rate in the MZ twin pairs of 0.64 compared with 0.38 in the DZ
elsewhere.4 pairs. Incomplete concordance in the MZ twin pairs clearly demonstrated
This chapter reviews evidence of the extent of genetic variation in both an environmental component to disease expression. A twin study focusing
common and rare forms of OA. on radiographic hip OA in female twins was carried out by the same
group.12 An h2 of approximately 50 per cent was determined for disease
defined by joint-space narrowing at the hip joint. A questionnaire based
Family studies: establishing familial twin study from Finland included both male (577 MZ and 1180 DZ) and
female (836 MZ and 1502 DZ) twin pairs.13 Recalled OA at any joint was
risk of OA used as the criterion for disease and an h2 of 44 per cent was obtained for
As early as 60 years ago, in 1941, Stecher5 observed the familial nature of women. Interestingly, this Finnish study detected no genetic component to
idiopathic Heberden’s nodes, which he concluded was explained best by male disease, with a concordance of 0.34 in MZ male twin pairs and 0.38 in
26 4    

Table 4.1 Family studies of OA with relative risk estimations

Study Proband Relatives Controls Relative risk

Stecher 5 Women with idiopathic Heberden’s nodes Sisters Women in the population 3-fold for sisters
mothers 2-fold for mothers
Kellgren et al.6 Men and women with primary generalized OA Siblings Age- and sex-matched controls 2-fold for siblings
Lindberg8 Men and women with hip OA Siblings Age- and sex-matched controls from 2-fold for siblings
the same hospital
Chitnavis et al.9 Men and women with hip and knee arthroplasty Siblings Spouses 2-fold for siblings
Lanyon et al.10 Men and women with severe hip OA Siblings Age- and sex-matched controls 7-fold for siblings

Table 4.2 Heritability (h2) of OA

Author Study group Controls h2

Twin studies
Spector et al.11 130 female MZ twin pairs 120 female DZ twin pairs 0.39 for knee OA
0.65 for hand OA
MacGregor et al.12 135 female MZ twin pairs 277 female DZ twin pairs 0.64 for hip JSN
Hip radiographic OA Hip radiographic OA 0.58 for OA overall (radiographic)
Kaprio et al.13 577 male MZ twin pairs 1180 male DZ pairs 0.44 for OA, any site in female twin
836 female MZ twin pairs 1502 female DZ pairs pairs only. No evidence of h2 in
male twin pairs
Population studies
Felson et al.16 337 nuclear families from the Spouses 0.42 for generalized OA
Framingham Heart Study ascertained
independently of OA status
Bijkerk et al.14 118 Probands with multiple affected 1587 indivduals randomly selected 0.75 for disk degeneration
joint sites and 257 of their siblings from the Rotterdam study of aging 0.56 for hand OA
Hirsch et al.15 167 families in the Baltimore Longitudinal Regression analysis of OA presence, Increased correlation identified
Study of Aging absence and severity in sibs and for hand and polyarticular OA:
offspring r  0.33  0.81, no direct h2 estimate
Chitnavis et al.9 402 probands with TKR or THR for Prevalence of TKR and THR for 0.27 for THR
idiopathic OA idiopathic OA in 1171 siblings and 0.31 for THR or TKR [all relatives
376 spouses included]

MZ, monozygotic (identical) twins; DZ, dizygotic (non-identical) twins; TKR, total knee replacement; THR, total hip replacement; JSN, joint-space narrowing.

DZ male twin pairs. This result suggests the possibility that genes play a To understand the mechanism by which OA genetic susceptibility is
more significant role in the development of OA in females. transmitted, Felson et al.16 conducted a segregation analysis of 337 nuclear
families and their adult offspring from the Framingham study. Parent–
offspring and sibling–sibling correlation were determined for primary
Population studies: OA of hand and knee radiographs and were in the range 0.115–0.306. In
segregation analyses, the best-fitting models were mixed models with a
heritability estimates Mendelian mode of inheritance and a residual multi-factorial component
representing either polygenic or environmental factors. The Mendelian
Other population-based studies also support a genetic contribution to
recessive model provided the best fit, although in a subsequent reanalysis
common OA risk (see Table 4.2). Bijkerk and co-workers14 compared the
Spector et al. found that other non-Mendelian models also fitted the data
joint specific frequency of OA between the probands and their siblings
equally.17
enrolled in the Rotterdam study and noted a high level of familial correla-
tion for hand OA (h2 of 0.56), but no correlation for hip or knee OA. The h2
for a score that summed the number of joints affected in the knees,
hips, hands, and spine was 0.78. Hirsch et al.15 studied the familial aggrega- Hereditary disorders causing
tion of OA by determining siblings’ correlations for the disease in a
cohort of patients collected by the Baltimore Longitudinal Study on Ageing.
premature OA
Increased correlation for hand and polyarticular OA demonstrated clear Osteochondrodysplasias, the developmental defects affecting cartilage
familial aggregation of OA. and bone, comprise a large and heterogeneous group of disorders. These
 4.1       27

heritable disorders predispose to OA by affecting joint shape, mobility, and Table 4.3 Candidate gene studies in OA
matrix composition. To date the genetic bases for only a few of the over 150
types have been identified.18 An online history of skeletal dysplasias and Gene Number Association result* Author(s)
their responsible genes can be found at www.csmc.edu/genetics/skeldys. COL2A1 n  86  Hull and Pope, 1989
The majority of these Mendelian genetic disorders can be diagnosed COL2A1 n  91 NS Vikkula, 1993
before the onset of adult OA symptoms on the basis of clinical and radio-
logical examinations during childhood and young adulthood (see practice COL2A1 n  21 pairs NS Priestly, 1991
point 2). A proper diagnosis is of practical importance with respect to COL2A1 n  38 pairs NS Loughlin, 1994
genetic and occupational counselling. The skeletal dysplasias are discussed CRTL1 n  38 pairs NS Loughlin, 1994
more fully in the chapter on specific gene defects associated with OA (see
Chapter 4.2). However, even in aggregate, the percentage of common OA CRTM n  38 pairs NS Loughlin, 1994
due to well-characterized genetic syndromes will be low (1 per cent) CRTM n  261  Muelenbelt, 1997
among the osteoarthritic population. The study of heritable arthropathies 2q n  99 pairs  Wright, 1996
not only illustrates the significant progress that has been made in the under-
VDR n  351  Keen, 1997
standing of the genetics of these diseases, but also highlights some import-
ant observations that are pertinent in considering the genetic variation VDR n  846  Uitterlinden, 1997
associated with common OA. These include: *NS: not significant.
1. Locus heterogeneity whereby similar clinical phenotypes can result
from mutations in different genes. An example is multiple epiphyseal loci in affected sibling pairs. They tested for linkage of the COL2A1 locus in 48
dysplasia (MED) (see practice point 3). sibling pairs with generalized OA (primary OA in 3 joint groups) but found no
2. Different mutations within the same gene can cause different pheno- evidence of linkage. Two other genes encoding secreted proteins in cartilage
types. For example, different mutations in the COL2A1 gene are asso- link protein (CRTL1) and cartilage matrix protein (CRTM) were also evalu-
ciated with the Stickler syndrome, neonatal achondrogenesis II/ ated for linkage, but neither gene could be substantially associated with OA
hypochondrogenesis, Kniest dysplasia as well as SED. risk. However, the sample size was too small and lacked power. Mustafa et al.32
examined a number of candidates using a cohort of 481 affected sibling pairs
3. Different mutations within the same gene can cause identical phenotypes. ascertained by joint replacement surgery for OA (hip, knee, or hip and knee).
For example, at least 20 different mutations within the COL2A1 gene on Suggestive linkage was obtained to the COL9A1 gene in female affected pairs
chromosome 12 have been associated with the Stickler syndrome.19,20 who were concordant for hip OA, with a LOD score of 2.3.
Meulenbelt et al.33 reported a large Dutch family that has primary OA of
the knees, hands, and the spine, in four generations with disease onset in
Identifying disease genes in the third and fifth decades. The transmittance of OA in this family appears
to be autosomal dominant and ten candidate genes were excluded as the
heritable OA mutant locus, including several that encode structural proteins of the car-
Linkage analysis. The first step for establishing the involvement of a candid- tilage extra-cellular matrix. A genome-wide scan has now linked the OA in
ate gene in heritable OA is the performance of genetic linkage analyses this family to micro-satellite markers on chromosome 2q.34 Roby et al.35
of large kindred populations. Linkage analysis has been utilized widely also excluded several candidate genes in a South African family of Dutch
for linkage studies of type II collagen gene COL2A1 mutations to heritable origin in which severe early-onset hip OA segregates as an autosomal dom-
cartilage diseases. The OA in these pedigrees is best described as secondary inant trait. The group subsequently mapped the disease to chromosome
and related to chondrodysplasia and not the common primary form. 4q35 (LOD score of 5.7).
Linkage analysis, in simple Mendelian disorders, tests for the Genome-wide linkage studies. Scans with the use of affected sibling pairs or
co-segregation of two polymorphic loci. A statistical calculation termed other affected relative pairs is a powerful approach especially if the genetic
the LOD score is a measure of the likelihood that a known normal allele variants under study have a high relative risk for predisposition to OA.
and the mutant allele in question for a disease syndrome co-segregate in a Siblings and other relatives who are concordant for OA because of genetic
family because they are physically close on the same chromosome versus predisposition will inherit their OA-predisposing alleles in common more
the two genes being on different chromosomes and travelling together by than would be expected by chance alone. On the other hand, siblings who
chance alone. For simple Mendelian genetic diseases, a LOD score above 3.0 are discordant for OA would be expected to inherit an OA predisposing
is considered strong statistical evidence of linkage. For complex genetic dis- allele in common less than would be expected by chance alone. Chapman
eases or traits dependent on the number of markers used and information and colleagues36 performed a two-stage genome-wide scan of 481 affected
content, higher LOD scores are required for strong statistical evidence of sibling pairs who were concordant for having undergone joint replacement
linkage. Lander and Kruglyak have reviewed the linkage analysis for com- surgery for primary OA. Analyses of their data suggested a potential OA sus-
plex traits in an excellent review.21 ceptibility region on chromosome 11q13 in the females with hip OA of the
Candidate gene linkage studies. Linkage analysis of affected sibling pairs gives a sibling pairs. Subsequent stratification of this cohort based on gender and
robust analysis especially in addressing uncertainties such as possible environ- site of involvement suggested other potential disease-predisposing genes on
mental con-founders in expression of phenotypes or uncertain inheritance chromosomes 4, 6, and 16.37 In a smaller cohort of 27 Finnish families with
patterns. Few studies have focused on families where OA is the only and pri- hand OA, Leppavuori et al.38 performed a genome-wide scan using
mary disease process without any chondrodysplasia. The other major limita- 302 micro-satellites. Eight regions supported linkage with LOD scores
tion to the studies has been the small numbers of sibling pairs, many with less 1.0 on chromosomes 2q, 4q, 7p, 8q, 9p, 9q, 10p, and 12q. The
than 80 pairs22–29 (See Table 4.3). Genetic studies have shown a positive link- X centromeric region also supported the linkage but at a lower level of
age to COL2A1 in two Finnish families having familial OA with classic clinical significance. In the second stage, additional markers and family members
and radiographic findings.23,30 Mutation analysis of the COL2A1 gene in 45 were genotyped in these chromosomal regions and supported linkage to
unrelated patients with familial OA found only a single putative COL2A1 chromosome 2q. Wright et al.27 genotyped 12 micro-satellite markers in
mutation.31 This result suggests that among patients with common OA, the 44 generalized nodal families in Nottingham, UK. Three of the twelve
incidence of COL2A1 mutations is low, probably even below 2 per cent. markers demonstrated linkage to chromosome 2q. However, the regions of
Loughlin and co-workers25 have investigated candidate genes as susceptibility chromosome 2, predicted to contain nodal OA susceptibility genes by
28 4    

Leppavuori and Wright, do not overlap suggesting that this chromosome investigated whether radiographic OA (ROA) is associated with specific
may contain more than one OA susceptibility gene. haplotypes of the COL2A1 gene in a large population-based study. The
Overall linkage analysis of affected sibling pairs has so far identified VNTR allele and the HindIII polymorphism showed significant association
regions of suggestive linkage on chromosomes 2q, 4q, 6, 7p, 11q, 16, and for ROA.
Xcen. Of these, chromosome 2q looks promising as it gave positive results The vitamin D receptor (VDR) has received considerable attention in
in genome screens performed in Oxford, Finland, and Nottingham as well OA. It is a gene implicated in regulating bone mass and density. Keen et al.28
as being linked in the Dutch pedigree. compared the allele frequencies and genotypes of an intragenic TaqI
Association analysis. Another contemporary approach for evaluating candi- dimorphism between females with primary radiographic knee OA and
date genes as risk factors for common OA has employed allelic association female controls. An allele of the Taq1 dimorphism was associated with
studies using case-control cohorts. Several OA association analyses have knee OA and particularly with osteophytes. Furthermore, the TaqI allele
been carried out22–24,26,28,29,39–54 (Table 4.4). Two main underlying was the haplotype associated with ROA at the knee joint in a study from
assumptions are pertinent in this approach. One is that OA risk in appar- Rotterdam.29 The COL2A1 and VDR genes are physically close on chro-
ently unrelated affected individuals is actually due to their inheriting the mosome 12q and one could surmise that an association to COL2A1 may
same OA-predisposing allele from a distant common ancestor. The second in fact be to the VDR gene and vice versa. Both genes alternatively, could
assumption is the prediction of the allele being disease causing. A number harbour susceptibility to OA. Uitterlinden looked at this possibility55 by
of different genes have been studied by tests of allelic association in the typing their subjects for the COL2A1. They have found that both genes
osteoarthritic population. These include the COL2A1 gene, CRTLI and appear to encode susceptibility for knee OA, with COL2A1 associated with
CRTM proteins, vitamin D receptor, oestrogen receptor, insulin-like joint-space narrowing and VDR associated with osteophytes.
growth factor 1, transforming growth factor -1 (TGF-1), and aggrecan. A number of other different genes have been studied by tests of allelic
The first COL2A1 study reported the association of an intragenic dimorph- association including CRTLI and CRTM proteins,26 estrogen receptor,49
ism with radiographic OA in a British cohort.22 Meulenbelt et al.43 insulin-like growth factor 1,50 aggrecan47 without firm conclusions or

Table 4.4 Primary OA association studies

Locus Site of OA Results Author

COL1A1 Female hip OA No association Aerssens, 1998
Hip or knee (both sexes) Association Loughlin, 2000
COL2A1 Female (more than one joint) Association Hull & Pope, 1989
PGOA/chondrodysplasia Association Knowlton, 1990
Nodal GOA No association Priestley, 1991
Hand or generalized OA No association Vikkula, 1993
Generalized OA both sexes Association Loughlin, 1995
Female hip No association Aerssens, 1998
Generalized OA both sexes Association Meulenbelt, 1999
Knee OA (JSN) Association Uitterlinden, 2000
COL9A1 Knees, hips, hands or spine (female) Association van Duijn, 1998
COL11A2 Knees, hips, hands or spine No association van Duijn, 1998
Hand OA, Association Leppavuori, 2000
CRTM Hip or knee Association (male) Meulenbelt, 1997
Hip or knee No Association Loughlin, 2000
CRTL1 Hip or knee No association Meulenbelt, 1997
Aggrecan Male hand or knee Association (hand) Horton, 1998
Female spine Association Kawaguchi, 1999
ER Generalized OA Association Ushiyama, 1998
IGF-1 Hand, hip, knee, or spine Association Meulenbelt, 1998
1
TGF- Female spine (osteophytes) Association Yamada, 2000
VDR Female knee Association Keen, 1997
Knee OA (osteophytes) Association Uitterlinden, 1997
Female hip OA No association Aerssens, 1998
Spine Association Jones, 1998
Male spine Association Videman, 1998
Hand, hip or knee No association Huang, 2000
Hip or knee No association Loughlin, 2000
Knee OA (osteophytes) Association Uitterlinden, 2000

CRTM, matrillin 1; CRTL, cartilage link protein; VDR, vitamin D receptor; ER, estrogen receptor; IGF-1, insulin-like growth factor-1; TGF-1, transforming growth factor 1.
 4.1       29

confirmation. A Japanese group has found TGF-1 to be related to disc Practice point 1: methods of genetic assessment in OA
degeneration.51 Lumbar disc degeneration has also been associated with
collagen 9 mutations in the Finnish population56 but its relationship to Relative risk to siblings (S)
OA at other sites is unclear. Of the collagen genes, association studies Risk estimates based on the presence of OA in siblings of patients with
have looked at COL1A1,39,40 COL2A1,39,42,57 COL9A1,45 COL11A2.45,46 OA and matched controls is obtained by:
Only COL9A1 and COL11A2 candidate genes reside on a chromosome,
chromosome 6, which has some evidence for linkage in the genome-wide per cent siblings with idiopathic OA
scans. (s) 
per cent controls with idiopathic OA
Allelic association studies are challenging to perform, as there may exist
a number of different genetic loci that independently contribute to the risk ◆ Controls in sibling and relative studies ideally resemble siblings
of OA. Also there are the difficulties encountered in selection/stratification
with respect to environmental risk factors but differ from them in
bias and ethnic or phenotypic differences, for instance comparing hip OA
regard to possible genetic determinants
with knee OA, or comparing ROA with end-stage OA. To overcome these
problems, a large number (thousands or tens of thousands) of patients and ◆ Controls should also be representative of the general population
control subjects could be required to identify an individual locus because in terms of their susceptibility to disease
it confers a risk in only a subset or small percentage of the osteoarthritic
population. h2 estimation
◆ A measure of the proportion of total variance of disease in a
population that is due to genetic influences

Defining the OA phenotype ◆ Genetic components include additive genetic effects (the effect
of individual alleles on the trait) and non-additive genetic effects
OA is now recognized as a heterogeneous group of conditions with a wide secondary to genetic dominance (the effect due to non-linear
variety of different pathological processes leading to a common outcome of interaction between alleles at a single locus)
joint destruction and disability. Linkage studies focusing on the end result
of OA, by examining the genetics of severe disease or joint replacement as
Twin pair studies
a categorical variable is a crude and underpowered approach to the prob-
lem. Even using global scores from joint radiographs may be too insensitive. ◆ In the classic twin method, the difference between intra-class
A more useful understanding of the physiology and genetic mechanisms of correlations for MZ twins and those for DZ twins is doubled to
the complex disease may be obtained by studying intermediate phenotypes provide a crude estimate of heritability with the remaining
individually or in combination. These are obtained by dividing OA into population variance attributed to environmental factors
its constituent parts; for example, bone turnover, subchondral bone sclero- h2  2(rMZ  rDZ)
sis, osteophyte formation, cartilage turnover and breakdown, cartilage
swelling, hydration, and volume. These intermediate phenotypes may oper-
ate independently or together in clusters determined by pleiotropic genes. Practice point 2: features of simple Mendelian disorders
A number of linkage simulations have suggested that the poor power of
linkage studies can be enhanced upto three-fold by combining the correct ◆ Onset of symptoms at childhood or teenage years
number of correlated phenotypes in analysis. The exact optimum number
◆ Multiple joints affected
and degree of correlation remains to be determined in practice. A common
thread among reported family studies has been that the genetic influences ◆ Unusual joint distribution such as elbow and ankle
on OA appear to be strongest for generalized OA, suggesting some common ◆ Other first degree relatives and family (siblings, offspring, parents)
susceptibility genes. The definition of ‘generalized OA’, however, has no
have similar findings
consensus by clinicians and epidemiologists and in general has weighted
heavily toward hand OA in the clinical studies.11,16 Furthermore, the preval- ◆ Radiographic evidence of skeletal disease at sites other than
ence of disease at any given joint site increases steeply with age, and it is not the symptomatic joint, such as the spine
clear that clustering between sites is greater than would be expected from ◆ There may be a disproportionate short stature or altered
the effects of age. There is a tendency towards polyarticular OA among
body habitus
women aged 45–64 years,58 but there is no single threshold number of joint
sites that can be used to define generalized OA. ◆ A proper diagnosis is of practical importance with respect to
genetic and occupational counselling

Conclusions Practice point 3

Twin pair, sibling risk, and segregation studies have demonstrated that ◆ Careful subdivision of OA phenotypes reduces the possibility that
primary OA is a complex genetic disease. Linkage analysis of affected sibling locus heterogeneity will confound the search for common genetic
pairs and rare pedigrees in which OA segregates as a Mendelian trait has variants that contribute to common OA.
revealed chromosomes 2, 4, and 16 as positive in more than one genome-
wide scan. Analysis of several known candidate genes has revealed evidence ◆ Any genetic mutation associated with a Mendelian disorder that
for COL2A1 and VDR genes. Tremendous effort and expense have already has OA as a component feature is capable of causing OA in
been committed to finding genes that confer risk for both rare and common isolation given the right mutation
forms of OA, and future efforts to identify OA genes will involve intense ◆ Persons with the same subtype of OA cannot be assumed to
investigations of those chromosomes highlighted by linkage studies. The have the same predisposing genetic variant; the genetic change
genetic study of OA is certain to lead to new therapeutic approaches for within the gene may have arisen independently among the
preventing and treating OA, as there undoubtedly will be a better under-
affected individuals
standing of joint development and homeostasis.
30 4    

25. Loughlin, J., Irven, C., Fergusson, C., and Sykes B. (1994). Sibling pair
References analysis shows no linkage of generalized osteoarthritis to the loci encoding
(An asterisk denotes recommended reading.) type II collagen, cartilage link protein or cartilage matrix protein. Br J
1. Cooper, C., McAlindon, T., Snow, S., et al. (1994). Mechanical and constitu- Rheumatol 33(12):1103–6.
tional risk factors for symptomatic knee osteoarthritis: differences between 26. Meulenbelt, I., Bijkerk, C., De Wildt, S.C., et al. (1997). Investigation of the
medial tibiofemoral and patellofemoral disease. J Rheumatol 21(2):307–13. association of the CRTM and CRTL1 genes with radiographically evident
2. Reich, T., James, J.W., and Morris, C.A. (1972). The use of multiple thresh- osteoarthritis in subjects from the Rotterdam study [see comments].
olds in determining the mode of transmission of semi-continuous traits. Arthritis Rheum 40(10):1760–5.
Ann Hum Genet 36(2):163–84. 27. Wright, G.D., Hughes, A.E., Regan, M., and Doherty, M. (1996).
3. Spector, T.D., Snieder, H., and MacGregor, A.J. (2000). Advances in twin Association of two loci on chromosome 2q with nodal osteoarthritis. Ann
and sib-pair analysis (1st ed.). Oxford: Oxford University Press. Rheum Dis 55(5):317–9.
4. *MacGregor, A.J., Spector, T.D. (1999). Twins and the genetic architecture of 28. Keen, R.W., Hart, D.J., Lanchbury, J.S., and Spector, T.D. (1997).
osteoarthritis. Rheumatology 38(7):583–8. Association of early osteoarthritis of the knee with a Taq I polymorphism of
A good review of the use of twin pairs in genetic studies with emphasis on OA. the vitamin D receptor gene. Arthritis Rheum 40(8):1444–9.
5. Stecher, R.M. (1941). Heberden’s nodes. Hereditary in hypertrophic arth- 29. Uitterlinden, A.G., Burger, H., Huang, Q., et al. (1997). Vitamin D receptor
ritis of finger joints. Am J Med Sci 201:801–9. genotype is associated with radiographic osteoarthritis at the knee. J Clin
6. Kellgren, J.H., Lawrence, J.S., and Bier, F. (1963). Genetic factors in general- Invest 100(2):259–63.
ized osteoarthrosis. Ann Rheum Dis 22:237–55. 30. Palotie, A., Vaisanen, P., Ott, J., et al. (1989). Predisposition to familial
7. Lawrence, J.S. and Moore, J. (1952). Generalized osteoarthritis and osteoarthrosis linked to type II collagen gene. Lancet 1(8644):924–7.
Heberden’s nodes. Br Med J 1:181–7. 31. Ritvaniemi, P., Korkko, J., Bonaventure, J., et al. (1995). Identification of
8. Lindberg, H. (1986). Prevalence of primary coxarthrosis in siblings of COL2A1 gene mutations in patients with chondrodysplasias and familial
patients with primary coxarthrosis. Clin Orthop 203:273–5. osteoarthritis. Arthritis Rheum 38(7):999–1004.
9. Chitnavis, J., Sinsheimer, J.S., Clipsham, K., et al. (1997). Genetic influences 32. Mustafa, Z., Chapman, K., Irven, C., et al. (2000). Linkage analysis of candid-
in end-stage osteoarthritis. Sibling risks of hip and knee replacement for ate genes as susceptibility loci for osteoarthritis-suggestive linkage of
idiopathic osteoarthritis. J Bone Joint Surg Br 79(4):660–4. COL9A1 to female hip osteoarthritis. Rheumatology 39(3):299–306.
10. Lanyon, P., Muir, K., Doherty, S., and Doherty, M. (2000). Assessment of 33. Meulenbelt, I., Bijkerk, C., Breedveld, F.C., and Slagboom, P.E. (1997).
a genetic contribution to osteoarthritis of the hip: sibling study. Br Med J Genetic linkage analysis of 14 candidate gene loci in a family with autosomal
321:1179–83. dominant osteoarthritis without dysplasia. J Med Genet 34(12):1024–7.
11. Spector, T.D., Cicuttini, F., Baker, J., Loughlin, J., and Hart, D. (1996). Genetic 34. Loughlin, J. (2001). Genetic epidemiology of primary osteoarthritis. Curr
influences on osteoarthritis in women: a twin study. Br Med J 312:940–3. Opin Rheumatol 13(2):111–16.
12. MacGregor, A.J., Antoniades, L., Matson, M., Andrew, T., and Spector, T.D. 35. Roby, P., Eyre, S., Worthington, J., et al. (1999). Autosomal dominant
(2000). The genetic contribution to radiographic hip osteoarthritis in (Beukes) premature degenerative osteoarthropathy of the hip joint maps to
women: results of a classic twin study. Arthritis Rheum 43(11):2410–16. an 11-cM region on chromosome 4q35. Am J Hum Genet 64(3):904–8.
13. Kaprio, J., Kujala, U.M., Peltonen, L., and Koskenvuo, M. (1996). Genetic lia- 36. *Chapman, K., Mustafa, Z., Irven, C., et al. (1999). Osteoarthritis-
bility to osteoarthritis may be greater in women than men. Br Med J 313:232. susceptibility locus on chromosome 11q, detected by linkage. Am J Hum
Genet 65(1):167–74.
14. Bijkerk, C., Houwing-Duistermaat, J.J., Valkenburg, H.A., et al. (1999).
The first published genome scan for OA susceptibility locus.
Heritabilities of radiologic osteoarthritis in peripheral joints and of disc
degeneration of the spine. Arthritis Rheum 42(8):1729–35. 37. *Loughlin, J., Mustafa, Z., Irven, C., et al. (1999). Stratification analysis
of an osteoarthritis genome screen-suggestive linkage to chromosomes 4, 6,
15. Hirsch, R., Lethbridge-Cejku, M., Hanson, R., et al. (1998). Familial aggrega-
and 16. Am J Hum Genet 65(6):1795–8.
tion of osteoarthritis: data from the Baltimore Longitudinal Study on Aging.
Additional loci detected by stratification of a genome scan.
Arthritis Rheum 41(7):1227–32.
38. Leppavuori, J., Kujala, U., Kinnunen, J., et al. (1999). Genome scan for pre-
16. Felson, D.T., Couropmitree, N.N., Chaisson, C.E., et al. (1998). Evidence for
disposing loci for distal interphalangeal joint osteoarthritis: evidence for a
a Mendelian gene in a segregation analysis of generalized radiographic
locus on 2q. Am J Hum Genet 65(4):1060–7.
osteoarthritis: the Framingham Study. Arthritis Rheum 41(6):1064–71.
39. Aerssens, J., Dequeker, J., Peeters, J., Breemans, S., and Boonen, S. (1998).
17. Spector, T.D., Snieder, H., and Keen, R. (1999). Interpreting the results of a
Lack of association between osteoarthritis of the hip and gene poly-
segregation analysis of generalized radiographic osteoarthritis: comment on
morphisms of VDR, COL1A1, and COL2A1 in postmenopausal women.
the article by Felson et al. Arthritis Rheum 42(5):1068–70.
Arthritis Rheum 41(11):1946–50.
18. International nomenclature and classification of the osteochondrodysplasias
40. Loughlin, J., Sinsheimer, J.S., Mustafa, Z., et al. (2000). Association analysis
(1997). (International Working Group on Constitutional Diseases of Bone
of the vitamin D receptor gene, the type I collagen gene COL1A1, and the
1998.) Am J Med Genet 79(5):376–82.
estrogen receptor gene in idiopathic osteoarthritis. J Rheumatol 27(3):
19. Annunen, S., Korkko, J., Czarny, M., et al. (1999). Splicing mutations of
779–84.
54-bp exons in the COL11A1 gene cause the Marshall syndrome, but other
41. Knowlton, R.G., Katzenstein, P.L., Moskowitz, R.W., et al. (1990). Genetic
mutations cause overlapping Marshall/Stickler phenotypes. Am J Hum
linkage of a polymorphism in the type II procollagen gene (COL2A1) to prim-
Genet 65(4):974–83.
ary osteoarthritis associated with mild chondrodysplasia. N Engl J Med
20. Snead, M.P. and Yates, J.R. (1999). Clinical and molecular genetics of
322(8):526–30.
Stickler syndrome. J Med Genet 36(5):353–9.
42. Loughlin, J., Irven, C., Athanasou, N., Carr, A., and Sykes, B. (1995).
21. Lander, E. and Kruglyak, L. (1995). Genetic dissection of complex traits:
Differential allelic expression of the type II collagen gene (COL2A1) in
guidelines for interpreting and reporting linkage results. Nat Genet
osteoarthritic cartilage. Am J Hum Genet 56(5):1186–93.
11(3):241–7.
43. Meulenbelt, I., Bijkerk, C., De Wildt, S.C., et al. (1999). Haplotype analysis
22. Hull, R. and Pope, F.M. (1989). Osteoarthritis and cartilage collagen genes.
of three polymorphisms of the COL2A1 gene and associations with general-
Lancet 1(8650):1337–8.
ized radiological osteoarthritis. Ann Hum Genet 63(Pt 5):393–400.
23. Vikkula, M., Palotie, A., and Ritvaniemi, P., et al. (1993). Early-onset
44. *Uitterlinden, A.G., Burger, H., van Duijn, C.M., et al. (2000). Adjacent
osteoarthritis linked to the type II procollagen gene. Detailed clinical pheno-
genes, for COL2A1 and the vitamin D receptor, are associated with separate
type and further analyses of the gene. Arthritis Rheum 36(3):401–9.
features of radiographic osteoarthritis of the knee. Arthritis Rheum
24. Priestley, L., Fergusson, C., Ogilvie, D., et al. (1991). A limited association of 43(7):1456–64.
generalized osteoarthritis with alleles at the type II collagen locus: COL2A1. Determines possible linkage disequilibrium between VDR and COL2A1
Br J Rheumatol 30(4):272–5. genes.
 4.1       31

45. van Duijn, C.M., Bijkerk, C., Houwing-Duistermaat, J.J., et al. (1998). 52. Jones, G., White, C., Sambrook, P., Eisman, J. (1998). Allelic variation in the
A population based study of the genetics of osteoarthritis [abstract]. Am J vitamin D receptor, lifestyle factors and lumbar spinal degenerative disease.
Hum Genet 63(Suppl.):1282. Ann Rheum Dis 57(2):94–9.
46. Leppavuori, J.K., Pastinen, T., Kujala, U. et al. (2000). Array-based candidate 53. Videman, T., Leppavuori, J., Kaprio, J., et al. (1998). Intragenic poly-
gene analysis identify COL11A2 on 6p21.3 as predisposing locus for distal morphisms of the vitamin D receptor gene associated with intervertebral
interphalangeal joint osteoarthritis [abstract]. Am J Hum Genet 67(Suppl.):125. disc degeneration. Spine 23(23):2477–85.
47. Horton, W.E. Jr, Lethbridge-Cejku, M., Hochberg, M.C., et al. (1998). An 54. Huang, J., Ushiyama, T., Inoue, K., Kawasaki, T., and Hukuda, S. (2000).
association between an aggrecan polymorphic allele and bilateral hand Vitamin D receptor gene polymorphisms and osteoarthritis of the hand, hip,
osteoarthritis in elderly white men: data from the Baltimore Longitudinal and knee: a case-control study in Japan. Rheumatology 39(1):79–84.
Study of Aging (BLSA). Osteoarthritis Cart 6(4):245–51. 55. Uitterlinden, A.G., Burger, H., van Duijn, C.M., et al. (2000). Adjacent
48. Kawaguchi, Y., Osada, R., Kanamori, M., et al. (1999). Association between genes, for COL2A1 and the vitamin D receptor, are associated with separate
an aggrecan gene polymorphism and lumbar disc degeneration. Spine features of radiographic osteoarthritis of the knee. Arthritis Rheum
24(23):2456–60. 43(7):1456–64.
49. Ushiyama, T., Ueyama, H., Inoue, K., Nishioka, J., Ohkubo, I., and 56. Paassilta, P., Lohiniva, J., Goring, H.H., et al. (2001). Identification of
Hukuda, S. (1998). Estrogen receptor gene polymorphism and generalized a novel common genetic risk factor for lumbar disk disease. JAMA
osteoarthritis. J Rheumatol 25(1):134–7. 285(14):1843–9.
50. Meulenbelt, I., Bijkerk, C., Miedema, H.S., et al. (1998). A genetic associ- 57. Vikkula, M., Nissila, M., Hirvensalo, E., et al. (1993). Multiallelic poly-
ation study of the IGF-1 gene and radiological osteoarthritis in a population- morphism of the cartilage collagen gene: no association with osteo-
based cohort study (the Rotterdam Study). Ann Rheum Dis 57(6):371–4. arthrosis. Ann Rheum Dis 52(10):762–4.
51. Yamada, Y., Okuizumi, H., Miyauchi, A., Takagi, Y., Ikeda, K., Harada, A. 58. Cooper, C., Egger, P., Coggon, D., et al. (1996). Generalized osteoarthritis
(2000). Association of transforming growth factor beta1 genotype with in women: pattern of joint involvement and approaches to definition for
spinal osteophytosis in Japanese women. Arthritis Rheum 43 (2):452–60. epidemiological studies. J Rheumatol 23(11):1938–42.
This page intentionally left blank
 4.2 Specific gene defects associated with
osteoarthritis
Thorvaldur Ingvarsson and Stefan Einar Stefansson

The recent publication of a first draft of the human genomic sequence will
facilitate the search for OA genes but the obstacles remain significant. For
The study of candidate genes for OA by
example, the identification of positional candidates still depends on association analysis
uncovering evidence of linkage between a chromosomal region and an OA
A candidate gene is defined as a gene whose protein product, based on its
phenotype. Once a chromosomal region has been found to link to OA it will
biological activity, can plausibly be assumed to influence the disease
be necessary to find a gene in the linked region and for this it will in many
(phenotype) under consideration. The candidate gene can be directly
cases be necessary to narrow the region down, by applying additional
screened for mutations in the affected family or individuals. The main
marker sets, to a more manageable size for DNA sequencing or other muta-
problem with this approach is that the number of these proteins is already
tion detection techniques. However, it is not sufficient merely to show an
large, and as more is learned about cartilage and bone biology the list of
association between a variant in a candidate gene and a phenotype, since
potential candidate genes keeps growing. The chances of a ‘lucky hit’ would
such an association may merely reflect linkage disequilibrium. The variant,
seem to be remote. Indeed studies of candidate genes have been disappoint-
such as a missense mutation, must also be shown to alter the biological
ing, with the exception of some of the chondrodysplasia families.
function of a protein in animal models or a tissue such as cartilage that
Most of the association analyses for OA, which have targeted candidate
relates to OA. For putative mutations in regulatory regions this chain of evi-
genes, have tended to be genes encoding structural proteins of the
dence becomes even more crucial. Current developments in molecular biol-
extracellular matrix of cartilage and bone or genes implicated in the regu-
ogy such as micro-array chips now allow the simultaneous large-scale
lation of bone density.
identification of thousands of genes with differential expression in disease.
Some examples of known mutations in candidate genes follow (see
This will enhance our ability to identify candidate molecules and processes
Table 4.5 for more information).
that are relevant to OA pathology.
Multiple pathogenetic mechanisms are implicated in the development
of OA. Continued studies of the kind outlined here will clarify the complex
genetic background of OA and identify genetic variation associated with
COL2A1
the disease. In addition to improving our understanding of the pathogen- A large number of mutations (over 100) have been reported in the
esis of OA and identifying new molecular targets for treatment, this know- COL2A1 gene on chromosome 12, which codes for II collagen.3 The clinical
ledge will also allow a better insight into the interactions between the genetic phenotypes have ranged widely in severity from the achondrogenesis
background and the environmental factors that initiate and drive OA. type II and hypochondrogenesis at the severe end of the spectrum to very
mild spondyloepiphyseal dysplasia (SED) with precocious OA, often called
late onset SED. Further phenotypes have included early onset premature
generalized OA (PGOA),4,5 families with crystal deposition disease pheno-
Is OA subject to major gene effects? types of intermediate severity,6 SED-congenita, and Kniest dysplasia.
Mutations have also been described in Stickler dysplasia, in which skeletal
Given that genetic factors may play a major role in the etiology of OA, the
phenotype is typical with OA rather than short stature, including eye
question arises whether these influences are exerted by one or a few ‘major
and inner ear abnormalities. See Table 4.5 and Refs 2 and 13 for further
genes’ each with a relatively large effect, such as the COL2A1 gene, or by a
details.
large number of ‘polygenes’, each with a relatively minor effect. Presumably,
in the monogenic form of OA the genes would be easier to identify and
could have greater relevance to public health. Complex segregation analysis
can be used to study the mode of inheritance of a trait and in particular to
COL11A2
infer whether the distribution of the trait within pedigrees is compatible Mutations have been described in the COL11A2 gene, which encode the
with the action of a major gene. This was done by Felson and co-workers in alpha-2 chain of type XI collagen3,7 in families with Stickler syndromes
1998,1 concluding that in generalized OA (in this study hand and knee OA) presenting mild SED, premature OA and sensorineural hearing loss, but
there was evidence to support a significant genetic contribution, with evid- lacking the eye involvement.
ence for a major recessive gene and a multi-factorial component, repres-
enting either polygenic or environmental factors. However, the evidence is
not yet conclusive that ‘common OA’ is influenced by major genes.2 On the COL9A1
contrary, several susceptible loci for OA have been suggested and each of Several mutations have been found in the gene for type IX collagen,7 all
them is thought to include a possible gene for OA. This could support the associated with Schmid metaphyseal chondrodysplasia, which is a relatively
notion that the inherited OA is a polygenic disease. mild chondrodysplasia.
34 4    

Table 4.5 Examples of known genes associated with OA mutations, their chromosomal location and the respective protein

Gene/locus Chromosome Gene product function Clinical phenotype Known mutations Reference Protein
COL11A1 1p21 Extracellular matrix protein Knee, hip, hand No 3,7 Alpha-1-chain type XI
collagen
COL11A2 6p21.3 Extracellular matrix protein Stickler dysplasia-SED Yes 3 Alpha-2-chain type XI
late onset, knee, collagen
hip, hand
COL2A1 12q13-q14 Extracellular matrix protein Achondrogenesis Yes 2 Alpha-1-chain type II
Hypochondrogenesis hundreds collagen
SED congenita
SED late onset
Kniest dysplasia
Stickler dysplasia
COL9A1 6q12-q13 Extracellular matrix protein Knee, hip, hand Yes 4 Alpha-1-chain type IX
collagen
COL9A2 1p32 Extracellular matrix protein Multiple epiphyseal Yes 3 Alpha-2-chain type IX
dysplasia collagen
COL10A1 6p Extracellular matrix protein Schmid metaphyseal Yes 3 Alpha-1-chain type X
hypertrophic cartilage chondrodysplasia several collagen
COMP 19p13.1 Extracellular matrix protein Multiple epiphyseal Yes 3 cartilage oligomeric matrix
dysplasia (Fairbanks) many protein
Pseudoachondroplasia
FGFR3 4p16.3 Tyrosine kinase transmembrane Thanotophoric dysplasia Yes 3,65–66
receptor for FGFs Achondrodysplasia many
Hypochondrodysplasia
PTHrPR 11p15.3 G-protein transmembrane Jansen metaphyseal Yes 3 G-protein
receptor for PTH and PTHrP chondrodysplasia
CRTL1 5q13-14.1 Cartilage link protein Hip or knee 53 Cartilage link protein
DTDST 5q31-q34 Transmembrane sulfate Achondrogenesis type IB Yes 3 diastrophic dysplasia sulfate
transporter Atelosteogenesis type II several transporter
Diastrophic dysplasia
VDR 12q13-q14 Vitamin D-receptor Hip, knee, hand, spine Unknown loci 9,10,54–56

COL9A2 MATN3 (Matrilin-3)

Mutations have been found in the type IX collagen in a multiple epiphyseal The gene MATN3 (matrilin-3) is located on chromosome 2p and mutations
dysplasia (MED) family (EDM2) on chromosome 1.41 Type I collagen is have been described associated with MED.8
a quantitatively minor cartilage collagen thought to participate in the regu-
lation of collagen fibril assembly in cartilage matrix.
VDR
FGFR3 A VDR haplotype was shown to be associated with higher risk for knee OA
Mutations in the gene for fibroblast growth factor receptor 3 are known in both women and men.9 The authors concluded that the two- to three-
in achondroplasia and hypochondroplasia.66,67 The gene is located on fold risk increase associated with this haplotype was due to increased osteo-
chromosome 4. phytosis rather than joint-space narrowing, suggesting that radiological
changes indicative of cartilage degeneration were not associated with this
COMP haplotype. In further support of an association of the VDR with OA,
Cartilage oligomeric matrix protein (COMP) is a member of the thrombo- a British survey showed that women with a specific VDR haplotype had
spondin protein family. It is found in the extracellular matrix of cartilage a three-fold increase of knee OA.10
and to a lesser extent in other connective tissues. The function of COMP is It is interesting to note, however, that the VDR gene locus is located in
not well-defined.3 Mutations have been described in the COMP gene close proximity (within 100 000 base pairs) of the type II collagen gene
located on chromosome 19 in pseudoachondroplasia and MED. It is postu- (COL2A1) on chromosome 12q. Accordingly, it was speculated that the
lated that pseudoachondroplasia and some forms of MED could share com- VDR is not the causative locus, but that it may be in linkage disequilibrium
mon pathogenetic features. with a neighboring disease-causing gene.11
Adding further complexity, a recent study suggested that the COL2A1
and the VDR genes are associated with separate features of ROA of the
DTDST knee. The investigators concluded that both genes are involved in knee OA,
The gene DTDST (diastrophic dysplasia sulfate transporter) is located but with separate features of radiographic knee OA: the COL2A1 pheno-
on chromosome 5q and is responsible for chondrodiastrophic dysplasia. type is associated with joint-space narrowing, while the VDR phenotype is
Several mutations are known.7 associated with osteophytes.12
 4.2       35

This gave evidence for the existence of a genetic predisposition to OA,

Association and linkage studies for example by a number of rare subtypes of OA including familial calcium
Before investing major resources in the detailed study of a potential candid- pyrophosphate deposition disease, Stickler syndrome, and some chondro-
ate gene, a strong prior hypothesis implicating the proposed candidate gene dysplasias that have a genetic basis.14 The spectrum of these particular
is needed, rather than just a general sense based on its biology. Such infor- hereditary forms of OA is quite varied encompassing mild disorders, which
mation can be gained in association studies, which can be done on both do not become clinically apparent until late adult life, to very severe forms
related and unrelated individuals. In this case one or several markers are that manifest during childhood. Many of these disorders have been classi-
selected for study, which are located within or adjacent to the candidate fied as secondary OA.15 They all have the common characteristic of being
gene and the association of these and other markers with the phenotype is associated with mutations in genes encoding macromolecules predomin-
investigated. Population stratification with the existence of more than one antly expressed in cartilage. Genes that may be involved in these diseases
ancestral source of population gene pool represents a major limitation of include those encoding cartilage-specific collagens (types II, IX, X, and XI),
this study design. If, as likely, the various ancestral sources differ both in proteoglycan core protein and link proteins, noncollagenous components
their susceptibility to various diseases and in the frequency of various of the cartilage matrix, growth factors involved in cartilage differentiation
genetic markers, false associations may be observed between genetic mark- or in the regulation of chondrocyte proliferation and specific gene expres-
ers and phenotypes. It is likely that population stratification is at least partly sion, and genes encoding enzymes involved in various cartilage-specific
responsible for some of the associations between candidate gene polymorph- metabolic pathways. Thus, these disorders may be seen to represent a dis-
isms and diseases that have later proven to be nonreplicable. Examples of tinct subgroup of OA that can be separated from secondary OA.14
loci that are associated with OA are shown in Table 4.6 and candidate gene After the initial descriptions of genetic linkage between the phenotype
polymorphisms associated with OA in Table 4.7. of precocious OA and the type II procollagen gene COL2A1 on chromo-
some 12,16 a large number of mutations in the genes encoding structural
and functional components of cartilage have been identified in various
hereditary diseases affecting this tissue.17–19 Progress on the elucidation of
Inherited specific forms of OA and gene mutations in hereditary OA has been most substantial for COL2A1,
mutations associated with OA in rare the gene encoding for type II collagen, the most abundant collagen in artic-
ular cartilage. Accordingly, the term ‘type II collagenopathies’ has been
Mendelian families used to describe hereditary cartilage diseases in which their primary defect
During the last decade several groups have reported linkage analysis of can- is mutation in COL2A1.20 Over 100 mutations are known in the COL2A1
didate genes to OA in families in which the disease was transmitted as a
dominant trait, often with incomplete penetrance. It soon became apparent, Table 4.7 Candidate gene polymorphisms associated with
that affected individuals in these families had mild osteochondrodysplasias osteoarthritis
or other rare diseases that can predispose to OA and that the OA in these
families is best described as secondary and not the common primary OA. Genetic Phenotype Association Reference
Recently, a number of groups have described families with early onset OA in polymorphism found?
the absence of chondrodysplasia.13 VDR Female knee OA Yes 10
Knee OA Yes 9
Table 4.6 Quantitative trait loci (QTL) associated with osteoarthritis (osteophytes)
Female OA No 54
Chromosome Joint(s) affected Phenotype Reference (hip replacements)
8q GOA Early-onset OA-CPDD 27 Hand, hip, knee OA No 55
(1 family)
Idiopathic OA No 56
2q23-35 Hand Nodal OA 47
COL2A1 GOA/chondrodysplasia Yes 57
11q Hip, knee Female 45
Nodal GOA No 58
2q Hip, knee OA of the hip 2
GOA, finger joints No 43
4q Hip, knee Female OA of the hip 2
GOA Yes 59
6p/6q Hip OA of the hip 2
Female OA No 54
11q Hip Female OA 2 (hip replacement)
16p/16q Hip Female OA of the hip 2 GOA Yes 60
2q12-13 Hand Dip OA 48 Knee OA Yes 12
4q26-27 Hand Dip OA 48 COL1A1 Female OA No 54
7p15-21 Hand Dip OA 48 (hip replacement)
X-cen Hand Dip OA 48 Idiopathic female OA Yes 56
4q35 Hip Premature degenerative 49 ER alfa GOA Yes 61
OA of the hip Idiopathic OA No 56
6q12-13 Hip, knee Female OA of the hip 50 TGFB1 Spine OA (osteophytes) Yes 62
6p21.3 Hip Female OA of the hip 50 IGF-I GOA Yes 63
2q31 Hip, knee Familial OA of the hip 46 Aggrecan Male bilateral hand OA Yes 64
16q Hip Familial OA of the hip 51 proteoglycan

CPDD, calcium pyrophosphate deposition disease; GOA, generalized osteoarthritis; OA, osteoarthritis; COL, collagen; ER, estrogen receptor; GOA, generalized osteoarthritis;
OA, osteoarthritis. IGF, insulin-like growth factor; TGF, transforming growth factor; JS, joint space.
36 4    

gene associated with different forms of hereditary cartilage aberrations and

OA. These mutations are of considerable scientific interest, but there is yet
Chondrodysplasias
little evidence that common forms of OA are due to mutations in type II These disorders represent a group of clinically heterogeneous hereditary
collagen, as a fact even the opposite.21 disorders, characterized by abnormalities in the growth and development of
articular and growth plate cartilages. The classifications are based on clin-
ical and radiographic features of the affected individuals. In the future, it
will be possible to classify those disorders by the nature of the genetic
Genome-wide scanning for linkage to defects, as the gene mutations responsible for these diseases are identified.
OA phenotypes Several of them are associated with the development of premature OA.
Linkage to COL2A1 has been demonstrated in some cases and excluded in
With the increasing availability of large numbers of highly informative
others.14 These disorders include SED, Stickler syndrome, Kniest dysplasia,
genetic markers, particularly micro-satellite markers that now span the
MED, and metaphyseal chondrodysplasias.
entire genome, the strategy of whole-genome scanning for linkage to
phenotypes of interest has become feasible.
Spondyloepiphyseal dysplasias
Primary generalized OA (PGOA) SED is a heterogeneous group of autosomal dominant disorders character-
ized by abnormal development of the axial skeleton and severe alterations of
PGOA is thought to be the most common form of inherited OA. It is char-
the epiphysis of long bones, often resulting in dwarfism with a marked
acterized by the familial trait of hand OA (Heberden’s and Bouchard’s
shortening of the trunk and to a lesser extent of the extremities.14 The pheno-
nodes) and premature alterations in multiple joints.22 Typically, the clinical
type of SED is quite varied, ranging from severe forms that are clinically
and radiological features have a precocious onset and an accelerated pro-
apparent at birth to milder forms, which manifest in childhood or early
gression. Generally, the loss of articular cartilage is concentric in the knees
adolescence. In the late onset form, early degenerative changes are seen in
and hips. The radiographic appearance is indistinguishable from that of
multiple joints, often as generalized OA. Hip OA may develop in adoles-
nonhereditary OA, except for the premature occurrence, increased severity,
cence, worsening in early adulthood. There may be mild epiphyseal abnor-
and rapid progression.14 In the hand the DIP, PIP, and CMC I joints are
mality in the peripheral joints, for example in the metacarpal phalangeal
involved. The hip is affected in early adult life.
joints. The most constant features in peripheral joints are the flattening of
Mutations in COL2A1 genes have been identified in affected members of
the articular surfaces of the ankles and knees and shallowness of femoral
several families with generalized OA with premature onset and rapid pro-
inter-condylar notches. In adult classic cases, platyspondyly may be severe,
gression. However, all these families also display evidence of mild SED.14
accompanied by severe coxa vara and small and irregular femoral epiphyses.
Conversely, however, a recent study analysing COL2A1 for mutations in
In milder cases the features of SED are not clinically obvious and many
47 families with early onset generalized OA without SED, identified a pos-
patients initially present with severe OA affecting multiple joints.14 Many
sible COL2A1 mutation in only one of the cases. Further screenings of fam-
different mutations have been described in families with SED. Several sup-
ilies with generalized OA showed that only some two per cent of them had
posedly unrelated families have demonstrated coinheritance of primary
a mutation in the COL2A1 gene, suggesting that only a small proportion of
generalized OA with mild SED with specific alleles of the gene for type II
OA can be explained by this genetic variation.23 Other reports suggest that
procollagen on chromosome 12.32–34 This allele has now been cloned and
COL2A1 is not the disease locus in families with premature generalized OA,
found to contain a single base mutation at position 519 of the alpha-1(II)
without evidence of SED.6
chain.33 It was speculated that three of the known families with OA and
A genetic predisposition in primary generalized OA has also been sug-
mild SED and which share this COL2A1 gene defect might be related
gested to be associated with HLA-A1B8 and HLA-B8 haplotypes24 and
through an early founder.34 In SED tarda a different mutation was recently
with alfa-1-antitrypsin isoform patterns, while other studies have failed to
described on chromosome Xp22.35
confirm this association.25

Familial calcium pyrophosphate deposition Stickler syndrome

This is a form of inherited OA characterized by ocular involvement asso-
This condition is also known as familial chondrocalcinosis. Following an
ciated with severe premature OA. Patients have myopia, hearing loss, and
initial description of the disease in five Czech families,26 multiple ethnic
epiphyseal dysplasia and even mandibular dysplasia. Linkage analysis of
series have been reported.14 The condition appears to be inherited in an
several Stickler’s syndrome kindreds has demonstrated that this disease is
autosomal dominant manner with precocious onset and severe clinical
linked to COL2A1 in some 25–50 per cent of the families and mutations
expression. Radiographs show chondrocalcinosis most frequently in the
have been shown.36 Linkage to collagen XI on chromosome 6 and a muta-
knees, symphysis pubis, and the wrist. These changes may precede OA
tion in the gene encoding the alpha-2 XI collagen chain was demonstrated
changes in the joint. No linkage to the COL2A1 gene on chromosome 12
in one family.37
has been proven, but susceptibility loci have been described on chromo-
some 5 and 8.27–29
A role for the ank gene in controlling tissue calcification in mice has Kniest dysplasia
been described.30 Ongoing investigations explore the role of the ank gene Kniest dysplasia is a disorder characterized by an autosomal dominant pat-
in human OA.28,29 tern of inheritance displaying shortening of the trunk and limbs, flattening
of the face, and severe joint abnormalities. The majority of affected indi-
Familial hydroxyapatite deposition disease viduals develop severe premature OA, most prominent in the hips and knee.
Mutations in COL2A1 have been identified in Kniest dysplasia.14
Another form of inherited crystal deposition disease is due to the deposi-
tion of hydroxyapatite crystals in articular cartilage. The mode of inherit-
ance is that of an autosomal dominant pattern with full penetrance. This Multiple epiphyseal dysplasia
disorder results in periarticular disease in the form of tendinitis or bursitis MED is a heterogeneous disorder characterized by alterations in epiphyseal
and less frequently, true articular disease. The most common locations are growth that cause irregularity and fragmentation of the epiphyses of
the shoulders, wrist, and hips. A recent study of a family from Argentina multiple long bones. Spinal alteration is absent or slight. The condition
with this condition and mild SED excluded several candidate genes includ- results in premature OA of both weight-bearing and nonweight-bearing
ing type II and X collagen.31 joints, often in childhood or early adulthood. Early studies of MED failed to
 4.2       37

identify an association with the genes for collagen types II and VI, chon- or both, for primary OA, Heberden’s nodes were noted in 38.5 per cent of
droitin sulfate, proteoglycan core protein, and cartilage link protein. the affected individuals. Using a genome-wide scan, a locus associating
Subsequent studies identified mutations in the gene encoding COMP,38,39 with OA in women was identified on chromosome 11q with a LOD score
in the COL9A2 gene,40,41 and lately in the MATN3 gene on chromosome 2.8 between 2 and 3. The authors suggested that a female-specific susceptibil-
It is likely that additional and different loci will be identified, since other ity gene for primary OA was present on chromosome 11q.45 Further
families with this phenotype have been shown to lack genetic linkage with reports on genome-wide scans have indicated the existence of multiple sus-
either COMP or COL9A2.42 ceptibility loci for OA by stratifying the same material by gender and joint.
It has even been shown that mutations in MED can in some cases be Thus, linkage to loci on chromosomes 2, 4, 6, and 16 was proposed.2,46
joint specific. The type of MED with predominant involvement of the
capital femoral epiphyses can have mutations in the COMP gene,43 and Chromosome 2
those patients with predominant involvement of knees with relative hip
sparing can have COL9A2 or COL9A3 mutations.43 This suggests that A suggestion of linkage was based on a LOD score of 1.22 that increased to
mutations can be joint specific with regard to phenotypic expression. 2.19 in hip patients, and it was stated that this suggestive linkage was greater
However, we do not as yet know if there are corresponding differences in for male hip patients. The proposed region 2q31 is at a similar location
the developmental or spatial expression patterns of these genes between described in nodal OA47 and near a locus, which was described for DIP
different joints. Table 4.7 lists examples of known genes and mutations OA.48 It appears possible that chromosome 2q contains at least one suscept-
associated with OA. ibility locus for OA.

Metaphyseal chondrodysplasias Chromosome 4

More than 150 different types of metaphyseal chondrodysplasias are The suggestion of linkage was based on a LOD score of 3.9 and centered on
known. The clinical features of affected individuals include short stature the chromosomal region 4q12-21.2, based on female sibling pairs with hip
with short limbs, bowed legs, and a waddling gait. Mutations have been OA.49 A separate study of a Dutch family with hip OA proposed linkage
shown in the COL2A1 gene.14 to a region of chromosome 4 more than 50 cM distal from the locus at
4q12-21.2. These authors concluded that it was unlikely that the two link-
ages had detected the same locus.2

Model free linkage analysis of affected Chromosome 6

Linkage for hip OA was based on an LOD score of 2.9 and centered on
sibling pairs by genome-wide scanning for a strong candidate gene for OA, COL9A1 at 6q12-13.2 This locus was
chromosomal loci associated with OA further confirmed in an additional report by the same group, suggesting that
the COL9A1 gene was associated with susceptibility for hip OA in females.50
The rarity of families in which primary OA segregates as a Mendelian trait,
combined with the potential of osteochondro dysplasias in these families,
has promoted the use of affected sibling pairs. Chromosome 16
As summarized in Table 4.5, a number of candidate genes have been pro- Linkage showed an LOD score of 2.1, but no candidate genes for OA were
posed to be associated with OA, but the results from studies published so far associated with this locus. Again, the material was stratified for hip OA in
are somewhat conflicting. This may be a reflection of the complexity of the women. The authors concluded that their analysis highlighted the potential
disease and/or be related to the limited power of some of these studies. The utility of genome-wide screens, and that stratification of the material
pathophysiology of OA is complex and the a priori choice of candidate genes revealed additional chromosomal regions that may harbor susceptibility
is increasingly difficult and prone to bias. Association studies may provide loci for OA in this British population.2
support for the role of specific candidate genes in an inherited disease, but A genome-wide scan was done on a large Icelandic family with mainly hip
population heterogeneity provides a limitation to this study design. OA, which showed linkage with a LOD score of 2.58 on chromosome 16 at
Systematic genome-wide scan for linkage of markers to phenotypes of a similar location as in the Loughlin study.51 This is the first independent
interest, using large numbers of randomly distributed anonymous poly- description of the same susceptibility locus in two different populations.
morphic micro-satellite markers together with DNA samples from large However, this raises the possibility that they have the same founder. In this
families or large numbers of sibling pairs with OA may be used to identify context, it is interesting to note that the Icelandic population gene pool
yet other, unknown, predisposing chromosomal loci for OA. However, has a significant component that appears to originate from Orkney, the
depending on the density and location of micro-satellite markers used in Western Isles, and the Isle of Skye (British Isles).52
the genome-wide scan, the chromosomal region implicated may sometimes
contain a large number of genes. Often, 300–900 micro-satellite markers
are used. The LOD score is used as a measure of support for linkage versus
absence of linkage,44 where a LOD score above 2.0 is sometimes referred to Conclusions
as ‘suggestive’, and a LOD score above 3.0 as ‘significant’ linkage. These are, Some conclusions may be drawn from the currently known mutations in
however, arbitrary LOD score cut-off levels and may need to be changed diseases associated with OA.
depending on the particular experimental design and statistical analysis First, no mutation is known today in the ‘common’ primary form of OA,
performed. and most of the mutations are associated with relatively rare syndromes
Genome-wide scans are often followed by further scans of the chromo- or diseases with what could be classified as secondary OA.
somal region of interest with higher density marker panels, to narrow down Second, most of the known mutations occur in collagen genes and it is
the chromosomal region (and number of genes) associating with the worth noting that mutations in, for example, COL2A1 genes can have clin-
phenotype of interest. However, assuming that a gene can eventually be ical phenotypes ranging from mild SED with premature OA of the hips to
identified through this search strategy, the challenge remains to identify severe crippling disease from childhood.
sequence variants that influence the function of the protein and result in Third, the number of newly discovered chondrodysplasia loci is decreas-
the phenotype. ing, but the number of ‘the garden-variety OA’ loci is increasing.
In a study using 481 families with at least two siblings each of whom had Fourth, chromosomes 2, 4, and 16 were identified in multiple genome
undergone one or more of total hip replacement or total knee replacement, scans and are therefore the most likely to encode susceptibility.
Another random document with
no related content on Scribd:
their Russian officers, on condition that the latter are
completely under the orders of the Minister of War. This
arrangement was ratified at a meeting between the commanders
and Colonel Liakhoff. Teheran is quiet, and the Persian
Cossacks are already fraternizing with the Fedai. The Sipahdar
has been appointed Minister of War, and the Sirdar Assad
Minister of the Interior." Being asked if he would represent
to the Russian Government the undesirability of advancing
Russian troops to Teheran, Sir Edward added: "In view of the
declarations already made by the Russian Government as to the
circumstances under which alone Russian troops would be sent
to Teheran and in view of the fact that no troops have been
sent to Teheran during the recent troubles, in spite of the
fact that at one time some apprehension, which happily proved
to be unfounded, was expressed for the safety of Russian
subjects, such representations would be most uncalled for."

On the 17th the Provisional Government gave notice to the

Anglo-Russian legations of the selection of the new Shah, and
asked that he should be delivered to their keeping; whereupon,
wrote the Times correspondent, "M. Sablin announced the
request to the Shah, who replied that he thought his mother
would not consent. The Shah then took M. Sablin to his mother
and an affecting scene ensued. Both the mother and father
broke down at the thought of parting with their favourite son
and offered their second son in his place. M. Sablin replied
that the selection had been made by the people and that he had
no voice in the matter. The boy wept bitterly in sympathy with
his parents and declined to leave his mother. Finally their
Majesties were persuaded to agree. On receiving the Shah’s
assent, the necessary proclamation was immediately promulgated
and it was arranged that the Regent and a Nationalist
deputation would receive the little Shah.

"An interested crowd witnessed his departure this morning from

the custody of his natural guardians. During the morning
Sultan Ahmed wept bitterly at the prospect of becoming a King,
and it required a stern message to the effect that crying was
not allowed in the Russian Legation before he dried his eyes.
Then the little man came out bravely, entered a large
carriage, and drove off alone, escorted by Cossacks, Sowars,
and Persian Cossacks and followed by a long string of
carriages. At Sultanatabad he was met by the Regent and the
deputation and ceremoniously notified of his high position and
of the hope entertained by the nation that he would prove to
be a good ruler.

‘Inshallah, I will,’ replied the lad. Arrangements for the

Coronation will be made hereafter. In the meanwhile the little
Shah, who is guarded by a Bakhtiari, remains with his tutors
at Sultanatabad, where his mother is free to visit him."

At Teheran, affairs settled quickly into quiet, but disorders

were prolonged in various parts of the provinces, being
especially serious at Shiraz. The deposed Shah remained for
weeks at the Russian Legation, while negotiations with him for
a pension or allowance in return for his surrender of jewels
and money to the State went on, and the unhappy child who
occupied his palace had more sorrow than he.

Early in August Colonel Liakhoff returned to Russia and was

appointed to a regimental command. On the 1st of September a
general amnesty, with a few exceptions, was proclaimed by the
new government at Teheran. On the 9th of September the deposed
Shah left the shelter of the Russian Legation and journeyed,
with his queen, four younger children and several friends,
under Russian escort, to a residence in Russia, at Odessa,
which was his choice. Persia was still waiting for the able
and much trusted constitutionalist statesman, Nasr-ul-Mulk, to
return from his exile at Paris and accept the offered
premiership in the government; but on the 21st of September
the report went out that he had definitely declined the post.
He returned to Persia, however, in October. On the 11th of
October the Russian Government made known that it had decided
 to withdraw the greater part of the troops it had been keeping
at Tabriz. A new Mejliss, for which the Regent had ordered
elections, was assembled on the 15th of November. On the 7th
of December the Mejliss unanimously approved the proposals of
the Government with regard to borrowing abroad and the
employment of Europeans in executive capacities for the
reorganization of the Finance Department. This, no doubt, will
improve the situation very greatly.

PERSIA: A. D. 1909 (January).

Destructive Earthquake in Luristan.

See (in this Volume)

EARTHQUAKES: PERSIA.

----------PERSIA: End--------

----------PERU: Start--------

PERU: A. D. 1899-1908.
Outline of History.

The leading events of Peruvian history are recorded in Volume

VI. of this work down to the election of President Eduardo de
Romaña, in 1899. "Romaña was a member of a prominent family of
Arequipa, and had been educated in England, at Stonyhurst. He
further had studied for, and taken a degree as, an engineer at
King’s College, London; and whilst he had not acquired much
experience in politics, he nevertheless successfully filled
the Presidential Chair throughout his term. He was alive to
the necessity for the development of the resources of the
country, and, fortunately, his administration was not
embarrassed by disturbances other than some small political
intrigues such as inevitably take place in a country which, as
Peru, was evolving a régime of civil government.
{491}
During this term there was some influx of North American
capitalists, who acquired important interests, in the copper
mines of Cerro de Pasco, and who commenced the construction of
a railway line thereto. … The presidency of Señor Romaña
uneventfully expired at its natural time; elections were held,
and Señor Manuel Candamo, who had already provisionally been
head of the State, was chosen as president in May, 1903.
Candamo had been successful in quieting political animosities
after the revolt against Caceres and in consolidating the
political situation. Peru now showed real evidences of
advancement. The old turbulent element was passing away; those
leaders who had placed purely personal ambition before the
true interests of their country had given place to the natural
talent and ability of the best citizens, whom the times were
calling to the front. Candamo’s rule promised well for the
country. He was surrounded by able men, among whom, as chief
cabinet minister, was Dr. Domingo Almenard, an upright lawyer.
The fiscal revenue was increased by taxes, against which there
were murmurings, but which the country was able to bear, and
the tax on tobacco was set apart for the construction of new
railways. Unfortunately, this able administrator, Señor
Candamo, continued but a short time in office, for he was
overtaken by illness, and died at Arequipa in May, 1904. This
event left the country under the temporary leadership of the
second vice-president, Señor Calderon, for the first
vice-president had died also. An election was at once called
according to law, the two candidates which were put forward
being Dr. Jose Pardo, son of the former president of the same
name, and Señor Nicolas Piérola, who had already been at the
head of the Government on two occasions. Rivalry between the
partisans of these two candidates became acute, and although
it was feared for a moment that some disturbance might occur,
good sense prevailed, and the elections proceeded without
interruption. Both contestants were good men—Piérola
representing the party known as the Democratas, whilst
Pardo headed the Civilistas. There were not very
radical differences of principle underlying these distinctions
of name; both were for civil government and for national
 progress. Piérola had done good work during his former term,
whilst Pardo had the prestige of the good name and
administration of his father, the former president of
1872-1876, and was also held in esteem personally among the
best element of the country. The result of the election—held,
probably, more fairly than ever in Peru before—fell to Dr.
Pardo, who took the presidential scarf and office in
September, 1904, and who still guides the affairs of his
country in a manner which has won the esteem of the nation, in
a general sense.

Dr. Pardo’s Cabinet was formed of some of the most capable men
in the country, prominent among whom was the minister of
Finance, Señor Leguia, to whose work is largely due the
improved financial situation. At the present time—1908—the
best elements of Peru are in the ascendant."

G. Reginald Enock,
Peru: Its Former, and Present Civilization,
History and Existing Conditions,
chapter 9 (Scribner’s Sons, New York).

PERU: A. D. 1901.
Broad Treaty of Arbitration with Bolivia.

See (in this Volume)

ARBITRATION, INTERNATIONAL: A. D. 1901 (NOVEMBER).

PERU: A. D. 1901-1906.
Participation in Second and Third International Conferences
of American Republics.

See (in this Volume)

AMERICAN REPUBLICS.

PERU: A. D. 1903-1909.
 Boundary disputes in the Acre region with Bolivia and Brazil.

See (in this Volume)

ACRE DISPUTES.

PERU: A. D. 1905.
Arbitration Treaties with Colombia and Ecuador.

In a message to the Peruvian Congress, July 28, 1906,

President Pardo communicated treaties of arbitration with
Colombia, one general in its nature, the other special for the
settlement of existing boundary questions. Of the latter the
message said:

"As in former treaties of the same character which have been

heretofore concluded with that Republic, the controversy is
submitted to the decision, to be based upon considerations of
equity, of His Holiness Pope Pius X. But as our question with
Colombia is connected with the one with Ecuador, it has been
agreed that the arbitration with Colombia shall only take
place after the termination of the one in which we are now
proceeding with Ecuador, upon the adjudication by the royal
Spanish arbitrator to Peru of territories which are likewise
claimed by Colombia."

PERU: A. D. 1906.
Decree for the Encouragement of Immigration.

See (in this Volume)

IMMIGRATION AND EMIGRATION: PERU.

PERU: A. D. 1907.
Diplomatic Relations with Chile reëstablished.
The Tacna and Arica questions remaining open.

See (in this Volume)

CHILE: A. D. 1907.
PERU: A. D. 1908-1909.
Seating of President Leguia.
Attempted Revolutions defeated.

On the 27th of May, 1908, Augusto B. Leguia became President,

succeeding Dr. Pardo. Señor Leguia had previously been Premier
and Minister of Finance and Commerce; prior to which he had
been managing director of a great English sugar company in
Peru. A revolutionary movement had been attempted a few weeks
before, in which Dr. Augusto Durand and Isaias Piérola were
engaged, and which suffered defeat.

A year later, on May 29, a similar attempt was announced from

Lima, and ascribed to the same "agitators," who, said the
despatch, "made an assault upon the palace and seized
President Leguia. The army, however, remained loyal and came
to his support. The revolutionists were obliged to liberate
the President, who immediately took measures to put down the
movement. Within an hour, although firing was still heard in
the streets, President Leguia seemed to be master of the
situation. Many shots were exchanged between the troops and
the revolutionists and it is believed that the casualties will
be heavy."

This was contradicted a week later, so far as concerned Dr.

Durand. "It has been proved," said the later statement, "that
the revolutionary outbreak of last week was engineered
entirely by the followers of the Piérola brothers. A committee
of the Liberal party to-day visited President Leguia, and,
declaring that neither Dr. Durand nor José Oliva had taken
part in the movement, requested that these men be set at
liberty. The country is quiet."

----------PERU: End--------

PETER I., King of Servia: His Election.

 See (in this Volume)
BALKAN AND DANUBIAN STATES: SERVIA.

{492}

PETIT, Archbishop Fulbert.

See (in this Volume)

FRANCE: A. D. 1905-1906.

PETROLEUM:
The Supply and the Waste in the United States.

See (in this Volume)

Conservation of Natural Resources.

PETROPALOVSK, SINKING OF THE.

See (in this Volume)

JAPAN: A. D. 1904 (February-August).

PHAGOCYTES: THEIR DEPENDENCE ON OPSONINS.

See (in this Volume)

SCIENCE AND INVENTION, RECENT: OPSONINS.

PHILADELPHIA: A. D. 1905.
A Spasm of Municipal Reform.

See (in this Volume)

MUNICIPAL GOVERNMENT.

PHILADELPHIA: A. D. 1909.
Defeat of Reform.
 See (in this Volume)
MUNICIPAL GOVERNMENT.

----------PHILIPPINE ISLANDS: Start--------

PHILIPPINE ISLANDS:
Gains to Spain from their Loss.

See (in this Volume)

SPAIN: A. D. 1898-1908.

PHILIPPINE ISLANDS: A. D. 1900-1902.

The Stamping Out of the Bubonic Plague.

See (in this Volume)

PUBLIC HEALTH.

PHILIPPINE ISLANDS: A. D. 1901.

Second Report of the Second Philippine Commission.
Collapse of the Insurrection.
Peace in all but five Provinces.
Organization of Provincial Governments.
Native Appointments.
Central Civil Government.
Appointment of Governor Taft.
Filipino Members added to Commission.

Down to the capture of Aguinaldo, leader of the Filipino

insurgents, on the 23d of March, 1901, and his submission to
"the sovereignty of the United States throughout the
Philippine Archipelago," as announced in an address to his
countrymen on the 19th of April, the history of American rule
in those islands is recorded in Volume VI. of this work. The
Second Philippine Commission, with the Honorable William H.
Taft at its head, had entered on the performance of its
extensive legislative duties on the 1st of the previous
September, while the Military Governor continued to exercise
administrative powers. The Commission had begun the
organization of provincial and municipal governments, and the
establishing of a system of public schools, as related in the
Volume referred to. From its second report, covering ten
months and a half, ending on the 15th of October, 1901, the
following statements are drawn, to continue the outline of
principal events and most important affairs down to that date:

"The collapse of the insurrection came in May, after many

important surrenders and captures, including that of
Aguinaldo. Cailles, in Laguna, surrendered in June, and
Belarmino, in Albay, on July 4.

"There are four important provinces in which the insurrection

still continues, Batangas, Samar, Cebu, and Bohol. Parts of
Laguna and Tayabas adjoining Batangas in the mountain region
are affected by the disturbances in Batangas. In Mindoro also,
a thinly settled and almost unexplored island, there are
insurrectos. … Outside of the five provinces named there is
peace in the remainder of the archipelago. …

"The work of the commission since it began to legislate in

September, 1900, has been constant. … We have passed since our
last report, in addition to numerous appropriation bills, a
municipal code, a provincial law, a school law, a law
prescribing an accounting system, acts organizing the various
bureaus of the central government, acts organizing the courts,
an act to incorporate the city of Manila, a code of civil
procedure for the islands, and a new tariff act. …

"The general provincial law provides for a provincial

government of five officers—the governor, the treasurer, the
supervisor, the secretary, and the fiscal, or prosecuting
attorney. The governing board is called the provincial board,
and includes as members the governor, the treasurer, and the
supervisor. The prosecuting attorney is the legal adviser of
the board and the secretary of the province is its secretary.
 The first function of the provincial government is to collect,
through the provincial treasurer, all the taxes, with few
exceptions, belonging to the towns or the province. Its second
and most important function is the construction of highways
and bridges and public buildings. Its third function is the
supervision, through the governor and the provincial
treasurer, of the municipal officers in the discharge of their
duties. Within certain limitations, the provincial board fixes
the rate of levy for provincial taxation.

"The governor has the power to suspend any municipal officer

found failing in his duty, and is obliged to visit the towns
of the province twice in a year, and hear complaints against
the municipal officers. … Under the act the offices are all to
be filled at first by appointment of the commission. The
governor holds his office until February, 1902, when his
successor is to be elected in a mass convention of the
municipal councilors of the towns of the province. The
secretary, treasurer, and supervisor after February next are
brought under the civil-service act, and all vacancies
thereafter arising are to be filled in accordance with the
terms of that act. The fiscal is appointed for an
indeterminate period, and is not subject to the civil-service
law. …

"The commission reached the conclusion that it would aid in

the pacification of the country; would make the members of
that body very much better acquainted with the country, with
the people, and with the local conditions, and would help to
educate the people in American methods, if the commission went
to the capital of each province and there passed the special
act necessary to create the provincial government and made the
appointments at that time. Accordingly, the commission visited
thirty-three provinces. …

{493}
"The policy of the commission in its provincial appointments
has been, where possible, to appoint Filipinos as governors
and Americans as treasurers and supervisors. The provincial
secretary and the provincial fiscal appointed have uniformly
been Filipinos. It will be observed that this makes a majority
of the provincial board American. The commission has, in
several instances, appointed to provincial offices former
insurgent generals who have been of especial aid in bringing
about peace, and in so doing it has generally acted on the
earnest recommendation of the commanding officer of the
district or province. We believe the appointments made have
had a good effect and the appointees have been anxious to do
their duty. …

"The central government of the islands established in

September, 1900, under the instructions of the President, with
a military governor as chief executive and the commission as
the legislative body with certain executive functions in
addition, continued until the 4th of July, 1901. At that time
Major General Adna R. Chaffee relieved Major-General MacArthur
as commanding general of this division and military governor.
By the order of June 21, previous, in all organized provinces
the civil executive authority theretofore reposed in the
military governor and in the commission was transferred on
July 4 to a civil governor. The president of the commission
was designated as civil governor. …

"By an order taking effect September 1, the purport of which

was announced the 4th day of July, there were added to the
commission, as a legislative body, three Filipinos, Dr. T. H.
Pardo de Tavera, Señor Benito Legarda, and Senor José
Luzuriaga. These gentlemen, the first two of them residents of
Manila and the last a resident of the island of Negros, had
been most earnest and efficient in bringing about peace in the
islands. Dr. Tavera was the first president of the Federal
party, had accompanied the commission in its trips to the
southern provinces, and was most useful in the effective
speeches which he delivered in favor of peace and good order
at every provincial meeting. Señor Legarda had been valuable
in the extreme to General Otis and to all the American
authorities by the wisdom of his suggestions, and the courage
and earnestness with which he upheld the American cause as the
cause most beneficial to his country. Señor José Luzuriaga was
a member of the first government of the island of Negros,
organized while there was insurrection rife throughout the
islands, as an independent government under the supervision of
a military governor, and was most active in preventing the
insurrection from gaining any foothold in that important
island. …

"The theory upon which the commission is proceeding is that

the only possible method of instructing the Filipino people in
methods of free institutions and self-government is to make a
government partly of Americans and partly of Filipinos, giving
the Americans the ultimate control for some time to come. In
our last report we pointed out that the great body of the
people were ignorant, superstitious, and at present incapable
of understanding any government but that of absolutism. The
intelligence and education of the people may be largely
measured by knowledge of the Spanish language. Less than 10
per cent of the people speak Spanish. With Spaniards in
control of these islands for four hundred years and with
Spanish spoken in all official avenues, nothing could be more
significant of the lack of real intelligence among the people
than this statement. The common people are not a warlike
people, but are submissive and easily—indeed much too
easily—controlled by the educated among them, and the power of
an educated Filipino politically ambitious, willing to plot
and use all the arts of a demagogue in rousing the people, is
quite dangerous. The educated people themselves, though full
of phrases concerning liberty, have but a faint conception of
what real civil liberty is and the mutual self-restraint which
is involved in its maintenance. They find it hard to
understand the division of powers in a government, and the
 limitations that are operative upon all officers, no matter
how high. In the municipalities, in the Spanish days, what the
friar did not control the presidente did, and the people knew
and expected no limit to his exercise of authority. This is
the difficulty we now encounter in the organization of the
municipality. The presidente fails to observe the limitations
upon his power, and the people are too submissive to press
them. In this condition of affairs we have thought that we
ought first to reduce the electorate to those who could be
considered intelligent, and so the qualifications for voting
fixed in the municipal code are that the voter shall either
speak, read, and write English or Spanish, or that he shall
have been formerly a municipal officer, or that he should pay
a tax equal to $15 a year or own property of the value of
$250."

Report of the U. S. Philippine Commission,

from December 1, 1900, to October 15, 1901,
part 1, pages 7-20.

PHILIPPINE ISLANDS: A. D. 1901-1902.

Report of Governor Taft.
Civil Government established in
all Christian Filipino Territory.
The Moros.
Destruction of the Carabao.
Cholera.
Ladrones.
The Native Constabulary.

"When our last report was submitted there was insurrection in

the province of Batangas, where the insurgent forces were
commanded by General Malvar, and in the adjacent provinces of
Tayabas and Laguna; in the province of Samar, where the
insurgent forces were commanded by General Lukban; in Cebu,
where the insurgent forces were under the insurgent leaders
Climaco and Maxilom; in Bohol, where the insurgent forces were
 commanded by the insurgent leader Samson; and in the island of
Mindoro. Vigorous campaigns were begun in November and
December by General Bell, in Batangas, Laguna, Tayabas, and
Mindoro, by General Smith in Samar, and by General Hughes in
Cebu and Bohol. In November and December the insurgents in
Cebu and Bohol surrendered, and conditions of peace were so
completely established that the Commission soon after received
the province of Cebu from the military authorities, and by act
numbered 322, passed December 20, 1901, restored the civil
government in that province to take effect January 1, 1902; in
Bohol the province was delivered over to the Commission early
in 1902, and the commission, by act of March 3, 1902, restored
civil government there to take effect April 1, 1902. General
Lukban, in Samar, was captured in February, 1902, and the
entire force of insurgents in that island under General
Guevara surrendered in April following.

{494}

"By an act passed June 17, 1902, Number 419, the Commission
organized the province of Samar, and established civil
government there. In April of 1902, General Malvar surrendered
with all his forces in Batangas, and by act passed June 23,
1902, the Commission restored civil government to that
province to take effect July 4, 1902. By act Number 424,
enacted July 1, 1902, the province of Laguna was organized
into a civil government. This completed the organization of
all the provinces in which insurrection had been rife during
the latter part of 1901, except Mindoro. There were, in
addition, certain tracts of territory occupied by Christian
Filipinos that had not received civil government, either
because of the remoteness of the territory or the scarcity of
population." The report then details the measures by which
civil government was given to these tracts of territory, and
proceeds:

"The question what shall be done with respect to Mindanao is

one which has not been definitely decided, first, because so
much has had to be done with respect to the northern and
Filipino provinces, and, second, because at present there is
an unsettled condition in the Lake Lanao country. The
hostility to the Americans does not reach beyond the Lake
Lanao Moros. The Moros of the Jolo group, of Zamboanga, and of
the Rio Grande de Mindanao Valley are all quiet, and all
entirely willing to submit to American supervision. It is very
possible that an arrangement can be brought about by which the
Sultan of Jolo can be induced to part with such rights as he
claims to have in the Jolo Archipelago, and in this way
questions which now present very perplexing difficulties with
respect to ownership of privileges, rights, and lands may be
obviated. … I think it wiser on the part of the Commission to
postpone the consideration of the Moro question until we have
passed legislation to meet needs that are more pressing
throughout the northern part of these possessions of the
United States. For a great many years to come there will be no
question of popular government in the Moro country; the Moros
do not understand popular government, do not desire it, and
are entirely content with the control by their dattos.
Possibly far in the future the control by dattos will cease.
There is room for material and industrial development among
the Moros, and with their material improvement may come a
change in their political views. For the present, however, it
is necessary only to provide a paternal, strong, but
sympathetic government for these followers of Mohammed.

"The civil government has assumed responsibility for the

preservation of order and the maintenance of law throughout
the Christian Filipino territory of this archipelago at a time
when the material conditions are most discouraging and present
every conceivable obstacle to the successful administration of
the affairs of 6,000,000 or 7,000,000 people. The war of six
years since 1896 has greatly interfered with the regular
pursuit of agriculture, which is almost the only source of
wealth in the islands. Many years ago there was sufficient
rice raised in the islands not only to feed the people but to
export it to other countries. For a number of years before the
American occupancy rice had been imported. The area of
cultivation of the rice has been much lessened during the war
and many fields which were formerly tilled are grown now with
the cogon grass because of neglect.

"The greatest blow to agriculture has been the loss of the

carabao or water buffalo, upon which the cultivation of rice,
according to the mode pursued in these islands, is wholly
dependent. The war in some degree, and the rinderpest in a
much larger degree, have destroyed about 90 per cent of the
carabaos; and the natives—never very active in helping
themselves—have simply neglected the rice culture, so that
now the islands are compelled to spend about $15,000,000 gold
to buy food upon which to live. The carabao is not so
necessary in the cultivation of the sugar crop or in the
cultivation of hemp. …

"The cholera has swept over these islands with fatal effect,
so that the total loss will probably reach 100,000 deaths.
Whole villages have been depopulated and the necessary
sanitary restrictions to avoid its spread have interfered with
agriculture, with intercommunication, and with all business.
The ravages of war have left many destitute, and a guerrilla
life has taken away from many all habits of industry. With no
means of carrying on agriculture, which is the only occupation
of these islands, the temptation to the less responsible of
the former insurgents after surrender to prey upon their
neighbors and live by robbery and rapine has been very great.
The bane of Philippine civilization in the past was ladronism,
and the present conditions are most favorable for its growth
and maintenance. … Many who were proscribed for political
offences in the Spanish times had no refuge but the mountains,
and being in the mountains conducted a free robber life, and
about them gathered legions not unlike those of the Robin Hood
days of England, so that they attracted frequently the
 sympathy of the common people. In the Spanish days it was
common for the large estate owners, including the friars, to
pay tribute to neighboring ladrones. Every Tagalog province
had its band of ladrones, and frequently each town had its
recognized ladrone whom it protected and through whom it
negotiated for immunity. …

"The insurrection is over. It is true that the ladrones,

though they live on nothing but cattle and rice stealing, and
never attack American soldiers, and prey only upon their own
people, do masquerade as insurrectos; but they recognize no
authority and have no characteristics other than those of
banditti. They have stirred up in some of the provinces the
organization of so-called secret societies for the purpose of
securing agencies with which successfully to conduct their
robbery and to sell the fruits of it. … The picture that I
have given of the depressed condition of agriculture, and the
tendency to ladronize in the Tagalog provinces and in some of
the Visayan provinces, does not apply to those provinces in
which hemp is the chief product. They are wealthy and
prosperous."

Report of Governor W. H. Taft

(Report of the Philippine Commission, 1902, part 1).

{495}

PHILIPPINE ISLANDS: A. D. 1902.

Padre Aglipay’s Secession from the Roman Catholic Church.
Organization of the Independent Filipino Catholic Church.

"Gregorio Aglipay is an Ilocano, and was an ordained priest of

the Roman Catholic Church in these islands before the
insurrection. During the insurrection he continued his
priestly functions at Mabolos and took such action as to bring
him into conflict with the hierarchy of the Church. What the
merits of this controversy were I do not know. Subsequently he
assumed the leadership of the insurrecto forces in Ilocos
Norte and carried on a very active campaign in the mountains
of that province. He was one of the last of the leaders to
surrender with his forces in North Luzon. Since his surrender
he has been quite active in spreading propaganda among the
native priests against the so-called Friar domination of the
church in these islands. The definite refusal of the Vatican
to withdraw the Spanish friars from the islands was made the
occasion for the formation of the Independent Filipino
Catholic Church. Actively engaged with Aglipay in this
movement was Isabelo de los Reyes, the former editor of an
insurrecto paper, published in Madrid, called Filipinas ante
Europa, and an agitator of irresponsible and irrepressible
character. … Padre Aglipay has secured the active and open
cooperation of a number of native priests, 15 of whom he has
appointed bishops, himself having the title of archbishop. He
has held mass in many different places in and about Manila;
his services have attracted large gatherings of people. …

"In order to prevent constant recurrence of disturbances of

the peace I have had to take a firm stand with the leaders of
the movement by impressing upon them that forcible
dispossession of a priest of the Roman Catholic Church, for
years in peaceable possession of the church and the rector’s
house, is contrary to law, and would be prevented by the whole
police power. The leaders of the movement assure me that they
have no desire to violate the law and wish to keep within it,
but that their followers at times are hard to control. I have
said to them that if they claim title to the churches they may
assert it through the courts, and if successful will secure
not only the confirmation of their title but actual
possession. …

"I have taken occasion to say, whenever an opportunity

occurred, that the insular government desired to take no part
whatever in the religious controversies thus arising; that it
would protect Father Aglipay and his followers in worshiping
 God as they chose just as it would protect the Roman Catholic
Church and its ministers and followers in the same rights. But
that, if the law was violated by either party, it would become
the duty of the government to step in and restrain such
lawlessness."

Governor Wm. H. Taft,

Report, 1902, pages 39-40.

PHILIPPINE ISLANDS: A. D. 1902-1903.

Governmental Purchase of the Friars’ Lands.

"As early as 1898, the Peace Commission, which negotiated the

treaty of Paris, became convinced that one of the most
important steps in tranquilizing the islands and in
reconciling the Filipinos to the American Government would be
the governmental purchase of the so-called friars’
agricultural lands in the Philippines, and the sale of these
lands to the tenants upon long, easy payments. … The Secretary
of War and the President concurred in the recommendations of
the Commission. Accordingly in May, 1902, the writer, as civil
governor of the Philippine Islands, was directed by the
Secretary of War to visit Rome and to confer with the Pope or
such agents as he might designate in respect to the question
of buying the friars’ agricultural lands and other questions
of a similar character which were pending between the Roman
Catholic Church and the Government. The negotiations which
were had on this subject in Rome were set forth in the
correspondence published by the Secretary of War in his report
to Congress for last year. In a word, the Pope approved the
purchase of the agricultural lands of the three great
religious orders that owned agricultural lands in the islands
and appointed an apostolic delegate with as full powers as he
could be invested with to bring about this result. …

"In order to determine the value of the estates, the

representatives of the various companies and other interests
were invited to attend a hearing, when various witnesses were
called to testify. The apostolic delegate was also present. …

"In accordance with the agreement reached in Rome, I sent to

the apostolic delegate a request for a statement of the exact
interests retained by the religious orders in the Philippines
in the lands which were the subject of negotiation. No formal
answer to this letter was ever received, but informally it was
stated to me by the delegate that the authorities in the
Philippines had informed him that they had so disposed of
their interests that they were unable to make a statement of
what their interests were, if any. The value of the lands, as
estimated according to the statements of the agents of the
companies, aggregated a sum between thirteen and fourteen
millions of dollars gold. The estimate of Villegas, the
surveyor employed by the Commission, showed the valuation of
the lands to be $6,043,000 gold, if his value in Mexican
should be reduced to gold at the rate of two to one, which was
the gold rate about the time of his survey and classification,
though the Mexican dollar fell considerably after that.
Considering the bad conditions which prevailed in agriculture,
the loss of cattle, the dispute concerning title, and the
agrarian question that must always remain in the management of
these estates and embarrass the owner, I considered—and I
believe the Commission generally agreed with me—that
$6,043,000 gold was a full price for the lands. The sum,
however, was scouted by the persons representing the owners,
and there appeared to be very little prospect of reaching an
agreement. …

"Not discouraged, however, by circumstances that seemed most

discouraging, the apostolic delegate bent his energies to
bringing the parties to a settlement. After some negotiation
the delegate first stated that he thought he could arrange a
sale for $10,500,000 gold. I told him there was no hope of
bringing about a purchase at that figure. … Then followed a
long and protracted discussion between the parties who were to
 be the venders as to how this sum should be divided, and there
was much difficulty in arriving at a solution—so great a
difficulty, indeed, that I was informed that unless $7,770,000
was paid there was no hope of reaching an agreement. With the
approval of the Secretary of War and the Commission, I replied
that $7,543,000 was our ultimatum, and that we would not give
more than that, and this was ultimately the basis upon which
the price was fixed."

Report of the Civil Governor of the Philippine Islands,

William H. Taft
(Fourth Report of the Philippine Commission).

{496}

PHILIPPINE ISLANDS: A. D. 1905.

Report of Committee on Methods of Dealing with the Sale and
Use of Opium.

See (in this Volume)

OPIUM PROBLEM.

PHILIPPINE ISLANDS: A. D. 1906-1907.

Resignation of Governor Ide.
Appointment and Inauguration of Governor Smith.
Complete Tranquility in the Islands.
Change in the Constitution of Provincial Boards.

"On September 20, 1906, the resignation of the Honorable Henry

Clay Ide as governor-general became effective, and on that
date the Honorable James F. Smith was inaugurated as
governor-general of the Philippine Islands. … Since April of
this year complete tranquility has prevailed in every part of
the archipelago, inclusive of the Moro province. In 21 of the
provinces peace has reigned supreme during the entire year. In
Bataan and Batangas there was some disturbance of the public
order, caused in the case of the first-named province by the
 escape of some provincial prisoners, and in the second by the
operations of six or seven brigands near the boundary line of
the provinces of La Laguna and Tayabas. All of the escaped
prisoners and all of the bandits with the exception of two in
each party have been captured. …

"The convention of provincial governors held in Manila in

October, 1906, recommended that the then existing law
providing that provincial boards shall be composed of a
provincial governor elected by the municipal councilors and
vice-presidents of the various municipalities of the province
and a provincial treasurer and a third member appointed by the
executive be so amended as to permit of the election of the
provincial governor and third member by direct vote of the
people. This recommendation was submitted to the Secretary of
War, and on receiving his approval thereof the provincial
government act was amended accordingly. This innovation in the
constitution and selection of provincial boards has been an
advantage both to the insular and to the local government. On
the one hand it has removed all cause for friction between the
provincial governor elected by the people and the two members
of the board named by the executive. On the other it has
imposed upon the provincial governor and the third member the
responsibility for the well-being of the province and has
removed from the insular government much of the responsibility
for conditions purely of local concern."

Report of the Philippine Commission,

December 31, 1907
(Abridgment, Message and Documents, 1907, pages 799-807).

PHILIPPINE ISLANDS: A. D. 1907.

The Philippine Election Law.
Election of a Popular Assembly.
Political Parties participating in it.
The first meeting of the Assembly.
Presence of Secretary Taft.
His account of the Assembly and of the Parties
represented in it.

"In January, 1907, the Philippine Commission passed the

Philippine election law. In framing this law the election
codes of Massachusetts, New York, the District of Columbia,
and California were consulted and features adopted from each,
modified in such a way as to meet insular conditions and to
avoid the mistakes and abuses that have arisen in some
provincial and municipal elections in the islands. The aim has
been to provide a law sufficiently explicit and not too
complicated for easy comprehension. Every effort has been made
to afford the necessary safeguards and machinery to insure
purity, secrecy, certainty, and expedition, without causing
too great a drain upon the resources of municipal and
provincial governments. The prominent features of this law as
amended are the division of those provinces not inhabited by
Moros or other non-Christian tribes into 78 assembly
districts, each province to constitute at least one district
and the more populous being divided into more districts, in
the ratio of 1 to every 90,000 of population and major
fraction thereof remaining. In accordance with this
apportionment there will be 80 delegates, two of whom will
represent the city of Manila, which is considered as a
province, within the meaning of the act of Congress, and
divided into two districts."

Report of the Chief of the Bureau of Insular Affairs,

October 31, 1907
(Abridgment, Message and Documents, 1907, page 781).

"On the 28th of March, 1907, the Commission by resolution,

unanimously adopted, certified to the President that for two
years following the publication of the census of the islands a
condition of general and complete peace had prevailed and then
existed in the territory of the islands not inhabited by Moros
or other non-Christian tribes. … By virtue of this certificate
 and in accordance with the provisions of the act of Congress of
July 1, 1902, the President on March 28, issued a proclamation
directing the Philippine Commission to call a general election
for the choice of delegates to a popular assembly. Accordingly
on the 30th of March, 1907, the Commission passed a resolution
ordering that an election be held for delegates on July 30 and
directing the governor-general to issue a proclamation
announcing the election for that date. The proclamation was
issued on April 1. By a strange coincidence the day of the
month fixed for holding the election was the same as that on
which the first legislative body in America, the house of
burgesses, met in the year 1619. Under the general election
law the delegates to the assembly elected at the elections
held on July 30th, 1907, and seated by the Philippine
assembly, will serve until January 1, 1910. Subsequent
elections for delegates will be held on the first Tuesday
after the first Monday in November, 1909, and on the first
Tuesday after the first Monday in November in each
odd-numbered year thereafter, delegates to take office on the
1st day of January next following their election and to hold
office for two years, or until their successors are elected
and qualified.

"The basis of representation in the Philippine assembly is one

delegate for every 90,000 of population and one additional
delegate for a major fraction thereof: Provided, however, that
each Christian province shall be entitled to at least one
delegate and that the total number of delegates shall at no
time exceed 100. Provinces entitled to more than one delegate
are divided into districts. The law declares Manila to be a
province within the meaning of the act of Congress authorizing
the assembly, and, it is allowed the same representation as
other provinces. Thirty-four provinces are represented in the
Philippine assembly, which is composed of 80 members.

{497}
"The act of Congress requires that delegates to the assembly
shall be qualified electors of the election district in which
they may be chosen, 25 years of age, and owing allegiance to
the United States. The act of Congress prescribes that the
qualifications of electors shall be the same as those
prescribed for electors in municipal elections under laws in
force at the time of the passage of the Congressional
enactment. As the municipal election laws in force at the time
of the passage of the act of Congress have undergone some
change in regard to the qualifications of electors, the
strange anomaly is presented of having certain
qualifications exacted from municipal and provincial officials
which are not required for delegates to the assembly. One of
the results is that felons, victims of the opium habit, and
persons convicted in the court of first instance for crimes
involving moral turpitude, but whose cases are pending on
appeal, are not eligible for election to any provincial or
municipal office, but may become delegates to the assembly.

"As announced by provincial governors the elections for

assemblymen held on the 30th of July, 1907, resulted in the
election of 32 Nacionalistas, 4 Independistas, 7
Inmediatistas, 16 Progresistas, 20 Independents, and 1 Centro
Catolico. The total number of voters registered for the
assembly elections was 104,966. The number of voters
registered for the provincial and municipal elections will be
very much larger than that for the assembly elections. The
difference in registration and votes cast at the two elections
seems to show with considerable certainty that there was far
more interest in the elections for provincial and municipal
officials than there was in the election for assemblymen. …

"The delegates to the Philippine assembly, in accordance with

the call of the governor-general as prescribed by the act of
Congress, met at the Grand Opera House in the city of Manila
on the 16th day of October at 9 o’clock A. M."
 Report of the Philippine Commission,
December 31, 1907
(Abridgment, Message and Documents, 1907, pages 810-811).

The Honorable William H. Taft, United States Secretary of War,

former Governor-General of the Philippine Islands, made the
long journey to the Islands on this occasion for the purpose
of opening the meeting of the Assembly and personally
inspecting the state of affairs. After returning, in the
following December, he made an extended report to the
President, in which he discussed the character of the Assembly
and of the parties represented in it at considerable length.
Recurring to the formation of the first political party that
arose in the Islands after they came under the control of the
United States, he said of it:

"It is a mistake to suppose that the war by the Filipinos

against the Americans had the sympathy of all the Filipinos.
On the contrary, there were many intelligent and conservative
men who favored American control and who did not believe in
the capacity of their people immediately to organize a
government which would be stable and satisfactory, but in the
face of a possible independence of the Islands, they were
still. Upon Mr. McKinley’s second election many of these
persons reached the conclusion that it was time for them to
act. Accordingly, they formed the Federal Party, the chief
platform of which was peace under American sovereignty and the
acceptance of the American promises to govern the Islands for
the benefit of the Filipinos and gradually to extend popular
self-government to the people. The Federal Party received
accessions by thousands in all parts of the Islands and in
every province, so that the Commission was enabled during the
year 1901, and under the auspices, and with the aid of, the
Federal Party, to organize civil government in some 32 or 33
provinces, or in substantially all of them. … The main purpose
and principle of the party was peace under the sovereignty of
the United States. In drafting a platform its leaders had
formulated a plank favoring the organization of the Islands
into a Territory of the United States, with a view to its
possibly becoming a State. From this plank it took its name.
In the first two or three years after its successful effort to
bring on peace, many prominent Filipinos having political
ambition became members, and in the gubernatorial elections
the great majority of governors elected were Federals. And so
substantially all who filled prominent offices in the
government by appointment, including the judges, were of that
party. Then dissension arose among prominent leaders and some
withdrew from the party. The natural opposition to a
government party led to the organization of other parties,
especially among those known as Intransigentes
[Irreconcilables]. The Federal Party had founded an organ, the
Democracia, early in its existence. The opponents of the
government looking to immediate independence founded a paper
called the Renacimiento. The latter was edited with especial
ability and with a partisan spirit against the American
Government.

"For two years before the election of the Assembly the

Filipinos who sympathized with the Renacimiento were
perfecting their organization to secure a majority in the
assembly. Many groups were formed, but they all were known as
the Partido Nacionalista. There was some difference as to
whether to this title should be added the word ‘inmediatista,’
but the great majority favored it. The party is generally
known as the Nacionalista Party. During much of these same two
years, the Federal Party was dormant. …

"Some six months before the elections, there sprung from the
ashes of the Federal Party a party which, rejecting the
statehood idea, declared itself in favor of making the
Philippines an independent nation by gradual and progressive
acquisition of governmental control until the people should
become fitted by education and practice under American
sovereignty to enjoy and maintain their complete independence.
 It was called the Partido Nacionalista Progresista. It is
generally known as the Progresista Party. …

{498}

"The campaign in the last two or three months was carried on

with great vigor. The Nacionalistas had the advantage of being
understood to be against the government. This, with a people
like the Filipino people, who had been taught to regard the
government as an entity separate from the people, taxing them
and prosecuting them, was in itself a strong reason for
popular sympathy and support. The Progresistas were denounced
as a party of office-holders. The government was denounced as
extravagant and burdensome to the people. In many districts
the Nacionalista candidates promised that if they were
returned immediate independence would follow. There were quite
a number of candidates in country and remote districts where
the controversy was not heated who did not declare themselves
on the main question, and maintained an independence of any
party. They were known as Independientes. Then, there were
other Independientes who declared themselves independent of
party, but in favor of immediate independence.

"The total vote registered and cast did not exceed 104,000,
although in previous gubernatorial elections the total vote
had reached nearly 150,000. The high vote at the latter
elections may be partly explained by the fact that at the same
elections town officers were elected, and the personal
interest of many candidates drew out a larger number of
electors. But the falling off was also in part due, doubtless,
to the timidity of conservative voters, who, because of the
heat of the campaign, preferred to avoid taking sides. This is
not a permanent condition, however, and I doubt not that the
meeting of the assembly and the evident importance of its
functions when actually performed will develop a much greater
popular interest in it, and the total vote will be largely
increased at the next election.
"I opened the assembly in your name. The roll of the members
returned on the face of the record was called. An appropriate
oath was administered to all the members and the assembly
organized by selecting Señor Sergio Osmeña as its speaker or
presiding officer. Señor Osmeña has been one of the most
efficient fiscals, or prosecuting attorneys, in the Islands,
having conducted the government prosecutions in the largest
province of the Islands, the province and Island of Cebu. He
was subsequently elected governor, and by his own activity in
going into every part of the island, he succeeded in enlisting
the assistance of all the people in suppressing ladronism,
which had been rife in the mountains of Cebu for thirty or
forty years, so that to-day there is absolute peace and
tranquillity throughout the island. He is a young man, not 30,
but of great ability, shrewdness, high ideals, and yet very
practical in his methods of dealing with men and things. The
assembly could have done nothing which indicated its good
sense so strongly as the selection of Senor Osmeña as its
presiding officer. …

"As a shibboleth—as a party cry—immediate independence has

much force, because it excites the natural pride of the
people; but few of their number have ever worked out its
consequences, and when they have done so they have been
willing to postpone that question until some of the immediate
needs of the people have been met. I may be wrong, but my
judgment is that the transfer of real power, by giving to the
people part of the legislative control of the Christian
provinces, sobers their leaders with the sense of
responsibility and teaches them some of the practical
difficulties of government I do not for a moment guarantee
that there will not at times be radical action by the
Assembly, which cannot meet the approval of those who
understand the legislative needs of the Islands, but all I
wish to say is that the organization and beginning of the life
of the Assembly have disappointed its would-be critics and