Dietary Reference Intakes for Energy

180 DIETARY REFERENCE INTAKES FOR ENERGY

for weight management. In addition to her U.S. work she coleads an international consortium of
scientists dedicated to addressing obesity worldwide. Dr. Roberts has published over 250
research papers and has an H–index of 63. She has been the awardee of preeminent awards for
national and international nutrition research including the 2009 E.V. McCollum Award of the
American Society for Nutrition and the 2016 W. O. Atwater Lecturer. Dr. Roberts completed her
Ph.D. in nutrition at the University of Cambridge in the U.K. and her postdoctoral training at the
Massachusetts Institute of Technology. Dr. Roberts was a member of the 2002 National
Academies’ Committee on Dietary Reference Intakes for Energy and Macronutrients and was a
member of the 2016 Committee to Review the Process to Update the Dietary Guidelines for
Americans.
Heidi J. Silver, Ph.D., is a Research Professor at Vanderbilt University Medical Center and a
Health Scientist in the Veteran’s Affairs Department of Research. Dr. Silver’s research focuses
on designing diet intervention trials that modify energy intake and the amounts and types of
macronutrients consumed to improve energy balance, body composition, inflammatory state, and
insulin sensitivity for cardiometabolic disease risk reduction. She established and directs the
Vanderbilt Diet, Body Composition, and Human Metabolism Core. In 2020, she was selected for
the Academy of Nutrition and Dietetics Excellence in Research Practice Award. She has
published over 50 peer-reviewed articles, taught several academic courses, and created several
webinars. Dr. Silver is an ad hoc reviewer for 21 journals, has been invited to speak at more than
75 national meetings, and has been invited to present lectures or workshops in 9 different
countries. Dr. Silver achieved her Ph.D. in nutrition in 2001 from Florida International
University where she was honored with Doctoral Recognition of the Year and Outstanding
Doctoral Scholarship Awards.
Janet A. Tooze, Ph.D., is Professor in the Department of Biostatistics and Data Science,
Division of Public Health Sciences, at the Wake Forest School of Medicine. She is a
biostatistician with expertise in statistical methods in nutrition, focused on dietary assessment
and measurement error. She has developed methods for estimating the usual intake of foods and
nutrients in a unified framework, termed the NCI Method, the foundation of which is a statistical
model developed by Dr. Tooze for repeated measures data with excess zeroes. This method is
used internationally to characterize population intakes of foods and nutrients and for risk
assessment. She led the statistical validation of the Healthy Eating Index-2015, a widely used
diet quality index. She has received three National Institutes of Health Merit Awards in
recognition of her work in the advancement of dietary assessment. Dr. Tooze received an M.P.H.
in public health from the Harvard School of Public Health and a Ph.D. in biometrics from the
University of Colorado. She was a member of the 2017–2019 National Academies’ Committee
to Review the Dietary Reference Intakes for Sodium and Potassium.
William W. Wong, Ph.D., is Distinguished Emeritus Professor of Pediatrics at Baylor College
of Medicine and Past Director of the Gas-Isotope-Ratio Mass Spectrometry Laboratory and
Chairman of the Center-wide Equipment Maintenance/Repair Program at the USDA/ARS
Children’s Nutrition Research Center. Other than performing the first whole-room indirect
calorimetric validation of the doubly labeled water (DLW) method, he was involved in many
studies defining the energy requirements in infants, toddlers, adolescents, pregnant and lactating
women, and women with twin pregnancy as well as adolescents with heart failure and cancer. In

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX B 181

addition to his expertise in the stable isotope methods, he was the project director of a multisite,
randomized, placebo-controlled clinical study to document the efficacy and safety of soy
isoflavones to prevent osteoporosis in menopausal women, as well as the project director of
Healthy Kids Houston, a community-based program to promote healthy lifestyles among
minority children with support from the City of Houston Parks and Recreation Department, the
Houston Metropolitan Transit Authority, YMCA, and the Houston Independent School District.
He also led a team of pediatricians, dietitians, nutritionists, and a psychotherapist to develop the
summer camp program, Kamp K’aana, to promote healthy lifestyles among obese children. The
program is now an official program at the YMCA in Houston and Wisconsin. He was one of the
original key scientists to help establish the International Atomic Energy Agency DLW Database.
Dr. Wong will serve on the Data and Safety Monitoring Board for the Nutritional Interventions
Planning Projects of the National Institute of Aging. He received his B.S. degree in Chemistry
and his M.S. and Ph.D. degrees in Oceanography. In spite of his lack of official training in
nutrition, biochemistry, and human physiology, he was able to develop research projects with
diverse research interests and worked effectively in a multidisciplinary setting.

Elizabeth A. Yetley, Ph.D., retired from the Office of Dietary Supplements (ODS) at the
National Institutes of Health (NIH) in 2008 having served as Senior Nutrition Research Scientist
for 4 years. Subsequently, she was contracted by ODS for the next 9 years to work on specific
projects that were of interest to the organization. From 1980 to 2004, she worked as a nutrition
scientist at the U.S. Food and Drug Administration (FDA), eventually attaining the rank of Lead
Nutrition Scientist. Dr. Yetley provided leadership for several projects at both NIH and FDA that
included health claims for nutrition labels, folic acid fortification, methodological challenges for
assessing folate and vitamin D biomarkers of status, and systematic reviews for Dietary
Reference Intakes, as well as other nutrition topics such as vitamin D and omega-3 fatty acids.
She also provided regulatory leadership for infant formulas, medical foods, and dietary
supplements. She received numerous awards from organizations including FDA, NIH, the Health
and Human Services’ Secretary, the American Society for Nutrition, the University of
Massachusetts, and Iowa State University. She received her B.S., M.S., and Ph.D. in nutrition
from Iowa State University. Dr. Yetley was a member of the 2017–2019 National Academies’
Committee to Review the Dietary Reference Intakes for Sodium and Potassium.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

Appendix C
Open-Session Agendas

February 4, 2022
Session 1

10:00 a.m. Introductions and Chair’s Statement

Anna Maria Siega-Riz, Committee Chair
10:10 Use of Doubly Labeled Water to Support Analysis of Energy Requirements
John Speakman, Institute of Biological and Environmental Sciences, University of
Aberdeen, Scotland
10:30 Q&A
Amy Luke, Committee Member
10:50 Physical Activity and Long-Term Body Weight Regulation
John Jakicic, Translational Research Institute at AdventHealth Orlando
11:10 NHANES and Physical Activity
Richard Troiano, Epidemiology and Genomics Research Program, National
Cancer Institute
11:30 Q&A
Susan Roberts, Committee Member
12:00 pm Break

Session 2

1:00 p.m. Introductions

Anna Maria Siega-Riz, Committee Chair
1:05 Impact of Dietary Macronutrient Content on Energy Metabolism
Kevin Hall, Integrative Physiology Section, National Institute of Diabetes and
Digestive and Kidney Diseases
1:45 Q&A
Heidi Silver, Committee Member
2:15 Closing Remarks
Anna Maria Siega-Riz, Committee Chair
2:30 pm Adjourn

March 31, 2022

PREPUBLICATION COPY: UNCORRECTED PROOFS

183

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

184 DIETARY REFERENCE INTAKES FOR ENERGY

12:30 p.m. Energy Values for Human Milk

Kellie Casavale, U.S. Food and Drug Administration
Kathryn Hopperton, Health Canada
12:50 Q&A
1:00 Energy Requirements to Support Pregnancy Weight Gain and Lactation at
Varying BMIs
Leanne Redman, Pennington Biomedical Research Center
1:20 Q&A
1:30 Adjourn Open Session

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

Appendix D
Literature Search Strategies and Results

Key Question: What is the association between body mass index (BMI) and chronic disease,
including all-cause mortality?
Date: March 24, 2022

Search Parameters:
Date: 2017–Present
Document Type: Systematic reviews
Language: English
Database: Embase (Ovid), Medline (Ovid), Cochrane Database of Systematic Reviews (Ovid)

Embase (Ovid):
Search No. Syntax Results
1 *body mass/ or *obesity/ or *underweight/ or *body weight/ or (body-mass 1,127,797
or BMI or body-mass-index or body-ban-mass or quetelet-index or z-score
or zscore or obesity or adipose-tissue-hyperplasia or adipositas or adiposity
or alimentary-obesity or body-weight-excess or excess-body-weight or
corpulency or fat-overload-syndrome or nutritional-obesity or obesitas or
overweight or thinness or weight-insufficiency or normal-weight or body-
weight or weight-body or weight-status).ti,ab.
2 *cardiovascular disease/ or *congenital heart disease/ or *non insulin 4,302,981
dependent diabetes mellitus/ or *malignant neoplasm/ or hypertension/ or
*hip fracture/ or *all cause mortality/ or *maternal mortality/ or
*preeclampsia/ or *infant mortality/ or *premature labor/ or (cardiovascular-
disease* or angiocardiopathy or angiocardiovascular-disease* or
cardiovascular-complication* or cardiovascular-disturbance* or
cardiovascular-lesion* or cardiovascular-syndrome* or cardiovascular-
vegetative-disorder* or complication-cardiovascular* or disease-
cardiovascular or major-adverse-cardiovascular-event* or congenital-heart-
disease* or congenital-cardiac-disease* or congenital-cardiac-distress* or
congenital-heart-distress* or congenital-heart-failure* or heart-congenital-
disease* or neonatal-cardiopathy or truncus-arteriosus-persistent* or adult-
onset-diabetes or diabetes-mellitus-type-2 or diabetes-mellitus-type-ii or
diabetes-mellitus-maturity-onset or diabetes-mellitus-non-insulin-dependent
or diabetes-type-2 or diabetes-type-II or diabetes-adult-onset or dm-2 or
insulin-independent-diabetes or ketosis-resistant-diabetes-mellitus or
maturity-onset-diabetes or NIDDM or non-insulin-dependent-diabetes or
non-insulin-dependent-diabetes-mellitus or noninsulin-dependent-diabetes or
T2DM or type2-diabetes or type-II-diabetes or cancer* or malignant-
neoplasia or malignant-neoplasm* or malignant-neoplastic-disease* or
malignant-tumor* or malignant-tumour* or neoplasia-malignant or tumor-
malignant or tumour-malignant or hypertension or blood-pressure-high or

PREPUBLICATION COPY: UNCORRECTED PROOFS

185

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

186 DIETARY REFERENCE INTAKES FOR ENERGY

high-blood-pressure or high-renin-hypertension or hypertensive-disease* or

hypertensive-effect* or hypertensive-effect* or hypertensive-response or
hip-fracture* or broken-hip* or fracture-hip* or fractured-hip* or all-cause-
mortality or maternal-mortality or mortality-maternal or mother-mortality or
preeclampsia or pre-eclampsia or pre-eclamptic-toxaemia or pre-eclamptic-
toxemia or preclampsia or preeclamptic-taxaemia or preeclamptic-toxemia
or toxaemia-preeclamptic or toxemia-preeclamptic or infant-mortality or
infantile-mortality or mortality-infant* or premature-labor or labor-
premature or labour-premature or obstetric-labor-premature or obstetric-
labour-premature or premature-delivery or premature-labour or premature-
obstetric-labor or premature-obstetric-labour or preterm-birth or preterm-
delivery or preterm-labor or preterm-labour or physical-growth).ti,ab.
3 risk factor/ or (risk-factor* or relative-risk*).ti,ab. or risk-of.ti. 1,972,996
4 1 and 2 and 3 117,461
5 limit 4 to (english language and yr="2017 -Current") 29,784
6 limit 5 to "systematic review" 1,374

Cochrane Database of Systematic Reviews (Ovid):

Search No. Syntax Results
1 (body-mass or BMI or body-mass-index or body-ban-mass or quetelet-index 310
or z-score or zscore or obesity or adipose-tissue-hyperplasia or adipositas or
adiposity or alimentary-obesity or body-weight-excess or excess-body-
weight or corpulency or fat-overload-syndrome or nutritional-obesity or
obesitas or overweight or thinness or weight-insufficiency or normal-weight
or body-weight or weight-body or weight-status).ti,ab,kw.
2 (cardiovascular-disease* or angiocardiopathy or angiocardiovascular- 2,258
disease* or cardiovascular-complication* or cardiovascular-disturbance* or
cardiovascular-lesion* or cardiovascular-syndrome* or cardiovascular-
vegetative-disorder* or complication-cardiovascular* or disease-
cardiovascular or major-adverse-cardiovascular-event* or congenital-heart-
disease* or congenital-cardiac-disease* or congenital-cardiac-distress* or
congenital-heart-distress* or congenital-heart-failure* or heart-congenital-
disease* or neonatal-cardiopathy or truncus-arteriosus-persistent* or adult-
onset-diabetes or diabetes-mellitus-type-2 or diabetes-mellitus-type-ii or
diabetes-mellitus-maturity-onset or diabetes-mellitus-non-insulin-dependent
or diabetes-type-2 or diabetes-type-II or diabetes-adult-onset or dm-2 or
insulin-independent-diabetes or ketosis-resistant-diabetes-mellitus or
maturity-onset-diabetes or NIDDM or non-insulin-dependent-diabetes or
non-insulin-dependent-diabetes-mellitus or noninsulin-dependent-diabetes or
T2DM or type2-diabetes or type-II-diabetes or cancer* or malignant-
neoplasia or malignant-neoplasm* or malignant-neoplastic-disease* or
malignant-tumor* or malignant-tumour* or neoplasia-malignant or tumor-
malignant or tumour-malignant or hip-fracture* or broken-hip* or fracture-
hip* or fractured-hip* or all-cause-mortality or maternal-mortality or
mortality-maternal or mother-mortality or preeclampsia or pre-eclampsia or
pre-eclamptic-toxaemia or pre-eclamptic-toxemia or preclampsia or

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX D 187

preeclamptic-taxaemia or preeclamptic-toxemia or toxaemia-preeclamptic or

toxemia-preeclamptic or infant-mortality or infantile-mortality or mortality-
infant* or premature-labor or labor-premature or labour-premature or
obstetric-labor-premature or obstetric-labour-premature or premature-
delivery or premature-labour or premature-obstetric-labor or premature-
obstetric-labour or preterm-birth or preterm-delivery or preterm-labor or
preterm-labour or physical-growth).ti,ab,kw.
3 risk*.ti,ab,kw. 6,674
4 1 and 2 and 3 88
5 limit 4 to last 5 years 40

Medline (Ovid):
Search No. Syntax Results
1 Body Mass Index/ or Overweight/ or Obesity/ or Thinness/ or (body-mass- 525,591
index or bmi or index-body-mass or index-quetelet or quetelet-index or
quetelets-index or zscore or z-score or overweight or obesity or leanness or
thinness or underweight).ti,ab.
2 Cardiovascular Diseases/ or Heart Defects, Congenital/ or Diabetes 4,814,116
Mellitus, Type 2/ or Neoplasms/ or Hypertension/ or Hip Fractures/ or
Maternal Mortality/ or Pre-Eclampsia/ or Infant Mortality/ or Premature
Birth/ or (cardiovascular-disease* or disease-cardiovascular or abnormality-
heart or congenital-heart-defect* or congenital-heart-disease* or defect-
congenital-heart or defects-congenital-heart or disease-congenital-heart or
heart-abnormalit* or heart-defect-congenital or heart-defects-congenital or
heart-disease-congenital or heart-malformation-of or malformation-of-
heart* or adult-onset-diabetes-mellitus or diabetes-maturity-onset or
diabetes-mellitus-adult-onset or diabetes-mellitus-ketosis-resistant or
diabetes-mellitus-maturity-onset or diabetes-mellitus-non-insulin-
dependent or diabetes-mellitus-noninsulin-dependent or diabetes-mellitus-
slow-onset or diabetes-mellitus-stable or diabetes-mellitus-type-2 or
diabetes-mellitus-type-ii or diabetes-type-2 or ketosis-resistant-diabetes-
mellitus or mody or maturity-onset-diabetes or maturity-onset-diabetes-
mellitus or niddm or non-insulin-dependent-diabetes-mellitus or slow-
onset-diabetes-mellitus or stable-diabetes-mellitus or type-2-diabetes or
neoplasm* or cancer* or malignancies or malignancy or neoplasia* or
tumor* or hypertension or blood-pressure-high or blood-pressures-high or
high-blood-pressure* or fractures-hip or fractures-intertrochanteric or
fractures-subtrochanteric or fractures-trochanteric or hip-fractures or
intertrochanteric-fracture* or trochanteric-fracture* or all-cause-mortality
or maternal-mortality or maternal-mortalities or mortalities-maternal or
mortality-maternal or preeclampsia-eclampsia or eph-complex or eph-
gestosis or eph-toxemia* or eclampsia-1-preeclampsia or eclampsia-1s-
preeclampsia or edema-proteinuria-hypertension-gestosis or gestosis-eph or
gestosis-edema-proteinuria-hypertension or gestosis-hypertension-edema-
proteinuria or gestosis-proteinuria-edema-hypertension or hypertension-
edema-proteinuria gestosis or pre-eclampsia or preeclampsia or pregnancy-

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

188 DIETARY REFERENCE INTAKES FOR ENERGY

toxemia* or proteinuria-edema-hypertension-gestosis or toxemia-eph or

toxemia-of-pregnanc* or toxemia-pregnancy or toxemias-pregnancy or
toxemias-eph or infant-mortalit* or mortalities-infant or mortality-infant or
mortalities-neonatal or mortalities-postneonatal or mortality-infant* or
mortality-neonatal or mortality-postneonatal or neonatal-mortalit* or
postneonatal-mortalit* or premature-birth* or birth-premature or birth-
preterm or births-premature or births-preterm or preterm-birth* or
growth*).ti,ab.
3 Risk Factors/ or (risk-factor* or factor-risk or health-correlate* or 2,427,333
population-at-risk or populations-at-risk or risk-of).ti,ab.
4 1 and 2 and 3 105,024
5 limit 4 to (english language and yr="2017 -Current" and "systematic 1,149
review")

Key Question: What is the effect of BMI (and other measures of adiposity) on energy balance
or energy expenditure?
Date: March 15, 2022
Search Parameters:
Date: 2000–Present
Document Type: Systematic reviews
Language: English
Database: Embase (Ovid), Medline (Ovid), Cochrane Database of Systematic Reviews (Ovid)

Embase (Ovid):
Search No. Syntax Results
1 Exp energy metabolism/ or exp energy expenditure/ or exp energy 215,665
balance/ or (energy-expenditure* or caloric-expenditure* or energy-
metabolism or metabolism-energy or energy-balance or balance-
energy).ti,ab.
2 body mass/ or waist hip ratio/ or waist circumference/ or weight height 716,486
ratio/ or body fat percentage/ or (body-mass or BMI or body-mass-index
or body-ban-mass or quetelet-index or waist-hip-ratio or hip-to-waist-
ratio or hip-waist-ratio or waist-to-hip-ratio or waist-circumference or
waist-size or weight-height-ratio or height-to-weight-ratio or height-
weight-ratio or weight-to-height-ratio or body-fat-percentage or %BF or
body-fat-percent or bodyfat-percentage or percent-body-fat or
percentage-of-body-fat or percentage-of-bodyfat).ti,ab.
3 1 and 2 18,256
4 limit 3 to (english language and "systematic review" and yr="2000 - 259
Current")

Medline (Ovid):
Search No. Syntax Results

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX D 189

1 exp Energy Metabolism/ or (bioenergetic* or energy-expenditure* or 438,295

energy-metabolism* or expenditure-energy or expenditures-energy or
metabolism-energy or metabolisms-energy or energy-balance or balance-
energy or caloric-expenditure*).ti,ab.
2 Body Mass Index/ or Waist Hip Ratio/ or Waist Circumference/ or (body- 267,883
mass-index or index-body-mass or index-quetelet or quetelets-index or
waist-hip-ratio or hip-to-waist-ratio or hip-waist-ratio or waist-to-hip-
ratio or waist-circumference or circumference-waist or weight-height-
ratio or height-to-weight-ratio or height-weight-ratio or weight-to-height-
ratio or body-fat-percentage or %BF or body-fat-percent or bodyfat-
percentage or percent-body-fat or percentage-of-body-fat or percentage-
of-bodyfat).ti,ab.
3 1 and 2 8,125
4 limit 3 to (english language and yr="2000 -Current" and "systematic 63
review")

Cochrane Database of Systematic Reviews (Ovid):

Search No. Syntax Results
1 (energy-expenditure* or caloric-expenditure* or energy-metabolism or 9
metabolism-energy or energy-balance or balance-energy).ti,ab,kw.
2 (body-mass or BMI or body-mass-index or body-ban-mass or quetelet- 117
index or waist-hip-ratio or hip-to-waist-ratio or hip-waist-ratio or waist-
to-hip-ratio or waist-circumference or waist-size or weight-height-ratio or
height-to-weight-ratio or height-weight-ratio or weight-to-height-ratio or
body-fat-percentage or %BF or body-fat-percent or bodyfat-percentage or
percent-body-fat or percentage-of-body-fat or percentage-of-
bodyfat).ti,ab,kw.
3 1 and 2 3

Key Question: What is the association of body composition on metabolic efficiency (energy
usage/expenditure)?

Date: February 23, 2022

Search Parameters:
Date: 2000–Present
Document Type: Systematic reviews
Language: English
Databases: Embase (Ovid) 1974 to 2022 February 22; Cochrane Database of Systematic Reviews
(Ovid)

Embase (Ovid)
1 *energy metabolism/ or *energy expenditure/ or *energy balance/ 30,030

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

190 DIETARY REFERENCE INTAKES FOR ENERGY

2 *body composition/ or *body fat distribution/ or *body weight/ or *body mass/ or 272,239
*skinfold thickness/ or *waist hip ratio/ or *obesity/ or *body weight change/ or
*underweight/
3 1 and 2 3,103
4 limit 3 to (human and english language and yr="2000 -Current") 1,406
5 systematic review.mp. or "systematic review"/ 434,492
6 4 and 5 25

Cochrane Database of Systematic Reviews (Ovid):

Search No. Syntax Results
1 energy.ti,ab,kw. 124
2 (body-composition or composition-body or body-fat-distribution or body- 261
weight or total-body-weight or weight-body or body-mass or body-mass-
index or body-ban-mass or quetelet-index or skinfold-thickness or skin-
fold-thickness or skin-fold-measurement* or skin-thickness or skinfold-
measurement* or waist-hip-ratio or hip-to-waist-ratio or hip-waist-ratio or
waist-to-hip-ratio or adipose-tissue-hyperplasia or adipositas or adiposity
or alimentary-obesity or body-weight-excess or corpulency or fat-
overload-syndrome or nutritional-obesity or obesitas or overweight or
body-weight-change* or weight-change* or underweight or thinness or
weight-insufficiency).ti,ab,kw.
3 1 and 2 35

Key Question: Identify DLW studies that may not be included in the IAEA database
Date: January 31, 2022
Search Parameters:
Date: 2000–Present
Document Type: All
Language: English
Database: Medline (Ovid)

Search:
Ovid MEDLINE(R) and Epub Ahead of Print, In-Process, In-Data-Review and Other
Nonindexed Citations, Daily and Versions <1946 to January 31, 2022>
Search Syntax Results
No.
1 "doubly labelled water".mp. 484
2 "doubly labeled water".mp. 1,069
3 1 or 2 1,542
4 limit 3 to (english language and humans and yr="2000 -Current") 885

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX D 191

Key Question: How does the increase in tissue deposition associated with growth during
infancy, childhood, adolescence, pregnancy, and lactation influence, effect, or contribute to
energy requirements?
Date: June 8, 2022

Search Parameters:
Date: 1980–Present
Document Type: All
Language: English
Databases: Medline (Ovid); Embase (Ovid)

Ovid MEDLINE(R)
Energy Cost of Pregnancy
1 exp Growth/ or exp Gestational Weight Gain/ or exp Weight Gain/ or exp Body 2,243,329
Composition/ or exp Anthropometry/ or exp "Body Weights and Measures"/ or
(growth or gain-weight or gains-weight or weight-gain* or body-composition* or
composition-body or compositions-body or doubly-labeled-water-method or doubly-
labelled-water-technique or anthropometry or body-measure* or body-weights-and-
measures or measures-body or measure-body or gestational-weight-gain or maternal-
weight-gain or postpartum-weight-retention or pregnancy-weight-gain or weight-
gain-gestational or weight-gain-maternal or weight-gain-pregnancy or weight-
retention-postpartum).ti,ab.
2 exp Pregnancy/ or exp Pregnant Women/ or (pregnancy or gestation or pregnancies or 1,050,557
pregnant-woman or pregnant-women or woman-pregnant or women-pregnant).ti,ab.
3 exp Energy Metabolism/ or exp Basal Metabolism/ or (energy-metabolism or 481,614
bioenergetic* or energy-expenditure* or energy-metabolism* or expenditure-energy
or expenditures-energy or metabolism-energy or metabolisms-energy or basal-
metabolic-rate* or basal-metabolism or rate-basal-metabolic or rate-resting-metabolic
or resting-metabolic-rate or resting-metabolic-rates or energy-cost* or energy-
consumption or energy-transfer* or energy-requirement* or energy-balance*).ti,ab.
4 1 and 2 and 3 2,805
5 Longitudinal studies/ 158,248
6 4 and 5 53
7 limit 6 to (english language and humans and yr="1980 -Current") 48

Ovid MEDLINE(R)
Energy Cost of Lactation
1 exp Energy Metabolism/ or exp Basal Metabolism/ or (energy-metabolism or 481,614
bioenergetic* or energy-expenditure* or energy-metabolism* or expenditure-energy
or expenditures-energy or metabolism-energy or metabolisms-energy or basal-
metabolic-rate* or basal-metabolism or rate-basal-metabolic or rate-resting-metabolic
or resting-metabolic-rate or resting-metabolic-rates or energy-cost* or energy-
consumption or energy-transfer* or energy-requirement* or energy-balance*).ti,ab.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

192 DIETARY REFERENCE INTAKES FOR ENERGY

2 exp Lactation/ or exp Milk, Human/ or (lactation or lactation-prolonged or lactations- 83,615

prolonged or milk-secretion* or prolonged-lactation* or breastmilk or breast-milk or
human-milk or milk-breast or milk-human).ti,ab.
3 Longitudinal studies/ 158,248
4 1 and 2 3,635
5 3 and 4 43
6 limit 5 to (english language and humans and yr="1980 -Current") 38

Ovid MEDLINE(R)
Energy Cost of Growth
1 exp Energy Metabolism/ or exp Basal Metabolism/ or (energy-metabolism or 481,614
bioenergetic* or energy-expenditure* or energy-metabolism* or expenditure-energy
or expenditures-energy or metabolism-energy or metabolisms-energy or basal-
metabolic-rate* or basal-metabolism or rate-basal-metabolic or rate-resting-metabolic
or resting-metabolic-rate or resting-metabolic-rates or energy-cost* or energy-
consumption or energy-transfer* or energy-requirement* or energy-balance*).ti,ab.
2 exp Growth/ or exp Child Development/ or exp Weight Gain/ or exp Body 2,300,924
Composition/ or exp Anthropometry/ or exp "Body Weights and Measures"/ or
(growth or child-development or development-child or development-infant or infant-
development or gain-weight or gains-weight or weight-gain* or body-composition*
or composition-body or compositions-body or doubly-labeled-water-method or
doubly-labelled-water-technique or anthropometry or body-measure* or body-
weights-and-measures or measures-body or measure-body).ti,ab.
3 exp Adolescent/ or exp Child/ or exp Infant/ or (adolescence or adolescent* or teen* 4,135,479
or teenager* or youth* or child* or infant*).ti,ab.
4 1 and 2 and 3 6,676
5 Longitudinal studies/ 158,248
6 4 and 5 266
7 limit 6 to (english language and humans and yr="1980 -Current") 254

Embase (Ovid)
Energy Cost of Lactation
1 exp "energy cost"/ or exp energy metabolism/ or exp basal metabolic rate/ or (energy- 296,447
cost* or energy-expenditure* or caloric-expenditure* or energy-metabolism* or
metabolism-energy or energy-consumption or energy-transfer* or energy-
requirement* or energy-balance or basal-metabolism or basal-metabolism-rate or
basal-oxygen-consumption or basic-metabolic-rate or basic-metabolism or bmr or
energy-content or energy-composition).ti,ab.
2 exp lactation/ or exp breast milk/ or (lactation or breast-secretion or lactic-secretion or 101,836
mammary-gland-secretion* or milk-excretion or milk-release* or milk-secretion* or
breast-milk or breastmilk or breast-fed-infant* or homogenized-pasteurized-human-

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX D 193

milk or human-milk or maternal-milk or milk-human or milk-mother or mother-milk

or woman-milk).ti,ab.
3 Longitudinal study/ 173,158
4 1 and 2 4,033
5 3 and 4 35
6 limit 5 to (human and english language and yr="1980 -Current") 26

Embase (Ovid)
Energy Cost of Pregnancy
1 exp "energy cost"/ or exp energy metabolism/ or exp basal metabolic rate/ or (energy- 296,447
cost* or energy-expenditure* or caloric-expenditure* or energy-metabolism* or
metabolism-energy or energy-consumption or energy-transfer* or energy-
requirement* or energy-balance or basal-metabolism or basal-metabolism-rate or
basal-oxygen-consumption or basic-metabolic-rate or basic-metabolism or bmr or
energy-content or energy-composition).ti,ab.
2 exp body growth/ or exp gestational weight gain/ or exp body weight gain/ or exp 2,145,031
body composition/ or exp growth rate/ or exp doubly labeled water technique/ or exp
anthropometry/ or (body-growth or growth-body or somatic-growth or growth or
body-weight-gain or body-weight-increase or weight-gain or weight-increase or
body-composition or composition-body or growth-rate* or growth-rate-relative or
growth-velocity or rate-growth or relative-growth-rate or velocity-growth or doubly-
labeled-water-method or doubly-labelled-water-technique or anthropometry or
anthropometric-index or anthropometrics or antropometry or body-measurement* or
tissue-deposition*or gestational-weight-gain or pregnancy-weight-gain).ti,ab.
3 exp pregnancy/ or exp pregnant woman/ or (pregnancy or child-bearing or 1,003,515
childbearing or gestation or gravidity or intrauterine-pregnancy or labor-presentation
or labour-presentation or pregnancy-maintenance or pregnancy-trimester* or
pregnant-woman or pregnant-women).ti,ab.
4 1 and 2 and 3 1,823
5 longitudinal study/ 173,158
6 4 and 5 31
7 limit 6 to (human and english language and yr="1980 -Current") 28

Embase (Ovid)
Energy Cost of Growth
1 exp "energy cost"/ or exp energy metabolism/ or exp basal metabolic rate/ or (energy- 296447
cost* or energy-expenditure* or caloric-expenditure* or energy-metabolism* or
metabolism-energy or energy-consumption or energy-transfer* or energy-
requirement* or energy-balance or basal-metabolism or basal-metabolism-rate or
basal-oxygen-consumption or basic-metabolic-rate or basic-metabolism or bmr or
energy-content or energy-composition).ti,ab.
2 exp body growth/ or exp postnatal growth/ or exp body weight gain/ or exp body 2148604
composition/ or exp growth rate/ or exp doubly labeled water technique/ or exp
anthropometry/ or (body-growth or growth-body or somatic-growth or postnatal-
growth or child-growth or infant-growth or growth or body-weight-gain or body-

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

194 DIETARY REFERENCE INTAKES FOR ENERGY

weight-increase or weight-gain or weight-increase or body-composition or

composition-body or growth-rate* or growth-rate-relative or growth-velocity or rate-
growth or relative-growth-rate or velocity-growth or doubly-labeled-water-method or
doubly-labelled-water-technique or anthropometry or anthropometric-index or
anthropometrics of anthropometry or body-measurement* or tissue-deposition*).ti,ab.
3 exp adolescent/ or exp child/ or (adolescent* or teenager* or child* or infant*).ti,ab. 4299109
4 1 and 2 and 3 5744
5 longitudinal study/ 173158
6 4 and 5 165
7 limit 6 to (human and english language and yr="1980 -Current") 155

Key Question: What equations are available for computing or calculating basal energy
expenditure (BEE), basal metabolic rate (BMR), and resting metabolic rate (RMR)?
Date: April 20, 2022

Search Parameters:
Date: 2012–Present
Document Type: Systematic reviews
Language: English
Database: Embase (Ovid), Medline (Ovid), Cochrane Database of Systematic Reviews (Ovid)

Embase (Ovid):
Search No. Syntax Results
1 exp energy metabolism/ or exp energy expenditure/ or exp energy balance/ or 264,392
exp basal metabolic rate/ or exp resting metabolic rate/ or exp resting energy
expenditure/ or (energy-expenditure* or caloric-expenditure* or energy-
metabolism or metabolism-energy or energy-balance or balance-energy or
basal-metabolism or basal-oxygen-consumption or basic-metabolic-rate* or
basic-metabolism or BMR or BEE or resting-metabolic-rate or RMR or
resting-energy-expenditure).ti,ab.
2 equation*.ti,ab,kw. 232,344
3 1 and 2 5,780
4 limit 3 to (english language and "systematic review" and yr="2012 -Current") 34

Medline (Ovid):
Search No. Syntax Results
1 exp Energy Metabolism/ or exp basal metabolic rate/ or (bioenergetic* or 453,951
energy-expenditure* or energy-metabolism* or expenditure-energy or
expenditures-energy or metabolism-energy or metabolisms-energy or energy-
balance or balance-energy or caloric-expenditure* or Basal-metabolism or
basal-metabolic-rate* or basal-metabolism or metabolic-rate-basal or
metabolic-rate-resting or metabolism-basal or rate-basal-metabolic or rate-

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX D 195

resting-metabolic or resting-metabolic-rate* or BMR or RMR or BEE or

basal-energy-expenditure* or resting-energy-expenditure*).ti,ab.
2 equation*.ti,ab,kw. 139,671
3 1 and 2 4,750
4 limit 3 to (english language and yr="2012 -Current" and "systematic review") 16

Cochrane Database of Systematic Reviews (Ovid):

Search No. Syntax Results
1 (bioenergetic* or energy-expenditure* or energy-metabolism* or 11
expenditure-energy or expenditures-energy or metabolism-energy or
metabolisms-energy or energy-balance or balance-energy or caloric-
expenditure* or Basal-metabolism or basal-metabolic-rate* or basal-
metabolism or metabolic-rate-basal or metabolic-rate-resting or metabolism-
basal or rate-basal-metabolic or rate-resting-metabolic or resting-metabolic-
rate* or BMR or RMR or BEE or basal-energy-expenditure* or resting-
energy-expenditure*).ti,ab,kw.
2 equation*.ti,ab,kw. 17
3 1 and 2 0

Key Question: What is the effect of race or ethnicity on energy expenditure?

Date: April 19, 2022 (Search 1)

Search Parameters:
Date: 2012–Present
Document Type: Systematic reviews
Language: English
Database: Embase (Ovid), Medline (Ovid), Cochrane Database of Systematic Reviews (Ovid)

Embase (Ovid):
Search No. Syntax Results
1 exp energy metabolism/ or exp energy expenditure/ or exp energy balance/ 249,873
or exp basal metabolic rate/ or exp resting metabolic rate/ or exp resting
energy expenditure/ or (energy-expenditure* or caloric-expenditure* or
energy-metabolism or metabolism-energy or energy-balance or balance-
energy or basal-metabolism or basal-oxygen-consumption or basic-
metabolic-rate* or basic-metabolism or BMR or resting-metabolic-rate or
RMR or resting-energy-expenditure).ti,ab.
2 exp "ethnic or racial aspects"/ or exp ancestry group/ or (ethnic-aspect* or 712,617
racial-aspect* or cultural-factor* or ethnic-difference* or ethnicity or race*
or racial-factor* or racial-difference* or ancestry-group or continental-
population-group* or racial-group*).ti,ab.
3 1 and 2 3,066

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

196 DIETARY REFERENCE INTAKES FOR ENERGY

4 limit 3 to (english language and "systematic review" and yr="2012 - 31

Current")

Medline (Ovid):
Search No. Syntax Results
1 exp Energy Metabolism/ or exp Basal Metabolism/ or (bioenergetic* or 443,449
energy-expenditure* or energy-metabolism* or expenditure-energy or
expenditures-energy or metabolism-energy or metabolisms-energy or
energy-balance or balance-energy or caloric-expenditure* or Basal-
metabolism or basal-metabolic-rate* or basal-metabolism or metabolic-rate-
basal or metabolic-rate-resting or metabolism-basal or rate-basal-metabolic
or rate-resting-metabolic or resting-metabolic-rate* or BMR or RMR or
basal-energy-expenditure* or resting-energy-expenditure* or REE).ti,ab.
2 exp Population Groups/ or (population-group* or group-population or 323,768
groups-population).ti,ab.
3 1 and 2 1,490
4 limit 3 to (english language and yr="2012 -Current" and "systematic 8
review")

Cochrane Database of Systematic Reviews (Ovid):

Search No. Syntax Results
1 (bioenergetic* or energy-expenditure* or energy-metabolism* or 9
expenditure-energy or expenditures-energy or metabolism-energy or
metabolisms-energy or energy-balance or balance-energy or caloric-
expenditure* or Basal-metabolism or basal-metabolic-rate* or basal-
metabolism or metabolic-rate-basal or metabolic-rate-resting or
metabolism-basal or rate-basal-metabolic or rate-resting-metabolic or
resting-metabolic-rate* or BMR or RMR or basal-energy-expenditure* or
resting-energy-expenditure* or REE).ti,ab,kw.
2 (population-group* or group-population or groups-population or race or 81
ethnicity).ti,ab,kw.
3 1 and 2 0

Date: June 2, 2022 (Search 2)

Search Parameters:
Date: 1980–Present
Document Type: All
Language: English
Database: PubMed

PubMed
1 African Americans [MeSH Terms] 60,282

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX D 197

2 Blacks [MeSH Terms] 94,672

3 Whites [MeSH Terms] 70,830
4 (Hispanic [MeSH Terms]) OR (latino [MeSH Terms]) 36,377
5 (American Indians [MeSH Terms]) OR (Alaska natives [MeSH Terms]) 1,106
6 Asian Americans [MeSH Terms] 8,634
7 (Native Hawaiian [MeSH Terms]) OR (Pacific Islander [MeSH Terms]) 1,1671
8 1 or 2 or 3 or 4 or 5 or 6 or 7 160,632
9 energy metabolism [MeSH Terms] 406,915
10 basal metabolism [MeSH Terms] 8,616
11 body composition [MeSH Terms] 61,583
12 9 or 10 or 11 461,143
13 8 AND 12 2,318
14 limit 13 to (English language and yr=”1980 -Current” and Humans) 2,245

Date: June 8, 2022

Search results were further restricted using Endnote
Title must contain: energy metabolism or energy requirement(s) or basal metabolism or body
composition or energy expenditure or energy balance or metabolic rate
Final results: 465 papers

Key Question: What is the association of macronutrient composition of the diet on metabolic
efficiency (energy usage/expenditure)?

Date: February 23, 2022

Search Parameters:
Date: 2000–Present
Document Type: Systematic reviews
Language: English
Databases: Embase (Ovid) 1974 to 2022 February 22; Cochrane Database of Systematic Reviews
(Ovid)

Embase (Ovid)
1 *energy metabolism/ or *energy expenditure/ or *energy balance/ or (energy- 103,565
expenditure* or caloric-expenditure* or energy-metabolism or metabolism-energy or
energy-balance or balance-energy).ti,ab.
2 exp macronutrient/ or macronutrient*.ti,ab. 17,270
3 exp diet therapy/ or exp low carbohydrate diet/ or exp carbohydrate loading diet/ or 402,614
exp ketogenic diet/ or exp low fat diet/ or exp protein diet/ or exp paleolithic diet/ or
exp protein restriction/ or (metabolic-efficienc* or diet-therap* or diet-intervention*

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

198 DIETARY REFERENCE INTAKES FOR ENERGY

or diet-treatment* or dietary-intervention* or dietary-therap* or dietary-treatment* or

nutrition-therap* or nutritional-therap* or carbohydrate-restricted-diet* or
carbohydrate-poor-diet* or diet-low-carbohydrate* or low-carb-diet* or carbohydrate-
loading or diet-carbohydrate-loading or diet-ketogenic* or keto-diet* or ketogenous-
diet* or ketotic-diet* or diet-fat-restricted or diet-low-fat or fat-diet-low or fat-
restricted-diet* or fat-restriction* or lipid-restricted-diet* or lipid-restriction or low-
lipid-diet* or diet-high-protein or diet-protein or high-protein-diet* or protein-meal*
or protein-enriched-diet* or protein-rich-diet* or caveman-diet* or diet-paleolithic or
hunter-gatherer-diet* or paleo-diet* or stone-age-diet* or borst-diet* or diet-borst or
diet-giovanetti or diet-protein-poor or diet-protein-restricted or giovanetti-diet* or
hypoprotein-diet* or low-protein-diet* or protein-free-diet* or protein-poor-diet* or
protein-restricted-diet*).ti,ab.
4 1 and 2 and 3 452
5 limit 4 to (human and english language and yr="2000 -Current") 354
6 "systematic review"/ or systematic review*.mp. 434,492
7 5 and 6 12

Cochrane Database of Systematic Reviews (Ovid):

Search No. Syntax Results
1 energy.ti,ab,kw. 124
2 macronutrient*.af. 111
3 (metabolic-efficienc* or diet-therap* or diet-intervention* or diet- 155
treatment* or dietary-intervention* or dietary-therap* or dietary-
treatment* or nutrition-therap* or nutritional-therap* or carbohydrate-
restricted-diet* or carbohydrate-poor-diet* or diet-low-carbohydrate* or
low-carb-diet* or carbohydrate-loading or diet-carbohydrate-loading or
diet-ketogenic* or keto-diet* or ketogenous-diet* or ketotic-diet* or diet-
fat-restricted or diet-low-fat or fat-diet-low or fat-restricted-diet* or fat-
restriction* or lipid-restricted-diet* or lipid-restriction or low-lipid-diet*
or diet-high-protein or diet-protein or high-protein-diet* or protein-meal*
or protein-enriched-diet* or protein-rich-diet* or caveman-diet* or diet-
paleolithic or hunter-gatherer-diet* or paleo-diet* or stone-age-diet* or
borst-diet* or diet-borst or diet-giovanetti or diet-protein-poor or diet-
protein-restricted or giovanetti-diet* or hypoprotein-diet* or low-protein-
diet* or protein-free-diet* or protein-poor-diet* or protein-restricted-
diet*).ti,ab,kw.
4 1 and 2 and 3 14

Key Question: What is the degree of systematic bias (random error, measurement error, or
misreporting) of energy intake as assessed by self-report (diet records, 24-hour recalls, or food
frequency questionnaires) compared to doubly labeled water studies?
Date: April 19, 2022

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX D 199

Search Parameters:
Date: 2012–Present
Document Type: Systematic reviews
Language: English
Database: Embase (Ovid), Medline (Ovid), Cochrane Database of Systematic Reviews (Ovid)

Embase (Ovid):
Search No. Syntax Results
1 exp caloric intake/ or (energy-intake* or caloric-intake* or calorie-intake* or 81,476
calory-intake* or dietary-energy or intake-caloric).ti,ab.
2 exp statistical bias/ or exp random error/ or exp error/ or (bias or statistical- 997,855
bias* or systematic-bias* or truncation-bias* or random-error* or error* or
error-stud* or human-error* or mistake* or misreport*).ti,ab.
3 exp self report/ or exp medical record/ or exp food frequency questionnaire/ 679,482
or exp doubly labeled water technique/ or (self-report* or diet-record* or
medical-record* or health-record* or patient-record* or 24-hour-recall* or
food-frequency-questionnaire* or doubly-labeled-water-stud* or doubly-
labeled-water-technique*).ti,ab.
4 1 and 2 and 3 594
5 limit 4 to (english language and "systematic review" and yr="2012 - 23
Current")

Medline (Ovid):
Search No. Syntax Results
1 exp Energy Intake/ or (energy-intake* or caloric-intake* or calorie- 65,015
intake*).ti,ab.
2 exp Bias/ or (bias* or outcome-measurement-error* or error-outcome- 274,133
measurement* or random-error* or measurement-error* or
misreport*).ti,ab.
3 exp Self Report/ or exp Diet Records/ or exp "Surveys and Questionnaires"/ 1,777,781
or (report-self or self-report* or diet-record* or diaries-food or diary-food
or diet-record* or dietary-record* or food-diar* or record-diet* or records-
diet* or survey* or questionnaire* or doubly-labeled-water).ti,ab.
4 1 and 2 and 3 595
5 limit 4 to (english language and yr="2012 -Current" and "systematic 14
review")

Cochrane Database of Systematic Reviews (Ovid):

Search No. Syntax Results
1 (energy-intake* or caloric-intake* or calorie-intake* or calory-intake* or 45
dietary-energy or intake-caloric).ti,ab,kw.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

200 DIETARY REFERENCE INTAKES FOR ENERGY

2 (bias* or statistical-bias* or systematic-bias* or truncation-bias* or 5,023

random-error* or error* or error-stud* or human-error* or mistake* or
misreport*).ti,ab,kw.
3 (self-report* or diet-record* or medical-record* or health-record* or 218
patient-record* or 24-hour-recall* or food-frequency-questionnaire* or
doubly-labeled-water-stud* or doubly-labeled-water-technique*).ti,ab,kw.
4 1 and 2 and 3 1

Key Questions:
How do physical activity and energy expenditure change across the life span?
What is the relationship between different measurements of physical activity and energy
expenditure?

Date: March 15, 2022

Search Parameters:
Date: 2000–Present
Document Type: Systematic reviews
Language: English
Database: Embase (Ovid), Medline (Ovid), Cochrane Database of Systematic Reviews (Ovid)

Embase (Ovid):

Search No. Syntax Results

1 exp energy metabolism/ or exp energy expenditure/ or exp energy balance/ 247,779
or (energy-expenditure* or caloric-expenditure* or energy-metabolism or
metabolism-energy or energy-balance or balance-energy).ti,ab.
2 physical activity/ or exercise/ or sedentary lifestyle/ or (physical-activity or 977,819
activity-physical or exercise or biometric-exercise or effort or exercise-
capacity or exercise-performance or exercise-training or exertion or fitness-
training or fitness-workout or physical-conditioning-human or physical-
effort or physical-exercise or physical-exertion or physical-work-out or
physical-workout or free-living-activit* or sedentary-behavior* or
sedentary-behaviour* or sedentary-lifestyle* or sedentary-life-style or
volume-of-activit* or physical-activity-intensit* or MetS or metabolic-
syndrome* or kcal* or total-activity-count*).ti,ab.
3 1 and 2 36,471
4 limit 3 to (english language and yr="2000 -Current") 29,334
5 Limit 4 to “systematic review” 514
6 accelerometer/ or self report/ or indirect calorimeter/ or oxygen 482,260
consumption/ or (wearable-physical-activity-monitor* or accelerometer* or
meter-accelero or objective-measure* or subjective-measure* or self-

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX D 201

report* or diary or physical-activity-questionnaire* or indirect-calorimeter*

or indirect-calorimetry or portable-calorimeter* or room-calorimeter* or
oxygen-consumption* or consumption-oxygen or consumptions-oxygen or
VO2 or doubly-labeled-water).ti,ab.
7 3 and 6 11,598
8 Limit 7 to (english language and yr=”2000-Current” and “systematic review” 135

Medline (Ovid):
Search No. Syntax Results
1 exp Energy Metabolism/ or (bioenergetic* or energy-expenditure* or 438,295
energy-metabolism* or expenditure-energy or expenditures-energy or
metabolism-energy or metabolisms-energy or energy-balance or balance-
energy or caloric-expenditure*).ti,ab.
2 Exercise/ or Sedentary Behavior/ or (physical-activit* or activities-physical 467,187
or activity-physical or exercise* or free-living-activit* or sedentary-
behavior* or behavior-sedentary or inactivity-physical or lack-of-physical-
activity or lifestyle-sedentary or physical-inactivity or sedentary-behaviour*
or sedentary-time* or time-sedentary or volume-of-activity or physical-
activity-intensit* or MetS or metabolic-syndrome* or kcal* or total-activity-
count*).ti,ab.
3 1 and 2 26,933
4 limit 3 to (english language and yr="2000 -Current") 19,391
5 limit 4 to "systematic review" 252
6 Wearable Electronic Devices/ or Accelerometry/ or Self Report/ or
Calorimetry, Indirect/ or Oxygen Consumption/ or (wearable-electronic-
device* or wearable-physical-activity-monitor* or device-wearable* or
devices-wearable* or electronic-device-wearable* or electronic-devices-
wearable* or electronic-skin* or skin-electronic or technologies-wearable
or technology-wearable or wearable-device* or wearable-electronic-
device* or wearable-technolog* or accelerometer* or accelerometr* or
objective-measure* or subjective-measure* or self-report* or report-self or
reports-self or diary or physical-activity-questionnaire* or calorimetry-
indirect or calorimetries-indirect or calorimetries-respiration* or
calorimetry-indirect or calorimetry-respiration or indirect-calorimetr* or
respiration-calorimetr* or indirect-calorimeter* or portable-calorimeter* or
room-calorimeter* or oxygen-consumption* or consumption-oxygen or
consumptions-oxygen or VO2 or doubly-labeled-water).ti,ab.
7 3 and 6 9,471
8 Limit 7 to (english language and yr="2000 -Current" and "systematic 64
review")

Cochrane Database of Systematic Reviews (Ovid):

Search No. Syntax Results

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

202 DIETARY REFERENCE INTAKES FOR ENERGY

1 (energy-expenditure* or caloric-expenditure* or energy-metabolism or 9

metabolism-energy or energy-balance or balance-energy).ti,ab,kw.
2 (physical-activity or activity-physical or exercise or biometric-exercise or 739
effort or exercise-capacity or exercise-performance or exercise-training or
exertion or fitness-training or fitness-workout or physical-conditioning-
human or physical-effort or physical-exercise or physical-exertion or
physical-work-out or physical-workout or free-living-activit* or sedentary-
behavior* or sedentary-behaviour* or sedentary-lifestyle* or sedentary-
life-style or volume-of-activit* or physical-activity-intensit* or MetS or
metabolic-syndrome* or kcal* or total-activity-count*).ti,ab,kw.
3 1 and 2 5
4 (wearable-physical-activity-monitor* or accelerometer* or meter-accelero 272
or objective-measure* or subjective-measure* or self-report* or diary or
physical-activity-questionnaire* or indirect-calorimeter* or indirect-
calorimetry or portable-calorimeter* or room-calorimeter* or oxygen-
consumption* or consumption-oxygen or consumptions-oxygen or VO2 or
doubly-labeled-water).ti,ab,kw
5 3 and 4 2

Key Question: What is the association between weight change and chronic disease outcomes?
Date: May 6, 2022
Search Parameters:
Date: 2017–Present
Document Type: Systematic reviews
Language: English
Database: Embase (Ovid), Medline (Ovid), Cochrane Database of Systematic Reviews (Ovid)

Embase (Ovid):
Search No. Syntax Results
1 exp body weight loss/ or exp body weight maintenance/ or exp body weight 310,134
gain/ or exp body weight change/ or exp weight cycling/ or exp body weight
fluctuation/ or (body-weight-loss or body-weight-decrease or body-weight-
reduction or weight-decrease or weight-los* or weight-reduc* or weight-
watch* or weight-maintenance or body-weight-gain* or body-weight-
increase* or weight-gain* or weight-increase* or weight-change* or body-
weight-change* or weight-cycling or yo-yo-diet* or yo-yo-effect* or yoyo-
diet* or body-weight-fluctuation* or weight-fluctuation*).ti,ab.
2 *cardiovascular disease/ or *congenital heart disease/ or *non insulin 4,226,771
dependent diabetes mellitus/ or *malignant neoplasm/ or *hypertension/ or
*hip fracture/ or *all cause mortality/ or *dementia/ or (cardiovascular-
disease* or angiocardiopathy or angiocardiovascular-disease* or
cardiovascular-complication* or cardiovascular-disturbance* or
cardiovascular-lesion* or cardiovascular-syndrome* or cardiovascular-
vegetative-disorder* or complication-cardiovascular* or disease-

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX D 203

cardiovascular or major-adverse-cardiovascular-event* or congenital-heart-

disease* or congenital-cardiac-disease* or congenital-cardiac-distress* or
congenital-heart-distress* or congenital-heart-failure* or heart-congenital-
disease* or neonatal-cardiopathy or truncus-arteriosus-persistent* or adult-
onset-diabetes or diabetes-mellitus-type-2 or diabetes-mellitus-type-ii or
diabetes-mellitus-maturity-onset or diabetes-mellitus-non-insulin-dependent
or diabetes-type-2 or diabetes-type-II or diabetes-adult-onset or dm-2 or
insulin-independent-diabetes or ketosis-resistant-diabetes-mellitus or
maturity-onset-diabetes or NIDDM or non-insulin-dependent-diabetes or non-
insulin-dependent-diabetes-mellitus or noninsulin-dependent-diabetes or
T2DM or type2-diabetes or type-II-diabetes or cancer* or malignant-
neoplasia or malignant-neoplasm* or malignant-neoplastic-disease* or
malignant-tumor* or malignant-tumour* or neoplasia-malignant or tumor-
malignant or tumour-malignant or hypertension or blood-pressure-high or
high-blood-pressure or high-renin-hypertension or hypertensive-disease* or
hypertensive-effect* or hypertensive-effect* or hypertensive-response or hip-
fracture* or broken-hip* or fracture-hip* or fractured-hip* or all-cause-
mortality or dementia or amentia or demention).ti,ab.
3 1 and 2 65,276
4 limit 3 to (english language and "systematic review" and yr="2017 -Current") 1,120

Medline (Ovid):
Search No. Syntax Results
1 exp Body Weight Changes/ or exp weight cycling/ or exp weight gain/ or 180,453
exp weight loss/ or (body-weight-change* or change-body-weight or
changes-body-weight or weight-change-body or weight-changes-body or
weight-cycling or cycling-weight or gain-weight or gains-weight or weight-
gain* or weight-loss or loss-weight or losses-weight or reduction-weight or
reductions-weight or weight-loss* or weight-reduction*).ti,ab.
2 Cardiovascular Diseases/ or Heart Defects, Congenital/ or Diabetes 3,890,945
Mellitus, Type 2/ or Neoplasms/ or Hypertension/ or Hip Fractures/ or
Dementia/ or (cardiovascular-disease* or disease-cardiovascular or
abnormality-heart or congenital-heart-defect* or congenital-heart-disease*
or defect-congenital-heart or defects-congenital-heart or disease-congenital-
heart or heart-abnormalit* or heart-defect-congenital or heart-defects-
congenital or heart-disease-congenital or heart-malformation-of or
malformation-of-heart* or adult-onset-diabetes-mellitus or diabetes-
maturity-onset or diabetes-mellitus-adult-onset or diabetes-mellitus-ketosis-
resistant or diabetes-mellitus-maturity-onset or diabetes-mellitus-non-
insulin-dependent or diabetes-mellitus-noninsulin-dependent or diabetes-
mellitus-slow-onset or diabetes-mellitus-stable or diabetes-mellitus-type-2
or diabetes-mellitus-type-ii or diabetes-type-2 or ketosis-resistant-diabetes-
mellitus or mody or maturity-onset-diabetes or maturity-onset-diabetes-
mellitus or niddm or non-insulin-dependent-diabetes-mellitus or slow-
onset-diabetes-mellitus or stable-diabetes-mellitus or type-2-diabetes or

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

204 DIETARY REFERENCE INTAKES FOR ENERGY

neoplasm* or cancer* or malignancies or malignancy or neoplasia* or

tumor* or hypertension or blood-pressure-high or blood-pressures-high or
high-blood-pressure* or fractures-hip or fractures-intertrochanteric or
fractures-subtrochanteric or fractures-trochanteric or hip-fractures or
intertrochanteric-fracture* or trochanteric-fracture* or all-cause-mortality
or dementia* or amentia*).ti,ab.
3 1 and 2 42,764
4 limit 3 to (english language and yr="2017 -Current" and "systematic 501
review")

Cochrane Database of Systematic Reviews (Ovid):

Search No. Syntax Results
1 (body-weight-change* or change-body-weight or changes-body-weight or 279
weight-change-body or weight-changes-body or weight-cycling or cycling-
weight or gain-weight or gains-weight or weight-gain* or weight-loss or
loss-weight or losses-weight or reduction-weight or reductions-weight or
weight-loss* or weight-reduction*).ti,ab,kw.
2 (cardiovascular-disease* or disease-cardiovascular or abnormality-heart or 2,572
congenital-heart-defect* or congenital-heart-disease* or defect-congenital-
heart or defects-congenital-heart or disease-congenital-heart or heart-
abnormalit* or heart-defect-congenital or heart-defects-congenital or heart-
disease-congenital or heart-malformation-of or malformation-of-heart* or
adult-onset-diabetes-mellitus or diabetes-maturity-onset or diabetes-
mellitus-adult-onset or diabetes-mellitus-ketosis-resistant or diabetes-
mellitus-maturity-onset or diabetes-mellitus-non-insulin-dependent or
diabetes-mellitus-noninsulin-dependent or diabetes-mellitus-slow-onset or
diabetes-mellitus-stable or diabetes-mellitus-type-2 or diabetes-mellitus-
type-ii or diabetes-type-2 or ketosis-resistant-diabetes-mellitus or mody or
maturity-onset-diabetes or maturity-onset-diabetes-mellitus or niddm or
non-insulin-dependent-diabetes-mellitus or slow-onset-diabetes-mellitus or
stable-diabetes-mellitus or type-2-diabetes or neoplasm* or cancer* or
malignancies or malignancy or neoplasia* or tumor* or hypertension or
blood-pressure-high or blood-pressures-high or high-blood-pressure* or
fractures-hip or fractures-intertrochanteric or fractures-subtrochanteric or
fractures-trochanteric or hip-fractures or intertrochanteric-fracture* or
trochanteric-fracture* or all-cause-mortality or dementia* or
amentia*).ti,ab,kw.
3 1 and 2 79
4 limit 3 to last 5 years 14

Key Question: What is the effect or association of weight cycling on metabolic efficiency
(energy usage/expenditure) and health outcomes?

Date: February 23, 2022 (Search 1)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX D 205

Search Parameters:
Date: 2000–Present
Document Type: Systematic reviews
Language: English
Databases: Embase (Ovid) 1974 to 2022 February 22; Cochrane Database of Systematic Reviews
(Ovid)

Embase (Ovid)
1 *energy metabolism/ or *energy expenditure/ or *energy balance/ or (energy- 103,565
expenditure* or caloric-expenditure* or energy-metabolism or metabolism-energy
or energy-balance or balance-energy).ti,ab.
2 exp weight cycling/ or (weight-cycling or yo-yo-diet* or yo-yo-effect* or yoyo- 573
diet*).ti,ab.
3 1 and 2 52
4 limit 3 to (human and english language and yr="2000 -Current") 26
5 "systematic review"/ or systematic review*.mp. 434,492
6 4 and 5 0

Cochrane Database of Systematic Reviews (Ovid):

Search No. Syntax Results
1 energy.ti,ab,kw. 124
2 (weight-cycling or yo-yo-diet* or yo-yo-effect* or yoyo-diet*).af. 9
3 1 and 2 0

Date: March 4, 2022 (Search 2)

Search Parameters:
Date: 2000–Present
Document Type: Systematic reviews
Language: English
Database: Embase (Ovid), Cochrane Database of Systematic Reviews (Ovid), PubMed

Embase (Ovid):
Search No. Syntax Results
1 exp weight cycling/ or exp body weight fluctuation/ or (weight-cycling or 1,015
yo-yo-diet* or yo-yo-effect* or yoyo-diet* or body-weight-fluctuation*
or weight-fluctuation*).ti,ab.
2 limit 1 to (human and english language and yr="2000 -Current") 598
3 "systematic review"/ 333,502
4 2 and 3 14

Cochrane Database of Systematic Reviews (Ovid):

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

206 DIETARY REFERENCE INTAKES FOR ENERGY

Search No. Syntax Results

1 (weight-cycling or yo-yo-diet* or yo-yo-effect* or yoyo-diet* or body- 1
weight-fluctuation* or weight-fluctuation*).ti,ab,kw.

PubMed:
(Weight cycling[mh] or body weight fluctuation[mh] or weight-cycling[tiab] or yo-yo-diet[tiab]
or yo-yo-diets[tiab] or yo-yo-effect[tiab] OR yo-yo-effects[tiab] or yoyo-diet[tiab] or yoyo-
diets[tiab] or body-weight-fluctuation[tiab] or body-weight-fluctuations[tiab] or weight-
fluctuation[tiab] or weight-fluctuations[tiab]) and systematic review[pt]
Limit: Humans, English
Results: 12

Key Question: What level of calorie intake is needed to produce weight gain in individuals with
underweight? What amount of calorie intake (deficit) is necessary to produce weight loss in
individuals with overweight or obesity? What level of calorie intake is needed to maintain weight
across the weight spectrum?

Date: February 23, 2022

Search Parameters:
Date: 2000–Present
Document Type: Systematic reviews
Language: English
Databases: Embase (Ovid) 1974 to 2022 February 22; Cochrane Database of Systematic Reviews
(Ovid)

Embase (Ovid)
1 *energy metabolism/ or *energy expenditure/ or *energy balance/ or (energy- 103,565
expenditure* or caloric-expenditure* or energy-metabolism or metabolism-energy
or energy-balance or balance-energy).ti,ab.
2 exp caloric intake/ or (calorie-intake* or calory-intake* or dietary-energy or 78,710
energy-intake* or intake-caloric*).ti,ab.
3 exp body weight loss/ or exp body weight maintenance/ or exp body weight gain/ or 306,354
exp caloric restriction/ or (body-weight-loss or body-weight-decrease or body-
weight-reduction or weight-decrease or weight-los* or weight-reduc* or weight-
watch* or weight-maintenance or body-weight-gain or body-weight-increase or
weight-gain or weight-increase or caloric-intake-restriction or calorie-
restriction*).ti,ab.
4 1 and 2 and 3 3,085
5 limit 4 to (human and english language and yr="2000 -Current") 1,762
6 "systematic review"/ or systematic review*.mp. 434,492
7 5 and 6 58

Cochrane Database of Systematic Reviews (Ovid):

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX D 207

Search No. Syntax Results

1 energy.ti,ab,kw. 124
2 (calorie-intake* or calory-intake* or dietary-energy or energy-intake* or 42
intake-caloric*).ti,ab,kw.
3 (body-weight-loss or body-weight-decrease or body-weight-reduction or 279
weight-decrease or weight-los* or weight-reduc* or weight-watch* or
weight-maintenance or body-weight-gain or body-weight-increase or
weight-gain or weight-increase or caloric-intake-restriction or calorie-
restriction*).ti,ab,kw.
4 1 and 2 and 3 22

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

Appendix E
Key Questions and Eligibility Criteria

Key Question: What is the association between body mass index (BMI) and chronic disease,
including all-cause mortality?
Population • Human only
• General population (not existing disease state)
Interventions/exposures and • Body weight category (by BMI, z-score): overweight,
comparators obese, underweight, normal weight

Study designs (of primary Include:

studies included within the • Longitudinal studies
systematic reviews)
Exclude:
• Cross-sectional studies

Systematic reviews must include a meta-analysis

Timing • No minimum “treatment” or exposure duration required

• No minimum follow-up duration (to when outcome is
measured)
Setting • Restrict to high- and upper middle-income countries
• Restricted to systematic reviews published 2017 or later

Outcomes • Incident cardiovascular disease/coronary heart disease

• Incident type 2 diabetes
• Incident cancer
• Incident hypertension
• Incident hip fracture
• All-cause mortality
• Pregnancy-related
o Maternal mortality
o Preeclampsia
o Infant mortality
o Premature delivery
• Pediatric
o Growth

Key Question: What is the effect of BMI (and other measures of adiposity) on energy balance
or energy expenditure?
Population • Human only
• Allow focus on health conditions: obesity, type 2 diabetes
o Exclude focus on other health conditions

PREPUBLICATION COPY: UNCORRECTED PROOFS

209

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

210 DIETARY REFERENCE INTAKES FOR ENERGY

• Allow focus on other population groups:

o High physical activity
o Diet type
• Subgroups of interest
o Age/life stage
o Sex
o Race/ethnicity

Interventions/exposures and • Body weight category (by BMI, z-score): overweight,

comparators obese, underweight, normal weight
• Other measures of adiposity
• Exclude any cointerventions/coexposures (e.g., low fat +
vitamin B12 vs. high fat; physical activity)

Study designs (of primary Include:

studies included within the • Randomized crossover trials
systematic reviews) • Randomized clinical trials (parallel)
• Nonrandomized comparative studies, multivariable
adjusted
• Association analyses (e.g., regression models, predictors,
risk factors), multivariable adjusted

Exclude:
• Nonrandomized comparative studies, unadjusted
• Single group (noncomparative between
interventions/exposures)
• Association analyses (e.g., models, predictors, risk
factors), univariate, unadjusted
• Other study designs

Timing • No minimum “treatment” or exposure duration required

• No minimum follow-up duration (to when outcome is
measured)
Setting • Restrict to high- and upper middle-income countries
• Restricted to studies published 2000 or later
Outcomes • Energy balance/imbalance/excess/deficit
o Energy intake
o Energy expenditure (TEE, REE, BMR, BEE,
RMR)
o Body weight
o BMI
• Energy utilization
o Fat/Carbohydrate/Protein oxidation
o Body fat
o Body weight
• Energy metabolism/metabolic efficiency/metabolic
flexibility

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX E 211

• Obesity risk
o Body weight
o BMI
o Body fat %
o Waist circumference
o Visceral fat
• Type 2 diabetes mellitus risk
o Body weight
o BMI
o Body fat %
o Waist circumference
o Visceral fat
o Glucose
o Insulin
o HOMA-IR
• Cardiovascular Disease risk

NOTE: BEE = basal energy expenditure; BMI = body mass index; BMR = basal metabolic rate; HOMA-IR =
Homeostatic Model Assessment of Insulin Resistance; REE = resting energy expenditure; RMR = resting metabolic
rate.

Key Question: What is the association of body composition on metabolic efficiency (energy
usage/expenditure)?
Population • Human only
• Exclude focus on other health conditions (e.g.,
malnourished population)
• Allow focus on other population groups:
o High physical activity
o Diet (e.g., vegetarian, vegan)
o Others (determined on a case-by-case basis)
• Subgroups of interest
o Age/life stage
o Sex
o Race/ethnicity
• Study must provide separate results data by sex (or be
specific to one sex)
Interventions/exposures and • Exposure: fat mass vs. fat-free mass measured within the
comparators same individual
• Within the exposure of interest, consider studies using
DLW, DXA, and/or underwater weighing to estimate FM
and FFM
• Allow any cointerventions/coexposures (e.g., low fat +
vitamin B12 vs. high fat; physical activity)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

212 DIETARY REFERENCE INTAKES FOR ENERGY

Study designs (of primary Include:

studies included within the • Randomized crossover trials (including N-of-1)
systematic reviews) • Randomized clinical trials (parallel)
• Single group (noncomparative between
interventions/exposures)
• Nonrandomized comparative studies, multivariable
adjusted

Exclude:
• Nonrandomized comparative studies, unadjusted
• Association analyses (e.g., regression models, predictors,
risk factors), multivariable adjusted
• Association analyses (e.g., models, predictors, risk
factors), univariate, unadjusted
• Other study designs

Timing • No minimum “treatment” or exposure duration required

• No minimum follow-up duration (to when outcome is
measured)

Setting • Restrict to high- and upper middle-income countries

• Restricted to studies published 2000 or later

Outcomes Usage/expenditure must be measured (e.g., DLW, direct and

indirect calorimetry)
• 24-hour energy expenditure
• Resting or basal metabolic rate
• Energy cost of physical activity
• Thermic effect of food
• Metabolic adaptation (mitochondria dynamics)

NOTE: DLW = doubly labeled water; DXA = dual-energy X-ray absorptiometry; FFM = fat-free mass; FM = fat
mass.

Key Question: How does the increase in tissue deposition associated with growth during
infancy, childhood, adolescence, pregnancy, and lactation influence, effect, or contribute to
energy requirements?
Population • Human only
• Do not allow focus on health conditions or other
population groups
• Study must provide separate results data by sex (or be
specific to one sex)
• Study must provide separate results data by age group (or
be specific to one age group or life stage)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX E 213

• Subgroups of interest
o Age/life stage
o Sex

Interventions/exposures and The life stage of infancy, childhood, adolescence, pregnancy, and
comparators lactation

Study designs Include:

• Randomized clinical trials
• Nonrandomized comparative studies, unadjusted
• Single group (noncomparative between
interventions/exposures)
• Nonrandomized comparative studies, multivariable
adjusted
• Longitudinal studies
• Systematic reviews

Exclude:
• Association analyses (e.g., regression models, predictors,
risk factors), multivariable adjusted
• Association analyses (e.g., models, predictors, risk
factors), univariate, unadjusted

Timing • No minimum “treatment” or exposure duration required

• No minimum follow-up duration (to when outcome is
measured)

Setting • Restrict to high- and upper middle-income countries

• Restricted to studies published 1980 or later

Outcomes • Energy usage or energy expenditure

o Measured by DLW, direct or indirect calorimetry
• Growth during infancy, childhood, or adolescence
o Measured by weight gain (anthropometry) and
composition of gain (body composition
techniques), energetic efficiency (encompassed
by DLW; also estimated by theoretical
biochemical efficiencies)
• Pregnancy
o Measured by rate of weight gain per trimester
(anthropometry) and composition of gain (body
composition techniques); energy efficiency
(encompassed by DLW)
• Lactation
o Measured by volume of milk production and
proximate analysis of milk (fat, protein,
carbohydrate) or energy content of milk (bomb

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

214 DIETARY REFERENCE INTAKES FOR ENERGY

calorimetry); energy efficiency (encompassed by

DLW)
NOTE: DLW = doubly labeled water

Key Question: What is the impact of race or ethnicity on energy expenditure?

Population • Human only
• Subgroups of interest
o Age/life stage
o Sex

Interventions/exposures and Race or ethnic group

comparators

Study designs (of primary • Randomized clinical trials

studies included within the • Nonrandomized comparative studies, unadjusted
systematic reviews and primary • Single group (noncomparative between
studies) interventions/exposures)
• Nonrandomized comparative studies, multivariable
adjusted
• Association analyses (e.g., regression models, predictors,
risk factors), multivariable adjusted
• Association analyses (e.g., models, predictors, risk
factors), univariate, unadjusted
• Other study designs

Timing • No minimum “treatment” or exposure duration required

• No minimum follow-up duration (to when outcome is
measured)
Setting • Restrict to high- and upper middle-income countries
• Restricted to systematic reviews published 2012 or later
• Restricted to primary studies published 1980 or later

Outcomes • BMR (basal metabolic rate)

• RMR (resting metabolic rate)
• BEE (basal energy expenditure)
• REE (resting energy expenditure)
• Energy metabolism
• Energy expenditure
• Energy balance
• Caloric expenditure

NOTE: BEE = basal energy expenditure; BMR = basal metabolic rate; REE = resting energy expenditure; RMR =
resting metabolic rate.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX E 215

Key Question: What is the association of macronutrient composition of the diet on metabolic
efficiency (energy usage/expenditure)?
Population • Human only
• Allow focus on health conditions: obesity or type 2
diabetes
o Exclude focus on other health conditions
• Allow focus on other population groups:
o High physical activity
o Diet (e.g., vegetarian, vegan)
o Others (determined on a case-by-case basis)
• Subgroups of interest
o Age/life stage
o Sex
o Race/ethnicity
Interventions/exposures and Diets with different macronutrient composition
comparators • High fat vs. low protein (holding carbohydrates stable
with comparator)
• High fat vs. low carbohydrate (holding protein stable with
comparator)
• Low fat vs. high protein (holding carbohydrates stable
with comparator)
• Low fat vs. high carbohydrate (holding protein stable
with comparator)
• High protein vs. low carbohydrate (holding fat stable with
comparator)
• Low protein vs. high carbohydrate (holding fat stable
with comparator)
• Within the intervention/exposure of interest, exclude if
diets not isocaloric

Study designs (of primary Include:

studies included within the • Randomized crossover trials (including N-of-1)
systematic reviews) • Randomized clinical trials (parallel)
• Single group (noncomparative between
interventions/exposures)
• Nonrandomized comparative studies, multivariable
adjusted

Exclude:
• Nonrandomized comparative studies, unadjusted
• Association analyses (e.g., regression models, predictors,
risk factors), multivariable adjusted
• Association analyses (e.g., models, predictors, risk
factors), univariate, unadjusted
• Other study designs

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

216 DIETARY REFERENCE INTAKES FOR ENERGY

Timing • No minimum “treatment” or exposure duration required

• No minimum follow-up duration (to when outcome is
measured)

Setting • Restrict to high- and upper middle-income countries

• Restricted to studies published 2000 or later

Outcomes Usage/expenditure must be measured (e.g., DLW, direct and

indirect calorimetry)
• 24-hour energy expenditure
• Resting or basal metabolic rate
• Energy cost of physical activity
• Thermic effect of food
• Metabolic adaptation (mitochondria dynamics)

NOTE: DLW = doubly labeled water.

Key Question: In the U.S. and Canadian free-living population, what is the degree of systematic
bias or random error of energy intake as assessed by diet records, 24-hour recalls, food frequency
questionnaires, compared to doubly labeled water? Does the bias differ by age group, sex, body
weight status, BMI, social economic status, race, ethnicity, presence of chronic disease, or
physical activity level?
Population • Human only
• Allow focus on health conditions: obesity, type 2 diabetes
• Allow focus on other population groups:
o High physical activity
o Diet (e.g., vegetarian, vegan)
• Subgroups of interest
o Age/life stage
o Sex
o Race/ethnicity
• Study must provide separate results data by age group (or
be specific to one age group/life stage).

Interventions/exposures and Participation in a doubly labeled water study measuring usual

comparators dietary intake

Study designs (of primary • Randomized clinical trials

studies included within the • Nonrandomized comparative studies, unadjusted
systematic reviews) • Single group (noncomparative between
interventions/exposures)
• Nonrandomized comparative studies, multivariable
adjusted
• Association analyses (e.g., regression models, predictors,
risk factors), multivariable adjusted

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX E 217

• Association analyses (e.g., models, predictors, risk

factors), univariate, unadjusted
• Other study designs

Timing • No minimum “treatment” or exposure duration required

• No minimum follow-up duration (to when outcome is
measured)
Setting • Restrict to high-income countries
• Restricted to systematic reviews published 2012 or later

Outcomes Percent of misreporting of energy intakes

Key Question: How do physical activity and energy expenditure change across the life span?
Population • Human only
• Allow focus on health conditions: obesity, type 2 diabetes
o Exclude focus on other health conditions
• Allow focus on other population groups:
o High physical activity
o Diet (e.g., vegetarian, vegan)
• Subgroups of interest
o Age/life stage
o Sex
o Race/ethnicity

Interventions/exposures and Different age or age categories

comparators

Study designs (of primary Include:

studies included within the • Randomized clinical trials
systematic reviews) • Nonrandomized comparative studies, multivariable
adjusted
• Association analyses (e.g., regression models, predictors,
risk factors), multivariable adjusted

Exclude:
• Nonrandomized comparative studies, unadjusted
• Single group (noncomparative between
interventions/exposures)
• Association analyses (e.g., models, predictors, risk
factors), univariate, unadjusted

Timing • No minimum “treatment” or exposure duration required

• No minimum follow-up duration (to when outcome is
measured)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

218 DIETARY REFERENCE INTAKES FOR ENERGY

Setting • Restrict to high- and upper middle-income countries

• Restricted to studies published 2000 or later

Outcomes • Physical activity

o Kcals
o METs
o Physical activity intensity (light, moderate,
vigorous)
o Sedentary behaviors
o Steps
o Total activity counts
o Volume
o Activity type
o Behavior
• Energy expenditure
o Total energy expenditure
o Net energy expenditure
o Activity energy expenditure
o Energy cost of physical activity
• Type of measurement
o Doubly labeled water
o Accelerometer
o Wearable sensors
o Questionnaire
o Objective
o Subjective
o Indirect calorimetry
o Room calorimeter
o Free living
• Obesity risk
o Body weight
o BMI
o Body fat %
o Waist circumference
o Visceral fat

NOTE: BMI = body mass index; METs = metabolic equivalent of task.

Key Question: What is the relationship between different measurements of physical activity
and energy expenditure?
Population • Human only
• Allow focus on health conditions: obesity, type 2 diabetes
o Exclude focus on other health conditions
• Allow focus on other population groups:
o High physical activity
o Diet (e.g., vegetarian, vegan)
• Subgroups of interest
o Age/life stage

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX E 219

o Sex
o Race/ethnicity

Interventions/exposures and Physical activity measurement method

comparators

Study designs (of primary Include:

studies included within the • Randomized clinical trials
systematic reviews) • Nonrandomized comparative studies, multivariable
adjusted
• Association analyses (e.g., regression models, predictors,
risk factors), multivariable adjusted
Exclude:
• Nonrandomized comparative studies, unadjusted
• Single group (noncomparative between
interventions/exposures)
• Association analyses (e.g., models, predictors, risk
factors), univariate, unadjusted

Timing • No minimum “treatment” or exposure duration required

• No minimum follow-up duration (to when outcome is
measured)

Setting • Restrict to high- and upper middle-income countries

• Restricted to studies published 2000 or later

Outcomes • Physical activity

o Kcals
o METs
o Physical activity intensity (light, moderate,
vigorous)
o Sedentary behaviors
o Steps
o Total activity counts
o Volume
o Activity type
o Behavior
• Energy expenditure
o Total energy expenditure
o Net energy expenditure
o Activity energy expenditure
o Energy cost of physical activity
• Type of measurement
o Doubly labeled water
o Accelerometer
o Wearable sensors
o Questionnaire
o Objective

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

220 DIETARY REFERENCE INTAKES FOR ENERGY

o Subjective
o Indirect calorimetry
o Room calorimeter
o Free living
• Obesity risk
o Body weight
o BMI
o Body fat %
o Waist circumference
o Visceral fat
NOTE: BMI = body mass index; kcals = kilocalories; METs = metabolic equivalent of task.

Key Question: What is the association between weight change and chronic disease outcomes?
Population • Human only
• General population (not existing disease state)
o Adults, including postpartum women (lactating
or not) as a subpopulation of interest
o Children

Interventions/exposures and • Body weight change (weight cycling, weight gain, weight
comparators loss, postpartum weight gain/retention). May include
weight maintenance or slowed weight gain
o Weight must be measured (not self-reported)
• Exclude studies focused on unintentional weight loss and
studies of bariatric surgery outcomes

Study designs (of primary Include:

studies included within the • Longitudinal studies (adults)
systematic reviews) • Cohort studies (adults)
• Randomized clinical trials of weight loss interventions

Exclude:
• Cross-sectional studies

Timing • Cohort studies: at least 12 months of follow-up

• Randomized clinical trials: at least 3 months of
intervention

Setting • Restrict to high- and upper middle-income countries

• Restricted to systematic reviews published 2017 or later

Outcomes • Incident cardiovascular disease/coronary heart disease

• Incident type 2 diabetes
• All-cause mortality
• Incident hypertension
• Incident cancer (adults)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX E 221

• Incident hip fracture (adults)

• Dementia (adults)
• cardiovascular disease risk factors (children)
• Metabolic syndrome or prediabetes (children)

Key Question: What is the effect or association of weight cycling on metabolic efficiency
(energy usage/expenditure) and health outcomes?
Population • Human only
• Allow focus on health conditions: obesity, type 2 diabetes
o Exclude focus on other health conditions
• Allow focus on other population groups:
o High physical activity
o Diet type
• Subgroups of interest
o Age/life stage
o Sex
o Race/ethnicity

Interventions/exposures and Weight cycling

comparators Exclude any cointerventions/coexposures (e.g., low fat + vitamin
B12 vs. high fat; physical activity)

Study designs (of primary Include:

studies included within the • Randomized crossover trials (including N-of-1)
systematic reviews) • RCTs (parallel)
• Nonrandomized comparative studies, multivariable
adjusted
• Association analyses (e.g., regression models, predictors,
risk factors), multivariable adjusted
Exclude:
• Nonrandomized comparative studies, unadjusted
• Single group (noncomparative between
interventions/exposures)
• Association analyses (e.g., models, predictors, risk
factors), univariate, unadjusted
• Other study designs

Timing • No minimum “treatment” or exposure duration required

• No minimum follow-up duration (to when outcome is
measured)

Setting • Restrict to high- and upper middle-income countries

• Restricted to studies published 2000 or later

Outcomes • Energy balance/imbalance/excess/deficit

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

222 DIETARY REFERENCE INTAKES FOR ENERGY

o Energy intake
o Energy expenditure (TEE, REE, BMR, BEE,
RMR)
o Body weight
o BMI
• Energy utilization
o Fat/carbohydrate/protein oxidation
o Body fat
o Body weight
• Body composition
o Lean mass
o Fat mass
o Body fat %
o Body weight
• Energy metabolism/metabolic efficiency/metabolic
flexibility
• Obesity risk
o Body weight
o BMI
o Body fat %
o Waist circumference
o Visceral fat
• Type 2 diabetes mellitus risk
o Body weight
o BMI
o Body fat %
o Waist circumference
o Visceral fat
o Glucose
o Insulin
o HOMA-IR

NOTE: BEE = basal energy expenditure; BMI = body mass index; BMR = basal metabolic rate; HOMA-IR =
Homeostatic Model Assessment of Insulin Resistance; REE = resting energy expenditure; RMR = resting metabolic
rate.

Key Questions:
What level of calorie intake is needed to produce weight gain in individuals with underweight?
What amount of calorie intake (deficit) is necessary to produce weight loss in individuals with
overweight or obesity? What level of calorie intake is needed to maintain weight across the
weight spectrum?
Population • Human only
• Allow focus on health conditions: obesity, type 2 diabetes
• Allow focus on other population groups:
o High physical activity
o Diet (e.g., vegan, vegetarian)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX E 223

• Subgroups of interest
o Age/life stage
o Sex
o Race/ethnicity

Interventions/exposures and Measure of caloric intake

comparators Exclude any cointerventions/coexposures (e.g., low fat + vitamin
B12 vs. high fat; physical activity)

Study designs (of primary Include:

studies included within the • Randomized crossover trials
systematic reviews) • Randomized clinical trials
• Nonrandomized comparative studies, multivariable
adjusted
• Association analyses (e.g., regression models, predictors,
risk factors), multivariable adjusted
Exclude:
• Nonrandomized comparative studies, unadjusted
• Single group (noncomparative between
interventions/exposures)
• Association analyses (e.g., models, predictors, risk
factors), univariate, unadjusted
• Other study designs

Timing • No minimum “treatment” or exposure duration required

• No minimum follow-up duration (to when outcome is
measured)

Setting • Restrict to high- and upper middle-income countries

• Restrict to studies published 2000 or later

Outcomes • Energy balance/imbalance/excess/deficit

o Energy intake
o Energy expenditure (TEE, REE, BMR, BEE,
RMR)
o Body weight
o BMI
• Energy use
o Fat/carbohydrate/protein oxidation
o Body fat
o Body weight
• Energy metabolism/metabolic efficiency/metabolic
flexibility
• Obesity risk
o Body weight
o BMI

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

224 DIETARY REFERENCE INTAKES FOR ENERGY

o Body fat %
o Waist circumference
o Visceral fat
• Type 2 diabetes mellitus risk
o Body weight
o BMI
o Body fat %
o Waist circumference
o Visceral fat
o Glucose
o Insulin
o HOMA-IR

NOTE: BEE = basal energy expenditure; BMI = body mass index; BMR = basal metabolic rate; HOMA-IR =
Homeostatic Model Assessment of Insulin Resistance; REE = resting energy expenditure; RMR = resting metabolic
rate.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

Appendix F
AMSTAR 2 Tool

During the data extraction process of the umbrella review, the methodological quality of
each systematic review was evaluated using the Assessment of Multiple Systematic Reviews 2
(AMSTAR 2) quality assessment tool, with some minor adaptations for clarity. The tool consists
of the following series of 15 questions. Alterations to or interpretations of the tool made by the
committee are noted in italic text.

(1) Did the research questions and inclusion criteria for the review include ALL the components
of PICO?
• Yes: Population, Intervention, Comparator, Outcome, AND Follow-up duration
described fully and adequately
• Partial: Described, but not adequately to sufficiently understand eligibility criteria
[Partial was added by the committee]
• No: Not all PICO elements

(2) Did the report of the review contain an explicit statement that the review methods were
established prior to the conduct of the review, and did the report justify any significant deviations
from the protocol?
• Partial: Protocol included "just" (1) review questions, (2) search strategy, (3) eligibility
criteria, AND (4) risk of bias assessment [and/or PROSPERO or other registry]
• Yes: Protocol included all Partial, PLUS (5) meta-analysis or synthesis plan, (6) plan to
investigate heterogeneity, AND (7) justifications for deviation from protocol
• No: Not all criteria met (for Partial) or no mention of a protocol
o Study authors were given the benefit of the doubt in edge cases

(3) Did the review authors explain their selection of the study designs for inclusion in the
review?
• Yes: Provided explanation for selecting study designs (and for not selecting excluded
study designs)
• No: No explanation
o Study authors were given the benefit of the doubt in edge cases. Implicit
explanations were acceptable.

(4) Did the authors use a comprehensive literature search strategy?

• Partial: (1) at least two relevant databases, (2) provided key words or search strategy,
AND (3) justified restrictions (e.g., language)
OR arbitrarily excluded studies (e.g., based on perceived risk of bias)
[latter item added by the committee]

PREPUBLICATION COPY: UNCORRECTED PROOFS

225

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

226 DIETARY REFERENCE INTAKES FOR ENERGY

• Yes: all Partial PLUS (4) searched reference lists, (5) searched study registries, OR (6)
consulted content experts
[Note, AMSTAR 2 says “AND” here]
• No: Not all criteria met (for Partial) OR used a clearly inadequate search strategy
[The latter two reasons were added by the committee]

(5) Did the review authors perform study selection in duplicate?

• Yes: At least two independent screeners, plus a method for reconciling conflicts OR
double screening of a sample with at least 80% agreement, followed by single screening
[Note, committee combined original Yes and Partial into simply Yes]
• No: Less stringent method used

(6) Did the review authors perform data extraction in duplicate?

• Yes: Double independent with reconciliation process OR Single with review by
experienced systematic reviewer
• No: Less stringent method used
o We gave researchers the benefit of the doubt regarding their systematic review
experience.

(7) Did the review authors provide a list of excluded studies and justify the exclusions?
• Yes: Listed and provided exclusion for each OR reported available access to such a list
[Note, committee added the second option here]
• Partial: Listed but did not explain each exclusion
• No: Did not list

(8) Did the review authors describe the included studies in adequate detail?
• Partial: Described “just” each PICOD element OR the studies were described, but with
some limitations for the needs of the committee
[Note, committee added the second option here]
• Yes: Described PICOD elements in detail, including setting and follow-up time OR the
descriptions of the studies were sufficient for the needs of the committee
[Note, committee added the second option here]
• No: Partial not met

(9) Did the review authors use a satisfactory technique for assessing the risk of bias (RoB) in
individual studies that were included in the review?
• Yes: Used a standard risk-of-bias tool (e.g., Cochrane Risk of Bias tool for randomized
trials, ROBINS-I for nonrandomized studies) or equivalent tool that addresses relevant
issues related to randomization/allocation concealment, confounding bias, selection bias,
outcome ascertainment, analytic method
o This framework is based on the concepts described by AMSTAR 2
• Partial: Used an appropriate tool but applied an arbitrary point system to determine level
of quality/risk of bias
o This revision was added, post hoc, upon reviewing eligible systematic reviews

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX F 227

• No: No or inadequate risk-of-bias tool applied

(10) Did the review authors report on the sources of funding for the studies included in the
review?
• Yes
• No

(11) If meta-analysis was performed, did the review authors use the appropriate methods for the
statistical combination of results?
• Yes: (1) Justified combining in meta-analysis, (2) used random effects model (or
equivalent), AND (3) analyzed heterogeneity
o Committee removed concepts related to whether unadjusted analyses were
included and whether studies of different designs were combined
• No: (1) Used fixed effect model (or equivalent) based on heterogeneity measures OR (2)
conducted meta-regression or subgroup analysis subject to ecological fallacy (i.e.,
regressed across the mean value for the sample, such as BMI)
If there were concerns regarding fixed effect models or ecological fallacy, the
relevant analyses were highlighted. The committee did not derive conclusions based on
analyses subject to ecological fallacy

(12) Did the review authors account for RoB in individual studies when interpreting/discussing
the results of the review?
• Yes
• No

(13) Did the review authors provide a satisfactory explanation for, and discussion of, any
heterogeneity observed in the results of the review?
• Yes: Heterogeneity was assessed AND, if present, assessed causes and included as part
of their interpretation of findings
• No: Did not assess heterogeneity OR only enumerated without assessing the impact on
findings

(14) If they performed quantitative synthesis, did the review authors carry out an adequate
investigation of publication bias (small study bias) and discuss its likely impact on the results of
the review?
• Yes: Reported and carried out plan to assess publication bias
[Note: The committee required reporting of a plan to assess publication bias]
• No: Did not report plan to assess publication bias

(15) Did the review authors report any potential sources of conflict of interest, including any
funding they received for conducting the review?
• Yes
• No

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

228 DIETARY REFERENCE INTAKES FOR ENERGY

OVERALL "QUALITY"

The system for determining the quality, or methodological adequacy, of the systematic
reviews was constructed by the committee based on concepts and terminology from AMSTAR 2.
While all AMSTAR 2 questions were answered, not all impacted the overall quality.

Well done/reported systematic reviews

• (1) Adequately reported PICO [Yes, not Partial], (2) had a protocol [Yes/Partial], (4) used
a comprehensive literature search [Yes, not Partial], (5) selected studies in duplicate
[Yes], (6) extracted studies in duplicate or equivalent [Yes], (8) adequately described
studies [Yes], (9) assessed risk of bias [Yes/Partial], (11) used appropriate meta-analysis
techniques if applicable [Yes], (12) accounted for risk of bias [Yes], (13) assessed
heterogeneity [Yes], and (15) reported sources of conflict of interest [Yes]
• Did not require (3) explanation of selected study designs, (7) listing of excluded studies,
(10) reporting of sources of funding for included studies, OR (14) assessment of
publication bias.

Partially well done/reported systematic reviews

Studies were downgraded to Partially well done/reported if
• (1) Partially adequate reporting of PICO [Partial], (2) did not report a protocol
[No], (4) partially adequate literature search [Partial], (6) did not extract in
duplicate or equivalent [No/Unclear], (8) did not adequately describe studies
[Partial/No], (13) did not adequately assess heterogeneity [No], OR (15) did
not report conflicts of interest.

Not well done/reported systematic reviews

Studies were downgraded to Not well-done/reported if
• (1) Inadequate reporting of PICO [No], (4) inadequate literature search [No],
(5) did not select studies in duplicate [No], (9) did not assess risk of bias [No],
(11) used inappropriate meta-analysis technique [No; this may apply only to
selected conclusions/findings within the systematic review], OR (12) did not
account for risk of bias [No].

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

Appendix G
DATA ANALYSIS REPORT

DATA ANALYSIS REPORT

Data Analysis for the Committee to Review the DRIs for Energy

October 19, 2022

Indiana University, School of Public Health – Bloomington
Dean David Allison, Dean and Distinguished Professor, Provost Professor,
Department of Epidemiology & Biostatistics
Dr. Carmen Tekwe, Associate Professor, Department of Epidemiology & Biostatistics
Dr. Roger Zoh, Associate Professor, Department of Epidemiology & Biostatistics
Stephanie Dickinson, Executive Director, Biostatistics Consulting Center
Lilian Golzarri Arroyo, Biostatistician II, Biostatistics Consulting Center
Aaron Cohen, Biostatistician I, Biostatistics Consulting Center
Jocelyn Mineo, Research Associate, Biostatistics Consulting Center

PREPUBLICATION COPY: UNCORRECTED PROOFS

229

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

230 DIETARY REFERENCE INTAKES FOR ENERGY

Table of Contents
Data Analysis for the Committee to Review the DRIs for Energy ... Error! Bookmark not defined.
1. Introduction ............................................................................................................231
1.1 Introduction and Summary of Work ..........................................................................231
2. IAEA Data Preparation ............................................................................................233
3. IOM Data Preparation .............................................................................................235
4. CNRC Data Preparation ...........................................................................................236
5. SOLNAS Data Preparation .......................................................................................237
6. Combined Data .......................................................................................................238
6.1 Age categories ..........................................................................................................238
6.2 PAL categories ..........................................................................................................239
6.3 Data screening ..........................................................................................................240
7. Statistical Methods .................................................................................................242
7.1 Multiple Imputation ..................................................................................................242
7.2 Statistical Modeling ..................................................................................................243
7.3 Model Performance and Evaluation...........................................................................246
7.4 Model Validation ......................................................................................................248
8. Results ....................................................................................................................248
8.1 Descriptive Statistics and plots ..................................................................................248
8.2 TEE Equations ...........................................................................................................250
8.3 Model Performance...................................................................................................254
8.4 External Validation ...................................................................................................256
9. Appendices..............................................................................................................257
Supplemental online files................................................................................................257

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 231

1. Introduction

1.1 Introduction and Summary of Work

The team at Indiana University, School of Public Health-Bloomington was engaged to perform
statistical analysis to derive equations for the estimation of energy expenditure in the general
human population, including pregnant and lactating women in the USA and Canada, based on
data collected across multiple studies using the Doubly Labeled Water (DLW) method for
measuring Total Energy Expenditure (TEE) under free-living conditions.

Four primary sources of data were sought:

1. Institute of Medicine (IOM)’s 2002/2005 Report for DRIs for Macronutrients,
2. International Atomic Energy Agency (IAEA),
3. Study of Latinos: Nutrition and Physical Activity Assessment Study (SOLNAS) data from
the National Heart, Lung and Blood Institute (NHLBI) Biologic Specimen and Data
Repository,
4. Harvard Men’s Lifestyle Study.

Data were obtained from IOM, IAEA, and SOLNAS as described below; however, data were not
obtained from Harvard Men’s Lifestyle Study in time for inclusion in this report. An additional
data source was added for pregnancy data from the Children’s Nutrition Research Center
(CNRC) to increase the sample sizes for pregnant and lactating women.

The first task was to request data from the relevant sources, preparing Data Use Agreements
(DUA’s) as needed, including lists of the specific variables requested, including TEE, Basal
Energy Expenditure (BEE, or basal metabolic rate, BMR), age, sex, height, weight, body
composition, physical activity, health status, athletic status, and country codes. The IU team
then worked diligently to harmonize variable names, recode classifications, and combine these
across datasets, resulting in 8,600 observations (cases) in the pooled dataset.

Multiple tables of descriptive statistics and visualizations were provided for each dataset to
allow Workgroup 1 (WG1) of the DRI energy committee to thoroughly inspect the data.

One challenge for the IU team was to consider how to perform predictive modeling of TEE
based on physical activity level (PAL) while PAL (=TEE/BEE) was missing for 54.2% of the data
(4,662 out of 8,600) where BEE was unavailable. Multiple Imputation was selected as the best
method to impute the data rather than simple estimates BEE (or more accurately BMR) based
on age, sex, and weight in equations such as that published by Schofield15 or others.

15
Schofield, W. N. 1985. Predicting basal metabolic rate, new standards and review of previous work.
Hum Nutr Clin Nutr 39(Suppl 1):5-41.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

232 DIETARY REFERENCE INTAKES FOR ENERGY

Additional work between IU and WG1 involved physical activity data and to consider how PAL
should be included in predictive equations of TEE. Initial models used the same cutoff criteria
and model forms used in the 2002/2005 IOM report, but methods were revised to use different
PAL criteria, which vary by age groups per discussion with WG1, as described below.

Prediction equations were then developed by fitting linear models on TEE based on sex, age,
PAL, weight, height, and body composition. Multiple imputation was used to estimate PAL
across 20 versions of imputed data, where models were fit to each of the 20 imputations, and
the results were pooled to identify final parameter estimates and standard errors (SE) as
defined by Rubin16.

Models were fit for the overall sample and were then separated by including different Body
Mass Index (BMI, kg/height or length2) groups: BMI 18.5 to 40 (removing extremes); BMI 18.5
to 25 (“healthy” only); and BMI 25+ (overweight/obese only) to compare how regression slopes
may differ by weight status groups.

The prediction of TEE with models using height and weight were also compared to those with
height, fat free mass (FFM), and fat mass (FM).

Models were fit separately in 7 strata:

• Infant/Toddler Boys (0 to 2.99 years old),
• Infant/Toddler Girls (0 to 2.99 years old),
• Boys (3.0 to 18.99 years old),
• Girls (3.0 to 18.99 years old),
• Men (19 or over),
• Women (19 and over),
• Pregnant.

Model validation was performed on the external data (described below) provided by WG1 as
summary data extracted from the literature. Parameter estimates from the TEE equations
developed on the main dataset were used to calculate predicted values of TEE on the external
data, and those predicted values were compared to the observed TEE values using measures
such as R-squared and correlation as a measure of model fit and performance.

16
Rubin, D. B. 1987. Multiple imputation for nonresponse in surveys. New York: John Wiley & Sons.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 233

2. IAEA Data Preparation

IAEA data were obtained by submitting an application to the database manager (Dr. John
Speakman) for the IAEA doubly labelled water database (https://doubly-labelled-water-
database.iaea.org/). After the proper procedures and approval process from the database
manager group, the data were transferred to Dr. Allison and his team in an Excel file “IAEA DLW
database 3.6.1 abbreviated for DRI group (allison).xlsx”.

Additional documents used for the data preparation of the IAEA data:

• “IAEA publications description IU 060722 - LGA notes.xlsx” – A list of studies to include

or remove as indicated by WG1
• “CLASS 2022-06-13.xlsx” – Categories of Income by Country to indicate high income
countries for inclusion, provided by WG1
• Growth charts downloaded from https://www.cdc.gov/growthcharts/index.htm:
o “Weightolength_WHO.xlsx” – Percentiles for weight-for-length for infants (0 to
2) per World Health Organization (WHO)
o “BMIcharts.xlsx” – BMI percentiles for children (2 to 18) per Center for Disease
Control (CDC)

The first step in preparation of the data was to apply the exclusion/inclusion criteria defined by
WG1:
1. The International Organization for Standardization (ISO) codes found in the “CLASS
2022-06-13.xlsx” file were used to exclude all studies being done in non-high-income
countries.
2. The codes under the column “Health” in the DLW database were used to include
subjects who were healthy, labeled as H. Subjects with a code beginning with a D, such
as D1 or D15 were excluded.
3. Professional athletes were removed from the dataset by excluding those listed as PA in
the ‘ath’ column.
4. Ineligible studies were removed according to the Excel file “IAEA publications
description IU 060722,” which WG1 highlighted yellow or indicated in workgroup
meetings, for which IU coded as “1” in the Excel file in a column “Remove” to remove
via SAS code. Those coded as “2” indicated special cases, which were inspected
manually to exclude low-income countries or non-healthy participants.

After ineligible studies were removed, pregnant and lactating females were identified in order
to be analyzed separately from other females in the study and coded for trimester and number
of weeks in gestation or in the postpartum period, as follows:

Reproductive Status for Females (rep_statF)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

234 DIETARY REFERENCE INTAKES FOR ENERGY

• L=Lactating
o Most 2020 study was coded as 25 weeks
o Motsiko study was coded as 12 weeks
• P1=Pregnant, 1st trimester (This did not exist in IAEA data after preparation steps
above.)
• P2=Pregnant, 2nd trimester
• P3=Pregnant, 3rd trimester

Next, while fat mass percentage (FM_pct) was provided in the IAEA data, fat mass (FM) was
calculated as:
FM=(FM_pct/100*FFM)/(1-FM_pct/100).

Table 1: Inclusion and exclusion criteria and sample size for the IAEA dataset.
IAEA Inclusion/Exclusion N
Read in the data 7696
Only keep High Income Countries 6989
Remove any subject with a Health Code beginning with D 6744
Remove Professional Athletes (PA) from PA category 6706
Remove other ineligible studies 5966
Remove those without age or sex 5805
Remove participants with PAL < 1 or > 2.5 as defined in section 6.3. 5717

Based on the above, two analysis-ready datafiles were created:

1. one including the 5805 participants for preliminary descriptive statistics and
visualizations before PAL exclusions (“IAEA”), and
2. one with the 5717 for final analysis (“IAEA_clean”).

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 235

3. IOM Data Preparation

IOM data were obtained by extracting data from the pdf versions of the data listed in the
appendix of the IOM’s 2005 publication17 and then converting into Excel.

The following datasets were read in for IOM data:

• “IOM2005_AppendixI_Tables_JM_LGA.xlsx” extracted from “IOM.2005.DRIs for

Macronutrients.pdf”
o TABLE I-1 Infants and Very Young Children (0 Through 2 Years of Age) Within the
3rd to 97th Percentile for Body Mass Index (BMI)
o TABLE I-2 Normal Weight Children, 3 Through 18 Years of Age with Body Mass
Index (BMI) > 85th Percentile
o TABLE I-3 Normal Weight Adults with Body Mass Index (BMI) from 18.5 up to 25
kg/m2
o TABLE I-4 Pregnant Women with Prepregnancy Body Mass Index (BMI) from 18.5
up to 25 kg/m2
o TABLE I-5 Lactating Women with Prepregnancy Body Mass Index (BMI) from 18.5
up to 25 kg/m2
o TABLE I-6 Overweight/Obese Children, 3 Through 18 Years of Age, with Body
Mass Index (BMI) > 85th Percentile
o TABLE I-7 Overweight/Obese Adults with Body Mass Index (BMI) > 25 kg/m2

Gestation weeks in IOM table I-4 were grouped into trimester (P_stage), and lactation months
in I-5 were grouped into 1-3 or 4-6 months postpartum.

Table 2: Inclusion and exclusion criteria and sample size for the IOM dataset.
IOM Inclusion/Exclusion N
Read in Table I-1 320
Read in Table I-2 525
Read in Table I-3 407
Read in Table I-4 22
Read in Table I-5 35
Read in Table I-6 319
Read in Table I-7 360
Additional Combined Pregnancy Lactation Data 382
Merge all tables together 2313

17
Institute of Medicine. 2005. Dietary Reference Intakes for energy, carbohydrate, fiber, fat, fatty acids,
cholesterol, protein, and amino acids. Washington, DC: The National Academies Press.
https://doi.org/10.17226/10490.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

236 DIETARY REFERENCE INTAKES FOR ENERGY

Remove participants with PAL<1 or >2.5 as defined in section 6.3 2283

Based on the above, two analysis-ready datafiles were created:

1. one including the 2313 participants for preliminary descriptive statistics and
visualizations before PAL exclusions (“IOM”), and
2. one including the 2283 for final analysis (“IOM_clean”).

4. CNRC Data Preparation

The following dataset was provided by WG1 and read in for the Children’s Nutrition Research
Center at Baylor College of Medicine (CNRC):

• “CNRC Pregnancy w weeks 2022-08-20.xlsx”

Data includes number of weeks pregnant and weeks of lactation post-partum. Non-pregnant
and non-lactating (NPNL) are also included, which were coded as weeks=0 for analysis.

The dataset included 222 observations across 60 women (with 3 or 4 time points per women at
preconception, second and third trimesters and six months postpartum)

Table 3: Inclusion and exclusion criteria and sample size for the CNRC dataset.
CNRC Inclusion/Exclusion N
Source data from CNRC 222
Remove participants with PAL<1 or >2.5 as defined in section 6.3 220

Based on the above, two analysis-ready datafiles were created:

1. one including the 222 participants for preliminary descriptive statistics and visualizations
before PAL exclusions (“CNRC”), and
2. one including the 220 for final analysis (“CNRC_clean”).

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 237

5. SOLNAS Data Preparation

DLW and physical activity data were obtained for Hispanic adults (19+) from The Study of
Latinos: Nutrition & Physical Activity Assessment Study (SOLNAS) from the National Heart, Lung,
and Blood Institute (NHLBI) BioLINCC site (https://biolincc.nhlbi.nih.gov/studies/hchssol/ ).

The following SAS datasets were read in for SOLNAS data:

• mysolnas.dlwa_lad1.sas7bdat
• mysolnas.vsea_lad1.sas7bdat
• mysolnas.biea_lad1.sas7bdat
• mysolnas.csea_lad1.sas7bdat

For the DLW dataset (DLWA), only urine data were kept for TEE. The following variables were
renamed according to the SOLNAS codebook.

TEE = DLWA33
BMI = DLWA34
FFM = DLWA35
FM = DLWA36
FM_pct = DLWA37

Height and weight were obtained from the main study visit 1 and visit 3 forms (VSEA) dataset,
and renamed as follows:

Height = VSEA3A
Weight = VSEA3B

For the body image (BIEA) data, data were only kept from the main study for gender, and the
following variables were defined according to the codebook:

if BIEA1=1 then Sex = 'M';

if BIEA1=2 then Sex = 'F';

For the calorimetry summary (CSEA) data, data were kept from the main study for age and
calorimeter weight, and the following variables were renamed:

Weight_calorim = CSEA2
Age = CSEA3
EE_mean_kcald= CSEA4D1

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

238 DIETARY REFERENCE INTAKES FOR ENERGY

EE_SD_kcald= CSEA4D2
EE_CV_kcald= CSEA4D3

The variable ‘EE_mean_kcald’ was relabeled as ‘BEE.’

The ethnicity for all participants in this study was coded as ‘Hispanic,’ and none of the
participants were pregnant or lactating.

Physical activity data were also explored where 69 subjects had physical activity data from
Actical. We used the “Actical derived variables at the participant level” data set for minutes per
day of sedentary, light, moderate, and vigorous activity to correlate with PAL from DLW data.

Table 4: Inclusion and exclusion criteria and sample size for the SOLNAS dataset.
SOLNAS Inclusion/Exclusion N
Merge datasets for DLW, VSEA, BIEA, CSEA 393
Remove those without age or sex 382
Remove participants with PAL<1 or >2.5 as defined in section 6.3 380

Based on the above, two analysis-ready datafiles were created:

1. one including the 382 participants for preliminary descriptive statistics and visualizations
before PAL exclusions (“SOLNAS”), and
2. one including the 380 for final analysis (“SOLNAS_clean”).

6. Combined Data
The datasets from IAEA, IOM, SOLNAS, and CNRC pregnancy were harmonized to use consistent
variable names and then combined. Variables are described in Appendix N: DLW Data
Codebook.docx:
SID, Age_cat, Age, Life_Stage, Ethnicity, Sex, BMI, BMIcat, Height, Weight, TEE, BEE,
Percentile, Percentile_group, Percentile_infant, Lactating, Pregnant, P_stage, Weeks,
PAL, PALCAT, BMR_kcal_Schofield, PAL_est, PALCAT_est, FFM, FM, FM_pct.

The combined data set included 8722 participants for preliminary descriptive statistics and
visualizations before PAL exclusions and 8,600 observations after removing participants with
PAL<1 or >2.5 as defined in section 6.3.

Data coding and preparations were performed as follows:

6.1 Age categories

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 239

Age categories were defined as follows for descriptive statistics reports, according to “Life
Stage” as indicated by WG1:
• Infants are 0 to 11.99 months
• Children are 12.0 months to 8.99 years
• Teenagers are 9.0 to 18.99 years
• Adults are 19.0 years to age 101 years

However, the following age categories were used for strata for the final TEE models:
• Infants are 0 to 2.99 years
• Children are 3.0 to 18.99 years
• Adults are 19.0 years and above

6.2 PAL categories

The IOM 2005 report previously classified people into physical activity categories using roughly
25th, 50th, and 75th quartiles of PAL values uniformly across all age groups:

If 1.0=<PAL<1.4 then PALCAT=Sedentary18

If 1.4=<PAL<1.6 then PALCAT=Low Active
If 1.6=<PAL<1.9 then PALCAT=Active
If 1.9=<PAL<2.5 then PALCAT=Very Active

Here, we used the same quartiles (25th, 50th, 75th) to group people into the four categories,
where the workgroup decided to calculate quartiles separately within age groups: 3 - 8.99
years, 9 - 13.99 years, and 14 - 18.99 years. For adults, PAL categories were defined by the
quartiles for 19-70 years, but these PAL categories were applied to all adults, including those
aged 71 and greater.

PAL percentiles as calculated on the raw data (before imputation) are shown in Appendix P
§4.10 as well as after multiple imputation as shown in Appendix Q §2.1 and also in results
section below.

After inspecting the PAL percentiles by age groups (shown below), PAL categories were defined
by WG1 accordingly, and categories (PALCAT) were calculated in the SAS code as follows:

If 3.0=<age<9.0 then do;

If 1.0=<PAL<1.31 then PALCAT=”S”;

18
Note that the term “Sedentary” and PALCAT=”S” is used in this report as well as the analytic
code and output, according to the labels in the 2005 IOM report before the committee relabeled
the lowest level as “inactive”.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

240 DIETARY REFERENCE INTAKES FOR ENERGY

If 1.31=<PAL<1.44 then PALCAT=”LA”;

If 1.44=<PAL<1.59 then PALCAT=”A”;
If 1.59=<PAL<2.5 then PALCAT=”VA”;
end;

If 9.0=<age<14.0 then do;

If 1.0=<PAL<1.44 then PALCAT=”S”;
If 1.44=<PAL<1.59 then PALCAT=”LA”;
If 1.59=<PAL<1.77 then PALCAT=”A”;
If 1.77=<PAL<2.5 then PALCAT=”VA”;
end;

If 14.0=<age<19.0 then do;

If 1.0=<PAL<1.56 then PALCAT=”S”;
If 1.56=<PAL<1.73 then PALCAT=”LA”;
If 1.73=<PAL<1.92 then PALCAT=”A”;
If 1.92=<PAL<2.5 then PALCAT=”VA”;
end;

If 19.0>=age then do;

*Note that these are based on percentiles of 19 to 70.99, but 71+ use these t
oo;
If 1.0=<PAL<1.53 then PALCAT=”S”;
If 1.53=<PAL<1.68 then PALCAT=”LA”;
If 1.68=<PAL<1.85 then PALCAT=”A”;
If 1.85=<PAL<2.5 then PALCAT=”VA”;
end;

6.3 Data screening

Because a PAL greater than 2.5 is considered unsustainable, participants with PAL>2.5 were
removed from analysis. A PAL less than 1 is considered unphysiological, as it’s not possible for
BEE to be larger than TEE. Where data for BEE and PAL were missing, BEE and TEE were
estimated using the Schofield equations19 for the purpose of data screening.

The SAS code for calculating BMR according to Schofield equations is as follows:

If .<Age<3 & Sex="M" then BMR_Mjd_Schofield = 0.0007*Weight + 6.349*Height/10

0 - 2.584;

19
Schofield, W. N. 1985. Predicting basal metabolic rate, new standards and review of previous work.
Hum Nutr Clin Nutr 39(Suppl 1):5-41. PMID: 4044297.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 241

if 3<=Age<10 & Sex="M" then BMR_Mjd_Schofield = 0.082*Weight + 0.545*Height/1

00 + 1.736;
if 10<=Age<18 & Sex="M" then BMR_Mjd_Schofield = 0.068*Weight + 0.574*Height/
100 + 2.157;
if 18<=Age<30 & Sex="M" then BMR_Mjd_Schofield = 0.063*Weight - 0.042*Height/
100 + 2.953;
if 30<=Age<60 & Sex="M" then BMR_Mjd_Schofield = 0.048*Weight - 0.011*Height/
100 + 3.670;
if 60<=Age & Sex="M" then BMR_Mjd_Schofield = 0.038*Weight + 4.068*Height/100
- 3.491;

if .<Age<3 & Sex="F" then BMR_Mjd_Schofield = 0.068*Weight + 4.281*Height/100

- 1.730;
if 3<=Age<10 & Sex="F" then BMR_Mjd_Schofield = 0.071*Weight + 0.677*Height/1
00 + 1.553;
if 10<=Age<18 & Sex="F" then BMR_Mjd_Schofield = 0.035*Weight + 1.948*Height/
100 + 0.837;
if 18<=Age<30 & Sex="F" then BMR_Mjd_Schofield = 0.057*Weight + 1.184*Height/
100 + 0.411;
if 30<=Age<60 & Sex="F" then BMR_Mjd_Schofield = 0.034*Weight + 0.006*Height/
100 + 3.530;
if 60<=Age & Sex="F" then BMR_Mjd_Schofield = 0.033*Weight + 1.917*Height/100
+ 0.074;

*Convert to kcal from MJ (BMR_kcal_Schofield);

BMR_kcal_Schofield = (BMR_Mjd_Schofield*1000)/4.184;

*Calculate PAL_est=TEE_kcal/BMR_kcal_Schofield;
PAL_est=TEE/BMR_kcal_Schofield;
label PAL_est= 'PAL estimated from BMR Schofield, in kcal/day';

The following decision criteria were used to screen high or low values of PAL:

For infants (0 to 11.9 mo) and children 1-3 years:

If PAL (observed) was <1 or >2.5, then that child was removed from the database.
If PAL was unobserved but PAL estimated from Schofield was <1 or >2.5, then that child
was removed from the database.

For children (4+) & adults:

If observed PAL >2.5, then that person was removed from the database.
If observed PAL < 1, then PAL was made missing, and it was filled with multiple
imputation (MI).

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

242 DIETARY REFERENCE INTAKES FOR ENERGY

If PAL was not observed, but PAL estimated from Schofield was >2.5, then that person
was removed from the database.
If PAL was not observed, but PAL estimated from Schofield is <1, the person remained in
the database and PAL was left missing to be imputed (and truncated at 1 during
imputation).
Note that while BEE and PAL estimates from Schofield were not used in TEE models, they were
retained in the dataset during multiple imputation as a “proxy” (or “auxiliary variables”) that
correlated with the variable to be imputed, which improves the precision of estimates.20,21

Table 5: Sample sizes for final analysis dataset, after exclusions, by data source.
Data source N
IAEA 5717
IOM 2283
CNRC 220
SOLNAS 380
Combined data for analysis 8600

Based on the above, two analysis-ready datafile were created:

1. one including the 8722 participants for preliminary descriptive statistics and
visualizations before PAL exclusions (“ALLDATA”), and
2. one including the 8600 for final analysis (“ALLDATA_clean”).

7. Statistical Methods

7.1 Multiple Imputation

PAL is a predictor in the TEE equations but was missing for 54.2% of the data (4,662 out of
8,600). Others in the field have estimated PAL using BEE (or actually, BMR) as estimated by
equations such as Schofield (1985)1 based on age, height, and weight. However, Dr. David
Allison and the IU team preferred to use multiple imputation (MI) to estimate a variety of
possible values of PAL using the information available in the other variables and maintaining
the variability of the true data.

20
Ejima, K., R. Zoh, C. Tekwe, D. Allison, and A. Brown. 2020. What proportion of planned missing
data is allowed for unbiased estimates of the association between energy intake and body weight using
multiple imputation? Curr Dev Nutr 4(Suppl 2):1167. doi: 10.1093/cdn/nzaa056_014. PMCID:
PMC7258036.
21
Cornish, R. P., J. Macleod, J. R. Carpenter, et al. 2017. Multiple imputation using linked proxy
outcome data resulted in important bias reduction and efficiency gains: a simulation study. Emerg Themes
Epidemiol 14(14). https://doi.org/10.1186/s12982-017-0068-0.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 243

Note that PAL estimates from Schofield were retained in the dataset during multiple imputation
as a “proxy” (or “auxiliary variables”) that correlated with the variable to be imputed, which
improved the precision of estimates.22,23,24

The SAS procedure ‘Proc MI’ was used for imputation with Markov Chain Monte Carlo (MCMC)
methods with multiple chains, using twenty imputations.

Proc MI data=df nimpute=pctmissing(min=5 max=20) out=outmi seed=5849975;

mcmc chain=multiple;

The combined clean dataset (n=8600), including all ages and weights and pregnant and
lactating women, was entered into the procedure with all variables in the dataset. All missing
data for all variables in the dataset were simultaneously imputed (e.g. PAL, FM, FFM) providing
twenty versions of a complete dataset.

PAL data were then evaluated again in the imputed data for unrealistic values, where PAL
values <1.0 (unphysiological) were truncated (Winsorized) at 1.0, and observations with values
>2.5 (unsustainable) were removed from that version of imputed data before analysis. Because
there were 20 imputed datasets, these values were only removed in that specific imputation,
and that person would remain in the other imputations where the values remained <=2.5.

7.2 Statistical Modeling

TEE models were fit separately for each strata:

• Infant/Toddler Boys (0 to 2.99 years old),

• Infant/Toddler Girls (0 to 2.99 years old),
• Child/Teen Boys (3.0 to 18.99 years old),
• Child/Teen Girls (3.0 to 18.99 years old),
• Men (19 and over),
• Women (19 and over),

22
Ejima, K., R. Zoh, C. Tekwe, D. Allison, and A. Brown. 2020. What proportion of planned missing
data is allowed for unbiased estimates of the association between energy intake and body weight using
multiple imputation? Curr Dev Nutr 4(Suppl 2):1167. doi: 10.1093/cdn/nzaa056_014. PMCID:
PMC7258036.
23
Cornish, R. P., J. Macleod, J. R. Carpenter et al. 2017. Multiple imputation using linked proxy outcome
data resulted in important bias reduction and efficiency gains: a simulation study. Emerg Themes
Epidemiol 14(4).
24
Li, P., and E. A. Stuart. 2019. Best (but oft-forgotten) practices: Missing data methods in randomized
controlled nutrition trials. Am J Clin Nutr 109(3):504-508. doi: 10.1093/ajcn/nqy271. PMID: 30793174;
PMCID: PMC6408317. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6408317/.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

244 DIETARY REFERENCE INTAKES FOR ENERGY

• Pregnant Women.

Within each strata, the linear models were fit separately on each version of the imputed
dataset, obtaining the relevant regression equations in each iteration.

TEE is estimated as a function of a person’s age, height and weight, and Physical Activity Level,
as a categorical measure “PALCAT”: Sedentary/Inactive, Low Active, Active, Very Active as
described above based on PAL (=TEE/BEE). Interaction terms are used to fit separate slopes for
the effect of height and weight within each PALCAT.

TEE = Intercept + Age (years) + Height (cm) + Weight (kg) + PALCAT + PALCAT x Weight +
PALCAT x Height

Because the raw parameter estimates for the model with interactions are not easily
interpretable to a non-statistical audience, the IU team used ‘estimate’ statements in SAS25,26 to
calculate the estimates (and Standard Errors) for slopes of weight and height within each
PALCAT.

The SAS code is pasted here, where &agecat. is a macro variable representing each age/sex
strata. Also note that when reading the code below, the levels of PALCAT occur in alphabetical
order in SAS (Active, Low Active, Sedentary, Very Active).

proc glm data=df&agecat.;

by _imputation_;
class PALCAT;
model TEE = Age PALCAT Weight Height PALCAT*Weight PALCAT*Height /solu
tion;

estimate 'Sedentary: Intercept' Intercept 1 PALCAT 0 0 1 0;

estimate 'Sedentary: Age' Age 1;
estimate 'Sedentary: Weight' Weight 1 PALCAT*Weight 0 0 1 0;
estimate 'Sedentary: Height' Height 1 PALCAT*Height 0 0 1 0;

estimate 'Low Active: Intercept' Intercept 1 PALCAT 0 1 0 0;

estimate 'Low Active: Age' Age 1;
estimate 'Low Active: Weight' Weight 1 PALCAT*Weight 0 1 0 0;

25
SAS Institute Inc. 2013. SAS/STAT 13.1 User’s Guide. Cary, NC: SAS Institute Inc., page 3459
https://support.sas.com/documentation/onlinedoc/stat/131/glm.pdf
26
Introduction to SAS. UCLA: Statistical Consulting Group, from https://stats.oarc.ucla.edu/sas/faq/how-
do-i-write-an-estimate-statement-in-proc-glm/

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 245

estimate 'Low Active: Height' Height 1 PALCAT*Height 0 1 0 0;

estimate 'Active: Intercept' Intercept 1 PALCAT 1 0 0 0;

estimate 'Active: Age' Age 1;
estimate 'Active: Weight' Weight 1 PALCAT*Weight 1 0 0 0;
estimate 'Active: Height' Height 1 PALCAT*Height 1 0 0 0;

estimate 'Very Active: Intercept' Intercept 1 PALCAT 0 0 0 1;

estimate 'Very Active: Age' Age 1;
estimate 'Very Active: Weight' Weight 1 PALCAT*Weight 0 0 0 1;
estimate 'Very Active: Height' Height 1 PALCAT*Height 0 0 0 1;

ods output Estimates=Est_&agecat. ParameterEstimates=Par_&agecat.;

output out=newdf&agecat. p=predicted UCL=UCL LCL=LCL STDI=STDI;
run;
quit;

Coefficients obtained from the ‘estimate’ statements can then be presented more simply, to
display an equation within each PAL category as:

TEE = Intercept + A x Age (years) + B x Height (cm) + C x Weight (kg)

where ‘A’, ‘B’, and ‘C’ are the model-based coefficients for slopes of Age, Height, and Weight,
respectively, for each PAL category.

Parameter estimates are then pooled across the 20 imputations using ‘proc mianalyze’ in SAS to
provide final parameter estimates and standard errors.
SAS code is pasted here, where &agecat. is a macro variable representing each age/sex
strata.

proc mianalyze data=Est_&agecat.;

by Parameter;
modeleffects Estimate ;
stderr StdErr;
ods output ParameterEstimates=Pooled_&agecat.;
run;

Children 0 to <3 years old did not have separate models by PAL category; all data were pooled.

Analysis of pregnancy data included longitudinal data for women by trimester. Non-pregnant
and non-lactating (NPNL) were included, coded as weeks=0 for analysis. Linear mixed models

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

246 DIETARY REFERENCE INTAKES FOR ENERGY

were performed with Proc Mixed in SAS using a repeated statement to account for the
correlation of data over time within women. A variable was also added to the model for weeks
of pregnancy.

7.3 Model Performance and Evaluation

Model performance is calculated as:

• R2
• Adjusted R2
• Shrunken R2 (Browne formula)27
• Pearson r correlation
• MAPE (Mean Absolute Percentage Error)
• MAE (Mean Absolute Error)
• Mean squared error (MSE)
• Root mean squared error (RMSE)

using the following equations28:

27
Yin, P., and X. Fan. 2001. Estimating R2 shrinkage in multiple regression: A comparison of different
analytical methods. J Experiment Educ 69(2), 203–224. http://www.jstor.org/stable/20152659.
28
Tibshirani, R., T. Hastie, G. James, and D. Witten. 2021. An introduction to statistical learning: With
applications in R. United States: Springer US.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 247

These are calculated in R code as follows, where ‘TEE’ is the true observed data (𝑦𝑖 ), and
‘TEE_pred.o’ is the predicted value for TEE (𝑦̂𝑖 ). In the code below, the ‘o’ in TEE_pred.o is for
the overweight/obese model, and the same is done for each BMI classification.

SSR = sum((TEE- TEE_pred.o)^2),

SST = sum((TEE- mean(TEE))^2),
R2= 1 - SSR/SST,
R2adj = 1 - ((1-R2) *(n-1)/(n-12-1)), #p=12, number of predi
ctors
R2shr = ((n-12-3) *(R2adj)^2 + R2adj) /( (n-2*12-2)*R2adj +
12), #p=12
r = cor(TEE, TEE_pred.o, method="pearson"),
MSE = SSR/n,
RMSE = sqrt(MSE),
MAE = sum(abs(TEE - TEE_pred.o))/n)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

248 DIETARY REFERENCE INTAKES FOR ENERGY

'MAPE (%)' = sum(abs((TEE-TEE_pred.o)/TEE))/n*100,

Note that both the RMSE and MAE are in the same units as the original TEE (kcal/day).

7.4 Model Validation

An out-of-sample model validation was performed on an external data set provided by WG1
(“Data extraction combined FINAL 081122”), which contained summary data (Means and SD) of
DLW studies extracted from the literature and not in the combined DLW database.

Parameter estimates from the TEE equations developed on the main dataset were used to
calculate the predicted values of TEE on the external data, and those predicted values were
compared to the observed (Mean) TEE values in the external validation data using the same
measures described above, such as the R-squared and Pearson correlation of observed vs
predicted values, as a measure of model fit and performance.

8. Results
All statistical output was stored as HTML (.html) files created in R markdown (.Rmd).
Additionally, to include results in the NASEM online appendix, the html output was also
converted to PDF (.pdf) format.

Four pieces of output (each in html and pdf) are:

1. “DRI-energy-data-prep-and-prelim-stats”
2. “DRI-energy-clean-analysis”
3. “DRI-energy-MI-glm”
4. “Performance-Report”

Key findings are presented below.

8.1 Descriptive Statistics and plots

Summary tables and descriptive plots are provided in:

(1) “DRI-energy-data-prep-and-prelim-stats” for all eligible data before excluding or truncating
based on PAL<1 or >2.5 (as described in section 6.3) (n=8,722), and
(2) “DRI-energy-clean-analysis” from final data for analysis (n=8,600)

Table 6: Sample sizes for final analysis dataset, by data source and age group. (Appendix P
§2.1)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 249

CNRC IAEA IOM SOLNAS TOTAL

Infants 0 378 177 0 555

Children 0 432 689 0 1121

Teenagers 0 425 279 0 704

Adults 0 4309 767 380 5456

Preg/Lac/NPNL29 220 173 371 0 764

TOTAL 220 5717 2283 380 8600

Detailed descriptive statistics for the 8,600 observations included are presented in Appendix P,
Section (§)3.

The non-linear relationship of TEE over age is shown in Figures 1 and 2 below (as well as in
Appendix P §4).

Figure 1. The relationship of TEE vs Age (Appendix P §4.18.5)

29
NPNL is non-pregnant non-lactating but were women included in the studies of pregnant or
lactating.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

250 DIETARY REFERENCE INTAKES FOR ENERGY

Figure 2. The relationship of PAL vs Age (Appendix P §4.11.5)

8.2 PAL percentiles

PAL percentiles calculated from the imputed data were used to inform the quartiles by age
group to use in classifying PAL levels. Bold numbers in Table 7 below were used to define the
new age-dependent PAL categories as described above (Appendix Q §2.1).

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 251

Table 7: PAL percentiles from imputed data by age categories (Appendix Q §2.1.1)
0-2.99 y 3-8.99 y 9-18.99 y 19-70.99 71+ y Lactating Pregnant
Percentile n=750 n=926 n=704 y n=4299 n=1281 n=203 n=431

10% 1.00 1.20 1.34 1.39 1.31 1.34 1.30

25% 1.11 1.31 1.50 1.53 1.46 1.50 1.46

50% 1.27 1.44 1.66 1.68 1.62 1.69 1.60

75% 1.44 1.59 1.85 1.85 1.79 1.83 1.77

90% 1.61 1.75 2.04 2.03 1.95 2.05 1.97

14-
1-3 4-8 9-13 18
0-6 7-11 year year year year 19-30 31-50 51-70 >=71
month month s s s s years years years years
Percenti s s n=24 n=87 n=30 n=40 n=141 n=199 n=151 n=128
le n=443 n=112 3 8 4 3 7 4 9 1

10% 1.00 1.08 1.06 1.20 1.29 1.40 1.35 1.39 1.39 1.31

25% 1.07 1.19 1.17 1.32 1.44 1.56 1.50 1.53 1.52 1.46

50% 1.23 1.31 1.33 1.44 1.59 1.73 1.67 1.69 1.67 1.62

75% 1.40 1.47 1.49 1.60 1.77 1.92 1.85 1.86 1.82 1.79

90% 1.58 1.65 1.64 1.76 1.92 2.11 2.05 2.03 1.99 1.95

The distribution of PAL within age group is shown in Figure 3 (and Appendix Q §2.2).

Figure 3. Histograms of PAL by age group (Appendix Q §2.2)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

252 DIETARY REFERENCE INTAKES FOR ENERGY

8.3 TEE Equations

Final TEE models with coefficients for age, height, and weight by PAL category (Sedentary, Low
Active, Active, Very Active) and age/sex strata are as follows (from Appendix Q §11):

Women, 19 years and above

• Sedentary: 584.90 -7.01 Age (y) + 5.72 Height (cm) + 11.71 Weight (kg)
• Low Active: 575.77 -7.01 Age (y) + 6.60 Height (cm) + 12.14 Weight (kg)
• Active: 710.25 -7.01 Age (y) + 6.54 Height (cm) + 12.34 Weight (kg)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 253

• Very Active: 511.83 -7.01 Age (y) + 9.07 Height (cm) + 12.56 Weight (kg)

Men, 19 years and above

• Sedentary: 753.07 -10.83 Age (y) + 6.50 Height (cm) + 14.10 Weight (kg)
• Low Active: 581.47 -10.83 Age (y) + 8.30 Height (cm) + 14.94 Weight (kg)
• Active: 1004.82 -10.83 Age (y) + 6.52 Height (cm) + 15.91 Weight (kg)
• Very Active: -517.88 -10.83 Age (y) + 15.61 Height (cm) + 19.11 Weight (kg)

Girls, 3-18 years old

• Sedentary: 55.59 -22.25 Age (y) + 8.43 Height (cm) + 17.07 Weight (kg)
• Low Active: -297.54 -22.25 Age (y) + 12.77 Height (cm) + 14.73 Weight (kg)
• Active: -189.55 -22.25 Age (y) + 11.74 Height (cm) + 18.34 Weight (kg)
• Very Active: -709.59 -22.25 Age (y) + 18.22 Height (cm) + 14.25 Weight (kg)

Boys, 3-18 years old

• Sedentary: -447.51 + 3.68 Age (y) + 13.01 Height (cm) + 13.15 Weight (kg)
• Low Active: 19.12 + 3.68 Age (y) + 8.62 Height (cm) + 20.28 Weight (kg)
• Active: -388.19 + 3.68 Age (y) + 12.66 Height (cm) + 20.46 Weight (kg)
• Very Active: -671.75 + 3.68 Age (y) + 15.38 Height (cm) + 23.25 Weight (kg)

Girls, 0-2 years old

• -69.15 + 80.00 Age (y) + 2.65 Height (cm) + 54.15 Weight (kg)

Boys, 0-2 years old

• -716.45 -1.00 Age (y) + 17.82 Height (cm) + 15.06 Weight (kg)

Pregnant women

• Sedentary: 1131.20 -2.04 Age (y) + 0.34 Height (cm) + 12.15 Weight (kg) + 9.16 Weeks
pregnant
• Low Active: 693.35 -2.04 Age (y) + 5.73 Height (cm) + 10.20 Weight (kg) + 9.16 Weeks
pregnant
• Active: -223.84 -2.04 Age (y) + 13.23 Height (cm) + 8.15 Weight (kg) + 9.16 Weeks
pregnant
• Very Active: -779.72 -2.04 Age (y) + 18.45 Height (cm) + 8.73 Weight (kg) + 9.16 Weeks
pregnant

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

254 DIETARY REFERENCE INTAKES FOR ENERGY

Model coefficients are shown here for the primary models including all BMI levels. Coefficients
for sensitivity analyses removing high and low BMI or separated for “healthy” or
“overweight/obese” are included for each strata in Appendix Q §3 (Women 19+) through §9
(Pregnant).

8.4 Model Performance

Model performance and validation is outlined in the “Performance-Report”. A summary of

model fit measures for the primary models including all BMI levels are listed here in Table 8
(Appendix R §1.8).

Table 8. Model fit for each age/sex strata (Appendix R §1.8)

Strata n R2 R2adj R2shr r MSE RMSE MAPE (%) MAE

Adult Women 19+ 1342 0.71 0.70 0.70 0.84 60393.43 245.75 8.67 190.89
Adult Men 19+ 1016 0.73 0.73 0.73 0.86 114615.05 338.55 9.35 265.54
Child Girls 3-18 477 0.84 0.84 0.83 0.92 56049.24 236.75 8.19 165.44
Child Boys 3-18 250 0.92 0.92 0.92 0.97 66831.33 258.52 7.11 163.25
Child Girls 0-2 432 0.83 0.83 0.83 0.91 9059.24 95.18 12.80 73.51
Child Boys 0-2 317 0.83 0.83 0.83 0.91 10732.61 103.60 13.56 79.47
Pregnancy 413 0.63 0.62 0.61 0.80 79769.92 282.44 8.80 222.10
R2adj
= adjusted R2, R2shr = shrunken R2, as described in methods above, along with MSE,
MAPE, and MAE.

Table 9 shows the mean and standard deviation of the difference in observed TEE minus-
predicted TEE (i.e. the error) from the primary models in each strata (Appendix R §1.10). The
mean of the error is useful as a measure of bias, indicating a general tendency for whether the
true values tend to be above or below the predicted values. Bland-Altman plots30 are displayed
in Appendix R §1.10 to visually display the differences in observed minus-predicted values

Table 9. Mean (Bias) and standard deviation of the difference in observed TEE minus-predicted
TEE, by strata (Appendix R §1.10)
Strata n Mean Std Dev.

Adult Women 19+ 1342 0.611 245.842

Adult Men 19+ 1016 29.387 337.437

30
P. S. Myles, J. Cui, I. 2007. Using the Bland–Altman method to measure agreement with repeated
measures, BJA: Brit J Anaesthesia 99(3):309–311. https://doi.org/10.1093/bja/aem214.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 255

Strata n Mean Std Dev.

Child Girls 3-18 477 11.452 236.718

Child Boys 3-18 250 43.162 255.400

Child Girls 0-2 432 0.510 95.289

Child Boys 0-2 317 0.059 103.762

Pregnancy 413 39.593 279.986

Prediction Error was calculated to show the precision of estimates if a new person’s TEE was
predicted based on their age, height, weight, and PAL. We first calculated the predicted TEE and
Standard Error (SE) based on a person at an average level of age, height and weight, with an
“Active” PAL level within each strata (Table 10A, Appendix Q §10.1) and then also for someone
above average (2 standard deviations above the mean) (Table 10B, Appendix Q §10.2).

Table 10A. Prediction Error for estimating a new person’s TEE at average levels of age, height
and weight, with “Active” PAL level (Appendix Q §10.1)

SE of the predicted
Strata Age Height Weight Weeks preg Predicted TEE
value

Adult Women 19+ 53.87 162.34 71.87 . 2280.94 240.93

Adult Men 19+ 50.25 175.92 83.10 . 2930.26 342.37
Child Girls 3-18 9.58 135.02 37.63 . 1872.65 221.06
Child Boys 3-18 8.65 134.03 37.06 . 2098.77 257.61
Child Girls 0-2 0.72 68.31 7.81 . 592.58 95.87
Child Boys 0-2 0.69 68.46 8.03 . 623.81 104.42
Pregnant 29.40 164.13 74.89 19.86 2679.76 302.35

Table 10B. Prediction Error for estimating a new person’s TEE at 2 SD above average levels of
age, height and weight, with “Active” PAL level (Appendix Q §10.2)

SE of the predicted
Strata Age Height Weight Weeks preg Predicted TEE
value
Adult Women
53.87 169.43 87.99 . 2526.23 241.28
19+

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

256 DIETARY REFERENCE INTAKES FOR ENERGY

Adult Men 19+ 50.25 183.46 99.55 . 3241.22 343.32

Child Girls 3-18 9.58 157.71 57.12 . 2496.52 223.35
Child Boys 3-18 8.65 161.26 58.34 . 2878.92 262.99
Child Girls 0-2 0.72 78.81 10.36 . 758.52 96.96
Child Boys 0-2 0.69 79.06 10.76 . 853.83 105.88
Pregnant 29.40 170.72 95.07 19.86 2931.34 306.19

8.5 External Validation

Model fit statistics were used to evaluate the out-of-sample data as described above. Table 11
shows the model fit from the predicted values after applying the TEE models to the study-level
data extracted from the literature (Appendix R §4.2). The analyses were performed first for the
studies with PAL available, with the second imputing PAL using Schofield equations based on
the study averages for age, height, and weight.

Table 11A. Model fit from external validation with complete data for PAL (Appendix R §4.2.1)
MAPE
Strata n R2 r MSE RMSE (%)

Boy 8 0.90 0.96 95651.47 309.28 7.90

Girl 7 0.82 0.97 56812.37 238.35 10.38

Man 14 0.85 0.93 45699.40 213.77 6.00

Woman 25 0.85 0.94 44940.87 211.99 6.83

Table 11B. Model fit from external validation with PAL imputed for Schofield equations
(Appendix R §4.2.2)

Strata n R2 r MSE RMSE MAPE (%)

Boy 21 0.92 0.96 61755.86 248.51 7.72

Girl 20 0.87 0.97 35342.89 188.00 8.01

Man 32 0.82 0.92 49684.74 222.90 5.57

Woman 71 0.82 0.93 28833.66 169.80 5.46

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX G 257

9. Appendices31

Supplemental online files

Appendix Description
Appendix N Codebook for DLW database
Appendix O R code and output for preliminary stats
Appendix P R code and output for descriptive stats for clean data
Appendix Q R code and output for MI and main models
Appendix R R code and output for model fits, performance, and
validation
Appendix S List of IAEA studies with inclusion/exclusion
Appendix T IOM Data extracted from 2002/2005 report pdf
Appendix U External validation data
Appendix V SAS code for Importing, harmonizing, and merging data
Appendix W SAS code for multiple imputation and models

31
All appendixes to this IU report are provided in Supplemental Appendixes N through W and
are available at: https://nap.nationalacademies.org/catalog/26818.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

Addendum to Appendix G
Details of Redefining of the TEE Model

As described in Chapter 5, a general model of TEE used age, height, weight, and PAL
category as predictors and also included interactions of the PAL category with height and
weight. The model performed to predict TEE used the following format:

TEE = Intercept0 + A × Age (years) + B0 × Height (cm) + C0 × Weight (kg)

+ Di × PALCATi + IBDi × PALCATi × Height (cm) + ICDi × PALCATi × Weight
(kg) + error

where PALCATi represents 3 indicator variables for PAL category (Active, Low Active,
Inactive) that are coded as 0 or 1; ‘A’, ‘B0’, ‘C0’, and ‘Di’ are the model coefficients for the
main effects of age, height, weight and the 3 PAL categories, respectively; and ‘IBDi’ and ‘IBCi’
are the model coefficients for the interaction of the 3 PAL categories with height and weight,
respectively. (The full model output including all the coefficients for interaction terms of height
and weight by PAL category are provided in supplemental Appendix Q32). Moving the main
effect of PAL category, and regrouping the terms by height and weight yields:

TEE = Intercept0 + D × PALCATi + A × Age (years) + (B0 + IBDi × PALCATi)

× Height (cm) + (C0 + ICDi × PALCATi ) × Weight (kg) + error

In this model, the intercept represents the mean TEE level when Age, Weight, and Height are
all 0. Obviously, this does not occur, and therefore it is not meaningful by itself, and could even be
negative. The coefficients for Age, Height, and Weight may be thought of as slopes, i.e., positive
slopes represent increasing energy expenditure, and negative slopes are decreasing energy expenditure
for a change in the corresponding variable holding the other values constant (e.g., for adult females,
there is on average a decrease of 10.83 kcal/d for each 1-year increase in age, for women of the same
weight, height, and physical activity level). The interaction terms allow the height and weight effects
to differ for each PAL category, and the interaction coefficient (IBDi for height, ICDi for weight)
represents the deviation from the referent group (Very Active).
Recognizing that PALCATi represents 3 indicator (0 or 1) variables (i=Active, Low Active,
Inactive) and that Very Active is the reference category (all 3 are 0), we can write the predicted value
for each category by substituting the 0 or 1 for PALCATi. For example, for the Active group:

TEEActive = (Intercept0 + DActive × 1) + (DInactive × 0) + (DLActive × 0) + A × Age

(years) + (B0 + IBDActive × 1 ) × Height (cm) + (B0 + IBDInactive × 0 )× Height
(cm) + (B0 + IBDLActive × 0 )× Height (cm) + (C0 + ICDActive × 1) × Weight (kg) +
(C0 + ICDInactive × 0) × Weight (kg) + (C0 + ICDLActive × 0) × Weight (kg)

which, after multiplying by the indicator values of 0 or 1, simplifies to:

32
Supplemental appendixes are available at: https://nap.nationalacademies.org/catalog/26818.

PREPUBLICATION COPY: UNCORRECTED PROOFS

259

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

260 DIETARY REFERENCE INTAKES FOR ENERGY

TEEActive = InterceptActive + A × Age (years) + (B0 + IBDActive)× Height (cm) +

(C0 + ICDActive ) × Weight (kg).

where InterceptActive = Intercept0 + DActive.

This equation can be written simply as:

TEEi = Intercept + A × Age (years) + B× Height (cm) + C × Weight (kg)

where ‘A’ is same as above, ‘Intercept’ represents the sum of the intercept in the full model
(Intercept0) and the ‘Di’ coefficient for the indicator for the PAL category, ‘B’ is the sum of the ‘B0’
coefficient from the full model and the ‘IBDi’ coefficient from the full model for the corresponding
PAL category, and ‘C’ is the sum of the ‘C0’ coefficient from the full model and the ‘ICDi’ coefficient
from the full model for the corresponding PAL category. All PAL levels could be predicted in the
same manner. For the special case of the referent group (Very Active), all indicator variables are 0, so
the prediction is simply:

TEEVActive = Intercept0 + A × Age (years) + B0× Height (cm) + C0 × Weight (kg)

which also simplifies to the equation directly above.

DIFFERENCE IN EQUATIONS COMPARED TO THE 2005 EER:

For the 2005 EER, the following prediction of TEE was used:

TEE = Intercept + A × Age (years) + PA × (B × Height (cm) + C × Weight (kg))

where ‘A’, ‘B’, and ‘C’ are the coefficients for age, height, and weight respectively, and ‘PA’ is a
coefficient for each PAL category that is multiplied by both height and weight. By substitution of the
4 coefficients for PA, this prediction could also be written separately for each PAL category, as above.
Also, similar to the TEE prediction equation above, the coefficient for Age remains constant for each
PAL category. However, in contrast to the TEE prediction equation above, the intercept also remains
constant, and, although the coefficients for Height and Weight vary by PAL category, they are
mutltiplied by the same PA coefficient, whereas in the equation above, the parameters represent a
deviation from the overall slope, which is not restricted to be the same for height and weight. A
comparison of the EER values from 2005 and 2023 is presented in Chapter 7.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

Appendix H
Characteristics of the DLW Database

Box H-1
Data Sets and Variables Requested from SOLNAS

Data Sets Requested: biea_lad1.sas7bdat; csea_lad1.sas7bdatt; dlwa_lad1.sas7bdat;

vsea_lad1.sas7bdat; vtea_lad1.sas7bdat
Variables requested:
Age (CSEA3)
Participant Gender (BIEA1)
Starting Weight (VSEA3B)
Height (VSEA3A)
BMI (DLWA34)
Ethnicity (all will be labeled as Hispanic)
Year of data collection (all should be coded as 2011-12)
Country (will be labeled as US)
Calorimetry body weight (CSEA2)
Final weight (VTEA2A)
kO (DLWA29)
kH (DLWA28)
18O space (DLWA25)
2H space (DLWA23)
VCO2 (DLWA32)
RQ mean, SD, CV (CSEA4C1, C2, C3)
RQ source (will be labeled as indirect calorimetry for all)
Lean body mass (DLWA35)
Body fat (DLWA36)
EE mean, SD, CV (CSEA4D1, D2, D3)
BEE method (calorimetry)
TEE0-12 days (DLWA33)
FQ (DLWA13)
FQ source (DLWA19)
Data Sets Requested: Physical Activity data sets: pa_derv_solnas_lad1.sas7bddat
Variables requested: all the physical activity derived variables for those included in SOLNAS
Data Sets Requested: Smoking variables from tbea_lad1.sas7bdat
Variables requested: TBEA1-smoke at least 100 cigs in lifetime; TBEA3-present smoking status

Data Sets Requested: Health status variables from mhea_lad1.sas7bdat

Variables Requested: MHEA1-HBP/hypertension-self reported; MHEA16-diabetes-self reported.

PREPUBLICATION COPY: UNCORRECTED PROOFS

261

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

262 DIETARY REFERENCE INTAKES FOR ENERGY

SOURCE: Data from SOLNAS are publicly available on NHLBI’s Biological Specimen and Data Repository
Information Coordinating Center (BIOLINCC) site: https://biolincc.nhlbi.nih.gov/home/.

TABLE H-1 Characteristics of All Participants, 0–100+ Years, Included in Combined DLW
Database, Except Pregnant or Lactating Women
Females Males Overall
(N = 5,025) (N = 2,919) (N = 7,944)
Age (years)
N 5,025 2,919 7,944
Mean (SD) 40.52 (27.16) 34.86 (26.36) 38.44 (27.01)
40.00 [0.03, 33.00 [0.02, 37.05 [0.02,
Median [Min, Max]
98.00] 101.00] 101.00]
Life Stage
0–6 months 266 (5.29%) 203 (6.95%) 469 (5.90%)
7–11 months 69 (1.37%) 45 (1.54%) 114 (1.44%)
1–3 years 145 (2.89%) 105 (3.60%) 250 (3.15%)
4–8 years 495 (9.85%) 384 (13.2%) 879 (11.1%)
9–13 years 184 (3.66%) 120 (4.11%) 304 (3.83%)
14–18 years 259 (5.15%) 158 (5.41%) 417 (5.25%)
19–30 years 633 (12.6%) 368 (12.6%) 1,001 (12.6%)
31–50 years 987 (19.6%) 703 (24.1%) 1,690 (21.3%)
51–70 years 1,079 (21.5%) 448 (15.3%) 1,527 (19.2%)
>= 71 years 908 (18.1%) 385 (13.2%) 1,293 (16.3%)
Sex
Females 5,025 (100%) 0 (0%) 5,025 (63.3%)
Males 0 (0%) 2919 (100%) 2919 (36.7%)
Ethnicity
African American 605 (12.08%) 235 (8.056%) 840 (10.526%)
Asian 103 (2.05%) 96 (3.29%) 199 (2.51%)
White 2,240 (44.6%) 1,282 (43.9%) 3,522 (44.3%)
Hispanic 350 (6.97%) 201 (6.89%) 551 (6.94%)
Other 44 (0.876%) 29 (0.993%) 73 (0.919%)
Unknown or data not available 1,683 (33.5%) 1,076 (36.9%) 2,759 (34.7%)
Height (cm)
N 5,013 2,901 7,914

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX H 263

Females Males Overall

(N = 5,025) (N = 2,919) (N = 7,944)
Mean (SD) 148.73 (29.99) 154.27 (38.27) 150.76 (33.37)
160.00 [46.40, 172.00 [47.30, 162.50 [46.40,
Median [Min, Max]
196.00] 204.70] 204.70]
Weight (kg)
N 5,022 2,913 7,935
Mean (SD) 59.55 (26.51) 63.83 (32.32) 61.12 (28.85)
63.30 [2.36, 72.95 [2.70, 66.00 [2.36,
Median [Min, Max]
164.55] 215.70] 215.70]
Fat-Free Mass (kg)
N 3,833 2,143 5,976
Mean (SD) 38.79 (12.48) 48.32 (20.46) 42.21 (16.46)
40.91 [1.80, 54.61 [2.28, 43.41 [1.80,
Median [Min, Max]
95.74] 97.81] 97.81]
Fat Mass (kg)
N 3,829 2,139 5,968
Mean (SD) 26.09 (13.62) 19.92 (12.56) 23.88 (13.58)
25.33 [0.12, 19.72 [0.07, 23.39 [0.07,
Median [Min, Max]
92.14] 90.27] 92.14]
Fat Mass %
N 3,829 2,139 5,968
Mean (SD) 37.84 (8.85) 27.67 (8.32) 34.20 (9.94)
38.67 [3.69, 27.88 [2.06, 34.53 [2.06,
Median [Min, Max]
70.08] 56.19] 70.08]
Body Mass Index
N 5,010 2,897 7,907
Mean (SD) 24.73 (6.78) 23.84 (6.10) 24.40 (6.55)
24.04 [10.77, 23.86 [10.89, 23.96 [10.77,
Median [Min, Max]
57.87] 61.70] 61.70]
BMI Category
< 18.5 kg/m2 986 (19.7%) 681 (23.5%) 1,667 (21.1%)
>= 18.5 and < 25 kg/m2 1,824 (36.4%) 998 (34.4%) 2,822 (35.7%)
2
>= 25 and < 30 kg/m 1,208 (24.1%) 808 (27.9%) 2,016 (25.5%)
>= 30 and < 35 kg/m2 605 (12.1%) 283 (9.77%) 888 (11.2%)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

264 DIETARY REFERENCE INTAKES FOR ENERGY

Females Males Overall

(N = 5,025) (N = 2,919) (N = 7,944)
>= 35 and < 40 kg/m2 250 (4.99%) 92 (3.18%) 342 (4.33%)
2
>= 40 kg/m 137 (2.73%) 35 (1.21%) 172 (2.18%)
BMI Percentile
N 1,406 997 2,403
Mean (SD) 59.13 (27.02) 58.88 (28.92) 59.03 (27.82)
60.40 [0.73, 60.61 [0.09, 60.43 [0.09,
Median [Min, Max]
99.38] 99.90] 99.90]
BMI Percentile Category
< 5th 20 (1.42%) 23 (2.31%) 43 (1.79%)
5th to 85th 1,067 (75.9%) 716 (71.8%) 1,783 (74.2%)
85th to 95th 161 (11.5%) 109 (10.9%) 270 (11.2%)
> 95th 158 (11.2%) 149 (14.9%) 307 (12.8%)
Total Energy Expenditure (kcal/d)
N 5,018 2,914 7,932
1,992.37 2,453.55 2,161.79
Mean (SD)
(648.51) (1,001.50) (826.96)
2,067.66 2,605.24 2,180.67
Median [Min, Max] [145.48, [151.10, [145.48,
4,873.09] 8,190.40] 8,190.40]
Basal Metabolic Rate Observed
(kcal/d)
N 2,031 1,412 3,443
1,236.78 1,519.16 1,352.59
Mean (SD)
(323.07) (438.80) (399.73)
1,260.00 1,570.27 1,353.00
Median [Min, Max] [263.00, [286.00, [263.00,
2,576.00] 3,035.00] 3,035.00]
Physical Activity Level Observed
N 2,027 1,411 3,438
Mean (SD) 1.65 (0.31) 1.69 (0.34) 1.66 (0.32)
Median [Min, Max] 1.63 [0.92, 3.35] 1.65 [0.83, 4.03] 1.64 [0.83, 4.03]
Physical Activity Level Category
Observed
PAL < 1 9 (0.444%) 7 (0.496%) 16 (0.465%)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX H 265

Females Males Overall

(N = 5,025) (N = 2,919) (N = 7,944)
Inactive 540 (26.6%) 323 (22.9%) 863 (25.1%)
Low Active 515 (25.4%) 357 (25.3%) 872 (25.4%)
Active 442 (21.8%) 305 (21.6%) 747 (21.7%)
Very Active 498 (24.6%) 386 (27.4%) 884 (25.7%)
PAL > 2.5 23 (1.13%) 33 (2.34%) 56 (1.63%)
Basal Metabolic Rate Predicted
(kcal/d)
N 5,010 2,895 7,905
1,259.80 1,509.07 1,351.09
Mean (SD)
(341.86) (524.23) (434.87)
1,676.26
1,332.52 [99.67, 1,384.63 [99.67,
Median [Min, Max] [100.69,
2,829.45] 3,346.80]
3,346.80]
Missing 15 (0.3%) 24 (0.8%) 39 (0.5%)
Physical Activity Level Observed +
Predicted (TEE/BMR)
N 5,007 2,896 7,903
Mean (SD) 1.59 (0.27) 1.63 (0.31) 1.60 (0.29)
Median [Min, Max] 1.57 [0.65, 3.35] 1.60 [0.72, 4.35] 1.58 [0.65, 4.35]
Physical Activity Level Category
Observed + Predicted
PAL < 1 27 (0.540%) 20 (0.693%) 47 (0.596%)
Inactive 1,767 (35.3%) 815 (28.2%) 2,582 (32.7%)
Low Active 1,358 (27.2%) 744 (25.8%) 2,102 (26.7%)
Active 971 (19.4%) 594 (20.6%) 1,565 (19.8%)
Very Active 853 (17.1%) 681 (23.6%) 1,534 (19.4%)
PAL > 2.5 23 (0.460%) 34 (1.18%) 57 (0.723%)
NOTE: BMI = body mass index; BMR = basal metabolic rate; cm = centimeter; DLW = doubly labeled water;
kcal/d = kilocalorie/day; kg = kilogram; m = meter; max = maximum; min = minimum; PAL = physical activity
level; SD = standard deviation; TEE = total energy expenditure.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

266 DIETARY REFERENCE INTAKES FOR ENERGY

TABLE H-2 Characteristics of Infants, 0–11 Months, Included in Combined DLW Database
Females Males Overall
(N = 335) (N = 248) (N = 583)
Age (years)
N 335 248 583
Mean (SD) 0.39 (0.25) 0.38 (0.26) 0.39 (0.25)
0.27 [0.03,
Median [Min, Max] 0.26 [0.02, 0.99] 0.27 [0.02, 1.00]
1.00]
Life Stage
0–6 months 266 (79.4%) 203 (81.9%) 469 (80.4%)
7–11 months 69 (20.6%) 45 (18.1%) 114 (19.6%)
Sex
Females 335 (100%) 0 (0%) 335 (57.5%)
Males 0 (0%) 248 (100%) 248 (42.5%)
Ethnicity
African American 0 (0%) 1 (0.403%) 1 (0.172%)
Asian 2 (0.597%) 1 (0.403%) 3 (0.515%)
White 111 (33.1%) 72 (29.0%) 183 (31.4%)
Hispanic 16 (4.78%) 13 (5.24%) 29 (4.97%)
Other 9 (2.69%) 14 (5.65%) 23 (3.95%)
Unknown or data not available 197 (58.8%) 147 (59.3%) 344 (59.0%)
Length (cm)
N 334 248 582
Mean (SD) 63.08 (6.50) 63.44 (6.81) 63.24 (6.63)
62.38 [46.40, 63.00 [47.30, 62.70 [46.40,
Median [Min, Max]
80.00] 78.80] 80.00]
Weight (kg)
N 335 248 583
Mean (SD) 6.62 (1.77) 6.84 (1.87) 6.72 (1.81)
6.46 [2.36, 6.80 [2.70, 6.60 [2.36,
Median [Min, Max]
11.30] 12.60] 12.60]
Fat-Free Mass (kg)
N 217 176 393
Mean (SD) 4.52 (1.07) 4.84 (1.18) 4.67 (1.13)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX H 267

Females Males Overall

(N = 335) (N = 248) (N = 583)
4.38 [1.80,
Median [Min, Max] 4.77 [2.28, 8.74] 4.54 [1.80, 8.74]
7.94]
Fat Mass (kg)
N 216 176 392
Mean (SD) 1.67 (0.77) 1.64 (0.83) 1.66 (0.80)
1.58 [0.12,
Median [Min, Max] 1.54 [0.07, 3.86] 1.56 [0.07, 4.28]
4.28]
Fat Mass %
N 216 176 392
Mean (SD) 25.98 (6.76) 23.98 (7.49) 25.08 (7.15)
26.72 [3.69, 24.40 [2.06, 25.65 [2.06,
Median [Min, Max]
50.61] 41.65] 50.61]
Weight-for-Length Percentile
N 334 248 582
Mean (SD) 51.10 (26.97) 50.93 (29.02) 51.03 (27.84)
51.76 [0.73, 50.03 [0.09, 51.04 [0.09,
Median [Min, Max]
99.38] 99.90] 99.90]
Weight-for-Length Percentile
Category
< 3rd 5 (1.50%) 8 (3.23%) 13 (2.23%)
3rd to 97th 322 (96.4%) 233 (94.0%) 555 (95.4%)
> 97th 7 (2.10%) 7 (2.82%) 14 (2.41%)
Total Energy Expenditure (kcal/d)
N 335 248 583
Mean (SD) 481.03 (159.86) 510.77 (190.95) 493.68 (174.23)
466.06 [145.48, 493.61 [151.10, 475.81 [145.48,
Median [Min, Max]
961.00] 1,187.32] 1,187.32]
Basal Metabolic Rate Observed
(kcal/d)
N 109 66 175
Mean (SD) 443.51 (91.47) 472.56 (95.08) 454.47 (93.65)
435.00 [263.00, 472.00 [286.00, 447.00 [263.00,
Median [Min, Max]
649.00] 649.00] 649.00]
Physical Activity Level Observed

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

268 DIETARY REFERENCE INTAKES FOR ENERGY

Females Males Overall

(N = 335) (N = 248) (N = 583)
N 109 66 175
Mean (SD) 1.27 (0.20) 1.27 (0.22) 1.27 (0.21)
1.26 [0.95,
Median [Min, Max] 1.28 [0.83, 1.80] 1.27 [0.83, 1.88]
1.88]
Physical Activity Level Category
Observed
PAL < 1 4 (3.67%) 6 (9.09%) 10 (5.71%)
Inactive 66 (60.6%) 32 (48.5%) 98 (56.0%)
Low Active 21 (19.3%) 13 (19.7%) 34 (19.4%)
Active 11 (10.1%) 9 (13.6%) 20 (11.4%)
Very Active 7 (6.42%) 6 (9.09%) 13 (7.43%)
PAL > 2.5 0 (0%) 0 (0%) 0 (0%)
Basal Metabolic Rate (kcal/d)
Predicted
N 334 248 582
Mean (SD) 339.70 (94.17) 346.25 (103.69) 342.49 (98.31)
326.67 [99.67, 339.55 [100.69, 333.19 [99.67,
Median [Min, Max]
577.34] 579.94] 579.94]
Physical Activity Level Observed +
Predicted
N 334 248 582
Mean (SD) 1.37 (0.26) 1.42 (0.30) 1.39 (0.28)
1.35 [0.65,
Median [Min, Max] 1.38 [0.77, 2.40] 1.36 [0.65, 2.41]
2.41]
Physical Activity Level Category
Observed + Predicted
PAL < 1 14 (4.19%) 13 (5.24%) 27 (4.64%)
Inactive 135 (40.4%) 79 (31.9%) 214 (36.8%)
Low Active 74 (22.2%) 49 (19.8%) 123 (21.1%)
Active 55 (16.5%) 39 (15.7%) 94 (16.2%)
Very Active 56 (16.8%) 68 (27.4%) 124 (21.3%)
PAL > 2.5 0 (0%) 0 (0%) 0 (0%)
NOTE: cm = centimeter; DLW = doubly labeled water; kcal/d = kilocalorie/day; kg = kilogram; m = meter; max =
maximum; min = minimum; PAL = physical activity level; SD = standard deviation.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX H 269

TABLE H-3 Characteristics of Children, 1–8 Years, Included in Combined DLW Database
Females Males Overall
(N = 640) (N = 489) (N = 1,129)
Age (years)
N 640 489 1,129
Mean (SD) 5.23 (2.28) 4.78 (1.81) 5.03 (2.10)
Median [Min, Max] 5.00 [1.00, 8.90] 5.00 [1.00, 8.50] 5.00 [1.00, 8.90]
Life Stage
1–3 years 145 (22.7%) 105 (21.5%) 250 (22.1%)
4–8 years 495 (77.3%) 384 (78.5%) 879 (77.9%)
Sex
Females 640 (100%) 0 (0%) 640 (56.7%)
Males 0 (0%) 489 (100%) 489 (43.3%)
Ethnicity
African American 15 (2.34%) 32 (6.54%) 47 (4.16%)
Asian 4 (0.625%) 0 (0%) 4 (0.354%)
White 100 (15.6%) 88 (18.0%) 188 (16.7%)
Hispanic 11 (1.72%) 23 (4.70%) 34 (3.01%)
Other 1 (0.156%) 2 (0.409%) 3 (0.266%)
Unknown or data not
509 (79.5%) 344 (70.3%) 853 (75.6%)
available
Height (cm)
N 633 475 1,108
Mean (SD) 110.86 (17.15) 109.30 (15.09) 110.19 (16.31)
113.40 [71.00, 112.00 [72.00, 113.00 [71.00,
Median [Min, Max]
155.00] 153.00] 155.00]
Weight (kg)
N 640 489 1,129
Mean (SD) 21.47 (8.41) 20.90 (7.46) 21.22 (8.01)
Median [Min, Max] 20.50 [8.20, 66.00] 20.00 [8.40, 68.40] 20.27 [8.20, 68.40]
Fat-Free Mass (kg)
N 208 223 431
Mean (SD) 13.54 (3.92) 14.83 (4.27) 14.20 (4.15)
Median [Min, Max] 13.42 [5.40, 31.43] 14.65 [6.47, 29.17] 14.09 [5.40, 31.43]

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

270 DIETARY REFERENCE INTAKES FOR ENERGY

Females Males Overall

(N = 640) (N = 489) (N = 1,129)
Fat Mass (kg)
N 208 223 431
Mean (SD) 5.59 (3.47) 5.31 (3.37) 5.44 (3.42)
Median [Min, Max] 4.73 [1.01, 24.57] 4.32 [1.22, 20.33] 4.62 [1.01, 24.57]
Fat Mass %
N 208 223 431
Mean (SD) 27.76 (6.77) 25.12 (6.84) 26.40 (6.92)
26.47 [10.48, 23.95 [10.93, 25.24 [10.48,
Median [Min, Max]
58.34] 50.31] 58.34]
Body Mass Index Percentile
N 633 475 1,108
Mean (SD) 60.17 (27.37) 60.56 (29.03) 60.33 (28.08)
Median [Min, Max] 61.45 [2.79, 98.45] 62.14 [2.88, 97.40] 61.56 [2.79, 98.45]
BMI Percentile Category
< 5th 9 (1.42%) 8 (1.68%) 17 (1.53%)
5th to 85th 462 (73.0%) 329 (69.3%) 791 (71.4%)
85th to 95th 75 (11.8%) 53 (11.2%) 128 (11.6%)
> 95th 87 (13.7%) 85 (17.9%) 172 (15.5%)
Total Energy Expenditure
(kcal/d)
N 640 489 1,129
Mean (SD) 1,359.80 (377.65) 1,424.46 (372.36) 1,387.80 (376.57)
1,343.11 [517.00, 1,421.00 [560.14, 1,376.49 [517.00,
Median [Min, Max]
2,699.00] 2,975.00] 2,975.00]
Basal Metabolic Rate Observed
(kcal/d)
N 300 194 494
Mean (SD) 934.61 (203.47) 1,001.27 (208.88) 960.79 (207.97)
969.50 [456.00, 1,030.00 [530.00, 985.00 [456.00,
Median [Min, Max]
1,490.00] 1,470.00] 1,490.00]
Physical Activity Level
Observed
N 300 194 494

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX H 271

Females Males Overall

(N = 640) (N = 489) (N = 1,129)
Mean (SD) 1.45 (0.24) 1.40 (0.17) 1.43 (0.21)
Median [Min, Max] 1.43 [0.92, 2.15] 1.38 [0.99, 1.91] 1.41 [0.92, 2.15]
Physical Activity Level
Category Observed
PAL < 1 5 (1.67%) 1 (0.515%) 6 (1.21%)
Inactive 74 (24.7%) 54 (27.8%) 128 (25.9%)
Low Active 74 (24.7%) 71 (36.6%) 145 (29.4%)
Active 72 (24.0%) 43 (22.2%) 115 (23.3%)
Very Active 75 (25.0%) 25 (12.9%) 100 (20.2%)
PAL > 2.5 0 (0%) 0 (0%) 0 (0%)
Basal Metabolic Rate Predicted
(kcal/d)
N 633 475 1,108
Mean (SD) 898.15 (191.71) 945.12 (193.07) 918.28 (193.61)
905.05 [456.00, 947.55 [476.41, 928.96 [456.00,
Median [Min, Max]
1,733.87] 1,954.74] 1,954.74]
Physical Activity Level
Observed + Predicted
N 635 477 1,112
Mean (SD) 1.47 (0.22) 1.46 (0.20) 1.46 (0.21)
Median [Min, Max] 1.45 [0.92, 2.43] 1.44 [0.99, 2.60] 1.44 [0.92, 2.60]
Physical Activity Level
Category Observed + Predicted
PAL < 1 5 (0.787%) 2 (0.419%) 7 (0.629%)
Inactive 135 (21.3%) 106 (22.2%) 241 (21.7%)
Low Active 163 (25.7%) 133 (27.9%) 296 (26.6%)
Active 175 (27.6%) 125 (26.2%) 300 (27.0%)
Very Active 157 (24.7%) 110 (23.1%) 267 (24.0%)
PAL > 2.5 0 (0%) 1 (0.210%) 1 (0.0899%)
NOTE: cm = centimeter; DLW = doubly labeled water; kcal/d = kilocalorie/day; kg = kilogram; max = maximum;
min = minimum; PAL = physical activity level; SD = standard deviation.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

272 DIETARY REFERENCE INTAKES FOR ENERGY

TABLE H-4 Characteristics of Children and Teens, 9–18 Years, Included in Combined DLW
Database
Females Males Overall
(N = 443) (N = 278) (N = 721)
Age (years)
N 443 278 721
Mean (SD) 13.86 (3.05) 13.56 (3.12) 13.75 (3.08)
Median [Min, Max] 14.60 [9.00, 18.80] 14.00 [9.00, 18.90] 14.20 [9.00, 18.90]
Life Stage
9–13 years 184 (41.5%) 120 (43.2%) 304 (42.2%)
14–18 years 259 (58.5%) 158 (56.8%) 417 (57.8%)
Sex
Females 443 (100%) 0 (0%) 443 (61.4%)
Males 0 (0%) 278 (100%) 278 (38.6%)
Ethnicity
African American 26 (5.871%) 17 (6.12%) 43 (5.966%)
Asian 2 (0.451%) 1 (0.360%) 3 (0.416%)
White 149 (33.6%) 141 (50.7%) 290 (40.2%)
Hispanic 0 (0%) 1 (0.360%) 1 (0.139%)
Other 11 (2.48%) 4 (1.44%) 15 (2.08%)
Unknown or data not
255 (57.6%) 114 (41.0%) 369 (51.2%)
available
Height (cm)
N 439 274 713
Mean (SD) 156.63 (12.85) 163.78 (18.37) 159.38 (15.59)
159.20 [116.00, 168.90 [125.00, 161.00 [116.00,
Median [Min, Max]
183.70] 201.00] 201.00]
Weight (kg)
N 443 277 720
Mean (SD) 54.43 (16.50) 58.78 (17.66) 56.10 (17.07)
54.25 [22.20, 60.00 [23.50, 56.30 [22.20,
Median [Min, Max]
133.20] 125.70] 133.20]
Fat-Free Mass (kg)
N 238 201 439
Mean (SD) 39.78 (8.73) 47.81 (13.54) 43.46 (11.88)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX H 273

Females Males Overall

(N = 443) (N = 278) (N = 721)
39.70 [14.65, 49.62 [19.26, 42.39 [14.65,
Median [Min, Max]
95.74] 73.58] 95.74]
Fat Mass (kg)
N 238 201 439
Mean (SD) 20.57 (9.82) 14.89 (8.84) 17.97 (9.79)
Median [Min, Max] 17.90 [6.05, 66.05] 12.94 [3.08, 54.33] 15.97 [3.08, 66.05]
Fat Mass %
N 238 201 439
Mean (SD) 33.07 (7.45) 23.36 (9.53) 28.62 (9.75)
Median [Min, Max] 32.21 [8.03, 59.57] 21.10 [6.01, 53.92] 28.33 [6.01, 59.57]
Body Mass Index Percentile
N 439 274 713
Mean (SD) 63.75 (25.20) 63.17 (27.30) 63.53 (26.01)
Median [Min, Max] 66.28 [5.09, 97.60] 64.43 [3.95, 97.57] 65.28 [3.95, 97.60]
BMI Percentile Category
< 5th 0 (0%) 3 (1.09%) 3 (0.421%)
5th to 85th 322 (73.3%) 191 (69.7%) 513 (71.9%)
85th to 95th 59 (13.4%) 28 (10.2%) 87 (12.2%)
> 95th 58 (13.2%) 52 (19.0%) 110 (15.4%)
Total Energy Expenditure
(kcal/d)
N 441 277 718
Mean (SD) 2,300.37 (544.75) 2,891.46 (789.38) 2,528.41 (710.54)
2,274.19 [1,000.00, 2,753.99 [1,178.00, 2,413.88 [1,000.00,
Median [Min, Max]
4,873.09] 5,555.39] 5,555.39]
Basal Metabolic Rate
Observed (kcal/d)
N 262 107 369
Mean (SD) 1,362.30 (233.26) 1,679.31 (378.90) 1,454.22 (317.31)
1,327.50 [881.00, 1,610.00 [1,025.00, 1,403.00 [881.00,
Median [Min, Max]
2,218.00] 3,010.00] 3,010.00]
Physical Activity Level
Observed

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

274 DIETARY REFERENCE INTAKES FOR ENERGY

Females Males Overall

(N = 443) (N = 278) (N = 721)
N 262 107 369
Mean (SD) 1.75 (0.34) 1.81 (0.41) 1.77 (0.36)
Median [Min, Max] 1.71 [1.02, 2.82] 1.73 [1.03, 4.03] 1.72 [1.02, 4.03]
Physical Activity Level
Category Observed
PAL < 1 0 (0%) 0 (0%) 0 (0%)
Inactive 57 (21.8%) 11 (10.3%) 68 (18.4%)
Low Active 62 (23.7%) 25 (23.4%) 87 (23.6%)
Active 53 (20.2%) 32 (29.9%) 85 (23.0%)
Very Active 81 (30.9%) 33 (30.8%) 114 (30.9%)
PAL > 2.5 9 (3.44%) 6 (5.61%) 15 (4.07%)
Missing 181 (40.9%) 171 (61.5%) 352 (48.8%)
Basal Metabolic Rate (kcal/d)
Predicted
N 439 273 712
Mean (SD) 1,384.22 (185.29) 1,690.05 (302.00) 1,501.48 (279.62)
1,391.68 [942.38, 1,717.34 [1,043.51, 1,446.60 [942.38,
Median [Min, Max]
2,196.50] 2,815.00] 2,815.00]
Physical Activity Level
Observed + Predicted
N 439 275 714
Mean (SD) 1.67 (0.31) 1.71 (0.34) 1.69 (0.32)
Median [Min, Max] 1.64 [1.02, 2.82] 1.68 [1.02, 4.03] 1.65 [1.02, 4.03]
Physical Activity Level
Observed + Predicted
PAL < 1 0 (0%) 0 (0%) 0 (0%)
Inactive 141 (32.1%) 65 (23.8%) 206 (28.9%)
Low Active 106 (24.1%) 72 (26.4%) 178 (25.0%)
Active 90 (20.5%) 67 (24.5%) 157 (22.1%)
Very Active 93 (21.2%) 63 (23.1%) 156 (21.9%)
PAL > 2.5 9 (2.05%) 6 (2.20%) 15 (2.11%)
NOTE: cm = centimeter; DLW = doubly labeled water; kcal/d = kilocalorie/day; kg = kilogram; max = maximum;
min = minimum; PAL = physical activity level; SD = standard deviation.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX H 275

TABLE H-5 Characteristics of Adults, 19 Years and Older, Included in Combined DLW
Database
Females Males Overall
(N = 3,607) (N = 1,904) (N = 5,511)
Age (years)
N 3,607 1,904 5,511
Mean (SD) 53.79 (19.79) 50.19 (19.35) 52.54 (19.71)
56.00 [19.00, 47.90 [19.00, 52.00 [19.00,
Median [Min, Max]
98.00] 101.00] 101.00]
Life Stage
19–30 years 633 (17.5%) 368 (19.3%) 1,001 (18.2%)
31–50 years 987 (27.4%) 703 (36.9%) 1,690 (30.7%)
51–70 years 1,079 (29.9%) 448 (23.5%) 1,527 (27.7%)
>= 71 years 908 (25.2%) 385 (20.2%) 1,293 (23.5%)
Sex
Females 3,607 (100%) 0 (0%) 3,607 (65.5%)
Males 0 (0%) 1,904 (100%) 1,904 (34.5%)
Race/Ethnicity
African American 564 (15.655%) 185 (9.713%) 749 (13.627%)
Asian 95 (2.63%) 94 (4.94%) 189 (3.43%)
White 1,880 (52.1%) 981 (51.5%) 2,861 (51.9%)
Hispanic 323 (8.95%) 164 (8.61%) 487 (8.84%)
Other 23 (0.638%) 9 (0.473%) 32 (0.581%)
Unknown or data not available 722 (20.0%) 471 (24.7%) 1,193 (21.6%)
Height (cm)
N 3,607 1,904 5,511
Mean (SD) 162.35 (7.09) 175.95 (7.60) 167.05 (9.73)
162.50 [137.00, 176.00 [147.00, 166.00 [137.00,
Median [Min, Max]
196.00] 204.70] 204.70]
Weight (kg)
N 3,604 1,899 5,503
Mean (SD) 71.86 (16.12) 83.06 (16.44) 75.73 (17.08)
69.00 [37.80, 80.10 [37.20, 73.68 [37.20,
Median [Min, Max]
164.55] 215.70] 215.70]
Fat-Free Mass (kg)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

276 DIETARY REFERENCE INTAKES FOR ENERGY

Females Males Overall

(N = 3,607) (N = 1,904) (N = 5,511)
N 3,170 1,543 4,713
Mean (SD) 42.72 (6.62) 58.19 (8.92) 47.79 (10.40)
41.99 [21.11, 57.66 [28.87, 45.52 [21.11,
Median [Min, Max]
81.31] 97.81] 97.81]

Fat Mass (kg)

N 3,167 1,539 4,706
Mean (SD) 29.52 (11.57) 24.79 (10.55) 27.97 (11.46)
27.59 [3.61, 23.62 [2.23, 26.22 [2.23,
Median [Min, Max]
92.14] 90.27] 92.14]
Fat Mass %
N 3,167 1,539 4,706
Mean (SD) 39.67 (7.94) 29.03 (8.04) 36.19 (9.41)
40.39 [6.55, 29.48 [2.91, 36.77 [2.91,
Median [Min, Max]
70.08] 56.19] 70.08]
Body Mass Index
N 3,604 1,900 5,504
Mean (SD) 27.24 (5.88) 26.82 (4.85) 27.10 (5.55)
26.08 [14.69, 25.99 [14.33, 26.03 [14.33,
Median [Min, Max]
57.87] 61.70] 61.70]
BMI Category
< 18.5 kg/m2 48 (1.33%) 20 (1.05%) 68 (1.24%)
>= 18.5 and < 25 kg/m2 1,451 (40.3%) 716 (37.7%) 2,167 (39.4%)
2
>= 25 and < 30 kg/m 1,143 (31.7%) 770 (40.5%) 1,913 (34.8%)
>= 30 and < 35 kg/m2 588 (16.3%) 275 (14.5%) 863 (15.7%)
>= 35 and < 40 kg/m2 245 (6.80%) 84 (4.42%) 329 (5.98%)
>= 40 kg/m2 129 (3.58%) 35 (1.84%) 164 (2.98%)
Total Energy Expenditure (kcal/d)
N 3,602 1,900 5,502
Mean (SD) 2,207.62 (424.76) 2,908.15 (638.17) 2,449.53 (608.00)
2,850.97
2,169.83 [833.81, 2,343.21 [833.81,
Median [Min, Max] [1,181.61,
4,450.48] 8,190.40]
8,190.40]

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX H 277

Females Males Overall

(N = 3,607) (N = 1,904) (N = 5,511)
Basal Metabolic Rate Observed
(kcal/d)
N 1,360 1,045 2,405
Mean (SD) 1,342.84 (225.32) 1,665.00 (306.63) 1,482.82 (308.30)
1,330.00 [773.00, 1,649.14 [884.08, 1,446.00 [773.00,
Median [Min, Max]
2,576.00] 3,035.00] 3,035.00]
Physical Activity Level Observed
(TEE/BMR)
N 1,356 1,044 2,400
Mean (SD) 1.70 (0.28) 1.76 (0.31) 1.73 (0.30)
Median [Min, Max] 1.68 [1.00, 3.35] 1.72 [1.02, 3.22] 1.70 [1.00, 3.35]
Physical Activity Level Category
Observed
PAL < 1 0 (0%) 0 (0%) 0 (0%)
Inactive 343 (25.3%) 226 (21.6%) 569 (23.7%)
Low Active 358 (26.4%) 248 (23.8%) 606 (25.3%)
Active 306 (22.6%) 221 (21.2%) 527 (22.0%)
Very Active 335 (24.7%) 322 (30.8%) 657 (27.4%)
PAL > 2.5 14 (1.03%) 27 (2.59%) 41 (1.71%)
Basal Metabolic Rate Predicted
(kcal/d)
N 3,604 1,899 5,503
Mean (SD) 1,393.44 (172.71) 1,775.97 (250.99) 1,525.45 (272.66)
1,374.35 1,769.52 1,455.59
Median [Min, Max] [1,000.33, [1,069.83, [1,000.33,
2,829.45] 3,346.80] 3,346.80]
Physical Activity Level Observed
+ Predicted
N 3,599 1,896 5,495
Mean (SD) 1.62 (0.26) 1.69 (0.31) 1.65 (0.28)
Median [Min, Max] 1.60 [0.70, 3.35] 1.65 [0.72, 4.35] 1.61 [0.70, 4.35]
Missing 8 (0.2%) 8 (0.4%) 16 (0.3%)
Physical Activity Level Category
Observed + Predicted

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

278 DIETARY REFERENCE INTAKES FOR ENERGY

Females Males Overall

(N = 3,607) (N = 1,904) (N = 5,511)
PAL < 1 8 (0.223%) 5 (0.265%) 13 (0.237%)
Inactive 1,356 (37.8%) 565 (29.9%) 1,921 (35.0%)
Low Active 1,015 (28.3%) 490 (25.9%) 1,505 (27.5%)
Active 651 (18.1%) 363 (19.2%) 1,014 (18.5%)
Very Active 547 (15.2%) 440 (23.3%) 987 (18.0%)
PAL > 2.5 14 (0.390%) 27 (1.43%) 41 (0.748%)
NOTE: BMI = body mass index; BMR = basal metabolic rate; cm = centimeter; DLW = doubly labeled water;
kcal/d = kilocalorie/day; kg = kilogram; m = meter; max = maximum; min = minimum; PAL = physical activity
level; SD = standard deviation; TEE = total energy expenditure.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX H 279

TABLE H-6 Characteristics of Pregnant Women Included in Combined DLW Database

1st Trimester 2nd Trimester 3rd Trimester NPNL Overall
(N = 31) (N = 201) (N = 206) (N = 133) (N = 571)
Sex
Females 31 (100%) 201 (100%) 206 (100%) 133 (100%) 571 (100%)
Pregnant
Yes 31 (100%) 201 (100%) 206 (100%) 0 (0%) 438 (76.7%)
No 0 (0%) 0 (0%) 0 (0%) 133 (100%) 133 (23.3%)
Lactating
No 31 (100%) 201 (100%) 206 (100%) 133 (100%) 571 (100%)
Pregnancy Stage
1st Trimester 31 (100%) 0 (0%) 0 (0%) 0 (0%) 31 (5.43%)
2nd Trimester 0 (0%) 201 (100%) 0 (0%) 0 (0%) 201 (35.2%)
3rd Trimester 0 (0%) 0 (0%) 206 (100%) 0 (0%) 206 (36.1%)
NPNL 0 (0%) 0 (0%) 0 (0%) 133 (100%) 133 (23.3%)
Weeks
N 31 201 206 133 571
Mean (SD) 9.29 (2.61) 19.14 (4.30) 34.83 (2.29) 0.00 (0.00) 19.81 (13.71)
9.00 [6.00, 18.00 [14.50, 36.00 [30.00, 20.86 [0.00,
Median [Min, Max] 0.00 [0.00, 0.00]
12.00] 27.00] 38.14] 38.14]
Age
N 31 201 206 133 571
Mean (SD) 28.83 (3.22) 29.06 (4.43) 29.28 (4.23) 30.24 (4.09) 29.40 (4.24)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

280 DIETARY REFERENCE INTAKES FOR ENERGY

1st Trimester 2nd Trimester 3rd Trimester NPNL Overall

(N = 31) (N = 201) (N = 206) (N = 133) (N = 571)
29.00 [21.00, 29.00 [18.00, 29.00 [18.00, 30.00 [21.00, 29.00 [18.00,
Median [Min, Max]
33.00] 41.00] 41.00] 41.00] 41.00]
Life Stage
19–30 years 19 (61.3%) 120 (59.7%) 124 (60.2%) 71 (53.4%) 334 (58.5%)
31–50 years 12 (38.7%) 79 (39.3%) 81 (39.3%) 62 (46.6%) 234 (41.0%)
14–18 years 0 (0%) 2 (0.995%) 1 (0.485%) 0 (0%) 3 (0.525%)
Height (cm)
N 30 200 205 133 568
5,991.15 4,752.45 6,988.55 5,469.86
Mean (SD) 164.13 (7.38)
(7,797.23) (7,328.27) (8,039.83) (7,617.11)
164.00 [155.00, 170.00 [142.10, 168.00 [142.10, 172.00 [149.81, 168.85 [142.10,
Median [Min, Max]
180.00] 17,820.00] 17,830.00] 17,680.00] 17,830.00]
Weight (kg)
N 30 200 205 133 568
Mean (SD) 63.23 (8.61) 78.31 (21.19) 82.75 (20.48) 60.95 (10.66) 75.05 (20.44)
62.21 [50.57, 74.50 [48.60, 76.20 [55.40, 59.30 [40.10, 69.90 [40.10,
Median [Min, Max]
77.70] 164.60] 170.70] 102.20] 170.70]
Fat-Free Mass (kg)
N 0 70 60 0 130
Mean (SD) NA (NA) 54.34 (8.11) 62.25 (9.69) NA (NA) 57.99 (9.68)
52.98 [40.74, 61.61 [45.09, 57.53 [40.74,
Median [Min, Max] NA [NA, NA] NA [NA, NA]
77.54] 89.83] 89.83]
Fat Mass (kg)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX H 281

1st Trimester 2nd Trimester 3rd Trimester NPNL Overall

(N = 31) (N = 201) (N = 206) (N = 133) (N = 571)
N 0 70 60 0 130
Mean (SD) NA (NA) 45.60 (12.07) 44.54 (10.65) NA (NA) 45.11 (11.41)
42.33 [29.22, 42.93 [28.20, 42.68 [28.20,
Median [Min, Max] NA [NA, NA] NA [NA, NA]
87.27] 82.56] 87.27]
Fat Mass %
N 0 70 60 0 130
Mean (SD) NA (NA) 45.17 (4.62) 41.45 (5.04) NA (NA) 43.45 (5.15)
45.09 [36.14, 40.92 [32.85, 43.18 [32.85,
Median [Min, Max] NA [NA, NA] NA [NA, NA]
58.25] 54.82] 58.25]
Body Mass Index
N 30 200 205 133 568
Mean (SD) 23.48 (2.98) 29.05 (7.39) 30.65 (7.31) 22.52 (3.59) 27.81 (7.27)
22.96 [18.80, 27.11 [18.75, 28.12 [20.79, 21.53 [17.43, 25.63 [17.43,
Median [Min, Max]
29.97] 56.31] 57.63] 35.61] 57.63]
BMI Category
< 18.5 kg/m2 0 (0%) 0 (0%) 0 (0%) 11 (8.27%) 11 (1.94%)
>= 18.5 and < 25 kg/m2 22 (73.3%) 79 (39.5%) 59 (28.8%) 93 (69.9%) 253 (44.5%)
>= 25 and < 30 kg/m2 8 (26.7%) 37 (18.5%) 55 (26.8%) 24 (18.0%) 124 (21.8%)
2
>= 30 and < 35 kg/m 0 (0%) 43 (21.5%) 36 (17.6%) 3 (2.26%) 82 (14.4%)
>= 35 and < 40 kg/m2 0 (0%) 24 (12.0%) 29 (14.1%) 2 (1.50%) 55 (9.68%)
>= 40 kg/m2 0 (0%) 17 (8.50%) 26 (12.7%) 0 (0%) 43 (7.57%)
Total Energy Expenditure
(kcal/d)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

282 DIETARY REFERENCE INTAKES FOR ENERGY

1st Trimester 2nd Trimester 3rd Trimester NPNL Overall

(N = 31) (N = 201) (N = 206) (N = 133) (N = 571)
N 31 201 206 133 571
2,289.11
Mean (SD) 2,553.66 (559.56) 2,779.89 (558.64) 2,478.24 (444.24) 2,603.35 (542.28)
(311.55)
2,323.00 2,544.22 2,736.52 2,431.17 2,574.00
Median [Min, Max] [1,628.11, [1,149.38, [1,526.29, [1,277.96, [1,149.38,
2,904.88] 7,543.25] 4,880.26] 3,937.13] 7,543.25]
Basal Metabolic Rate Observed
N 31 131 141 130 433
1,454.83
Mean (SD) 1,464.87 (197.34) 1,710.76 (213.91) 1,363.62 (163.61) 1,513.82 (238.85)
(173.23)
1,447.00 1,698.14
1,435.55 [848.47, 1,362.21 [859.50, 1,483.99 [848.47,
Median [Min, Max] [1,046.00, [1,040.39,
1,975.71] 1,843.11] 2,392.59]
1,833.41] 2,392.59]

Physical Activity Level

Observed (TEE/BMR)
N 30 56 85 73 244
Mean (SD) 1.60 (0.17) 1.67 (0.40) 1.66 (0.35) 1.79 (0.34) 1.69 (0.35)
Median [Min, Max] 1.64 [1.19, 1.92] 1.61 [1.05, 4.00] 1.60 [1.12, 2.91] 1.75 [1.17, 2.58] 1.64 [1.05, 4.00]
Physical Activity Level Category
Observed
PAL < 1 0 (0%) 0 (0%) 0 (0%) 0 (0%) 0 (0%)
Inactive 10 (33.3%) 15 (26.8%) 35 (41.2%) 20 (27.4%) 80 (32.8%)
Low Active 11 (36.7%) 24 (42.9%) 17 (20.0%) 14 (19.2%) 66 (27.0%)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX H 283

1st Trimester 2nd Trimester 3rd Trimester NPNL Overall

(N = 31) (N = 201) (N = 206) (N = 133) (N = 571)
Active 7 (23.3%) 6 (10.7%) 13 (15.3%) 14 (19.2%) 40 (16.4%)
Very Active 2 (6.67%) 9 (16.1%) 17 (20.0%) 23 (31.5%) 51 (20.9%)
PAL > 2.5 0 (0%) 2 (3.57%) 3 (3.53%) 2 (2.74%) 7 (2.87%)
Basal Metabolic Rate Predicted
(kcal/d)
N 30 200 205 133 568
1,396.41 8,693.97 6,785.12 9,304.05 7,762.45
Mean (SD)
(108.07) (16,536.12) (14,459.64) (17,303.25) (15,628.54)
1,398.26 1,656.29 1,651.91 1,473.57 1,588.13
Median [Min, Max] [1,255.12, [1,204.61, [1,300.54, [1,141.94, [1,141.94,
1,642.45] 51,563.86] 51,717.50] 51,034.18] 51,717.50]
Physical Activity Level
Observed + Predicted
N 31 201 206 133 571
Mean (SD) 1.58 (0.20) 1.32 (0.62) 1.42 (0.56) 1.44 (0.70) 1.40 (0.61)
Median [Min, Max] 1.64 [0.99, 1.92] 1.49 [0.04, 4.00] 1.53 [0.04, 2.91] 1.60 [0.04, 2.58] 1.53 [0.04, 4.00]
Physical Activity Level Category
Observed + Predicted
PAL < 1 1 (3.23%) 34 (17.0%) 26 (12.6%) 24 (18.0%) 85 (14.9%)
Inactive 10 (32.3%) 81 (40.5%) 77 (37.4%) 30 (22.6%) 198 (34.7%)
Low Active 11 (35.5%) 44 (22.0%) 51 (24.8%) 27 (20.3%) 133 (23.3%)
Active 7 (22.6%) 23 (11.5%) 23 (11.2%) 23 (17.3%) 76 (13.3%)
Very Active 2 (6.45%) 16 (8.00%) 26 (12.6%) 27 (20.3%) 71 (12.5%)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

284 DIETARY REFERENCE INTAKES FOR ENERGY

1st Trimester 2nd Trimester 3rd Trimester NPNL Overall

(N = 31) (N = 201) (N = 206) (N = 133) (N = 571)
PAL > 2.5 0 (0%) 2 (1.00%) 3 (1.46%) 2 (1.50%) 7 (1.23%)
NOTE: BMI = body mass index; BMR = basal metabolic rate; cm = centimeter; DLW = doubly labeled water; kcal/d = kilocalorie/day; kg = kilogram; m =
meter; max = maximum; min = minimum; PAL = physical activity level; SD = standard deviation; TEE = total energy expenditure.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX H 285

TABLE H-7 Characteristics of Lactating Women Included in Combined DLW Database

Lactating 1-3 mo Lactating 4-6 mo Overall
(N = 93) (N = 114) (N = 207)
Sex
Females 93 (100%) 114 (100%) 207 (100%)
Pregnant
No 93 (100%) 114 (100%) 207 (100%)
Lactating
Yes 93 (100%) 114 (100%) 207 (100%)
Lactation Stage
Lactating 1–3 mo 93 (100%) 0 (0%) 93 (44.9%)
Lactating 4–6 mo 0 (0%) 114 (100%) 114 (55.1%)
Weeks
N 93 114 207
Mean (SD) 9.24 (2.98) 24.85 (2.18) 17.83 (8.20)
25.00 [19.00, 20.60 [4.30,
Median [Min, Max] 8.70 [4.30, 13.00]
27.00] 27.00]
Age
N 93 114 207
Mean (SD) 30.39 (3.90) 30.23 (4.12) 30.30 (4.01)
30.00 [21.00, 30.00 [21.00, 30.00 [21.00,
Median [Min, Max]
41.00] 41.00] 41.00]
Life Stage
19–30 years 54 (58.1%) 59 (51.8%) 113 (54.6%)
31–50 years 39 (41.9%) 55 (48.2%) 94 (45.4%)
Height (cm)
N 92 114 206
5,266.09 2,987.79
Mean (SD) 164.69 (6.07)
(7,547.46) (6,153.29)
165.00 [149.81, 168.85 [142.10, 167.00 [142.10,
Median [Min, Max]
180.00] 17,570.00] 17,570.00]
Weight (kg)
N 92 114 206
Mean (SD) 62.18 (9.86) 76.40 (22.70) 70.05 (19.43)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

286 DIETARY REFERENCE INTAKES FOR ENERGY

Lactating 1-3 mo Lactating 4-6 mo Overall

(N = 93) (N = 114) (N = 207)
60.10 [45.50, 69.86 [42.50, 63.50 [42.50,
Median [Min, Max]
87.90] 139.20] 139.20]
Fat-Free Mass (kg)
N 5 39 44
Mean (SD) 41.19 (3.34) 52.93 (7.57) 51.60 (8.12)
41.18 [36.36, 51.86 [34.67, 51.07 [34.67,
Median [Min, Max]
44.80] 73.00] 73.00]
Fat Mass (kg)
N 5 39 44
Mean (SD) 22.02 (13.64) 48.37 (12.43) 45.38 (15.02)
14.57 [14.10, 47.96 [25.21, 46.65 [14.10,
Median [Min, Max]
45.77] 77.43] 77.43]
Fat Mass %
N 5 39 44
Mean (SD) 32.90 (11.96) 47.22 (6.17) 45.59 (8.25)
28.00 [23.94, 46.61 [33.10, 45.74 [23.94,
Median [Min, Max]
53.44] 62.23] 62.23]
Body Mass Index
N 92 114 206
Mean (SD) 22.89 (3.18) 28.15 (8.15) 25.80 (6.93)
22.42 [18.21, 25.85 [17.87, 23.50 [17.87,
Median [Min, Max]
31.57] 49.25] 49.25]
BMI Category
< 18.5 kg/m2 3 (3.26%) 4 (3.51%) 7 (3.40%)
>= 18.5 and < 25 kg/m2 70 (76.1%) 52 (45.6%) 122 (59.2%)
>= 25 and < 30 kg/m2 13 (14.1%) 15 (13.2%) 28 (13.6%)
>= 30 and < 35 kg/m2 6 (6.52%) 15 (13.2%) 21 (10.2%)
>= 35 and < 40 kg/m2 0 (0%) 19 (16.7%) 19 (9.22%)
>= 40 kg/m2 0 (0%) 9 (7.89%) 9 (4.37%)
Total Energy Expenditure (kcal/d)
N 93 114 207
Mean (SD) 2,319.47 (402.15) 2,516.04 (490.24) 2,427.72 (462.24)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX H 287

Lactating 1-3 mo Lactating 4-6 mo Overall

(N = 93) (N = 114) (N = 207)
2,333.39 2,448.56 2,380.00
Median [Min, Max] [1,609.00, [1,519.36, [1,519.36,
3,380.26] 3,772.94] 3,772.94]
Basal Metabolic Rate Observed
(kcal/d)
N 88 74 162
Mean (SD) 1,367.42 (131.50) 1,380.24 (161.78) 1,373.27 (145.78)
1,353.23 1,349.64 1,351.37
Median [Min, Max] [1,108.00, [1,053.71, [1,053.71,
1,873.09] 1,865.21] 1,873.09]

Physical Activity Level Observed

(TEE/BMR)
N 88 39 127
Mean (SD) 1.70 (0.30) 1.76 (0.33) 1.72 (0.31)
Median [Min, Max] 1.69 [1.02, 2.81] 1.75 [1.08, 2.75] 1.71 [1.02, 2.81]
Physical Activity Level Category
Observed
PAL < 1 0 (0%) 0 (0%) 0 (0%)
Inactive 26 (29.5%) 9 (23.1%) 35 (27.6%)
Low Active 17 (19.3%) 6 (15.4%) 23 (18.1%)
Active 26 (29.5%) 11 (28.2%) 37 (29.1%)
Very Active 17 (19.3%) 11 (28.2%) 28 (22.0%)
PAL > 2.5 2 (2.27%) 2 (5.13%) 4 (3.15%)
Basal Metabolic Rate Predicted
(kcal/d)
N 92 114 206
6,315.37 4,109.79
Mean (SD) 1,376.79 (108.79)
(13,812.46) (10,546.39)
1,355.57 1,596.89 1,459.41
Median [Min, Max] [1,162.38, [1,154.20, [1,154.20,
1,718.94] 48,810.25] 48,810.25]
Physical Activity Level Observed
+ Predicted
N 93 114 207

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

288 DIETARY REFERENCE INTAKES FOR ENERGY

Lactating 1-3 mo Lactating 4-6 mo Overall

(N = 93) (N = 114) (N = 207)
Mean (SD) 1.70 (0.30) 1.43 (0.56) 1.55 (0.48)
Median [Min, Max] 1.69 [1.02, 2.81] 1.51 [0.04, 2.75] 1.62 [0.04, 2.81]
Physical Activity Level Category
Observed + Predicted
PAL < 1 0 (0%) 13 (11.4%) 13 (6.28%)
Inactive 27 (29.0%) 46 (40.4%) 73 (35.3%)
Low Active 18 (19.4%) 19 (16.7%) 37 (17.9%)
Active 27 (29.0%) 16 (14.0%) 43 (20.8%)
Very Active 19 (20.4%) 18 (15.8%) 37 (17.9%)
PAL > 2.5 2 (2.15%) 2 (1.75%) 4 (1.93%)
NOTE: BMI = body mass index; BMR = basal metabolic rate; cm = centimeter; DLW = doubly labeled water;
kcal/d = kilocalorie/day; kg = kilogram; m = meter; max = maximum; min = minimum; mo = months; PAL =
physical activity level; SD = standard deviation; TEE = total energy expenditure.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

Appendix I
Characteristics of the DLW Validation Studies

TABLE I-1 Characteristics of 144 Cohorts from 65 Published Studies Included in the Model
Validation
Boy Girl Man Woman Overall
(N = 21) (N = 20) (N = 32) (N = 71) (N = 144)
Sex
Female 0 (0%) 20 (100%) 0 (0%) 71 (100%) 91 (63.2%)
Male 21 (100%) 0 (0%) 32 (100%) 0 (0%) 53 (36.8%)
Age
Mean (SD) 9.97 (4.68) 9.53 (4.14) 39.7 (16.6) 34.7 (11.9) 28.7 (16.8)
Median 8.50 [3.00, 8.10 [3.00, 35.5 [19.0, 33.7 [19.4, 30.0 [3.00,
[Min, Max] 18.6] 18.0] 94.0] 94.0] 94.0]
Weight
Mean (SD) 40.2 (22.4) 37.3 (19.3) 82.5 (11.9) 71.7 (19.0) 64.7 (24.8)
Median 30.0 [16.4, 29.3 [15.9, 82.7 [65.0, 67.3 [38.5, 67.2 [15.9,
[Min, Max] 88.4] 85.8] 107] 162] 162]
Height
Mean (SD) 140 (24.9) 135 (21.1) 176 (4.38) 162 (16.9) 158 (22.2)
Median 133 [100, 129 [103, 177 [167, 165 [24.0, 165 [24.0,
[Min, Max] 181] 175] 182] 170] 182]
BMI
Mean (SD) 18.7 (3.91) 18.9 (4.36) 26.6 (3.38) 27.0 (6.77) 24.6 (6.54)
Median 16.7 [14.4, 17.2 [14.9, 26.0 [21.9, 24.3 [20.9, 24.0 [14.4,
[Min, Max] 27.4] 33.0] 34.9] 60.3] 60.3]
TEE
Mean (SD) 2,410 (884) 1,970 (525) 3,040 (532) 2,330 (399) 2,450 (636)
Median 1,990 [1,290, 1,840 [1,150, 3,030 [1,940, 2,230 [1,510, 2,420 [1,150,
[Min, Max] 4,370] 3,050] 4,370] 3,850] 4,370]
PAL
Mean (SD) 1.84 (0.156) 1.81 (0.221) 1.66 (0.130) 1.67 (0.161) 1.71 (0.172)
Median 1.77 [1.69, 1.72 [1.62, 1.70 [1.36, 1.69 [1.19, 1.71 [1.19,
[Min, Max] 2.07] 2.21] 1.90] 1.98] 2.21]

PREPUBLICATION COPY: UNCORRECTED PROOFS

289

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

290 DIETARY REFERENCE INTAKES FOR ENERGY

Boy Girl Man Woman Overall

(N = 21) (N = 20) (N = 32) (N = 71) (N = 144)
Missing 13 (61.9%) 13 (65.0%) 18 (56.3%) 46 (64.8%) 90 (62.5%)
PALCAT
Active 15 (71.4%) 6 (30.0%) 14 (43.8%) 17 (23.9%) 52 (36.1%)
Very Active 6 (28.6%) 0 (0%) 4 (12.5%) 6 (8.45%) 16 (11.1%)
Low Active 0 (0%) 14 (70.0%) 9 (28.1%) 25 (35.2%) 48 (33.3%)
Inactive 0 (0%) 0 (0%) 5 (15.6%) 23 (32.4%) 28 (19.4%)
NOTE: BMI = body mass index; DLW = doubly labeled water; max = maximum; min = minimum; mo = months;
PAL = physical activity level; PALCAT = PAL category; SD = standard deviation; TEE = total energy expenditure

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

Appendix J

Summary of Data Extracted from Systematic Reviews and other

Reviewed Literature

PREPUBLICATION COPY: UNCORRECTED PROOFS

291

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

292 DIETARY REFERENCE INTAKES FOR ENERGY

TABLE J-1 Evidence on the Relationship Between Different Measurements of Physical Activity and Energy Expenditure: Systematic
Reviews
Author, Year Number Sample Predictor or Primary Quantitative or Qualitative Risk of Hetero- Overall
of Character- Intervention or Outcome Finding(s) Bias geneity of AMSTAR2
Studies istics Comparator Studies Rating
Adamo et al., 5 47 males Indirect Mean Data from studies/substudies Overall, - Partially well
2009 and measures of difference reporting on combined male 19/24 done/reported
females physical from DLW and female data that studies
1–18 y; activity in boys and compared an indirect measure unclearly
white included girls to DLW indicated that reported or
European, activity diaries combined indirect measures failed to
U.S. or logs, overestimated physical report
African questionnaires, activity or energy expenditure between
American, surveys, and with a mean percent one and
U.S. recall difference of 22% and a range five of the
White interviews of -25% to 78%. 16
components
Adamo et al., 13 110 males Indirect Mean Results for male-only had Overall, - Partially well
2009 1–18 y; measures of difference mean percent differences of 0 19/24 done/reported
white physical from DLW (range: -33% to 56%). studies
European, activity in boys unclearly
U.S. included reported or
African activity diaries failed to
American, or logs, report
U.S. questionnaires, between
White surveys, and one and
recall five of the
interviews 16
components
Adamo et al., 13 93 Indirect Mean Results for female-only had Overall, - Partially well
2009 females measures of difference mean percent differences of - 19/24 done/reported
1–18 y; physical from DLW 1.2 (range: -43% to 95%). studies
white activity in girls unclearly
European, reported or

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 293

Author, Year Number Sample Predictor or Primary Quantitative or Qualitative Risk of Hetero- Overall
of Character- Intervention or Outcome Finding(s) Bias geneity of AMSTAR2
Studies istics Comparator Studies Rating
U.S. included failed to
African activity diaries report
American, or logs, between
U.S. questionnaires, one and
White surveys, and five of the
recall 16
interviews components
Dowd et al., 27 Males and Self-reported Criterion Mean percent differences for Mean - Well
2018 females ≥ measures of PA validity of PA diaries ranged from - AMSTAR done/reported
19y; high- included 7-day EE 12.9% to 20.8%, self-reported score was
income recall estimates PA energy expenditure 5.4 (out of
countries questionnaires, compared recalled from the previous 7 11)
past year recall to 8–15 days (or typical week)
questionnaires, days of ranging from -59.5% to
typical week DLW 62.1%, self-reported PA
questionnaires, measureme energy expenditure for the
and PA nt previous month ranged from -
logs/diaries 13.3% to 11.4%, self-reported
PA from the previous 12
months ranged from -77.6%
to 112.5%.
Dowd et al., 24 Males and Activity DLW The range of MPD observed Mean - Well
2018 females ≥ monitor in studies that examined the AMSTAR done/reported
19y; high- determined criterion validity of activity score was
income energy monitor-determined energy 5.4 (out of
countries expenditure expenditure ranged from - 11)
56.59% to 96.84%. However,
a trend was apparent for
activity monitor-determined
energy expenditure to

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

294 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Year Number Sample Predictor or Primary Quantitative or Qualitative Risk of Hetero- Overall
of Character- Intervention or Outcome Finding(s) Bias geneity of AMSTAR2
Studies istics Comparator Studies Rating
underestimate the criterion
measure.
Dowd et al., 9 Males and Activity Indirect For LIPA, the MPD ranged Mean - Well
2018 females ≥ monitor calorimetry from -79.8% to 429.1%. For AMSTAR done/reported
19 y; determined PA and whole MPA, MPD ranged from - score was
high- intensity room 50.4% to 454.1%, while 5.4 (out of
income calorimetry estimates for VPA ranged 11)
countries PA from -100% to 163.6%.
intensity Energy expenditure estimates
from activity monitoring
devices for total PA were
compared against indirect
calorimetry estimates, where
MPD ranged from -41.4% to
115.7%. The MPD range for
activity monitor-determined
total energy expenditure
compared with whole room
calorimetry were narrower (-
16.7% to -15.7%).
Dowd et al., 31 Males and Activity Indirect Estimated energy expenditure Mean - Well
2018 females monitor calorimetry was compared between AMSTAR done/reported
≥19 y; determined EE activity monitors and indirect score was
high- energy calorimetry (kcal over 5.4 (out of
income expenditure specified durations; [-68.5% 11)
countries to 81.1%]); (METs over
specified durations; [-67.3%
to 48.4%]). A single study
compared the estimated
energy expenditure from 5
different activity monitors

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 295

Author, Year Number Sample Predictor or Primary Quantitative or Qualitative Risk of Hetero- Overall
of Character- Intervention or Outcome Finding(s) Bias geneity of AMSTAR2
Studies istics Comparator Studies Rating
and indirect calorimetry at
incremental speeds (54, 80,
107, 134, 161, 188, and 214
m.min-1) in both men and
women (MPD ranged from -
60.4% to 90.8%).
Dowd et al., 3 Males and Pedometer DLW In free-living studies that Mean - Well
2018 females ≥ determined EE examined the criterion AMSTAR done/reported
19y; high- validity of pedometer score was
income determined energy 5.4 (out of
countries expenditure, pedometers were 11)
worn for 2 to 8 days (-62.3%
to 0.8%).
Helmerhorst 2 111 males Physical DLW For PAEE, Spearman r - - Not well
et al., 2012 and activity ranged from 0.09 to 0.45 and done/reported
females < questionnaires MD was 0.46 to 0.76 kg/kg/d.
18 y; For TEE, Spearman r ranged
high- from 0.49 to 0.65; MD 2,800
income kJ/day.
countries
Helmerhorst 6 239 males Physical DLW For PAEE, Spearman r = 0.39 - - Not well
et al., 2012 and activity and MD was −12.9 kJ/day done/reported
females questionnaires from one study. Four studies
18–65 y; reported TEE and Spearman r
high- ranging from 0.15 to 0.67;
income Pearson r ranged from 0.12 to
countries 0.58; MD ranged from -
3,451.9 to 7,455 kJ/day. One
study reported PAL, and the
Pearson r ranged from 0.34 to
0.69.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

296 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Year Number Sample Predictor or Primary Quantitative or Qualitative Risk of Hetero- Overall
of Character- Intervention or Outcome Finding(s) Bias geneity of AMSTAR2
Studies istics Comparator Studies Rating
Helmerhorst 2 86 males Physical DLW For TEE, Spearman r ranged - - Not well
et al., 2012 and activity from 0.10 to 0.64; Pearson r done/reported
females > questionnaires ranged from 0.11 to 0.65; MD
65 y; ranged from 435 to 3,146
high- (men) and 37 to 2,037
income (women) kJ/day.
countries
Jeran et al., 24 1,148 Assess whether Crude R2 Crude R2 ranged from 0.043 - - Not well
2016 males and study or acceleromet to 0.80 with a median of 0.26. done/reported
females ≥ accelerometer er output Crude R2 did not significantly
19 y; mix device vs. AEE or differ by accelerometer
of general characteristics AEE per kg recording period (≤1 week
populatio influence the vs. 41 week), body position
n, association (trunk vs. limbs), wear time
soldiers, between (waking hours vs. 24 hours),
and accelerometer- accelerometer output type
patients derived (uniaxial vs. triaxial outputs)
(COPD physical or accelerometer output
and activity output metrics (counts vs. steps vs.
cancer); and DLW- other) (all P-values of Mann–
high- derived AEE Whitney U-test and Kruskal–
income Wallis test 40.05). There was
countries a significant inverse
association between crude R2
and sample size (r = -0.45, p
= .03). There was no
significant correlation
between crude R2 and mean
age of participants (r = 0.16,
p = .44).

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 297

Author, Year Number Sample Predictor or Primary Quantitative or Qualitative Risk of Hetero- Overall
of Character- Intervention or Outcome Finding(s) Bias geneity of AMSTAR2
Studies istics Comparator Studies Rating
O’Driscoll et 60 1,946 EE estimate of - Overall, devices
al., 2020 males and wrist worn or underestimated EE (ES: -
females ≥ arm devices 0.23, 95% CI -0.44 to -0.03; n
19 y; (40 different = 104; p = .03) and showed
high- devices; 33 significant heterogeneity
income wrist worn) between devices (I2 =
countries 92.18%; p ≤ .001).
O’Driscoll et 60 1,946 TEE estimate - The pooled effect for TEE median - Partially well
al., 2020 males and of wrist-worn showed a significant score of 13, done/reported
females ≥ or arm devices underestimation of EE (ES: - 1 low
19 y; (10 different 0.68, 95% CI -1.15 to -0.21; n quality, 48
high- devices) = 16; p = .005), and moderate,
income significant heterogeneity was and 11 high
countries observed between devices (I2 quality
= 92.71%; p < .01). The SWA
p3 did not differ significantly
from criterion measures and
showed significant
heterogeneity (I2 = 94.20%; p
= .001).
Pisanu et al., 5 734 males REE estimated - One study obtained an Risk of bias - Well
2020 and from wearable underestimation of REE was judged done/reported
females ≥ accelerometer- SWA although the statistical as low (or partially
19 y with based devices significance was not well
overweigh specified. However, a done/reported
t and significant overestimation of if
obesity; SWA was observed in all the heterogeneity
high- other four studies. issue is
income Pearson’s correlation important)
countries coefficient was reported in
three studies, in which it

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

298 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Year Number Sample Predictor or Primary Quantitative or Qualitative Risk of Hetero- Overall
of Character- Intervention or Outcome Finding(s) Bias geneity of AMSTAR2
Studies istics Comparator Studies Rating
ranged between 0.58
(obtained in women) and 0.88
(obtained in the whole
population).
Results of Bland–Altman
analysis revealed the
tendency of the bias to
increase as the REE increased
across participants. Authors
did not find any relationship
between the bias and age,
BMI, fat-free mass, total body
water, and extracellular water
of individuals.
Bland–Altman plots indicated
that SWA systematically
overestimated REE in women
displaying low REE values
and underestimated REE in
women displaying high REE
values.
Pisanu et al., 9 339 males PAEE - A general trend toward Risk of bias - Well
2020 and estimated from overestimation can be was judged done/reported
females ≥ wearable noticed. However, the study as low (or partially
19 y with accelerometer- protocol differs greatly well
overweigh based devices among the included studies. done/reported
t and during different if
obesity; structured heterogeneity
high- physical issue is
income activities important)
countries

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 299

Author, Year Number Sample Predictor or Primary Quantitative or Qualitative Risk of Hetero- Overall
of Character- Intervention or Outcome Finding(s) Bias geneity of AMSTAR2
Studies istics Comparator Studies Rating
Pisanu et al., 5 185 males TEE or PAEE - The accuracy of the Caltrac Risk of bias - Well
2020 and free-living uniaxial accelerometer in the was judged done/reported
females ≥ from wearable measurement of TEE was as low (or partially
19 y with accelerometer- evaluated: even if the well
overweigh based devices accuracy of the instrument done/reported
t and was good at a group level, at if
obesity; individual level differences heterogeneity
high- were large. issue is
income An underestimation of EE in important)
countries free-living conditions was
obtained in one study. RT3
limits of agreement were
smaller than TriTrac-R3D,
but presented limitations at
individual levels.
Bland–Altman plots showed
that SWA and IDEEA
accurately estimated TEE,
and the IDEEA accelerometer
accurately measured AEE. On
the other hand, the
performance of Actical was
low. Accuracy of TEE and
AEE estimates of the SWA,
using software versions 6.1
and 5.1 in a sample of older
participants (78–89 years
old), which were overweight
as a group. Both versions
showed high Pearson’s
correlation coefficients (r >

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

300 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Year Number Sample Predictor or Primary Quantitative or Qualitative Risk of Hetero- Overall
of Character- Intervention or Outcome Finding(s) Bias geneity of AMSTAR2
Studies istics Comparator Studies Rating
0.75) for TEE. On the other
hand, AEE was
underestimated by both
versions 6.1 and 5.1.
Nevertheless, Bland–Altman
plots revealed no systematic
bias when considering both
TEE and AEE.

Plasqui et al., 25 944 males Validity of - Mean differences in TEE or - - Not well
2013 and wearable PA AEE between DLW and the done/reported
females; monitor accelerometer were often
high- estimates of EE small on the group level, but
income the limits of agreement (2
countries SD) were usually large.
Observed correlations
between PAL and activity
counts vary between 0.06
(Lifecorder) and 0.68
(TracmorD). Interpreting
correlations between AEE or
TEE and activity counts
becomes more difficult as
body mass and other
characteristics are the main
determinants of EE. Output
from the 3dNX accelerometer
significantly increased the
prediction of TEE in addition
to fat-free mass. The Tracmor
significantly contributed to

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 301

Author, Year Number Sample Predictor or Primary Quantitative or Qualitative Risk of Hetero- Overall
of Character- Intervention or Outcome Finding(s) Bias geneity of AMSTAR2
Studies istics Comparator Studies Rating
the prediction of TEE after
correcting for sleeping
metabolic rate, body mass, or
FFM. Likewise, the RT3
significantly contributed to
the prediction of TEE and
AEE after correction for
subject characteristics. When
AEE is expressed per kg body
mass, correlations with
activity counts vary between
0.37 (Actigraph) and 0.79
(Tracmor).
Sharifzadeh 30 3,877 Physical - The weighted mean - - Not well
et al., 2021 males and activity difference was not significant done/reported
females; questionnaire between TEE DLW -TEE
high- TEE (50 PAQ (WMD: -243, 95% CI (-
income questionnaires) 841.4 to 354.6), I2 = 97.9%, p
countries < .0001).
Sharifzadeh 15 2,058 Physical - A significant difference was - - Not well
et al., 2021 males and activity found between AEEs done/reported
females; questionnaire examined by various indirect
high- AEE (35 measures and the direct
income questionnaires) measures derived from DLW
countries (WMD: 414.6, 95% CI (78.7
to 750.5), I2 = 92%, p < .001)
in which AEE assessed by
DLW was higher than that of
measured by PAQ.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

302 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Year Number Sample Predictor or Primary Quantitative or Qualitative Risk of Hetero- Overall
of Character- Intervention or Outcome Finding(s) Bias geneity of AMSTAR2
Studies istics Comparator Studies Rating
Tudor-Locke 8 Males and Pedometer - Although a single study - - Not well
et al., 2002 females; versus energy comparing pedometer outputs done/reported
high- expenditure with energy expenditure
income derived from doubly labeled
countries water reported a significant
correlation of r = 0.61 in a
patient population, two other
studies reported no significant
correlations in different
populations (no reported r
values).
Tudor-Locke 8 Males and Pedometer - Pedometers generally - - Not well
et al., 2002 females; versus energy correlate with indirect done/reported
high- expenditure calorimetry from r = 0.49 to
income 0.81
countries
AEE = activity energy expenditure; COPD = Chronic obstructive pulmonary disease; DLW = doubly labeled water; EE = energy expenditure; ES = effect size;
FFM = fat-free mass; kg = kilogram; kJ = kilojoule; LIPA = light intensity physical activity; MD = mean difference; MET = metabolic equivalent of task; MPA
= moderate intensity physical activity; MPD = mean percentage difference; PA = physical activity; PAL = physical activity level; PAQ = physical activity
questionnaire; REE = resting energy expenditure; SD = standard deviation; SWA = SenseWear Armband; TEE = total energy expenditure; VPA = vigorous
intensity physical activity; WMD = weighted mean difference; y = years.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 303

TABLE J-2 Evidence on the Association of Macronutrient Composition of the Diet on Metabolic Efficiency (Energy Usage or
Energy Expenditure): Systematic Reviews
Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Character- Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR2
Studies istics or Studies Rating
Comparator
Ludwig et 29 617 male Low vs. high TEE Lower Among trials < 2.5 - I2 = Not well
al., 2021 and female carbohydrate carbohydrate weeks, the lower- 69.8%, p done/reported
adults 19– diet diet had lower carbohydrate diets < .001
50 y TEE for studies slightly reduced TEE.
< 2.5 weeks (-
50.0 kcal (-
77.4, -22.6)

Higher TEE Among trials of > 2.5

among > 2.5 weeks, the lower- I2 =
weeks 135.5 carbohydrate diet 26.4%, p
kcal (72.0, substantially = .255
198.7). increased TEE—by
Sensitivity ∼50 kcal/d for every
analysis 10% decrease in
produced carbohydrate as %
similar results EI—with minimal
residual
heterogeneity.
Park et 15 Adults 19– - - - Many studies - Not well
al., 2020 50 y with reported that the main done/reported
obesity or determinant of DIT is
lean/normal the energy content of
weight food, followed by the
protein fraction of
food. The thermic
effect of alcohol is
similar to that of

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

304 DIETARY REFERENCE INTAKES FOR ENERGY

protein. Therefore,
the main
determinants of DIT
are the energy
content and protein
fraction of the diet.
Quatela et 19 Male and Total energy DIT; The effect of This model shows - Not well
al., 2016 (related female intake RMR energy intake that DIT (kJ) done/reported
to adults 19 y on DIT increases
energy) and older (coefficient significantly (p <
0.011, standard .001) when the kJ
error 0.0013, p content of meals
< .001, 95% CI increases, although
0.0083; 0.014) this increase is of a
small magnitude
(coefficient 0.011).
This model predicts
that for every 100 kJ
increase in energy
intake, DIT increases
by 1.1 kJ/h. Model 2
produced similar
results. (47.4%
variance explained in
model 1; 70.6% in
model 2)
Cisneros 15 210 male type of fatty DIT or No conclusion - - - Not well
et al., and female acid EE can be drawn done/reported
2019 adults 19 y
and older
Wycherle 4 40 high protein REE >= 12 weeks There was Provide I2 = 64% Not well
y et al., participants (low fat) vs (secondar mean significantly less d risk of done/reported
2012 standard y difference was reduction in REE bias for
outcome) 130 kJ/day (- each

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 305

protein (low- 205.13, with a high-protein included

fat) 465.13); < 12 diet study
weeks 838
kJ/day (228.83,
1,447.17).
Across all time
595.50 kJ/day
(66.95,
1,124.05)
NOTE: CI = confidence interval; DIT = diet-induced thermogenesis; EE = energy expenditure; EI = energy intake; kJ = kilojoule; REE = resting energy
expenditure; RMR = resting metabolic rate; TEE = total energy expenditure; y = years.

TABLE J-3 Evidence on the Association of Body Composition on Metabolic Efficiency (Energy Usage or Energy Expenditure):
Systematic Reviews
Author, Number of Sample Predictor or Primary Quantitative Qualitative Finding(s) Risk of Hetero- Overall
Year Studies Character- Intervention or Outcome Finding(s) Bias geneity AMSTAR2
istics Comparator of Rating
Studies
Bailly et al., 29 for any Male and Constitutional TEE, See Table 7 in See Table 9 in Bailly et - - Partially
2021 meta- female thinness (CT) RMR, Bailly et al., al., 2021: CT well
analysis. 15 adults 19– vs. anorexia RMR/FF 2021: individuals have a done/report
assessed 50 y; nervosa or M, RQ, comparison of lower TEE, REE ed
TEE (2 included normal BMI AEE, CT vs. controls compared to normal
using pregnant PAL (C) weight; No diff in RQ,
DLW), women AEE, PAL between CT
RMR and normal weight;
indirect RMR/FFM trend of
calorimeter significant difference
n = 14 and such that C < CT (p =
9 with .083)
portable Comment:
devices, Meta-analysis
physical only done in

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

306 DIETARY REFERENCE INTAKES FOR ENERGY

activity women, no
measured cohort studies
with included
acceleromet because risk of
er n = 5 bias too high

NOTE: AEE = activity energy expenditure; BMI = body mass index; C = controls; CT = constitutional thinness; DLW = doubly labeled water; FFM = fat-free
mass; PAL = physical activity level; REE = resting energy expenditure; RMR = resting metabolic rate; RQ = respiratory quotient; TEE = total energy
expenditure; y = years.

TABLE J-4 Evidence on the Effect or Association of Weight Cycling on Metabolic Efficiency (Energy Usage/Expenditure) and
Health Outcomes: Systematic Reviews and Observational Studies
Author, Number Sample Predictor or Primary Quantitative Qualitative Finding(s) Risk of Hetero- Overall
Year of Characteristics Intervention or Outcome Finding(s) Bias geneity AMSTAR
Studies Comparator of 2 Rating
Studies
Zou et al., 14 253,766 males Weight cycling Type 2 RR = 1.23 Weight cycling increases - I2 = Partially
2021 and females 19 diabetes (1.07 to 1.41, p risk for new-onset 73.9% well-
y and older mellitus = .003) diabetes by 23% in done/repo
persons with BMI < 30 rted
Zou et al., 20 341,395 males Weight cycling All-cause RR = 1.41 Weight cycling increases - I2 = Well
2019 and females 19 mortality (1.27 to 1.57, p risk for all-cause mortality 78.1% done/repo
y and older < .001) by 41%, CVD mortality rted
by 36%, and risk for
hypertension by 35% in
adults
El Ghoch - 38 males and Weight cycling REE No change in Weight cycling does not - - -
et al., females 19–50 REE: 1,840.2 ± appear to adversely affect
2018 y with obesity 397.9 vs. REE in adults with morbid
1,831.9 ± obesity (BMI = 40)
408.9, p = .78

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 307

Author, Number Sample Predictor or Primary Quantitative Qualitative Finding(s) Risk of Hetero- Overall
Year of Characteristics Intervention or Outcome Finding(s) Bias geneity AMSTAR
Studies Comparator of 2 Rating
Studies
Nymo et - 38 males and Weight cycling REE REE only 70 Although weight loss - - -
al., 2019 females 19–50 kcal lower than associated with reduced
y baseline REE, there was no
association between REE
and weight cycling in
adults with class I/II
obesity
Bosy- - 47 males and Very low REE REE adjusted In overweight and obese - - -
Westphal females 19–50 calorie diet for changes in adults age 22–45, weight
et al., y with obesity organ and cycling shows a reduced
2013 tissue masses, REE when adjusted for
remains organ and tissue mass.
reduced on
weight cyclers,
p < .01.
Dombrow 45 7,788 males Diet Weight N/A Behavioral interventions - I2 = 75% Well
ski et al., and females cycling for weight loss done/
2014 19–50 y with maintenance in obese reported
overweight and adults reduces risk for
obesity weight regain/cycling.
Turichi et 43 2,379 males Diet Weight Amount of When controlling for the 1 study Tau2 Not done/
al., 2019 and females 19 cycling weight loss: R2 rate of weight loss, the high risk reported
y and older = 0.29, p < amount of weight loss of bias,
with .001; Rate of significantly predicts 4 studies
overweight and weight loss (R2 weight regain. low risk
obesity = 0.06, p = of bias,
.049) 38
medium
risk of
bias

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

308 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Finding(s) Risk of Hetero- Overall
Year of Characteristics Intervention or Outcome Finding(s) Bias geneity AMSTAR
Studies Comparator of 2 Rating
Studies
Fothergill - 14 males and Diet and TEE and REE reduced Metabolic adaptation in - - -
et al., females 19–50 exercise REE 704 ± 427 morbid obesity is
2016 y with class III kcal/d below associated with the degree
obesity baseline at 6 of weight loss; REE and
years after TEE remain reduced for 6
weight loss, p < years after weight loss
.0001 even with weight regain or
increased physical
activity.
Zhang et 4 92,063 females Weight cycling Endomet Odds ratio 1.23 Weight cycling is - - Partially
al., 2019 19 y and older rial to 2.33 associated with 1.2 to 2.3 well-
cancer fold increased risk for done/repo
endometrial cancer in rted
females age ≥ 18.

NOTE: BMI = body mass index; CVD = cardiovascular disease; DLW = doubly labeled water; N/A = not applicable; REE = resting energy expenditure; RR =
relative risk; TEE = total energy expenditure; y = years.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 309

TABLE J-5 Evidence on the Effect of Race or Ethnicity on Energy Expenditure

Author, Year Populations Sex Life Stage Conclusion Category Mean General Conclusions
Difference
(kcal/d)
Albu et al., 1997 B/W F Adults REE difference - adjusted 180 Lower REE in B vs. W
Foster et al., 1999 B/W F Adults REE difference - adjusted 135 Lower EE in B vs. W
Jakicic and Wing, B/W F Adults REE difference - adjusted 172 Lower EE in B vs. W
1998
Mika Horie et al., B/W F Adults REE difference - adjusted 200 Lower EE in B vs. W
2009
Reneau et al., 2019 B/W F/M Adults REE difference - adjusted 144 Lower REE in B vs. W, attenuated
with inclusion of trunk lean body
mass
Shook et al., 2014 B/W F Adults REE difference - adjusted 101 Lower EE in B vs. W, also lower
fitness levels
Olivier et al., 2016 B/W F Adults REE difference - adjusted 140 Lower EE in B vs. W
Sharp et al., 2002 B/W F/M Adults REE difference - adjusted 80 Lower EE in B vs. W, CARDIA
study
Spaeth et al., 2015 B/W F/M Adults REE difference - adjusted 100 Lower EE in B vs. W
Vander Weg et al., B/W F Adults REE difference - adjusted 65 Prediction equation, lower EE in B
2004
Wang et al., 2010 B/W F Adults REE difference - adjusted 121 Lower EE in B vs. W
Adzika Nsatimba et B/W F/M Adults REE difference - adjusted 250 Lower EE in B vs. W
al., 2016
Forman et al., 1998 B/W F Adults REE difference - adjusted 200 Lower EE in B vs. W
Santa Clara et al., B/W F Adults REE difference - adjusted 80 Lower EE in B vs. W
2006
Vander Weg et al., B/W F/M Adults REE difference - adjusted 78 Lower EE in B vs. W, no body
2000 composition; smokers
Martin et al., 2004 B/W F/M Adults REE difference - adjusted 135 Lower EE in B vs. W; diabetes
status
Most et al., 2018 B/W F Adults REE difference - adjusted 81 Early pregnancy; lower REE in B
vs. W

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

310 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Year Populations Sex Life Stage Conclusion Category Mean General Conclusions
Difference
(kcal/d)
Manini et al., 2011 B/W F/M Adults REE difference - adjusted 50 European admixture associated with
higher REE; elderly
Desilets et al., 2006 B/W F/M Adults REE difference - adjusted 110 Lower EE in B vs. W
Rush et al., 1997 Maori/W F Adults REE difference - adjusted 119 Lower REE in Maori vs. W
Wouters-Adriaens Asian/W F/M Adults REE no difference - 0 Equal REE after adjusting for body
and Westerterp, adjusted composition
2008
Byrne et al., 2003 B/W F Adults REE no difference - 0 Equal REE after adjusting for
adjusted detailed composition
Hunter et al., 2000 B/W F Adults REE no difference - 0 Equal EE after adjusting for trunk
adjusted lean body mass
Deemer et al., 2010 Hispanic/W F Adults REE no difference - 0 Equal REE but unadjusted
adjusted
Soares et al., 1998 Indian/W F/M Adults REE no difference - 0 Equal EE after adjusting for body
adjusted composition
Weyer et al., 1999 Pima/W F/M Adults REE no difference - 0 Higher TEE in Pima vs. W, equal
adjusted SMR
Javed et al., 2010 B/W F/M Adults REE no difference - 0 Equal after adjusting for organ
HMRO metabolic rate
Jones et al., 2002 B/W F Adults REE no difference - 0 Equal after adjusting for skeletal
HMRO muscle mass
Gallagher et al., B/W F/M Adults REE no difference - 0 Organ sizes/metabolic rates
2006 HMRO
Gallagher et al., B/W F/M Adults REE no difference - 0 Body composition differences
1997 HMRO
Song et al., 2016 Chinese/Indian/ M Adults REE no difference - 0 Lower EE in Asians, equal when
Malay HMRO adjusting for trunk lean body mass
Tranah et al., 2011 B/W F/M Adults REE no difference - 0 Equal EE after adjusting for mtDNA
mtDNA haplotypes; elderly
Glass et al., 2002 B/W F Adults REE no difference - 0 Equal EE
unadjusted

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 311

Author, Year Populations Sex Life Stage Conclusion Category Mean General Conclusions
Difference
(kcal/d)
DeLany et al., 2014 B/W F Adults TEE difference - adjusted 233 Lower EE B vs. W
Dugas et al., 2009 B/W F Adults TEE difference - adjusted 105 Lower EE in B vs. W
Lam et al., 2014 B/W F/M Adults TEE difference - adjusted 52 Lower EE in B vs. W, develop
predictive equation
Weinsier et al., B/W F Adults TEE difference - adjusted 138 Lower EE in B vs. W
2000
Most et al., 2018 B/W F Adults TEE difference - adjusted 230 Lower SMR and TEE in B vs. W;
early pregnancy
Blanc et al., 2004 B/W F/M Adults TEE difference - adjusted 200 Lower TEE and REE in B vs. W;
elderly
Walsh et al., 2004 B/W F Adults TEE difference - 116 Lower TEE in B vs. W, unadjusted
unadjusted
Weyer et al., 1999 Pima/W F/M Adults TEE difference (Pima -44
higher)
Katzmaryk et al., B/W F/M Adults TEE no difference - 0 Lower EE in B vs. W
2018 adjusted
Hunter et al., 2000 B/W F Adults TEE no difference - 0
adjusted
Kushner et al., 1995 B/W F Adults TEE no difference - 0 Equal TEE after adjusting body
adjusted composition
Lovejoy et al., 2001 B/W F Adults TEE no difference - 0 Lower SMR in B vs. W, equal TEE
adjusted
Saad et al., 1991 Pima/W M Adults TEE no difference - 0 Equal 24-hr EE, difference in
adjusted sympathetic nervous system activity
Christin et al., 1993 Pima/W M Adults TEE no difference - 0 Equal EE, norepinephrine turnover
adjusted as predictor
Fontveille et al., Pima/W F/M Adults TEE no difference - 0 Lower SMR in Pimas
1994 adjusted
Tershakovec et al., B/W F/M Children REE difference - adjusted 77 Lower EE in B vs. W, attenuated
2002 with inclusion of trunk lean body
mass

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

312 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Year Populations Sex Life Stage Conclusion Category Mean General Conclusions
Difference
(kcal/d)
Wong et al., 1996 B/W F Children REE difference - adjusted 52 Testing REE predictive equations;
greater overestimation in B
Bandini et al., 2002 B/W F Children REE difference - adjusted 62 Lower REE,TEE, AEE in B vs. W
Morrison et al., B/W F Children REE difference - adjusted 120 Lower REE in B vs. W
1996
Yanovski et al., B/W F Children REE difference - adjusted 92 Lower REE in B vs. W
1997
Wong et al., 1999 B/W F Children REE difference - adjusted 79 Lower REE in B vs. W
Sun et al., 2001 B/W F/M Children REE difference - adjusted 45 Lower REE in B vs. W
McDuffie et al., B/W F/M Children REE difference - adjusted 36 Lower EE in B vs. W; developed
2004 predictive equation
Pretorius et al., B/W F/M Children REE difference - adjusted 91 Lower EE in B vs. W
2021
Sun et al., 1998 B/W F/M Children REE no difference - 0 Equal EE
adjusted
Broadney et al., B/W F/M Children REE no difference - 0 Equal REE after adjusting for
2018 adjusted truncal composition
Hanks et al., 2015 B/W M Children REE no difference - 0 Looking at BMD as predictor
adjusted
Rush et al., 2003 Maori/Pacific F/M Children REE no difference - 0 Equal REE across groups
Islander/W adjusted
Spurr et al., 1992 Mestizo/B/ F/M Children REE no difference - 0 Equal EE across groups
Amerindian adjusted
Goran et al., 1995 Mohawk/W F/M Children REE no difference - 0 Lower EE in W vs. Mohawk
adjusted
Fontveille et al., Pima/W F/M Children REE no difference - 0 Equal REE
1992 adjusted
Bandini et al., 2002 B/W F Children TEE difference - adjusted 110 Lower TEE in B vs. W; prepubertal
and pubertal
DeLany et al., 2002 B/W F/M Children TEE difference - adjusted 62 Lower EE B vs. W

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 313

Author, Year Populations Sex Life Stage Conclusion Category Mean General Conclusions
Difference
(kcal/d)
Dugas et al., 2008 Hispanic/W F Children TEE difference - adjusted 60 Equal REE, lower AEE in Hispanic
Sun et al., 1998 B/W F/M Children TEE no difference - 0
adjusted
Goran et al., 1998 B/W/Mohawk/ F/M Children TEE no difference - 0 Equal REE across groups, lower
Guatemalan adjusted AEE in Guatemalans
Goran et al., 1995 Mohawk/W F/M Children TEE no difference - 0
adjusted
NOTE: AEE = activity energy expenditure; B = Black; BMD = bone mineral density; EE = energy expenditure; F = female; HMRO = high metabolic rate
organs; kcal/d = kilocalorie per day; M = male; mtDNA = mitochondrial DNA; REE = resting energy expenditure; SMR = sleeping metabolic rate; TEE = total
energy expenditure; W = White.

TABLE J-6 Evidence on How Physical Activity and Energy Expenditure Change Across the Life Span: Systematic Reviews
Author, Number Sample Predictor or Primary Quantitative Finding(s) Risk of Hetero- Overall
Year of Characteristics Intervention or Outcome Bias geneity AMSTAR2
Studies Comparator of Rating
Studies
Craigie et 22 11,889 males Association - In general, the correlation - - Not well
al., 2011 and females, between coefficients tended to be done/reported
children and physical stronger in the European
adults from activity levels studies (ranging from -0.01 to
high-income at baseline and 0.47), compared with
countries follow-up Canadian (-0.1 to 0.24),
United States (0.01 to 0.17)
or Australian studies (0.04 to
0.07).
In males coefficients varied
between -0.1 (nonsignificant,
at 22-year follow-up) and
0.47 (p < 0.001 for frequency
of activity over 8 years). In

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

314 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Finding(s) Risk of Hetero- Overall
Year of Characteristics Intervention or Outcome Bias geneity AMSTAR2
Studies Comparator of Rating
Studies
females these ranged between
-0.04 (nonsignificant over 7
years) and 0.37 (p < .001 over
6 years).
Craigie et 13 4,999 males Maintenance of - Over 5–8 years follow-up - - Not well
al., 2011 and females, relative from adolescence between done/reported
children and position— 44% and 59% maintained
adults from physical their tertile position for
high-income activity activity, with higher
countries proportions for males than for
females. In the
Cardiovascular Risk in
Young Finns study
participants, the probability of
9–18-year-olds remaining
active 6 years later (44% of
all participants) was
significantly weaker than the
probability of remaining
sedentary (57% of all
participants) (p = .002).
Craigie et 10 17,654 males The probability - Four studies reported the - - Not well
al., 2011 and females, of being probability of being done/reported
children and physically physically active in adulthood
adults from active at using odds ratios. However a
high-income follow-up comparison of their findings
countries according to is complicated by the
activity at variation in categories used in
baseline their analyses. The
Amsterdam Growth and

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 315

Author, Number Sample Predictor or Primary Quantitative Finding(s) Risk of Hetero- Overall
Year of Characteristics Intervention or Outcome Bias geneity AMSTAR2
Studies Comparator of Rating
Studies
Health Longitudinal Study
reported general daily
physical activity: those in the
lowest quartile for daily
physical activity at 13 years
old were 3.6 times more
likely (95% CI 2.4 to 5.4) to
be in the lowest quartile 14
years later than those in the 3
higher quartiles at baseline.
Foulds et 8 915 males and Average PAL via Overall average total energy Citations - Partially well
al., 2013 females; physical DLW and expenditure among Native included done/reported
Native activity metabolic American adults was found to in the
American levels—adults chamber be 10.53 megajoules (MJ), physical
population in with 2.28 MJ of activity activity
Canada and energy expenditure. Overall, behaviour
United States Native American adults were assessmen
found to have PAL ratios t consisted
averaging 1.48. of a range
of grades
from 1A
to 3B and
an
average
quality
score of
11 out of
15 (range
6–14)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

316 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Finding(s) Risk of Hetero- Overall
Year of Characteristics Intervention or Outcome Bias geneity AMSTAR2
Studies Comparator of Rating
Studies
Foulds et 2 408 males and Average PAL via Among children at age 5 Citations - Partially well
al., 2013 females; physical DLW and years, overall average total included done/reported
Native activity metabolic energy expenditure was found in the
American levels—adults chamber to be 5.93 MJ, with 1.17 MJ physical
population in of activity energy activity
Canada and expenditure, resulting in a behaviour
United States PAL ratio of 1.42. Results assessmen
among other ages of t consisted
children/youth are not of a range
available in the literature. of grades
from 1A
to 3B and
an
average
quality
score of
11 out of
15 (range
6–14)
Foulds et 5 > 100,000 Physical PAL via More recent reports of Citations - Partially well
al., 2013 publishe males and activity change self-report physical activity behavior included done/reported
d from females; over time among Native American in the
1980 to Native adults identify individuals as physical
1989, 14 American being less active than the activity
publishe population in 1990s. Overall, greater behaviour
d from Canada and proportions of Native assessmen
1990 to United States American adults from 2000 to t consisted
1999, 2011 reported inactive levels of a range
and 20 of activity compared to earlier of grades
from assessments, with lower from 1A

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 317

Author, Number Sample Predictor or Primary Quantitative Finding(s) Risk of Hetero- Overall
Year of Characteristics Intervention or Outcome Bias geneity AMSTAR2
Studies Comparator of Rating
Studies
2000 to proportions reporting to 3B and
2011 insufficient PALs. an
average
quality
score of
11 out of
15 (range
6–14).
Tanaka et 10 7,238 males Longitudinal Average The follow-up duration Study - Partially well
al., 2014 and females; changes in sedentary ranged from 1.0 to over 10.0 methodol done/reported
children and overall behavior years. The age of the ogical
adolescents; sedentary change per participants at baseline quality
from high- behavior year via ranged from 3.8 to 13.2 years. was rated
income wearable The overall percentage daily as high
countries devices sedentary behavior change with all
per year ranged from -3.8% to 10 papers
12.5% for boys and from - rated as
2.5% to 12.7% for girls, with >=70%
a weighted mean increase of
daily sedentary behavior of
+5.7% in boys and 5.8% in
girls, equivalent to additional
approximately 30 min of
daily accelerometer-measured
sedentary behavior per year.

NOTE: CI = confidence interval; DLW = doubly labeled water; mJ = megajoule; PAL = physical activity level.

TABLE J-7 Evidence on the Effect of BMI (and Other Measures of Adiposity) on Energy Balance or Energy Expenditure: Systematic
Reviews

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

318 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Year Number Sample Predictor or Primary Quantitative Qualitative Finding(s) Risk of Hetero- Overall
of Characteristics Intervention or Outcome Finding(s) Bias geneity AMSTAR
Studies Comparator of 2 Rating
Studies
Ashtary- 7 361 males and Gradual weight Weight Gradual weight Gradual weight loss 3/7 low - Partially
Larky et al., females 19 y loss change loss preserved produces less reduction well done/
2020 and older with REE by ~100 in REE than rapid reported
overweight and kcals compared weight loss and a
obesity to rapid weight greater loss of fat mass
loss and percent body fat.
Cheng et al., 12 1,499 males Pubertal REE REE increases Both REE and TEE are medium - Partially
2016 and females 9– 12% and TEE significantly higher well done/
18 y increases 16% during puberty. reported
during puberty
Nunes et al., 33 2,528 males Weight loss REE or REE and TEE In adults, there is Low to - Partially
2022 and females 19 TEE show up to adaptive thermogenesis medium well done/
y and older 20% greater with weight loss reported
decrease than leading to a greater
predicted. than predicted decrease
in energy expenditure.
Schwartz and 90 2,996 males Diet or diet REE REE decreases The 15 kcal/kg - - Not well
Doucet, 2010 and females 19 plus exercise or 15 kcal/kg decrease in REE during done/
y and older diet plus during weight weight loss does not reported
with pharmacologic loss. differ by sex. Short
overweight and al intervention interventions (2–6
obesity weeks) have greater
decrease in REE than
long intervention (> 6
weeks).
Dhurandar et 32 1,680 males Diet Compens Diet restriction Energy compensation medium - Not well
al., 2015 and females ation results in 12– (intake and/or done/
19–50 y with 44% less expenditure) leads to reported
normal weight, weight loss less weight loss than
than predicted.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 319

Author, Year Number Sample Predictor or Primary Quantitative Qualitative Finding(s) Risk of Hetero- Overall
of Characteristics Intervention or Outcome Finding(s) Bias geneity AMSTAR
Studies Comparator of 2 Rating
Studies
overweight, predicted with diet
and obesity restriction.
Kee et al., 20 Males and BMI REE REE ranges REE increases with - - Not well
2012 females 19–50 1,800–2,600 increasing BMI in done/
y with morbid kcal in adults morbid obesity (BMI ≥ reported
obesity (BMI ≥ with BMI ≥ 40 40).
40)
Nunes et al., 94 males and Diet; calorie REE Reduction in Adaptive - - Partially
2021 females 19 y restriction REE ranges -70 thermogenesis occurs well done/
and older with averaged 270 to -220 kcal/d with moderate weight reported
overweight and kcal/d more than loss of 5%.
obesity predicted.
Schwartz et 90 815 males and Diet or diet REE Reduction in Reduction in REE - - Not well
al., 2012 females 19 y plus exercise or REE 29.1% greater than predicted done/
and older with diet plus greater than from Harris Benedict reported
overweight and weight loss predicted by equation, but Harris
obesity intervention Harris Benedict Benedict equation after
equation. weight loss may
overestimate energy
intake needs for weight
maintenance.

NOTE: BMI = body mass index; kcal = kilocalorie; kg = kilogram; REE = resting energy expenditure; TEE = total energy expenditure; y = years.

TABLE J-8 Evidence on How the Increase in Tissue Deposition Associated with Growth During Infancy, Childhood, and
Adolescence Influences, Effects, or Contributes to Energy Requirements
Author, N Sex Age (SD) Ethnicity Weight Gain Protein FFM Gain FM Gain Energy
Year g/day (SD) Gain g/day (SD) g/day Deposition
(SD)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

320 DIETARY REFERENCE INTAKES FOR ENERGY

g/day kcal/day
(SD) (SD)
DeLany et 28 F 10.7 (0.7) Black 10.7 (4.3) 8.1 (1.6) 2.6 (3.6) 32.72
al., 2006 25 F 10.6 (0.4) White 10.80 (4.7) 7 (2.3) 3.8 (3.3) 42.64
31 M 10.9 (0.8) Black 12.8 (5.2) 9.2 (4.3) 3.5 (5) 42.22
29 M 10.9 (0.6) White 9.7 (6.1) 7.5 (4.3) 2.2 (5) 28.38

Plachta- 680 M 6–10 y 12.2 kg/4 y 10.6 kg/4 y 1.8 kg/4 y 19.3 (50)
Danielzik et 684 F 6–10 y 12.7 kg/4 y 10.0 kg/4 y 2.7 kg/4 y 24.5 (50)
al., 2008
254 M 10–14 y 21.5 kg/4 y 18.5 kg/4 y 2.9 kg/4 y 31.8 (50)
260 F 10–14 y 18.4 kg/4 y 12.5 kg/4 y 5.7 kg/4 y 45.6 (50)
Wells and 49 41% M 1.5 mo White 0.24 kg/wk (0.08) 3.3 (1.4) 14.4 (3.2) 152.0 (4.8)
Davies, 92 59% F 0.2 (0.1) 2.8 (1.7) 12.8 (3.7) 134.3 (4.3)
1998 37 0.12 (0.1) 2.5 (1.7) 3.7 (4) 46.6 (7.6)
36 0.11 0.11) 2.4 (1.9) 3.1 (5.2) 42.8 (9.1)
18 0.09 (0.09) 2.1 (1.6) 1.7 (3.3) 28.0 (9.1)

NOTE: F = female; g = gram; kcal = kilocalorie/day; kg = kilogram; M = male; mo = months; SD = standard deviation; wk = weeks; y = years.

TABLE J-9a Evidence on How the Increase in Tissue Deposition Associated with Pregnancy Influences, Effects, or Contributes to
Energy Requirements: Nonsystematic Reviews
Author, N Age BMI Ethnicity Gestational Protein Gain FFM Gain FM Gain
Year (SD) Status Weight Gain g/day (SD) g/day g/day (SD)
g/day (SD) (SD)
Catalano et 6 normal, 10 31.8 y 20.8 - 13.5 7.3 kg 2 kg from
al., 1998 GDM/IGT (5.5) from pre preconception to
to 36 36 weeks
weeks
Kopp- 10 29.1 y 23.1 - 11.6 kg at 36 4.5 kg from
Hoolihan et (5) weeks (4.3) preconception to
al., 1999 34/36 weeks

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 321

Berggren et 11 29 y 23.8 10 White 17.5 median 12.2 3.5 kg (median)

al., 2015 (median) 1 non- from (median)
White preconception to
34/36 weeks
Okereke et 8 NGT, 7 NGT obese > - 12.7 kg NGT at 5.8 NGT 6.9
al., 2004 GDM 31.6 y 25% body 36 weeks
(3.4) fat, 8
NGT 26.2
Abeysekera 26 - - - 10.8 (3.9 kg) 3.9 (2.4) 7.0 (3.6) kg
et al., 2016 from 12–14 to kg
34–36 weeks

NOTE: BMI = body mass index; g = gram; FFM = fat-free mass; FM = fat mass; GDM = gestational diabetes mellitus; IGT = impaired glucose tolerance; kcal =
kilocalorie; kg = kilogram; NGT = normal glucose tolerance; SD = standard deviation; y = years.

TABLE J-9b Evidence on How the Increase in Tissue Deposition Associated with Pregnancy Influences, Effects, or Contributes to
Energy Requirements: Systematic Reviews
Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity AMSTAR2
Studies ics or of Rating
Comparator Studies
Savard 32 Pregnant Pregnancy REE/TEE Increases in REE REE and TEE Huge - Partially well
et al., women, ranged increase during variability. done/reported
2021 mostly from 0.5% to 18.3% pregnancy, mainly Inclusion of
White (8 to 239 kcal) from early to late women
between early and and from mid- to with
midpregnancy, from late pregnancy. excessive
3.0% to 24.1% (45 to Great variability gestational
327 kcal) between in the extent to weight gain
mid- and late which REE and and sample
pregnancy, and from TEE increase with small
6.4% to 29.6% (93 to throughout number of
416 kcal) between pregnancy. overweight

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

322 DIETARY REFERENCE INTAKES FOR ENERGY

early and late or obese

pregnancy. women
The median increases may have
in REE were 5.3% lead to
(72 kcal), 9.9% (153 over-
kcal), and 18.0% estimation
(252 kcal) between of energy
early and mid-, mid- requirement
and late, and early s.
and late pregnancy,
respectively.

Increases in TEE
ranged from 4.0% to
17.7% (84 to 363
kcal) between early
and midpregnancy,
from 0.2% to 30.2%
(5 to 694 kcal)
between mid- and
late pregnancy, and
from 7.9% to 33.2%
(179 to 682 kcal)
between early and
late pregnancy,
respectively. The
median increases in
TEE were 6.2% (144
kcal), 7.1% (170
kcal), and 12.0%
(290 kcal) between
early and mid-, mid-
and late, and early

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 323

and late pregnancy,

respectively.

NOTE: kcal = kilocalorie; REE = resting energy expenditure; TEE = total energy expenditure.

TABLE J-10a Evidence on How the Increase in Tissue Deposition Associated with Lactation Influences, Effects, or Contributes to
Energy Requirements: Nonsystematic Reviews
Author, Year N Age BMI Status Ethnicity Weight Gain Findings
(SD) g/day (SD)
Pereira et al., 52 and 49 32 y (4) 27.3 (5.6) White Negative 0.8 BMI units FFM gain of 0.4 g from 3 to 9
2019 from 3 to 9 months months pp
postpartum (pp)
FM loss of 2 grams

REE and TEE measure by whole

body calorimetry. REE increased
significantly by 48 (108 kcal day)
3.2% at 3 months; breast milk
volume 771 (261) g/d for BM
energy output of 678 (230)
kcal/day. At 9 months breast milk
vol 530 (225) g/d for breast milk
energy output of 465 (198) kcal/d.
41/52 and 28/49 were BF at 3 and
9 mo. TEE at 9 months 2,028
(286) kcal/d. No difference in TEE
between lactating and
nonlactating.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

324 DIETARY REFERENCE INTAKES FOR ENERGY

Thakkar et 50 28–33 y Asian Energy content of HM at 1 months

al., 2013 was 65.92 (9.43) kcal/100 ml, at 3
months 70.24 (22.0). Energy
content for milk produced for male
infants was greater. Figure 1
shows significant difference at 3
months of 14.8 kcal/100 ml or
24% difference.

Nielsen et al., 47 and in the end 33.7 y 25.0 (3.9) White Mean weight at 15 Mean milk intake (DLW) 923 (SD
2011 n = 30 with 26 (4.3) days was male 6.72 = 122) g/day at 15 weeks and 997
EBF (0.78) and female 6.30 (SD = 142) g/day at 25 weeks for
(0.64); male 7.84 all infants. For EBF 999 (SD =
(0.91) and female 7.37 146) g/day at 25 weeks. Milk
(0.75) at 25 weeks energy content 2.72 (SD = 0.38) at
15 weeks, and 2.62 (SD = 0.40)
kg/g at 25 weeks. No different by
sex. Energy intakes male 2,582
(SD = 362) and females 2,403 (SD
= 215) kJ/day at 15 weeks and
males 2,748 (SD = 480) and
females 2,449 (SD = 312) kJ/day
at 25 weeks. Significant difference
by sex at 25 weeks (Table 2 in
paper). However, milk and energy
intake decreased from 15 weeks to
25 weeks (Table 3).

Nielsen et al., - - - See Table 2 in Nielsen Same as above (Nielsen et al.,

2013 et al., 2013 2011) but now used DLW to
measure TEE

NOTE: BF = breast feeding; BMI = body mass index; DLW = doubly labeled water; EBF = exclusively breast feeding; FFM = fat-free mass; FM = fat mass; g =
gram; kg = kilogram; kJ = kilojoule; ml = milliliter; pp = postpartum; SD = standard deviation; TEE = total energy expenditure; y = years.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 325

TABLE J-10b Evidence on How the Increase in Tissue Deposition Associated with Lactation Influences, Effects, or Contributes to
Energy Requirements: Systematic Reviews
Author, Number Sample Predictor or Primary Quantitative Finding(s) Qualitative Risk of Hetero- Overall
Year of Characteristics Intervention Outcome Finding(s) Bias geneity AMSTA
Studies or of R2
Comparator Studies Rating
Reilly et 3–4 3–4 months Not Milk At 3–4 months: The Cross-sectional Risk of - Partially
al., 2005 months 1,041; 5–6 applicable transfer weighted mean milk studies of milk bias was well
33; 5–6 months 99; at 6 transfer was 779 g/d (SD = transfer suggest that provided done/
months months 72 40), and the unweighted it typically varies for reported
6; 6 mom–infant mean was 796 g/d (SD = between included
months dyads; 48) (95% CI 778, 812 approximately 779 studies
5 exclusively g/day). At 5–6 months: g/d at age 3–4 months
breast feeding Weighted mean milk (for which there was
transfer was 826 g/d (SD = a great deal of
39). The unweighted mean evidence: 33 studies
was 816 g/d (SD 42) (95% of 1,041 mother–
CI 772, 860 g/d. At 6 infant pairs and
months: Weighted mean approximately 894
milk transfer was 894 g/d g/d at age 6 months
(SD = 87) and unweighted (for which evidence
mean transfer 883 g/d (SD = was limited: five
89) (95% CI 790, 975 g/d). studies with 72
Changes in BM transfers possibly highly
between 2 to 5 months from selected mother–
nine studies reported no infant pairs;
marked increase in milk longitudinal studies,
transfer over the periods of in contrast, did not
time measured, and most suggest any marked
described the pattern of increase in milk
change in intake over time transfer over time
as a ‘plateau’ in milk over the period of 3–

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

326 DIETARY REFERENCE INTAKES FOR ENERGY

transfer after 3 months. The 6 months. The

weighted mean metabolizable energy
metabolizable energy content of breast milk
content of milk from 25 is approximately 2.6
papers of 777 mom–infant kJ/g. They speculate
pairs was 2.6 (SD = 0.2) that using lower
kJ/g (equivalent to 0.62 values for breastmilk
kcal/g) (see Table 4 in energy content than
Reilly et al., 2005). the 0.67 to 0.68
kcal/g used in WHO
reviews might alter
the apparent
adequacy of
exclusive
breastfeeding to 6
months of age.
NOTE: CI = confidence interval; d = day; g = gram; kcal = kilocalorie; kJ = kilojoule; SD = standard deviation; WHO = World Health Organization.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 327

TABLE J-11 Evidence on the Calorie Intake Needed to Achieve Weight Loss (if Overweight), Weight Maintenance (for All
Individuals), or Weight Gain (if Underweight): Systematic Reviews
Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Character- Intervention or Outcome Finding(s) Finding(s) Bias geneity AMSTAR
Studies istics Comparator of 2 Rating
Studies
Heymsfield 10 150 obese Relationship TEE- Mean difference Limited literature, - - Not well
et al., 2007 subjects on between measured DLW or between measured but findings done/
low-calorie and predicted TEE indirect and predicted TEE support that low reported
diet and among reduced calorimet for all reduced patient adherence
patients obesity after long- er obesity subjects 20.1 is the main basis
with term (>= 26 kcal/day (-58, -155) for modest weight
reduced weeks) weight % difference 1.3% (- loss associated
obesity loss treatment 1.7, -8.5). From the with low calorie
reduced DLW studies— diet (LCD). Obese
obesity difference in -518 subjects have
kcal/day. Reduction weight loss < 50%
in energy intake of of expected for the
~500 kcal/day had a degree of
weight loss of 30 kg. prescribed LCD
energy deficit.
TEE in the
reduced obesity
state is close to
predicted in never
obese subjects
(1%).
NOTES: DLW = doubly labeled water; kcal = kilocalorie; kg = kilogram; LCD = low calorie diet; TEE = total energy expenditure.

TABLE J-12 Evidence on the Association Between Weight Change and Chronic Disease Outcomes: Systematic Reviews

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

328 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
Alharbi 2 715 Intentional All-cause RR (95% CI) In this small good Moderate Well
et al., community- weight loss mortality = 0.92 (0.54 sample of older heterogen done/
2021 dwelling risk to 1.54) adults, intentional eity reported
male and weight loss was
females 65 not associated p = .99; I2
y and older; with all-cause = 56%
not all from mortality.
high-
income More research is
countries needed on the
effect of
intentional weight
loss on all-cause
mortality or the
reasons for
intentional weight
loss in older
community-
dwelling adults.

Older,
community-
dwelling adults
with very small
sample size and
no information on
how weight loss
was measured

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 329

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
Alharbi 23 1,210,116 Weight All-cause RR (95% CI) No information on Most were Low Well
et al., community- gain mortality = 1.10 (1.02 whether weight good heterogen done/
2021 dwelling risk to 1.17) gains or losses eity reported
male and were intentional
females 65 p = .01; I2
y and older; Weight gain had a = 41%
not all from small, but
high- significant
income association with
countries all-cause
mortality.

In community-
dwelling older
adults, weight
gains are
associated with an
increased risk of
all-cause mortality
relative to stable
weight.

Weight gain data

were a mixture of
measured and
self-reported.
Need research on
reason for weight
gain.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

330 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
Alharbi 4 6,901 Weight All-cause RR (95% CI) No information on Most were No Well
et al., community- fluctuation mortality = 1.66 (1.28, whether weight good significant done/
2021 dwelling risk 2.15) gains or losses heterogen reported
male and were intentional eity
females 65
y and older; A 63% increased p = .31; I2
not all from risk of all-cause = 14.6%
high- mortality with
income weight fluctuation
countries compared to
stable weight
reference

In community-
dwelling older
adults, weight
fluctuations are
associated with an
increased risk of
all-cause mortality
relative to stable
weight.

Weight fluctuation
data were a
mixture of
measured and
self-reported.
Need research on
effect of

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 331

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
intentional vs.
unintentional
weight
fluctuations.

Capristo 17 39,875 Weight loss All-cause OR (95% No significant Suboptimal No Not well
et al., males and associated mortality CI): 1.03 reduction in risk quality significant done/
2021 females ≥ with anti- (0.87 to 1.21) of all-cause heterogen reported
18 y with obesity mortality with eity
overweight medications weight lowering
or obesity; drugs compared I2 = 0%; p
not all from with placebo or no = 1.0
high- treatment.
income
countries There was a weak,
but statistically
significant linear
association
between all-cause
mortality and
magnitude of
weight loss (ß =
0.0007, p = .045).
A weight loss of
20 kg would lower
mortality by 1.4%
and a 30 kg
weight loss by
2.1%.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

332 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
Although unable
to demonstrate a
superiority of anti-
obesity
medications over
placebo, meta-
regression showed
that even a small
weight reduction
tends to reduce
all-cause mortality
in obesity.

The health status

of participants is
not described.
Capristo 8 28,657 Weight loss Cardio- OR (95% No significant Suboptimal No Not well
et al., males and associated vascular CI): 0.92 decrease in the quality significant done/
2021 females ≥ with mortality (0.72 to 1.18) risk of CVD death heterogen reported
18 y with antiobesity with antiobesity eity
overweight medications drugs
or obesity; I2 = 0%; p
not all from Linear association = .79.
high- between CVD
income mortality and
countries magnitude of
weight loss was
not significant.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 333

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
Unable to
demonstrate an
effect of weight
loss medications
on CVD mortality
in trials with an
average of 52
weeks follow-up.

Unclear as to the
health status of
participants

Capristo 7 30,404 Weight loss Myocardi OR (95% No significant Suboptimal No Not well
et al., males and associated al CI): 1.01 decrease in the quality heterogen done/
2021 females ≥ with anti- infarction (0.86, 1.19 risk of myocardial eity reported
18 y with obesity infarction with
overweight medications anti-obesity drugs. I2 = 0%, t2
or obesity; = 0, p =
not all from Unable to .87.
high- demonstrate an
income effect of weight
countries loss medications
on myocardial
infarction in trials
with an average of
52 weeks follow-
up.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

334 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
Unclear as to the
health status of
participants or if
these were
incidence cases

Capristo 4 21,584 Weight loss Stroke OR (95% Unable to Suboptimal No Not well
et al., males and associated CI): 0.93 demonstrate effect quality heterogen done/
2021 females ≥ with anti- (0.72 to 1.20) of weight loss eity reported
18 y with obesity medications on
overweight medications stroke I2 = 0%, t2
or obesity; = 0, p =
not all from Unclear as to the .49
high- health status of
income participants or if
countries these were
incidence cases

Chan et 8 1,373 Adult Premeno RR (95% Inverse Most I2 = 0%, p Not well
al., 2019 females ≥ weight loss pausal CI): 0.85 associations for studies = .93. done/
18 y; of unknown breast (0.74 to 0.99) premenopausal considered reported
underweigh intention cancer breast cancers average to
t women when comparing good
(BMI < any weight loss of quality.
18.5 kg/m2) unknown Higher or
were intention from age lower RoB
excluded; 18 y to study studies on
not all from baseline with average did
high- stable weight. not find

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 335

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
income statistically
countries The results were different
not robust and associations
require further in the
confirmation. subgroup
meta-
analyses.
Chan et 14 8,283 Adult Postmen RR (95% Inverse Most I2 =24%, Not well
al., 2019 females ≥ weight loss opausal CI): 0.90 associations for studies p done/
18 y; of unknown breast (0.81 to 0.99) postmenopausal considered heterogen reported
underweigh intention cancer breast cancers average to eity =
t women when comparing good 0.20.
(BMI < any weight loss of quality.
18.5 kg/m2) unknown Higher or
were intention from age lower RoB
excluded; 18 y to study studies on
not all from baseline with average did
high- stable weight. not find
income statistically
countries The results were different
not robust and associations
require further in the
confirmation. subgroup
meta-
analyses.
Chan et 9 Females ≥ Adult Premeno RR (95% CI) No association of Most I2 = Not well
al., 2019 18 y; weight gain pausal = 1.00 (0.97 weight gain and studies 20.7%, p done/
underweigh per 5 kg (of breast to 1.03) breast cancer in considered = .265 reported
t women unknown cancer premenopausal average to
(BMI < intention) women good

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

336 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
18.5 kg/m2) quality.
were Higher or
excluded; lower RoB
not all from studies on
high- average did
income not find
countries statistically
different
associations
in the
subgroup
meta-
analyses.
Chan et 16 Females ≥ Adult Postmen RR (95% CI) Positive Most I2 = 64%; Not well
al., 2019 18 y; weight gain opausal = 1.07 (1.11 association of studies p done/
underweigh per 5 kg (of breast to 1.23) weight gain and considered heterogen reported
t women unknown cancer breast cancer in average to eity ≤
(BMI < intention) postmenopausal good .001
18.5 kg/m2) women quality.
were Higher or
excluded; lower RoB
not all from studies on
high- average did
income not find
countries statistically
different
associations
in the
subgroup

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 337

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
meta-
analyses.
Hao et 19 862,177 Highest Onset of Highest vs. Weight gain in No data Highest Partially
al., 2021 females ≥ adult breast lowest Asian women had vs. lowest well done/
19 y; weight gain cancer or weight gain a much stronger weight reported
American, since early total and effect (34%) than gain in
European, adulthood cancer premenopaus in other countries. premenop
Australia, for both al risk: RR = No significant ausal
Asian whole 1.00 (95% findings among women: I2
(Japanese, adulthood CI, 0.83, premenopausal = 24.9%.
Chinese) and 1.21); women: RR = Postmeno
hormone- postmenopau 1.00, 95% CI, pausal
changed sal risk: RR = 0.83, 1.21 women I2
menopause 1.55 (95% = 47.2%.
stages CI, 1.40, Dose–response Dose–
1.71). Dose– analysis response:
response: RR confirmed a postmeno
per 5 mg significant pausal I2
weight gain: increased risk of = 69.4%.
1.08 (95% 8% of developing
CI, 1.07, postmenopausal
1.09). Weight breast cancer with
gain since each 5 kg
menopause: increment in adult
RR = 1.59 weight gain for
(95% CI, Western women,
1.23, 2.05). but about a 34%
stronger risk in
Asian women. No
significant finding

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

338 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
among
premenopausal
women. Higher
weight gain since
menopause
associated with
increased
postmenopausal
breast cancer risk
based on
comparison of
highest vs. lowest
adult weight gain.

For
postmenopausal
women, there was
a significant effect
of weight gain
since menopause
on breast cancer
risk. The effect is
strongest in Asian
women. No effect
of weight gain on
breast cancer risk
in premenopausal
women.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 339

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
The majority of
participants came
from Europe,
United States,
UK, Canada,
Australia. Only a
small minority
were from China
or Japan.
Jayedi 5 134,247 Weight Hyperten There was a There was a linear No data I2 = Well
et al., males and gain equal sion linear association 77.8%. P done/
2018 females; to a 1-unit incidence association between weight hetogeneit reported
general increment between gain and risk of y = 0.001.
population in BMI weight gain hypertension (Pnon-
> 18 y with (both self- and risk of linearity = 0.58)
>1y reported hypertension
follow-up; and (P non- Adjustment for
high- measured linearity = baseline blood
income weights) 0.58) pressure
countries attenuated the
associations, but
results remained
significant
indicating that
adiposity
increases the risk
of hypertension
independent of
baseline blood
pressure. Greater

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

340 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
risk in self-
reported subgroup
compared with
measured.

Preventing weight
gain in adults is a
useful approach
for reducing the
risk of
hypertension in
adults.

The study
provided evidence
of the role of
weight gain in
hypertension risk.
One limitation
was the failure of
included studies to
control for salt
intake or renal
function. Some
evidence of
publication bias.
Jayedi of 11 < 505,802 Weight CVD RR (95% CI) A nonlinear dose– Out of a I2 = 84%, Partially
et al., studies males and gain during mortality = 1.14 (1.02 response analysis possible P well done/
2020 with females ≥ adulthood in to 1.26) for a indicated that the score of heterogen reported
data on 18 y persons 5 kg risk of CVD "9", 1/3 of eity = <

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 341

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
CVD reporting without increment in mortality did not the studies 0.001; P
mortality unintended pre- body weight change materially were rated heterogen
,5 weight gain existing with weight gain as "7" and eity
studies during CVD of 0 to 5 kg and 2/3 as "8". between
had data adulthood then increased subgroups
on or before sharply at weight = 0.15.
participa assessment; gain > 6 kg.
nts Europe
without (13), United Measuring weight
pre- States (8), gain during
existing Asia (2), adulthood may be
CVD Australia better than static,
(1), Middle cross-sectional
East (1) assessment of
weight because it
considers trend
over time, and
thus, can be used
as a
supplementary
approach to
predict CVD.

Adult weight gain

could increase the
risk of CVD
incidence and
mortality.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

342 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
Slightly more than
half of the studies
relied on self-
reported weight
gain, which could
have attenuated
relationships.
Jayedi 2 118,140 Weight CVD RR (95% CI) In five studies in Out of a I2 = 6%. Partially
et al., males and gain during incidence = 1.12 (1.10, which participants possible P well done/
2020 females ≥ adulthood 1.13) for a 5 with preexisting score of heterogen reported
18 y kg increment CVD were "9", 1/3 of eity =
reporting in body excluded, the RR the studies 0.30.
unintended weight (95% CI) = 1.14 were rated
weight gain (1.02 to 1.26).I2 = as "7" and
during 84% (P < .001) 2/3 as "8".
adulthood and between
or before group
assessment; heterogeneity =
Europe 0.15.
(13), United
States (8), Measuring weight
Asia (2), gain during
Australia adulthood may be
(1), Middle better than a
East (1) static, cross-
sectional
measurement of
weight (e.g., BMI)
for predicting
CVD risk.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 343

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator

Adult weight gain

may be associated
with a higher risk
of CVD.

Karahali 18 Healthy Weight at All-cause No data Used results from No data I2 = Partially
os et al., adults baseline mortality a subgroup of 64.4%, well done/
2017 measured and follow- participants whose tau2 = reported
between up based on weights were 0.16.
middle and measured based on Ratio of
older age weight measured values HRs =
(subgroup rather on the full 1.00.
No data on analysis) sample that
number of combined
participants measured and
self-reported
weights.

Weight gain in
middle-aged to
older adults is
associated with
muscle mass
decreases and fat
mass increases
with the largest
increase in

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

344 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
visceral and
abdominal fat.

Weight gain from

middle to older
adulthood was
associated with a
slightly increased
risk of all-cause
mortality.

Studies using self-

reported measures
of weight at
baseline and
follow-up had
higher HRs than
studies with
measured weight.
None of the
participants were
underweight at
baseline.
Karahali 11 Healthy Measured CVD RR (95% CI) Studies that used No data I2 = Partially
os et al., adults weights at mortality = 1.14 (0.97, self-reported 58.2%, well done/
2017 measured baseline 1.35) measures of tau2 = reported
between and follow- weight gain had 0.029.
middle and up. Largest higher HRs (HR = Ratio of
older age weight gain 1.41, 95% CI = HRs =
from 0.97, 2.05. 1.00.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 345

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
No data on baseline to
number of follow-up. Studies with The time
participants Included normal weight or between
both overweight/obese weight
intentional participants gave measurem
and similar HRs to ents (i.e.,
unintention studies that > 10 y or
al weight combined all < 10 y)
gain. participants. The explained
Excluded effect of baseline much of
studies that weight on the
investigated association is heterogen
weight gain unknown. eity.
from early Studies
adulthood Weight gain in with > 10
to middle midlife is y between
age; associated with weight
included increased risk of measurem
studies of CVD mortality. ents had
weight gain higher
from HRs than
middle age studies
to older with < 10
age. y.
Karahali 2 Healthy Intentional All-cause HR = 1.44; Results from No data No data Partially
os et al., adults weight loss mortality 95% CI = weight loss well done/
2017 measured (measured 1.03, 2.00. studies with reported
between and self- measured weights
middle and reported) and including both
older age intentional and

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

346 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
unintentional
No data on weight loss were
number of similar: HR =
participants 1.40 (1.14, 1.71);

Unintentional
weight loss might
reflect an
underlying
disease, resulting
in excess
mortality. Only
two studies had
data on intentional
weight loss.

LeBlanc 9 Males and Behavior- Diabetes RR (95% Weight loss - I2 = Partially

et al., females ≥ based incidence CI): 0.67 interventions 49.2%, p well done/
2018 19 y; high- weight loss in (0.51 to 0.89) associated with a = .46. reported
income prediabet decreased risk of
countries ic type 2 diabetes in The
participa prediabetic consistenc
Included nts participants up to y across
studies with 36 months of interventi
≥ 12 follow-up. ons and
months subgroups
follow-up Behavior-based suggests
and weight loss that
participants interventions were benefits

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 347

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
≥ 18 y with associated with are likely
above more weight loss dependent
normal than controls. on
weight. Weight loss individual
Excluded maintenance , social,
studies with interventions were and
participants associated with environm
with less weight regain ental
chronic than control factors
diseases or conditions over 12 more than
secondary to 18 months interventi
causes of on
obesity. Behavior-based characteri
weight loss stics.
interventions were
associated with
more weight loss
and a lower risk of
developing
diabetes than
control conditions.
Weight loss
medications were
associated with
higher rates of
harms than
behavior-based
interventions.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

348 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
Infrequent
reporting of CVD,
cancer, and all-
cause mortality
precluded
summarizing data
for these
outcomes.
Ma et 24 15,176 Dietary New RR (95% Similar results - I2 = 0%, p Partially
al., 2017 males and weight loss CVD CI): 0.93 when using = .829. well done/
females age ± physical events (0.83, 1.04 ACC/AHA reported
≥ 19 y with activity. All definitions. "New
obesity but 1 of the CVD events" was
diets were a secondary
low fat. outcome.
Follow-up
for ≥ 1 y. In predominantly
middle-aged
adults, the authors
were unable to
show effects of
weight loss on
new CVD events.
There were fewer
trials and much
uncertainty for
this outcome.

Because all but

one study used a

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 349

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
low-fat, weight-
reducing diet, the
results are
relevant only to
this cause of
weight loss.
Ma et 19 6,330 males Dietary New RR (95%CI) "New cancers" - I2 = 0%; p Partially
al., 2017 and females weight loss cancers = 0.92 (0.63, was a secondary = .992. well done/
age ≥ 19 y ± physical 1.36) outcome. reported
with activity. All
obesity but 1 of the In predominantly
diets were middle-aged
low fat. adults, the authors
Follow-up were unable to
for ≥ 1 y. show effects of
weight loss on
new cancer
events. There
were fewer trials
and much
uncertainty for
this outcome.

Because all but

one study used a
low-fat, weight-
reducing diet, the
results are
relevant only to

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

350 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
this cause of
weight loss.
Ma et 34 Males and Dietary All-cause RR (95% CI) In predominantly - I2 = 0%. p Partially
al., 2017 females age weight loss mortality = 0.82 (0.71, middle-aged = .945 well done/
≥ 19 y with ± physical 0.95) adults, weight loss reported
obesity activity. All diets, usually low
but 1 of the in fat and
diets were saturated fat, with
low fat. or without
Follow-up exercise advice or
for ≥ 1 y. programs, may
reduce premature
all-cause mortality
in adults with
obesity.

Because all but

one study used a
low-fat, weight-
reducing diet, the
results are
relevant only to
this cause of
weight loss.
Sun et 6 studies 128,164 Those with Type 2 Compared to Those who Study Heterogen Partially
al., 2021 in meta- males and (1) normal Diabetes normal developed excess quality eity or Not
analysis females, weight in weight in weight in ranged assessed. well-done
from childhood childhood adulthood or were from 6-8 After well-
childhood and and persistently out of 9 subgroup done/repo
to overweight/ adulthood, overweight/obese (moderate analyses rted

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 351

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
adulthood; obese in ORs (95% in childhood and to high by child
mixed adulthood; CI) of adult adulthood had quality) age (< 11
race/ethnici (2) T2D were: increased risk of and > 11
ty; not all overweight/ (1) 3.40 (2.71 T2D. Those with years) and
from high- obese in to 4.25) for excess child adult age
income childhood normal child weight but normal (< 30 and
countries and weight but adult weight had a > 30
adulthood; overweight/o much reduced years);
(3) bese adult increase in risk. definition
Age at overweight/ weight (2) of
baseline obese in 3.94 (3.05 to NOTE: They also childhood
weight childhood 5.08) for assessed a number overweigh
assessment and normal overweight/o of other CVD risk t and
< 20 y weight in bese in factors, including obesity
adulthhood childhood dyslipidemia, (U.S.
and nonalcoholic fatty CDC and
adulthood; liver disease, internatio
(3) 1.37 (1.10 metabolic nal BMI
to 1.70) for syndrome, percentile
overweight/o inflammation, left );
bese in ventricular measured
childhood but hypertrophy, and vs. self-
normal subclinical CVD reported
weight in markers. All weight
adulthood showed increased and
OR in the incident height, the
and persistent heterogen
obesity groups, eity
and most were NS disappear
ed.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

352 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
for resolved
obesity.
Sun et 4 30,309 Those with Hyperten Compared to Incident and Study Heterogen Partially
al., 2021 studies males and (1) normal sion normal persistent quality eity or Not
in meta- females, weight in weight in overweight/obesit ranged assessed. well-done
analysis from childhood childhood y are associated from 6-8 After well-
(vs. 10 childhood and and with increased out of 9 subgroup done/repo
in to overweight/ adulthood, risk of adult (moderate analyses rted
review) adulthood; obese in ORs (95% hypertension. to high by child
mixed adulthood; CI) of adult Resolved obesity quality) age (< 11
race/ethnici (2) hypertension is not. and > 11
ty; not all overweight/ were: (1) years) and
from high- obese in 2.69 (2.07 to adult age
income childhood 3.49) for (< 30 and
countries and normal child > 30
adulthood; weight but years);
(3) overweight/o definition
Age at Overweight bese adult of
baseline / obese in weight (2) childhood
weight childhood 3.49 (2.21 to overweigh
assessment and normal 5.05) for t and
< 20 y weight in overweight/o obesity;
adulthhood bese in measured
childhood vs. self-
and reported
adulthood; weight
(3) 1.25 (0.73 and
to 2.13) for height, the
overweight/o heterogen
bese in eity

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 353

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
childhood but disappear
normal ed.
weight in
adulthood
Sun et 4 studies 87,556 Those with Adult Compared to Incident and - Heterogen Partially
al., 2021 in meta- males and (1) normal cardiovas normal persistent eity or Not
analysis females, weight in cular weight in overweight/obesit assessed. well-done
from childhood disease childhood y are associated After well-
childhood and (CHD, and with increased subgroup done/repo
to overweight/ CVD, adulthood, risk of adult CVD. analyses rted
adulthood; obese in stroke, ORs (95% Resolved obesity by child
mixed adulthood; heart CI) of adult is not. age (< 11
race/ethnici (2) failure) CVD were: and > 11
ty; not all overweight/ (1) 2.76 (1.79 years) and
from high- obese in to 4.27) for adult age
income childhood normal child (< 30 and
countries and weight but > 30
adulthood; overweight/o years);
(3) bese adult definition
Age at overweight/ weight (2) of
baseline obese in 3.04 (1.69– childhood
weight childhood 5.46) for overweigh
assessment and normal overweight/o t and
< 20 y weight in bese in obesity;
adulthhood childhood measured
and vs. self-
adulthood; reported
(3) 1.22 (0.92 weight
to 1.62) for and
overweight/o height, the

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

354 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
bese in heterogen
childhood but eity
normal disappear
weight in ed.
adulthood

Wang et 20 38,141 Weight loss Diagnosi RR = 1.26, Subgroup analysis 12 studies Subgroup Well
al., 2021 males and s of 95% CI 1.15 by baseline BMI were high analyses done/
females ≥ dementia to 1.38 identified that quality conducted reported
19 y; from weight loss in (score of 7- (degree of
United normal weight 9) and 8 weight
States, participants had were loss,
Europe, similar dementia medium dementia
Nigeria, risk (1.21, 95% CI quality (4- subtype,
Australia, 1.06 to 1.38) to 6) diagnostic
South weight loss in criteria for
Korea overweight/obese dementia,
individuals (1.22, country,
1.11 to 1.34). gender,
age,
Weight loss may baseline
be associated with BMI,
increased risk of baseline
dementia. health
Maintaining stable status,
weight may help duration
prevent dementia. of follow-
up, and
Information was adjusted
not available on factors).

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 355

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
whether weight In most
loss was cases,
intentional or not. results
were
consistent
among
subgroups
.
Zhang 15 623,973 Weight All-cause Overall HR Weight fluctuation Newcastle Heterogen Partially
et al., males and fluctuation mortality for group might be scores eity well done/
2019 females ≥ episodes with greatest associated with an ranged assessed reported
19 y; from weight increased risk of from 5-9 by meta-
United fluctuation all-cause (moderate regression
States, (vs. group mortality. to high ,
South with most quality) sensitivity
Korea, stable analyses,
Australia, weight) was and
Germany, 1.45 (95% CI stratified
UK 1.29 to 1.63) analyses
according
to
prespecifi
ed study
characteri
stics.
Overall
conclusio
n was not
changed.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

356 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator

Zou et 20 341,395 Weight All-cause RR = 1.41 Relationship Most Analysis Partially

al., 2019 males and fluctuation mortality (95% CI between weight studies of well done/
females ≥ (studies 1.27, 1.57) fluctuation and were high heterogen reported
19 y varied in all-cause mortality quality eity was
how this did not differ by significant
was BMI or age, or by .
measured) how weight Contributi
fluctuation was ng factors
measured included
(continuous or study
categorical) location,
duration,
Body weight quality,
fluctuation is weight
associated with ascertain
higher all-cause ment
mortality. Future measured
study needed to or self-
determine causal reported,
links. adjustmen
t for
Studies included physical
weight fluctuation activity
measured either as and
categorical energy
(episodes of a intake.
given magnitude)
or continuous
(e.g., intrapersonal

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 357

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
variation of
weight). Most
studies did not
indicate if weight
fluctuation was
intentional or not.
Zou et 11 245,109 Weight CVD RR = 1.36 Relationship 11 of 11 Heterogen Partially
al., 2019 males and fluctuation mortality (95% CI between weight studies eity NS well done/
females ≥ (studies 1.22, 1.52) fluctuation and were high reported
19 y varied in CVD mortality quality
how this was observed in
was those with normal
measured) weight and
overweight, but
not with obesity
or by how weight
fluctuation was
measured
(continuous or
categorical)

Zou et 6 172,709 Weight Cancer RR = 1.01 Body weight 6 of 6 Heterogen Partially

al., 2019 males and fluctuation mortality (95% CI, fluctuation is studies eity NS well done/
females ≥ (studies 0.90, 1.13) NOT associated were high reported
19 y varied in with cancer quality
how this mortality.
was
measured)

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

358 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Predictor or Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characterist Intervention Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies ics or Studies 2 Rating
Comparator
Zou et 5 122,920 Weight CVD RR = 1.49 Body weight 3 of 5 Significan Partially
al., 2019 males and fluctuation morbidit (95% CI, fluctuation is studies t. well done/
females ≥ (studies y 1.26, 1.76) associated with were high Appeared reported
19 y varied in CVD quality to be
how this affected
was by method
measured) of weight
ascertain
ment
Zou et 4 144,256 Weight Hyperten RR = 1.35, Body weight Not Heterogen Partially
al., 2019 males and fluctuation sion 95% CI, fluctuation is reported eity NS well done/
females ≥ (studies 1.14, 1.61 associated with reported
19 y varied in hypertension
how this
was
measured)
NOTE: ACC = American College of Cardiology; AHA = American Heart Association; BMI = body mass index; CHD = coronary heart disease; CI = confidence
interval; CVD = cardiovascular disease; HR = hazard ratio; kg = kilogram; m = meter; OR = odds ratio; RR = relative risk; T2D = type 2 diabetes; y = year.

TABLE J-13 Evidence on the Association Between BMI and Chronic Disease, Including All-Cause Mortality: Systematic Reviews
and Observational Studies
Author, Number Number of Age or Sex BMI Cut Primary Quantitative Clinical Interpretation Risk of Overall
Year of Participants Life Point for Outcome Finding(s) Bias AMSTAR
Studies Stage Risk 2 Rating
Azizpour 16 8,397 1–18 y Females ≥ 25.0 and Asthma Overweight Risk for asthma in P= -
et al., including and ≥ 30.0 1.64 (1.13– children and adolescents 0.312; P
2018 3,577 cases males 2.38); Obese who are overweight or = 0.09
1.92 (1.39– obese is 64–92% higher
2.65) compared to

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 359

Author, Number Number of Age or Sex BMI Cut Primary Quantitative Clinical Interpretation Risk of Overall
Year of Participants Life Point for Outcome Finding(s) Bias AMSTAR
Studies Stage Risk 2 Rating
underweight/normal
weight.
Sharma et 52 1,553,683 5–13 y Females ≥ 85th Child/Ad Prediabetes: Children and adolescents - Partially
al., 2019 and percentile olescent 1.4 (1.2‐1.6); (age 5–13) with well
males prediabet HTN: 4.0 (2.8– overweight or obesity (≥ done/repo
es, HTN, 5.7); NAFLD: 85th percentile) are 1.4 rted
NAFLD 26.1 (9.4–72.2) times more likely to
have prediabetes, those
with obesity are 4.4
times more likely to
have high blood
pressure, and 26.1 times
more likely to have
NAFLD.
Hidayat et 6 13,510 Pregna Females ≥ 25.0 Child- Overweight Each 5kg/m2 increase in P = 0.23 -
al., 2019 cases ncy onset 1.09 (1.03– maternal BMI associated
T1DM 1.15); Obese with 10% increased risk
1.25 (1.16– for child-onset T1DM.
1.34) Association was
nonlinear, with steeper
increase in risk at BMI ≥
26.0
Xiao et 103 1,826,454 Prepre Females ≥ 25.0 Gestation 2.64 (1.56– Prepregnancy - Partially
al., 2021 including gnancy al 4.45) overweight or obesity well
120,696 diabetes increases risk 2.64-fold done/repo
cases for having gestational rted
diabetes.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

360 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Number of Age or Sex BMI Cut Primary Quantitative Clinical Interpretation Risk of Overall
Year of Participants Life Point for Outcome Finding(s) Bias AMSTAR
Studies Stage Risk 2 Rating
Ibe et al., BRFSS 1,168,418 18–64 Females ≥ 25.0 T2DM 3.57 (3.52– Adjusting for age, race, - -
2014 (Behavio 3.63) physical activity, and
ral Risk year of survey response,
Factor results indicate a 3.5-
Surveilla fold increase in diabetes
nce in females with BMI >
System) 25.
Jayedi et 182 5,585,850 > 18 Females > 20 T2DM 1.72 (1.65– Each 5 kg/m2 increase in - -
al., 2022 including and 1.81) BMI above 20.0
228,695 males associated with 72%
cases increased risk for
T2DM, with steep
upward curve at BMI >
25 in younger adults.
Khadra et 11 60,118 19–50 Females ≥ 25.0 T2DM 1.38 (1.27– Sarcopenic obesity is - -
al., 2019 and 1.50) associated with a 38%
males increased risk for T2DM
compared to
nonsarcopenic obesity
Larsson et 47 218,792 > 18 Females ≥ 25.0 T2DM 2.03 (1.88– Mendelian - -
al., 2021 and 2.19) randomization
males (genetically predicted)
studies show high adult
BMI is a causal risk
factor for T2DM, with a
2-fold increased risk for
T2DM when BMI ≥ 25.
Yu et al., 82 2,690,000 > 18 Females ≥ 25.0 Prediabet Prediabetes Overweight and obesity - Partially
2022 and es, overweight and are associated with a well
males T2DM obesity: 1.24 24% increased risk for done/repo
(1.19–1.28); prediabetes. Overweight rted

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 361

Author, Number Number of Age or Sex BMI Cut Primary Quantitative Clinical Interpretation Risk of Overall
Year of Participants Life Point for Outcome Finding(s) Bias AMSTAR
Studies Stage Risk 2 Rating
T2DM is associated with a 2-
Overweight: fold increased risk and
2.24 (1.95– obesity a 4.5-fold
2.56); Obese: increased risk for
4.56 (3.69– T2DM.
5.64)
Jayedi et 50 2,255,067 > 18 Females > 20 HTN 1.49 (1.41– Each 5 kg/m2 increase in 0.0001 -
al., 2018 including and 1.58) BMI above 20.0
190,320 males associated with 49%
cases increased risk for HTN
Zhou et 57 830,685 > 18 Females HTN BMI 18.5: 1.27 Risk for HTN increases - Partially
al., 2018 including and (1.20–1.35), at least 50% for every 5- well
125,071 males BMI 25.0: 2.07 unit increase in BMI. done/repo
cases (1.34–2.46), rted
BMI 30: 3.13
(2.49–3.93)
Rexrode Physicia 16,164 40–84 Males ≥ 27.6 CHD 1.73 (1.29– Males with BMI ≥ 27.6 - -
et al., ns including 2.32) have a 73% increased
2001 Health 552 cases risk for a coronary heart
Study disease event.
Kim et al., Framing 1,882 30–62 Males ≥ 23.8 CHD 1.28 (1.00– In males, the relative risk - -
2000 ham 1.65) for CHD is 28% at BMI
Heart ≥ 23.8, 45% at BMI ≥
Study 25.9 and 53% at BMI ≥
28.2
Kim et al., Framing 2,373 30–62 Females ≥ 27.6 CHD 1.56 (1.16– In females with BMI ≥ - -
2000 ham 2.08) 27.6, there is a 56%
Heart increased risk for
Study developing CHD.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

362 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Number of Age or Sex BMI Cut Primary Quantitative Clinical Interpretation Risk of Overall
Year of Participants Life Point for Outcome Finding(s) Bias AMSTAR
Studies Stage Risk 2 Rating
Liu et al., 43 4,432,475 > 18 Females > 23.5 Stroke 1.10 (1.06– Risk of stroke increases P = 0.06 Well
2018a including and 1.13) by 10% for every 5-unit done/repo
102,466 males increase in BMI > 23.5, rted
cases and is greater for males
than females.
Dugani et 16 12,700,000 > 18 Females ≥ 25.0 and Prematur Males 1.94 Males in overweight or - -
al., 2021 (18–65) ≥ 30.0 e MI (1.47–2.56); obese BMI categories
and Females 1.28 have almost a 2-fold
males (0.95–1.73) increased risk for
(18–55) premature MI.
Meigs et Commu 2,902 > 18 Females ≥ 25.0 CVD Overweight: Adults with - -
al., 2006 nity and 3.01 (1.68– overweight/obesity have
Longitud males 5.41) a 3-fold increased risk
inal for CVD.
Study
Darbandi 38 137,256 > 18 Females ≥ 30.0 CVD BMI: AUC BMI, WC, and WHR Ps < -
et al., and 0.66 (0.63– have moderate power to 0.001
2020 males 0.69); WC: identify risk for CVD. In
AUC 0.69 adults, WC and WHR
(0.64–0.74); predict CVD better than
WHR: AUC BMI.
0.69 (0.66–
0.73) males,
0.71 (0.68 =
0.73) females
Kim et al., 77 30,000,000 > 18 Females > 20 CVD 1.10 (1.01– Mendelian - -
2021 and 1.210 for randomization
males hemorrhagic (genetically predicted)
stroke; 1.49 studies show high BMI
(1.40–1.60) for is a causal risk factor for
HTN CVD outcomes; each 5

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 363

Author, Number Number of Age or Sex BMI Cut Primary Quantitative Clinical Interpretation Risk of Overall
Year of Participants Life Point for Outcome Finding(s) Bias AMSTAR
Studies Stage Risk 2 Rating
kg/m2 increase in BMI
increases risk for CVD
events.
Church et Aerobics 2,316 > 20 Males ≥ 25.0 CVD 2.70 (1.40– Overweight and obese - -
al., 2005 Center with mortality 5.10) males with diabetes have
Longitud T2DM similar 2.7-fold
inal increased risk for CVD-
Study mortality.
Jarvis et 14 1,930,000 > 18 Females > 30.0 NAFLD 1.20 (1.12– BMI > 30 is associated - -
al., 2020 including and 1.28) with 20% increased risk
49,451 males for severe liver disease.
cases
Campbell 14 1,570,000 > 18 Females ≥ 25.0 Hepatoce 1.21 (1.09– Compared with normal - -
et al., including and llular 1.35) weight BMI, persons
2016 2,162 cases males carcinom with overweight, class I
a obesity, class II obesity,
and class III obesity
were associated with
21%, 87%, 142%, and
116% increased risk of
liver cancer.
Sohn et 28 8,135,906 > 18 Females ≥ 25.0 Hepatoce 1.69 (1.50– Risk for liver cancer - Well
al., 2021 and llular 1.90) increases in a BMI- done/repo
males carcinom dependent manner with a rted
a 36% increased risk for
BMI > 25, 77%
increased risk for BMI >
30, a 3-fold increased
risk for BMI > 35 (and a
70% increased risk
overall for BMI ≥ 25.0).

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

364 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Number of Age or Sex BMI Cut Primary Quantitative Clinical Interpretation Risk of Overall
Year of Participants Life Point for Outcome Finding(s) Bias AMSTAR
Studies Stage Risk 2 Rating
Byun et 37 1,849,875 ≤ 30 Females 13.2–32.5 Breast 0.84 (0.81– Each 5 kg/m2 increase in P< -
al., 2022 including cancer 0.87) early-life BMI 0.001
39,733 (premeno associated with 16%
cases pausal) reduced premenopausal
breast cancer risk
Byun et 10 662,779 ≤ 30 Females 15.3–32.5 Endomet 1.40 (1.25– Each 5 kg/m2 in BMI P< -
al., 2022 including rial 1.57) increase in early life (age 0.001
4,539 cases cancer ≤ 25 y) BMI associated
with 1.4-fold increased
endometrial cancer risk
Byun et 6 496,391 ≤ 30 Females 14.6–32.5 Ovarian 1.15 (1.07– Each 5 kg/m2 in BMI P< -
al., 2022 including cancer 1.23) increase in early life (age 0.001
2,692 cases ≤ 25 y) BMI associated
with 15% increased risk
for ovarian cancer
Fang et 325 1,525,052 > 18 Females > 20.0 Cancer Endometrial: Every 5 kg/m2 increase - -
al., 2018 and (23 tissue 1.48 in BMI is associated
males types) with increased risk for
18 types of tissue
cancers. The strongest
positive association is
between BMI and
endometrial cancer (RR
= 1.48). BMI was
negatively associated
with the risk of oral
cavity, lung, and
premenopausal breast
cancers.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 365

Author, Number Number of Age or Sex BMI Cut Primary Quantitative Clinical Interpretation Risk of Overall
Year of Participants Life Point for Outcome Finding(s) Bias AMSTAR
Studies Stage Risk 2 Rating
Gao et al., 27 28,784,269 > 18 Females ≥ 25.0 Lung BMI: 0.77 BMI is inversely P= -
2019 including and cancer (0.72–0.82); associated with lung 0.005
127,161 males WC: 1.24 cancer risk. When
cases (1.13–1.35) controlling for BMI,
high waist circumference
associates with lung
cancer risk.
Gu et al., 52 279,499 > 18 Males ≥ 25.0 Prostate 0.99 (0.99– Higher BMI associated - -
2022 including cancer 1.00) with 1% decreased risk
51,704 for localized prostate
cases cancer.
Hidayat et 56 56,744 ≤ 30 Females ≥ 20.0 Cancer (8 Each 5kg/m2 increase in - -
al., 2018a and types) BMI in early life (≤ 30
males y) associated with 1.88-
fold increased risk for
esophageal cancer, 1.31-
fold increased risk for
liver cancer, 1.17-fold
increased risk for
pancreatic cancer, 1.59-
fold increased risk for
gastric cancer, 1.22-fold
for kidney cancer, and
1.45-fold increased risk
for endometrial cancer.
Hidayat et 22 7,000,000 > 18 Females ≥ 20.0 Non- 1.13 (1.06– Each 5 kg/m2 increase in - -
al., 2018b including and Hodgkin’ 1.20) BMI associated with 6%
20,000 males s increased risk for NHL,
cases lymphom with no difference by
a sex. Further, each 5
kg/m2 increase in BMI in

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

366 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Number of Age or Sex BMI Cut Primary Quantitative Clinical Interpretation Risk of Overall
Year of Participants Life Point for Outcome Finding(s) Bias AMSTAR
Studies Stage Risk 2 Rating
early adulthood (18–21
y) associated with 11%
increased risk for NHL.
Li et al., 12 5,902 cases > 18 Females ≥ 25.0 Gallblad Overweight: The pooled risk for - -
2016 and der 1.10 (0.98– gallbladder cancer at
males cancer 1.23); Obese BMI ≥ 25 for overweight
1.58 (1.43– is 10% and obesity 58%,
1.75) and risk increases by 4%
for each 1-unit increase
in BMI.
O’Sulliva 20 47,692 ≤ 50 Females ≥ 30.0 Colorecta Obese: 1.54 Obesity (BMI ≥ 30 - Well
n et al., cases and l (1.01 – 2.35) kg/m2 ) is associated done/repo
2021 males cancer— with a 54% increased rted
early risk of early onset (≤ 50
onset y) colorectal cancer,
with males at higher risk
than females.
Li et al., 6 8,150,473 ≤ 55 Females ≥ 25.0 Colorecta Overweight Overweight and obesity P = 0.60 -
2021 including and l 1.32 (1.19– (BMI ≥ 25 kg/m2) are
11,299 males cancer— 1.47); Obese associated with a 42%
cases early 1.88 (1.40– increased risk of early
onset 2.54) onset (age ≤ 55)
colorectal cancer.
Liu et al., 24 8,953,478 > 18 Females > 20 Kidney Overweight: Risk of kidney cancer - Well
2018b including and cancer RR 1.35 (1.27– increases 6% for every done/repo
15,535 males 1.43); Obese 1-unit increase in BMI > rted
cases RR 1.76 (1.61– 20.
1.91)
Youssef et 31 24,489,477 > 18 Females < 18.5, ≥ Thyroid Underweight: Overweight and obesity - Not well
al., 2021 including and 25.0 cancer 0.68 (0.65– are associated with a done/repo
males 0.72); 26% and 50% increased rted

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 367

Author, Number Number of Age or Sex BMI Cut Primary Quantitative Clinical Interpretation Risk of Overall
Year of Participants Life Point for Outcome Finding(s) Bias AMSTAR
Studies Stage Risk 2 Rating
86,097 Overweight: risk of thyroid cancer,
cases 1.26 (1.24– with risk greater in
1.28); Obese: females than males.
1.50 (1.45– Having an underweight
1.55) BMI decreases risk by
32%.
Jiang et 9 96,213 ≥ 65 Females > 28 Disabilit 1.19 (1.01– BMI 24.0–28.0 - -
al., 2019 and y 1.40) decreases risk by 4% for
males disability in adults age ≥
65 years, but BMI > 28
increases disability risk
by 19%.
Mortense 35 1,508,366 > 50 Females < 18.5 Fragility 2.83 (1.82– BMI under 18.5 is Partially
n et al., and hip 4.39) associated with almost a well
2021 males fracture 3-fold increased risk for done/repo
fragility hip fracture, rted
where as BMI over 30
may be protective.
Jiang et 37 320,594 ≥ 65 Females < 23 and > All-cause BMI < 18.5: BMI < 23.0 and > 33.0 - -
al., 2019 and 33.0 mortality 1.69 (1.57– increase risk for all-
males 1.83); BMI cause mortality in adults
18.5–22.9: 1.17 ≥ 65 years
(1.12–1.22);
BMI 23.0–
27.9: 0.91
(0.88–0.94);
BMI 28.0–
32.9: 0.98
(0.94–1.03);
BMI > 33.0:

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

368 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Number of Age or Sex BMI Cut Primary Quantitative Clinical Interpretation Risk of Overall
Year of Participants Life Point for Outcome Finding(s) Bias AMSTAR
Studies Stage Risk 2 Rating
1.32 (1.15–
1.51)
Kitahara 20 9,564 > 18 Females Class III All-cause BMI 40–59: Adults with BMI 40–49 - -
et al., and obesity mortality 1.40 (1.31– have a 2.3- to 3.3-fold
2014 males 1.51) increased risk for death,
those with BMI 50–59
have a 3.5 to 5.9
increased risk for death,
and risks are greater for
males vs. females.
NOTE: AUC = area under the curve; BMI = body mass index; BRFSS = Behavioral Risk Factor Surveillance System; CHD = coronary heart disease; CVD =
cardiovascular disease; HTN = hypertension; kg = kilogram; m = meter; MI = myocardial infarction; NAFLD = nonalcoholic fatty liver disease; RR = relative
risk; NHL = non-Hodgkin’s lymphoma; T1DM = type 1 diabetes mellitus; T2DM = type 2 diabetes mellitus; WC = waist circumference; WHR = waist–hip ratio;
y = year.

TABLE J-14 Evidence on the Degree of Systematic Bias or Random Error of Energy Intake as Assessed by Self-Report Compared to
Doubly Labeled Water Studies: Systematic Reviews
Author, Number Sample Intervention / Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characteristics Comparator Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies Studies 2 Rating
Burrows 36 2,834 male and Food EI-TEE Most studies found The food record is 29/36 - Partially
et al., female adults record/TEE underreporting by likely to positive well done/
2019 including from DLW 11–41% significantly quality; reported
pregnant underreport EI 7/36 neutral
women; not all when compared to quality
high-income TEE measured via
countries the DLW method.
Burrows 24 3,295 male and 24-hour EI-TEE EI underreported by EI tends to be 16/24 - Partially
et al., female adults recall/TEE 8–30% in almost all underreported on positive; well done/
2019 including from DLW studies 24-hour recalls. 8/24 neutral reported
pregnant

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 369

Author, Number Sample Intervention / Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characteristics Comparator Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies Studies 2 Rating
women; not all
high-income
countries
Burrows 21 2,997 male and FFQ/TEE EI-TEE Significant FFQs tend to 14/21 - Partially
et al., female adults from DLW underreporting underestimate positive; well done/
2019 including found in all studies energy intake, 7/21 neutral reported
pregnant using an FFQ particularly at the
women; not all individual level.
high-income
countries
Burrows 5 71 male and Diet EI-TEE Underreporting in 4 Diet histories tend 4/5 - Partially
et al., female adults history/TEE of 5 studies ranging to underreport EI. positive; well done/
2019 including from DLW from 1 to 47% 1/5 neutral reported
pregnant
women; not all
high-income
countries
Burrows 5 106 male and FFQ/TEE EI-TEE Significant FFQ has limitations 4/5 positive - Partially
et al., female children from DLW underreporting in 3 for assessing EI, quality; 1 well done/
2020 and adolescents of 5 studies (-7% to especially at the neutral reported
-23% of estimated individual level. quality
EI); other 2 studies
were small (n = 9 or
12), one had a
higher mean EI on
FFQ vs. TEE from
DLW, the other was
lower

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

370 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Intervention / Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characteristics Comparator Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies Studies 2 Rating
Burrows 4 66 male and Weighted EI-TEE Significant Only 1 study 4/4 positive - Partially
et al., female children food underreporting in 1 concluded the tool well done/
2020 and adolescents record/TEE of 4 studies (-10% may be useful in reported
from DLW of estimated EI) individual children;
it may not be
accurate at the
individual level.
Burrows 3 108 male and Remote food EI-TEE Differences ranged There is limited - - Partially
et al., female children photography/ from -16% to +7%. ability to assess EI well done/
2020 and adolescents TEE from One study found no at the individual reported
DLW significant level.
difference between
reported and
measured values;
one found remote
food photography
method was not
valid at the
individual or group
level.
Burrows 2 52 male and 24-hour EI-TEE One study found a The 24-hour recall 1/2 - Partially
et al., female children recall /TEE difference of -23 was valid on the positive; well done/
2020 and adolescents from DLW (+/- 442 kcal); the group level, but not 1/2 neutral reported
second found a at the individual
difference of -0.9% level.
Burrows 1 29 male and Precoded EI-TEE Overreporting by Method 1/1 positive - Partially
et al., female children food +24% (p < .0001); overestimated EI. well done/
2020 and adolescents record/TEE mean difference of reported
from DLW 726 kJ/day

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 371

Author, Number Sample Intervention / Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characteristics Comparator Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies Studies 2 Rating
Capling et 11 109 adolescent Food record/ EI-TEE Mean difference EI- The food record is fair to - Not well
al., 2017 and adult male DLW TEE: -19%; -2,793 likely to moderate done/
and female +/- 1,134 kJ/day significantly for most reported
athletes; absolute difference; underreport studies
includes Effect size -1.01 estimated EI when
pregnant (95% CI, -1.3, -0.7) compared with TEE
women; not all estimated via DLW
from high- in athletes.
income countries
Gemming 2 82 male and Image-based EI-TEE Remote food Image-based food - - Not well
et al., female adults; food record photography records are likely to done/
2015 not all from /TEE from underestimated by - underestimate EI. reported
high-income DLW 6% to -26% in
countries overweight and
obese adults
Gemming 1 14 male and Image- EI-TEE Image-assisted 24- Image-assisted - - Not well
et al., female adults; assisted 24- hour recall methods may done/
2015 not all from hour recall overestimated by overestimate EI. reported
high-income /TEE from +7.6%
countries DLW
Ho et al., 6 205 children and Image-based Total Four studies A large weighted The overall Heterogen Partially
2020 adults, males dietary energy reported a lower mean difference in quality of eity well-
and females; assessment intake mean EI as energy intake the 6 between done/repo
includes method/DLW estimated by the showed significant studies studies rted
pregnant women IBDA method; two energy ranged was very
studies reported underreporting on from good high (I2 =
agreement and no the IBDA methods, to very 95%),
bias between the when compared good. 2 indicating
IBDA and DLW. with DLW. studies substantia
The weighted mean were rated l
difference for IBDA as very variability

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

372 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Intervention / Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characteristics Comparator Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies Studies 2 Rating
and DLW methods good with between
was -448.04 kcal (- 9-10 pts., studies.
755.52, -140.56), and 4
but heterogeneity studies
between studies was were rated
very high (I2 = as good
95%), indicating quality with
substantial 7-8 pts.
variability between
studies.
Ho et al., 4 142 children and Image-based Total One study showed a No statistically The overall Heterogen Partially
2020 adults, males dietary energy significant positive significant quality of eity was well-
and females; assessment intake correlation for EI differences were the 4 high (I2 = done/repo
includes method/24- between the IBDA found in the studies 76%), rted
pregnant women hour dietary and 24-hour weighted mean ranged indicating
recall methods, another differences of from good some
study showed that energy intake to very variability
the IBDA method between the IBDAs good. 1 between
under-reported EI and the 24-hour study was studies.
when compared recalls. rated as
with the 24-hour very good
method, and the with 9-10
other two studies pts., and 3
provided mean studies
estimates but not were rated
statistical analyses. as good
Weighted mean quality with
difference in EI for 7-8 pts.
IBDAs and 24-hour
recalls was -91.53
kcal (-151.45,

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 373

Author, Number Sample Intervention / Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characteristics Comparator Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies Studies 2 Rating
46.13);
Heterogeneity was
high (I2 = 76%),
indicating some
variability between
studies.
Ho et al., 6 266 children and Image-based Total Three studies No statistically The overall Heterogen Partially
2020 adults, males dietary energy reported good significant quality of eity was well-
and females; assessment intake agreement in differences were the 6 high (I2 = done/repo
includes method/weig estimated EI, two found in the studies 66%), rted
pregnant women hted food studies reported an weighted mean ranged indicating
record underestimation of differences of from good some
EI using the IBDA energy intake to very variability
methods, and one between the IBDAs good. 2 between
study reported an and the WFRs. studies studies.
overestimation of were rated
EI using the IBDA as very
method. Weighted good with
mean difference in 9-10 pts.,
EI for IBDA and and 4
WFR was -52.66 studies
kcal (-151.45, were rated
46.13); as good
Heterogeneity was quality with
high (I2 = 66%), 7-8 pts.
indicating some
variability between
studies.
Ho et al., 3 103 children and Image-based Macronut One study showed a No statistically The overall Heterogen Partially
2020 adults, males dietary rients significant positive significant quality of eity was well-
and females; assessment correlation for all differences in the the 3 high (I2 =

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

374 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Intervention / Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characteristics Comparator Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies Studies 2 Rating
includes method/24- three weighted mean studies 66%) for done/repo
pregnant women hour dietary macronutrients, one difference of ranged carbohydr rted
recall study observed a carbohydrate, from good ate intake,
significant protein, or fat intake to very indicating
difference in were observed good. 1 some
carbohydrate but between the IBDA study was variability
not protein or fat and 24HR methods. rated as between
intake, and the other very good studies,
study provided with 9-10 but was
mean estimates but pts., and 2 not
not statistical studies present
analyses. WMD in were rated for protein
carbohydrate intake as good (I2 = 0%;
was -15.52 g (95% quality with p = .95) or
CI: -41.34, 10.30); 7-8 pts. fat intake
heterogeneity was I2 (I2 = 0%;
= 66% (p = .05). p = .44).
WMD in protein
intake was 2.06 g (-
3.16, 7.28);
heterogeneity was I2
= 0% (p = .95).
WMD in fat intake
was -2.90 g (-8.34,
2.55); heterogeneity
was I2 = 0% (p =
.44).
Ho et al., 6 256 children and Image-based Macronut Three studies No statistically The overall Heterogen Partially
2020 adults, males dietary rients reported good significant quality of eity was well-
and females; assessment agreement in differences in the the 6 moderate done/repo
method/weig estimated weighted mean studies to high for rted

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 375

Author, Number Sample Intervention / Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characteristics Comparator Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies Studies 2 Rating
includes hted food macronutrients difference of ranged carbohydr
pregnant women record between the two carbohydrate, from good ate (I2 =
methods, two protein, or fat intake to very 63%; p =
studies reported no were observed good. 2 .02) and
difference in between the IBDA study was protein
macronutrient and WFR methods. rated as intake (I2
intake between the very good = 77%; p
IBDA and WFR, with 9-10 < .01), but
and one study pts., and 4 low for fat
reported that the studies intake (I2
IBDA were rated = 21%; p
overestimated as good = .28).
carbohydrate, quality with
protein, and fat 7-8 pts.
intake. Weighted
mean difference
(WMD) in
carbohydrate intake
for IBDAs and
WFRs was -6.71 g
(-20.2, 6.79);
heterogeneity was I2
= 63% (p = 0.02).
WMD in protein
intake for IBDAs
and WFRs was -
0.85 g (-6.10, 4.40);
heterogeneity was
high (I2 = 77%).
WMD in fat intake
for IBDAs and

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

376 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Intervention / Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characteristics Comparator Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies Studies 2 Rating
WFRs was -0.30 g
(-2.65, 2.05);
heterogeneity was
low (I2 = 21%; p =
.28).
Ho et al., 2 53 children and Image-based Micronut One study showed a No statistically 1 study was Heterogen Partially
2020 adults, males dietary rients significant positive significant rated as eity was well-
and females; assessment correlation with differences were very good not done/repo
includes method/24- iron and vitamin C, found in the WMDs with 9-10 present rted
pregnant women hour dietary and the other study of iron or vitamin C pts., and 1 for iron
recall provided mean intake. study was (I2 = 0%;
estimates but not rated as p = .38) or
statistical analyses. good vitamin C
Weighted mean quality with intake (I2
difference in iron 7-8 pts. = 0%; p =
intake for IBDAs .56).
and 24HRS was
0.39 mg (95% CI: -
0.81, 1.59);
heterogeneity was I2
= 0% (p = .38).
WMD in vitamin C
intake was 9.14 mg
(-13.16, 31.45);
heterogeneity was I2
= 0% (p = .56).
Ho et al., 3 152 children and Image-based Micronut One study showed a No statistically The overall Heterogen Partially
2020 adults, males dietary rients significant positive significant quality of eity was well-
and females; assessment correlation with differences were the 3 minimal done/repo
includes method/weig iron and vitamin C found in the WMDs studies for iron rted
pregnant women for the IBDA and ranged intake (I2

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 377

Author, Number Sample Intervention / Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characteristics Comparator Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies Studies 2 Rating
hted food the WFR, another of iron or vitamin C from good = 3%; p =
record study showed a intake. to very 0.36), but
significant positive good. 1 quite
correlation with study was substantia
vitamin C, and the rated as l for
other study showed very good vitamin C
no difference in with 9-10 intake (I2
micronutrient intake pts., and 2 = 89%; p
(both iron and studies < .01).
vitamin C) between were rated
the two methods. as good
The WMD in iron quality with
intake was -0.19 g 7-8 pts.
(95% CI: -0.78,
0.40); heterogeneity
was I2 = 3% (p =
.36). The WMD in
vitamin C intake
was -10.97 g (-
39.95, 18.01);
heterogeneity was I2
= 89% (p < .01).
Tugault- 15 2,576 school- School meal Relative Poor accuracy for The relative - - Not well-
Lafleur et aged children recalls/observ accuracy individual foods accuracy of school done/repo
al., 2017 ational reported (omission meal recalls is poor rted
method (i.e., and intrusion rates for individual foods
in-person > 15%, n = 8 of 12 reported but is
meal studies). Acceptable acceptable for
observations, accuracy when reporting the
digital reporting amounts estimated energy
photography, consumed (n = 4 of intake of a group.

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

378 DIETARY REFERENCE INTAKES FOR ENERGY

Author, Number Sample Intervention / Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characteristics Comparator Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies Studies 2 Rating
weighted 5 studies).
food records Acceptable energy
report rates (n = 2
of 3 studies).
Tugault- 1 24 school-aged Estimated Relative Pearson correlations The estimated food - - Not well-
Lafleur et children food accuracy ranged from r = record had done/repo
al., 2017 records/obser 0.16 to 0.85 for acceptable accuracy rted
vational different meal with daily
method (i.e., components (mean monitoring but poor
in-person r = 0.66) under a accuracy with
meal daily monitoring weekly monitoring.
observation approach. For the
weekly monitoring
approach, Pearson
correlation
coefficients ranged
from r = -0.21 to
0.69 (mean r =
0.25)
Tugault- 1 46 school-aged FFQs/4-day Relative The Pearson Acceptable - - Not well-
Lafleur et children estimated accuracy correlation accuracy for done/repo
al., 2017 food record coefficients were r measuring select rted
= 0.71, 0.70, an beverages and
0.69 for beverages, snack foods; the
snacks, and total majority of the 19
fruits and questions assessing
vegetables, in-school dietary
respectively. Mean intakes were
r = 0.69 for all food significantly
and beverage items; associated with
p < .05. amounts obtained

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.

Dietary Reference Intakes for Energy

APPENDIX J 379

Author, Number Sample Intervention / Primary Quantitative Qualitative Risk of Hetero- Overall
Year of Characteristics Comparator Outcome Finding(s) Finding(s) Bias geneity of AMSTAR
Studies Studies 2 Rating
from the estimated
food record.
Tugault- 2 1,149 school- Digital Relative In the first study, The findings from - - Not well-
Lafleur et aged children photography accuracy correlation the two studies done/repo
al., 2017 (DP) coefficients suggest that the DP rted
methods/wei indicated strong method is a valid
ghted food positive method for
recalls correlations, estimating the
ranging from 0.89 dietary intakes, in
to 0.97, and no terms of the types
statistically and amounts of
significant foods consumed, of
differences were both home-packed
found in mean and school lunches.
amounts for
differences in lunch
meal components
estimated by using
the DP and the
weighted FRs.
Bland-Altman
analyses suggested
a tendency to
slightly
underestimate fruit
(mean bias: -4.27g)
and vegetables
(mean bias:-6.19g).
In the second study,
all 11 school meal
items had a

PREPUBLICATION COPY: UNCORRECTED PROOFS

Copyright National Academy of Sciences. All rights reserved.