You are on page 1of 54


The Science of Obesity Management: An Endocrine

Society Scientific Statement
George A. Bray,1 William E. Heisel,2 Ashkan Afshin,2 Michael D. Jensen,3 William H. Dietz,4 AFFILIATIONS
1Department of Clinical Obesity,
Michael Long,4 Robert F. Kushner,5 Stephen R. Daniels,6 Thomas A. Wadden,7 Adam G. Tsai,8 Pennington Biomedical Research
Frank B. Hu,9 John M. Jakicic,10 Donna H. Ryan,1 Bruce M. Wolfe,11 and Thomas H. Inge12,13 Center, Louisiana State University,
Baton Rouge, Louisiana 70808
2Institute of Health Metrics

and Evaluation University of

ABSTRACT The prevalence of obesity, measured by body mass index, has risen to unacceptable levels in both Washington, Seattle, Washington
men and women in the United States and worldwide with resultant hazardous health implications. Genetic, 3Department of Medicine, Mayo

environmental, and behavioral factors influence the development of obesity, and both the general public and Clinic, Rochester, Minnesota
health professionals stigmatize those who suffer from the disease. Obesity is associated with and contributes to 55905
4Redstone Global Center for
a shortened life span, type 2 diabetes mellitus, cardiovascular disease, some cancers, kidney disease, obstructive Prevention and Wellness, Milken
sleep apnea, gout, osteoarthritis, and hepatobiliary disease, among others. Weight loss reduces all of these Institute School of Public Health,
diseases in a dose-related manner—the more weight lost, the better the outcome. The phenotype of George Washington University,
Washington, District of Columbia
“medically healthy obesity” appears to be a transient state that progresses over time to an unhealthy
phenotype, especially in children and adolescents. Weight loss is best achieved by reducing energy intake and 5Northwestern Feinberg School of

increasing energy expenditure. Programs that are effective for weight loss include peer-reviewed and approved Medicine, Chicago, Illinois 60611
6Department of Pediatrics,
lifestyle modification programs, diets, commercial weight-loss programs, exercise programs, medications, and
University of Colorado Children
surgery. Over-the-counter herbal preparations that some patients use to treat obesity have limited, if any, data Hospital, Denver, Colorado 80218
documenting their efficacy or safety, and there are few regulatory requirements. Weight regain is expected in 7Department of Psychiatry,

all patients, especially when treatment is discontinued. When making treatment decisions, clinicians should University of Pennsylvania
Perelman School of Medicine,
consider body fat distribution and individual health risks in addition to body mass index. (Endocrine Reviews Philadelphia, Pennsylvania 19104
39: 1 – 54, 2018) 8Kaiser Permanente Colorado,

Denver, Colorado 80111

9Department of Nutrition and

Epidemiology, Harvard T.H. Chan

School of Public Health, Boston,
Massachusetts 02115
INTRODUCTION Strategies for preventing obesity aimed Diet in managing obesity—food is more
10University of Pittsburgh,
What’s past is prologue at the entire population than calories
Pittsburgh, Pennsylvania 15260
Headwinds in the management of Food Introduction
11Oregon Health and Science
obesity Increasing physical activity Very low–calorie diets
University, Portland, Oregon
Introduction Obesity and risk of death low-carbohydrate diets, and
12Department of Surgery,
Limitations of the BMI The mechanism of obesity-associated sugar-sweetened beverages
University of Colorado Denver,
Adults morbidity Dietary fat, energy density, and
Aurora, Colorado 80045
Children Type 2 diabetes mellitus low-fat diets
Children’s Hospital Colorado,
Are there metabolically healthy obese Cancer Low–glycemic index diets
Aurora, Colorado 80045
people? Myocardial infarction High-protein diets
Adults Heart failure and the obesity paradox Mediterranean-style diets
Children Hypertension and stroke Balanced-deficit diets
Age and obesity Obstructive sleep apnea Comparison of diets with different
Adults Hepatobiliary disease macronutrient composition ISSN Print: 0163-769X
Children Gallbladder disease Commercial programs for weight loss ISSN Online: 1945-7189
PREVENTION OF OBESITY Nonalcoholic fatty liver disease Maintenance of long-term weight loss Printed: in USA
Strategies for preventing obesity in Gout and osteoarthritis Future considerations/summary Copyright © 2018
pregnancy Gout Exercise in managing obesity Endocrine Society
Strategies aimed at children Osteoarthritis Introduction Received: 1 December 2017
Early care and education Effects of obesity during pregnancy Genetic factors of physical activity Accepted: 2 December 2017
Schools DIET, EXERCISE, AND LIFESTYLE IN Resistance vs aerobic exercise First Published Online:
Strategies aimed at adults: worksites MANAGING OBESITY Vigorous vs moderate exercise 6 March 2018

doi: 10.1210/er.2017-00253 1

Downloaded from
by guest
on 17 March 2018

Physical activity declines with age Lorcaserin SURGERY IN MANAGING OBESITY

Sedentary behavior Liraglutide Introduction
Treatment of patients who are PHEN/TPM ER Sleeve gastrectomy
overweight or obese using exercise Naltrexone/bupropion combination Roux-en-Y gastric bypass
with and without diet Comparison of medications approved Laparoscopic adjustable gastric banding
BEHAVIORAL THERAPY IN MANAGING for chronic weight management Biliopancreatic diversion with (or
OBESITY Drugs approved by the FDA for without) duodenal switch
Lifestyle methods short-term treatment of patients with Vagal blockade
Self-monitoring obesity Gastrointestinal endoscopic
Stimulus control Phentermine interventions or devices
Goal setting Efficacy of phentermine Liposuction
Problem solving Safety of phentermine Indications for bariatric surgery
Short-term efficacy Best practices for medications approved Criteria for bariatric surgery
New developments in the delivery of for weight management Preoperative assessment
behavioral treatment Medicating the patient for other Outcomes of bariatric surgery
Telephone-delivered programs chronic conditions who is also Safety
Digitally-delivered programs overweight or obese Weight loss
MEDICATION IN MANAGING OBESITY What is the current status of clinical Related outcomes/remission of T2DM
Early history adoption of medications for chronic Bariatric/metabolic surgery in
More recent drug development: weight management? adolescents
continuing difficulties Dietary supplements, over-the-counter WHERE DO WE GO FROM HERE?
Medications approved by the FDA for products, and other treatments with
treating obesity unproven efficacy and unknown safety
Orlistat Drug targets

Introduction the Fat” of obesity using theriac—a mixture of more

than  ingredients ().
What’s past is prologue The first English language texts dealing with

S ome would say that the obesity epidemic began

in the s, but history provides a broader view
(–). Evidence of obesity in humans can be found in
obesity treatment were published in  and 
and recommended chamomile soap and vinegar, as
well as other remedies (, , ).
primitive art that dates back to the Paleolithic age (). By the beginning of the th century, doctors were
Two thousand five hundred years ago, Hippocrates using a number of medications for treating obesity
cautioned that sudden death is more common in those (e.g., thyroid extract, dinitrophenol, and amphet-
who are naturally fat than lean (). amine), often with unfortunate outcomes ().
In , the English physician Malcolm Fle- The discovery of leptin in  () (a peptide
myng wrote that obesity can be called a disease, produced in adipose tissue) marks the beginning of the
because it obstructs the free exercise of the animal “molecular era” for obesity. People who are deficient in
functions and can shorten life (). In , William this peptide become massively obese. Leptin replacement
Wadd (Secretary of the Royal College of Surgeons therapy completely reverses obesity for these individuals.
in London) stated that the increase of wealth and However, leptin treatment has proven ineffective in the
the refinement of modern times may have banished typical obese patient who is not leptin deficient.
plague and pestilence, but it has introduced ner- Rapid advances in basic science related to main-
vous disorders and increased the frequency of taining an appropriate amount of body fat have
corpulence (). provided insights into potential treatments for obesity.
Modern concepts of the pathophysiology of obesity This newer understanding of the regulation of food
date back to the end of the th century when Antoine intake and body weight provides the basis for promising
Lavoisier established that life was synonymous with future developments (, ).
oxidation (). More than  years later, Atwater and
Rosa () applied the laws of thermodynamics to hu- Headwinds in the management of obesity
man beings, and during the th century, researchers Despite progress in understanding obesity, advance-
discovered that hypothalamic tumors and tumors of ments in the clinical management of the disease
the pituitary gland could cause obesity (–). struggle against several headwinds.
Obesity treatments date as far back as Hippocrates, First, obesity is a stigmatized condition. The
who recommended lifestyle changes to obese patients general public and health professionals often respond
(, ). Two thousand years later, William Banting negatively to overweight persons, which can negatively
(an undertaker living in London in the th century) affect treatment ().
wrote one of the first “popular” diet books (). Second, the desire for the cosmetic effects of weight
“Drug” therapies can be traced back at least to the loss often far exceeds the desire for the health benefits
th century when Hisdai ibn Shaprut cured “Sancho associated with reducing weight (–). This may

2 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

well account for the fact that there are more women This stigma of obesity as a cosmetic issue vs
seeking help in managing obesity than men, even a health issue also affects how the U.S. Food and
though the health issues related to obesity are similar Drug Administration (FDA) reviews drugs that
between the sexes (–). manage weight loss. The FDA holds antiobesity
Although a modest % to % weight loss has drugs to a higher standard of review than other
proven health benefits, it often does not provide the drugs, requiring that the risks from these medica-
cosmetic benefit that patients are looking for. This tions be very low compared with drugs of other
results in a mismatch between the patient’s goals for classes ().
weight loss and what diet and exercise can realistically Finally, the lack of reimbursement by health in-
achieve (). The same is true with surgical approaches surers has resulted in poor sales of drug therapies for
to weight loss; patients often value the appearance of obesity, which only further dampens the pharma-
lost weight much more than the health benefits (, , ceutical industry’s interest in developing drug thera-
–). pies for obesity ().

Defining Obesity Overweight and obesity are worldwide problems ()

that affect . million Americans or .% of the
Introduction adult population. The most recent data from the  to
Historically, the medical community defined excess  U.S. National Health and Nutrition Examination
weight and its associated health consequences using Survey indicate that obesity (defined as BMI $  kg/m)
population-based anthropometric measurements, affects ~.% of men and .% of women in the United
(i.e., sex-specific body weight and height using life States (). Among children and adolescents aged  to 
insurance tables) (, ). However, these data only years, the prevalence of obesity in  to  was .%
represented insured individuals based on norma- (). Fig.  shows the percentage of U.S. men and women
tive standards without considering adiposity, and categorized as overweight, obese, or extremely obese
clinicians eventually abandoned these tables in between  and . The category of extreme obesity
favor of body mass index (BMI), which is a mea- (BMI .  kg/m) shows the greatest proportional
sure of body weight adjusted for height [weight change and is the most difficult group to effectively treat
(kg)/height (m)]. without surgery.
The National Institutes of Health and the World Among adult men, the prevalence of obesity is:
Health Organization have both adopted BMI as a cri- Hispanic, .%; black, .%; white, .%; and Asian,
terion for defining obesity (, ). This made in- .%. In women, the prevalence of obesity is: black,
terpretation simpler, eliminated the need for sex-specific .%; Hispanic, .%; white, .%; and Asian,
height/weight tables, and provided a measurement that .%. In children and adolescents, .% of - to
is better correlated with other estimates of adiposity. -year-olds are obese, with males and females equally
The measurement is based on the observation that affected (). The prevalence of obesity among chil-
dren and adolescents is: Hispanic, .%; black, .%;
body weight is proportional to the squared height in
white, .%; and Asian, .% ().
adults with normal body frames. In adults, classifi-
cation systems () and obesity guidelines (, )
define healthy body weight as a BMI between . and 50 Men
. kg/m, overweight between . and . kg/m , Women
and obesity $ kg/m . In children and adoles- 40

cents, the U.S. Centers for Disease Control and

Prevention (CDC) BMI-for-age growth charts
define overweight as a BMI at or above the th

percentile of standard weight and obesity as a BMI
above the th percentile of standard weight. 20
BMI provides the most useful population-level
measurement of overweight and obesity, and nu- 10 Obesity Extreme obesity
merous large population studies across multiple
Extreme obesity
continents have demonstrated its utility as an estimate 0
of risk (–). Additionally, current assessment and 1960- 1971- 1976- 1988- 1999- 2003- 2007- 2011-
1962 1974 1980 1994 2000 2004 2008 2012
management guidelines from the United States,
© 2018 Endocrine Reviews ENDOCRINE SOCIETY
Canada, and Europe recommend measuring BMI as
a first screening step in evaluating adult and pediatric Figure 1. Trends in the United States for adults with obesity or overweight, 1960–1962 to
patients for obesity (, –). 2011–2012 (48).

doi: 10.1210/er.2017-00253 3

Downloaded from
by guest
on 17 March 2018

Limitations of the BMI BMI was significantly associated with adverse car-
diometabolic traits and higher risks for both TDM
Adults and coronary heart disease (). A  standard de-
Although research had demonstrated the utility of viation increase in WHR adjusted for BMI was as-
BMI in assessing population-based mortality and sociated with a % higher risk of TDM (odds ratio,
disease-specific morbidity, there are two major limi- . [% confidence interval (CI), . to .]) and
tations in using BMI alone to diagnose obesity in an a % higher risk of coronary heart disease [odds ratio,
individual. . (% CI, . to .)].
The first is the inability of BMI to distinguish
weight associated with muscle vs fat. Children
Population studies have demonstrated a high There has also been concern about the association
specificity of using BMI cutoff values to diagnose between obesity and visceral or central adiposity
obesity but low sensitivity to identify adiposity, thus among children and adolescents, which has led to
missing approximately half of people with excess fat suggestions for using waist circumference in pediatric
(). This is particularly concerning in the elderly patients as well (). However, there are many issues
population, where a reduced lean body mass (sarco- with the implementation of this in routine pediatric
penia) might be misclassified as a healthy BMI (). practice, such as lack of standardized definitions of
Dual-energy x-ray absorptiometry or air displacement waist circumference and the inability of waist cir-
plethysmography are both accurate methods to assess cumference to add much to the strong association
lean body mass and body fat, but they are expensive between BMI and comorbidity in children (). If
and thus impractical for routine clinical application. clinicians are going to use waist circumference to help
Using bioelectric impedance to measure body define obesity in children, it is likely that we will need
water provides a relatively inexpensive measure of population-based percentile values, similar to those for
body fat mass vs fat-free mass (as body fat contains BMI (–).
more water). However, this method has large in- Because of these limitations, BMI has also emerged as
terindividual variations, suggesting that this method the most useful approach in children . years of age ().
may be insufficient for estimating individual body fat
mass and fat-free mass (). Are there metabolically healthy people
BMI also does not distinguish body fat distribution, with obesity?
a known determinant of metabolic risk. Measuring fat
distribution helps identify higher risk individuals, because Adults
increased visceral fat predicts the development of the In cross-sectional studies, many individuals with
metabolic syndrome, type  diabetes mellitus (TDM), obesity do not manifest “associated” comorbidities,
and total and cardiovascular mortality risk better than such as prediabetes, dyslipidemia, hypertension, or
total body fat alone (–). Several anthropometric other comorbidities (). These individuals often have
techniques are available to estimate the distribution of a predominantly lower body fat distribution and
body fat, such as waist circumference alone, the ratio of normal insulin regulation of adipose tissue lipolysis
waist circumference divided by hip circumference (waist- (, ). The phenotype “metabolically healthy obe-
to-hip ratio ([WHR]), and the ratio of waist circum- sity” (MHO) meets the standard BMI cutoff point for
ference divided by height (waist-to-height ratio). These obesity ($ kg/m) but does not have other elements
measures have been associated with the risk of de- of the metabolic syndrome, such as insulin resistance
veloping heart disease, TDM, and other chronic (, ). They have lower levels of visceral and ectopic
problems associated with obesity (, ). Combining fat, less liver steatosis (), and a lower degree of
waist circumference with BMI provides a way to in- systemic inflammation. Among the  studies iden-
corporate weight distribution into measures of obesity. tified by Rey-López et al. (), there were  definitions
Studies have demonstrated a strong link between waist of metabolic health that relied on four criteria: blood
circumference and BMI for both cardiovascular disease pressure (BP), high-density lipoprotein (HDL) cho-
(CVD) and TDM (, ). Waist circumference is most lesterol, triglycerides, and plasma glucose. BMI $
useful in individuals with a BMI of # kg/m ().  kg m was the main criterion for obesity. In this
However, despite its promise, most clinicians only use group of studies, the prevalence of MHO ranged
BMI and not waist circumference as a gauge of risk from between % and % (, –).
obesity. Beyond recommending annual BMI and waist Whereas short-term cross-sectional studies suggest
circumference testing, the American Association of that MHO men and women are not at increased risk of
Clinical Endocrinologists also recommends evaluating CVD, longitudinal studies suggest that this phenotype
other potential associated events (). may not be benign, and that this group is at higher risk
Genetic factors are involved in the relationship of for increased carotid artery intima-media thickness,
waist circumference to risk of CVD or TDM. A coronary calcification, impaired vasoreactivity, and/or
polygenic risk score for increased WHR adjusted for other cardiovascular events, as well as all-cause

4 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

mortality (, –). Therefore, clinicians should muscle, which suggests that efforts should be made to
view MHO as a transient or intermediary state that “conserve” muscle mass in old age ().
may progress over time to an unhealthy phenotype in
many people. Cardiorespiratory fitness is one factor Children
related to MHO. Research has shown that cardio- During childhood and adolescence, there are sub-
respiratory fitness lowers the risk of all-cause stantial changes in growth, body composition, and
mortality for metabolically unhealthy individuals pubertal status. During periods of rapid growth,
with obesity and those with and without the MHO weight and height may be somewhat mismatched, with
phenotype (–), suggesting that the inclusion of weight gain preceding growth in height. However, in
cardiorespiratory fitness along with BMI and waist the past three decades, children are often gaining
circumference may improve the assessment of risk weight at a pace much faster than what could be
status. Several systems are available for evaluating considered healthy or normal.
and staging obesity when assessing risk (–). Another critical period is the time when growth in
Increasing physical activity might thus be a valuable height ceases and caloric requirements decrease. If
recommendation for individuals with MHO. Ad- calorie intake does not adjust, weight gain is the likely
ditionally, clinicians should observe these in- result.
dividuals for the risk of developing cardiometabolic Furthermore, adolescence is a time of relative in-
disease (–). sulin resistance (). Because of this insulin resistance,
adolescents who are obese become more susceptible to
Children the development of TDM.
There also has been interest in whether children and
adolescents can be obese but metabolically healthy.
Some pediatric patients with obesity, even some with Prevention of Obesity
severe obesity, have few metabolic or clinical abnor-
malities (). However, the presence of obesity tends to Recent trends suggest that we are making some
track from childhood to adolescence and on to progress in the prevention and control of the obesity
adulthood. Thus, there is a high likelihood that a child epidemic using several strategies outlined below. First,
with obesity will become an obese adult, often with the the prevalence of obesity among - to -year-old
severity of obesity increasing over time with ongoing children has decreased significantly since  to
weight gain. This makes it likely that children and  (). Second, it has plateaued among - to
adolescents with obesity, even when metabolically -year-olds (). In contrast, however, obesity has
healthy at presentation, will develop associated dis- continued to increase in adult women ().
eases over time.
Strategies for preventing obesity in pregnancy
Age and obesity Three systematic reviews relating weight gain during
pregnancy and pregnancy outcomes found that dietary
Adults interventions reduced gestational weight gain and the
The current guidelines for assessing obesity among risks of preeclampsia, hypertension, and shoulder
adults do not consider age as an independent criterion. dystocia in infants. No differences occurred in the
However, there are physiological and functional incidence of small-for-gestational-age infants as a re-
changes that occur among the aging population that sult of these treatments (–).
may confound the interpretation of BMI and risk A  Cochrane review found that diet, exercise,
estimates in older people. Body composition changes or both reduced excessive gestational weight gain by an
associated with aging include sarcopenia, reduced average of %. Dietary interventions—including low
bone mineral density, and the accumulation of visceral glycemic index diets, supervised or unsupervised ex-
fat; BMI alone will not detect these changes (). BMI ercise programs, and diet combined with exercise—all
values associated with the lowest relative mortality are had comparable effects. Maternal hypertension was
slightly higher in older than in younger adults, which is reduced, but preeclampsia was not. No differences
often misinterpreted to suggest that obesity is not as were found between intervention and control groups
harmful in the elderly. BMI may be a less appropriate in the risk of preterm births or macrosomia. However,
index in the elderly because of sarcopenia (). Centrally a % reduction in macrosomia occurred among
located fat (waist circumference) and relative loss of fat- women who were overweight or had obesity. Newborn
free mass may become more important than BMI in respiratory distress syndrome was also decreased in the
determining the health risk associated with obesity in intervention groups among mothers who were either
the elderly (). The importance of loss of muscle mass overweight or obese (). Maternal consumption of
was clearly shown in the Health ABC Study where older sugar-sweetened beverages, similar to maternal smok-
adults with greater thigh muscle loss had a higher risk of ing, may also have long-term detrimental effects on
mortality compared with those with preserved thigh their offspring. Gillman et al. () reported that at an

doi: 10.1210/er.2017-00253 5

Downloaded from
by guest
on 17 March 2018

average age of . years children of mothers who A number of long-term studies lasting $ months
consumed two or more servings per day during the provide more specific information on the effects of
second trimester of pregnancy were both fatter and school-based interventions. We summarized these in
heavier. This provides an additional important piece of Table  (–).
information to provide to the pregnant woman. Randomized controlled trials (RCTs) have shown
that the reduction or elimination of sugar-sweetened
Strategies aimed at children drinks (often through the substitution of calorie-free
The two most important settings for the prevention of beverages) has effectively reduced rates of weight gain
obesity in children and adolescents are early care and in children and adolescents (). These observations
education (ECE) and schools. Children spend a lot of are consistent with the association between re-
time in these settings, where there are great oppor- ductions in sugar-sweetened drinks and both the
tunities for instilling positive behaviors regarding decrease in the prevalence of obesity in - to -year-
nutrition and physical activity. old US children and the plateau in the prevalence of
obesity in - to -year-old US children. The absence
Early care and education of a significant effect in several of these studies may
Although millions of young children are enrolled in ECE, indicate that a significant caloric deficit relative to the
there are only a few intervention studies on preventing or control condition was not established or sustained
mitigating obesity in these settings (). One of these ().
studies is the Romp and Chomp Intervention conducted Compared with efforts in specific settings, clinical
in Australia. This study used multiple ECE and com- interventions aimed at prevention have had limited
munity interventions directed at children  to  years of impact ().
age. The interventions included logical and proven targets
for weight control, such as reducing sugar-sweetened Strategies aimed at adults: worksites
drinks and energy-dense foods, increasing fruit and In , the Center for Disease Control’s Task Force
vegetable intake and active play, and reducing television on Community Preventive Services concluded that
time. The study reported significant reductions in obesity worksite health promotion programs that improved
prevalence in - and .-year-old children compared with physical activity and/or nutrition were effective in
children who did not receive the interventions (). reducing body weight and BMI (). Studies were
Because of the immense impact that policy and limited to those with at least  months of follow-up. A
environmental changes in ECE could have on child- pooled effect of nine RCTs found a weight decrease of
hood obesity, widespread efforts are underway to de- . kg, and a pooled effect of six RCTs found a de-
velop and incorporate standards and programs to crease of . BMI units (). Most of the studies
increase physical activity and improve diets in ECE combined informational and behavioral strategies to
settings (). One such program is the U.S. Department influence diet and physical activity; fewer studies
of Agriculture’s Child and Adult Care Food Program, modified the work environment (e.g., cafeteria, ex-
which helps child care institutions provide nutritious ercise facilities) to promote healthy choices. Recent
foods that contribute to the wellness, healthy growth, efforts to reduce sugar-sweetened drink consumption
and development of young children (). in hospitals have effectively used labeling and choice
architecture as environmental strategies to reduce
Schools sugar-sweetened drink consumption (, ).
A recent Cochrane meta-analysis of  studies (in-
cluding , children) () found beneficial effects Strategies for preventing obesity aimed at the
of a number of components of school-based in- entire population
terventions. These included: school curricula that Population-based messages aimed at the public con-
incorporate healthy eating, physical activity, and cerning food and exercise require individual com-
body image; increased sessions for physical activity and mitment if they are to be effective (). If individuals
the development of fundamental movement skills follow the advice in the message, this strategy would be
throughout the school week; improvements in the nu- sufficient to overcome the epidemic of obesity. However,
tritional quality of the food that schools supply; envi- positive nutritional messages are often dwarfed by al-
ronments and cultural practices that support children ternative messages urging consumption of less healthful
eating healthier foods and being active throughout each foods, and the built environment is often a barrier to
day; support for teachers and other staff to implement healthful exercise behaviors.
health promotion strategies and activities (e.g., pro- One approach might be to re-engineer the built
fessional development, capacity building activities); and environment to displace car use with physically active
parental support and home activities that encourage transportation options (such as walking and biking)
children to be more active, eat more nutritious foods, and and increase the number of accessible healthful food
spend less time in screen-based activities (). Beneficial options (). A systematic review by Papas et al. ()
effects were most notable in children  to  years old. identified  studies that examined the association

6 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

Table 1. Long-Term Studies or Preventive Interventions in Children and Adolescents

Reference Sample Design Results

Epstein et al., 2001 (102) 26 children 12-mo RCT 1.1% decrease in overweight prevalence with
increased fruits and vegetables vs 2.4% with
decreased fat and sugar

6–11 y old Increased fruits and vegetables vs Differences not significant

decreased fat and sugar

James et al., 2004 (103) 644 children 1-y intervention; classrooms randomized No significant difference in BMI z-score
to reduce sugar drink consumption
7–11 y old

Ebbeling et al., 2012 (104) 224 overweight or obese RCT Significantly lower rates of weight gain in
adolescents; mean age intervention group
15 y Sugar-free drinks and behavior modification
vs untreated control

de Ruyter et al., 2012 (105) 641 children 18-mo RCT Significantly lower rates of weight gain among
group receiving sugar-free drinks
5–12 y old Sugar-free drinks vs drinks containing
sugar at lunch

Sallis et al., 2009 (106) 995 4th and 5th grade PE taught by PE instructor or teacher Some fitness measures improved in girls
students vs control
No significant differences in changes in skinfolds

Caballero et al., 2003 (107) 1704 Native American 3-y study randomized by schools to No significant difference in body composition
children 8–11 y old control or intervention (41 schools); or PA

changes in dietary intake, increased

PA, classroom curriculum changes,
family involvement

Gortmaker et al., 1999 (108) 1295 6th–7th grade 2-y RCT with five intervention and Decreased prevalence of obesity in girls
students five control schools

Decreased TV, decreased fat and

increased fruit and vegetable
intakes, and PA

Plachta-Danielzek et al., 240 intervention and 952 Nutrition intervention delivered within No significant difference in increases in overweight
2011 (109) nonintervention schools and daily running games vs between intervention and control students in
children controls 8-y follow-up

Mean age 6 y old Significant decreases in BMI z-scores with upper

income students

Sahota et al., 2001 (110) 636 children, 7–11 y old Randomized by school. Teacher training, No significant differences in BMI in intervention
changes in school meals, and compared with control schools
314 intervention development of school actions plans
to promote healthy eating and PA
322 control

Abbreviations: PA, physical activity; PE, physical education; RCT, randomized controlled trial; TV, television.

between obesity and the numbers of outlets for changes among people living in poverty. Moving from
physical activity and food,  of which were cross- a high-poverty neighborhood to a neighborhood of lower
sectional. Seventeen of these studies found a sig- poverty was associated with a reduced prevalence of severe
nificant relationship between the built environment obesity ().
(food outlets or physical activity opportunities) and Use of public transit has also been associated with
the risk of obesity. The number of recreational increased levels of physical activity (). For example,
facilities and likelihood of overweight in adolescents the implementation of a light rail system in Charlotte,
were significantly related. However, few studies have North Carolina, was associated with a higher odds of
examined the impact of changes in the built envi- meeting the daily physical activity requirement and
ronment with changes in the prevalence of obesity. a lower BMI (). Neighborhood walkability appears
One exception is a study of the impact of housing to have much the same effect ().

doi: 10.1210/er.2017-00253 7

Downloaded from
by guest
on 17 March 2018

Food sugar-sweetened beverages. Colchero et al. () re-

Faith et al. () concluded that manipulating the ease ported that purchases of these taxed beverages de-
of food access and/or restricting access to certain foods creased .% in  and .% in , yielding an
might influence food purchases, consumption, and average reduction of .% during the study period.
possibly weight change, although this requires further Whether this translates into improvements in health
research. In contrast, the food industry favors the outcome is currently unknown.
hypothesis that obesity results from reduced levels of
physical activity and strongly supports providing more Increasing physical activity
places for people to exercise and more healthful food As indicated above, physical activity levels in both
alternatives as a strategy to help overcome the obesity children and adults have declined substantially.
problem (, ). However, the expense of healthy Helping incorporate exercise into how people get from
food items and limited access to healthful foods in where they live to where they learn, work, shop, play,
many lower income communities pose significant and pray has become a prominent strategy to reverse
challenges. To address access to healthful food options, this trend. Table  lists  strategies that the CDC’s
the Healthy Food Financing Initiative introduced Guide for Community Preventive Services identified
supermarkets to underserved communities. However, for increasing physical activity ().
this did not increase the consumption of healthful The CDC has also released a convenient guide that
foods (). Ideally, improved access needs to be focuses on how to implement these strategies ().
accompanied by pricing and promotion strategies to China provides an interesting example of how ur-
increase consumption of more healthful products. banization and improved incomes reduces physical ac-
Some of the strategies for introducing healthful tivity (, ). As recently as  years ago, the bicycle
food options include introducing farmer’s markets, was a major mode of transport for Chinese. Since then,
subsidizing the availability of fresh fruits and vege- the automobile has relegated bicycles to museums.
tables to school children, lowering the cost of fruits
and vegetables while increasing the price of high-fat or
high-sugar foods in school or worksite cafeterias, and/ Obesity and Disease
or changing marketing strategies. These strategies, for
the most part, increase fruit and vegetable con- Obesity and risk of death
sumption (–). Importantly, however, note that For many illnesses related to obesity, there is a cur-
addressing fruit and vegetable consumption alone vilinear increase in risk as a function of weight (Fig. )
might not be enough, as the impact of fruit and (). The Global Burden of Disease project ()
vegetable consumption on obesity prevention is un- reported this relationship between BMI and all-cause
certain. However, increased fruit and vegetable con- mortality in  prospective studies that included .
sumption does confer significant health benefits. Diets million people with a median follow-up of . years.
high in fruits and vegetables and low in fat and sugar Nearly  million subjects who survived  years were
lowered BP across the range of salt intake in in- free of chronic diseases at recruitment. There was
dividuals who were maintaining their body weight a clear J-shaped relationship between the BMI of the
(, ). , who died and all-cause mortality. The lowest
Taxation provides another strategy to reduce
consumption of less healthful products by increasing
their price. Smed et al. () has shown that among Table 2. Evidence-Based Recommendations To Increase
Physical Activity in Communities
Europeans, increasing the tax or reducing the subsidies
on unhealthful items and reducing the tax on healthful Community-wide campaigns to promote physical activity
items through the value-added tax system could shift
Point of decision prompts for stair use
consumption toward healthier foods (). Because of
their contribution to obesity, taxation of sugar- Individually adapted health behavior change programs
sweetened drinks has become a major focus in the
Enhanced school-based physical education
United States. Although many municipalities have
imposed sales taxes on sugar-sweetened drinks, this Social support in community settings
approach is less effective than an excise tax, which
Creation of or enhanced access to places for physical activity
increases the price of the product on the shelf. In , combined with informational outreach activities
Berkeley, California, passed a sugar drink tax of $.
per ounce. A study of sugar-sweetened drinks in that Street-scale urban design and land-use policies
city reported that consumption in low-income Community-scale urban design and land-use policies
neighborhoods (compared with two neighboring
communities) declined by % and water consump- Active transport to school
tion increased by % (). In January of , Transportation and travel policies and practices
Mexico imposed an excise tax of  peso per liter on

8 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

mortality was with a BMI of . to . kg/m. Below new, healthy subcutaneous adipocytes are relatively
this (BMI, . to . kg/ m), mortality significantly protected from many of the metabolic consequences
increased by % [hazard ratio (HR), .]. In the of obesity. Those without this ability will store excess
individuals with a BMI of . to . kg/m, all-cause fat in ectopic depots, including liver, visceral fat, and
mortality increased by % (HR, .), and with a BMI muscle (). This is supported by the finding that
of . to . kg/m, it increased by % (HR, .). larger subcutaneous fat cells are associated with more
For grade- obesity (BMI . to ,. kg/m), all- accumulation of visceral fat during overfeeding, be-
cause mortality increased by % (HR, .), and for cause they cannot expand to store more fat ().
grade- obesity (BMI, . to ,. kg/m), it in- One study reported that the predisposition for
creased by % (HR, .). For those with grade- TDM was associated with impaired recruitment of
obesity (BMI, . to ,. kg/m), all-cause mortality new adipose cells to store excess lipids in subcutaneous
rose by % (HR, .). For each  BMI unit increase, adipose tissue (). Another study reported that
total mortality rose by %, mortality from chronic adults who develop more leg adipocytes in response to
kidney disease rose by %, and mortality from TDM overfeeding have a lesser increase in abdominal
rose by % (). subcutaneous adipose size (). Fabbrini et al. ()
Just as weight gain can increase the risk of mor- showed that those with MHO are resistant to the
tality, weight loss can reduce the risk of mortality in adverse metabolic effects of moderate weight gain,
obese individuals. The results from the Swedish Obese whereas metabolically unhealthy people are predis-
Subjects Study (which compared long-term follow-up posed to such adverse effects. These authors concluded
of obese patients after surgical intervention for obesity that increased adipose tissue capacity for lipogenesis
with a matched but unoperated control group) showed might help protect people with MHO from weight
a % reduction in overall mortality after . years gain–induced metabolic dysfunction, at least with
(). Individuals in the Look AHEAD trial had modest weight gain during shorter periods of time.
a similar outcome after a median follow-up of . In addition to the known toxic effects of excess
years. Those who lost at least % of their body weight fatty acids, abnormalities in the hormonal function of
in the first year of the study had a % lower risk of the adipose tissue may contribute to metabolic disease.
primary CVD outcome [HR, . (% CI, . to Adiponectin is the most abundant peptide produced
.); P = .] and a % reduced risk of the sec- by adipose tissue (). It improves insulin sensitivity
ondary outcome [HR, . (% CI, . to .); P = and vascular function. Adiponectin concentrations are
.] compared with individuals who were weight inversely related to adipocyte size and visceral fat mass.
stable or gained weight. Participants in the intensive In contrast, most adipokines are secreted in larger
lifestyle intervention group who lost at least % of quantities as fat cells increase in size.
their body weight had a % lower risk of the primary Researchers have discovered a large number of
CVD outcome [HR, . (% CI, . to .); P = adipokines, but their exact role in disease is often
.] and a % lower risk of the secondary CVD unknown. The angiotensinogen produced by adipose
outcome [HR, . (% CI, . to .); P = .] tissue is a precursor for angiotensin, which can con-
compared with the control group (). tribute to the risk for hypertension. Additionally, the
aromatase enzymes in adipose tissue can convert
The mechanism of obesity-associated morbidity sterols (androstenedione) to estrogen, which may
The effects of obesity on the body appear to be me- explain the greater risk of breast and endometrial
diated by several major pathways. Fig.  shows how cancer in women with obesity, particularly post-
obesity as a disease process might lead to a variety of menopausal women with obesity where estrogens
other diseases (). derived from fat are their principal source of estrogens
A variety of types of adipose tissue dysfunction clearly ().
play a role in the genesis of many obesity-related diseases.
These include impairments in adipocyte storage and Type 2 diabetes mellitus
release of fatty acids, overproduction or underproduction There is overwhelming evidence that BMI, central
of “adipokines” and cytokines (), hormonal conver- adiposity, and the increase in body weight predict future
sion, and the adverse mechanical effects of greater tissue TDM (). A meta-analysis of prospective studies
mass. provided evidence that as upper body adiposity in-
The pathology of obesity is closely linked with body creases, both the risk of the metabolic syndrome and of
fat distribution. Upper body/visceral or ectopic fat developing TDM also increase (). The duration of
accumulation is a much better predictor of insulin obesity in younger compared with older individuals is
resistance, dyslipidemia, and such than total fat. also associated with a greater risk for TDM ().
Visceral fat is considered one of the “ectopic” fat Weight gain in adult life increases the risk of developing
depots, along with hepatic, intramyocellular, in- TDM, particularly in the age range  to  years ().
tramuscular, and pericardial fat. Humans with the The duration of increased body weight is also a risk
ability to respond to excess energy intake by recruiting factor for TDM. For a given level of excess BMI-years

doi: 10.1210/er.2017-00253 9

Downloaded from
by guest
on 17 March 2018

in the National Longitudinal Survey, younger in- Breast cancer is not only related to total body fat
dividuals compared with older ones (and Hispanic and but also may have a more important relationship to
black compared with white individuals) had a higher central body fat (). This relationship to body fat
risk of developing TDM (). may also help explain why breast cancer risk is in-
Weight loss is clearly beneficial in reducing the risk creased at age  in women in the highest vs the lowest
of converting to diabetes. In the Diabetes Prevention quartile of BMI (). Circulating, unconjugated es-
Program, a median weight loss of .% during . years tradiol may mediate the relationship between in-
reduced the risk of converting from prediabetes to creased body fat and breast cancer (), as well as the
diabetes by % (). Similarly, bariatric surgery has relationship between increased body fat and the risk of
repeatedly reversed diabetes to normal glucose tol- endometrial cancer ().
erance (–).
Myocardial infarction
Cancer Many studies show that as BMI increases, there is an
Certain forms of cancer are significantly increased in increased risk for heart disease (, ) and heart
individuals who are overweight (, ). Males face failure (). Data from the Nurses’ Health Study
increased risk for neoplasms of the colon, rectum, and indicate that the risk for U.S. women developing
prostate. In women, cancers of the reproductive sys- coronary artery disease is increased .-fold with
tem, including breast (), endometrium (), and a BMI .  kg/m compared with women with
gallbladder, are more common. Women who gained a BMI ,  kg/m (). A BMI of  to , kg/m
 kg or more after age  were at increased risk of increases the relative risk to .. Weight gain also
breast cancer (RR . P , .). Women who strongly affects this risk at any initial BMI. That is, at
gained  kg or more after menopause were also at all levels of initial BMI (and within BMI categories)
increased risk for breast cancer compared with women there was a graded increase in risk of heart disease with
whose weight remained stable. Women who lost and increasing waist circumference. Similarly, within waist
maintained $ kg and who did not use post- circumference categories there was an increased risk of
menopausal hormones were at lower risk than those heart disease with increasing BMI (). Major risk of
who maintained weight (RR, .) (). CVD was increased % for each . kg/m increase in
BMI among  British men ().
Central adiposity, as reflected in waist circumfer-
8.0 Studies Participants Deaths HR per 5 kg/m2 ence, is also a strong predictor of the risk for CVD
Men 157 913174 115328 1.51 (1.46–1.56) (). When increased central adiposity is added to
Women 141 2743371 264657 1.30 (1.26–1.33) other components of the metabolic syndrome, the
prediction is even higher. Using the National Health
and Examination Survey data, Janssen et al. ()
showed that BMI predicted the risk of the metabolic
syndrome in men. However, when BMI is adjusted for
waist circumference as a continuous variable, waist
circumference accounted for essentially all of the risk
for the metabolic syndrome. In a meta-analysis in-
cluding  studies, indices of abdominal obesity (in-
Hazard ratio

cluding WHR and waist circumference) were better

discriminators than BMI of cardiovascular risk factors,
including TDM, hypertension, and dyslipidemia
Both atrial fibrillation (, ) and congestive
heart failure (, ) have a higher risk in subjects
who are overweight. In the Multi-Ethnic Study of
Atherosclerosis, the risk of congestive heart failure in
obesity was associated with elevated levels of in-
flammatory markers (interleukin- and C-reactive
protein) and albuminuria ().
Heart failure and the obesity paradox
15 20 25 30 35 40 45 Obesity increases the risk of heart failure, yet some
studies have found that elevated BMI may improve
Mean body-mass index (kg/m2)
survival in individuals who already have congestive
Figure 2. BMI and all-cause mortality. Vertical bars are 95% CI. The Global Mortality Collaboration, heart failure, a phenomenon called “the obesity par-
2016 (142). adox” (–). This appears to contradict the

10 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

Figure 3. A schematic model of the intermediary mechanisms for dyslipidemia, insulin resistance, T2DM, heart disease, hypertension,
some forms of cancer, OSA, NAFLD, and osteoarthritis.

Healthy body composition

Positive fat balance

fat expansion

Constitutional limit to Additional positive Heart

subcutaneous fat expansion fat balance

Increased cytokine release Increased FFA release

Decreased adiponectin release subcutaneous fat Decreased TG storage
Increased angiotensinogen

Ectopic lipids
Further subcutaneous Increase portal FFA
fat expansion and IL6
vascular volume
Expansion of ectopic
(visceral) adipose stores Liver Muscle
Increased Increased Islets
estradiol body mass
Dyslipidemia Peripheral insulin
Increased increased CRP resistance
pharyngeal fat Increased glucose Reduced insulin
production secretion
Breast/Endometrial Osteoarthritis
Sleep apnea Diabetes mellitus
© 2018 Endocrine Reviews ENDOCRINE SOCIETY

curvilinear relationship of BMI to body weight (, ~ million of them in the United States (). Hy-
–, ). pertension is the most important of  risk factors for
One possible explanation is “selection bias.” This worldwide risk of coronary heart disease, stroke, renal
occurs when studies select individuals as higher risk disease, and all-cause mortality (). Furthermore,
because they are identified after the disease develops antihypertensive therapy results in reductions of in-
rather than before. A simple way to eliminate this bias cidence of stroke, myocardial infarction, and heart
is to match the start of exposure to the start of follow- failure ().
up. The same is true regarding the effect of obesity on Among hypertensive individuals who reduced their
the risk of mortality (, –). Alternatively, the BP levels following a successful weight-loss in-
obesity paradox may reflect some capacity of the in- tervention, those who maintained weight loss also
dividual with obesity to overcome cardiovascular risk. maintained lower BP levels, and those who regained
Still another explanation for this paradox may be the weight returned to their baseline BP levels (). In
difference between what BMI tells us and what the a meta-analysis of  studies, Neter et al. () found
underlying fat distribution is doing. In a recent study, that weight loss averaging . kg after diet and/or
Padwal et al. () found that BMI and body fat have exercise programs reduced BP by ./. mm Hg
different predictive values for cardiovascular risk. If fat (systolic BP/diastolic BP). The studies with weight
is the culprit, then measuring BMI may lead to an losses . kg showed larger decreases in BP than those
erroneous conclusion (). with less weight loss.

Hypertension and stroke Obstructive sleep apnea

Hypertension is a global public health problem. In contrast to the relatively benign effects of excess
Roughly  billion people worldwide are estimated to weight on most components of respiratory function,
have clinically significant elevations in BP (), with overweight predisposes to obstructive sleep apnea

doi: 10.1210/er.2017-00253 11

Downloaded from
by guest
on 17 March 2018

(OSA), which can be severe and life-threatening (). found that obesity produced a .-fold increased risk
OSA is more common in men than women. An in- of developing NAFLD. Moreover, there was a dose
creased snoring index and increased maximal noc- response to rising BMI, with the relative risk increasing
turnal sound intensity are characteristic. Nocturnal . for each  unit increase in BMI. Another meta-
oxygen saturation is significantly reduced (). A analysis () found that for each  unit increase in
study of obese patients with diabetes using poly- waist circumference, the odds ratio of NAFLD in-
somnography showed that .% of the participants creased ., and for each  unit increase in BMI, the
had moderate OSA, and .% had the severe form. odds ratio increased .. The prevalence is greater in
Waist circumference was significantly related to the Hispanic than white populations and less in blacks
presence of OSA, and severe OSA was most likely in than whites. NAFLD and NASH are also more
individuals with a higher BMI (). Independently of common in persons with TDM.
obesity, OSA is associated with features of the met-
abolic syndrome, including hypertension, TDM, and Gout and osteoarthritis
increased cardiovascular risk, possibly mediated by
stress responses and hypoxia. Excess daytime sleepi- Gout
ness is an important consequence and can be a risk for Aune et al. () reported on the relationship of BMI
driving and other tasks that require alertness (). to the risk of gout in  prospective studies that in-
cluded , cases of gout among a population of
Hepatobiliary disease , (median follow-up of . years). The sum-
mary relative risk for a -unit increment in BMI was
Gallbladder disease . for all studies combined (% CI, . to .).
Obesity is associated with an increased risk of gall- The summary relative risk per -unit increase in BMI
bladder disease. In a meta-analysis of gallbladder was . for men (% CI, . to .) and . for
disease and obesity, Aune et al. () reported that the women (% CI, . to .). The relative risks were
risk of gallbladder disease increased even within ., ., ., and . for persons with a BMI of ,
normal BMI ranges. For each -unit increase in BMI, , , and  kg/m, respectively, compared with
the relative risk of gallbladder disease increased %. persons with a BMI of  kg/m. The study also as-
For a  cm increase in waist circumference, the in- sociated increased risk with BMI in young adulthood,
crease in relative risk was %. WHR, and weight gain from age  to  to midlife,
but the analyses included few studies.
Nonalcoholic fatty liver disease
Fatty liver disease is often associated with obesity (). Osteoarthritis
Excess liver fat without inflammation/hepatocellular Osteoarthritis is likewise significantly increased in
injury is called nonalcoholic fatty liver disease (NAFLD), individuals who are overweight or obese. The osteo-
which may progress to nonalcoholic steatohepatitis arthritis that develops in the knees and ankles may be
(NASH) and eventually cirrhosis. The diagnosis of directly related to the trauma associated with the
NAFLD requires evidence of excess liver fat in the degree of excess body weight (). However, the
absence of secondary causes. NASH is diagnosed when increased osteoarthritis in non–weight-bearing joints
there is evidence of hepatocellular injury (most often in suggests that some components of the excess weight
the context of fatty liver) and is of greater concern may alter cartilage and bone metabolism independent
because it poses a genuine risk of progression to fibrosis, of weight bearing. Increased rates of osteoarthritis
cirrhosis, greater risk for hepatocellular carcinoma, and account for a significant component of the cost of
cirrhosis-related liver failure. overweight and for the associated disability ().
The prevalence of NAFLD ranges from % to %, Okoro et al. () found that class- obesity (BMI .
depending on the diagnostic approaches and pop-  kg/m) was associated with survey-reported dis-
ulations studied. The estimated prevalence of NASH is ability among individuals . years of age who re-
% to %. Both liver fat and fibrosis were increased as ported arthritis, as well as those who did not report
a function of time in nonhuman primates fed a high- arthritis.
fructose diet vs nonhuman primates without the
added fructose (). Effects of obesity during pregnancy
NAFLD is considered by some to be the hepatic A narrative analysis of  reviews on pregnancy in
manifestation of the metabolic syndrome (). Fatty women with obesity () showed that gestational
liver is extremely common in patients undergoing diabetes, preeclampsia, gestational hypertension, de-
bariatric surgery (prevalence % to %). The pression, instrumental and cesarean birth, and
prevalence of fatty liver in the United States has been surgical-site infection are more likely to occur in
increasing steadily from  to  with obesity as pregnant women with obesity compared with women
an independent predictor (). In a meta-analysis of with a healthy weight. Obesity in pregnancy is also
 studies ( of which were prospective), Li et al. () linked to greater risk of preterm birth, large-for-

12 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

gestational-age babies, fetal defects, congenital Strategies (POUNDS Lost) Study (–), in-
anomalies, and perinatal death. Additionally, breast- dividuals with the A genotype of the fat mass and
feeding initiation rates are lower, and there is greater obesity-associated (FTO) gene had greater weight loss
risk of early breastfeeding cessation in women with when assigned the high-protein diets but not when
obesity compared with healthy-weight women. eating the low-protein diets (). Another analysis,
which examined eight clinical trials in overweight or
adults with obesity, reported that the FTO genotype
Diet, Exercise, and Lifestyle in did not modify the response to diet (). Using
Managing Obesity genetic profiles may thus be of value in the future for
developing personalized dietary regimens for man-
Diet in managing obesity—food is more aging obesity, but more evidence is needed for any
than calories clinical applications.

Introduction Very low–calorie diets

The idea that single food items or diets are able to We define very low-calorie diets (VLCDs) as those
promote and maintain weight loss has stimulated having an energy level between  and  kcal/d. In
numerous studies to investigate different proportions a review comparing low-calorie diets with VLCDs,
of dietary fat, protein, or carbohydrates as weight-loss Tsai and Wadden () reported that VLCDs pro-
diets () (Table ). Underlying all of these dietary duced significantly greater short-term weight loss
approaches, however, is the fact that to lose weight, (.%) than did low-calorie diets (.%) but similar
energy balance must be negative. Although calories are longer-term weight loss.
the essential component of energy balance, and re-
ducing them is important for weight loss, food consists Carbohydrate subtypes, low-carbohydrate diets,
of more than calories. When choosing a diet, it is and sugar-sweetened beverages
important to select foods that you enjoy and substitute Carbohydrates, such as sugar or high-fructose corn
lower calorie healthy foods that can improve the syrup, create additional challenges to a weight-loss
quality of your diet. Macronutrient composition aside, diet, because added sugar in beverages provides extra
a reduction of energy intake is still an essential energy with reduced satiety, thus increasing the total
component of the effectiveness of any diet. In the energy intake ().
Diabetes Prevention Program, calorie reduction was In a meta-analysis, Nordmann et al. () found that
the major predictor of weight loss (). Reduced weight loss was greater at  months with low-
intake in fat was the second predictor, and physical carbohydrate diets (defined as carbohydrate intake
activity was only an important predictor when the of , g/d) but not at  months (compared with other
calorie intake was unchanged (). diets). In a meta-analysis of longer trials by Tobias et al.
A calorie deficit of  kcal/d produces a weekly (), interventions with similar intensity led to a sig-
deficit of ~ kcal, which is roughly equivalent to the nificantly greater weight loss of . kg on the low-
energy in  pound (. kg) of fat tissue (). Al- carbohydrate diets. This is in line with a meta-analysis by
though this calculation would predict linear weight Bueno et al. (), which showed a greater weight loss of
loss, weight loss is not linear; it is curvilinear. At the . kg with very low–carbohydrate ketogenic diets.
initial stage, weight loss tends to be more rapid, and Although both are statistically significant, the absolute
then slows until it reaches a plateau (–). The difference in weight loss was quite small (~ kg weight
initial reduction of calorie intake initiates a number of reduction in a -kg individual). These studies over the
compensatory mechanisms, which tend to drive food long term are hindered by the participants’ lack of ad-
intake up and reduce weight loss (–). herence to the prescribed dietary regimens.
Several factors contribute to the different patterns of To circumvent the problem of variable effects of
response during weight loss. The first is the initial rate of dietary protein in evaluating low-carbohydrate and
weight loss (). In the Look AHEAD trial, a multicenter low-fat diets (), Hall and Guo () performed
clinical trial in individuals with diabetes, those in the a meta-analysis of isocaloric low-carbohydrate/high-fat
highest tertile of initial weight loss in the first  months diets vs high-carbohydrate/low-fat diets where protein
had nearly twice as much weight loss at  and  years consumption was held constant. This analysis included
compared with those in the lowest tertile of weight loss in  studies ( subjects total), which provided all food to
the first  months. This could be explained by the fact the subjects. Dietary carbohydrate ranged from % to
that adherence to any dietary program is critical to % and dietary fat from % to % of total energy intake.
successful weight loss (, –). There was a small but significant  kcal/d weighted
Genetic variation can also influence weight loss, as mean energy expenditure difference favoring the low-fat/
can the biological response to different diets (, ). high-carbohydrate diets (not shown) and a small but
In both the Diabetes Prevention Program (, ) significant  g/d weighted mean body fat difference
and the Preventing Overweight Using Novel Dietary favoring the low-fat/high-carbohydrate diets (Fig. )

doi: 10.1210/er.2017-00253 13

Downloaded from
by guest
on 17 March 2018

Table 3. A Comparison of Various Diet Programs and Eating Plans to a Typical American Diet
General Dietary AHA/ACC/TOS Evaluation
Type of Diet Example Characteristics Comments and Others

Typical American Carb: 50% Low in fruits and vegetables, dairy,

diet and whole grains

Protein: 15% High in saturated fat and unrefined

Fat: 35%

Average of 2200 kcal/d

Balanced-nutrient, DASH Diet or diet based Carb: 55%–60% Based on set pattern of selections Meta-analysis showing DASH
moderate-calorie on MyPyramid food guide. from food lists using regular approach better than
approach Commercial diet plans grocery store foods or control or healthy diets
such as: Diet Center, prepackaged foods supplemented (weight mean difference
Jenny Craig, Nutrisystem, by fresh food items 0.87–1.5 kg).
Physician’s Weight Loss,
Shapedown Pediatric Protein: 15%–20% Low in saturated fat and ample in
Program, Weight Watchers, fruits, vegetables, and fiber
Setpoint, Sonoma, Fat: 20%–30% Recommended reasonable weight-
Volumetrics loss goal of 0.5–2.0 pounds/wk

Usually 1200-1800 kcal/d Prepackaged plans may limit

food choices

Most recommend exercise plan

Many encourage dietary record


Some offer weight-maintenance


Low- and very low–fat, Ornish Diet (Eat More, Weigh Carb: 65% Long-term compliance with some Same weight loss at 6 mo
high-carbohydrate Less), Pritikin Diet, T-factor plans may be difficult because of comparing 30% fat to .
approach Diet, Choose to Lose Diet, low level of fat 40% fat; strength of
Fit or Fat Diet evidence: moderate
Protein: 10%–20% Diet can be low in calcium

Fat: #10%–19% Some plans restrict healthful foods

(seafood, low-fat dairy, poultry)

Limited intake of animal Some encourage exercise and stress

protein, nuts, seeds, management techniques
other fats

Low energy density Volumetrics Diet Carb: 55% Four food categories: More weight loss at 6 mo
with low energy-dense
Protein: 10%–25% (1) Very low density—nonstarchy diet; strength of evidence:
fruits and vegetables, nonfat milk, RCT
broth-based soups

Fat: 20%–35% (2) Low density—starchy fruits/

vegetables, grains, breakfast cereal,
low-fat meats, and mixed dishes

Focus on fruits, vegetables, (3) Medium density—meat,

and soups cheese, pizza, fries, dressings,
bread, and such

(4) High density—desserts, nuts,

butter, oils

Focus on categories 1 and 2, some

from 3, minimum from 4
(Continued )

14 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

Table 3. Continued
General Dietary AHA/ACC/TOS Evaluation
Type of Diet Example Characteristics Comments and Others

Portion controlled Use of meal replacements Weight loss at 1 year in Look

both liquid and solid AHEAD trial related to
meals frequency of consuming
portion-control meals

Mediterranean- Carb: 35%–40% Eat primarily plant-based foods (fruits,

style diets vegetables, whole grains, legumes,
and nuts)

Protein: 12%–20% Healthy oils instead of saturated fats

Fat: 40%–50% Limit red meat to a few times a month

Approximately 25%–30% Eat fish and poultry at least twice

of energy from a week
monounsaturated fat
Red wine in moderation, for individuals
who choose to drink alcohol

Be active and enjoy meals with

family and friends

Low-carbohydrate, Atkins New Diet Revolution, Carb: #20% Promote quick weight loss (much is Same weight loss at 6 mo
high-protein, Protein Power Diet, Stillman water loss rather than fat loss) comparing ,30 g/d vs
high-fat approach Diet (The Doctor’s Quick 55% Carb–15% protein
Weight Loss Diet), or 40% Carb and 30%
Carbohydrate Addict’s protein
Diet, Scarsdale Diet
Protein: 25%–40% Ketosis causes loss of appetite Strength of evidence: low

Fat: $55%–65% Can be too high in saturated fat

Strictly limits carbohydrates Low in carbohydrates, vitamins,

to ,100–125 g/d minerals, and fiber

Not practical for long term because of

rigid diet or restricted food choices

Higher protein, The Zone Diet, Sugar Busters Carb: 40%–50% Diet rigid and difficult to maintain Same weight loss at 6 mo
moderate- Diet, South Beach Diet comparing 25%–30% vs
carbohydrate, Protein: 25%–40% Enough carbohydrates to avoid ketosis 15% protein; strength
moderate-fat Fat: 30%–40% Low in carbohydrates; can be low in of evidence: high
approach vitamins and minerals

Glycemic load The Glycemic-Load Carb: 40% to .55% Focus on low-glycemic-load foods Same weight loss at 6 mo
Diet—Rob Thompson comparing high vs low
Protein: 15%–30% glycemic load; strength
Fat: 30% of evidence: low

Low-sugar or non– Not really a diet but just No recommendation other Meta-analyses show that consumption Weight loss less in adolescents
sugar-sweetened a call to reduce sugar- than to reduce/remove of sugar-sweetened beverages is comparing artificial vs sugar-
beverages sweetened beverages sugar-sweetened related to risk of obesity, T2DM, sweetened drinks; strength
intake as a preventive beverages from your and heart disease of evidence: RCT comparing
strategy overall diet plan artificial sweetener vs
sugar-sweetened beverages

Novelty diets Immune Power Diet, Most promote certain No scientific basis for recommendations
Rotation Diet, Cabbage foods, or combinations
Soup Diet, Beverly Hills of foods, or nutrients as
Diet, Dr. Phil Diet having allegedly magical
(Continued )

doi: 10.1210/er.2017-00253 15

Downloaded from
by guest
on 17 March 2018

Table 3. Continued
General Dietary AHA/ACC/TOS Evaluation
Type of Diet Example Characteristics Comments and Others

Very low–calorie Health Management ,800 kcal/d Requires medical supervision

diets Resources Program,
Medifast Diet, Optifast For clients with BMI $ 30 or
Diet BMI $ 27 with other risk factors

May be difficult to transition to

regular meals

Weight-loss online Cyberdiet, Dietwatch, eDiets, Meal plans and other Recommend reasonable weight
diets tools available online loss of 0.5–2.0 pounds/wk

Most encourage exercise

Some offer weight-maintenance


Abbreviations: AHA, American Heart Association; ACC, American College of Cardiology; Carb, carbohydrate; RCT, randomized controlled trial; TOS, The Obesity Society.

(). This analysis does not support the concept of difference in weight loss of . kg that favored low–
a metabolic advantage for lower carbohydrate, higher fat glycemic index diets. Additionally, both total and low-
diets, suggesting that any benefits of such diets probably density lipoprotein (LDL) cholesterol fell more with
involve differences in energy intake. low–glycemic index diets. The long-term effects of
low–glycemic index diets warrant further evaluation.
Dietary fat, energy density, and low-fat diets Fasting glucose may provide a clue to dietary se-
For decades, dietary recommendations for weight loss lection. Hjorth et al. () have reported that in-
have emphasized a reduction in fat intake because of its dividuals with higher fasting glucose who are
high-energy content ( kcal/g) compared with carbo- prediabetic may respond better to a lower glycemic
hydrates ( kcal/g) (). A meta-analysis of six trials index diet with more fiber and whole grain.
reported no significant differences between low-fat diets
( to  g/d or % of total energy) vs other weight-loss High-protein diets
diets in terms of sustained weight loss (). A -year study comparing % and % protein diets
A recent systematic review and meta-analysis as part of a % fat diet (, ) reported that weight
compared the effects of low-fat interventions loss during  weeks was substantially greater with the
(,% total fat) vs other dietary interventions on higher protein diet, and that this result was maintained
long-term ($ year) weight changes. It found that up to  weeks but not at  weeks.
when the groups differed by .% fat content, the A meta-analysis of energy-restricted, high-protein/
higher fat interventions led to slightly greater weight low-fat diets compared with standard-protein/low-fat
loss and better adherence, although the magnitude of diets showed that the high-protein diet was better at
the differences in weight loss was small (). The reducing body weight (2. kg; % CI, 2.
important message is that “adherence” rather than to 2. kg), fat mass (2. kg; % CI, 2.
a specific diet is the important ingredient in success. to 2. kg), and triglycerides (2. mmol/L; %
Another strategy for reducing energy density CI, 2. to 2. mmol/L) and resulted in less of
(besides reducing dietary fat intake) is to substitute a decrease in fat-free mass (. kg; % CI, . to
foods with higher water content. One trial has com- . kg) and resting energy expenditure (. kJ/d; %
pared a reduced-fat diet to a diet with extra fruits and CI, . to . kJ/d) (). In the intent-to-treat
vegetables with lower energy density. In this trial, the analysis of the POUNDS Lost Study (), which
addition of fruits and vegetables led to greater weight compared % and % protein diets, there was no dif-
loss compared with lowering fat only (). Diets with ference in weight loss between these diets. However, those
a higher intake of fruits and vegetables evolved into the who adhered to a higher protein diet lost more weight.
Volumetrics diet (). The efficacy of the Volumetrics When this study used urinary nitrogen loss as a measure of
diet warrants further investigation. protein intake, those with the greater increase in protein
intake lost significantly more weight ().
Low–glycemic index diets
The glycemic index is based on the rise in blood glucose Mediterranean-style diets
in response to test foods (, ). A meta-analysis by Mediterranean-style diets are characterized by
Thomas et al. () reported a significant but small enhanced consumption of olive oil, nuts, whole

16 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

grain, fruits, and vegetables. In diabetic in- a weight loss of . kg compared with controls, and
dividuals, the Mediterranean diet produced the weight-loss effect lasted up to  years.
a greater weight loss during  years than did a low- In a -month intervention, the daily use of
fat diet (). Another meta-analysis () re- a commercially available portion-control plate was
ported that Mediterranean diets reduced body effective in promoting weight loss among patients with
weight . kg compared with low-fat diets. The obesity and TDM when compared with a usual-care
Prevención con Dieta Mediterránea (PREDIMED) dietary group. A meta-analysis of six studies using
study from Spain showed that consumption of meal replacements showed more weight loss than low-
a high-fat Mediterranean diet (.% calorie from calorie diets at  months (). Data from another trial
fat) resulted in a . kg weight loss (P = .) and showed that portion control can increase diet quality
a . cm waist circumference reduction (P = .) while maintaining significant weight loss during
vs a comparison diet (.% calorie from fat)  months ().
during . years of follow-up ().
Comparison of diets with different
Balanced-deficit diets macronutrient composition
Diets with a reduced content of carbohydrates, pro- Several RCTs have compared diets head-to-head (,
teins, and fat (so-called “balanced-deficit diets”) have , , , ). We summarize these in Table  (,
been widely used in managing obesity. In a meta- , –). These studies show improvements in
analysis, Avenell et al. () reported that intervention hemoglobin Ac (HbAc) in patients with TDM and
diets with an average deficit of  kcal/d led to improvements in triglycerides and HDL cholesterol in

ES (g/d) LCL (g/d) UCL (g/d) Weight (%)

Hall et al 2016 -15.48 -11.10 -19.85 20

Davy et al 2001 -39.34 -23.45 -55.24 1.5

Hall et al 2015 -39.31 -33.32 -45.29 11

Yang et al 1976 -3.30 9.57 -16.17 2.3

VerboeketVan et al 1994 -31.16 -26.71 -35.60 19

Bogardus et al 1981 21.43 86.43 -43.57 .09

Horton et al 1995 -2.14 14.47 -18.75 1.4

Treuth et al 2003 -22.49 -19.01 -25.97 32

Schrauwen et al 1997 -13.36 -10.18 -16.54 38

Roy et al 1998 -6.32 3.35 -15.99 4.1

Astrup et al 1994 -13.39 -10.90 -15.89 62

Golay et al 1996 34.07 65.82 2.32 .38

Shepard et al 2001 -49.28 -22.24 -76.32 .53

Eckel et al 2006 -44.57 -36.11 -53.04 5.4

Rumpler et al 1991 -7.14 91.85 -106.1 .04

McDvitt et al 2000 -4.77 -1.55 -8.00 37

Goldberg et al 1998 -12.76 -8.35 -17.16 20

Galgani et al 2010 -35.89 -27.26 -44.52 5.2

Kahlhofer et al 2014 4.64 19.66 -10.38 1.7

Smith et al 2000 -14.11 -10.59 -17.64 31

Pooled WMD -16.36 -15.21 -17.50

-100 -50 0 50 100

Favors low fat diet Favors low CHO diet

WMD in body fat (g/d)

Figure 4. Weight loss comparing isocaloric low-carbohydrate/high-fat and high-carbohydrate/low-fat diets where meals were provided
and protein consumption was the same. 95% horizontal CI. CHO, carbohydrate; ES, effect size; LCL, lower confidence limit; UCL, upper
confidence limit; WMD, weighted mean difference. See Hall and Guo, 2017 (235).

doi: 10.1210/er.2017-00253 17

Downloaded from
by guest
on 17 March 2018

Table 4. Weight Losses from Randomized Controlled Trials That Compared Diets With Varying Macronutrient Compositions
No. of Lifestyle
No., Sex, and Sessions Weight
Study Completers Provided Dietary Intervention Change Month Comments/Other Results

Bazzano et al., 148, 88% F, 80% 10 Low Carb (,40 g/d) 26.5 kg a
12 Participants without CVD or diabetes;
2014 (256) completed low carbohydrate diet group had
Low fat (,30% fat) 22.6 kgb greater decrease in body fat and
triglycerides and greater increase in
HDL cholesterol than did the
low-fat group. C-reactive protein
and 10-Year Framingham Risk Score
improved more in low-carbohydrate
group. No difference in BP response.
Low-fat group had lower protein
intake than in the low-carbohydrate

Dansinger et al., 160, 51% F, 58% 4 Atkins (Low Carb) 22.1 kga 12 All patients had hypertension, dyslipidemia,
2005 (216) completed and/or fasting hyperglycemia.

Zone (30% fat) 23.2 kga Weight loss was associated with level of

Weight Watchers 23.0 kga Each diet decreased LDL/HDL ratio.

(Low calorie)

Ornish (10% fat) 23.3 kga There were no significant effects on BP or

blood glucose at 12 mo.

Das et al., 2007 34, % F unknown, 52 Low glycemic load 27.8%a 12 Triglycerides and total, HDL, and LDL
(257) 85% completed cholesterol decreased in both groups.
High glycemic load 28.0% a

Fabricatore et al., 79, 80% F, 63% 30 Low glycemic load 24.5%a 9 All patients had T2DM.
2011 (258) completed
Low fat 26.4% a
There were larger reductions in HbA1c in
the low–glycemic load group.

Foster et al., 2003 63, 68% F, 59% 3 Low carbohydrate 24.4%a 12 HDL cholesterol increased more in the
(259) completed (high protein, high fat) low-carbohydrate group, and
triglycerides were lower only in the
low-carbohydrate group.

Conventional (high 22.5%a Diastolic BP decreased in both groups.

carbohydrate, low fat)
Area under the insulin curve decreased in
both groups.

Foster et al., 2010 307, 68% F, 63% 38 Low carbohydrate 26.3 kga 24 HDL cholesterol increased more in the
(260) completed low-carbohydrate group.
Low fat 27.4 kg a

Gardner et al., 2007 311, 100% F, 80% 8 Atkins (low carb) 24.7 kga 12 Increase in HDL cholesterol was larger
A to Z Study completed in the Atkins than in the Ornish
(254) group. Triglyceride levels decreased
more in the Atkins than in the Zone

Zone (30% fat) 21.6 kgb There were no differences in insulin or

blood glucose between groups.

LEARN (calorie restricted) 22.2 kga,b Systolic BP decreased more in Atkins than in
all other groups.

Ornish (,10% fat) 22.6 kga,b Diastolic BP decreased more in Atkins group
than in Ornish group.
(Continued )

18 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

Table 4. Continued
No. of Lifestyle
No., Sex, and Sessions Weight
Study Completers Provided Dietary Intervention Change Month Comments/Other Results

Sacks et al., 2009 811, 64% F, 80% 66 Low fat, average protein 22.9 kg a
24 LDL cholesterol decreased significantly
POUNDS Lost completed (highest carbohydrate) more in lowest fat/highest carbohydrate
Study (218) group than in highest fat/lowest
carbohydrate groups.

Low fat, high protein 23.8 kga HDL cholesterol increased more with
lowest carbohydrate than with the
highest carbohydrate diet.

High fat, average protein 23.1 kga All diets decreased triglyceride levels

High fat, high protein 23.5 kga All diets, except the highest carbohydrate
(lowest carbohydrate) diet, decreased fasting insulin (greater
decrease in the high-protein vs
average-protein diets).

Shai et al., 2008 322, 14% F, 85% 24 Low fat 22.9 kga 24 No significant change in LDL cholesterol
DIRECT Study completed in any group.
Mediterranean 24.4 kgb HDL cholesterol increased in all groups,
significantly more in the low-
carbohydrate than low-fat group.

Low carbohydrate 24.7 kgb Triglyceride levels decreased more

in the low-carbohydrate than in
the low-fat group.

In diabetic patients, only the

Mediterranean diet group had
a decrease in fasting glucose.

Insulin decreased in all groups for both

diabetic and nondiabetic patients.

All groups had a significant decrease

in BP.

Adiponectin levels increased and leptin

levels decreased in all groups.

Stern et al., 132, 17% F, 66% 15 Low carbohydrate 25.1 kga 12 Triglyceride levels decreased more in the
2004 (261) completed low-carbohydrate group than in the
low-fat group.

Conventional (low fat) 23.1 kga HDL cholesterol decreased less in the
low-carbohydrate group than in the
low-fat group.

Changes in total and LDL cholesterol were

not significant between groups.

Yancy et al., 120, 76% F, 66% 9 Low-fat diet 26.5%a 6 All patients were hyperlipidemic.
2004 (262) completed
Low-carbohydrate, ketogenic 212.9%b Triglycerides decreased more and HDL
diet with nutritional cholesterol increased more in the
supplements low-carbohydrate group.

Different letters (in superscript) indicate statistically significant differences (P # 0.05) in weight loss between groups.
Abbreviations: MR, meal replacements; VLDL, very low–density lipoprotein.

doi: 10.1210/er.2017-00253 19

Downloaded from
by guest
on 17 March 2018

the groups assigned to the low-carbohydrate diet arms. loss) randomized participants to one of four diets,
One trial randomized  individuals with obesity to one with % of patients providing data on body weight
of four popular diets, including the Atkins diet (), The at the end of  years. The diets were: () % fat/%
Ornish diet (), the Weight Watchers diet (), and protein; () % fat/% protein; () % fat/%
the Zone diet (). At the end of  months, each protein; or () % fat/% protein. The foods in all
diet produced similar weight losses (~ kg). Adher- four diets were the same, although they differed in
ence to the diets was the single most important quantity. At the end of  months,  months, and 
criterion of success in these trials. In one study, a low- years, the weight loss was similar for all four diets
fat diet was compared with a low-carbohydrate diet (); however, those who achieved the largest
(Atkins diet) and a Mediterranean-style diet (). increase in protein intake lost more weight ().
Compared with the low-fat diet, individuals assigned The similarity of the mean weight loss in all four diet
to the Mediterranean diet and low-carbohydrate diet groups obscures the wide range of individual weight
had significantly greater weight loss and maintenance losses shown in Fig.  (). The data from
by  months (). In a meta-analysis of numerous the POUNDS Lost Study are consistent with the
popular diets that included  unique trials, low- recommendations of the American College of
carbohydrate diets performed equally with low-fat Cardiology/American Heart Association/Obesity
diets after  months, with the low-carbohydrate diets Society Guideline for the Management of Over-
resulting in . kg of weight loss (% CI, . to . kg) weight and Obesity in Adults, which states that “a
compared with . kg of weight loss in the low-fat diet variety of dietary approaches can produce weight
groups (% CI, . to . kg) (). loss in overweight and obese adults, and that the
The POUNDS Lost Study (the largest trial ex- choice should be based on the patient’s preferences
amining macronutrient composition and weight and health status” ().

(a) Red: Adherers to fat & protein month 6 weight changes (b) Gold: Adherers to fat & protein month 6 weight changes
5 5
Weight change (kg)

Weight change (kg)

-20 -15

-25 -20

-30 -25
0 5 10 15 20 25 0 5 10 15 20 25 30

Individuals Individuals

(c) Purple: Adherers to fat & protein month 6 weight changes (d) Blue: Adherers to fat & protein month 6 weight changes

Weight change (kg)

Weight change (kg)

-5 -10

-10 -20

-15 -30

-20 -40
0 5 10 15 0 5 10 15 20

Individuals Individuals

Panel (a) (n = 38) is the adequate-protein/low-fat group (15% protein, 20% fat, diet, 65% carbohydrate); Panel (b) (n = 43) is the
high-protein/low-fat group (25% protein, 20% fat, 55% carbohydrate); Panel (c) (n = 28) is the high-protein/low-fat group (15% protein,
40% fat, 45% carbohydrate), and Panel (d) (n = 30) is the high-protein/high-fat group (25% protein, 40% fat, 35% carbohydrate).

Figure 5. Weight change from baseline to 6 months for each individual participant in the four dietary assignment groups ranked from
the largest loser on the left to the most weight gain on the right. (a) (n = 38) Adequate-protein/low-fat group (15% protein, 20% fat,
65% carbohydrate); (b) (n = 43) high-protein/low-fat group (25% protein, 20% fat, 55% carbohydrate); (c) (n = 28) high-protein/low-fat
group (15% protein, 40% fat, 45% carbohydrate); (d) (n = 30) high-protein/high-fat group (25% protein, 40% fat, 35% carbohydrate).

20 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

Commercial programs for weight loss a higher positive carbohydrate balance on day , were
In a meta-analysis, Gudzune et al. () reported that inactive, and ate an isocaloric high-carbohydrate diet
the Weight Watchers diet resulted in at least a .% gained less fat mass during a -year follow-up period
greater weight loss than those assigned to control/ ().
education after  months. The Jenny Craig diet
resulted in a .% greater weight loss during a - Future considerations/summary
month period vs groups receiving control/education Diets with many different macronutrient composi-
and counseling. The Nutrisystem diet resulted in tions can result in short-term weight loss. However,
a .% greater weight loss at  months vs control/ weight loss reaches a plateau within the first  to
education and counseling. VLCDs (Health Manage-  months. After that, weight is regained and often
ment Resources, Medifast, and Optifast) resulted in returns to baseline by  to  years.
a .% greater short-term weight loss than counseling, Maintenance of long-term weight loss is strongly
and the weight-loss effect lasted up to  months. The influenced by the ability to adhere to the dietary
Atkins diet (not technically a commercial program, but program. Behavioral support can significantly improve
one with affiliated diet products) resulted in .% to outcomes. There are variations among individuals in
.% greater weight loss at  months compared with the response to each diet, which are larger than the
counseling (). The differences in the amount of difference in mean weight loss between comparison
weight loss among various commercial diets were diets. Clinicians should consider genetic differences
relatively small, and the long-term effects of these diets regarding dietary response to weight loss, as person-
on weight control and chronic disease risk are still alized dietary regimens might improve the efficacy of
unclear. long-term weight-loss regimens.
Current data indicate that some (but not all) in-
Maintenance of long-term weight loss dividuals can achieve modest long-term weight loss
As previously discussed and illustrated in the Diabetes with any one of the diets evaluated herein. Additional
Prevention Program (, ) and the Look AHEAD research is needed to identify optimal diets for weight
trial (), maintaining weight loss is a challenge. control and long-term health, which should extend
One study (, ) assigned participants to beyond macronutrient composition and examine
weight loss with a VLCD for  weeks before ran- food quality and overall dietary patterns, as well as
domizing them to either a control diet or study diet factors that can improve long-term compliance. The
supplemented with . g/d of protein. At the end of Nurses Health Study and Health Professionals
 months, the group receiving the protein supplement Follow-up Study reported that improving diet quality
(to bring protein to % of total energy) had a % was associated with less weight gain, especially in
reduction in body-weight regain. younger women or individuals who are overweight
Data from the Women’s Health Initiative indicate ().
that reducing dietary fat intake may be of value for
long-term weight maintenance (, ). The study Exercise in managing obesity
reported that body weight in the low-fat diet group
and the control-diet group was similar after an average Introduction
of . years of follow-up (). However, those who There is a significant body of evidence supporting the
maintained the lowest quintile of fat intake were . kg effect of physical activity in both short-term and long-
lighter compared with those in the top quintile of fat term weight loss in adults (, , –).
intake, who were . kg heavier after  years. A recent The main components of energy expenditure (by
comprehensive meta-analysis indicated that long-term order of magnitude) are resting energy expenditure,
effects of low-fat diets on body weight depended on physical activity, and the thermic effect of food.
the intensity of intervention in the comparison group. Resting energy expenditure is the amount of energy
When compared with other dietary interventions of required for a -hour period by the body during
similar intensity, evidence from RCTs does not sup- resting conditions. Physical activity is composed of
port low-fat diets over other dietary interventions both nonexercise activity thermogenesis and ther-
(). mogenesis due to volitional activity of muscle groups.
The National Weight Control Registry identifies The thermic effect of food is the amount of energy
additional strategies for maintaining weight loss (), (above the resting rate) used for processing and storing
which include engaging in higher levels of physical food.
activity (e.g.,  to  min/wk), eating a low-fat, low- Energy expenditure from physical activity is di-
calorie diet ( to  kcal/d for women), and rectly related to body weight. However, it is unclear to
weighing themselves frequently (once a week or more) what extent reductions in energy expenditure from
(, ). physical activity relate to the epidemic of obesity that
Prediction of weight gain may also be related to the has developed during the last  years. Most mea-
ability to metabolize carbohydrates. Subjects who had surements of energy expenditure are not precise or

doi: 10.1210/er.2017-00253 21

Downloaded from
by guest
on 17 March 2018

easy to use. Therefore, reliable longitudinal data are Sedentary behavior

lacking. There is keen interest in the influence of sedentary
Two recent studies have concluded that the current behavior on a variety of health-related outcomes,
epidemic of obesity is more the result of an increase of including overweight and obesity. Energy expenditure
energy intake than a decrease in energy expenditure in occupational activities has declined by ~ kcal/
(–), but this is not the universal opinion (). d since  in the United States, and this reduction in
energy expenditure accounts for a significant portion
Genetic factors of physical activity of the increase in mean U.S. body weights for women
There is an important genetic component associated and men since  ().
with the extent to which individuals engage in physical Much of the early literature in this area focused on
activity (). In a study examining regular exercise the association between television viewing as an in-
among identical and fraternal twins that included both dicator of sedentary behavior and the risk of obesity.
same and opposite sex pairs, environmental factors Television viewing is positively associated with the risk
shared by children at age  accounted for % to % of gaining weight and the development of obesity (,
of sports participation, whereas genetic differences ).
provided almost no contribution. By age  to  the
genetic influences represented % of the variance in Treatment of patients who are overweight or obese
the level of participation in sports, and by age  to , using exercise with and without diet
genetic factors accounted for almost all (%) of the Studies on obesity have evaluated exercise as a sole
differences in participation in sports (, ). treatment, in combination with diets, and as a way to
maintain weight loss. Östman et al. () performed
Resistance vs aerobic exercise a Medline search for studies related to physical exercise
Although most research on the effects of physical and overweight and identified six relevant RCTs. Five
activity on body weight has focused on aerobic types of had a treatment interval of  months, and all had
physical activity, there is also evidence suggesting that a dropout rate of ,%. Table  (–) has been
resistance exercise may have some effect on weight adapted from this study with the addition of two
loss. Resistance exercise may influence body weight by newer trials, one  months long and one  months
increasing lean body mass, which will result in an long. The effects from diet are significantly greater
increase in resting metabolic rate. Resistance exercise than those from exercise, but increasing physical ac-
also improves one’s strength, which may result in more tivity may have important benefits on improving BP
free-living physical activity and thus increased total and cardiometabolic risk factors.
daily energy expenditure (). However, the vast
majority of data indicate that resistance exercise only
results in minimal reductions in body weight or body Behavioral Therapy in Managing Obesity
fatness (–).
Behavioral modifications and/or lifestyle interventions
Vigorous vs moderate exercise have been an important part of weight-loss programs
A study of . adults from the U.S. National for more than half a century (–). Data from two
Health and Nutrition Examination Survey showed large RCTs, the Look AHEAD trial and the Diabetes
that greater physical activity was associated with Prevention Program, support the efficacy of these
a lower BMI (). This relationship only existed with approaches. These studies are the gold standard and
moderate- to vigorous-intensity physical activity and are notable for the frequency of contact, the emphasis
not with low-intensity physical activity. These data on individualizing therapy, and the long-term em-
imply that there is an intensity threshold of physical phasis on maintaining weight loss. Fig.  shows data
activity that is necessary to affect body weight and from the Look AHEAD trial. The best outcomes are
prevent excessive weight gain. with frequent, face-to-face interventions. However,
incorporating this in primary care is challenging.
Physical activity declines with age In a meta-analysis of behavioral weight-loss pro-
Despite the benefit of physical activity in weight loss, grams, LeBlanc et al. () reported a mean weight loss
physical activity appears to decline during adolescence of 2. kg (% CI, 2. to 2. kg) favoring the
and remains low in most adults (, ). In a lon- behavioral strategy, but the range of mean values was
gitudinal study of adolescent girls, the level of activity quite large (2. to 2. kg).
declined in both black and white girls each year during Lifestyle interventions may also be effective for
adolescence. By age , black girls engaged in almost preventing weight regain (, ). Patients who
no spontaneous physical activity and white girls only participated in group sessions every other week for
engaged in very modest amounts of spontaneous  year after weight reduction maintained  kg of their
physical activity (). We do not have a comparable . kg end-of-treatment weight loss (). The most
study in adolescent males. successful patients monitor their weight frequently

22 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

and respond quickly to small increases in weight (). reduction in initial weight, a criterion for clinically
This can be daily or several times a week, but some meaningful weight loss (). Individuals with the best
daily variation (2. to . kg) is to be expected from attendance and greatest consistency in keeping self-
fluctuations in body water. monitoring records achieve the largest weight losses
Lifestyle methods
New developments in the delivery of
Self-monitoring behavioral treatment
Self-monitoring involves recording the type and
amount of foods and beverages consumed, along with Telephone-delivered programs
their calorie content and weight gain. Self-monitoring Sherwood et al. () demonstrated that during a -
helps patients identify their eating patterns (including month period, patients who received  intervention
times and places associated with consumption) and session phone calls lost an average of . kg; those who
also helps patients select targets for reducing calorie received  intervention calls lost . kg, and those
intake () (Table ). who were self-directed lost . kg. Appel et al. ()
reported that a group that received weekly telephone
Stimulus control coaching for  months, an Internet program for re-
Techniques of stimulus control teach patients to cording food intake and physical activity, and monthly
manage external cues, such as the sight or smell of coaching for an additional  months lost a mean of
food, as well as times, places, and events associated . kg at  months. The weight loss was generally well
with eating (, ). By decreasing exposure to maintained at  months (. kg) and was not sig-
problem foods, patients are less likely to overeat. nificantly different than what another group achieved
using an intensive in-person intervention (. kg at
Goal setting  months) ().
Goal setting helps patients make objective, measurable Perri et al. () demonstrated that women who
changes in eating, activity, and related behaviors (, were enrolled in extended-care programs that included
). They are guided in setting specific targets for problem-solving counseling delivered in  biweekly
calorie intake, minutes of physical activity, and fre- sessions via telephone or face-to-face regained only
quency of self-monitoring. . kg in  year of treatment, vs . kg for those in
a newsletter-only group.
Problem solving Several studies that used structured dietary in-
Problem solving teaches patients to analyze chal- terventions (i.e., meal replacements and/or portion-
lenges they have in adhering to their diet and ac- controlled entrees) reported roughly equivalent weight
tivity prescriptions (, , ). Patients learn to losses when the same behavioral intervention was
identify a number of possible solutions to the delivered in person or by telephone (–).
problem, pick the most promising one, and then
implement it. They learn to identify cognitive dis- Digitally-delivered programs
tortions (e.g., “I will never be able to lose weight Tate et al. () demonstrated that patients who were
because I ate that dessert”) and to replace them with provided with a directory of Internet resources for
rational responses (e.g., “One hundred fifty calories weight management and also received  weekly
of cake is not going to hinder my weight loss, lessons over  months via e-mail (where patients
particularly if I walk after dinner”) (). It is im- submitted their food and activity records online and
portant for patients to remember that the  received online feedback from an interventionist) lost
calories needs to be “subtracted” in the future either . kg vs patients who only received the directory. In
with exercise or by reducing intake of some other a -year follow-up study, Tate et al. () demon-
carbohydrate/fat-containing foods. strated that patients assigned to a low-intensity In-
ternet intervention with the addition of weekly
Short-term efficacy behavioral counseling lost . kg, whereas those re-
The structured behavioral programs, as described ceiving only the low-intensity Internet intervention
above, produce an average loss of  to  kg in the first lost . kg.
 months but with great variability. Some lose no Harvey-Berino et al. () compared the same
weight; others lose .%. Seven to  kg weight loss is -session intervention provided either via Internet
generally equivalent to a reduction of % to % of or on site. In  months, the on-site program resulted
initial weight, because  kg is the average weight for in . kg weight loss vs . kg for the Internet-only
patients in many studies (, ). Patients require group.
a high-intensity intervention to achieve these losses; These studies underscore the importance of patients
lower intensity treatment is not as effective (). keeping records of their food intake and physical activity
Approximately % to % of patients achieve a $% and receiving feedback from a trained interventionist.

doi: 10.1210/er.2017-00253 23

Downloaded from
by guest
on 17 March 2018

Table 5. Clinical Trials of Exercise in Individuals Who Are Overweight or Obese

No. Patients/
Authors Inclusion Criteria Intervention Groups Duration No. Follow-up Results Comments

Wood et al., Men 120%–160% (1) Diet (21 kg/wk); fat 1 year (1) 51/42 (1) BW 27.2 kg TG and HDL cholesterol
1988 (297) overweight reduced by 30% improved

(2) Individual instruction (21 (2) 52/47 Fat 25.9 kg Diet and physical activity
kg/wk); exercise (60%–80% yield the same reduction
of maximal physical in weight and fat at the
capacity, 40–50 min, three same negative calorie balance
to four times per week)

(3) Control (3) 52/42 (2) BW 24.0 kg

Fat 24.2 kg

(3) BW +0.6 kg

Fat 20.3 kg

Wing et al., Women: 30–60 y with (1) Diet (21000 kcal/d) + 12 mo (1) 12 (1) BW 27.9 kg HbA1C was reduced and
1988 (298) T2DM; .20% above walking (3 miles, three medications were
ideal weight times per week) reduced in groups 1 and 2

(2) Free diet + walking (2) 15 (2) BW 27.9 kg

(3 miles, four times
per week)

(3) Diet + stretching (3) 13 (3) BW 23.8 kg

Wood et al., Men and women: (1) Diet (moderate reduction Bottom of (1) 87/71 Men: BP decreased in groups 1 and 2
1991 (299) 25–49 y; overweight of energy, fat, cholesterol) form (both men and
120%–160% women)

(2) Diet (as above) + exercise (2) 90/81 (1) BW 25.1 kg Cholesterol decreased in
(60%–80% of maximal groups 1 and 2 (women)
physical capacity, 25-45 min,
three times per week)

(3) Control (3) 87/79 (2) BW 28.7 kg HDL cholesterol increased

in group 2 (both men and

(3) BW +1.7 kg TG decreased in group 2


(1) BW 24.2 kg

(2) BW 25.5 kg

(3) BW +1.3 kg

Svendsen et al., Women: 49-58 y; (1) Diet (4.2 MJ/d = 2 wk with (1) 51/47 12 wk: TG decreased and
1994 (300) BMI 25–42 kg/m2 1000 kcal/d) 6 mo
(2) Diet (as above) + exercise (2) 49/47 (1) BW 26.6 kg HDL cholesterol increased
(submaximal aerobics and
body building)

(3) Control (3) 21/16 (2) BW 210.9 kg There was no effect from
physical activity
(3) BW 0.0 kg

6 mo:

(1) BW 28.0 kg

(2) BW 28.0 kg

(3) BW 0.0 kg
(Continued )

24 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

Table 5. Continued
No. Patients/
Authors Inclusion Criteria Intervention Groups Duration No. Follow-up Results Comments

Pritchard et al., Men: overweight mean (1) Diet (2500 kcal/d) + 12 mo 66/60 (1) BW 26.3 kg Diet was “self-controlled”
1997 (301) BMI 29 kg/m2 low fat

(2) Exercise (65%–75% of (2) BW 22.6 kg

maximal physical capacity,
45 min, three to seven
times per week)

(3) Control (3) BW +0.9 kg

Irwin et al., Overweight nonsmoking (1) Exercise [at least 45 min, 12 mo (1) 87/84 (1) 3 mo: Participants were advised
2003 (302) postmenopausal moderate intensity, 5 d/wk, to maintain usual diet
women age 50–75 y 12 mo (months 1–3 they
with a BMI . 25 kg/m2 attended three sessions
or BMI 24–25 and per week; months 4–12
body fat . 33% by they attended one session
DXA who were per week)]
sedentary at baseline
(,60 min/wk of (2) Stretching (weekly sessions (2) 86/86 BW 20.5 Weight loss was related
moderate to vigorous of 45 min for 12 mo) to degree of exercise
activity) and maximal 12 mo:
oxygen uptake of
,25 mL/kg/min BW 21.3 kg

Fat 21.4 kg

VAT 28.5 cm2

(2) 3 mo:

BW 0.0 kg

12 mo:

BW 0.1 kg

Fat 20.1 kg

VAT 0.1 cm2

Donnelly et al., Overweight men and (1) Exercise (400 kcal/d, 5 d/wk 16 mo of (1) 87/41 (1) Men Exercise produced weight
2003 (303) women age 17–36 y with walking on a treadmill verified loss in men and prevented
with a BMI of at 55%–70% of maximal exercise weight gain in women
25.0–34.9 kg/m2 physical capacity uptake)

(2) Control (2) 44/33 BW 25.2 kg

Fat 24.9 kg

VAT 222.4 cm2


BW +0.4 kg

(1) Fat 20.2 kg

VAT 23.2 cm2

(2) Men

BW 20.5 kg

Fat 20.7 kg

VAT 26.3 cm2

(Continued )

doi: 10.1210/er.2017-00253 25

Downloaded from
by guest
on 17 March 2018

Table 5. Continued
No. Patients/
Authors Inclusion Criteria Intervention Groups Duration No. Follow-up Results Comments


BW +2.9 kg

Fat +2.0 kg

VAT +3.1 cm2

Slentz et al., Men and women age (1) Exercise [high amount/ 8 mo of (1) 44/17 (1) BW 23.5 kg Subjects were counseled not
2004 (304) 40–60 y and BMI of vigorous intensity, observed to change diets and
25–35 kg/m2 and mild calorically equivalent to exercise encouraged to maintain
to moderate lipid ~20 miles (32.0 km) of body weight.
abnormalities jogging per week at
65%–80% maximal
physical capacity]

(2) Exercise [low amount/ (2) 52/24 Fat 4.9 kg

vigorous intensity,
equivalent to ~12 miles
(19.2 km) of jogging per
week at 65%–80%]

(3) Exercise [low amount/ (3) 42/14 Waist 23.4 cm

moderate intensity,
equivalent to ~12 miles
(19.2 km) of walking per
week at 40%–55%]

(4) Control (4) 44/7 (2) BW 21.1 kg

Fat 22.6 kg

Waist 21.4 cm

(3) BW 21.3 kg

Fat 22.0 kg

Waist 21.1 cm

(4) BW 1.1 kg

Fat 0.5 kg

Waist 0.8 cm

Abbreviations: BW, body weight; DXA, dual x-ray absorptiometry; TG, triglyceride; VAT, visceral adipose tissue.
Adapted and updated from Östman et al., 2004 (296).

Educational instruction (i.e., information) alone is not application revealed essentially no weight-loss dif-
sufficient to induce clinically meaningful weight loss. ference between the two approaches during  months
These studies also suggest that the most successful ().
Internet programs are those in which therapists provide
weekly e-mail feedback to patients. However, on-site
behavioral programs still provide better results (). Medication in Managing Obesity
The reduced efficacy of Internet programs, how-
ever, is offset by the potentially greater accessibility and Early history
affordability of this approach, compared with tradi- Medications for managing obesity have a long and
tional behavioral treatment. checkered history (). Treatment in the th century
Despite their popularity, little is known about the included soap (, ) and vinegar mixed with
effectiveness of smart-phone applications for weight a number of purgatives (). Some treatments also
management. A recent study that compared usual used tobacco, a strategy people still use today to
primary care with or without the MyFitnessPal prevent weight gain.

26 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

In the late th and early to mid-th century, release (PHEN/TPM ER), and the combination of
three major groups of medications came into use: naltrexone and bupropion sustained release (SR).
thyroid hormone, dinitrophenol, and amphet- The second group consists of older, sympatho-
amine. Clinicians prescribed both thyroid extract mimetic drugs that are FDA approved for short-term
and dinitrophenol (a product of the aniline use, usually considered , weeks. The FDA did not
dye industry) until negative side effects became use modern standards to evaluate these “short-term”
evident (). medications for safety and efficacy. The FDA approved
Amphetamine became popular after  when them using only data from small, short-term studies,
Nathanson () noted that  of  patients treated and there are no cardiovascular outcome studies for
with amphetamine for narcolepsy had marked loss of these agents.
appetite and weight. However, the abuse potential of Importantly, note that in all the clinical trials eval-
amphetamines soon became apparent (), and cli- uating these agents, the drug-vs-placebo study also in-
nicians stopped prescribing them as a way to manage cluded lifestyle interventions, such as diet and/or exercise,
obesity. which contribute to the overall weight loss reported.
Aminorex, another member of the amphetamine- Also important to note, these drugs are all con-
like group, emerged in Austria and Switzerland in traindicated for pregnant women, as is weight loss per
, but it was removed from the market in  due se. Because weight loss can increase fertility, all women
to associated pulmonary hypertension (). Table  in a weight-management program that use medica-
lists several drugs for obesity management that were tions should be cautioned about the need for con-
associated with significant detrimental side effects traception. If pregnancy does occur while a patient is
(). taking any of these medications, the patient should
From the end of World War II through , there immediately stop the medication and contact a med-
was considerable research on monoaminergic drugs. ical professional.
Researchers discovered that injecting norepinephrine Listed below are brief assessments of these drugs’
into the central nervous system of experimental ani- action, efficacy, and safety. More detailed information
mals reduced food intake and activated thermogenesis. is in Figs.  () and  and Table .
This resulted in a search for thermogenic drugs that
could work through monoaminergic receptors. Orlistat
During this period, researchers also synthesized Orlistat is a potent and selective inhibitor of pancreatic
many derivatives of amphetamine for treating obesity lipase that reduces intestinal digestion of fat. One
(), along with serotonergic drugs and multiple clinical trial resulted in weight loss of % of body
monoamine reuptake inhibitors. weight at  year, compared with ~.% in the placebo

More recent drug development:

continuing difficulties 0
The discovery of leptin in  () marks the be- DSE
ginning of modern approaches to identifying drugs for
treating obesity. Leptin is a peptide made primarily in -2
Percent reduction in initial weight

adipose tissue. Its absence is associated with massive -2.1%

obesity in animals and human beings. Treatment with ILI
leptin reverses the obesity caused by leptin deficiency,
indicating that there is a clear-cut molecular–genetic
mechanism and a highly effective treatment of at least -4.7%
one type of obesity. However, because leptin failed to
show adequate weight loss in obese persons who are P < 0.001 for comparisons at all years
not leptin deficient, trials were stopped (, ). The -8
discovery of leptin opened a flood of research to
discover new treatments, some of which were with-
drawn from the market due to health risks (). –10
0 1 2 3 4 5 6 7 8

Medications approved by the FDA for Years

treating obesity
In Table  () we list medications that are FDA Abbreviations: DSE, diabetes support and education; ILI, intensive lifestyle intervention
approved for weight management in patients with © 2018 Endocrine Reviews ENDOCRINE SOCIETY

obesity and divide them into two groups. First are the Figure 6. Mean (6SE) weight losses during 8 years for participants randomly assigned to an
agents approved for long-term treatment of obesity. intensive lifestyle intervention or diabetes support and education (usual-care group). Differences
These include orlistat, lorcaserin, liraglutide, the between groups were significant (P , 0.001) at all years. DSE, diabetes support and education; ILI,
combination of phentermine/topiramate extended intensive lifestyle intervention.

doi: 10.1210/er.2017-00253 27

Downloaded from
by guest
on 17 March 2018

Table 6. Key Components of Comprehensive Behavioral Weight-Loss Interventions to Achieve a 7% to 10% Weight Loss
Component Weight Loss Weight-Loss Maintenance

Frequency and duration of treatment • Weekly contact, in person or by telephone, for • Every-other-week contact for 52 wk (or longer)
contact 20–26 wk (Internet/e-mail contact yields smaller
weight loss)

• Group or individual contact • (Monthly contact likely adequate)

• Group or individual contact

Dietary prescription • Low-calorie diet (1200-1500 kcal for those ,250 pounds; • Consumption of a hypocaloric diet to maintain
1500–1800 kcal for those $250 pounds reduced body weight

• Typical macronutrient composition: #30% fat (#7% • Typical macronutrient composition similar to that
saturated fat), 15%–25% protein, remainder from for weight loss
carbohydrate (diet composition based on individual
needs or preferences)

Physical activity prescription • 180 min/wk of moderately vigorous aerobic activity • 200-300 min/wk of moderately vigorous aerobic
(e.g., brisk walking), strength training also desirable activity (e.g., brisk walking), strength training also

Behavior therapy prescription • Daily monitoring of food intake and physical activity • Occasional to daily monitoring of food intake and
by use of paper or electronic diaries physical activity by use of similar diaries

• Weekly monitoring of weight • Twice weekly to daily monitoring of weight

• Structured curriculum of behavior change (e.g., Diabetes • Curriculum of behavior change, including relapse
Prevention Program) prevention and individualized problem solving

• Regular feedback from an interventionist • Periodic feedback from an interventionist

group () (Fig. ). Another study achieved a weight studies using lorcaserin (Table ), the maximal weight
loss of % compared with % in the placebo-treated loss (by modeling) was 2. kg, and half the maximal
group and a reduction of % in the development of effect occurred by . weeks (). They also showed
TDM in patients who had impaired glucose tolerance improvements in cardiovascular risk factors (–).
(). In a meta-analysis of  studies using orlistat In preclinical toxicology studies in rats, there were
(Table ), the maximal weight loss (by modeling) more brain and mammary tumors. This may reflect
was 2. kg, and half the maximal effect occurred by the fact that the drug does not reach the high con-
. weeks (). centrations in the central nervous system of human
Orlistat is the only medication the FDA approved beings that is does in rats ().
for weight management in adolescents with obesity
(). Adherence to orlistat use falls off rapidly after Liraglutide
initial prescription (). Orlistat can cause small but Liraglutide is a GLP- agonist that has a % ho-
significant decreases in fat-soluble vitamins, and cli- mology to GLP-. The molecular change extends the
nicians should advise patients to take vitamin sup- circulating half-life from  to  minutes to  hours.
plements. Rare cases of severe liver injury have been Clinicians prescribe this drug in combination with
reported with patients taking orlistat. A causal re- a reduced-calorie diet and increased physical activity
lationship has not been established, but patients who for chronic weight management in adult patients with
take orlistat should contact their health care provider if an initial BMI of . kg/m or in adult patients with
itching, jaundice, pale color stools, or anorexia develop a BMI of . kg/m who have TDM, hypertension,
(). or dyslipidemia.
One study () that administered daily sub-
Lorcaserin cutaneous injections of liraglutide at ., ., ., or
Lorcaserin selectively targets the serotonin-c re- . mg produced mean weight losses of ., ., .,
ceptors to reduce food intake (), but it has low and . kg, respectively, after  year of treatment,
affinity for the serotonin-b receptors on heart valves. compared with a loss of . kg in the placebo-treated
The three clinical studies that provided the data for group and . kg in the orlistat-treated comparator
lorcaserin’s approval reported modest weight loss (see group. Another larger trial reported that after
Fig.  for one of these trials). In a meta-analysis of five  weeks, liraglutide reduced body weight by . kg

28 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

compared with . kg in the placebo-treated group (on nonfatal myocardial infarction, or nonfatal stroke
average) () (see Fig. ). In another trial (), those (). Semaglutide lowered the rate of cardiovascular
receiving liraglutide for weight maintenance (after death, nonfatal myocardial infarction, or nonfatal
initially losing weight from a low-calorie diet) lost an stroke ().
additional . kg compared with no additional weight
loss in the placebo group. Furthermore, only about PHEN/TPM ER
half of the placebo group was able to maintain the PHEN/TPM ER has lower doses of phentermine than
weight they lost due to diet. In a meta-analysis of three clinicians usually prescribe for phentermine alone.
studies using liraglutide (Table ), the maximal weight Phentermine acts to reduce appetite through in-
loss (by modeling) was 2. kg, and half the maximal creasing norepinephrine in the hypothalamus. Top-
effect occurred by . weeks (). iramate may reduce appetite through its effect on
Liraglutide is contraindicated in people with GABA receptors.
a family history of medullary thyroid carcinoma or Two clinical studies (, ) provided the efficacy
multiple endocrine neoplasia syndrome type . Cli- and safety data for the approval of PHEN/TPM ER
nicians should not prescribe liraglutide for patients () (see Fig.  for one of these trials). The patients in
with a history of pancreatitis and should discontinue these two studies had higher risk profiles due to excess
liraglutide if acute pancreatitis develops. If weight loss weight. PHEN/TPM ER produced weight losses of
does not exceed % by  weeks, patients should stop .% and .% with the middle and high doses, re-
taking liraglutide. Two cardiovascular outcome trials spectively, compared with .% in the placebo group.
studied liraglutide (. mg/d) () and the long-acting This weight loss is larger than observed in clinical trials
version, semaglutide (. or . mg weekly) (). In with single drugs (). In a meta-analysis of six
patients with TDM, liraglutide lowered the rate of the studies using phentermine/topiramate (Table ), the
first occurrence of death from cardiovascular causes, maximal weight loss (by modeling) was . kg, and

Table 7. Some Medications Used in the Past for Managing Body Weight That Were Withdrawn or Are Not Approved in the
United States
Year Introduced or
Drug Withdrawn Comments

Thyroid 1892 Mimics endogenous thyroxine/triiodothyronine

Associated with tachycardia and increase in metabolic rate

Dinitrophenol 1932 Uncouples oxidative phosphorylation

Associated with cataracts, neuropathy, and death

Amphetamine 1937 Noradrenergic-dopaminergic drug

Associated with recreational abuse and pulmonary hypertension

Aminorex 1965 Noradrenergic drug

Associated with pulmonary hypertension

Fenfluramine, 1997 Serotonergic drugs

Both associated with cardiac valvulopathy and primary pulmonary

Phenylpropanolamine 1998 Noradrenergic agonist

Associated with strokes and cardiovascular deaths

Ephedra alkaloids 2003 Noradrenergic drugs

Associated with heart attacks, strokes, and death

Rimonabant 2008 Cannabinoid receptor antagonist

Associated with depression and suicidality

Sibutramine 2010 Norepinephrine-serotonin reuptake inhibitor

Associated with elevated BP and death

doi: 10.1210/er.2017-00253 29

Downloaded from
by guest
on 17 March 2018

half the maximal effect occurred by . weeks (some  days of treatment with monoamine oxidase in-
of which was related to the titration schedule) (). hibitors and in patients with hypersensitivity to any of
Improvements in BP, glycemic measures, HDL the ingredients in the medication. Topiramate is
cholesterol, and triglycerides occurred with both the a carbonic anhydrase inhibitor that often produces tin-
recommended and the top doses of the medication in gling in the fingers and may change the taste for car-
these trials (, ). Improvements in risk factors bonated beverages. Other potential issues include risk of
were related to the amount of weight loss. In patients kidney stones (associated with topiramate) and increased
with OSA, this combination reduced the severity of heart rate in patients susceptible to phentermine.
symptoms ().
Taking topiramate in the first trimester of preg- Naltrexone/bupropion combination
nancy may increase risk of cleft lip/cleft palate in Bupropion is approved as a single agent for depression
infants. Therefore, clinicians must inform women of and for smoking cessation. It reduces food intake by
childbearing potential of this risk and conduct acting on adrenergic and dopaminergic receptors in
a pregnancy test before prescribing PHEN/TPM ER. the hypothalamus. It has a modest effect on weight
Glaucoma is a rare side effect of topiramate, and the loss. Bupropion stimulates the pro-opiomelanocortin
drug is contraindicated in glaucoma. PHEN/TPM ER neurons in the hypothalamus to produce pro-
is also contraindicated in hyperthyroidism within opiomelanocortin, which is further processed to

Table 8. Drugs Approved by the FDA for Managing Patients With Obesity
Generic Name (Year of Approval) Trade Name(s) Dosage DEA Schedule

Pancreatic lipase inhibitors FDA approved for long-term use

Orlistat (1999) Xenical 120 mg, three times daily before meals Not scheduled

Orlistat (2007) Alli (over-the-counter) 60 mg three times daily before meals Not scheduled

Serotinin-2C receptor agonists FDA approved for long-term use

Lorcaserin (2012) Belviq 10 mg, two times daily or 20 mg/d IV

Glucagon-like peptide-1 agonists FDA approved for long-term use

Liraglutide (2015) Saxenda 3 mg/d: begin at 0.6 mg/d for week 1 and increase by Not scheduled
0.6 mg/d each week to reach 3 mg/d at week 4

Combination drugs FDA approved for long-term use

PHEN/TPM ER (2012) Qsymia 3.75 mg/23 mg IV

7.5 mg/46 mg

11.25 mg/69 mg

15 mg/92 mg

Naltrexone SR/bupropion SR (2014) Contrave 8 mg/32 mg tablets: one in AM for week 1, one in AM and one Not scheduled
in PM for week 2, two in AM and one in PM for week 3,
two in AM and two in PM for week 4

Noradrenergic drugs FDA approved for short-term use

Diethylpropion (1959) Tenuate 25 mg, three times daily IV

Tenuate dospan 75 mg, every morning

Phentermine (1959) Adipex and many others 15–30 mg/d IV

Benzphetamine (1960) Didrex 25–50 mg, three times daily III

Phendimetrazine (1959) Bontril 17.5–70 mg, three times daily III

Prelu-2 105 mg/d

Abbreviation: DEA, Drug Enforcement Agency.

See Bray and Ryan, 2012 (330).

30 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

Figure 7. Diagram of the sites within the central nervous system where medications can have their effects. See Apovian et al., 2015

Anorexigenic signaling Anorexigenic signaling Orexigenic signaling

Paraventricular Lateral
nucleus hypothalamic area
Nucleus accumbens neuron
D1 D2
Y1/Y5R MC3/4R MC3/4R Y1/Y5R

Dopamine NPY
AGRP neuron AGRP
neuron Serotonergic
DAT Serotonin
Bupropion NPY
Dorsal Pramlintide nucleus
vagal LepR LepR
LepR complex
Y4R GLP1R Exenalide, Liraglutide Lorcaserin HCl
Leptin Leptin
Vagal afferents
Receptors: Stimulating Inhibitory
© 2018 Endocrine Reviews ENDOCRINE SOCIETY

produce both a-melanocyte stimulating hormone cardiometabolic risk markers, weight-related quality of
(which reduces food intake) and b-endorphin (which life, and control of eating.
stimulates feeding). Naltrexone blocks this effect of Finally, naltrexone/bupropion use in patients with
b-endorphin, thus allowing the inhibitory effects of TDM resulted in significantly greater weight re-
a-melanocyte stimulating hormone to reduce food duction (.% vs .% in the placebo group) and
intake by acting on the melanocortin- receptor significantly greater reductions in HbAc (2.
system (). vs 2.%; P , .) (). There was also im-
Three studies of the combination drug naltrexone/ provement in triglycerides and HDL cholesterol
bupropion provided the basis for its approval. compared with placebo.
In one study (), weight loss at  weeks was .% Efficacy of weight loss with the naltrexone/
for a lower dose of naltrexone/bupropion ( mg per day/ bupropion combination at  year is higher than lor-
 mg per day) and .% for a higher dose ( mg per caserin but not as high as PHEN/TPM ER and is
day/ mg per day), compared with placebo. Treatment associated with improvements in risk factors (, ,
also improved waist circumference, fasting glucose, ). In a meta-analysis of six studies using naltrexone/
fasting insulin, homeostasis assessment model of insulin bupropion (Table ), the maximal weight loss (by
resistance (HOMA-IR), and HDL cholesterol, but there modeling) was 2. kg, and half the maximal effect
was a transient increase in BP. occurred by . weeks (probably related to the ti-
In a second study that included an intensive be- tration schedule) ().
havioral modification program (), weight loss at Because bupropion increases pulse and both
 weeks was about % for naltrexone/bupropion bupropion and naltrexone increase BP, an ongoing
( mg per day/ mg per day) vs about .% study is examining cardiovascular outcomes ().
(Fig. ) for placebo. The study also reported significant
improvements in weight, waist circumference, insulin, Comparison of medications approved for chronic
homeostatic model assessment of insulin resistance, weight management
HDL cholesterol, triglycerides, and quality of life. There are no head-to-head comparisons of these
In a third study, weight loss at week  was .% medications. However, there is an analysis of  RCTs
with naltrexone/bupropion ( mg per day/ mg per of weight-loss medications that included trials with
day) compared with .% with placebo (). As in orlistat, lorcaserin, liraglutide, naltrexone/bupropion,
the other studies, there were improvements in and PHEN/TPM ER. The inclusion criteria and

doi: 10.1210/er.2017-00253 31

Downloaded from
by guest
on 17 March 2018

Figure 8. Randomized controlled trial data showing weight loss with orlistat, lorcaserin, liraglutide, phentermine/topiramate, and
naltrexone/bupropion. NB, naltrexone/bupropion; Phen, phentermine; SE, standard error; SR, sustained release; tid, three times a day;
Top, topiramate.

Orlistat Lorcaserin Liraglutide

0 0
0 Placebo tid
Percent change in body weight (±SE)

Orlistat 120 mg tid

-2 -2

Weight change (%)

Weight loss (kg)
-4 -4 -5

-6 -6

-8 -8 -10
-10 -10 Cross-over Placebo lifestyle
Continuous Liraglutide lifestyle
-12 -12 -15
-10 0 10 20 30 40 50 60 70 80 90 100 110 0 8 16 24 32 40 48 56 64 72 80 88 96104 0 4 8 12 16 20 24 28 32 36 40 44 48 52 56
Week Weeks of treatment Weeks
Mildly hypocaloric Weight maintenance
diet (eucaloric) diet

Phentermine/Topiramate Naltrexone/Bupropion
0 0

Weight loss (%)
Weight loss (%)

-6 Phen 7.5 + Top 46 mg -4 Placebo

Phen 15.0 + Top 82.0 mg NB 16
-8 NB 32
-10 -15
0 3 18 24 32 40 48 68 0 8 16 24 32 40 48 56
Weeks of treatment Weeks of treatment

Abbreviations: NB, naltrexone/bupropion; Phen, phentermine; SE, standard error; SR, sustained release; tid, three times a day; Top, topiramate
© 2018 Endocrine Reviews ENDOCRINE SOCIETY

background lifestyle interventions differed across drugs that the FDA tested and approved before .
studies, so we must interpret results with caution. The U.S. Drug Enforcement Agency classifies phen-
Attrition rates were % to % across these trials. All termine and diethylpropion as schedule IV drugs and
five agents were associated with significantly greater benzphetamine and phendimetrazine as schedule III
weight loss at  year than placebo. Collectively, these drugs. This regulatory classification indicates the
studies reported a weight loss of .% in % of patients government’s idea that these drugs have the potential
treated with placebo, % of patients treated with orlistat, for abuse, although this potential appears to be low
% of patients treated with lorcaserin, % of patients (). These drugs are approved for only a “few weeks”
treated with naltrexone/bupropion SR, % of patients (usually  weeks).
treated with liraglutide, and % of patients treated with
PHEN/TPM ER. The highest odds ratio for treatment- Phentermine
related discontinuation of the trial was with liraglutide Efficacy of phentermine. The FDA approved
and naltrexone/bupropion (). phentermine as a single agent in , and it remains
the most commonly prescribed drug for weight loss in
Drugs approved by the FDA for short-term the United States (). There are few current data to
treatment of patients with obesity evaluate its long-term efficacy.
We group the sympathomimetic drugs benzphet- A -month study of phentermine reported that 
amine, diethylpropion, phendimetrazine, and phen- mg/d resulted in .% weight loss at  months
termine together, because they are noradrenergic compared with .% for placebo (). In another

32 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

-month study of phentermine, weight loss was .% Endocrinologists/American College of Endocrinol-
for phentermine at . mg/d and .% for phenter- ogy “comprehensive clinical practice guidelines for
mine at  mg/d compared with .% for the placebo medical care of patients with obesity” from 
group (). Finally, a study from Korea () re- () indicate that clinicians may consider phar-
ported that after  weeks, mean weight loss for macotherapy as a first-line treatment of weight
phentermine was . 6 . kg vs . 6 . kg for reduction if patients present with one or more se-
placebo patients. Weight loss with phentermine may vere comorbidities and would benefit from weight
not be greatly enhanced by increasing doses beyond loss of $%. Those guidelines do not require that
 mg (). patients fail lifestyle therapy before clinicians pre-
Safety of phentermine. Phentermine is part of scribe medications.
a group of drugs called sympathomimetic drugs. These
drugs produce central excitation, manifested as dry Medicating the patient for other chronic conditions
mouth, insomnia, or nervousness. This effect is most who is also overweight or obese
obvious shortly after the drug is started and wanes For patients who are overweight or obese, the 
substantially with continued use. Sympathomimetic Endocrine Society clinical practice guidelines on
drugs may also increase heart rate and BP. The pre- obesity pharmacotherapy () recommended that
scribing information usually recommends that the providers consider body weight when prescribing
drugs not be given to individuals with a history of medications for other chronic health conditions, so
CVD (–). that at-risk patients can avoid medications that pro-
Lacking good quantitative measures of the effects mote weight gain. The guideline recommends that
of sympathomimetic drugs on heart rate and pulse, we patients use medications that are weight neutral or
recommend caution in prescribing drugs in this group. associated with weight loss.
According to the Endocrine Society Guidelines (), In managing patients with obesity, the guideline
clinicians should not prescribe sympathomimetic also advises that providers review medications at every
drugs to persons with a history of CVD and elevated visit and discuss weight effects with patients, so that
BP. patients at risk for weight gain can share in the de-
cision process when choosing medications. Addi-
Best practices for medications approved for tionally, the guideline cautions against prescribing
weight management medications known to be associated with weight loss if
The  American Heart Association/American they have no proven beneficial effect on the patient’s
College of Cardiology/The Obesity Society “guide- other identified health issues ().
line for the management of overweight and obesity
in adults” () and the  Endocrine Society What is the current status of clinical adoption of
clinical practice guideline on obesity pharmaco- medications for chronic weight management?
therapy () both agree that clinicians may con- According to the Awareness, Care and Treatment in
sider prescribing weight-reducing drug therapies for Obesity Management study (), there are a number
patients who: () struggle to achieve weight goals, () of misconceptions regarding obesity shared by pro-
meet label indications (BMI .  kg/m or BMI . viders and patients alike, specifically that obesity is not
 kg/m with comorbidity), and () need to lose a disease, that patients have the primary responsibility
weight for health reasons (such as osteoarthritis, for their problem and for its treatment, that prevention
prediabetes, fatty liver, or other conditions). Fur- is more important than treatment, and that the risks of
thermore, the American Association of Clinical treatment should be low.

Table 9. Weight Loss Associated With Use of Orlistat, Lorcaserin, Liraglutide, Topiramate/Phentermine, and

Maximal Weight Weeks to Half Maximal

Drug/Placebo No. of Trials Loss (kg) Weight Loss (wk) Drop Rate (%)
Maximal weight loss is the
Orlistat 31 26.65 35.4 29.0 modeled maximal effect and
does not contain the placebo
Lorcaserin 5 25.39 19.3 40.9 effect.
Data are from Dong et al., 2017
Liraglutide 3 27.68 12.7 24.3 (334).

Topiramate/phentermine 6 215.6 29.8 34.9

Naltrexone/bupropion 6 213.2 35.2 49.1

Placebo 51 22.71 12.3

doi: 10.1210/er.2017-00253 33

Downloaded from
by guest
on 17 March 2018

Table 10. Complementary and Over-the-Counter Products Used for Weight Loss
Proposed Mechanism of
Ingredient Action Evidence of Efficacy Safety Concerns

Chromium Increases lean muscle mass; Several clinical trials of varying No safety concerns reported at recommended
promotes fat loss; and reduces methodological quality intakes (25-45 mg/d for adults)
food intake, hunger levels, and
fat cravings Research findings: minimal effect Reported adverse effects: headache, watery stools,
on body weight and body fat constipation, weakness, vertigo, nausea, vomiting,
and urticaria (hives)

b-Hydroxy b-methylbutyrate Metabolite of leucine, produced Used in conditions of muscle In humans, no reported adverse in young adults or
in 0.3 g/d, but taken in doses wasting and to augment older adults when b-Hydroxy b-methylbutyrate
of 30–60 g/d muscle in athletes is taken in doses of 3 g/d for up to 1 y

b-Hydroxy b-methylbutyrate
supplementation can preserve
lean muscle mass in older adults
(according to a 2015 meta-

Pyruvate Increases lipolysis and energy Few clinical trials of weak Few safety concerns reported
expenditure methodological quality

Research findings: possible minimal Reported adverse effects: diarrhea, gas, bloating,
effect on body weight and body and (possibly) decreased HDL levels

Conjugated linoleic acid Promotes apoptosis in adipose Several clinical trials Few safety concerns reported
Research findings: minimal effect Reported adverse effects: abdominal discomfort
on body weight and body fat and pain, constipation, diarrhea, loose stools,
dyspepsia, and (possibly) adverse effects on
blood lipid profiles

Calcium Increases lipolysis and fat Several large clinical trials No safety concerns reported at recommended
accumulation, decreases fat intakes (1000-1200 mg/d for adults)
Research findings: no effect on body Reported adverse effects: constipation, kidney
weight, weight loss, or prevention stones, and interference with zinc and iron
of weight gain based on clinical absorption at intakes .2000–2500 mg for adults

Green tea (Camellia sinensis) Increases energy expenditure Several clinical trials of good No safety concerns reported when used as a
and green tea extract and fat oxidation methodological quality studied beverage
green tea catechins with and
without caffeine

Reduces lipogenesis and fat Research findings: possible modest Contains caffeine
absorption effect on body weight
Some safety concerns reported for
green tea extract

Reported adverse effects (for green tea extract):

constipation, abdominal discomfort,
nausea, increased BP, liver damage

Green coffee bean extract Inhibits fat accumulation Few clinical trials, all of poor Few safety concerns reported but
(Coffea aribica, Coffea methodological quality not rigorously studied
canephora, Coffea robusta)
Modulates glucose metabolism Research findings: possible modest Contains caffeine
effect on body weight
Reported adverse effects: headache and
urinary tract infections
(Continued )

34 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

Table 10. Continued

Proposed Mechanism of
Ingredient Action Evidence of Efficacy Safety Concerns

Caffeine (as added caffeine Stimulates central nervous system Short-term clinical trials of Safety concerns not usually reported at doses
or from guarana, kola nut, combination products ,400 mg/d for adults, significant safety
yerba mate, or other herbs) Increases thermogenesis and concerns at higher doses
fat oxidation
Research findings: possible modest Reported adverse effects: nervousness,
effect on body weight or jitteriness, vomiting, and tachycardia
decreased weight gain over time

Forskolin (Plectranthus Activates the enzyme adenylyl cyclase One clinical trial Forskolin should be used with caution or avoided
barbatus) altogether in women who are pregnant
Increases intracellular levels of Research findings: oral ingestion
cyclic adenosine monophosphate of forskolin (250 mg of 10%
forskolin extract twice a day)
for a 12-wk period was shown
to favorably alter body
composition while concurrently
increasing bone mass and serum
free testosterone levels in
overweight and obese men

Fucoxanthin Increases energy expenditure Studied only in combination with No safety concerns reported but not rigorously
and fatty acid oxidation pomegranate seed oil in one trial studied
in humans

Suppresses adipocyte differentiation Research findings: insufficient Reported adverse effects: none known
and lipid accumulation research to draw firm conclusions

Hydroxycitric acid (Garcinia Inhibits lipogenesis Several short-term clinical trials of Few safety concerns reported
cambogia) varying methodological quality

Suppresses food intake Research findings: little to no effect Reported adverse effects: headache, nausea,
on body weight upper respiratory tract symptoms, and
gastrointestinal symptoms

Yohimbe (Pausinystalia Has hyperadrenergic effects Very little research on yohimbe Significant safety concerns reported
yohimbe, yohimbine) for weight loss

Research findings: no effect on Reported adverse effects: headache, anxiety,

body weight; insufficient research agitation, hypertension, and tachycardia
to draw firm conclusions

Hoodia (Hoodia gordonii) Suppresses appetite Very little published research Some safety concerns reported, increases heart
in humans rate and BP

Reduces food intake Research findings: no effect on Reported adverse effects: headache, dizziness,
energy intake or body weight nausea, and vomiting
based on results from one study

Raspberry ketone Alters lipid metabolism Studied only in combination with No safety concerns reported but not rigorously
other ingredients studied

Research findings: insufficient Reported adverse effects: none known

research to draw firm conclusions

Guar gum Acts as bulking agent in gut, delays Several clinical trials of good Few safety concerns reported with currently
gastric emptying methodological quality available formulations

Increases feelings of satiety Research findings: no effect on Reported adverse effects: abdominal pain, flatulence,
body weight diarrhea, nausea, and cramps

Bitter orange (synephrine) Increases energy expenditure Small clinical trials of poor Some safety concerns reported
and lipolysis methodological quality

Acts as a mild appetite suppressant Research findings: possible effect Reported adverse effects: chest pain, anxiety,
on resting metabolic rate and and increased BP and heart rate
energy expenditure; inconclusive
effects on weight loss
(Continued )

doi: 10.1210/er.2017-00253 35

Downloaded from
by guest
on 17 March 2018

Table 10. Continued

Proposed Mechanism of
Ingredient Action Evidence of Efficacy Safety Concerns

Chitosan Binds dietary fat in the digestive Small clinical trials, mostly of Few safety concerns reported, could cause
tract poor methodological quality allergic reactions

Research findings: minimal effect Reported adverse effects: flatulence, bloating,

on body weight constipation, indigestion, nausea, and heartburn

Glucomannan Increases feelings of Several clinical trials of varying Significant safety concerns reported with
satiety and fullness methodological quality, mostly tablet forms, which might cause esophageal
focused on effects on lipid and obstructions, but few safety concerns with
blood glucose levels other forms

Prolongs gastric emptying time Research findings: little to no Reported adverse effects: loose stools, flatulence,
effect on body weight diarrhea, constipation, and abdominal discomfort

White kidney bean (Phaseolus Interferes with breakdown Several clinical trials of varying Few safety concerns reported
vulgaris) and absorption of carbohydrates methodological quality
by acting as a “starch blocker”
Research findings: possible modest Reported adverse effects: headache, soft
effect on body weight and body stools, flatulence, and constipation
See National Institutes of Health, Office of Dietary Supplements, 2015 (365).

At present, the FDA has approved nine agents (five approves pharmaceutical preparations for patients
for long-term use and four for short-term use). For who are overweight or obese.
newer drugs, the time since approval of these medi- Blanck et al. () reported that .% of adults
cations is too short to know whether and how they will (.% of women and .% of men) have used
be used. However, older data (which predate the a weight-loss supplement, and .% have used one in
current medication landscape) indicate there are some the past year (.% of women and .% of men).
serious concerns about how diet medications are used, Almost % (.%) used them for $ months.
such as: patients using prescription weight-loss pills Pillitteri et al. () reported that females, the
who do not meet the BMI criterion for these medi- young, the less educated, and those with lower in-
cations; family, friends, and other nonphysicians comes are more likely to use these products. Many
providing medications; the use of nonprescription diet respondents thought that dietary supplements are
products; using pills after they were withdrawn from safer than prescription drugs, and many overestimated
the market; low -year persistent use rates; and co- the degree of regulatory screening of these products.
using narcotic and antidepressants (, , , ). Clinicians should be aware and knowledgeable
about these products when they begin discussing
Dietary supplements, over-the-counter products, weight management with patients, since patients have
and other treatments with unproven efficacy and likely taken them or may currently be taking them.
unknown safety Table  () provides a list of herbal and comple-
The Dietary Supplement Health Education Act of mentary medications and treatments that claim to im-
 provided the framework for an expansion in the prove weight loss. Evidence to support the effectiveness
use of non–FDA-approved, over-the-counter products for weight loss or the safety of these preparations is
in the United States billed as “dietary supplements.” As usually nonexistent. Moreover, variability in the com-
a result, there has been a proliferation in the use of position of these products adds an additional uncertainty
these products. to their use. We thus think that the public would be better
This legislation helped undercut the credibility of served if the dietary supplements were held to a higher
legitimate weight-management practices by allowing standard and were overseen by the FDA.
the promotion of agents that are often unsafe, in-
effective, and have unproven health claims. As long as Drug targets
the claim is not for disease treatment per se, and Effective drugs to treat obesity have been slow to arise,
products are generally recognized as safe, they can be but efforts are still underway to develop novel, ef-
promoted for health claims. These agents are regulated fective, and transformative medications that would
by the U.S. Federal Trade Commission but not by the have the effect on treating obesity that statins had for
FDA, and thus they do not undergo the rigorous high cholesterol or thiazides had for hypertension
testing and review exercised by the FDA when it ().

36 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

Surgery in Managing Obesity The lack of gastrojejunal anastomosis has theoretic

benefits, such as reducing the risk of micronutrient
Introduction deficiencies and peptic ulcer disease. Although some
Surgical strategies (including the use of medical devices) restriction of food intake may occur, gastric emptying
for the purpose of inducing and maintaining clinically is accelerated.
significant weight loss have emerged and evolved during
the last  years. Surgeons performed ~, bariatric Roux-en-Y gastric bypass
procedures in  in the United States. RYGB refers to procedures in which a small (~ to
Sleeve gastrectomy (SG) is the most common  mL) gastric pouch is created just distal to the
procedure (.%), followed by Roux-en-Y gastric gastroesophageal junction with a stapling device. Most
bypass (RYGB), .%; laparoscopic adjustable gastric of the stomach is therefore disconnected (but not
banding (LAGB), .%; biliopancreatic diversion with excised) from the alimentary stream.
or without duodenal switch, .%; and revision and The small gastric pouch is the restrictive compo-
others, .% (). SG and RYGB together are the nent of this procedure. RYGB permits ingested food to
most popular procedures (%), whereas LAGB has pass directly from the esophagus through the small
become less popular due to poor long-term results. We stomach pouch and proceed directly into the jejunum,
list the three most common surgical procedures in Fig. with little or no gastric or duodenal phase of digestion,
 (). because food never enters the body of the stomach or
Evidence now indicates that that some of these the duodenum.
bariatric procedures (which were intended to either RYGB became a predominant weight-loss pro-
physically limit the ingestion of food or produce cedure in the s and is used worldwide today. The
malabsorption of energy-containing nutrients) actu- development and demonstration of the safety and
ally produce durable weight loss and health benefits by efficacy of minimally invasive (laparoscopic) tech-
altering metabolic processes, reducing appetite, and niques, the recognition of severe obesity as a disease,
inducing satiety early after meal ingestion. and the health benefits of bariatric surgery have led to
a progressive increase in the number of gastric bypass
Sleeve gastrectomy procedures performed (, , , ).
In SG, surgeons use a linear cutting stapler to make
a narrow gastric tube along the lesser curvature of the Laparoscopic adjustable gastric banding
stomach and remove the remaining % to % of the LAGB constricts the upper stomach by placing
gastric body and fundus (, ). a mechanical device encircling the stomach just

(a) (b) (c) Pouch

Esophagus Esophagus

Pouch Gastric
Adjustable band “sleeve”
Duodenum intestine

Stomach Resected Stomach


Access port

© 2018 Endocrine Reviews


Figure 9. The three most commonly performed bariatric surgical operations. (a) The laparoscopic gastric band is placed around the
upper stomach to restrict the transit of ingested food. (b) Laparoscopic sleeve gastrectomy involves separation of the greater curvature
from the omentum and splenic attachments. (c) RYGB involves the rearrangement of the alimentary canal, such that injected food
bypasses most of the stomach, all of the duodenum, and a portion of the proximal jejunum. See Nielsen et al., 2014 (368).

doi: 10.1210/er.2017-00253 37

Downloaded from
by guest
on 17 March 2018

beyond the gastroesophageal junction, thus creating stomach. An anastomosis between the proximal du-
a small ( to  mL) pouch. The tightness of the band odenum and bypassed intestine creates a degree of
is adjusted by inflating a linear balloon fixed within the malabsorption of nutrients. This procedure is in-
wall of the band. The balloon is connected to a sub- frequently performed because of a relatively high
cutaneous port, so clinicians can tighten the band via incidence of short-term and long-term complications,
a relatively simple percutaneous injection procedure. including micronutrient deficiencies ().
The band is intended to reduce the amount of food
consumed (). Vagal blockade
In this procedure, leads are placed about the vagal
Biliopancreatic diversion with (or without) trunks at the diaphragm to produce intermittent vagal
duodenal switch blockade. Weight loss occurs by reducing appetite and
Biliopancreatic diversion with or without duodenal inducing early satiety. The intermittent blockade is
switch is a complex procedure in which ~% of the designed to avoid the neural adaptation that occurred
body of the stomach is resected, creating a tubular with truncal vagatomy for peptic ulcer disease. Weight
stomach (SG) based on the lesser curvature of the loss, although modest, is superior to sham-treated
controls yet less successful than conventional surgi-
cal procedures, such as SG and gastric bypass ().
5 Roux-en-Y Gastric Bypass Despite a better safety profile than adjustable banding,
0 intermittent vagal blockade has limited efficacy. This
-5 coupled with adverse events make it a less desirable
-10 intervention for resolving obesity and associated
Group 1 (n = 36, 2.1%)
comorbidities ().
Percent weight change

Group 2 (n = 368, 21.5%) Gastrointestinal endoscopic interventions
or devices
-30 Group 3 (n = 796, 46.5%) Several devices, placed either by gastrointestinal en-
-35 doscopy or suturing procedures, have become avail-
Group 4 (n = 408, 23.8%)
-40 able. The FDA approved two gastric balloons in 
-45 Group 5 (n = 103, 6.0%) and another in . Clinicians can fill the Orbera
-50 intragastric balloon system with  to  mL of
-55 saline. The ReShape integrated dual balloon system
0 0.5 1 2 3 contains two connected, saline-filled balloons. In 
the FDA approved the Obalon balloon system, which
Years of follow-up
expands with air after insertion. Technical improve-
ments to these devices have resulted in a favorable
safety profile (). The present protocol requires
5 Laparoscopic Adjustable Gastric Band removal of the intragastric balloon  to  months after
0 placement, which is a limitation to the long-term
Group 1 (n = 115, 18.9.%)
-5 efficacy of this intervention. The balloon can be
-10 replaced for those who regain weight (). In August
, the FDA sent a letter to health care providers
Percent weight change

-15 Group 2 (n = 379, 62.4%)

-20 noting seven deaths associated with liquid-filled
intragastric balloon systems used to treat obesity.
Group 3 (n = 82, 13.5%) Four of the reports involved the Orbera intragastric
Group 4 (n = 7, 1.2%) balloon system and one with the ReShape integrated
Group 5 (n = 24, 4.0%) dual balloon system. Two earlier deaths were also
-45 Researchers have also developed a specially
-50 designed percutaneous gastrostomy tube and ap-
-55 paratus, called the AspireAssist device, that allows
0 0.5 1 2 3 patients to directly remove ingested food from the
Years of follow-up stomach (). After  year with this device, patients
lost .% compared with .% in the control group.
Median and interquartile range (observed)
This aspiration technique requires available facili-
Group trajectory (modeled) © 2018 Endocrine Reviews ENDOCRINE SOCIETY
ties to discard the aspirated food and is not for
Figure 10. (a) Percentage weight trajectories. See Courcoulas et al., 2013 (383). (b) Percentage of everyone.
participants in the intensive lifestyle intervention and diabetes support and education groups who Additionally, endoscopic placement of a duodenal–
achieved different categorical weight losses at year 8. See Look AHEAD Research Group, 2014 (271). jejunal luminal sleeve is under evaluation (). In

38 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

a study that examined endoscopic ablation of duodenal considerable variability. In this study, the bottom %
mucosa to enhance glycemic control of TDM (), of participants lost ,% of their body weight in
reduction of HbAc persisted  months after ablation. contrast to the top % who lost  kg (Fig. b) ().
During active weight loss after surgery, BP decreases
Liposuction to a point where antihypertensive drugs may be dis-
Liposuction (also known as lipoplasty or suction- continued. In addition, the requirement for hypoglyce-
assisted lipectomy) is the most common esthetic pro- mic medications in patients with TDM may also be
cedure performed in the United States, with ., diminished or discontinued. However, after weight sta-
cases performed annually (). Although not generally bilization, the results are less clear, as hypertension
considered to be a bariatric procedure, clinicians commonly reoccurs. Additionally, if weight is regained,
remove and contour subcutaneous fat by aspiration comorbidities that were present at baseline may reap-
after injecting physiologic saline. As techniques have pear. As a result, the question of what constitutes
improved, it is now possible to remove significant “successful medical therapy” is open to interpretation.
amounts of subcutaneous adipose tissue without af- Therefore, additional criteria for surgical interventions
fecting the amount of visceral fat. In a study to examine should include an understanding of operative risk and
the effects of this procedure, Klein et al. () studied the ability to manage obesity and comorbid disease after
seven diabetic women who were overweight and eight surgery.
women with normal glucose tolerance that were A recent joint statement by international diabetes
overweight before and after liposuction. One week after organizations has indicated that bariatric or metabolic
assessing insulin sensitivity, the subjects underwent surgery procedures are a consideration for patients
large volume tumescent liposuction, which consists of with poorly controlled TDM and a BMI of  to
removing . L of aspirate injected into the fat beneath  kg/m (). The Endocrine Society has also re-
the skin. There was a significant loss of subcutaneous leased pediatric guidelines for bariatric surgery ().
fat, but no change in the visceral fat. Subjects were
reassessed  to  weeks after the surgery. The non- Preoperative assessment
diabetic women lost . kg of body weight and . kg of Preoperative assessment of potential bariatric surgical
body fat, which reduced body fat by .%. The diabetic candidates includes confirming the patient’s un-
women had a similar response with a weight loss of derstanding of the basic procedure(s) proposed and
. kg, a reduction in body fat of . kg, and a re- what he or she needs to do to help make the treatment
duction in percentage fat of .%. Waist circumference successful. It also includes determining the patient’s
was also significantly reduced. Despite these signif- dedication and motivation to make the behavioral
icant reductions in body fat, there were no changes in changes necessary for a satisfactory outcome. Fig. 
BP, lipids, or cytokines (tumor necrosis factor-a, () contains a flowchart for managing bariatric
interleukin-), or C-reactive protein. There was also patients with obesity.
no improvement in insulin sensitivity, suggesting that The patient must also understand the risks asso-
removal of subcutaneous adipose tissue without ciated with the procedure, and clinicians need to assess
reducing ectopic fat depots has little influence on the all related comorbid conditions and manage these
risk factors related to being overweight. conditions preoperatively. At a minimum, clinicians

Indications for bariatric surgery

80% Intensive lifestyle intervention (ILI)
Criteria for bariatric surgery 73.6% Diabetes support and education (DSE)
The National Institutes of Health Consensus Panel in 70%
Percentage of participants at year 8

 established the initial criteria for surgical in- 60.8%

terventions for obesity (). The panel concluded that
individuals with BMI $  kg/m with a related 50%
comorbidity or BMI $  kg/m were appropriate
candidates for bariatric surgery. An additional crite- 40%
rion was failure of medical treatment to accomplish
30% 26.9%
sustained weight loss. These criteria have been variably
interpreted for many years but have remained es- 20% 17.2%
sentially unchanged until the present. 11.0%
In evaluating the outcome for any procedure, we 10% 7.0%
need criteria for “successful” treatment. Weight loss is
highly variable with all interventions. For example, >0% ≥5% ≥10% ≥15%
intense lifestyle interventions in the Look AHEAD Percent reduction in initial weight
© 2018 Endocrine Reviews ENDOCRINE SOCIETY
trial produced an average of .% weight loss at  year
and ~% at  years. However, this average covers Figure 10. (Continued)

doi: 10.1210/er.2017-00253 39

Downloaded from
by guest
on 17 March 2018

Figure 11. Obesity management flow. Summarized from the 2013 AHA/ACC/TOS guideline for the management of overweight and
obesity in adults (39). * refers to comorbid conditions. Reproduced with permission from Beamish et al., 2016 (384).

Primary assessment Intervention Follow-up

BMI ≥30 or ≥25 with Advice on weight gain Follow-up and weight-
comorbid disease* No avoidance and treatment of loss maintenance with
any identified risk factors medical management of
Yes CV risk factors and
a High-intensity comprehensive
lifestyle intervention treating
CV risk factors and
Determine readiness Not ready a: ≥25 and <27 Yes
for lifestyle changes and
establish weight b No
Ready b: ≥30 or ≥27 Consider pharmacotherapy Weight loss 5% and Continue intensive
loss goals
with comorbid disease* as an adjunct to lifestyle sufficient health medical management
intervention improvements of CV risk factors and
c: ≥40 or ≥35 obesity-related
with comorbid disease* comorbidities and
consider escalation
c Offer referral to specialist
of treatment
bariatric surgical MDT for
evaluation as an adjunct to
lifestyle intervention

Abbreviations: CV, cardiovascular; MDT, multidisciplinary team.

© 2018 Endocrine Reviews ENDOCRINE SOCIETY

should meet standard guidelines for cancer screening, Metabolic and Bariatric Surgery and the American
given the increased risk for common cancers in those College of Surgeons.
with obesity (including breast and colon cancers). The addition of laparoscopic procedures also
Clinicians should also identify and correct micro- contributed to improved safety. Recently, the Longi-
nutrient deficiencies. tudinal Assessment of Bariatric Surgery (LABORA-
Many centers require preoperative weight loss, TORIES) (a multicenter bariatric surgery research
which may decrease the risk of perioperative medical consortium funded by the National Institutes of
complications of anesthesia and abdominal surgery. Health) reported a -day overall bariatric surgery
However, the role of preoperative weight loss in de- mortality rate of .%. For laparoscopic RYGB, it
termining longer term outcomes (such as weight loss reported a -day mortality rate of .% ().
beyond  year) has not been demonstrated. A serious complication occurred in .% of all
patients. Factors that predicted a major complication
Outcomes of bariatric surgery include high BMI, extreme OSA, inability to walk 
feet, and a history of deep vein thrombosis. Other
Safety studies have reported different risk profiles. Studies
There is little or no disagreement about the benefits of consistently report that the experience of both the
weight loss among individuals with severe obesity, surgeon and the surgical center are predictors of safety
particularly those with comorbid conditions. These ().
benefits, however, must be considered in the context of Mid-term and longer term complications have
potential surgical complications. A population-based been well described, although determining their in-
study in  reported % mortality after RYGB (). cidence is limited by a progressively greater number of
In response, the bariatric surgical community enacted patients lost to follow-up (). These include, but are
several changes to improve safety. This included not limited to, intestinal obstruction, marginal ulcer,
identifying the importance of surgeon experience and ventral hernia, and gallstones. Metabolic complica-
the experience of the particular surgical center; the tions reported include nephrolithiasis, osteoporosis,
establishment of pathways, care protocols, and and hypoglycemia. Mineral and vitamin deficiencies
quality initiatives; and the incorporation of all these and weight regain are reported in variable numbers of
aspects of care into the Metabolic and Bariatric patients. Micronutrient deficiencies following gastric
Surgery Accreditation and Quality Improvement bypass include: iron, % to %; calcium/vitamin D,
Program administered by the American Society for % to %; vitamin B, % to %; copper, % to

40 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

%; and thiamine, ,% (). Established guidelines The single best predictors of sustained post-
recommend routine nutrient supplementation to in- operative weight loss (identified by the LABORA-
clude multivitamins, vitamin B, iron, minerals, cal- TORIES Consortium) are postoperative eating and
cium, and vitamin D (, ). lifestyle behaviors. Specifically, subjects who self-
Perioperative complications specific to LAGB are monitor (e.g., frequent weighing), avoid eating when
less frequent, with near zero mortality. Longer term full, and who avoid snacking between meals appear to
complications, however, continue to occur at a rate of experience the greatest weight loss (). The weight
~% per year. These longer term complications include loss following RYGB, compared with interventions
erosion of the gastric wall by the band and slippage or other than surgery (Fig. a and b), demonstrates
herniation of the body of the stomach, thereby creating that even the poorest weight loss following gastric
obstructions within the band. Inadequate weight loss is bypass is comparable to the best reported weight loss
the most common cause of LAGB failure. Compli- for nonsurgical interventions (). A third study
cations following other device placement procedures found changes from baseline after  years in the
occur but are infrequent and generally less severe. surgical groups were superior to the changes seen with
However, there is a tradeoff between reduced com- medical therapy. Body weight decreased % with
plication rates and the severity of complications vs gastric bypass, % with SG, and % with drug
efficacy of weight loss. therapies (). We must interpret these outcomes
In summary, both perioperative and longer term with the caveat that the requirement for surgical in-
complications occur after all bariatric surgical procedures. tervention is the failure of patients to accomplish
Multiple steps have been taken in recent years to reduce sustained weight loss via other means, thereby creating
perioperative mortality and serious complications. Pro- a selected population.
spective data collection, analysis, and reporting to indi-
vidual centers through the accreditation program will Related outcomes/remission of T2DM
continue to identify complications and stimulate ap- The remarkable remission of TDM following RYGB
propriate quality improvement initiatives. has generated much interest given the prevalence of
TDM and the severity of this disease (, )
Weight loss (Table ) (, ).
The high degree of variability of weight loss following Analysis of the data from the LABORATORIES
all interventions (including intense lifestyle in- Consortium has demonstrated that both weight loss
tervention, medications, and virtually all bariatric and the neuroendocrine effects specific to gastric
surgical procedures) speaks to the complexity of severe bypass contribute to the remission of TDM ().
obesity (Fig. a and b). We can define obesity in The durability of the remission in many participants
terms of excess weight represented by BMI. However, was sustained through year  ().
the genetic, behavioral, and environmental factors that The Swedish Obese Subjects study reported that
underlie excessive weight gain in life are exceedingly ~% of gastric bypass patients who were in TDM
complex and variable. remission at year  had recurrence by year  ().
The same factors influence variability observed in Weight loss in these groups is shown in Fig. . Gastric
the weight-loss phenotypes following surgery. After banding was the predominant procedure in the
RYGB, for example, the LABORATORIES Consor- Swedish Obese Subjects trial. This suggests that
tium reported patterns of similar and rapid weight loss maintaining weight loss, as well as the incretin
among patients  months after surgery by stratifying stimulation associated with RYGB, contributes to the
weight loss into five separate trajectories, ranging from durability of the remission. Reports of TDM re-
% to % total body weight loss  years after surgery mission among patients with considerably less severe
() (Fig. a). These weight-loss trajectories persist
through  years ().
Weight loss following LAGB is similarly variable,
but only one-half of the total body weight loss seen
Table 11. Weight Loss and Reversal of Diabetes Mellitus
with RYGB is seen after LAGB (on average). Thus,
after Metabolic/Bariatric Surgery
the commonly reported mean weight loss among
populations undergoing bariatric surgery is of limited Excess Weight Resolution of
use for predicting results for individuals who are Procedure Loss (%) T2DM (%)
Including data from Buchwald
contemplating surgical treatment. Gastric banding 46.2 56.7 et al., 2009 (396) and Bray, 2011
There have been many efforts to identify pre- (397).
Gastroplasty 55.5 79.7
operative clinical predictors of postoperative weight
loss. However, although research has established some RYGB 59.7 80.3
statistical correlations, the extent of the variability
Biliopancreatic 63.6 95.1
explained by a number of clinical covariates has been
disappointing ().

doi: 10.1210/er.2017-00253 41

Downloaded from
by guest
on 17 March 2018

obesity (BMI  to  kg/m) have led to several RCTs Bariatric/metabolic surgery in adolescents
in this patient population (). Owing to the lack of effectiveness of nonsurgical
Most recently, SG has taken a dominant place in options for treating severe obesity in young patients
the spectrum of procedures used for weight loss and the demonstrated safety and efficacy of bariatric
worldwide (, ). Although the weight loss and surgery in adults, clinicians increasingly use surgical
TDM remission following SG appear to be slightly procedures to induce weight loss in selected adoles-
less than that following gastric bypass, lower peri- cents with severe obesity. The rationale for and ex-
operative complication rates, shorter lengths of stay, pectation of bariatric treatment in adolescents are to
and lower costs have made SG an attractive bariatric provide significant and durable weight reduction,
surgical procedure (, ). The mechanism and correct existing health problems, and prevent expected
durability of this improved glycemic control, including comorbidities in those at risk.
the role of diet-induced weight loss, have not been Lifestyle modification and even pharmacotherapy
determined (). in adolescents with severe obesity are associated with
Overall, there is considerable evidence favoring unsatisfactory outcomes, and any weight reduction
RYGB, LAGB, and now SG as superior methods for seen may not be sustained. Conversely, growing evi-
controlling or inducing remission of TDM, vs in- dence indicates that surgery results in % to %
tense medical treatment (). As a result, the term weight reduction in severely obese adolescents. In the
“metabolic” surgery has become popular. The con- Teen-LABORATORIES study, SG and RYGB were the
cept that clinicians should consider surgical in- most commonly performed procedures in adolescents,
tervention for patients with poorly controlled TDM and -year outcomes demonstrated a similar weight
and patient with less severe obesity (class I) with loss of nearly % for these procedures ().
TDM (rather than having BMI be the primary The lower complexity of SG and the lower theo-
indication for surgery) has gained widespread in- retical risk of at least some micronutrient deficiencies
ternational support (). Remission of dyslipidemia associated with RYGB make SG an attractive option
is also seen in most patients following effective for most adolescents, despite fewer published studies
surgical weight loss, whereas remission of hyper- of SG in adolescent age groups.
tension is less frequent. In , investigators in the United States ()
and Sweden () simultaneously reported long-term
outcomes for weight loss and comorbidities in ado-
lescents who underwent RYGB. Eight-year (United
5 States) and -year (Sweden) post-RYGB surgery
Control follow-up assessments indicated % and % BMI
reductions, respectively. Both research groups docu-
-5 mented important improvements in health.
In the U.S. study, remission of TDM occurred in
Percent weight change

-10 % (n = ). The study did not report any incident
TDM during the  years. The study also reported
VBG dyslipidemia remission in % (n = ) and inci-
-20 dent dyslipidemia in four of eight subjects who did not
have dyslipidemia at baseline. The study reported
-25 hypertension remission in % (n = ) and incident
hypertension in only % ( of ) participants
without hypertension at baseline.
-35 The Swedish study reported similar health im-
0 1 2 3 4 6 8 10 15 20 provements, with remission of comorbid conditions in
Years of follow-up % to % of participants. The study reported re-
Number examined mission of TDM in  of  participants, disturbed
Control 2037 1490 1242 1267 556 176 glucose homeostasis in  of , dyslipidemia in  of
Banding 376 333 284 284 150 50 , elevated BP in  of , inflammation (high-
VBG 1369 1086 967 1007 489 82 sensitivity C-reactive protein $  mg/L) in  of
GBP 265 209 184 180 37 13
© 2018 Endocrine Reviews ENDOCRINE SOCIETY
, and elevated liver enzymes in  of  participants.
Both studies also reported long-term nutritional
Figure 12. Mean weight change percentages from baseline for controls and the three surgery effects. The U.S. study reported mild anemia in %
groups during 20 years in the Swedish obese subjects study. Data shown for controls obtaining
(n = ), hyperparathyroidism in % (n = ), and low
usual care and for surgery patients obtaining banding, vertically banded gastroplasty, or
gastric bypass at baseline. Percentage weight changes from the baseline examination and
vitamin B levels in % (n = ) at long-term follow-
onward are based on data available on 1 July 2011. Error bars represent 95% CIs. Vertical error up. At  years in the Swedish study, % ( of ) had
bars represent SEM. GBP, gastric bypass; VBG, vertically banded gastroplasty. See Sjöström vitamin D (-hydroxy vitamin D) insufficiency (,
et al., 2012 (393). nmol/L) and % ( of ) had low ferritin and/or

42 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

iron levels. The prevalence of anemia rose from % ( (). These observations may provide the potential
of ) to % ( of ), and % had low vitamin B for more effective preventive strategies utilizing
levels. social engineering.
In summary, the two long-term, prospective Genetic factors also play a role. Recently, in-
studies demonstrate excellent durability of weight loss sufficiency in the gene TRIM was shown to produce
and response of comorbidities for adolescents who polyphenic obesity in both mice and humans. In this
have RYGB surgery. These studies also reported the setting, both lean and obese phenotypes can arise from
typical nutritional consequences of RYGB that we see identical genotypes through dysregulation of an
in studies in adults, and this must be taken into imprinted gene network (). This finding and other
consideration when counseling patients about long- genetic research into the mechanisms behind obesity
term risks of RYGB. may provide new genetic strategies for helping this
Current expert opinion recommends that clinicians segment of the population.
should use criteria similar to those used for adults when The hazards of obesity are many, including
selecting adolescents for weight-loss surgery (, ). a shortened life span, TDM, CVD, some cancers,
Surgery is generally recommended for adolescents with kidney disease, OSA, gout, osteoarthritis, and hep-
a BMI $  kg/m and a weight-related comorbid atobiliary disease, among others. As might be expected,
condition or impairment in quality of life. It is also weight loss reduces all of these diseases in a dose-
recommended for those with a BMI of $ kg/m with related manner.
significant current comorbidities, such as TDM, dysli- The phenotype of MHO appears to be a transient
pidemia, OSA, hypertension, NASH, or pseudotumor state that progresses over time to an unhealthy phe-
cerebri (, ). notype, especially in children and adolescents. Un-
derstanding in more detail how complications of
obesity develop will provide new opportunities for
Where Do We Go From Here? prevention of these negative outcomes.
Of particular interest are reports that two diabetes
In this Endocrine Society Scientific statement titled medications (liraglutide and empagliflozin) (, )
“The Science of Obesity Management: An Endocrine also produce weight loss and are cardioprotective.
Society Scientific Statement,” we have documented the Particularly striking is the fact that these two drugs
rising prevalence of obesity in both men and women in reduce cardiovascular death to a greater extent than
the United States and worldwide with resultant haz- statins. This opens a whole new paradigm for man-
ardous health implications. The prevalence of obesity aging patients with obesity and diabetes in relationship
is correlated with income disparity both between to their complications.
developed countries and between the states of the One of the unexplained issues in all treatment
United States (). strategies is the marked variability in response of any
Obesity results in part from environmental and form of treatment of obesity. Efforts to understand the
behavioral factors, and both the public and health biological basis of this variability may provide new
care professionals alike stigmatize the condition. insights into its treatment. The POUNDS Lost Study
The opportunity to move from a neighborhood with population of  individuals randomized to one of
a high level of poverty to one with a lower level of four diets (% vs % fat and % vs % protein) has
poverty was associated with modest but potentially provided many genetic clues to help us better un-
important reductions in the prevalence of extreme derstand factors that modulate dietary response (,
obesity and diabetes (), supporting the re- ). The ability to combine several measures to
lationship between income inequality and obesity. predict responses to environmental factors may ex-
Because the prevalence of obesity has strong social pand the option for personalized medicine. An algo-
and environmental components, this may provide rithm that integrates blood parameters, dietary habits,
a basis for future approaches. The study by anthropometrics, physical activity, and gut microbiota
Christakis and Fowler () showed that “friends” of measured in a sample cohort () showed that these
an individual with obesity were more likely to also factors accurately predict personalized postprandial
be obese. glycemic response to real-life meals. Similar strategies
Obesity is lower when there are more oppor- might well be developed for obesity. Many genes affect
tunities for physical activity as part of everyday life, the response to diets, opening the possibility of
as shown by the slower rise in obesity among more “personalized medicine” for managing obesity.
active individuals during  years (). Sleep time The public commonly uses over-the-counter
is a modifiable behavior, and the observation that herbal preparations to manage obesity, but evidence
preschool-aged children with early weekday bed- documenting their efficacy or safety is usually absent.
times were one-half as likely as children with late We think that the public would be well served by more
bedtimes to be obese as adolescents offers further regulatory requirements regarding sale and use of
opportunities for intervention in the environment these products.

doi: 10.1210/er.2017-00253 43

Downloaded from
by guest
on 17 March 2018

We can expect to see weight regain in all patients peptides have been shown to enhance weight loss and the
when they discontinue obesity treatments. When decline in glucose, opening a fascinating new horizon
making treatment decisions, clinicians should consider (, ).
body fat distribution and individual health risks in Finally, improved techniques for modulating food
addition to BMI. Because all treatments have con- transit through the gastrointestinal track and its ab-
siderable variability in their outcome, it is important to sorption also offer new strategies for dealing with the
know when to stop treatment as well as when to begin. devastating epidemic posed by obesity. It is clear that
Surgical strategies have demonstrated greater weight food is more than calories and that dietary choices and
loss that outlasts other treatment options. diet quality play a role in long-term weight change
As the knowledge base underpinning obesity con- (), and this provides other opportunities for public
tinues to expand, the options for treating patients with health programs. The so-called “obesogens” in the
obesity should also expand, offering hope for future food supply offer another opportunity for making the
conquest of this problem. One fascinating new strategy is food supply less likely to contribute to obesity ().
the combination of peptides acting on receptors in the Control of obesity is the most important public health
gastrointestinal track into a single molecule acting on two strategy for the prevention of diabetes and its dev-
or more receptors, called coagonists and triagonists. astating consequences. With all of these opportunities
Using glucagon-like peptide-, glucagon, and glucose- on the horizon, we are optimistic about the future of
insulin peptide as the background for these molecules, treatment and prevention of obesity.

1. Bray GA. Obesity: historical development of sci- 16. Rössner S. Sancho the Fat: King of León, Spain. Obes 28. Lee MD, Somerville EM, Kennett GA, Dourish CT,
entific and cultural ideas. Int J Obes. 1990;14(11): Rev. 2011;12(11):995. Clifton PG. Reduced hypophagic effects of d-fen-
909–926. 17. Short T. Discourse Concerning the Causes and Effects fluramine and the 5-HT2C receptor agonist mCPP in
2. Bray G. Battle of the Bulge: A History of Obesity Research. of Corpulency: Together With the Method for Its 5-HT1B receptor knockout mice. Psychopharma-
Philadelphia, PA: Dorrance Publishing; 2007. Prevention and Cure. 2nd ed. London, England: cology (Berl). 2004;176(1):39–49.
3. Gilman S. Obesity: The Biography. Oxford, U.K.: J. Roberts; 1727. 29. Heal DJ, Gosden J, Smith SL. A review of late-stage
Oxford University Press; 2010. 18. Bonetus T. Sepulchretum, Sive Anatomia Practica, ex CNS drug candidates for the treatment of obesity.
4. Angeli W. Die Venus von willendorf. Vienna, Austria: Cadaveribus Morbo Denatis, Proponens Historias Int J Obes. 2013;37(1):107–117.
Edition Wien; 1989. Omnium Humani Corporis Affectum. Geneva, 30. Foster GD, Wadden TA, Vogt RA, Brewer G. What is
5. Hippocrates. Oeuvres Complètes: Traduction Nouvelle Switzerland: Sumptibus Cramer & Perachon; 1700. a reasonable weight loss? Patients’ expectations and
Avec le Texte Grec en Regard, Collationné sur les 19. Colman E. Dinitrophenol and obesity: an early evaluations of obesity treatment outcomes.
Manuscrits et Toutes les Éditions; Accompagnée d’une ... twentieth-century regulatory dilemma. Regul Toxicol J Consult Clin Psychol. 1997;65(1):79–85.
Suivie d’une Table Générale. Paris, France: J.B. Bailliere; Pharmacol. 2007;48(2):115–117. 31. Smiertka JK, MacPherson BH. Bariatric surgery
1839. 20. Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, postoperative behavioral change: The importance
6. Flemyng M. A Discourse on the Nature, Causes, and Friedman JM. Positional cloning of the mouse obese of ongoing assessment and teaching. Obes Surg.
Cure of Corpulency. Illustrated by a Remarkable Case. gene and its human homologue. Nature. 1994; 1995;5(3):337–340.
Read Before the Royal Society November 1757. 372(6505):425–432. 32. Rabner JG, Greenstein RJ. Obesity surgery: expec-
London, England: L. Davis and C. Reymers; 1760. 21. Schwartz L, ed. Preface. The Cambridge Companion tation and reality. Int J Obes. 1991;15(12):841–845.
7. Wadd W. Cursory Remarks on Corpulence; or Obesity to Paradise Lost. New York, NY: Cambridge Uni- 33. Wee CC, Jones DB, Davis RB, Bourland AC, Hamel
Considered as a Disease: With a Critical Examination versity Press; 2014:xi–xiv. MB. Understanding patients’ value of weight loss
of Ancient and Modern Opinion, Relative to Its Causes 22. Schwartz MW, Seeley RJ, Zeltser LM, Drewnowski A, and expectations for bariatric surgery. Obes Surg.
and cure. London: J. Callow; 1816. Ravussin E, Redman LM, Leibel RL. Obesity path- 2006;16(4):496–500.
8. Lavoisier AL. Traité Élémentaire de Chimie: Présenté ogenesis: an Endocrine Society scientific statement. 34. Weintraub M, Bray GA. Drug treatment of obesity.
dans un Ordre Nouveau et d’Après les Découvertes Endocr Rev. 2017;38(4):267–296. Med Clin North Am. 1989;73(1):237–249.
Modernes: Avec Figures. Paris, France: Cuchet; 1789. 23. Puhl RM, Latner JD. Stigma, obesity, and the health 35. Bolen SD, Clark JM, Richards TM, Shore AD,
9. Atwater WO, Rosa EB. A new respiration calo- of the nation’s children. Psychol Bull. 2007;133(4): Goodwin SM, Weiner JP. Trends in and patterns of
rimeter and experiments on the conservation of 557–580. obesity reduction medication use in an insured
energy in the human body, II. Phys Rev Ser I. 1899;9: 24. Linné Y, Hemmingsson E, Adolfsson B, Ramsten J, cohort. Obesity (Silver Spring). 2010;18(1):206–209.
214–251. Rössner S. Patient expectations of obesity 36. Harrison GG. Height-weight tables. Ann Intern Med.
10. Cushing H. The pituitary body and its disorders. treatment-the experience from a day-care unit. Int J 1985;103(6, Pt 2):989–994.
Clinical states produced by disorders of the hy- Obes Relat Metab Disord. 2002;26(5):739–741. 37. Bray GA, Greenway FL, Molitch ME, Dahms WT,
pophysis cerebri. Am J Med Sci. 1912;144(6):891. 25. Kaly P, Orellana S, Torrella T, Takagishi C, Saff-Koche Atkinson RL, Hamilton K. Use of anthropometric
11. Babinski M. Tumeur du corps pituitaire sans L, Murr MM. Unrealistic weight loss expectations in measures to assess weight loss. Am J Clin Nutr. 1978;
acromégalie et avec de développement des organes candidates for bariatric surgery. Surg Obes Relat Dis. 31(5):769–773.
génitaux. Rev Neurol. 1900;8:531–533. 2008;4(1):6–10. 38. Obesity: preventing and managing the global epi-
12. Fröhlich A. Ein Fall von Tumor der Hypophysis cerebri 26. Dutton GR, Perri MG, Stine CC, Goble M, Van demic. Report of a WHO consultation. World Health
ohne Akromegalie. Wiener Klin Rdsch; 1901:883–886. Vessem N. Comparison of physician weight loss Organ Tech Rep Ser. 2000;894:i–xii, 1–253.
13. Precope J. Hippocrates on Diet and Hygiene. London, goals for obese male and female patients. Prev Med. 39. Jensen MD, Ryan DH, Apovian CM, Ard JD,
England: Zeno; 1952. 2010;50(4):186–188. Comuzzie AG, Donato KA, Hu FB, Hubbard VS,
14. Foxcroft L. Calories and Corsets: A History of Dieting 27. Haddock CK, Poston WSC, Dill PL, Foreyt JP, Ericsson Jakicic JM, Kushner RF, Loria CM, Millen BE, Nonas
Over 2000 Years. London, U.K.: Profile Books; 2011. M. Pharmacotherapy for obesity: a quantitative CA, Pi-Sunyer FX, Stevens J, Stevens VJ, Wadden TA,
15. Banting W. Letter on corpulence, addressed to the analysis of four decades of published randomized Wolfe BM, Yanovski SZ, Jordan HS, Kendall KA, Lux
public. 1869. Obes Res. 1993;1(2):153–163. clinical trials. Int J Obes Relat Metab Disord. 2002; LJ, Mentor-Marcel R, Morgan LC, Trisolini MG,
26(2):262–273. Wnek J, Anderson JL, Halperin JL, Albert NM,

44 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

Bozkurt B, Brindis RG, Curtis LH, DeMets D, Accuracy of body mass index in diagnosing obesity BMI, waist circumference, and waist-to-height ratio
Hochman JS, Kovacs RJ, Ohman EM, Pressler SJ, in the adult general population. Int J Obes. 2008; for identifying cardio-metabolic risk in children and
Sellke FW, Shen WK, Smith SC, Jr, Tomaselli GF; 32(6):959–966. adolescents. PLoS One. 2016;11(2):e0149351.
American College of Cardiology/American Heart 51. Berstad P, Randby A, Seim Ekeland G, Ulveland H, 64. Ogden CL, Kuczmarski RJ, Flegal KM, Mei Z, Guo S,
Association Task Force on Practice Guidelines Omland T, Almendingen K. Body fat and fat-free Wei R, Grummer-Strawn LM, Curtin LR, Roche AF,
Obesity Society. 2013 AHA/ACC/TOS guideline for mass measured by bioelectric impedance spec- Johnson CL. Centers for disease control and pre-
the management of overweight and obesity in troscopy and dual-energy X-ray absorptiometry in vention 2000 growth charts for the united states:
adults: a report of the American College of obese and non-obese adults. Br J Nutr. 2012;107(8): improvements to the 1977 national center for
Cardiology/American Heart Association Task Force 1192–1200. health statistics version. Pediatrics. 2002;109(1):
on Practice Guidelines and The Obesity Society. 52. Després JP. Body fat distribution and risk of car- 45–60.
Circulation. 2014;129(25, Suppl 2):S102–S138. diovascular disease: an update. Circulation. 2012; 65. Kahn HS, Imperatore G, Cheng YJ. A population-
40. Clinical guidelines on the identification, evaluation, 126(10):1301–1313. based comparison of BMI percentiles and waist-to-
and treatment of overweight and obesity in adults: 53. Matsuzawa Y, Funahashi T, Nakamura T. The height ratio for identifying cardiovascular risk in
executive summary. Expert Panel on the Identifi- concept of metabolic syndrome: contribution of youth. J Pediatr. 2005;146(4):482–488.
cation, Evaluation, and Treatment of Overweight in visceral fat accumulation and its molecular 66. Roy SM, Spivack JG, Faith MS, Chesi A, Mitchell JA,
Adults. Am J Clin Nutr. 1998;68(4):899–917. mechanism. J Atheroscler Thromb. 2011;18(8): Kelly A, Grant SFA, McCormack SE, Zemel BS. Infant
41. Whitlock G, Lewington S, Sherliker P, Clarke R, 629–639. BMI or weight-for-length and obesity risk in early
Emberson J, Halsey J, Qizilbash N, Collins R, Peto R; 54. Sahakyan KR, Somers VK, Rodriguez-Escudero JP, childhood. Pediatrics. 2016;137(5):e20153492.
Prospective Studies Collaboration. Body-mass index Hodge DO, Carter RE, Sochor O, Coutinho T, Jensen 67. Wildman RP, Muntner P, Reynolds K, McGinn AP,
and cause-specific mortality in 900 000 adults: MD, Roger VL, Singh P, Lopez-Jimenez F. Normal- Rajpathak S, Wylie-Rosett J, Sowers MR. The obese
collaborative analyses of 57 prospective studies. weight central obesity: implications for total and without cardiometabolic risk factor clustering and
Lancet. 2009;373(9669):1083–1096. cardiovascular mortality. Ann Intern Med. 2015; the normal weight with cardiometabolic risk factor
42. GBD 2015 Disease and Injury Incidence and Prev- 163(11):827–835. clustering: prevalence and correlates of 2 pheno-
alence Collaborators. Global, regional, and national 55. Vague J. The degree of masculine differentiation of types among the US population (NHANES
incidence, prevalence, and years lived with disability obesities: a factor determining predisposition to 1999–2004). Arch Intern Med. 2008;168(15):
for 310 diseases and injuries, 1990–2015: a system- diabetes, atherosclerosis, gout, and uric calculous 1617–1624.
atic analysis for the Global Burden of Disease Study disease. 1956. Obes Res. 1996;4(2):204–212. 68. Jensen MD. Role of body fat distribution and the
2015. Lancet. 2016;388(10053):1545–1602. 56. Pouliot M-C, Després J-P, Lemieux S, Moorjani S, metabolic complications of obesity. J Clin Endocrinol
43. Berrington de Gonzalez A, Hartge P, Cerhan JR, Flint Bouchard C, Tremblay A, Nadeau A, Lupien PJ. Metab. 2008;93(11, Suppl 1):S57–S63.
AJ, Hannan L, MacInnis RJ, Moore SC, Tobias GS, Waist circumference and abdominal sagittal di- 69. Stefan N, Häring HU, Hu FB, Schulze MB. Meta-
Anton-Culver H, Freeman LB, Beeson WL, Clipp SL, ameter: best simple anthropometric indexes of bolically healthy obesity: epidemiology, mecha-
English DR, Folsom AR, Freedman DM, Giles G, abdominal visceral adipose tissue accumulation nisms, and clinical implications. Lancet Diabetes
Hakansson N, Henderson KD, Hoffman-Bolton J, and related cardiovascular risk in men and women.
Endocrinol. 2013;1(2):152–162.
Hoppin JA, Koenig KL, Lee IM, Linet MS, Park Y, Am J Cardiol. 1994;73(7):460–468.
70. Roberson LL, Aneni EC, Maziak W, Agatston A,
Pocobelli G, Schatzkin A, Sesso HD, Weiderpass E, 57. Lean MEJ, Han TS, Morrison CE. Waist circumfer-
Feldman T, Rouseff M, Tran T, Blaha MJ, Santos RD,
Willcox BJ, Wolk A, Zeleniuch-Jacquotte A, Willett ence as a measure for indicating need for weight
Sposito A, Al-Mallah MH, Blankstein R, Budoff MJ,
WC, Thun MJ. Body-mass index and mortality management. BMJ. 1995;311(6998):158–161.
Nasir K. Beyond BMI: the “metabolically healthy
among 1.46 million white adults. N Engl J Med. 2010; 58. Balkau B, Deanfield JE, Després JP, Bassand JP, Fox
obese” phenotype & its association with clinical/
363(23):2211–2219. KAA, Smith SC, Jr, Barter P, Tan CE, Van Gaal L,
subclinical cardiovascular disease and all-cause
44. Yumuk V, Tsigos C, Fried M, Schindler K, Busetto L, Wittchen HU, Massien C, Haffner SM. International
mortality—a systematic review. BMC Public
Micic D, Toplak H; Obesity Management Task Day for the Evaluation of Abdominal Obesity
Health. 2014;14(1):14.
Force of the European Association for the Study of (IDEA): a study of waist circumference, cardiovas-
71. Camhi SM, Katzmarzyk PT. Differences in body
Obesity. European guidelines for obesity manage- cular disease, and diabetes mellitus in 168,000
composition between metabolically healthy obese
ment in adults. Obes Facts. 2015;8(6):402–424. primary care patients in 63 countries. Circulation.
45. Lau DCW, Douketis JD, Morrison KM, Hramiak IM, and metabolically abnormal obese adults. Int J Obes.
Sharma AM, Ur E; Obesity Canada Clinical Practice 59. Cerhan JR, Moore SC, Jacobs EJ, Kitahara CM, 2014;38(8):1142–1145.
Guidelines Expert Panel. 2006 Canadian clinical Rosenberg PS, Adami H-O, Ebbert JO, English DR, 72. Rey-López JP, de Rezende LF, Pastor-Valero M, Tess
practice guidelines on the management and pre- Gapstur SM, Giles GG, Horn-Ross PL, Park Y, Patel BH. The prevalence of metabolically healthy obesity:
vention of obesity in adults and children [sum- AV, Robien K, Weiderpass E, Willett WC, Wolk A, a systematic review and critical evaluation of the
mary]. CMAJ. 2007;176(8):S1–S13. Zeleniuch-Jacquotte A, Hartge P, Bernstein L, definitions used. Obes Rev. 2014;15(10):781–790.
46. Barlow SE; Expert Committee. Expert committee Berrington de Gonzalez A. A pooled analysis of 73. van Vliet-Ostaptchouk JV, Nuotio M-L, Slagter
recommendations regarding the prevention, as- waist circumference and mortality in 650,000 SN, Doiron D, Fischer K, Foco L, Gaye A, Gögele
sessment, and treatment of child and adolescent adults. Mayo Clin Proc. 2014;89(3):335–345. M, Heier M, Hiekkalinna T, Joensuu A, Newby C,
overweight and obesity: summary report. Pediatrics. 60. Garvey WT, Mechanick JI, Brett EM, Garber AJ, Pang C, Partinen E, Reischl E, Schwienbacher C,
2007;120(Suppl 4):S164–S192. Hurley DL, Jastreboff AM, Nadolsky K, Pessah-Pollack Tammesoo M-L, Swertz MA, Burton P, Ferretti V,
47. Flegal KM, Kruszon-Moran D, Carroll MD, Fryar CD, R, Plodkowski R. American Association of Clinical Fortier I, Giepmans L, Harris JR, Hillege HL,
Ogden CL. Trends in obesity among adults in the Endocrinologists and American College of Endocri- Holmen J, Jula A, Kootstra-Ros JE, Kvaløy K,
united states, 2005 to 2014. JAMA. 2016;315(21): nology comprehensive clinical practice guidelines for Holmen TL, Männistö S, Metspalu A, Midthjell K,
2284–2291. medical care of patients with obesity. Executive Murtagh MJ, Peters A, Pramstaller PP, Saaristo T,
48. Fryar C. Prevalence of overweight, obesity, and summary. Complete guidelines available at https:// Salomaa V, Stolk RP, Uusitupa M, van der Harst P,
extreme obesity among adults: United States, Endocr Pract. van der Klauw MM, Waldenberger M, Perola M,
trends 1960–1962 through 2009–2010. Available at: 2016;22(7):842–884. Wolffenbuttel BHR. The prevalence of metabolic 61. Emdin CA, Khera AV, Natarajan P, Klarin D, Zekavat syndrome and metabolically healthy obesity in
adult_09_10/obesity_adult_09_10.pdf. Accessed SM, Hsiao AJ, Kathiresan S. Genetic association of Europe: a collaborative analysis of ten large co-
4 April 2017. waist-to-hip ratio with cardiometabolic traits, type hort studies. BMC Endocr Disord. 2014;14(1):9.
49. Okorodudu DO, Jumean MF, Montori VM, 2 diabetes, and coronary heart disease. JAMA. 2017; 74. Stefan N, Kantartzis K, Machann J, Schick F, Thamer
Romero-Corral A, Somers VK, Erwin PJ, Lopez- 317(6):626–634. C, Rittig K, Balletshofer B, Machicao F, Fritsche A,
Jimenez F. Diagnostic performance of body mass 62. Styne DM, Arslanian SA, Connor EL, Farooqi IS, Häring H-U. Identification and characterization of
index to identify obesity as defined by body adi- Murad MH, Silverstein JH, Yanovski JA. Pediatric metabolically benign obesity in humans. Obstet
posity: a systematic review and meta-analysis. Int J obesity-assessment, treatment, and prevention: An Gynecol Surv. 2009;64(1):30–31.
Obes. 2010;34(5):791–799. Endocrine Society clinical practice guideline. J Clin 75. Kramer CK, Zinman B, Retnakaran R. Are meta-
50. Romero-Corral A, Somers VK, Sierra-Johnson J, Endocrinol Metab. 2017;102(3):709–757. bolically healthy overweight and obesity benign
Thomas RJ, Collazo-Clavell ML, Korinek J, Allison 63. Sardinha LB, Santos DA, Silva AM, Grøntved A, conditions?: A systematic review and meta-analysis.
TG, Batsis JA, Sert-Kuniyoshi FH, Lopez-Jimenez F. Andersen LB, Ekelund U. A comparison between Ann Intern Med. 2013;159(11):758–769.

doi: 10.1210/er.2017-00253 45

Downloaded from
by guest
on 17 March 2018

76. Bell JA, Hamer M, Sabia S, Singh-Manoux A, Batty 92. Thangaratinam S, Rogozińska E, Jolly K, Glinkowski 107. Caballero B, Clay T, Davis SM, Ethelbah B, Rock BH,
GD, Kivimaki M. The natural course of healthy S, Duda W, Borowiack E, Roseboom T, Tomlinson J, Lohman T, Norman J, Story M, Stone EJ, Stephenson
obesity over 20 years. J Am Coll Cardiol. 2015;65(1): Walczak J, Kunz R, Mol BW, Coomarasamy A, Khan L, Stevens J; Pathways Study Research Group.
101–102. KS. Interventions to reduce or prevent obesity in Pathways: a school-based, randomized controlled
77. Eckel N, Meidtner K, Kalle-Uhlmann T, Stefan N, pregnant women: a systematic review. Health trial for the prevention of obesity in American
Schulze MB. Metabolically healthy obesity and Technol Assess. 2012;16(31):iii–iv, 1–191. Indian schoolchildren. Am J Clin Nutr. 2003;78(5):
cardiovascular events: a systematic review and 93. Thangaratinam S, Rogozinska E, Jolly K, Glinkowski 1030–1038.
meta-analysis. Eur J Prev Cardiol. 2016;23(9): S, Roseboom T, Tomlinson JW, Kunz R, Mol BW, 108. Gortmaker SL, Peterson K, Wiecha J, Sobol AM,
956–966. Coomarasamy A, Khan KS. Effects of interventions Dixit S, Fox MK, Laird N. Reducing obesity via
78. Hwang YC, Hayashi T, Fujimoto WY, Kahn SE, in pregnancy on maternal weight and obstetric a school-based interdisciplinary intervention
Leonetti DL, McNeely MJ, Boyko EJ. Visceral ab- outcomes: meta-analysis of randomised evidence. among youth: Planet Health. Arch Pediatr Adolesc
dominal fat accumulation predicts the conversion BMJ. 2012;344:e2088. Med. 1999;153(4):409–418.
of metabolically healthy obese subjects to an un- 94. Tanentsapf I, Heitmann BL, Adegboye ARA. Sys- 109. Plachta-Danielzik S, Landsberg B, Lange D, Seiberl J,
healthy phenotype. Int J Obes. 2015;39(9): tematic review of clinical trials on dietary in- Müller MJ. Eight-year follow-up of school-based
1365–1370. terventions to prevent excessive weight gain during intervention on childhood overweight—the Kiel
79. Jae SY, Franklin B, Choi Y-H, Fernhall B. Metaboli- pregnancy among normal weight, overweight and Obesity Prevention Study. Obes Facts. 2011;4(1):
cally healthy obesity and carotid intima-media obese women. BMC Pregnancy Childbirth. 2011; 35–43.
thickness: effects of cardiorespiratory fitness. 11(1):81. 110. Sahota P, Rudolf MC, Dixey R, Hill AJ, Barth JH, Cade
Mayo Clin Proc. 2015;90(9):1217–1224. 95. Muktabhant B, Lawrie TA, Lumbiganon P, Laopaiboon J. Randomised controlled trial of primary school
80. Ortega FB, Lee DC, Katzmarzyk PT, Ruiz JR, Sui X, M. Diet or Exercise, or Both, for Preventing Excessive based intervention to reduce risk factors for obesity.
Church TS, Blair SN. The intriguing metabolically Weight Gain in Pregnancy. Cochrane Database of BMJ. 2001;323(7320):1029–1032.
healthy but obese phenotype: cardiovascular Systematic Reviews. Wiley-Blackwell; 2015. 111. Malik VS, Pan A, Willett WC, Hu FB. Sugar-
prognosis and role of fitness. Eur Heart J. 2013;34(5): 96. Gillman MW, Rifas-Shiman SL, Fernandez-Barres S, sweetened beverages and weight gain in children
389–397. Kleinman K, Taveras EM, Oken E. Beverage intake and adults: a systematic review and meta-analysis.
81. Ricketts TA, Sui X, Lavie CJ, Blair SN, Ross R. Ad- during pregnancy and childhood adiposity. Pedi- Am J Clin Nutr. 2013;98(4):1084–1102.
dition of cardiorespiratory fitness within an obesity atrics. 2017;140(2):e20170031. 112. Dietz WH, Gortmaker SL. New strategies to pri-
risk classification model identifies men at increased 97. Larson N, Ward DS, Neelon SB, Story M. What role oritize nutrition, physical activity, and obesity in-
risk of all-cause mortality. Am J Med. 2016;129(5): can child-care settings play in obesity prevention? A terventions. Am J Prev Med. 2016;51(5):e145–e150.
536.e13–536.e20. review of the evidence and call for research efforts. 113. Kamath CC, Vickers KS, Ehrlich A, McGovern L,
82. McAuley PA, Blaha MJ, Keteyian SJ, Brawner CA, Al Johnson J, Singhal V, Paulo R, Hettinger A, Erwin PJ,
J Am Diet Assoc. 2011;111(9):1343–1362.
Montori VM. Clinical review: behavioral in-
Rifai M, Dardari ZA, Ehrman JK, Al-Mallah MH. 98. de Silva-Sanigorski AM, Bell AC, Kremer P, Nichols
terventions to prevent childhood obesity: a sys-
Fitness, fatness, and mortality: the FIT (Henry Ford M, Crellin M, Smith M, Sharp S, de Groot F,
tematic review and metaanalyses of randomized
exercise testing) project. Am J Med. 2016;129(9): Carpenter L, Boak R, Robertson N, Swinburn BA.
trials. J Clin Endocrinol Metab. 2008;93(12):
960–965.e1. Reducing obesity in early childhood: results from
83. Daniels SR. The consequences of childhood over- Romp & Chomp, an Australian community-wide
114. Task Force on Community Preventive Services. A
weight and obesity. Future Child. 2006;16(1):47–67. intervention program. Am J Clin Nutr. 2010;91(4):
recommendation to improve employee weight
84. Villareal DT, Apovian CM, Kushner RF, Klein S; 831–840.
status through worksite health promotion pro-
American Society for Nutrition; NAASO, The 99. Evaluating obesity prevention efforts: what have we
grams targeting nutrition, physical activity, or both.
Obesity Society. Obesity in older adults: technical learned? Available at:
Am J Prev Med. 2009;37(4):358–359.
review and position statement of the American sites/default/files/downloads/Redstone-Center/
115. Anderson LM, Quinn TA, Glanz K, Ramirez G,
Society for Nutrition and NAASO, The Obesity Evaluating Obesity Prevention Efforts What
Kahwati LC, Johnson DB, Buchanan LR, Archer WR,
Society. Obes Res. 2005;13(11):1849–1863. Have We Learned.pdf. Accessed 31 July 2016. Chattopadhyay S, Kalra GP, Katz DL; Task Force on
85. National Clinical Guideline Centre (U.K.). Obesity: 100. United States Department of Agriculture Food and Community Preventive Services. The effectiveness
Identification, Assessment and Management of Nutrition Services. Child and adult care food pro- of worksite nutrition and physical activity in-
Overweight and Obesity in Children, Young People gram. Available at: terventions for controlling employee overweight
and Adults. Partial Update of cg43. London, U.K.: child-and-adult-care-food-program. Accessed 12 and obesity: a systematic review. Am J Prev Med.
National Institute for Health and Care Excellence March 2017. 2009;37(4):340–357.
(U.K.); 2014. 101. Waters E, de Silva-Sanigorski A, Hall BJ, Brown T, 116. Thorndike AN, Sonnenberg L, Riis J, Barraclough S,
86. Sharma AM, Kushner RF. A proposed clinical Campbell KJ, Gao Y, Armstrong R, Prosser L, Levy DE. A 2-phase labeling and choice architecture
staging system for obesity. Int J Obes. 2009;33(3): Summerbell CD. Interventions for preventing intervention to improve healthy food and beverage
289–295. obesity in children. Cochrane Database Syst Rev. choices. Am J Public Health. 2012;102(3):527–533.
87. Han TS, Tajar A, Lean ME. Obesity and weight 2011; (12):CD001871. 117. Boston Public Health Commission. Healthy bever-
management in the elderly. Br Med Bull. 2011;97(1): 102. Epstein LH, Gordy CC, Raynor HA, Beddome M, ages. Available at:
169–196. Kilanowski CK, Paluch R. Increasing fruit and veg- healthy-eating-active-living/healthy-beverages/Pages/
88. Zamboni M, Mazzali G, Zoico E, Harris TB, Meigs JB, etable intake and decreasing fat and sugar intake in Healthy-Beverages.aspx. Accessed 7 February 2014.
Di Francesco V, Fantin F, Bissoli L, Bosello O. Health families at risk for childhood obesity. Obes Res. 2001; 118. Jeffery RW, French SA. Preventing weight gain in
consequences of obesity in the elderly: a review of 9(3):171–178. adults: the pound of prevention study. Am J Public
four unresolved questions. Int J Obes. 2005;29(9): 103. James J, Thomas P, Cavan D, Kerr D. Preventing Health. 1999;89(5):747–751.
1011–1029. childhood obesity by reducing consumption of 119. Papas MA, Alberg AJ, Ewing R, Helzlsouer KJ, Gary
89. Santanasto AJ, Goodpaster BH, Kritchevsky SB, carbonated drinks: cluster randomised controlled TL, Klassen AC. The built environment and obesity.
Miljkovic I, Satterfield S, Schwartz AV, Cummings trial. BMJ. 2004;328(7450):1237. Epidemiol Rev. 2007;29(1):129–143.
SR, Boudreau RM, Harris TB, Newman AB. Body 104. Ebbeling CB, Feldman HA, Chomitz VR, Antonelli 120. Ludwig J, Sanbonmatsu L, Gennetian L, Adam E,
composition remodeling and mortality: the health TA, Gortmaker SL, Osganian SK, Ludwig DS. A Duncan GJ, Katz LF, Kessler RC, Kling JR, Lindau ST,
aging and body composition study. J Gerontol A Biol randomized trial of sugar-sweetened beverages and Whitaker RC, McDade TW. Neighborhoods, obe-
Sci Med Sci. 2017;72(4):513–519. adolescent body weight. N Engl J Med. 2012;367(15): sity, and diabetes—a randomized social experiment.
90. Kelly LA, Lane CJ, Weigensberg MJ, Toledo-Corral 1407–1416. N Engl J Med. 2011;365(16):1509–1519.
CM, Goran MI. Pubertal changes of insulin sensi- 105. de Ruyter JC, Olthof MR, Seidell JC, Katan MB. A trial 121. Besser LM, Dannenberg AL. Walking to public
tivity, acute insulin response, and b-cell function in of sugar-free or sugar-sweetened beverages and transit: steps to help meet physical activity rec-
overweight Latino youth. J Pediatr. 2011;158(3): body weight in children. N Engl J Med. 2012;367(15): ommendations. Am J Prev Med. 2005;29(4):273–280.
442–446. 1397–1406. 122. MacDonald JM, Stokes RJ, Cohen DA, Kofner A,
91. Dietz WH, Economos CD. Progress in the control 106. Sallis JF, Glanz K. Physical activity and food envi- Ridgeway GK. The effect of light rail transit on body
of childhood obesity. Pediatrics. 2015;135(3): ronments: solutions to the obesity epidemic. Mil- mass index and physical activity. Am J Prev Med.
e559–e561. bank Q. 2009;87(1):123–154. 2010;39(2):105–112.

46 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

123. Creatore MI, Glazier RH, Moineddin R, Fazli GS, 137. Colchero MA, Rivera-Dommarco J, Popkin BM, Ng 147. Danforth E, Jr. Failure of adipocyte differentiation
Johns A, Gozdyra P, Matheson FI, Kaufman-Shriqui SW. In Mexico, evidence of sustained consumer causes type II diabetes mellitus? Nat Genet. 2000;
V, Rosella LC, Manuel DG, Booth GL. Association of response two years after implementing a sugar- 26(1):13.
neighborhood walkability with change in over- sweetened beverage tax. Health Aff (Millwood). 2017; 148. Bray GA, Redman LM, de Jonge L, Rood J, Smith SR.
weight, obesity, and diabetes. JAMA. 2016;315(20): 36(3):564–571. Effect of three levels of dietary protein on metabolic
2211–2220. 138. Centers for Disease Control and Prevention. Envi- phenotype of healthy individuals with 8 weeks of
124. Faith MS, Fontaine KR, Baskin ML, Allison DB. ronmental and policy approaches to increase overfeeding. J Clin Endocrinol Metab. 2016;101(7):
Toward the reduction of population obesity: physical activity: creation of or enhanced access to 2836–2843.
macrolevel environmental approaches to the places for physical activity combined with in- 149. Arner P, Arner E, Hammarstedt A, Smith U. Genetic
problems of food, eating, and obesity. Psychol Bull. formational outreach activities. Available at: http:// predisposition for type 2 diabetes, but not for
2007;133(2):205–226. overweight/obesity, is associated with a restricted
125. James W, Gill T. Prevention of obesity. In: Bray GA, policy/improvingaccess.html. Accessed 5 April 2015. adipogenesis. PLoS One. 2011;6(4):e18284.
Bouchard C, eds. Handbook of Obesity: Clinical Ap- 139. Centers for Disease Control and Prevention. 150. Tchoukalova YD, Votruba SB, Tchkonia T, Giorgadze
plications. 3rd ed. Boca Raton, FL: CRC Press; 2008: Strategies to Prevent Obesity and Other Chronic N, Kirkland JL, Jensen MD. Regional differences in
157–175. Diseases: The CDC Guide to Strategies to Increase cellular mechanisms of adipose tissue gain with
126. Dubowitz T, Ghosh-Dastidar M, Cohen DA, Beck- overfeeding. Proc Natl Acad Sci USA. 2010;107(42):
Physical Activity in the Community. Atlanta, GA: U.S.
man R, Steiner ED, Hunter GP, Flórez KR, Huang C, 18226–18231.
Department of Health and Human Services; 2011.
Vaughan CA, Sloan JC, Zenk SN, Cummins S, Collins 151. Fabbrini E, Yoshino J, Yoshino M, Magkos F,
140. Zhai F, Wang H, Du S, He Y, Wang Z, Ge K, Popkin
RL. Diet and perceptions change with supermarket Tiemann Luecking C, Samovski D, Fraterrigo G,
BM. Prospective study on nutrition transition in
introduction in a food desert, but not because of Okunade AL, Patterson BW, Klein S. Metabolically
China. Nutr Rev. 2009;67(Suppl 1):S56–S61.
supermarket use. Health Aff (Millwood). 2015;34(11): normal obese people are protected from adverse
141. Monda KL, Adair LS, Zhai F, Popkin BM. Longitu-
1858–1868. effects following weight gain. J Clin Invest. 2015;
dinal relationships between occupational and do-
127. Sacks FM, Svetkey LP, Vollmer WM, Appel LJ, Bray 125(2):787–795.
mestic physical activity patterns and body weight in 152. Wang ZV, Scherer PE. Adiponectin, the past two
GA, Harsha D, Obarzanek E, Conlin PR, Miller ER III, China. Eur J Clin Nutr. 2008;62(11):1318–1325.
Simons-Morton DG, Karanja N, Lin P-H, Aickin M, decades. J Mol Cell Biol. 2016;8(2):93–100.
142. Di Angelantonio E, Bhupathiraju ShN, Wormser D, 153. Hvidtfeldt UA, Gunter MJ, Lange T, Chlebowski RT,
Most-Windhauser MM, Moore TJ, Proschan MA, Gao P, Kaptoge S, Berrington de Gonzalez A, Cairns
Cutler JA; DASH-Sodium Collaborative Research Lane D, Farhat GN, Freiberg MS, Keiding N, Lee JS,
BJ, Huxley R, Jackson ChL, Joshy G, Lewington S, Prentice R, Tjønneland A, Vitolins MZ, Wassertheil-
Group. Effects on blood pressure of reduced dietary Manson JE, Murphy N, Patel AV, Samet JM,
sodium and the Dietary Approaches to Stop Hy- Smoller S, Strickler HD, Rod NH. Quantifying me-
Woodward M, Zheng W, Zhou M, Bansal N, diating effects of endogenous estrogen and insulin
pertension (DASH) diet. N Engl J Med. 2001;344(1): Barricarte A, Carter B, Cerhan JR, Smith GD, Fang X, in the relation between obesity, alcohol con-
3–10. Franco OH, Green J, Halsey J, Hildebrand JS, Jung KJ, sumption, and breast cancer. Cancer Epidemiol
128. Kennedy BM, Paeratakul S, Ryan DH, Bray GA. Korda RJ, McLerran DF, Moore SC, O’Keeffe LM, Biomarkers Prev. 2012;21(7):1203–1212.
Socioeconomic status and health disparity in the Paige E, Ramond A, Reeves GK, Rolland B, Sacerdote 154. DeFronzo RA, Ferrannini E, Groop L, Henry RR,
United States. J Hum Behav Soc Environ. 2007;
C, Sattar N, Sofianopoulou E, Stevens J, Thun M, Herman WH, Holst JJ, Hu FB, Kahn CR, Raz I,
Ueshima H, Yang L, Yun YD, Willeit P, Banks E, Beral Shulman GI, Simonson DC, Testa MA, Weiss R. Type
129. Glanz K, Yaroch AL. Strategies for increasing fruit
V, Chen Zh, Gapstur SM, Gunter MJ, Hartge P, Jee 2 diabetes mellitus. Nat Rev Dis Primers. 2015;1:
and vegetable intake in grocery stores and com-
SH, Lam TH, Peto R, Potter JD, Willett WC, 15019.
munities: policy, pricing, and environmental change.
Thompson SG, Danesh J, Hu FB; The Global 155. Galassi A, Reynolds K, He J. Metabolic syndrome
Prev Med. 2004;39(Suppl 2):S75–S80.
Mortality Collaboration. Body-mass index and and risk of cardiovascular disease: a meta-analysis.
130. Glanz K, Hoelscher D. Increasing fruit and vegetable
all-cause mortality: individual-participant-data Am J Med. 2006;119(10):812–819.
intake by changing environments, policy and
meta-analysis of 239 prospective studies in four 156. Sasai H, Sairenchi T, Iso H, Irie F, Otaka E, Tanaka K,
pricing: restaurant-based research, strategies, and
continents. Lancet. 2016;388:776–786. Ota H, Muto T. Relationship between obesity and
recommendations. Prev Med. 2004;39(Suppl 2): incident diabetes in middle-aged and older Japa-
143. Sjöström L, Narbro K, Sjöström CD, Karason K,
S88–S93. nese adults: the Ibaraki Prefectural Health Study.
Larsson B, Wedel H, Lystig T, Sullivan M, Bouchard C,
131. Buzby M. Evaluation of the USDA fruit and vegetable Mayo Clin Proc. 2010;85(1):36–40.
program: report to congress. Available at: https://www. Carlsson B, Bengtsson C, Dahlgren S, Gummesson A,
Jacobson P, Karlsson J, Lindroos A-K, Lönroth H, 157. Wang Y, Rimm EB, Stampfer MJ, Willett WC, Hu FB. Comparison of abdominal adiposity and overall
efan03006_002.pdf. Accessed 8 January 2018. Näslund I, Olbers T, Stenlöf K, Torgerson J, Ågren G,
Carlsson LMS; Swedish Obese Subjects Study. Ef- obesity in predicting risk of type 2 diabetes among
132. French SA, Wechsler H. School-based research and men. Am J Clin Nutr. 2005;81(3):555–563.
initiatives: fruit and vegetable environment, policy, fects of bariatric surgery on mortality in Swedish
obese subjects. N Engl J Med. 2007;357(8):741–752. 158. Knowler WC, Fowler SE, Hamman RF, Christophi
and pricing workshop. Prev Med. 2004;39(Suppl 2): CA, Hoffman HJ, Brenneman AT, Brown-Friday JO,
S101–S107. 144. Gregg EW, Jakicic JM, Blackburn G, Bloomquist P,
Goldberg R, Venditti E, Nathan DM; Diabetes
133. Appel LJ, Clark JM, Yeh H-C, Wang N-Y, Coughlin Bray GA, Clark JM, Coday M, Curtis JM, Egan C,
Prevention Program Research Group. 10-year
JW, Daumit G, Miller ER III, Dalcin A, Jerome GJ, Evans M, Foreyt J, Foster G, Hazuda HP, Hill JO,
follow-up of diabetes incidence and weight loss in
Geller S, Noronha G, Pozefsky T, Charleston J, Horton ES, Hubbard VS, Jeffery RW, Johnson KC,
the Diabetes Prevention Program Outcomes Study.
Reynolds JB, Durkin N, Rubin RR, Louis TA, Brancati Kitabchi AE, Knowler WC, Kriska A, Lang W, Lewis
Lancet. 2009;374(9702):1677–1686.
FL. Comparative effectiveness of weight-loss in- CE, Montez MG, Nathan DM, Neiberg RH, Patricio J,
159. Purnell JQ, Selzer F, Wahed AS, Pender J, Pories W,
terventions in clinical practice. N Engl J Med. 2011; Peters A, Pi-Sunyer X, Pownall H, Redmon B,
Pomp A, Dakin G, Mitchell J, Garcia L, Staten MA,
365(21):1959–1968. Regensteiner J, Rejeski J, Ribisl PM, Safford M, McCloskey C, Cummings DE, Flum DR, Courcoulas
134. Smed S, Jensen JD, Denver S. Socio-economic Stewart K, Trence D, Wadden TA, Wing RR, A, Wolfe BM. Type 2 diabetes remission rates after
characteristics and the effect of taxation as Yanovski SZ; Look AHEAD Research Group. As- laparoscopic gastric bypass and gastric banding:
a health policy instrument. Food Policy. 2007; sociation of the magnitude of weight loss and results of the longitudinal assessment of bariatric
32(5–6):624–639. changes in physical fitness with long-term cardio- surgery study. Diabetes Care. 2016;39(7):1101–1107.
135. Brownell KD, Farley T, Willett WC, Popkin BM, vascular disease outcomes in overweight or obese 160. Sjöström L, Lindroos A-K, Peltonen M, Torgerson J,
Chaloupka FJ, Thompson JW, Ludwig DS. The public people with type 2 diabetes: a post-hoc analysis of Bouchard C, Carlsson B, Dahlgren S, Larsson B,
health and economic benefits of taxing sugar- the Look AHEAD randomised clinical trial. Lancet Narbro K, Sjöström CD, Sullivan M, Wedel H;
sweetened beverages. N Engl J Med. 2009;361(16): Diabetes Endocrinol. 2016;4(11):913–921. Swedish Obese Subjects Study Scientific Group.
1599–1605. 145. Heymsfield SB, Wadden TA. Mechanisms, patho- Lifestyle, diabetes, and cardiovascular risk factors 10
136. Falbe J, Thompson HR, Becker CM, Rojas N, physiology, and management of obesity. N Engl J years after bariatric surgery. N Engl J Med. 2004;
McCulloch CE, Madsen KA. Impact of the Berkeley Med. 2017;376(3):254–266. 351(26):2683–2693.
excise tax on sugar-sweetened beverage con- 146. Halberg N, Wernstedt-Asterholm I, Scherer PE. The 161. Schauer PR, Bhatt DL, Kirwan JP, Wolski K, Aminian
sumption. Am J Public Health. 2016;106(10): adipocyte as an endocrine cell. Endocrinol Metab A, Brethauer SA, Navaneethan SD, Singh RP, Pothier
1865–1871. Clin North Am. 2008;37(3):753–768, x–xi. CE, Nissen SE, Kashyap SR; STAMPEDE Investigators.

doi: 10.1210/er.2017-00253 47

Downloaded from
by guest
on 17 March 2018

Bariatric surgery versus intensive medical therapy 178. Tsang TSM, Barnes ME, Miyasaka Y, Cha SS, Bailey 194. Young T, Peppard PE, Taheri S. Excess weight and
for diabetes—5-year outcomes. N Engl J Med. 2017; KR, Verzosa GC, Seward JB, Gersh BJ. Obesity as a risk sleep-disordered breathing. J Appl Physiol (1985).
376(7):641–651. factor for the progression of paroxysmal to per- 2005;99(4):1592–1599.
162. Brawer R, Brisbon N, Plumb J. Obesity and cancer. manent atrial fibrillation: a longitudinal cohort 195. Foster GD, Sanders MH, Millman R, Zammit G,
Prim Care. 2009;36(3):509–531. study of 21 years. Eur Heart J. 2008;29(18): Borradaile KE, Newman AB, Wadden TA, Kelley D,
163. Calle EE, Rodriguez C, Walker-Thurmond K, Thun 2227–2233. Wing RR, Sunyer FX, Darcey V, Kuna ST; Sleep
MJ. Overweight, obesity, and mortality from cancer 179. Aune D, Sen A, Norat T, Janszky I, Romundstad P, AHEAD Research Group. Obstructive sleep apnea
in a prospectively studied cohort of U.S. adults. Tonstad S, Vatten LJ. Body mass index, abdominal among obese patients with type 2 diabetes. Di-
N Engl J Med. 2003;348(17):1625–1638. fatness, and heart failure incidence and mortality: abetes Care. 2009;32(6):1017–1019.
164. Rosner B, Eliassen AH, Toriola AT, Hankinson SE, a systematic review and dose-response meta- 196. Clark JM. The epidemiology of nonalcoholic fatty
Willett WC, Natarajan L, Colditz GA. Short-term analysis of prospective studies. Circulation. 2016; liver disease in adults. J Clin Gastroenterol. 2006;
weight gain and breast cancer risk by hormone 133(7):639–649. 40(Suppl 1):S5–S10.
receptor classification among pre- and post- 180. Bahrami H, Bluemke DA, Kronmal R, Bertoni AG, 197. Cydylo MA, Davis AT, Kavanagh K. Fatty liver
menopausal women. Breast Cancer Res Treat. 2015; Lloyd-Jones DM, Shahar E, Szklo M, Lima JAC. Novel promotes fibrosis in monkeys consuming high
150(3):643–653. metabolic risk factors for incident heart failure and fructose. Obesity (Silver Spring). 2017;25(2):290–293.
165. Aune D, Navarro Rosenblatt DA, Chan DSM, their relationship with obesity: the MESA (Multi- 198. Than NN, Newsome PN. A concise review of non-
Vingeliene S, Abar L, Vieira AR, Greenwood DC, Ethnic Study of Atherosclerosis) study. J Am Coll alcoholic fatty liver disease. Atherosclerosis. 2015;
Bandera EV, Norat T. Anthropometric factors and Cardiol. 2008;51(18):1775–1783. 239(1):192–202.
endometrial cancer risk: a systematic review and 181. Lavie CJ, Milani RV, Artham SM, Patel DA, Ventura 199. Younossi ZM, Stepanova M, Afendy M, Fang Y,
dose-response meta-analysis of prospective studies. HO. The obesity paradox, weight loss, and coronary Younossi Y, Mir H, Srishord M. Changes in the
Ann Oncol. 2015;26(8):1635–1648. disease. Am J Med. 2009;122(12):1106–1114. prevalence of the most common causes of chronic
166. Eliassen AH, Colditz GA, Rosner B, Willett WC, 182. Lavie CJ, Sharma A, Alpert MA, De Schutter A, liver diseases in the United States from 1988 to
Hankinson SE. Adult weight change and risk of Lopez-Jimenez F, Milani RV, Ventura HO. Update 2008. Clin Gastroenterol Hepatol. 2011;9(6):524–530.e1.
postmenopausal breast cancer. JAMA. 2006;296(2): on obesity and obesity paradox in heart failure. Prog 200. Li L, Gan Y, Li W, Wu C, Lu Z. Overweight, obesity
193–201. Cardiovasc Dis. 2016;58(4):393–400. and the risk of gallbladder and extrahepatic bile
167. van Kruijsdijk RCM, van der Wall E, Visseren FLJ. 183. Lavie CJ, Arena R, Alpert MA, Milani RV, Ventura duct cancers: a meta-analysis of observational
Obesity and cancer: the role of dysfunctional adi- HO. Management of cardiovascular diseases in studies. Obesity (Silver Spring). 2016;24(8):
pose tissue. Cancer Epidemiol Biomarkers Prev. 2009; patients with obesity. Nat Rev Cardiol. 2018;15(1): 1786–1802.
18(10):2569–2578. 45–56. 201. Pang Q, Zhang JY, Song SD, Qu K, Xu XS, Liu SS, Liu
168. Sweeney C, Blair CK, Anderson KE, Lazovich D, 184. Lajous M, Banack HR, Kaufman JS, Hernán MA. C. Central obesity and nonalcoholic fatty liver
Folsom AR. Risk factors for breast cancer in elderly Should patients with chronic disease be told to gain
disease risk after adjusting for body mass index.
women. Am J Epidemiol. 2004;160(9):868–875. weight? The obesity paradox and selection bias. Am
World J Gastroenterol. 2015;21(5):1650–1662.
169. Schairer C, Fuhrman BJ, Boyd-Morin J, Genkinger JM, J Med. 2015;128(4):334–336.
202. Aune D, Norat T, Vatten LJ. Body mass index and
Gail MH, Hoover RN, Ziegler RG. Quantifying the 185. Adams KF, Schatzkin A, Harris TB, Kipnis V, Mouw
the risk of gout: a systematic review and dose-
role of circulating unconjugated estradiol in me- T, Ballard-Barbash R, Hollenbeck A, Leitzmann MF.
response meta-analysis of prospective studies. Eur J
diating the body mass index-breast cancer associ- Overweight, obesity, and mortality in a large pro-
Nutr. 2014;53(8):1591–1601.
ation. Cancer Epidemiol Biomarkers Prev. 2016;25(1): spective cohort of persons 50 to 71 years old. N Engl
203. Felson DT, Anderson JJ, Naimark A, Walker AM,
105–113. J Med. 2006;355(8):763–778.
Meenan RF. Obesity and knee osteoarthritis. The
170. Kenchaiah S, Evans JC, Levy D, Wilson PWF, Ben- 186. Srikanthan P, Karlamangla AS. Muscle mass index as
Framingham Study. Ann Intern Med. 1988;109(1):
jamin EJ, Larson MG, Kannel WB, Vasan RS. Obesity a predictor of longevity in older adults. Am J Med.
and the risk of heart failure. N Engl J Med. 2002; 2014;127(6):547–553.
204. Okoro CA, Hootman JM, Strine TW, Balluz LS,
347(5):305–313. 187. Padwal R, Leslie WD, Lix LM, Majumdar SR. Re-
Mokdad AH. Disability, arthritis, and body weight
171. Canoy D, Cairns BJ, Balkwill A, Wright FL, Green J, lationship among body fat percentage, body mass
Reeves G, Beral V; Million Women Study Collab- among adults 45 years and older. Obes Res. 2004;
index, and all-cause mortality. Ann Intern Med. 2016;
orators. Body mass index and incident coronary 164(8):532–541. 12(5):854–861.
heart disease in women: a population-based pro- 188. World Health Organization. World health statistics 205. Marchi J, Berg M, Dencker A, Olander EK, Begley C.
spective study. BMC Med. 2013;11(1):87. 2015. Available at: Risks associated with obesity in pregnancy, for the
172. Aune D, Sen A, Prasad M, Norat T, Janszky I, Tonstad publications/world_health_statistics/2015/en/. mother and baby: a systematic review of reviews.
S, Romundstad P, Vatten LJ. BMI and all cause Accessed March 2017. Obes Rev. 2015;16(8):621–638.
mortality: systematic review and non-linear dose- 189. Nwankwo T, Yoon SS, Burt V, Gu Q. Hypertension 206. Bray GA, Siri-Tarino PW. The role of macronutrient
response meta-analysis of 230 cohort studies with among adults in the united states: National health content in the diet for weight management.
3.74 million deaths among 30.3 million participants. and nutrition examination survey, 2011–2012. Endocrinol Metab Clin North Am. 2016;45(3):
BMJ. 2016;353:i2156. NCHS Data Brief. 2013;2013(133):1–8. 581–604.
173. Flint AJ, Hu FB, Glynn RJ, Caspard H, Manson JE, 190. Ettehad D, Emdin CA, Kiran A, Anderson SG, 207. Hamman RF, Wing RR, Edelstein SL, Lachin JM, Bray
Willett WC, Rimm EB. Excess weight and the risk of Callender T, Emberson J, Chalmers J, Rodgers A, GA, Delahanty L, Hoskin M, Kriska AM, Mayer-Davis
incident coronary heart disease among men and Rahimi K. Blood pressure lowering for prevention of EJ, Pi-Sunyer X, Regensteiner J, Venditti B, Wylie-
women. Obesity (Silver Spring). 2010;18(2):377–383. cardiovascular disease and death: a systematic re- Rosett J. Effect of weight loss with lifestyle in-
174. Emberson JR, Whincup PH, Morris RW, Wanna- view and meta-analysis. Lancet. 2016;387(10022): tervention on risk of diabetes. Diabetes Care. 2006;
methee SG, Shaper AG. Lifestyle and cardiovascular 957–967. 29(9):2102–2107.
disease in middle-aged British men: the effect of 191. Stevens VJ, Obarzanek E, Cook NR, Lee IM, Appel LJ, 208. Antonetti VW. The equations governing weight
adjusting for within-person variation. Eur Heart J. Smith West D, Milas NC, Mattfeldt-Beman M, change in human beings. Am J Clin Nutr. 1973;26(1):
2005;26(17):1774–1782. Belden L, Bragg C, Millstone M, Raczynski J, Brewer 64–71.
175. Janssen I, Katzmarzyk PT, Ross R. Waist circum- A, Singh B, Cohen J; Trials for the Hypertension 209. Thomas DM, Gonzalez MC, Pereira AZ, Redman
ference and not body mass index explains obesity- Prevention Research Group. Long-term weight loss LM, Heymsfield SB. Time to correctly predict the
related health risk. Am J Clin Nutr. 2004;79(3): and changes in blood pressure: results of the Trials amount of weight loss with dieting. J Acad Nutr Diet.
379–384. of Hypertension Prevention, phase II. Ann Intern 2014;114(6):857–861.
176. Lee CMY, Huxley RR, Wildman RP, Woodward M. Med. 2001;134(1):1–11. 210. Thomas DM, Weedermann M, Fuemmeler BF,
Indices of abdominal obesity are better discrimi- 192. Neter JE, Stam BE, Kok FJ, Grobbee DE, Geleijnse JM. Martin CK, Dhurandhar NV, Bredlau C, Heymsfield
nators of cardiovascular risk factors than BMI: Influence of weight reduction on blood pressure: SB, Ravussin E, Bouchard C. Dynamic model pre-
a meta-analysis. J Clin Epidemiol. 2008;61(7):646–653. a meta-analysis of randomized controlled trials. dicting overweight, obesity, and extreme obesity
177. Wang TJ, Parise H, Levy D, D’Agostino RB, Sr, Wolf Hypertension. 2003;42(5):878–884. prevalence trends. Obesity (Silver Spring). 2014;22(2):
PA, Vasan RS, Benjamin EJ. Obesity and the risk of 193. Coccagna G, Pollini A, Provini F. Cardiovascular 590–597.
new-onset atrial fibrillation. JAMA. 2004;292(20): disorders and obstructive sleep apnea syndrome. 211. Hall KD, Sacks G, Chandramohan D, Chow CC,
2471–2477. Clin Exp Hypertens. 2006;28(3–4):217–224. Wang YC, Gortmaker SL, Swinburn BA.

48 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

Quantification of the effect of energy imbalance on 224. Qi Q, Xu M, Wu H, Liang L, Champagne CM, Bray 238. Ello-Martin JA, Roe LS, Ledikwe JH, Beach AM, Rolls
bodyweight. Lancet. 2011;378(9793):826–837. GA, Sacks FM, Qi L. IRS1 genotype modulates BJ. Dietary energy density in the treatment of
212. Sumithran P, Prendergast LA, Delbridge E, Purcell K, metabolic syndrome reversion in response to 2-year obesity: a year-long trial comparing 2 weight-loss
Shulkes A, Kriketos A, Proietto J. Long-term per- weight-loss diet intervention: the POUNDS LOST diets. Am J Clin Nutr. 2007;85(6):1465–1477.
sistence of hormonal adaptations to weight loss. trial. Diabetes Care. 2013;36(11):3442–3447. 239. Rolls BJ, Barnett RA. Volumetrics. A Systematic
N Engl J Med. 2011;365(17):1597–1604. 225. Xu M, Qi Q, Liang J, Bray GA, Hu FB, Sacks FM, Qi L. Lifetime Approach to Eating. Proven Methods for
213. Hron BM, Ebbeling CB, Feldman HA, Ludwig DS. Genetic determinant for amino acid metabolites Satisfying Hunger. Increase Food Volume Without
Hepatic, adipocyte, enteric and pancreatic hor- and changes in body weight and insulin re- Gaining Weight. Sound Recipes and Menus for
mones: response to dietary macronutrient com- sistance in response to weight-loss diets: the Weight Loss. New York, NY: Harper-Collins, Inc;2000.
position and relationship with metabolism. Nutr Preventing Overweight Using Novel Dietary 240. Ebbeling CB, Leidig MM, Sinclair KB, Hangen JP,
Metab (Lond). 2017;14(1):44. Strategies (POUNDS LOST) trial. Circulation. Ludwig DS. A reduced-glycemic load diet in the
214. Gavrieli A, Mantzoros CS. Novel molecules regu- 2013;127(12):1283–1289. treatment of adolescent obesity. Arch Pediatr
lating energy homeostasis: physiology and regula- 226. Mirzaei K, Xu M, Qi Q, de Jonge L, Bray GA, Sacks F, Adolesc Med. 2003;157(8):773–779.
tion by macronutrient intake and weight loss. Qi L. Variants in glucose- and circadian rhythm- 241. Ebbeling CB, Leidig MM, Feldman HA, Lovesky MM,
Endocrinol Metab (Seoul). 2016;31(3):361–372. related genes affect the response of energy ex- Ludwig DS. Effects of a low-glycemic load vs low-fat
215. Espeland MA, Bray GA, Neiberg R, Rejeski WJ, penditure to weight-loss diets: the POUNDS LOST diet in obese young adults: a randomized trial.
Knowler WC, Lang W, Cheskin LJ, Williamson D, Trial. Am J Clin Nutr. 2014;99(2):392–399. JAMA. 2007;297(19):2092–2102.
Lewis CB, Wing R; Look Ahead Study Group. De- 227. Zheng Y, Huang T, Zhang X, Rood J, Bray GA, Sacks 242. Thomas DE, Elliott EJ, Baur L. Low glycaemic index
scribing patterns of weight changes using principal FM, Qi L. Dietary fat modifies the effects of fto or low glycaemic load diets for overweight and
components analysis: results from the Action for genotype on changes in insulin sensitivity. J Nutr. obesity. Cochrane Database Syst Rev. 2007; (3):
Health in Diabetes (Look AHEAD) research group. 2015;145(5):977–982. CD005105.
Ann Epidemiol. 2009;19(10):701–710. 228. Livingstone KM, Celis-Morales C, Papandonatos 243. Hjorth MF, Ritz C, Blaak EE, Saris WH, Langin D,
216. Dansinger ML, Gleason JA, Griffith JL, Selker HP, GD, Erar B, Florez JC, Jablonski KA, Razquin C, Marti Poulsen SK, Larsen TM, Sørensen TI, Zohar Y, Astrup
Schaefer EJ. Comparison of the Atkins, Ornish, A, Heianza Y, Huang T, Sacks FM, Svendstrup M, Sui A. Pretreatment fasting plasma glucose and insulin
Weight Watchers, and Zone diets for weight loss X, Church TS, Jääskeläinen T, Lindström J, Tuomilehto modify dietary weight loss success: results from 3
and heart disease risk reduction: a randomized trial. J, Uusitupa M, Rankinen T, Saris WH, Hansen T, randomized clinical trials. Am J Clin Nutr. 2017;
JAMA. 2005;293(1):43–53. Pedersen O, Astrup A, Sørensen TI, Qi L, Bray GA, 106(2):499–505.
217. Alhassan S, Kim S, Bersamin A, King AC, Gardner Martinez-Gonzalez MA, Martinez JA, Franks PW, 244. Skov AR, Toubro S, Rønn B, Holm L, Astrup A.
CD. Dietary adherence and weight loss success McCaffery JM, Lara J, Mathers JC. FTO genotype and Randomized trial on protein vs carbohydrate in ad
among overweight women: results from the A TO Z libitum fat reduced diet for the treatment of
weight loss: systematic review and meta-analysis of
weight loss study. Int J Obes. 2008;32(6):985–991. obesity. Int J Obes Relat Metab Disord. 1999;23(5):
9563 individual participant data from eight rand-
218. Sacks FM, Bray GA, Carey VJ, Smith SR, Ryan DH, 528–536.
omised controlled trials. BMJ. 2016;354:i4707.
Anton SD, McManus K, Champagne CM, Bishop 245. Due A, Toubro S, Skov AR, Astrup A. Effect of
229. Tsai AG, Wadden TA. The evolution of very-low-
LM, Laranjo N, Leboff MS, Rood JC, de Jonge L, normal-fat diets, either medium or high in protein,
calorie diets: an update and meta-analysis. Obesity
Greenway FL, Loria CM, Obarzanek E, Williamson on body weight in overweight subjects: a rando-
(Silver Spring). 2006;14(8):1283–1293.
DA. Comparison of weight-loss diets with different mised 1-year trial. Int J Obes Relat Metab Disord.
230. Bray GA, Popkin BM. Dietary sugar and body
compositions of fat, protein, and carbohydrates. 2004;28(10):1283–1290.
weight: have we reached a crisis in the epidemic of
N Engl J Med. 2009;360(9):859–873. 246. Wycherley TP, Moran LJ, Clifton PM, Noakes M,
obesity and diabetes?: health be damned! Pour on
219. Delahanty LM, Pan Q, Jablonski KA, Watson KE, Brinkworth GD. Effects of energy-restricted high-
the sugar. Diabetes Care. 2014;37(4):950–956.
McCaffery JM, Shuldiner A, Kahn SE, Knowler WC, protein, low-fat compared with standard-protein,
231. Nordmann AJ, Nordmann A, Briel M, Keller U,
Florez JC, Franks PW; Diabetes Prevention Program low-fat diets: a meta-analysis of randomized con-
Yancy WS, Jr, Brehm BJ, Bucher HC. Effects of low-
Research Group. Genetic predictors of weight loss trolled trials. Am J Clin Nutr. 2012;96(6):1281–1298.
carbohydrate vs low-fat diets on weight loss and
and weight regain after intensive lifestyle modifi- 247. Bray GA, Ryan DH, Johnson W, Champagne CM,
cation, metformin treatment, or standard care in cardiovascular risk factors: a meta-analysis of ran- Johnson CM, Rood J, Williamson DA, Sacks FM.
the Diabetes Prevention Program. Diabetes Care. domized controlled trials. Arch Intern Med. 2006; Markers of dietary protein intake are associated
2012;35(2):363–366. 166(3):285–293. with successful weight loss in the POUNDS Lost
220. Papandonatos GD, Pan Q, Pajewski NM, Delahanty 232. Tobias DK, Chen M, Manson JE, Ludwig DS, Willett trial. Clin Obes. 2017;7(3):166–175.
LM, Peter I, Erar B, Ahmad S, Harden M, Chen L, W, Hu FB. Effect of low-fat diet interventions versus 248. Esposito K, Ida Maiorino M, Ciotola M, Di Palo C,
Fontanillas P, Wagenknecht LE, Kahn SE, Wing RR, other diet interventions on long-term weight Scognamiglio P, Gicchino M, Petrizzo M, Saccomanno
Jablonski KA, Huggins GS, Knowler WC, Florez JC, change in adults: a systematic review and meta- F, Beneduce F, Ceriello A, Giugliano D. Effects of
McCaffery JM, Franks PW; Diabetes Prevention analysis. Lancet Diabetes Endocrinol. 2015;3(12): a Mediterranean-style diet on the need for anti-
Program and the Look AHEAD Research Groups. 968–979. hyperglycemic drug therapy in patients with newly
Genetic predisposition to weight loss and regain 233. Bueno NB, de Melo ISV, de Oliveira SL, da Rocha diagnosed type 2 diabetes: a randomized trial. Obstet
with lifestyle intervention: analyses from the di- Ataide T. Very-low-carbohydrate ketogenic diet v. Gynecol Surv. 2010;65(6):379–380.
abetes prevention program and the look ahead low-fat diet for long-term weight loss: a meta- 249. Nordmann AJ, Suter-Zimmermann K, Bucher HC,
randomized controlled trials. Diabetes. 2015;64(12): analysis of randomised controlled trials. Br J Nutr. Shai I, Tuttle KR, Estruch R, Briel M. Meta-analysis
4312–4321. 2013;110(7):1178–1187. comparing Mediterranean to low-fat diets for
221. Qi Q, Bray GA, Hu FB, Sacks FM, Qi L. Weight-loss 234. Bray GA, Smith SR, de Jonge L, Xie H, Rood J, Martin modification of cardiovascular risk factors. Am J
diets modify glucose-dependent insulinotropic CK, Most M, Brock C, Mancuso S, Redman LM. Med. 2011;124(9):841–851.e2.
polypeptide receptor rs2287019 genotype effects Effect of dietary protein content on weight gain, 250. Estruch R, Martı́nez-González MA, Corella D, Salas-
on changes in body weight, fasting glucose, and energy expenditure, and body composition during Salvadó J, Fitó M, Chiva-Blanch G, Fiol M, Gómez-
insulin resistance: the Preventing Overweight Using overeating: a randomized controlled trial. JAMA. Gracia E, Arós F, Lapetra J, Serra-Majem L, Pintó X,
Novel Dietary Strategies trial. Am J Clin Nutr. 2012; 2012;307(1):47–55. Buil-Cosiales P, Sorlı́ JV, Muñoz MA, Basora-Gallisá J,
95(2):506–513. 235. Hall KD, Guo J. Obesity energetics: body weight Lamuela-Raventós RM, Serra-Mir M, Ros E;
222. Zhang X, Qi Q, Zhang C, Smith SR, Hu FB, Sacks FM, regulation and the effects of diet composition. PREDIMED Study Investigators. Effect of a high-fat
Bray GA, Qi L. FTO genotype and 2-year change in Gastroenterology. 2017;152(7):1718–1727.e3. Mediterranean diet on bodyweight and waist cir-
body composition and fat distribution in response 236. Astrup A, Grunwald GK, Melanson EL, Saris WHM, cumference: a prespecified secondary outcomes
to weight-loss diets: the POUNDS LOST Trial. Di- Hill JO. The role of low-fat diets in body weight analysis of the PREDIMED randomised controlled
abetes. 2012;61(11):3005–3011. control: a meta-analysis of ad libitum dietary in- trial. Lancet Diabetes Endocrinol. 2016;4(8):666–676.
223. Zhang X, Qi Q, Bray GA, Hu FB, Sacks FM, Qi L. tervention studies. Int J Obes Relat Metab Disord. 251. Avenell A, Broom J, Brown TJ, Poobalan A, Aucott L,
APOA5 genotype modulates 2-y changes in lipid 2000;24(12):1545–1552. Stearns SC, Smith WC, Jung RT, Campbell MK,
profile in response to weight-loss diet intervention: 237. Pirozzo S, Summerbell C, Cameron C, Glasziou P. Grant AM. Systematic review of the long-term
the Pounds Lost Trial. Am J Clin Nutr. 2012;96(4): Should we recommend low-fat diets for obesity? effects and economic consequences of treat-
917–922. Obes Rev. 2003;4(2):83–90. ments for obesity and implications for health

doi: 10.1210/er.2017-00253 49

Downloaded from
by guest
on 17 March 2018

improvement. Health Technol Assess. 2004;8(21): 265. Nidetch J. The Story of Weight Watchers. New York, 281. Bray GA. From farm to fat cell: why aren’t we all fat?
iii–iv, 1–182. NY: New American Library; 1972. Metabolism. 2015;64(3):349–353.
252. Heymsfield SB, van Mierlo CAJ, van der Knaap HCM, 266. Sears B. The Zone: A Revolutionary Life Plan to Put 282. Westerterp KR, Speakman JR. Physical activity en-
Heo M, Frier HI. Weight management using a meal Your Body in Total Balance for Permanent Weight ergy expenditure has not declined since the 1980s
replacement strategy: meta and pooling analysis Loss. New York, NY: HarperCollins Publishers; 1995. and matches energy expenditures of wild mammals.
from six studies. Int J Obes Relat Metab Disord. 2003; 267. Johnston BC, Kanters S, Bandayrel K, Wu P, Naji F, Int J Obes. 2008;32(8):1256–1263.
27(5):537–549. Siemieniuk RA, Ball GD, Busse JW, Thorlund K, 283. Swinburn B, Sacks G, Ravussin E. Increased food
253. Ptomey LT, Willis EA, Goetz JR, Lee J, Szabo-Reed Guyatt G, Jansen JP, Mills EJ. Comparison of weight energy supply is more than sufficient to explain the
AN, Sullivan DK, Donnelly JE. Portion-controlled loss among named diet programs in overweight US epidemic of obesity. Am J Clin Nutr. 2009;90(6):
meals provide increases in diet quality during and obese adults: a meta-analysis. JAMA. 2014; 1453–1456.
weight loss and maintenance. J Hum Nutr Diet. 2016; 312(9):923–933. 284. Stubbe JH, Boomsma DI, De Geus EJ. Sports par-
29(2):209–216. 268. National Institutes of Health. Pounds lost study. ticipation during adolescence: a shift from envi-
254. Gardner CD, Kiazand A, Alhassan S, Kim S, Stafford Available at: ronmental to genetic factors. Med Sci Sports Exerc.
RS, Balise RR, Kraemer HC, King AC. Comparison of resources/heart/other-pounds-lost. Accessed 31 2005;37(4):563–570.
the Atkins, Zone, Ornish, and LEARN diets for January 2017. 285. Samaras K, Kelly PJ, Chiano MN, Spector TD,
change in weight and related risk factors among 269. Gudzune KA, Doshi RS, Mehta AK, Chaudhry ZW, Campbell LV. Genetic and environmental in-
overweight premenopausal women: the A TO Z Jacobs DK, Vakil RM, Lee CJ, Bleich SN, Clark JM. fluences on total-body and central abdominal fat:
Weight Loss Study: a randomized trial. JAMA. 2007; Efficacy of commercial weight-loss programs: an the effect of physical activity in female twins. Ann
297(9):969–977. updated systematic review. Ann Intern Med. 2015; Intern Med. 1999;130(11):873–882.
255. Shai I, Schwarzfuchs D, Henkin Y, Shahar DR, 162(7):501–512. 286. Donnelly JE, Blair SN, Jakicic JM, Manore MM,
Witkow S, Greenberg I, Golan R, Fraser D, Bolotin A, 270. Venditti EM, Bray GA, Carrion-Petersen ML, Rankin JW, Smith BK; American College of Sports
Vardi H, Tangi-Rozental O, Zuk-Ramot R, Sarusi B, Delahanty LM, Edelstein SL, Hamman RF, Hoskin MA, Medicine. American College of Sports Medicine
Brickner D, Schwartz Z, Sheiner E, Marko R, Katorza Knowler WC, Ma Y; Diabetes Prevention Program position stand. Appropriate physical activity in-
E, Thiery J, Fiedler GM, Blüher M, Stumvoll M, Research Group. First versus repeat treatment with tervention strategies for weight loss and prevention
Stampfer MJ; Dietary Intervention Randomized a lifestyle intervention program: attendance and of weight regain for adults. Med Sci Sports Exerc.
Controlled Trial (DIRECT) Group. Weight loss with weight loss outcomes. Int J Obes. 2008;32(10): 2009;41(2):459–471.
a low-carbohydrate, Mediterranean, or low-fat diet. 1537–1544. 287. Swift DL, Johannsen NM, Lavie CJ, Earnest CP,
N Engl J Med. 2008;359(3):229–241. 271. Look AHEAD Research Group. Eight-year weight Church TS. The role of exercise and physical activity
256. Bazzano LA, Hu T, Reynolds K, Yao L, Bunol C, Liu Y, losses with an intensive lifestyle intervention: the in weight loss and maintenance. Prog Cardiovasc Dis.
Chen CS, Klag MJ, Whelton PK, He J. Effects of low- look AHEAD study. Obesity (Silver Spring). 2014; 2014;56(4):441–447.
288. US Department of Health and Human Services.
carbohydrate and low-fat diets: a randomized trial. 22(1):5–13.
Physical activity guidelines advisory committee
Ann Intern Med. 2014;161(5):309–318. 272. Westerterp-Plantenga MS, Lejeune MPGM, Nijs I,
report 2008. Available at:
257. Das SK, Gilhooly CH, Golden JK, Pittas AG, Fuss PJ, van Ooijen M, Kovacs EMR. High protein intake
paguidelines/guidelines/. Accessed 10 March 2017.
Cheatham RA, Tyler S, Tsay M, McCrory MA, sustains weight maintenance after body weight loss
289. Fan JX, Brown BB, Hanson H, Kowaleski-Jones L,
Lichtenstein AH, Dallal GE, Dutta C, Bhapkar MV, in humans. Int J Obes Relat Metab Disord. 2004;28(1):
Smith KR, Zick CD. Moderate to vigorous physical
Delany JP, Saltzman E, Roberts SB. Long-term effects 57–64.
activity and weight outcomes: does every minute
of 2 energy-restricted diets differing in glycemic 273. Larsen TM, Dalskov S-M, van Baak M, Jebb SA,
count? Am J Health Promot. 2013;28(1):41–49.
load on dietary adherence, body composition, and Papadaki A, Pfeiffer AFH, Martinez JA, Handjieva-
290. Kimm SY, Glynn NW, Kriska AM, Fitzgerald SL,
metabolism in CALERIE: a 1-y randomized con- Darlenska T, Kunešová M, Pihlsgård M, Stender S,
Aaron DJ, Similo SL, McMahon RP, Barton BA.
trolled trial. Am J Clin Nutr. 2007;85(4):1023–1030. Holst C, Saris WHM, Astrup A; Diet, Obesity, and
Longitudinal changes in physical activity in a biracial
258. Fabricatore AN, Wadden TA, Ebbeling CB, Thomas Genes (Diogenes) Project. Diets with high or low
cohort during adolescence. Med Sci Sports Exerc.
JG, Stallings VA, Schwartz S, Ludwig DS. Targeting protein content and glycemic index for weight-loss 2000;32(8):1445–1454.
dietary fat or glycemic load in the treatment of maintenance. N Engl J Med. 2010;363(22): 291. DeLany JP, Bray GA, Harsha DW, Volaufova J. Energy
obesity and type 2 diabetes: a randomized con- 2102–2113. expenditure in African American and white boys
trolled trial. Diabetes Res Clin Pract. 2011;92(1): 274. Howard BV, Manson JE, Stefanick ML, Beresford SA, and girls in a 2-y follow-up of the Baton Rouge
37–45. Frank G, Jones B, Rodabough RJ, Snetselaar L, Children’s Study. Am J Clin Nutr. 2004;79(2):
259. Foster GD, Wyatt HR, Hill JO, McGuckin BG, Brill C, Thomson C, Tinker L, Vitolins M, Prentice R. Low-fat 268–273.
Mohammed BS, Szapary PO, Rader DJ, Edman JS, dietary pattern and weight change over 7 years: the 292. Kimm SY, Glynn NW, Kriska AM, Barton BA,
Klein S. A randomized trial of a low-carbohydrate Women’s Health Initiative Dietary Modification Kronsberg SS, Daniels SR, Crawford PB, Sabry ZI, Liu
diet for obesity. N Engl J Med. 2003;348(21): Trial. JAMA. 2006;295(1):39–49. K. Decline in physical activity in black girls and white
2082–2090. 275. Hooper L, Abdelhamid A, Moore HJ, Douthwaite W, girls during adolescence. N Engl J Med. 2002;347(10):
260. Foster GD, Wyatt HR, Hill JO, Makris AP, Rosenbaum Skeaff CM, Summerbell CD. Effect of reducing total 709–715.
DL, Brill C, Stein RI, Mohammed BS, Miller B, Rader DJ, fat intake on body weight: systematic review and 293. Church TS, Thomas DM, Tudor-Locke C, Katz-
Zemel B, Wadden TA, Tenhave T, Newcomb CW, meta-analysis of randomised controlled trials and marzyk PT, Earnest CP, Rodarte RQ, Martin CK, Blair
Klein S. Weight and metabolic outcomes after 2 years cohort studies. BMJ. 2012;345:e7666. SN, Bouchard C. Trends over 5 decades in U.S.
on a low-carbohydrate versus low-fat diet: a ran- 276. Wing RR, Hill JO. Successful weight loss mainte- occupation-related physical activity and their as-
domized trial. Ann Intern Med. 2010;153(3):147–157. nance. Annu Rev Nutr. 2001;21(1):323–341. sociations with obesity. PLoS One. 2011;6(5):e19657.
261. Stern L, Iqbal N, Seshadri P, Chicano KL, Daily DA, 277. Wing RR, Tate DF, Gorin AA, Raynor HA, Fava JL. A 294. Ball K, Brown W, Crawford D. Who does not gain
McGrory J, Williams M, Gracely EJ, Samaha FF. The self-regulation program for maintenance of weight weight? Prevalence and predictors of weight
effects of low-carbohydrate versus conventional loss. N Engl J Med. 2006;355(15):1563–1571. maintenance in young women. Int J Obes Relat
weight loss diets in severely obese adults: one-year 278. Steinberg DM, Tate DF, Bennett GG, Ennett S, Metab Disord. 2002;26(12):1570–1578.
follow-up of a randomized trial. Ann Intern Med. Samuel-Hodge C, Ward DS. The efficacy of a daily 295. Hu FB, Li TY, Colditz GA, Willett WC, Manson JE.
2004;140(10):778–785. self-weighing weight loss intervention using smart Television watching and other sedentary behaviors
262. Yancy WS, Jr, Olsen MK, Guyton JR, Bakst RP, scales and e-mail. Obesity. 2013;21(9):1789–1797. in relation to risk of obesity and type 2 diabetes
Westman EC. A low-carbohydrate, ketogenic diet 279. Eckel RH, Hernandez TL, Bell ML, Weil KM, Shepard mellitus in women. JAMA. 2003;289(14):1785–1791.
versus a low-fat diet to treat obesity and hyper- TY, Grunwald GK, Sharp TA, Francis CC, Hill JO. 296. Östman M, Britton E, Jonsson E. Physical exercise. In:
lipidemia: a randomized, controlled trial. Ann Intern Carbohydrate balance predicts weight and fat gain Östman M, Britton E, Jonsson E, eds. Treating and
Med. 2004;140(10):769–777. in adults. Am J Clin Nutr. 2006;83(4):803–808. Preventing Obesity. Weinheim, Germany: Wiley-
263. Atkins R. Dr. Atkins’ Diet Revolution: The High Calorie 280. Fung TT, Pan A, Hou T, Chiuve SE, Tobias DK, VCH Verlag; 2004:142–143.
Way to Stay Thin Forever. New York, NY: David Mozaffarian D, Willett WC, Hu FB. Long-term 297. Wood PD, Stefanick ML, Dreon DM, Frey-Hewitt B,
McKay; 1972. change in diet quality is associated with body Garay SC, Williams PT, Superko HR, Fortmann SP,
264. Ornish D. What if Americans ate less fat? JAMA. weight change in men and women. J Nutr. 2015; Albers JJ, Vranizan KM, Ellsworth NM, Terry RB,
1992;267(3):362, author reply 363–364. 145(8):1850–1856. Haskell WL. Changes in plasma lipids and

50 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

lipoproteins in overweight men during weight loss mail and telephone. J Consult Clin Psychol. 1984; 326. Bray GA, Hollander P, Klein S, Kushner R, Levy B,
through dieting as compared with exercise. N Engl J 52(3):404–413. Fitchet M, Perry BH. A 6-month randomized,
Med. 1988;319(18):1173–1179. 310. Alamuddin N, Wadden TA. Behavioral treatment of placebo-controlled, dose-ranging trial of topiramate
298. Wing RR, Epstein LH, Paternostro-Bayles M, Kriska the patient with obesity. Endocrinol Metab Clin for weight loss in obesity. Obes Res. 2003;11(6):
A, Nowalk MP, Gooding W. Exercise in a behav- North Am. 2016;45(3):565–580. 722–733.
ioural weight control programme for obese patients 311. Jakicic JM, Marcus BH, Gallagher KI, Napolitano M, 327. Heil GC, Ross ST. Agents affecting appetite. Annu
with type 2 (non-insulin-dependent) diabetes. Lang W. Effect of exercise duration and intensity on Rep Med Chem. 1973;(8):42–51.
Diabetologia. 1988;31(12):902–909. weight loss in overweight, sedentary women: 328. Heymsfield SB, Greenberg AS, Fujioka K, Dixon RM,
299. Wood PD, Stefanick ML, Williams PT, Haskell WL. a randomized trial. JAMA. 2003;290(10):1323–1330. Kushner R, Hunt T, Lubina JA, Patane J, Self B, Hunt
The effects on plasma lipoproteins of a prudent 312. Wadden TA, Vogt RA, Andersen RE, Bartlett SJ, P, McCamish M. Recombinant leptin for weight loss
weight-reducing diet, with or without exercise, in Foster GD, Kuehnel RH, Wilk J, Weinstock R, in obese and lean adults: a randomized, controlled,
overweight men and women. N Engl J Med. 1991; Buckenmeyer P, Berkowitz RI, Steen SN. Exercise in dose-escalation trial. JAMA. 1999;282(16):1568–1575.
325(7):461–466. the treatment of obesity: effects of four in- 329. Ravussin E, Smith SR, Mitchell JA, Shringarpure R,
300. Svendsen OL, Hassager C, Christiansen C. Six months’ terventions on body composition, resting energy Shan K, Maier H, Koda JE, Weyer C. Enhanced
follow-up on exercise added to a short-term diet in expenditure, appetite, and mood. J Consult Clin weight loss with pramlintide/metreleptin: an in-
overweight postmenopausal women—effects on body Psychol. 1997;65(2):269–277. tegrated neurohormonal approach to obesity
composition, resting metabolic rate, cardiovascular risk 313. Sherwood NE, Jeffery RW, Welsh EM, Vanwormer J, pharmacotherapy. Obesity (Silver Spring). 2009;
factors and bone. Int J Obes Relat Metab Disord. 1994; Hotop AM. The drop it at last study: six-month 17(9):1736–1743.
18(10):692–698. results of a phone-based weight loss trial. Am J 330. Bray GA, Ryan DH. Medical therapy for the patient
301. Pritchard JE, Nowson CA, Wark JD. A worksite Health Promot. 2010;24(6):378–383. with obesity. Circulation. 2012;125(13):1695–1703.
program for overweight middle-aged men achieves 314. Perri MG, Limacher MC, Durning PE, Janicke DM, 331. Apovian CM, Aronne LJ, Bessesen DH, McDonnell
lesser weight loss with exercise than with dietary Lutes LD, Bobroff LB, Dale MS, Daniels MJ, Radcliff ME, Murad MH, Pagotto U, Ryan DH, Still CD;
change. J Am Diet Assoc. 1997;97(1):37–42. TA, Martin AD. Extended-care programs for weight Endocrine Society. Pharmacological management of
302. Irwin ML, Yasui Y, Ulrich CM, Bowen D, Rudolph RE, management in rural communities: the treatment obesity: an Endocrine Society clinical practice
Schwartz RS, Yukawa M, Aiello E, Potter JD, of obesity in underserved rural settings (TOURS) guideline. J Clin Endocrinol Metab. 2015;100(2):
McTiernan A. Effect of exercise on total and intra- randomized trial. Arch Intern Med. 2008;168(21): 342–362.
abdominal body fat in postmenopausal women: 2347–2354. 332. Sjöström L, Rissanen A, Andersen T, Boldrin M,
a randomized controlled trial. JAMA. 2003;289(3): 315. Donnelly JE, Smith BK, Dunn L, Mayo MM, Jacobsen Golay A, Koppeschaar HP, Krempf M; European
323–330. DJ, Stewart EE, Gibson C, Sullivan DK. Comparison Multicentre Orlistat Study Group. Randomised
303. Donnelly JE, Hill JO, Jacobsen DJ, Potteiger J, Sullivan of a phone vs clinic approach to achieve 10% weight placebo-controlled trial of orlistat for weight loss
DK, Johnson SL, Hellan K, Hise M, Fennessey PV, loss. Int J Obes. 2007;31(8):1270–1276.
and prevention of weight regain in obese patients.
Sonko B, Sharp T, Jakicic JM, Blair SN, Tran ZV, Mayo 316. Donnelly JE, Goetz J, Gibson C, Sullivan DK, Lee R,
Lancet. 1998;352(9123):167–172.
M, Gibson C, Washburn RA. Effects of a 16-month Smith BK, Lambourne K, Mayo MS, Hunt S, Lee JH,
333. Torgerson JS, Hauptman J, Boldrin MN, Sjöström L.
randomized controlled exercise trial on body Honas JJ, Washburn RA. Equivalent weight loss for
XENical in the prevention of diabetes in obese
weight and composition in young, overweight men weight management programs delivered by phone
subjects (XENDOS) study: a randomized study of
and women: the Midwest Exercise Trial. Arch Intern and clinic. Obesity (Silver Spring). 2013;21(10):
orlistat as an adjunct to lifestyle changes for the
Med. 2003;163(11):1343–1350. 1951–1959.
prevention of type 2 diabetes in obese patients.
304. Slentz CA, Duscha BD, Johnson JL, Ketchum K, 317. Rock CL, Flatt SW, Sherwood NE, Karanja N, Pakiz B,
Diabetes Care. 2004;27(1):155–161.
Aiken LB, Samsa GP, Houmard JA, Bales CW, Kraus Thomson CA. Effect of a free prepared meal and
334. Dong Z, Xu L, Lv Y, Zheng Q, Li L. Comparative
WE. Effects of the amount of exercise on body incentivized weight loss program on weight loss
efficacy of five long-term weight loss drugs:
weight, body composition, and measures of central and weight loss maintenance in obese and over-
quantitative information for medication guidelines.
obesity: STRRIDE—a randomized controlled study. weight women: a randomized controlled trial.
Obes Rev. 2017;18(12):1377–1383.
Arch Intern Med. 2004;164(1):31–39. JAMA. 2010;304(16):1803–1810.
335. Rucker D, Padwal R, Li SK, Curioni C, Lau DCW. Long
305. Bray GA, Frühbeck G, Ryan DH, Wilding JP. Man- 318. Tate DF, Wing RR, Winett RA. Using Internet
agement of obesity. Lancet. 2016;387(10031): term pharmacotherapy for obesity and overweight:
technology to deliver a behavioral weight loss
1947–1956. program. JAMA. 2001;285(9):1172–1177. updated meta-analysis. BMJ. 2007;335(7631):
306. Jensen MD, Ryan DH, Donato KA, Apovian CM, Ard 319. Harvey-Berino J, West D, Krukowski R, Prewitt E, 1194–1199.
JD, Comuzzie AG, Hu FB, Hubbard VS, Jakicic JM, VanBiervliet A, Ashikaga T, Skelly J. Internet de- 336. Padwal R, Kezouh A, Levine M, Etminan M. Long-
Kushner RF; American College of Cardiology/ livered behavioral obesity treatment. Prev Med. term persistence with orlistat and sibutramine in
American Heart Association Task Force on Prac- 2010;51(2):123–128. a population-based cohort. Int J Obes. 2007;31(10):
tice Guidelines, Obesity Expert Panel, 2013. Exec- 320. Knowler WC, Hamman RF, Edelstein SL, Barrett- 1567–1570.
utive summary: guidelines (2013) for the Connor E, Ehrmann DA, Walker EA, Fowler SE, 337. Halford JCG, Harrold JA, Boyland EJ, Lawton CL,
management of overweight and obesity in adults: Nathan DM, Kahn SE; Diabetes Prevention Program Blundell JE. Serotonergic drugs: effects on appetite
a report of the American College of Cardiology/ Research Group. Prevention of type 2 diabetes with expression and use for the treatment of obesity.
American Heart Association Task Force on Practice troglitazone in the Diabetes Prevention Program. Drugs. 2007;67(1):27–55.
Guidelines and the Obesity Society published by Diabetes. 2005;54(4):1150–1156. 338. Smith SR, Weissman NJ, Anderson CM, Sanchez M,
the Obesity Society and American College of 321. Svetkey LP, Batch BC, Lin P-H, Intille SS, Corsino L, Chuang E, Stubbe S, Bays H, Shanahan WR; Be-
Cardiology/American Heart Association Task Force Tyson CC, Bosworth HB, Grambow SC, Voils C, Loria havioral Modification and Lorcaserin for Over-
on Practice Guidelines. Based on a systematic re- C, Gallis JA, Schwager J, Bennett GG. Cell phone weight and Obesity Management (BLOOM) Study
view from the The Obesity Expert Panel, 2013. intervention for you (CITY): a randomized, con- Group. Multicenter, placebo-controlled trial of
Obesity (Silver Spring). 2014;22(Suppl 2):S5–S39. trolled trial of behavioral weight loss intervention lorcaserin for weight management. N Engl J Med.
307. Wadden TA, Webb VL, Moran CH, Bailer BA. for young adults using mobile technology. Obesity 2010;363(3):245–256.
Lifestyle modification for obesity: new de- (Silver Spring). 2015;23(11):2133–2141. 339. Fidler MC, Sanchez M, Raether B, Weissman NJ,
velopments in diet, physical activity, and behavior 322. Bray GA, Greenway FL. Pharmacological treatment Smith SR, Shanahan WR, Anderson CM; BLOSSOM
therapy. Circulation. 2012;125(9):1157–1170. of the overweight patient. Pharmacol Rev. 2007; Clinical Trial Group. A one-year randomized trial of
308. LeBlanc ES, O’Connor E, Whitlock EP, Patnode CD, 59(2):151–184. lorcaserin for weight loss in obese and overweight
Kapka T. Effectiveness of primary care-relevant 323. Nathanson MH. The central action of betaami- adults: the BLOSSOM trial. J Clin Endocrinol Metab.
treatments for obesity in adults: a systematic evi- nopropylbenzene (benzedrine): clinical observa- 2011;96(10):3067–3077.
dence review for the U.S. Preventive Services Task tions. J Am Med Assoc. 1937;108(7):528–531. 340. O’Neil PM, Smith SR, Weissman NJ, Fidler MC,
Force. Ann Intern Med. 2011;155(7):434–447. 324. Kasant O. The Amphetamines: Toxicity and Addic- Sanchez M, Zhang J, Raether B, Anderson CM,
309. Perri MG, Shapiro RM, Ludwig WW, Twentyman tion. Ontario, Canada: Alcoholism and Drug Ad- Shanahan WR. Randomized placebo-controlled
CT, McAdoo WG. Maintenance strategies for the diction Research Foundation of Ontario; 1966. clinical trial of lorcaserin for weight loss in type 2
treatment of obesity: an evaluation of relapse 325. Fishman AP. Aminorex to fen/phen: an epidemic diabetes mellitus: the BLOOM-DM study. Obesity
prevention training and posttreatment contact by foretold. Circulation. 1999;99(1):156–161. (Silver Spring). 2012;20(7):1426–1436.

doi: 10.1210/er.2017-00253 51

Downloaded from
by guest
on 17 March 2018

341. Food and Drug Administration. FDA briefing in- Erickson JS, Maier HN, Kim DD, Dunayevich E. 366. George M, Rajaram M, Shanmugam E. New and
formation: meeting of the endocrinologic and metabolic Weight loss with naltrexone SR/bupropion SR emerging drug molecules against obesity.
drugs advisory committee. Available at: http://www.fda. combination therapy as an adjunct to behavior J Cardiovasc Pharmacol Ther. 2014;19(1):65–76.
gov/downloads/AdvisoryCommittees/Committees- modification: the COR-BMOD trial. Obesity (Silver 367. Ponce J, DeMaria EJ, Nguyen NT, Hutter M, Sudan R,
MeetingMaterials/Drugs/Endocrinologicand- Spring). 2011;19(1):110–120. Morton JM. American Society for Metabolic and
MetabolicDrugsAdvisoryCommittee/UCM303198.pdf. 353. Apovian CM, Aronne L, Rubino D, Still C, Wyatt H, Bariatric Surgery estimation of bariatric surgery
Accessed 8 January 2018. Burns C, Kim D, Dunayevich E; COR-II Study Group. procedures in 2015 and surgeon workforce in the
342. Astrup A, Carraro R, Finer N, Harper A, Kunesova M, A randomized, phase 3 trial of naltrexone SR/ United States. Surg Obes Relat Dis. 2016;12(9):
Lean ME, Niskanen L, Rasmussen MF, Rissanen A, bupropion SR on weight and obesity-related risk 1637–1639.
Rössner S, Savolainen MJ, Van Gaal L; NN8022-1807 factors (COR-II). Obesity (Silver Spring). 2013;21(5): 368. Nielsen S, Svane MS, Bojsen-Møller KN, Madsbad S.
Investigators. Safety, tolerability and sustained 935–943. Effects of bariatric surgery on weight loss and
weight loss over 2 years with the once-daily human 354. Hollander P, Gupta AK, Plodkowski R, Greenway F, quality of life. Anaplastology. 2014;3:136.
GLP-1 analog, liraglutide. Int J Obes. 2012;36(6): Bays H, Burns C, Klassen P, Fujioka K; COR-Diabetes 369. Gagner M, Rogula T. Laparoscopic reoperative
843–854. Study Group. Effects of naltrexone sustained- sleeve gastrectomy for poor weight loss after bil-
343. Pi-Sunyer X, Astrup A, Fujioka K, Greenway F, release/bupropion sustained-release combination iopancreatic diversion with duodenal switch. Obes
Halpern A, Krempf M, Lau DCW, le Roux CW, therapy on body weight and glycemic parameters in Surg. 2003;13(4):649–654.
Violante Ortiz R, Jensen CB, Wilding JPH; SCALE overweight and obese patients with type 2 diabetes. 370. Zundel N. Laparoscopic sleeve gastrectomy: Tech-
Obesity and Prediabetes NN8022-1839 Study Diabetes Care. 2013;36(12):4022–4029. nique and outcomes. In: Nguyen NT, Blackstone RP,
Group. A randomized, controlled trial of 3.0 mg of 355. Ali KF, Shukla AP, Aronne LJ. Bupropion-SR plus Morton JM, Ponce J, Rosenthal R, eds. American
liraglutide in weight management. N Engl J Med. naltrexone-SR for the treatment of mild-to- Society for Metabolic and Bariatric Surgery: Textbook
2015;373(1):11–22. moderate obesity. Expert Rev Clin Pharmacol. of Bariatric Surgery. Vol. 1. New York, NY: Springer;
344. Wadden TA, Hollander P, Klein S, Niswender K, 2016;9(1):27–34. 2015:205–210.
Woo V, Hale PM, Aronne L; NN8022-1923 In- 356. Khera R, Murad MH, Chandar AK, Dulai PS, Wang Z, 371. Nguyen NT, Goldman C, Rosenquist CJ, Arango A,
vestigators. Weight maintenance and additional Prokop LJ, Loomba R, Camilleri M, Singh S. Asso- Cole CJ, Lee SJ, Wolfe BM. Laparoscopic versus open
weight loss with liraglutide after low-calorie-diet- ciation of pharmacological treatments for obesity gastric bypass: a randomized study of outcomes,
induced weight loss: the SCALE Maintenance with weight loss and adverse events: a systematic quality of life, and costs. Ann Surg. 2001;234(3):
randomized study. Int J Obes. 2013;37(11): review and meta-analysis. JAMA. 2016;315(22): 279–289.
1443–1451. 2424–2434. 372. DeMaria EJ, Pate V, Warthen M, Winegar DA.
345. Marso SP, Daniels GH, Brown-Frandsen K, Kristensen 357. U.S. Food and Drug Administration. New drug Baseline data from American Society for Metabolic
P, Mann JF, Nauck MA, Nissen SE, Pocock S, Poulter applications 22580: Vi-0521 qnexa (phentermine/ and Bariatric Surgery-designated bariatric surgery
NR, Ravn LS, Steinberg WM, Stockner M, Zinman B,
topiramate). Available at: centers of excellence using the bariatric outcomes
Bergenstal RM, Buse JB; LEADER Steering Committee;
downloads/AdvisoryCommittees/CommitteesMeeting- longitudinal database. Surg Obes Relat Dis. 2010;6(4):
LEADER Trial Investigators. Liraglutide and cardio-
Materials/Drugs/EndocrinologicandMetabolicDrugs- 347–355.
vascular outcomes in type 2 diabetes. N Engl J Med.
AdvisoryCommittee/UCM292315.pdf. Accessed 15 373. O’Brien PE, Dixon JB. Lap-band: outcomes and re-
June 2017. sults. J Laparoendosc Adv Surg Tech A. 2003;13(4):
346. Marso SP, Bain SC, Consoli A, Eliaschewitz FG, Jódar
358. Kang JG, Park CY, Kang JH, Park YW, Park SW. 265–270.
E, Leiter LA, Lingvay I, Rosenstock J, Seufert J, Warren
Randomized controlled trial to investigate the ef- 374. Shikora SA, Wolfe BM, Apovian CM, Anvari M,
ML, Woo V, Hansen O, Holst AG, Pettersson J,
fects of a newly developed formulation of phen- Sarwer DB, Gibbons RD, Ikramuddin S, Miller CJ,
Vilsbøll T; SUSTAIN-6 Investigators. Semaglutide
termine diffuse-controlled release for obesity. Knudson MB, Tweden KS, Sarr MG, Billington CJ.
and cardiovascular outcomes in patients with
Diabetes Obes Metab. 2010;12(10):876–882. Sustained weight loss with vagal nerve blockade but
type 2 diabetes. N Engl J Med. 2016;375(19):
359. Addy C, Jumes P, Rosko K, Li S, Li H, Maes A, not with sham: 18-month results of the recharge
Johnson-Levonas AO, Chodakewitz J, Stoch SA, trial. J Obes. 2015;2015:365604.
347. Gadde KM, Allison DB, Ryan DH, Peterson CA,
Wagner JA. Pharmacokinetics, safety, and tolera- 375. Shikora SA, Bergenstal R, Bessler M, Brody F, Foster
Troupin B, Schwiers ML, Day WW. Effects of low-
dose, controlled-release, phentermine plus top- bility of phentermine in healthy participants re- G, Frank A, Gold M, Klein S, Kushner R, Sarwer DB.
iramate combination on weight and associated ceiving taranabant, a novel cannabinoid-1 receptor Implantable gastric stimulation for the treatment of
comorbidities in overweight and obese adults (CB1R) inverse agonist. J Clin Pharmacol. 2009; clinically severe obesity: results of the SHAPE trial.
(CONQUER): a randomised, placebo-controlled, 49(10):1228–1238. Surg Obes Relat Dis. 2009;5(1):31–37.
phase 3 trial. Lancet. 2011;377(9774):1341–1352. 360. Kim KK, Cho H-J, Kang H-C, Youn B-B, Lee K-R. 376. Saber AA, Shoar S, Almadani MW, Zundel N,
348. Allison DB, Gadde KM, Garvey WT, Peterson CA, Effects on weight reduction and safety of short- Alkuwari MJ, Bashah MM, Rosenthal RJ. Efficacy of
Schwiers ML, Najarian T, Tam PY, Troupin B, Day term phentermine administration in Korean obese first-time intragastric balloon in weight loss: a sys-
WW. Controlled-release phentermine/topiramate people. Yonsei Med J. 2006;47(5):614–625. tematic review and meta-analysis of randomized
in severely obese adults: a randomized controlled 361. Blanck HM, Khan LK, Serdula MK. Prescription controlled trials. Obes Surg. 2017;27(2):277–287.
trial (EQUIP). Obesity (Silver Spring). 2012;20(2): weight loss pill use among Americans: patterns 377. Mion F, Ibrahim M, Marjoux S, Ponchon T,
330–342. of pill use and lessons learned from the fen- Dugardeyn S, Roman S, Deviere J. Swallowable
349. Garvey WT, Ryan DH, Look M, Gadde KM, Allison phen market withdrawal. Prev Med. 2004;39(6): Obalon® gastric balloons as an aid for weight loss:
DB, Peterson CA, Schwiers M, Day WW, Bowden 1243–1248. a pilot feasibility study. Obes Surg. 2013;23(5):
CH. Two-year sustained weight loss and metabolic 362. Hampp C, Kang EM, Borders-Hemphill V. Use of 730–733.
benefits with controlled-release phentermine/ prescription antiobesity drugs in the United States. 378. Klein S, Kinney J, Jeejeebhoy K, Alpers D, Hellerstein
topiramate in obese and overweight adults (SE- Pharmacotherapy. 2013;33(12):1299–1307. M, Murray M, Twomey P. Nutrition support in
QUEL): a randomized, placebo-controlled, phase 3 363. Blanck HM, Serdula MK, Gillespie C, Galuska DA, clinical practice: review of published data and
extension study. Am J Clin Nutr. 2012;95(2):297–308. Sharpe PA, Conway JM, Khan LK, Ainsworth BE. Use recommendations for future research directions.
350. Hendricks EJ, Rothman RB, Greenway FL. How of nonprescription dietary supplements for weight Clin Nutr. 1997;16(4):193–218.
physician obesity specialists use drugs to treat loss is common among Americans. J Am Diet Assoc. 379. Gersin KS, Keller JE, Stefanidis D, Simms CS, Abraham
obesity. Obesity (Silver Spring). 2009;17(9):1730–1735. 2007;107(3):441–447. DD, Deal SE, Kuwada TS, Heniford BT. Duodenal-
351. Greenway FL, Fujioka K, Plodkowski RA, Mudaliar S, 364. Pillitteri JL, Shiffman S, Rohay JM, Harkins AM, jejunal bypass sleeve: a totally endoscopic device for
Guttadauria M, Erickson J, Kim DD, Dunayevich E; Burton SL, Wadden TA. Use of dietary supplements the treatment of morbid obesity. Surg Innov. 2007;
COR-I Study Group. Effect of naltrexone plus for weight loss in the United States: results of 14(4):275–278.
bupropion on weight loss in overweight and obese a national survey. Obesity (Silver Spring). 2008;16(4): 380. Rajagopalan H, Cherrington AD, Thompson CC,
adults (COR-I): a multicentre, randomised, double- 790–796. Kaplan LM, Rubino F, Mingrone G, Becerra P,
blind, placebo-controlled, phase 3 trial. Lancet. 2010; 365. National Institutes of Health Office of Dietary Rodriguez P, Vignolo P, Caplan J, Rodriguez L, Galvao
376(9741):595–605. Supplements. Dietary supplements for weight Neto MP. Endoscopic duodenal mucosal resurfac-
352. Wadden TA, Foreyt JP, Foster GD, Hill JO, Klein S, loss. ing for the treatment of type 2 diabetes: 6-month
O’Neil PM, Perri MG, Pi-Sunyer FX, Rock CL, HealthProfessional/. Accessed 31 January 2016. interim analysis from the first-in-human proof-of-

52 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018

concept study. Diabetes Care. 2016;39(12): effective therapy for adult-onset diabetes mellitus. Lam B, Larder R, Yeo GS, O’Rahilly S, Vavouri T,
2254–2261. Ann Surg. 1995;222(3):339–350. Whitelaw E, Penninger JM, Jenuwein T, Cheung C-L,
381. Klein S, Fontana L, Young VL, Coggan AR, Kilo C, 395. Mingrone G, Panunzi S, De Gaetano A, Guidone C, Ferguson-Smith AC, Coll AP, Körner A, Pospisilik JA.
Patterson BW, Mohammed BS. Absence of an effect Iaconelli A, Nanni G, Castagneto M, Bornstein S, Trim28 haploinsufficiency triggers bi-stable epige-
of liposuction on insulin action and risk factors for Rubino F. Bariatric-metabolic surgery versus con- netic obesity. Cell. 2016;164(3):353–364.
coronary heart disease. N Engl J Med. 2004;350(25): ventional medical treatment in obese patients with 410. Zinman B, Wanner C, Lachin JM, Fitchett D,
2549–2557. type 2 diabetes: 5 year follow-up of an open-label, Bluhmki E, Hantel S, Mattheus M, Devins T,
382. Gastrointestinal surgery for severe obesity con- single-centre, randomised controlled trial. Lancet. Johansen OE, Woerle HJ, Broedl UC, Inzucchi SE;
sensus statement. Nutrition Today 1991;26: 2015;386(9997):964–973. EMPA-REG OUTCOME Investigators. Empagli-
32–35. 396. Buchwald H, Estok R, Fahrbach K, Banel D, Jensen flozin, cardiovascular outcomes, and mortality in
383. Courcoulas AP, Christian NJ, Belle SH, Berk PD, MD, Pories WJ, Bantle JP, Sledge I. Weight and type 2 diabetes. N Engl J Med. 2015;373(22):
Flum DR, Garcia L, Horlick M, Kalarchian MA, type 2 diabetes after bariatric surgery: systematic 2117–2128.
King WC, Mitchell JE, Patterson EJ, Pender JR, review and meta-analysis. Am J Med. 2009;122(3): 411. Huang T, Zheng Y, Hruby A, Williamson DA, Bray
Pomp A, Pories WJ, Thirlby RC, Yanovski SZ, 248–256. GA, Shen Y, Sacks FM, Qi L. Dietary protein modifies
Wolfe BM; Longitudinal Assessment of Bariatric 397. Bray GA. A Guide to Obesity and the Metabolic the effect of the MC4R genotype on 2-year changes
Surgery (LABS) Consortium. Weight change and Syndrome: Origins and Treatment. New York, NY: in appetite and food craving: the POUNDS Lost
health outcomes at 3 years after bariatric surgery CRC Press; 2011:265. Trial. J Nutr. 2017;147(3):439–444.
among individuals with severe obesity. JAMA. 398. Laferrère B, Heshka S, Wang K, Khan Y, McGinty J, 412. Zeevi D, Korem T, Zmora N, Israeli D, Rothschild D,
2013;310(22):2416–2425. Teixeira J, Hart AB, Olivan B. Incretin levels and effect Weinberger A, Ben-Yacov O, Lador D, Avnit-Sagi T,
384. Beamish AJ, Olbers T, Kelly AS, Inge TH. Cardio- are markedly enhanced 1 month after Roux-en-Y Lotan-Pompan M, Suez J, Mahdi JA, Matot E, Malka
vascular effects of bariatric surgery. Nature Rev gastric bypass surgery in obese patients with type 2 G, Kosower N, Rein M, Zilberman-Schapira G,
Cardiol 2016;13(12):730–743. diabetes. Diabetes Care. 2007;30(7):1709–1716.
Dohnalová L, Pevsner-Fischer M, Bikovsky R, Halpern
385. Flum DR, Salem L, Elrod JA, Dellinger EP, Cheadle A, 399. Michalsky M, Reichard K, Inge T, Pratt J, Lenders
Z, Elinav E, Segal E. Personalized nutrition by pre-
Chan L. Early mortality among Medicare benefi- C; American Society for Metabolic and Bariatric
diction of glycemic responses. Cell. 2015;163(5):
ciaries undergoing bariatric surgical procedures. Surgery. ASMBS pediatric committee best
JAMA. 2005;294(15):1903–1908. practice guidelines. Surg Obes Relat Dis. 2012;
413. Müller TD, Finan B, Clemmensen C, DiMarchi RD,
386. Smith MD, Patterson E, Wahed AS, Belle SH, Bessler 8(1):1–7.
Tschöp MH. The new biology and pharmacology of
M, Courcoulas AP, Flum D, Halpin V, Mitchell JE, 400. Sieber P, Gass M, Kern B, Peters T, Slawik M, Peterli R.
glucagon. Physiol Rev. 2017;97(2):721–766.
Pomp A, Pories WJ, Wolfe B. Relationship between Five-year results of laparoscopic sleeve gastrectomy.
414. Tschöp MH, Finan B, Clemmensen C, Gelfanov V,
surgeon volume and adverse outcomes after RYGB Surg Obes Relat Dis. 2014;10(2):243–249.
in Longitudinal Assessment of Bariatric Surgery Perez-Tilve D, Müller TD, DiMarchi RD. Unim-
401. Schauer PR, Bhatt DL, Kirwan JP, Wolski K, Brethauer
(LABS) study. Surg Obes Relat Dis. 2010;6(2): SA, Navaneethan SD, Aminian A, Pothier CE, Kim olecular polypharmacy for treatment of diabetes
118–125. ESH, Nissen SE, Kashyap SR; STAMPEDE In- and obesity. Cell Metab. 2016;24(1):51–62.
387. Puzziferri N, Roshek TB III, Mayo HG, Gallagher R, vestigators. Bariatric surgery versus intensive med- 415. Mozaffarian D. Dietary and policy priorities for
Belle SH, Livingston EH. Long-term follow-up after ical therapy for diabetes—3-year outcomes. N Engl J cardiovascular disease, diabetes, and obesity:
bariatric surgery: a systematic review. JAMA. 2014; Med. 2014;370(21):2002–2013. a comprehensive review. Circulation. 2016;133(2):
312(9):934–942. 402. Inge TH, Courcoulas AP, Jenkins TM, Michalsky MP, 187–225.
388. Wolfe BM, Kvach E, Eckel RH. Treatment of obesity: Helmrath MA, Brandt ML, Harmon CM, Zeller MH, 416. Janesick AS, Blumberg B. Obesogens: an emerging
weight loss and bariatric surgery. Circ Res. 2016; Chen MK, Xanthakos SA, Horlick M, Buncher CR; threat to public health. Am J Obstet Gynecol. 2016;
118(11):1844–1855. Teen-LABS Consortium. Weight loss and health 214(5):559–565.
389. Gletsu-Miller N, Wright BN. Mineral malnutrition status 3 years after bariatric surgery in adolescents.
following bariatric surgery. Adv Nutr. 2013;4(5): N Engl J Med. 2016;374(2):113–123.
506–517. 403. Inge TH, Jenkins TM, Xanthakos SA, Dixon JB, Acknowledgments
390. Mechanick JI, Youdim A, Jones DB, Garvey WT, Daniels SR, Zeller MH, Helmrath MA. Long-term We acknowledge science writer Eric A. Vohr for his con-
Hurley DL, McMahon MM, Heinberg LJ, Kushner R, outcomes of bariatric surgery in adolescents with tribution to this scientific statement.
Adams TD, Shikora S, Dixon JB, Brethauer S. Clinical severe obesity (FABS-5+): a prospective follow-up Financial Support: Work on this paper was supported
practice guidelines for the perioperative nutritional, analysis. Lancet Diabetes Endocrinol. 2017;5(3): in part by Bill and Melinda Gates Foundation Grant
metabolic, and nonsurgical support of the bariatric 165–173. OPP1070441 to A.A., National Institutes of Health Grants
surgery patient—2013 update: cosponsored by 404. Olbers T, Beamish AJ, Gronowitz E, Flodmark C-E, DK40484 and DK45343 to M.D.J., and by National Institutes
American Association of Clinical Endocrinologists, Dahlgren J, Bruze G, Ekbom K, Friberg P, Göthberg of Health Grant DK046200 to F.B.H.
The Obesity Society, and American Society for G, Järvholm K, Karlsson J, Mårild S, Neovius M, Correspondence and Reprint Requests: George A. Bray,
Metabolic and Bariatric Surgery. Obesity (Silver Peltonen M, Marcus C. Laparoscopic Roux-en-Y MD, Pennington Biomedical Research Center, Louisiana State
Spring). 2013;21(Suppl 1):S1–S27. gastric bypass in adolescents with severe obesity University, Baton Rouge, Louisiana 70808. E-mail: george.
391. Courcoulas AP, Christian NJ, O’Rourke RW, Dakin G, (AMOS): a prospective, 5-year, Swedish nationwide
Patchen Dellinger E, Flum DR, Melissa Kalarchian study. Lancet Diabetes Endocrinol. 2017;5(3): Disclaimer Statement: The Endocrine Society develops
PD, Mitchell JE, Patterson E, Pomp A, Pories WJ, 174–183. scientific statements to be of assistance to clinicians by
Spaniolas K, Steffen K, Wolfe BM, Belle SH. Pre- 405. Pratt JSA, Lenders CM, Dionne EA, Hoppin AG, Hsu providing guidance and recommendations for particular
operative factors and 3-year weight change in the GLK, Inge TH, Lawlor DF, Marino MF, Meyers AF, areas of practice. One should not consider this scientific
Longitudinal Assessment of Bariatric Surgery (LABS) Rosenblum JL, Sanchez VM. Best practice updates statement inclusive of all proper approaches or methods or
consortium. Surg Obes Relat Dis. 2015;11(5): for pediatric/adolescent weight loss surgery. Obesity exclusive of others. It cannot guarantee any specific outcome,
1109–1118. (Silver Spring). 2009;17(5):901–910. nor does it establish a standard of care. It is not intended to
392. Mitchell JE, Christian NJ, Flum DR, Pomp A, Pories 406. Pickett K, Wilkinson RP. The Spirit Level: Why dictate the treatment of a particular patient. Health care
WJ, Wolfe BM, Courcoulas AP, Belle SH. Post- Greater Equality Makes Societies Stronger. New York, providers must make treatment decisions based on their best
operative behavioral variables and weight change 3 NY: Bloomsbury Press; 2011. judgment and each patient’s individual circumstances, needs,
years after bariatric surgery. JAMA Surg. 2016;151(8): 407. Christakis NA, Fowler JH. The spread of obesity in and preferences. The Endocrine Society makes no warranty,
752–757. a large social network over 32 years. N Engl J Med. express or implied, regarding this scientific statement and
393. Sjöström L. Review of the key results from the 2007;357(4):370–379. specifically excludes any warranties of merchantability and
Swedish Obese Subjects (SOS) trial a prospective 408. Anderson SE, Andridge R, Whitaker RC. Bedtime in fitness for a particular use or purpose. The Endocrine Society
controlled intervention study of bariatric surgery. preschool-aged children and risk for adolescent shall not be liable for direct, indirect, special, incidental, or
J Intern Med. 2013;273(3):219–234. obesity. J Pediatr. 2016;176:17–22. consequential damages related to the use of the information
394. Pories WJ, Swanson MS, MacDonald KG, Long SB, 409. Dalgaard K, Landgraf K, Heyne S, Lempradl A, contained herein.
Morris PG, Brown BM, Barakat HA, deRamon RA, Longinotto J, Gossens K, Ruf M, Orthofer M, Disclosure Summary: G.B. is on the Herbalife Nutrition
Israel G, Dolezal JM, Dohm L. Who would have Strogantsev R, Selvaraj M, Lu TT, Casas E, Teperino R, Advisory Council, is a consultant to Medifast, and is a con-
thought it? An operation proves to be the most Surani MA, Zvetkova I, Rimmington D, Tung YCL, sultant to Novo Nordisk. M.D.J. serves as a consultant to

doi: 10.1210/er.2017-00253 53

Downloaded from
by guest
on 17 March 2018

Novo Nordisk, Merck, and Jannsen, serves on the advisory J.M.J. reports personal fees from Weight Watchers In- cardiovascular disease; ECE, early care and education; ER,
board of Weight Watchers International, and serves on the ternational for serving on the Scientific Advisory Board. D.H.R. extended release; FDA, Food and Drug Administration;
board of directors of the Partnership for a Healthier America. serves as consultant to Novo Nordisk, Orexigen, Janssen, HbA1c, hemoglobin A1c; HDL, high-density lipoprotein; HR,
R.F.K. serves on the advisory board to Novo Nordisk, Weight Merck, Astra Xenica, and KVK Tech, serves as speaker for hazard ratio; LAGB, laparoscopic adjustable gastric banding;
Watchers, Retrofit, and Zafgen. S.R.D. serves as a consultant Novo Nordisk, Orexigen, Eisai, and Kwang Dong, and has an LDL, low-density lipoprotein; MHO, metabolically healthy
for Sanofi and is on the Data Monitoring Committee for equity position in Scientific Intake and Gila Therapeutics. The obesity; NAFLD, nonalcoholic fatty liver disease; NASH,
Novo Nordisk. T.A.W. serves on advisory boards for Novo remaining authors have nothing to disclose. nonalcoholic steatohepatitis; OSA, obstructive sleep apnea;
Nordisk, Nutrisystem, and Weight Watchers, and has re- PHEN/TPM ER, phentermine/topiramate extended release;
ceived grant support on behalf of the University of Penn- RCT, randomized controlled trial; RYGB, Roux-en-Y gastric
sylvania from Eisai Pharmaceutical, Novo Nordisk, and Abbreviations bypass; SG, sleeve gastrectomy; SR, sustained release; T2DM,
Weight Watchers. F.B.H. receives research support from BMI, body mass index; BP, blood pressure; CDC, Centers for type 2 diabetes mellitus; VLCD, very low–calorie diet; WHR,
Metagenics and the California Walnut Commission. Disease Control and Prevention; CI, confidence interval; CVD, waist-to-hip ratio.

54 Bray et al Management of Obesity Endocrine Reviews, April 2018, 39(2):1–54

Downloaded from
by guest
on 17 March 2018