OBES SURG (2014) 24:487–519

DOI 10.1007/s11695-014-1214-1

OTHER

Bariatric Surgery in Class I Obesity

A Position Statement from the International Federation for the Surgery of Obesity
and Metabolic Disorders (IFSO)

Luca Busetto & John Dixon & Maurizio De Luca &

Scott Shikora & Walter Pories & Luigi Angrisani

Published online: 18 March 2014

# Springer Science+Business Media New York 2014

Abstract Class I obesity conveys an increased risk of comor- comorbidity burden similar to, or greater than, patients with
bidities, impairs physical and mental health-related quality of more severe obesity. Therefore, the denial of bariatric surgery
life, and it is associated to an increased psychosocial burden, to a patient with class I obesity suffering from a significant
particularly in women. The need for effective and safe thera- obesity-related health burden and not achieving weight con-
pies for class I obesity is great and not yet met by nonsurgical trol with nonsurgical therapy simply on the basis of the BMI
approaches. Eligibility to bariatric surgery has been largely level does not appear to be clinically justified. A clinical
based on body mass index (BMI) cut points and limited to decision should be based on a more comprehensive evaluation
patients with more severe obesity levels. However, obese of the patient’s current global health and on a more reliable
patients belonging to the same BMI class may have very prediction of future morbidity and mortality. After a careful
different levels of health, risk, and impact of obesity on quality review of available data about safety and efficacy of bariatric
of life. Individual patients in class I obesity may have a surgery in patients with class I obesity, this panel reached a
consensus on ten clinical recommendations.

The Position Statement has been written by a working group formed by Keywords Class I obesity . BMI . Eligibility
members of the International Federation for the Surgery of Obesity and
Metabolic Disorders (IFSO)-The Position Statement has been discussed
and approved by the Executive Board of the International Federation for
the Surgery of Obesity and Metabolic Disorders (IFSO). Executive Summary and Final Recommendations
L. Busetto (*)
Department of Medicine, University of Padua, Padua, Italy Class I obesity [body mass index (BMI) 30–35 kg/m2] con-
e-mail: luca.busetto@unipd.it veys an increased risk of comorbidities, impairs physical and
mental health-related quality of life, and it is associated to an
J. Dixon
Clinical Obesity Research, Baker IDI Heart & Diabetes Institute, increased psycho-social burden, particularly in women. The
Melbourne, Australia need for effective and safe therapies for class I obesity is great
and not yet met by nonsurgical approaches.
M. De Luca
Department of Surgery, San Bortolo Hospital, Vicenza, Italy
Eligibility to bariatric surgery has been largely based on
BMI cut-points and limited to patients with more severe
S. Shikora obesity levels (BMI > 40 kg/m2 or BMI 35–40 kg/m2 with
Center for Metabolic and Bariatric Surgery, obesity-related comorbidities). However, obese patients be-
Brigham and Women’s Hospital, Boston, MA, USA
longing to the same BMI class may have very different levels
W. Pories of health, risk, and impact of obesity on quality of life.
Brody School of Medicine, East Carolina University, Greenville, NC, Individual patients in class I obesity may have a comorbidity
USA burden similar to, or greater than, patients with more severe
obesity. Therefore, the denial of bariatric surgery to a patient
L. Angrisani
General And Endoscopic Surgery Unit, San Giovanni Bosco with class I obesity suffering from a significant obesity-related
Hospital, Naples, Italy health burden and not achieving weight control with
488 OBES SURG (2014) 24:487–519

nonsurgical therapy simply on the basis of the BMI level does (9) National and regional health providers need to consider
not appear to be clinically justified. A clinical decision should the current evidences favoring the application of bariatric
be based on a more comprehensive evaluation of the patient’s surgery in class I obesity in the context of local health
current global health and on a more reliable prediction of resources and deliver services that are locally
future morbidity and mortality. appropriate.
After a careful review of available data about safety and (10) Published literature on bariatric surgery in class I obe-
efficacy of bariatric surgery in patients with class I obesity, sity is small and hampered by many factors related to
this panel reached a consensus on these recommendations: poor study design, short follow-up, and diversity of
clinical definitions. Accrual of controlled long-term
(1) The impact on health of class I obesity varies greatly data is strongly advised. The introduction in clinical
between subjects. However, the physical, psychological practice of novel procedures and new devices should
and social health burden imposed by class I obesity may be guided by the results of appropriately designed re-
be great at an individual level. search protocols conducted with the highest levels of
(2) Nonsurgical therapies may achieve a clinically ethical behavior.
meaningful weight loss in a significant number of
patients with class I obesity, but this weight loss is Position Statements
maintained in the long term only in a smaller
proportion of them. Introduction
(3) Bariatric surgery is a highly effective weight loss strategy
in patients with class I obesity at least in the medium The Global Pandemic
term. Adverse event’s rate in class I obese patients ap-
pears to be the same than in morbid obesity. The global pandemic of obesity continues to progress global-
(4) Access to bariatric surgery should not be denied to ly. The causes of this pandemic are complex. To date, any
a patient with class I obesity associated to signifi- attempts to control the trends of the pandemic appear
cant obesity-related co-morbidity simply on the ba- ineffective.
sis of the BMI level, which alone is an inaccurate
index of adiposity and a poor health risk predictor.
Achieving and Sustaining Weight Loss is Difficult
Patients with class I obesity who are not able to
achieve adequate weight loss after a reasonable
Regulatory processes that maintain body fatness are highly
period of nonsurgical therapy should be considered
efficient and any increase in weight is defended physiologi-
for bariatric surgery.
cally. Multiple interventions assist in inducing and maintain-
(5) Bariatric surgery should be considered in patients with
ing weight loss including lifestyle changes, specific diets,
class I obesity on an individual basis and after a compre-
medications, devices, and surgery. The extent to which
hensive clinical evaluation of the patient’s global health
sustained weight loss can be achieved varies with the
and a prediction of its future disease risk. The use of
intervention.
bariatric surgery in patients with class I obesity should be
considered only after failure of proper nonsurgical
therapy. Need for Effective Treatment
(6) Indication to bariatric surgery in class I obesity
should be based more on the comorbidity burden Obesity is a chronic disorder requiring a chronic disease
than on BMI levels. Comorbidities should be eval- model of care. Combining interventions and scaling-up ther-
uated considering their likely response to surgery apy for serious or resistant disease are usual parts of chronic
and in relation to how they can be treated by disease models of care.
established medical therapies.
(7) The use of bariatric surgery should be avoided in Responsibility
patients with class I obesity and advanced obesity-
related or obesity-unrelated comorbidities (frailty The International Federation for the Surgery of Obesity
patients), in which intentional weight loss may and Metabolic Disorders (IFSO) recognizes its responsi-
not have any beneficial effect on prognosis or bility in developing evidence-based position statements
may be harmful. regarding new and emerging areas related to bariatric
(8) The use of bariatric surgery cannot be currently recom- and metabolic surgery. This draft position statement
mended in children/adolescents or in elderly obese pa- examines the use of bariatric surgery in the class I
tients with class I obesity. obesity range (BMI 30–35 kg/m2).
OBES SURG (2014) 24:487–519 489

Impact of Class I Obesity on Health Current Pharmacologic Approaches

Mortality Pharmacotherapy of obesity is rapidly evolving, with many

new drugs or combination drugs moving closer to clinical use.
Most recent epidemiologic data suggest that the BMI range Many of these drugs seem to have the capability to potentiate
with the lowest mortality is in the overweight range, with the significantly the effects of life-style modifications, with 25–
risk for the normal weight and that of class I obesity being 50 % of patients achieving the 10 % weight loss target. Weight
similar. Mortality rates are increased only in patients with maintenance seems also to be facilitated. However, long-term
BMI > 35 kg/m2. The effect of intentional weight loss on results (>2 years) are currently available only for orlistat.
mortality is unclear for patients with class I obesity. Long-term efficacy, tolerability, and adverse events of new
combination and single drug regimens remain to be
established.
Risk of Comorbidity
Endoscopic Procedures
The risk of developing type 2 diabetes, hypertension, dyslip-
idemia, metabolic syndrome, obstructive sleep apnea, poly- The intra-gastric balloon, several forms of transoral/
cystic ovary syndrome, depression, osteoarthritis, and nonal- endoscopic gastric partitioning, and novel endoscopic devices
coholic fatty liver is increased in class I obesity. Class I obesity mimicking the effect of the exclusion of the proximal intestine
is clearly associated with an increased risk of many cancers. have been proposed as less invasive alternatives to bariatric
Class I obesity impairs physical and mental health-related surgery. The efficacy, safety, durability, and long-term clinical
quality of life and it is associated to an increased psychosocial utility of these procedures remain to be established.
burden, particularly in women.
The Current Position of Bariatric and Metabolic Surgery
for Class I Obesity
Nonsurgical Therapy for Class I Obesity
International Diabetes Federation
Treatment Goals
The International Diabetes Federation (IDF) experts sug-
The objective of nonsurgical obesity therapy is to achieve and
gested that diabetic patients with class I obesity may be
sustain a weight loss of as much as 10 % of initial body
eligible, but not prioritized, for surgery if they have poorly
weight. This degree of weight loss is considered to be safe
controlled diabetes despite fully optimized conventional ther-
and sufficient to obtain a significant improvement of general
apy, especially if their weight is increasing or other weight-
health in patients who are overweight or have class I obesity.
responsive comorbidities are not achieving targets on conven-
tional therapy.
Lifestyle Modification Programs
American Society for Metabolic and Bariatric Surgery
Randomized controlled trials (RCTs) demonstrated that life-
style modification programs may achieve a weight loss of 5– The American Society for Metabolic and Bariatric Surgery
7 % in approximately half of the patients. This modest weight (ASMBS) Clinical Issue Committee stated that bariatric sur-
loss is only partly maintained over time, but may convey gery should be an available option for suitable patients with
important health benefits such as diabetes prevention in at- BMI 30–35 kg/m2 who do not achieve substantial and durable
risk populations and improved metabolic control in patients weight and comorbidities improvement with nonsurgical
with type 2 diabetes. Prevention of cardiovascular events was methods.
not achieved.
Beyond BMI in the Selection/Prioritization of Obese Patients
for Surgery
Meal replacements and Very Low Calorie Diets
BMI Alone is a Poor Index of Adiposity
The use of meal replacement programs and very low energy
diets may induce greater weight loss than conventional diets in BMI is used in epidemiological and clinical practice to diag-
the short term. However, it is not clear whether the initial nose and categorize obesity. Eligibility to bariatric surgery has
weight loss advantage obtained would result in a better weight been based so far largely on BMI cut-offs. However, the use of
loss maintenance over the long term. BMI as a proxy for adiposity, the true determinant of the obese
490 OBES SURG (2014) 24:487–519

state, is misleading, giving that its value is influenced also by Prospective Observational Studies and Retrospective Studies
skeletal muscle and organs mass.
Multicenter or single-site, observational, prospective, or ret-
rospective studies analyzed the outcome of class I obese
Beyond BMI in the Definition of Cardiometabolic Risk
patients sorted out from general bariatric surgery series and
in Obesity
reported satisfactory weight loss, with resolution or improve-
ments of comorbidities.
Patients in class I obesity may have very different levels of
health and risks at the same BMI level. Visceral fat accumu-
Final Summary
lation and the presence of ectopic fat deposition in relevant
organs are the most important determinants of cardiometabol-
A comprehensive evaluation of the randomized control trials,
ic risk in class I obesity. BMI does not convey any information
meta-analysis, and prospective or retrospective studies dem-
on these biological body components.
onstrated that overall weight loss was excellent in patients
with class I obesity after all the most established bariatric
Beyond BMI in Phenotyping Obese Patients procedures, with some studies reporting better excess weight
loss in this group of patients compared to patients with morbid
The use of only BMI in the selection of obese patients for obesity. Adverse event’s rate in class I obese patients appears
surgery should be abandoned. A clinical decision should be to be the same than in morbid obesity, with some studies
based on a more comprehensive evaluation of the patient’s reporting serious adverse events.
global health and on a more reliable prediction of its future
disease risk. Limitations

RCTS are few and observational studies contain several meth-

Obesity Scoring Systems odological deficiencies, with lack of control data, propensity
to bias, and lack of information. Length of follow-up is short
The use of a score that could quantitatively represent the (<2 years) in most of the studies in patients with class I
actual and future health burden that obesity induces in the obesity. Shortness of follow-up limits our knowledge on the
individual patient would be an important tool for clinicians for long-term risk / benefit ratio of surgery in this subset of
phenotypization beyond BMI levels and for guiding therapeu- patients. In particular, potentially serious effects of the pro-
tic choices. The validation and application of obesity scoring found weight loss produced by surgical procedures on nutri-
systems or algorithms should be implemented. tional status and body composition cannot be evaluated. Fi-
nally, reliable information about the effects of bariatric surgery
on longevity in patients with class I obesity remains complete-
Surgery in Class I Obesity: “What Do We Know”
ly lacking.
and “Identify Gaps”

Special Considerations Regarding Patient Selection

Randomized Controlled Trials
Ethnicity
Four RCTs evaluated the results of bariatric surgery in sam-
ples including patients with class I obesity and one was
The risk of metabolic syndrome and type 2 diabetes varies
specifically conducted in patients with class I obesity. All
with ethnicity. Adjusted BMI action cut points for Asians or
the studies reported consistent weight loss and comorbidities
other high-risk ethnic groups are recommended.
improvements.
Age
Meta-Analysis and Systematic Reviews in Patients with Type 2
Diabetes The use of surgery should not be extended in children and
adolescents with BMI < 35 kg/m2 as long as its efficacy and
Comprehensive reviews evaluated nonrandomized prospec- safety would be not more firmly proved in adults. Data about
tive and retrospective studies in patients with type 2 diabetes efficacy and safety of surgery in class I obese adolescents are
and BMI < 35 kg/m2. Both traditional and experimental lacking. The optimal weight for lowest mortality appear to be
procedures were included. Weight loss, diabetes remission in the overweight/class I obesity range in the elderly and there
rates, and improvements in lipids and metabolic syndrome is no clear guidance regarding intentional weight loss in older
were judged to be as good as in morbid obese patients. adults, as it is unclear that benefits outweigh risks. The use of
OBES SURG (2014) 24:487–519 491

bariatric metabolic surgery cannot be currently recommended designed clinical human trials, with small open-label feasibil-
in older adults with a BMI < 35 kg/m2. ity trials performed first and larger-scale investigations with
sufficient follow-up thereafter. If feasible, a randomized sham-
Regional, Economic, and Equity Considerations controlled trial should be performed. Each procedure or de-
vice, having different safety profiles, degree of complexities,
There are regional variations in access to bariatric metabolic and outcome results, should be judged by its own defined set
surgery and regional differences in the regulatory and eco- of criteria. Procedures that are less radical, less complex, and/
nomic conditions that may limit the direction of surgery for or less risky can be acceptable even if they result in signifi-
patients with class I obesity. National and regional health cantly less benefit than more complex procedures that have
services providers need to consider the evidence and deliver higher complication profiles.
services that are locally appropriate.

Comorbidity Reporting Weight Loss Outcomes

Metabolic, mechanical, and psychological comorbidities of There is still no scientifically validated and universally ac-
obesity often cluster and are associated with increased risk cepted method for measuring and recording weight loss out-
of morbidity and mortality that is poorly related to BMI. In comes. Professional medical societies and medical journals
indicating bariatric metabolic surgery in patients with class I still differ on the preferred method.
obesity, comorbidities should be evaluated considering their
likely response to surgery and in relation to how they can be
Measuring and Reporting Comorbidity Outcomes
treated by established medical therapies.
Universal standard definitions for comorbidity outcomes need
Low BMI as a Consequence of Previous Medical or Surgical
to be instituted for clinical practice and research protocols.
Therapy
This would include a uniform acceptance of the definition of
each disease state, a uniform definition of the chemical
BMI criterion for election to bariatric metabolic surgery
markers used to label a patient with suffering from a particular
should be the current BMI or a documented previous BMI
disease, a uniform terminology for determining the severity of
of this severity. Surgery may be indicated at low BMI levels in
the disease, uniform and scientifically based criteria for the
patients who exhibited a substantial weight loss in a conser-
various outcomes after surgery, and consensus on what con-
vative treatment program but started to gain weight again or in
stitutes “best” medical therapy.
bariatric patents having reached a low BMI after a first inter-
vention, but requiring redo surgery for complications or side
effects. Is There a Need for a Large RCT Looking at Hard Outcomes?

Research Gaps and Priorities RCTs represent the highest standard in clinical investigation.
Randomizing patients to different study groups dramatically
Long-Term Outcomes reduces differences, inequalities, and biases between study
and control subjects. However, RCTs are difficult to conduct
One of the biggest deficiencies of the prevailing literature in the field or bariatric surgery and large long-term RCTs
concerning both conventional and nonconventional surgical present formidable challenges.
procedures for BMI < 35 kg/m2 is the lack of long-term
outcome data. It would be reasonable to suggest that adequate
postoperative follow-up for the sake of investigational data Ethics of Surgery for BMI < 35 kg/m2
collection and procedure evaluation should be no less than
3 years and preferably 5 years. The ethical behavior for studying or treating patients whose
BMI < 35 kg/m2 by surgical interventions should be rigorous.
How to Assess new Procedures, Devices, and Techniques While there is an overwhelming body of evidence that con-
cludes that bariatric surgery is safe and effective for patients
All new procedures, devices, and techniques mandate rigor- whose BMI ≥ 35 kg/m2, it cannot be assumed that the results
ous assessment before being offered to patients. To minimize would be the same for patients with BMIs < 35 kg/m2. Novel
harm, new procedures should undergo extensive preclinical metabolic operative procedures and devices are still investi-
investigation. After demonstration of efficacy and safety, the gational and must be treated as such. Standard rules of ethical
procedure should be rigorously evaluated in carefully research needs to be applied to patients in class I obesity.
492 OBES SURG (2014) 24:487–519

Introduction protecting life in a similar way to maintaining blood pressure,

temperature, and blood glucose. It is extraordinary how accu-
Global Pandemic—Complex Determinants—Obesity rate regulation is for body weight and what drives the dysreg-
Prevention ulation that leads to a tiny but regular increase in weight in
obese individuals is poorly understood. Unfortunately, any
The global pandemic of obesity with its associated comorbid- increase in weight is defended physiologically in a similar
ity has progress steadily and inexorably since the late 1970s way in which a person with hypertension defends and inap-
and foreseeably the most serious and costly health issue for propriately high blood pressure and therapy needed. Multiple
this century. The magnitude of rise has varied with region, interventions assist in generating and maintaining weight loss
country, and with gender; however, stabilization of the obesity including lifestyle changes, specific diets, physical activity,
prevalence is rare, and of great concern, the rise has acceler- medications, devices, and surgery. The extent to which
ated globally over the last decade. The global age- sustained weight loss can be achieved varies with the inter-
standardized prevalence of obesity [body mass index (BMI) vention used, and there is also great variability within inter-
≥ 30 kg/m2] nearly doubled from 6.4 % in 1980 to 12.0 % in ventions suggesting individual differences in response are
2008. Half of this rise occurred in the 20 years between 1980 important.
and 2000 and half occurred in the 8 years between 2000 and
2008 [1]. With increasing levels of obesity we see an expo- Need for Effective Treatment
nential rise in class III obesity (BMI > 40 kg/m2). In the US
between 2000 and 2005, the prevalence of obesity increased Obesity and related disorders that it generates are chronic
by 24 %, class III obesity by 50 % and BMI > 50 kg/m2 conditions requiring a chronic disease model care. As for
increased by 75 %, two and three times faster, respectively [2]. any chronic disorder, a cure of the condition is an ultimate
The resultant exponential increase in class III obesity and goal, but difficult to achieve. Obesity is no exception. We can
super obesity is an expected trend as the mean BMI for a manage obesity, but to date have no cure and we need a range
community steadily increases. Sadly, no part of the globe is of effective therapies. As for all chronic disease management,
protected from the obesity pandemic, and the transition from we engage the informed patient actively in the management
undernourished to overnourished occurs with alarming speed decision-making process for long-term care with the ultimate
in developing countries in both urban and rural regions [3]. aim of using evidence-based interventions to improve long-
The causes of this pandemic are complex and extend beyond term outcomes. Combining interventions and scaling-up ther-
the relevant, but overly simplistic, view that it is driven by apy for serious or resistant disease are usual parts of this
western fast food diets and sedentary nature of modern living. continuum of care.
Genetics, metabolic programming, epigenetic changes, increas- All interventions have a range of benefits and risks, and this
ing maternal age, and assertive mating set the baby born today need to be balanced for each individual. Medications used to
up for an aberrant interaction with today’s environment gener- treat diabetes and cardiovascular risks sometimes appear to
ating an almost inevitable problematic weight trajectory for the have benefits beyond their initial primary target: metformin an
majority. Arguably the most important 4 years for an individual effect beyond glycemic control, statins effects beyond an
weight trajectory for life are the 4 years prior to their third influence on LDL-cholesterol, and angiotensin-converting-
birthday. In addition many aspects of modern living impercep- enzyme inhibitors beyond blood pressure. Similarly bariatric
tibly and passively contribute to the issue: shorter sleeping procedures originally designed primarily for weight loss have
hours, increase screen time, temperature-controlled environ- effects beyond that weight loss. These additional benefits may
ments, endocrine modifiers, and medications for many chronic alter the indications for therapy, but can also come at a cost
conditions increasing weight. with unexpected downsides. One only needs to look at the
The complexity of the determinants of this pandemic gen- medications used for weight management and to treat type 2
erate major challenges in prevention and arguments regarding diabetes for unexpected downsides, and it would be naive to
the degree of personal responsibility and blame vs. the need for expect that interventions targeting the GI tract using surgery or
a widespread environmental makeover with greater regulation devices and having effect on complex biological systems
lead to philosophical stalemates, commercial conflicts, and would be exempt from similar unfavorable downsides.
systematic inertia. To date, any attempts to control the trends The International Federation for the Surgery of Obesity and
of the pandemic appear piecemeal, tokenistic, and ineffective. Metabolic Disorders (IFSO) recognizes its responsibility in
developing evidence-based position statements to guide its
Achieving and Sustaining Weight Loss is Difficult members, and health services providers generally, regarding
new and emerging areas related to bariatric and metabolic
Regulatory processes that maintain body weight and body surgery. Extending the indications for bariatric–metabolic sur-
fatness are highly efficient and physiologically critical for gery beyond traditional boundaries defined by age, body mass
OBES SURG (2014) 24:487–519 493

index, level of comorbidity, and operative risk is such an area. obesity are not associated with a reduced life expectancy [5].
This position statement examines the use of bariatric–meta- In those of East Asian ethnicity, there was an increased mor-
bolic surgery and devices in the class I obesity range (BMI tality with class I obesity, but in South Asians, there does not
30–35 kg/m2) with the appropriate adjustment for ethnicity. appear to be an increase [6]. However, action BMI cut points
The process, led by IFSO president-elect Luigi Angrisani for Asia are reduced by 2.5 kg/m2 to BMI 27.5, 32.5, and
in 2013–2014, commenced with the development of an expert 37.5 kg/m2, respectively [7]. There are difficulties looking at
working group: specific causes of mortality as obesity is often not split into
BMI subclasses. As class I obesity is not associated with a
(1) To critically review the current knowledge regarding the major overall increased mortality risk, then are there sub-
epidemiology, health risks, and current therapies for groups that are at increased risk?
those with class I obesity; Type 2-diabetes is often the disease that generates interest
(2) To review the evidence for bariatric metabolic surgery in in lower BMI bariatric–metabolic surgery, however, those
those with class I obesity: efficacy and safety, relative with diabetes and a BMI in the class I obese range may have
risk and benefit, and effect on obesity-related the lowest risk of mortality. The weight status at the time of
comorbidity; diabetes diagnosis was examined in relation to mortality in
(3) To examine the broader issues involved from a health five major longitudinal studies. Overweight and obese had
care prioritization and delivery perspective in of extend- lowered all cause, cardiovascular, and noncardiovascular haz-
ing bariatric–metabolic surgery to the class I obese ard ratios. These were not altered after adjustment for demo-
range; graphics, smoking, and cardiovascular risk factors [8]. Two
(4) To develop practical recommendations for clinicians; additional studies have demonstrated a nadir of mortality in
(5) To identify gaps in our present knowledge and identify class I obese individuals [9, 10]. Data from the PROactive
priorities for clinical research in using established bariat- study, a randomized trial using pioglitazone, demonstrated an
ric–metabolic procedures; and increased mortality with weight loss and a reduction with
(6) To provide guidance regarding development and intro- weight gain [10]. A follow-up study of men with diabetes
duction of novel procedures and devices for bariatric attending US Veteran Affairs clinic found an inverse associa-
metabolic surgery. tion between BMI and mortality [11]. This pattern extends to
Taiwan where there was also an inverse relationship between
BMI and mortality in those with type 2 diabetes [12]. It is of
note that this mortality data in those with diabetes has largely
Impact of Class I Obesity on Health been published since 2011 and therefore represent a recent
clinical consideration. In addition, the early cessation of the
To assess the place of bariatric–metabolic surgery in those Look Ahead study for failing to demonstrate hard cardiovas-
with class I obesity (BMI 30–35 kg/m2), we need to under- cular and mortality advantage in overweight and obese partic-
stand the risk associated with this BMI range. It is also ipants to an intensive lifestyle program including intentional
important to consider that the great majority of obese individ- weight loss [13] raises questions about the value of intentional
uals are in the class I obese category with considerable public weight loss in overweight and obese (class I) individuals with
health and health economic impact. In the US Centers for type 2 diabetes. Of course bariatric surgery has demonstrated a
Disease Control and Prevention (CDC), data indicate that mortality advantage in severely obese patients with and with-
approximately 2/3 of obese men are in the class I obese range, out diabetes, but these studies were all restricted to those with
and 50 % of obese women are in the class I obese range. conventional BMI indications for surgery (BMI > 35) and
where an increased overall mortality is demonstrated [14].
Mortality The last decade has seen a puzzling array of data demon-
strating a mortality advantage in overweight and obese pa-
Most recent epidemiologic data suggest that the BMI range tients with cardiac failure, myocardial infarction, coronary
with the lowest mortality is actually the overweight range artery disease, post-coronary bypass surgery, and renal failure
(BMI 25–30 kg/m2) with the risk for the normal weight range compared with those of normal weight [15, 16]. Indeed there
(BMI 18.5–25 kg/m2) and that of class I obesity range similar appears to be a U-shaped relationship between BMI and
[4]. Mortality rates are increased in those with a BMI of 35 kg/ mortality with the nadir shifted up in BMI with many diseases
m2 or greater. These findings are common across multiple and with aging [17].
countries and ethnic groups. This data comes from 97 studies In summary, mortality risk for those with class I obesity
providing a combined sample size of more than 2.88 million may be higher than for overweight individuals, but similar to
individuals and more than 270,000 deaths. When examining those of the normal weight range, when all of the adult studies
years of life lost related to obesity, overweight and class I are combined. However, with aging and in many disease
494 OBES SURG (2014) 24:487–519

states, the nadir for mortality risk increases to a higher BMI Quality of Life, Psychosocial Burden, and Costs
level, and under these circumstances, mortality risk of those
with class I obesity may appear to be lower than those in the There is consistent evidence that obesity impairs phys-
normal weight range. Many issues may confound these re- ical and mental aspects of health-related quality of life.
sults, for example, smoking and issues related to illness and The effect is graded with increasing levels of obesity,
unintentional weight loss. The effect of intentional weight loss and utility-based quality of life measures are important
on mortality is unclear for those in the class I obese range and when wanting to perform health economic studies of
requires careful research. From a mortality perspective, the cost effectiveness [27]. Patients seeking bariatric–meta-
risk with surgery vs. the benefit with surgery is shifted toward bolic surgery report poorer health-related quality of life
risk, when compared with those with classes II and III obesity, than matched controls not seeking a surgical solution
as surgery itself is always associated with risk, and any addi- [28].
tional benefits and risks of surgery are unknown. Bariatric– There are many psychosocial demographic factors
metabolic surgery in patients in the class I obese range gener- associated with obesity and these can vary with ethnic-
ates major improvements in diabetes and cardiovascular risk ity. Depression, low self-esteem, binge-eating disorder,
factors, however, this may not translate to mortality benefit lower employment opportunity, and stigmatization and
and there is the potential that intentional weight loss may discrimination all tend to have greater impact on women
increase mortality risk for some patient groups. and all increase with increasing levels of obesity. Obe-
sity, diabetes, and depression are conditions that all
cluster together in low socioeconomic groups [29].
Risk of Comorbidity Levels of conditions such as depression and binge-
eating disorder are higher in those seeking surgery than
The risk of developing type 2 diabetes increases progressively those of the same BMI in the general community [30,
with increasing BMI. The adjusted relative risk of developing 31]. Taken together, these issues raise concern about
type 2 diabetes is 93 times higher in women with BMI 35 kg/ equity of access to bariatric–metabolic surgery and are
m2 than in women with normal BMI, and 42 times higher in important considerations to those providing national
men with BMI 35 kg/m2 than in women with normal BMI, but health care service delivery.
a significant increase in prevalence is observed also in class I The overall health costs related to obesity are esti-
obesity [18, 19]. There is also consistent and impressive data mated to be 4–8 % of health budgets and growing. The
from lifestyle programs and bariatric surgery that weight loss personal costs to the individual are considerable and
reduces the risk of developing type 2 diabetes [20–22]. The include additional health costs, reduced employment,
risk of developing hypertension also increases progressively employment opportunity, employment discrimination
with BMI and the prevalence is 18 % in those with a normal and lower income, and increased disability, injury, and
BMI and 39 % in those with class I obesity [23]. Dyslipidemia likelihood of requiring social support through pensions
has a complex association with BMI, with the highest [32]. These costs are partly borne by the community,
risk in the BMI 30–40 kg/m2 range and a reduction in and in addition, lost productivity related to obesity
risk at greater BMI levels [24]. These risk prevalence through absenteeism and presenteeism are considerable
data would suggest that cardiovascular mortality would and also grade up with increasing BMI and especially
be considerably increased in the class I obese popula- in the classes II and III obese categories [33].
tion but the risk, if any, is modest when looking at
epidemiological data. The risk of obstructive sleep ap-
nea, polycystic ovary syndrome, metabolic syndrome, Conclusion
depression, osteoarthritis, and nonalcoholic fatty liver
are all increased with class I obesity when compared The impact of class I obesity on mortality is consider-
with a normal weight population [25]. ably less than that of classes II and III obesity and the
BMI is linked in a positive way to the risk of many cancers, benefit of bariatric metabolic surgery in terms of lon-
and cancer incidence increases progressively with increasing gevity may be far more difficult to define. The role of
levels of obesity. In men, a 5 kg/m2 increase in BMI is strongly substantial intentional weight loss on total mortality,
associated with esophageal adenocarcinoma and with thyroid, while providing a logical therapy, is yet to be deter-
colon, and renal cancers. In women, strong associations are mined in the overweight and class I obese range. Class
with endometrial, gallbladder, esophageal adenocarcinoma, I obesity conveys an increased risk of comorbidities,
and renal cancers. Other cancers have weaker associations impairs physical and mental health-related quality of
with obesity [26]. However, class I obesity is clearly associ- life, and it is associated to an increased psychosocial
ated with an increased risk of many cancers. burden, particularly in women.
OBES SURG (2014) 24:487–519 495

Nonsurgical Therapy for Class I Obesity effect of lifestyle modifications on diabetes was observed
years after the termination of the trials, when most of the
Introduction effects of the intervention program on body weight were no
longer evident [21]. In the Diabetes Prevention Program
Current treatment guidelines set the objective of nonsurgical (DPP) [38], 3,234 overweight or obese patients with impaired
obesity therapy to achieve and sustain a weight loss of as glucose tolerance (BMI ≥25 kg/m2; mean BMI 34.0±6.7 kg/
much as 10 % of initial body weight for a period of time [34, m2) were randomly assigned to standard care, to standard care
35]. This degree of weight loss is considered to be safe and plus metformin, or to an intensive lifestyle intervention pro-
sufficient to obtain a significant improvement of general gram. The goals for the participants assigned to the intensive
health in patients who are overweight or have class I obesity. lifestyle intervention were to achieve and maintain a weight
More ambitious levels of weight loss are generally required reduction of at least 7 % of initial body weight through a
for patients with class II or class III obesity, the population in healthy low-calorie, low-fat 4diet and to engage in physical
which bariatric surgery is currently recommended [34, 35]. activity of moderate intensity for at least 150 min/week. Only
A systematic review of randomized controlled trials half of the participants in the lifestyle intervention group
(RCTs) conducted with traditional diet programs tends to achieved a 7 % weight loss during the program and this weight
demonstrate that the mean weight loss obtained in the first loss was only partly maintained over the 4-year follow-up.
year of treatment is generally inferior to the abovementioned However, the incidence of diabetes was again 58 % lower in
therapeutic objectives, and this weight loss is rarely main- the lifestyle intervention group than in the placebo group and
tained over time [36]. The adherence of obese patients to also 39 % lower than in the metformin group [38]. Addition-
dietary regimens tends to reduce after the first 6 months, and ally, the DPP population has been followed beyond the 4 year
a total or partial weight regain was usually observed thereafter. initial study period and continues to demonstrate an advantage
However, some patients were able to achieve and maintain a in diabetes prevention with life-style modification at even
10 % weight loss [36]. Additionally, some health benefits 6 years out [22]. A broad range of diabetes prevention studies
were observed at even less than a 10 % body weight loss. In have been conducted in developing countries in high-risk
this chapter, we briefly review the results of current nonsur- populations and all have been shown to be effective [39].
gical treatment options for class I obesity including long-term The Look AHEAD study was a very large RCT that also
lifestyle modification programs, meal replacements, very low examined the effects of an intensive lifestyle intervention on
calorie diets, current pharmacologic approaches, and novel the incidence of major CVD events in 5,145 overweight or
endoscopic procedures. obese individuals with type 2 diabetes (BMI ≥25 kg/m2; mean
BMI in the intervention group: 36.3±6.2 kg/m2 in women and
Lifestyle Modification Programs 35.3 ± 5.7 kg/m2 in men) [40]. Patients were randomly
assigned to conventional diabetes support and education or
Two large seminal randomized control trials (RCTs) tested the to an intensive lifestyle intervention program with a 10 % weight
efficacy of lifestyle modification programs for the prevention loss goal. Over the first year, the intensive lifestyle intervention
of type 2 diabetes in high-risk populations [37, 38]. These group lost an average of 8.6 % of initial body weight, with
trials were not specifically conducted in the overweight or 37.8 % of participants having a greater than 10 % weight loss,
patients with class I obesity, but were instead conducted on and 55.2 % of subjects achieved a greater than 7 % weight loss
participants whose mean baseline BMI was in the 30–35 kg/ [40]. Participants in the intensive lifestyle intervention main-
m2 range [4, 5]. In the Finnish Diabetes Prevention Study [37], tained a mean weight loss of 4.7 % at year 4 of the study [41].
523 overweight or obese patients with impaired glucose tol- This moderate but sustained weight loss was associated to
erance (BMI ≥ 25 kg/m2; mean BMI in the intervention group: improvements in fitness, glycemic control, and CVD risk factors
31.3±4.6 kg/m2) were randomly assigned to standard care or [41] and to very small rates of complete diabetes remission [42].
to an intensive lifestyle intervention with specific dietary and However, the National Institutes of Health decided to prema-
physical activity goals including the achievement and main- turely halt the Look AHEAD trial because of a failure to achieve
tenance of a weight loss of at least 5 % of initial body weight. a significant reduction in the occurrence of cardiovascular events
During the first year of the study, body weight decreased by a in the intervention group [13].
mean of 4.7 % in the intervention group. However, only 43 % In summary, RCTs demonstrated that lifestyle modification
of patients in the intervention group had a weight loss greater programs may achieve a modest weight loss of 5–7 % but only
than 5 % of initial body weight and in the second year of the in approximately half of the patients. This modest weight loss
study, weight regain was observed. On a positive note, the is only partly maintained over time, but may still convey
study did demonstrate that the 4-year cumulative incidence of important health benefits, such as diabetes prevention in at-
diabetes was 58 % lower in the intervention than in the control risk populations and improved metabolic control in patients
group (p<0.001) [37]. Moreover, a prolonged protective with type 2 diabetes.
496 OBES SURG (2014) 24:487–519

Meal Replacements and Very Low Calorie Diets patients (BMI ≥ 30 kg/m2; mean BMI in the intervention
group: 37.4±4.5 kg/m2) were randomly assigned to orlistat
Meal replacement programs are low-calorie diet plans where- plus lifestyle changes or to placebo plus lifestyle changes [53].
by one or two meals of the day are replaced by commercially Mean weight loss was significantly greater for the orlistat
available, energy-reduced products that are vitamin- and group than the placebo group at 1 year (10.6 vs. 6.2 kg) and
mineral-fortified. These programs have been proposed as remained significantly greater at the end of the 4-year study
more effective weight reduction strategies in obese patients (5.8 vs. 3.0 kg). Significantly more orlistat patients (41.0 %)
than conventional diets. A meta-analysis that analyzed six than placebo patients (20.8 %) achieved a weight loss greater
short-term studies of liquid meal replacements indicated that than or equal to 10 % after 1 year of treatment and for those
weight loss was greater in the meal replacement groups when patients who completed 4 full years of treatment, 26.2 and
compared to the calorie equivalent traditional diets, with an 15.6 %, respectively, lost greater than or equal to 10 % of
average 7–8 % body weight loss in the meal replacement baseline body weight [53]. Additionally, a 37.3 % decrease in
group compared with an average 3–7 % body weight loss in the risk of developing diabetes was observed in the orlistat
the conventional diet group [43]. However, it is not clear group as compared to placebo [53].
whether the initial weight loss advantage obtained by meal Apart from orlistat, many new drugs or combination of
replacement strategy would result in a better weight loss drugs are now in advanced phases of clinical research and/or
maintenance over the long term. entering clinical practice. Lorcaserin, a selective serotonin 2C
Very low energy diets (VLEDs) are defined as diets that receptor agonist, has been tested against placebo in the Be-
provide less than 800 kcal/day. They are usually prescribed as havioral Modification and Lorcaserin for Overweight and
a synthetic or food-based formulation of 450–80 kcal/day, Obesity Management (BLOOM) trial, a 2-year RCT that
provided high levels of protein, and supplemented with vita- enrolled 3,182 obese patients with BMI of 30–45 or 27–
mins, minerals, electrolytes, and fatty acids. VLEDs are sec- 45 kg/m2 with at least one coexisting condition [54]. At the
ond to surgery in generating weight loss and the ketosis end of the first year of the study, patients in the lorcaserin
generated by low carbohydrate intake and in utilizing stored group lost an average of 5.8 % of the baseline body weight, as
fat for energy provides suppression of appetite by altering compared with 2.1 % in the placebo group (P<0.001). In
some of the physiological changes to weight loss [44]. VLEDs addition, a greater percentage of patients lost 10 % or more
have been proposed as a more effective method for weight of their baseline body weight in the lorcaserin group (22.6 %)
loss in obese patients. The efficacy and safety of modern than in the placebo group (7.7 %). For year 2, patients who
VLEDs in obese patients [45] and in obese patients with type had been receiving placebo continued to receive it, whereas
2 diabetes [46] have been recently revisited. In summary, the patients who had been receiving lorcaserin were again ran-
use of a course of VLED may safely produce a large initial domly assigned either to continue to receive lorcaserin or to
weight loss, in the order of 1.5–2.5 kg/week. Obese patients begin to receive placebo. Among patients in the lorcaserin
treated by VLEDs may have better long-term weight mainte- group who had weight loss of 5 % or more at year 1, the loss
nance than patients treated by more conventional diet [47], was maintained in a greater proportion of patients who con-
and even a more effective weight maintenance may be ob- tinued to receive lorcaserin in year 2 compared with those who
served in those obese subjects with higher initial weight loss were reassigned to receive placebo (67.9 % vs. 50.3 %) [54].
results [48]. In summary, VLEDs require careful physician Lorcaserin was approved for clinical use in the US by the
supervision, but generally provide better sustained weight loss Food and Drug Administration (FDA) in 2012.
than other dietary methods and can be used intermittently or In the CONQUER study, a combination drug regimen with
on demand to maintain weight loss [49]. VLCDs have an low dose of phentermine and topiramate was tested for two
important role in reducing liver size prior to bariatric surgery different formulations (7.5 mg phentermine/46 mg controlled
[50]. release topiramate or 15 mg phentermine/92 mg controlled
release topiramate) against placebo in 2,487 subjects with a
Current Pharmacologic Approaches BMI greater than or equal to 27 and less than or equal to 45 kg/
m2 as well as suffering from more than two weight-related
The history of drug treatments for obesity has been one of comorbidities [55]. Phentermine is a central norepinephrine-
recurrent optimism followed by failure, with many promising releasing drug approved in some countries for short-term
drugs coming into clinical practice only for being subsequent- treatment of obesity as monotherapy and topiramate is an
ly withdrawn for unexpected side effects [51]. Currently, the anticonvulsant that has shown unexpected weight-loss prop-
only drug worldwide approved for weight loss treatment with erties. After 56 weeks, the change in body weight was −1·4,
a 15-year-long clinical experience is orlistat, a gastrointestinal −8·1, and −10·2 kg in the patients assigned to placebo,
lipase inhibitor [52]. In the XENDOS study, the longest du- phentermine 7.5 mg plus topiramate 46 mg, and phentermine
ration RCT published thus far with orlistat, 3,505 obese 15 mg plus topiramate 92 mg, respectively. The percent of
OBES SURG (2014) 24:487–519 497

patients achieving a weight loss greater than or equal to 10 % On the basis of this short overview of recent anti-obesity
was 7, 37, and 48 %, respectively [55]. Of the 866 completers drugs RCTS, we can conclude that pharmacotherapy of obesity
of the CONQUER study, 676 was subsequently enrolled in an is now rapidly moving forward, with many new drugs or
extension SEQUEL study and continued receiving their combination drugs with a very diverse spectrum of mechanisms
blinded treatments for an additional 52 weeks [56]. In this of action moving closer to clinical use. Many of these drugs
extension trial, the difference in weight loss observed in the seem to have the capability to potentiate significantly the effects
three treatment arms in the first year was maintained even in of lifestyle modifications on body weight, whereas 25–50 % of
the second year [56]. In 2012, the combination drug patients can achieve the 10 % weight loss target. Weight
phentermine/topiramate was approved for limited clinical maintenance seems also to be facilitated. However, long-term
use in the US by the FDA. This combination medication results (>2 years) are currently known only for orlistat (the least
was rejected by the EMA for use in Europe because further potent of the medications described above). Long-term effica-
evidence of cardiovascular safety was required. cy, tolerability, and adverse events of these new combination
Finally, the efficacy and safety of a combined treatment and single drugs regimens remain to be established.
with sustained–released naltrexone and bupropion was tested
in patients with BMI 30–45 kg/m2 and uncomplicated obesity Endoscopic Procedures
or with BMI 27–45 kg/m2 and controlled hypertension or
dyslipidemia. In two independent 56-week RCTs (COR-I The intra-gastric balloon, a temporary 6-month endoscopic
and COR-II) [57, 58], weight loss was significantly higher gastric restriction procedure where an inflatable balloon is
in the combination therapy groups than in the placebo groups. endoscopically inserted into the stomach, has been widely
Weight loss greater than or equal to 5 % was achieved by 16 % used for weight loss purposes in some countries for patients
participants assigned to the placebo group and 48 % assigned with class I obesity [63]. However, no prospective controlled
to naltrexone 32 mg plus bupropion in COR-I [57] and in 17.1 observations have supported the hypothesis that intra-gastric
and 55.6 %, respectively, in COR-II [58]. This combination balloon treatment would achieve better weight loss than diet in
preparation was rejected by the US FDA because further the long term and the procedure itself is invasive and not free
evidence of cardiovascular safety was required. from side effects and mortality [63]. Several forms of
Liraglutide is a glucagon-like peptide-1 (GLP-1) analog transoral/endoscopic gastric partitioning using various loca-
with a 97 % structural homology to human GLP-1. Native tions, techniques, and devices have been proposed as less
GLP-1 has a short elimination half-life of 1–2 min, whereas invasive alternatives to surgical gastric restriction [64]. There
liraglutide has a much longer half-life and can be administered are also under development novel endoscopic [65] or mixed
once daily by subcutaneous injection. Liraglutide was initially endoscopic laparoscopic [66] devices meant to mimic the
developed for the treatment of type 2 diabetes mellitus as it effect of the exclusion of the proximal intestine as achieved
was shown to decrease glycosylated hemoglobin and, at the with bariatric surgical procedures such as the gastric bypass.
same time, to reduce body weight [59, 60]. The weight loss However, the development of these techniques has been ham-
observed by liraglutide treatment in patients with type 2 pered by technical problems and side effects and long-term
diabetes supported the investigation of the drug as an anti- durability and sustainability remain completely undetermined
obesity treatment, for overweight and obese subjects without [64].
type 2 diabetes. Astrup and coworkers [61] randomized 563 In summary, the results of weight loss endoscopic proce-
obese patients (BMI 30–40 kg/m2) without type 2 diabetes to dures to date have been mixed, with some devices providing
liraglutide (1.2, 1.8, 2.4, or 3.0 mg once a day by subcutane- inadequate weight loss and others promising results. In addi-
ous injection), placebo or orlistat for a 20-week trial. Partici- tion, there are only limited published studies, most being small
pants on liraglutide lost significantly more weight than did series with short follow-up. These are considered less invasive
those on placebo and orlistat. Additionally, a greater number than most conventional bariatric surgical procedures but more
of individuals lost more than 5 % weight with liraglutide 3· invasive than medical therapies. However, while there is
0 mg (76 %) than with placebo (30 %) or orlistat (44 %). excitement in the novel medical device area, the efficacy,
Completers of this first short study entered an extension 2- safety, durability, and clinical utility of many of these proce-
year RCT, continuing on randomized treatment for 1 year, dures in the management of obese people diabetes is still to be
after which liraglutide- or placebo-treated individuals established and the procedures need to be considered still
switched to liraglutide 2.4–3.0 mg and orlistat was continued investigational at this stage.
as the only comparator. At the end of the trial, 52 % of the
patients on liraglutide 3.0 mg had a weight loss greater than 5, Combining Therapies
and 26 % of subjects had a weight loss greater than 10 %,
while corresponding figures in the orlistat group was 29 and As obesity is a chronic condition, selectively combining the
16 %, respectively [62]. therapies listed above may provide a more logical and
498 OBES SURG (2014) 24:487–519

sustained approach to therapy. Lifestyle modification forms The first significant attempts to consider bariatric surgery for
the basis for all weight management including bariatric sur- patients with BMI < 35 kg/m2 were stimulated by the increas-
gery. Achieving weight loss with low-calorie diets, meal re- ing awareness of the effects of surgery on type 2 diabetes. In
placements, VLEDs, and intra-gastric devices can be substan- 2007, a multidisciplinary group of experts convened in Rome,
tial and satisfactory for the class I obese range. However, all Italy, for the first international Diabetes Surgery Summit [69].
these approaches are followed by the challenge of weight After an extensive discussion on the evidences, the experts
regain. It is here that emerging pharmacotherapy may play released a clear position statement supporting the role of sur-
its most significant role. By combining currently available gery for the treatment of type 2 diabetes in acceptable surgical
therapy with medications expected in the near future, within candidates with BMI ≥ 35 kg/m2 not achieving adequate met-
a chronic disease framework, we may be on the verge of abolic control by lifestyle and medical therapy and, for the first
adequately treating large numbers of people in this BMI time, suggested that a surgical approach may also be appropri-
range. ate as a nonprimary alternative to treat inadequately controlled
type 2 diabetes in suitable surgical candidates with mild-to-
moderate obesity (BMI 30–35 kg/m2) [69]. Randomized con-
Conclusion
trolled trials on the field were strongly encouraged [69]. The
position statements released by the Diabetes Surgery Summit
According to the data revised here, we can conclude that the
were endorsed by several scientific societies, including IFSO
need for effective and safe therapies for class I obesity is great
[69]. Shortly after, the “bariatric surgery” section of the clinical
and not yet met by nonsurgical approaches. However, the field
recommendations for the standard of care in diabetes released
is rapidly evolving. Structured and feasible lifestyle modifica-
by the American Diabetes Association in 2009 recommended
tion programs may achieve modest weight loss in the range of
bariatric surgery for adults with BMI ≥ 35 kg/m2 and type 2
5–7 % of body weight in about half the patients and important
diabetes, especially if the diabetes is difficult to control with
health benefits. Established and novel pharmacologic treat-
lifestyle and pharmacologic therapy, but considered current
ments may significantly potentiate the effects of lifestyle
evidences insufficient to recommend surgery in patients with
modifications, in that 25–50 % of patients may obtain the
BMI < 35 kg/m2 outside of a research protocol [70]. Finally, the
10 % weight loss target. Definitive conclusions about long-
International Diabetes Federation (IDF) Taskforce on Epidemi-
term efficacy and safety of new combination and single drugs
ology and Prevention of Diabetes in 2011, 20 years after the
remain pending. Endoscopic alternatives are underdevelop-
NIH guidelines, released an important document on this subject
ment but will need more study to better understand the safety
[71]. After reviewing the accumulating studies on the role of
and efficacy.
surgery on diabetes, a consensus working group of diabetolog-
ists, endocrinologists, surgeons, and public health experts con-
cluded that there was clear evidence that bariatric surgery is a
Current Position of Bariatric and Metabolic Surgery very effective therapy for obese patients with type 2 diabetes
for Those with a BMI < 35 kg/m2 and attempted to position this therapeutic option in diabetes
treatment algorithms [71]. According to IDF experts, while
In 1991, a panel of experts endorsed by the National Institutes the indications for bariatric surgery typically classify those
of Health produced the first set of guidelines for the criteria for who are eligible for surgery, recommendations for surgical
selecting obese patients for bariatric surgery [67]. At that time referral as best practice or prioritization has not been widely
and without the support of evidence-based data, the panel considered. IDF suggested that conditional eligibility or pri-
decided to restrict the use of bariatric procedures to patients oritization for surgery should be assessed by a team special-
with severe obesity (BMI > 40 kg/m2) or with less severe izing in diabetes. Working on this framework, the IDF sug-
obesity (BMI 35–40 kg/m2) accompanied by severe obesity- gested the consideration for bariatric treatment for morbid
related comorbidities. This decision was driven by a prudent obese patients (BMI > 40 kg/m2) with type 2 diabetes not
evaluation of the risk / benefit ratio of bariatric surgery in an adequately controlled (HbA1c > 53 mmol/mol or 7 %) by
era in which open surgery was the rule and procedures were lifestyle measures and metformin. Less severe obese diabetic
limited to a very few options. However, the decision was patients (BMI 35–40 kg/m2) should be eligible for surgery and
subsequently supported by the accumulation of high-quality may be prioritized if they have poorly controlled diabetes
prospective data that confirmed that bariatric surgery resulted (HbA1c > 53 mmol/mol or 7.5 %) despite fully optimized
in an improvement of metabolic cardiovascular risk factors, a conventional therapy, especially if their weight is increasing or
reduction of coronary events, a lower incidence of cancer, and other weight responsive comorbidities (blood pressure, dys-
a reduction of total mortality [20]. The support for offering lipidemia, and obstructive sleep apnea) are not achieving
surgery to patients with BMI > 35 kg/m2 was subsequently targets on conventional therapy. Finally, diabetic patients with
endorsed by other independent entities [34, 35, 68]. class I obesity (BMI 30–35 kg/m2) may be eligible, but not
OBES SURG (2014) 24:487–519 499

prioritized, for surgery if they fall in the same metabolic and Beyond BMI in the Selection/Prioritization of Obese
clinical conditions warranting prioritization in the 35–40 BMI Patients for Surgery
class [71]. Australian clinical practice guidelines for the man-
agement of overweight and obesity also recently suggested BMI Alone Is a Poor Index of Adiposity
that bariatric surgery may be a consideration for people with a
BMI > 30 kg/m2 who have poorly controlled type 2 diabetes Body mass index (BMI) is used in epidemiological and clin-
and are at increased cardiovascular risk, taking into account ical practice to define underweight, normal weight, over-
the individual situation [72], and a similar position has been weight and obesity [34]. Categorization of obese patients for
also included in the updated version of the clinical practice the eligibility to bariatric surgery has been based so far largely
guidelines for the perioperative nutritional, metabolic, and on BMI cut-offs [67]. However, BMI is not a biological trait,
nonsurgical support of the bariatric surgery patient but a calculated value based on body weight. The use of BMI
cosponsored by the American Association of Clinical Endo- as a proxy for adiposity, the true determinant of the obese
crinologists, the Obesity Society, and the American Society state, may be misleading, given that body weight is the sum of
for Metabolic & Bariatric Surgery [73]. individual organs and tissues, and therefore it includes adipose
More recently, the American Society for Metabolic and tissue, skeletal muscle mass, and organs mass. Moreover, BMI
Bariatric Surgery (ASMBS) issued a very relevant position does not convey any information on fat distribution (e.g.,
statement in response to numerous inquiries made to the visceral fat accumulation and fatty infiltrations in individual
society by patients, physicians, society members, hospitals, organs) that is now considered an important determinant of
and others regarding the safety profile and efficacy of bariatric metabolic and cardiovascular risk [75].
surgery for patients with class I obesity (BMI 30–35 kg/m2) On a population level, a strong positive correlation between
[74]. By reviewing and summarizing available data, the BMI and overall body fat content has been extensively report-
ASMBS Clinical Issue Committee stated that class I obesity ed [76]. However, this can mask significant variations in the
is a well-defined deserving treatment disease causing or ex- relationship between BMI and adiposity on an individual
acerbating multiple other diseases, decreasing the duration level. For instance, at a given BMI value (24 kg/m2), the body
and the quality of life. Current options of nonsurgical treat- fat content has been demonstrated to vary in male and female
ment for class I obesity were considered by the ASMBS to subjects from 7.8 to 38.3 % and from 29.9 to 44.2 %, respec-
generally not be effective in achieving a substantial and dura- tively [77]. This large variability implies that an individual
ble weight reduction in the majority of patients treated by subject may have a BMI corresponding to an obese state (e.g.,
these measures. Therefore, the ASMBS concluded that bariat- 32 kg/m2) both having a low fat-free mass and a substantial fat
ric surgery should be an available option for suitable patients accumulation or having a large skeletal muscle mass and
with BMI 30–35 kg/m2 who do not achieve substantial normal fat mass. This latter condition typically occurs in
and durable weight and comorbidity improvement with athletes, in which high BMI may simply reflect increased
nonsurgical methods [74]. The ASMBS document muscle mass, which does not infer less favorable health [78].
stressed the fact that the existing cut-off of BMI, which The poor performance of BMI as a marker adiposity is further
excludes those with class I obesity, was established emphasized by the large differences in percentage body fat
arbitrarily nearly 20 years ago [67] and that on the observed between men and women at almost every BMI point
basis of currently available data, there is no current [77] and by the fact that a BMI of 20–25 kg/m2, which would
justification on grounds of evidence of clinical effec- be considered lean and by inference “healthy” within a Cau-
tiveness, cost-effectiveness, ethics, or equity that this casian population, may correspond to an elevated body fat
group should be excluded from surgery [74]. The content in other ethnic groups [79].
ASMBS Clinical Issue Committee indicated that gastric The deleterious and confusing consequences of the use of
banding, sleeve gastrectomy, and gastric bypass have BMI as a simple clinical and epidemiological marker for
been shown in RCTs to be safe, well-tolerated and obesity/adiposity may be better understood by considering
effective treatment for patients with BMI 30–35 kg/m2 the so-called “Obesity Paradox”. The term Obesity Paradox
in the short and medium term [74]. Finally, the ASMBS refers to a body of epidemiological observations in which
statement concluded that, before considering surgical treat- having a BMI level in the overweight or class I obesity range
ment for obesity for any individual, an adequate trial of seems to confer a survival advantage with respect to normal
nonsurgical therapy should always be required. If, however, weight and underweight patients in selected clinical situations.
the attempts at treating their obesity and obesity-related co- Indeed, a survival advantage in people with overweight or
morbidities have not been effective, we must recognize that moderate obesity, when compared with underweight or nor-
the individual has a disease that threatens their health and life mal weight subjects, has been described in patients with
expectancy and therefore must seek an effective, durable chronic heart failure [80–82]; in end-stage renal disease [83];
therapy such as bariatric surgery [74]. after major vascular surgery for peripheral arterial disease
500 OBES SURG (2014) 24:487–519

[84]; in patients who underwent a percutaneous coronary In the opposite corner, even if obesity is known to be
intervention for coronary artery disease [85]; in patients who related to numerous metabolic disturbances, a substantial pro-
are medically treated for non-ST-segment elevation acute portion of obese subjects appear to have a favorable metabolic
coronary syndrome [86]; and in the first 30 days after general profile: normal insulin sensitivity levels and blood pressure,
non bariatric surgery [87]. Several complex possible explana- high HDL, low plasma triglycerides levels, and absence of
tions for the Obesity Paradox have been advocated, but one inflammation. These subjects have been referred to as “meta-
common suggestion is that the very concept of the obesity bolically healthy obese” (MHO) [92]. The incidence of MHO
paradox may be driven by the deleterious effects of cachexia varies according to the criteria used for its definition, but it is
and not by salutary effects of obesity [16]. The protective substantial, covering about 25–35 % of the obese population
effect of a relatively high BMI level in such stressful clinical [92, 93]. This phenotype seems to be characterized by elevat-
situations may be therefore driven by having a good fat-free ed fat content and subcutaneous adipose tissue, but reduced
mass and a good nutritional status, instead of by a protective visceral and ectopic fat deposition.
effect of adiposity alone. However, we must also accept that These observations, coming from the characterization of
quality studies exclude early deaths to limit the effect of “extreme” or “outlier” metabolic phenotypes, emphasizes
disease-driven cachexia, and that a higher BMI at a population once more the role of fat distribution in the determinism of
level is largely associated with fatness and therefore, in some the metabolic and cardiovascular complications of obesity.
conditions, increased fatness may have mortality advantage. Visceral fat accumulation may, however, be difficult to quan-
These considerations may stress the need for a more accurate tify at a clinical level, and surrogate anthropometric indexes
description of body composition, fat distribution, and global have been proposed. Waist circumference has been proposed
health in patients with moderate obesity, which are considered as a reliable clinical indicator of visceral fat accumulation [34]
to be candidates for bariatric surgery. and having a large waist is associated to a higher prevalence of
metabolic disorders and cardiovascular diseases [34]. There-
Beyond BMI in the Definition of Cardiometabolic Risk fore, the measurement of the waist circumference is suggested
in Obesity to determine cardiovascular risk of overweight and obese
patients [34] and specific ethnic cut-off levels for waist cir-
The use of only BMI and obesity-related comorbidity for the cumference have been defined [94]. The simple measurement
selection and prioritization of patients to surgery has been of waist circumference has replaced the use of the waist-to-hip
frequently criticized, but no alternative options have been circumference ratio (WHR), originally proposed as a powerful
proposed. This issue will become even more important if we marker of fat distribution. More recently, on the basis of
move to include class I obesity in bariatric surgery. Indeed, several epidemiological studies showing that having a large
patients in this class may have very different levels of health hip circumference may confer some BMI-independent protec-
and risk even though at the same BMI level. tion from metabolic and cardiovascular diseases, particularly
Evidence in favor of the use of body composition and fat in women, a return to the measurement of hip circumference
distribution analyses in the categorization of metabolic risk has been proposed [95]. The presence of ectopic fat deposition
comes from data on a subgroup of normal weight subjects in the relevant organs may be even more difficult to quantify
with low subcutaneous but increased visceral fat mass. This than visceral fat accumulation in clinical practice. However,
thin-on-the-outside, fat-on-the-inside (TOFI) subphenotype liver fat infiltration (hepatic steatosis) may be roughly, albeit
has been observed in both male and female subjects and imprecisely, estimated by ultrasound [96] and precisely mea-
increases an individual’s risk of metabolic disease [77]. The sured by more advanced imaging techniques [77]. Increased
elevated visceral fat found in individuals classified as TOFI is liver fat has been suggested to be a more crucial determinant
accompanied by increased levels of ectopic fat deposition both of multiorgan insulin resistance than visceral fat [97]. An
in the liver and in the skeletal muscle. Lipid accumulation in alternative approach to the quantification of ectopic fat accu-
nonadipose cells (ectopic fat) may impair the normal function mulation may be represented by the ultrasonographic mea-
of some tissues through a process known as “lipotoxicity” surement of epicardial fat, which has been suggested as a
[85]. Ectopic storage of excess lipids in organs such as the further marker of metabolic and cardiovascular risk [98].
liver, skeletal muscle and pancreatic beta-cells may be the
causative link between fat distribution and the metabolic Beyond BMI in Phenotyping Obese Patients
syndrome [88]. Ectopic fat deposition has also been shown
to affect cardiovascular function and contribute to the devel- The use of only BMI in the selection of obese patients for
opment of cardiovascular diseases [89]. Similar findings have surgery appears now a clear oversimplification of the problem
been already reported in obese individuals, where obese sub- [99]. A clinical decision based on a more comprehensive
jects with a disproportionate accumulation of visceral fat had evaluation of the patient’s global health and on a more reliable
increased incidence of metabolic disorders [90, 91]. prediction of its future disease risk may be more sensible than
OBES SURG (2014) 24:487–519 501

neglecting or suggesting surgery to someone simply on the Table 1 A list of clinical factors that may potentially be integrated in a
comprehensive evaluation system for the selection or the prioritization of
basis of the calculated ratio between body weight and squared
obese patents for bariatric surgery
height. On the basis of the above considerations, a more
precise phenotypization of obese patients should include a Body composition BMI (% body fat, as determined by DEXA)
determination of percentage body fat with reliable techniques Fat distribution Waist circumference
(DEXA), particularly in cases where the BMI value may be Hip circumference
misleading, and an estimation of fat distribution and ectopic Visceral fat accumulation
fat deposition (waist circumference, hip circumference, hepat- Ectopic fat deposition Liver fat infiltration (hepatic steatosis)
ic steatosis, epicardial fat, etc.). Phenotyping should obviously Epicardial fat
be completed by the determination of cardiovascular risk Cardiovascular risk LDL-cholesterol, HDL-cholesterol, triglycerides
factors and clinical status of obesity-related comorbidities factors Fibrinogen
[73], and a comprehensive medical history for those factors hs-PCR
that may increase the risk of metabolic diseases in the future Obesity-related Type 2 diabetes
(family history of type 2 diabetes, gestational diabetes, poly- comorbidities Hypertension
cystic ovary syndrome, impaired glucose tolerance, or im- Obesity-related cardiomyopathy
paired fasting glucose) or may represent early markers of Sleep apnea syndrome
atherosclerosis (plaques or increased intima-media thickness Obesity/hypoventilation syndrome
at carotid ultrasonography, low ankle-brachial index, and high Disabling weight-bearing joint disease
coronary artery calcium score) or initial signs of organ damage Obesity-related infertility
(left-sided cardiac hypertrophy and microalbuminuria). Psy- Urinary stress incontinence
chological issues, eating behavior disorders, and quality-of- Severe gastro-esophageal reflux disease
life impairment should also probably be included. A list of High risk for type 2 Family history of type 2 diabetes
clinical data that should potentially be integrated in the com- diabetes Previous gestational diabetes
prehensive evaluation of the obese patients beyond BMI Polycystic ovary syndrome
values is reported in Table 1. Impaired glucose tolerance/impaired fasting
The integration of this large set of clinical information in a glucose
comprehensive picture would be highly facilitated by the Hyperinsulinemia/insulin resistance (HOMA)
adoption of an obesity scoring system. The use of a score that Early markers of Plaques or increased intima-media thickness at
could quantitatively represent the actual and future health atherosclerosis carotid ultrasonography
burden that obesity induces in every single patient would be Low ankle-brachial index
an important tool for clinicians for the phenotypization of the High coronary artery calcium score
patients beyond simple BMI level and for guiding therapeutic Initial signs of organ Left sided cardiac hypertrophy
choices. A scoring system should also be helpful for prioriti- damage Micro-albuminuria/proteinuria
zation and resources allocation in a health system with limited Socio and Depressive symptoms
resources. However, at this stage, we do not have an obesity psychological Eating behavior disorders
scoring system of this type already implemented and the issues
Reduced work capacity
relative weight to assign to each factor in the construction of Impaired quality of life
this score would be largely arbitrary in the absence of reliable
prognostic data. An alternative option would be the use of a
more simple but integrated staging system. The Edmonton
Obesity Staging System (EOSS) has been proposed by The application of a staging system for the selection/
Sharma and Kushner [100] as a clinical staging system for prioritization of obese patients to bariatric surgery beyond
obesity. EOSS classified obesity in five stages (0 to 4) accord- BMI values does not automatically imply that patients in the
ingly to worsening clinical and functional status (Table 2) most advanced stages should represent the best candidates for
[100]. EOSS stage has been shown to be a more stringent surgical procedures. Patients in EOSS stage 4 have a poor
predictor of total mortality than BMI levels in large epidemi- prognosis, a very high surgical and anesthesiological risk, and
ological databases [101, 102], and its application for the disputable benefits from intentional weight loss. Even patients
selection/prioritization of obese patients to bariatric surgery in stage 3, for instance, a patient with a recent myocardial
has been suggested [103]. The validation and application of infarction, may have clinical conditions that suggest surgery
EOSS or other alternative staging systems for the selection/ should be avoided or postponed. The clinical decision to
prioritization of obese patients to bariatric surgery beyond indicate a bariatric procedure should obviously also take into
BMI values should be a focus of future clinical research in account individual surgical risk. The surgical risk of bariatric
the field. procedures is generally low [104], but risk can be stratified by
502 OBES SURG (2014) 24:487–519

Table 2 Edmonton obesity scoring system: a proposed clinical and obesity with or without T2DM. The level of evidence from
functional staging of obesity (modified by Ref # [100])
these trials is high and the importance of surgical operations to
Stage Description reduce weight and to treat comorbidities is critical. Additional-
ly, two large meta-analysis/systematic reviews analyzed the
0 No apparent obesity-related risk factors (e.g., blood pressure, outcomes of several prospective and retrospective studies con-
serum lipids, fasting glucose, etc., within normal range), no
physical symptoms, no psychopathology, no functional
ducted in patients with class I obesity and T2DM, showing that
limitations and/or impairment of well-being surgical treatment is able to determine significant changes in
1 Presence of obesity-related subclinical risk factors (e.g., body weight, fasting plasma glucose, HbA1c, and lipid levels
borderline hypertension, impaired fasting glucose, elevated in diabetic patients with class I obesity. In the reviews, the rate
liver enzymes, etc.), mild physical symptoms (e.g., dyspnea on of complications and adverse were also evaluated in studies
moderate exertion, occasional aches and pains, fatigue, etc.),
mild psychopathology, mild functional limitations, and/or mild
with follow-up ranging from 6 months up to over 10 years.
impairment of well-being Finally, nine other observational studies, seven prospective, and
2 Presence of established obesity-related chronic disease (e.g., four retrospective, evaluated the effects of bariatric surgery on
hypertension, type 2 diabetes, sleep apnea, osteoarthritis, reflux class I obese patients with and without T2DM in terms of
disease, polycystic ovary syndrome, anxiety disorder, etc.), weight loss, diabetes remission, improvements in lipid and
moderate limitations in activities of daily living and/or well–
being
metabolic syndrome, and rate of complications.
3 Established end-organ damage such as myocardial infarction,
heart failure, diabetic complications, incapacitating Randomized Controlled Trials
osteoarthritis, significant psychopathology, significant
functional limitations and/or impairment of well–being Five RCTs evaluated the results of bariatric surgery in samples
4 Severe (potentially end-stage) disabilities from obesity-related including patients with class I obesity. Four of these trials also
chronic diseases, severe disabling psychopathology, severe
included patients with higher BMI levels and one was specif-
functional limitations, and/or severe impairment of well–being
ically conducted in patients with class I obesity. Considered
operations are gastric banding, gastric bypass, sleeve gastrec-
tomy. and mini-gastric bypass. Primary endpoint was diabetic
a range of simple clinical factors [105, 106]. Recently, differ- remission in four trials and weight loss in one. However, all
ent scoring systems have been proposed for more accurately the studies reported consistent weight loss and co-morbidity
predicting surgical complications in bariatric surgery reduction data. The principal characteristics of these five
[106–109]. The integrated use of more accurate instruments RCTs are reported in Table 3.
for a more complete clinical description of the obese patient In 2006, O’Brien et al. [110] randomized 80 patients with a
and for a more precise estimation of surgical risk may help BMI range of 30–35 to laparoscopic gastric banding or to
clinicians to base the clinical decision on a more logical and medical weight loss therapy. Duration of the study was 2 years
appropriate basis than the simple BMI level. and follow-up rate was 97 %. The surgical group achieved
greater weight loss than the medical group at 2 years (87.2 %
EWL vs. 21.8 %; p<0.001). Mean BMI decreased from 33.7
Surgery in Class I Obesity: “What Do We Know” to 26.4 in the surgical group and from 33.5 to 31.5 in the
and “Identify Gaps” medical group (p<0.001). The metabolic syndrome, defined
by the Adult Treatment Panel III Criteria, was initially present
To date, there is a robust body of literature to support safety, in 38 % patients in each group and was present in 3 % of
efficacy, and comorbidity benefits for patients with class I patients of surgical group and 24 % of patients of medical
obesity. This chapter performs a review of current evidences group at the end of the study (p<0.002). Quality-of-life chang-
on the role of bariatric surgery on class I obesity and one or es were measured with the Short Form 36 Health Survey (SF-
more obesity-related comorbidities, including type 2 diabetes 36). Quality of life improved more in surgical group (eight of
mellitus (T2DM), hypertension, hyperlipemia, obstructive eight subscales of Short Form-36) than in nonsurgical group
sleep apnea (OSA), obesity-hypoventilation syndrome (three of eight subscales). The gastric banding group had
(OHS), Pickwickian syndrome (combination of OSA and significantly greater improvements than the nonsurgical group
OHS), nonalcoholic fatty liver disease (NAFLD), nonalcohol- for physical functioning, vitality, and mental health. Four
ic steato-hepatitis (NASH), pseudotumor cerebri, gastro- serious adverse event were reported in surgical group
esophageal reflux disease (GERD), asthma, venous stasis (10 %); these patients developed posterior gastric wall pro-
disease, urinary incontinence, debilitating arthritis, or consid- lapse and needed revisional surgery [110].
erably impaired quality of life. After 2 years, in 2008, Dixon et al. [111] randomized 60
Five randomized controlled trials (RCTs) have been con- patients with a BMI range of 30–40 and recent-onset type 2
ducted to evaluate the role of bariatric surgery on class I diabetes (<2 years duration) to laparoscopic gastric banding or
OBES SURG (2014) 24:487–519 503

Table 3 Principal characteristics of randomized controlled trials including patients with class I obesity

Reference Pts. no. Pts. BMI/characteristics Arms FU length FU rate

O’Brien et al. [110] 80 30–35 Adjustable gastric banding vs. medical weight loss therapy 2 years 97 %
Dixon et al. [111] 60 30–40 (BMI < 35 in 13 pts.) Adjustable gastric banding vs. conventional diabetes therapy 2 years 92 %
with type 2 diabetes
Lee et al. [112] 60 25–35 with type 2 diabetes Mini gastric bypass vs. sleeve gastrectomy 1 year 100 %
Schauer et al. [113] 150 27–43 (BMI < 35 in 51 pts.) Roux-en-Y gastric bypass vs. sleeve gastrectomy vs. 1 year 93 %
with type 2 diabetes Intensive medical therapy
Ikramuddin et al. [114] 120 30–40 (BMI < 35 in 71 pts.) Roux-en-Y gastric bypass vs. intensive lifestyle-medical 1 year 95 %
with type 2 diabetes management

to conventional diabetes therapy focused mainly on weight gastrectomy (LSG) or intensive medical therapy (IMT). Du-
management. Duration of the study was again follow-up ration of the study was 1 year and follow-up rate 93 %. EWL
2 years and follow-up rate is 92 %. The surgical group was 88 % in LRYGB patients, 81 % in LSG patients, and
achieved significantly greater weight loss at 2 years (20 % 13 % in IMT patients (p<0.001). BMI change was −10.2,
of baseline body weight vs. 1.4 %; p<0.001). Mean BMI −8.9, and −1.9, respectively. Remission of diabetes was the
changed from 36.9 to 29.5 in the surgical group and from primary end point of this study. Baseline value of HbA1c was
37.1 to 36.6 in the conventional diabetes therapy group 9.3 % for LRYGB group, 9.5 % for SG group and 8.9 % for
(p<0.001). Moreover surgical group achieved greater results IMT group, while at 1 year, the values were 6.4 (change −2.9),
in terms of diabetes remission. Rate of diabetes remission in 6.6 (change −2.9), 7.5 (change −1.4), respectively, (p<0.001).
surgical group was 73 % (fasting glucose level <126 mg/dl The reduction in prevalence of the metabolic syndrome was
and glycated hemoglobin (HbA1c) value <6.2 % while taking significantly greater in the two surgical group than in medical
no glycemic therapy) compared with 13 % of the medical therapy group. Considering use of cardiovascular medica-
treatment group (p<0.001). There was a significant reduction tions, lipid lowering drugs were required at baseline in 86
in the use of drugs for glycemic control in the surgical group at and 78 % of patients assigned to LRYGB and SG, respective-
2 years and no decrease in the conventional managed group. ly, but use declined to 27 and 39 % at 1 year, as compared with
Weight loss and diabetes remission results were not separately 92 % at 1 year for IMT group (p<0.001). There was no
reported for patients with class I obesity and patients with significant difference in systolic and diastolic blood pressure
class II obesity in this study. No serious adverse events were levels among the three groups at 1 year, but there was a
reported in either group [111]. significant reduction in the number of anti-hypertension med-
In 2011, Lee et al. [112] randomized 60 patients with a ications after the two bariatric procedures. Weight loss, diabe-
BMI range of 25–35 and poorly controlled diabetes to lapa- tes remission, and metabolic syndrome results were not divid-
roscopic mini-gastric bypass (LMGB) or to laparoscopic ed for patients with BMI 27–35 and patients with BMI 35–43.
sleeve gastrectomy (LSG). There was no medical treatment In the surgical group, 15.5 % of patients had one or more than
arm in this study. The primary endpoint was diabetes remis- one serious adverse events requiring hospitalization including
sion (fasting glucose <126 mg/dl and HbA1c <6.5 % without four cases (4 %) of revisional surgery. In the IMT group,
glycemic therapy). Duration of the study was 1 year and hospitalization was required in 9 % of the patients [113].
follow-up rate 100 %. LMGB patients achieved 94 % EWL, Finally, in 2013, Ikramuddin et al. [114] randomized 120
LSG patients achieved 76 % EWL. BMI changes for LMGB patients with T2DM and a BMI range of 30–40 (59 % of
and LSG were −7.2 (from 30 to 22.8) and −5.6 (from 30 to patients with BMI < 35) to intensive lifestyle-medical man-
24.4), respectively. Significantly more LMGB patients achieved agement and Roux-en-Y gastric bypass (LRYGB) or intensive
diabetes remission compared to LSG patients (93 vs. 47 %, lifestyle-medical management alone. Medications for hyper-
respectively; p<0.02). HbA1c decreased from 9.9 to 5.4 % for glycemia, hypertension, and dyslipidemia were prescribed
LMGB patients and from 10.2 to 7.2 % for LSG patients. according to protocol. Duration of the study was 1 year and
LMGB patients achieved significantly greater improvements follow-up rate 95 %. Main outcome of the study was a
in lipids levels and metabolic syndrome than LSG patients. composite goal of HbA1c less than 7.0 %, low-density lipo-
There were no serious adverse events in either group. Minor protein cholesterol less than 100 mg/dl, and systolic blood
complications were recorded in 10 % of patients (three cases in pressure less than 130 mm Hg. After 12 months, 28 partici-
LMGB group and three cases in LSG group) [112]. pants (49 %; 95 %CI: 36 %–63 %) in the gastric bypass group
In 2012, Schauer et al. [113] randomized 150 patients with and 11 (19 %; 95 %CI: 10–32 %) in the lifestyle-medical
T2DM and a BMI range of 27–43 (34 % of patients with BMI management group achieved the primary end points (odds
< 35) to Roux-en-Y gastric bypass (LRYGB), sleeve ratio: 4.8; 95 %CI: 1.9–11.7). Participants in the gastric bypass
504 OBES SURG (2014) 24:487–519

group lost 26.1 vs. 7.9 % of their initial body weigh compared Reis et al. [129] recently conducted a literature review on
with the lifestyle-medical management group (difference: the role of bariatric–metabolic surgery in the treatment of
17.5 %; 95 %CI: 14.2–20.7 %). There were 22 serious adverse obese type 2 diabetes with body mass index < 35 kg/m2. A
events in the gastric bypass group, including one cardiovas- total of 29 studies, with 1,209 class I obese T2DM patients,
cular event, and 15 in the lifestyle-medical management were included. Twelve studies were from Brazil, five from
group. There were four perioperative complications and six Italy, two from the US, two 2 from China, two from South
late postoperative complications. The gastric bypass group Korea, two from Chile, and one from Australia. Effects of
experienced more nutritional deficiency than the lifestyle- laparoscopic ileal interposition were evaluated in nine studies,
medical management group [114]. laparoscopic duodenal–jejunal bypass in five studies, laparo-
scopic gastric bypass in five studies, biliopancreatic diversion
in four studies, laparoscopic adjustable gastric banding in
Metanalysis and Systematic Reviews in Patients with Type 2 three studies, laparoscopic mini gastric bypass in two studies,
Diabetes and laparoscopic sleeve gastrectomy (LSG) in one study. In
the pooled analysis, BMI was reduced from 29.9 to 24.8 kg/
Li et al. [115] recently published a large meta-analysis of m2 (p<0.001). Fasting plasma glucose was reduced from
prospective and retrospective non randomized studies on the 207.8 to 113.5 mg/dl (p<0.001) and Hb1Ac from 8.9 to
metabolic effects of bariatric surgery in T2DM patients with 6.3 % (p<0.001). The withdrawal of T2DM medications
class I obesity. A total of 13 studies, including 357 patients, was obtained in 84 % patients [129]. A more extended series
were systematically evaluated. Both traditional and experi- of patients treated with stomach-sparing duodenal–jejunal
mental bariatric/metabolic procedures were included. Laparo- bypass has been recently published [130].
scopic Roux-en-Y gastric bypass was performed in four stud- In summary, the two systematic reviews [115, 129] evalu-
ies [116–119], duodenal–jejunal bypass in three studies ated 13 and 29 nonrandomized prospective and retrospective
[120–122], bilio-pancreatic diversion in three studies studies, with 359 and 1,209 diabetic patients, respectively.
[123–125], laparoscopic mini-gastric bypass in two studies Both traditional (Gastric Banding, Gastric Bypass, Sleeve
[126, 127] and laparoscopic ileal interposition with diverted Gastrectomy, Biliopancreatic Diversion) and experimental
sleeve gastrectomy in one study [128]. Nine studies were (Duodeno Jejunal Bypass and Ileal Interposition) procedures
prospective and four studies were retrospective. Five studies were included. In the first meta-analysis, weight loss, diabetes
were conducted in Brazil, three in Taiwan, three in Italy, one in remission and improvements in lipids and metabolic syn-
India, and 1 in the US. The principal characteristics and the drome were analyzed and the results were judged to be as
main results of the 13 studies included in this meta-analysis good as in morbid obese patients for all the four parameters
were reported in Table 4. Follow-up length ranged from 6 to [115]. Perioperative 30 days, complications rate was 3.2 % in
48 months for 11 studies, while the remaining two studies the total series. Different complications rates were observed in
lasted more than 5 years, even 18 years in 1 study. The median different operations, but class I obesity surgical complications
duration of follow-up was 26.8 months. Total weight loss, as rate was considered, in general, lower than the complication
derived by the meta-analysis of the five studies reporting this rates observed in morbid obesity [115]. The second meta-
outcome, was 17.23 kg (p<0.00001). Mean BMI reduction analysis focused diabetic remission as the primary endpoint
(data reported in 12 studies) was 5.18 kg/m2 (p<0.00001). and reported for this specific outcome good results, as ob-
Resolution of diabetes was defined as a normal fasting plasma served in morbid obesity [129]. Further systematic reviews
glucose (<100 mg/dl), a normal HbA1c (<6 %), and no need and meta-analysis on the topic have been recently published
for diabetic medications. The majority of patients (80 %) [131, 132], and others will probably appear in the next future
reached a HbA1c value <7 % and these patients were off as new evidences are rapidly accumulating.
T2DM medications. Mean reduction in fasting plasma glucose
levels was −4.4 mmol/L (p<0.00001) in 12 studies and mean Prospective Observational Studies and Retrospective Studies
reduction of HbA1c was 2.59 % (p<0.00001) in 11 studies.
These important effects on glucose metabolism were accom- In 2009, Flum et al. [133] evaluated the 30-day outcomes in
panied by a significant reduction in LDL-cholesterol levels 623 patients with BMI 30–40 sorted out from 4,776 consec-
(−36.7 mg/dl in four studies; p<0.00001) and triglycerides utive patients undergoing bariatric surgical procedure at ten
levels (−56.7 mg/dl in five studies; p<0.00001) and by a clinical sites in the US from 2005 through 2007. Most of the
nonsignificant increment in HDL-cholesterol levels (+ procedures were represented by Roux-en-Y Gastric Bypass
5.37 mg/dl in six studies; p=0.08). Perioperative (<30 days) (GBP) (71.4 % of cases; laparoscopic GBP 87.2 % and open
serious adverse events were recorded in 11 cases (3.2 % of GBP 12.8 %) and by Laparoscopic Gastric Banding (LGB)
patients). None resulted in late complications of the whole (25.1 % of cases), with 3.5 % of the procedures represented by
series [115]. other techniques, mainly laparoscopic sleeve gastrectomy
Table 4 Principal characteristics and main results of the prospective and retrospective non randomized studies included in the Li et al. [115] meta-analysis on the metabolic effects of bariatric surgery in
T2DM patients with class I obesity

Reference Pts no. BMI Type FU length FU rate Main results

range

Roux-en-Y gastric bypass

Shah et al. [116] 15 22–35 Prospective observational 9 months 100 % BMI from 28.9 to 23.0. Fasting glucose from 233 to 89 mg/dl. HbA1c from 10.1 to 6.1 %.
OBES SURG (2014) 24:487–519

Off T2DM medication in 100 % of patients. Improvements in lipids and metabolic syndrome
Huang et al. [117] 22 25–35 Prospective observational 12 months 100 % BMI from 30.8 to 23.7. Fasting glucose from 204 to 103 mg/dl. HbA1c from 9.2 to 5.9 %.
Off T2DM medication in 90.9 % of patients
De Sa et al. [118] 27 30–35 Retrospective observational 20 months 100 % BMI from 33.5 to 25.6. Fasting glucose from 176.1 to 93.9 mg/dl. HbA1c from 8.4 to 5.9 %.
Off T2D medication in 74 % of patients
Cohen et al. [119] 66 30–35 Prospective observational 5 years 100 % Total weight loss: 36 %. T2DM remission: 88 %, HbA1c from 9.7 to 5.9 %. Improvements in
lipids and metabolic syndrome
Duodenal–jejunal bypass
Ferzli et al. [120] 7 21–33 Prospective observational 12 months 100 % BMI loss not recorded. Fasting glucose from 208 to 154 mg/dl. HbA1c from 9.4 to 8.5 %.
Off T2DM medication in 14.2 % of patients. Reduction T2DM medication in 85.8 % of
patients. Improvements in lipid and metabolic syndrome
Geloneze et al. [121] 12 25–29.9 Prospective nonrandomized 24 weeks 100 % Surgical patients compared to 162 medically treated patients. HbA1c from 8.7 to 7.4 % in
case control surgical, 8.9 to 8.7 % in control group. Reduction in insulin therapy in 93 % of surgical
group and 29 % in control group
Ramos et al. [122] 20 20–30 Prospective observational 6 months 100 % BMI from 27.1 to 24.4. Fasting glucose from 171.3 to 107.1 mg/dl. HbA1c from 8.8 to 6.8 %.
Off T2D medication in 90 % of patients
Biliopancreatic diversion
Scopinaro et al. [123] 7 32–34.6 Retrospective observational 13 years 100 % Body weight from 92 to 75 kg. 100 % of patients <125 mg/dl fasting glucose at 3 years.
(10–18 years) Off T2DM medication in 100 % of patients. Improvements in lipid and metabolic syndrome
Chiellini et al. [124] 5 27–33 Prospective observational 18 months 100 % BMI from 30.9 to 25.1. HbA1c from 8.5 to 5.7 %. Off T2DM medication in 100 % of patients.
Improvements in lipids and metabolic syndrome
Scopinaro et al. [125] 30 25–30 Prospective controlled 100 % BMI from 28.1 to 24.6 (25–30) and from 33.1 to 27.4 (30–35). HbA1c from 9.1 to 6.9 % (25–30)
and and from 9.5 to 5.9 % (30–35). Off T2DM medication in 83.3 % of patients. Improvements in
30–35 lipid and metabolic syndrome
Mini gastric bypass
Lee et al. [126] 44 BMI < 35 Retrospective observational 12 months 100 % Fasting glucose reached normal range in 89.5 % of patients. HbA1c < 7 % in 76.5 % of patients.
Off T2DM medication in 90 % of patients. Improvements in lipids
Lee et al. [127] 62 23–35 Retrospective observational 2 years 100 % BMI from 30.1 to 23. Fasting glucose from 195 to 106 mg/dl. HbA1c from 9.7 to 5.9 %.
Off T2DM medication in 55 % of patients
Ileal interposition and diverted sleeve gastrectomy
De Paula et al. [128] 69 21–29 Prospective observational 22 months 100 % BMI from 25.7 to 21.8. Fasting glucose from 218 to 102 mg/dl. HbA1c from 8.7 to 5.9 %.
Off T2DM medication in 95.7 % of patients. Improvements in lipid and metabolic syndrome
505
506 OBES SURG (2014) 24:487–519

(LSG) and biliopancreatic diversion (BPD). In this large pro- gastrectomy (SG) in seven patients, biliopancreatic diversion
spective multicenter observational study, perioperative com- (BPD) in one patient and other surgical operations in nine
plications were recorded in 23 out of 623 patients with BMI patients. Results at 1 year were evaluated and compared in the
30–40 (3.7 %), with no clear differences in adverse event rate 109 AGB patients and the 109 GBP patients. One year follow-
comparing bariatric surgery in patients with BMI 30–40 and up was 62 % for AGB patients and 69 % for GBP patients.
surgery in patients with BMI, with the exclusion of super- BMI levels were reduced from 33.9 to 30.9 kg/m2 for AGB
obese patients having a clearly higher adverse event’s rate. In (p<0.001) and from 33.7 to 27.1 kg/m2 for GBP (p<0.0001).
the general sample, predictors of adverse events were the Off 2DM medication was observed in 27.5 % for AGB
presence of obstructive sleep apnea, a positive history for deep patients (p<0.05) and in 55.2 % for GBP patients (p<0.05).
vein thrombosis and an older age. Patients treated with GBP Complication rate was 3.3 % in AGB patients and 18 % in
reported more complications than patients treated with LGB. GBP patients [134].
However, specific procedure-related adverse events were not The results of other smaller and/or single site prospective
separately reported for patients with class I obesity and pa- observational studies conducted in patients with class I obesity
tients with class II obesity [133]. [135–139] are summarized in Table 5 and the results of some
A second large retrospective review multicentre study was retrospective studies specifically analyzing the outcome of
published in 2010. De Maria et al. [134] evaluated the data class I obese patients sorted out from general bariatric surgery
from 66,264 patients with a primary bariatric surgery proce- series [140–144] are reported in Table 6.
dures belonging to the Bariatric Outcomes Longitudinal Da- All the observational studies report satisfactory weight
tabase (BOLD). BOLD was created by Surgical Review Cor- loss, resolution or improvements of type 2 diabetes mellitus,
poration as a tool to monitor and track outcomes of surgeries improvements in lipids and metabolic syndrome. The details
performed by participants of the American Society for Meta- regarding specific parameters were variable. Several studies
bolic and Bariatric Surgery (ASMBS), Bariatric Surgery Cen- focused primarily on the effects of surgery on type 2 diabetes
tre of Excellence (BSCOE) program. Patients’ recruitment and evaluated remission rates based on different definitions.
started on June 2007 and ended on June 2009. Class I obesity However, all the observational studies reported positive ef-
patients with type 2 diabetes patients were specifically con- fects on glycemic control and diabetes remission rate. These
sidered. A total of 235 patients met these inclusion criteria, effects were higher in patients with higher BMI. Other cardio-
with adjustable gastric banding (AGB) performed in 109 vascular risk factors, when reported, improved after bariatric
patients, gastric bypass (GBP) in 109 patients, sleeve surgery. The systematic reviews and the observational studies,

Table 5 Principal characteristics and main results of prospective observational single site studies on the role of bariatric surgery in patients with class I
obesity

Reference Pts N. BMI range Procedure FU length FU rate Main results

Parikh 93 25–30 Adjustable 8 years 89 % %EWL: 53.8 at 3 years. BMI from 32.7 to 27.2 at
et al. [135] gastric banding 3 years. Significant improvements in lipids and
metabolic syndrome
Sultan 53 28–35 Adjustable 2 years 81 % %EWL: 69.7. BMI from 33.1 to 25.8. Off T2DM
et al. [136] gastric banding medication in 50 % of patients. Significant
improvements in lipids and metabolic syndrome.
Major and minor overall complications rate: 13.2 %
Choi 66 30–40 (30–35 if Adjustable 18 months 100 % %EWL: 42.2. T2DM resolution /improvement in
et al. [137] comorbidities) gastric banding 33.3 % of patients. Significant improvements in
lipids and metabolic syndrome
Kakoulidis 23 30–35 Sleeve gastrectomy 6 months 100 % %EWL: 100 %. BMI from 33.8 to 25.0. Significant
et al. [138] improvements in lipids and metabolic syndrome.
Quality of life excellent in 50 % of patients, very
good in 16.6 %, good in 25 %, fair in 4 % and
poor in 0 %.
Abbatini 9 30–35 Sleeve gastrectomy 1 year 100 % 9 surgical patients compared with 9 patients treated with
et al. [139] conventional medical therapy. BMI from 32.7 to 21.1
in LSG group, and from 32.9 to 31.7 in medical group.
HbA1c from 8.1 % to 5.9 % in LSG group, and from
7.5 to 8.2 in medical group. Off T2DM medication in
89 % of patients in LSG group and 0 % in medical
group. Significant improvements in lipids and metabolic
syndrome in LSG group, no changes in medical group
OBES SURG (2014) 24:487–519 507

Table 6 Principal characteristics and main results of retrospective studies specifically analysing the outcome of class I obese patients sorted out from
general bariatric surgery series

Reference Pts N. Procedure FU length FU rate Main results

Angrisani 225 class I obese patients AGB 5 years 72 % %EWL at 5 years: 71.9 %. BMI from 33.9 to 28.2. T2DM
et al. [140] out of 3319 bariatric remission in 100 %. Comorbidities reduction in 89.1 %
patients of patients. Perioperative complication in 8.1 % of cases.
Mortality in 1 case (0.4 %).
Serrot 17 patients with class I GBP 1 year 100 % 17 surgical patients compared with 17 patients treated with
et al. [141] obesity and T2DM medical therapy. BMI from 34.6 to 25.8 in GBP group and
from 34 to 34.3 in medical group. HbA1c from 8.2 % to
6.1 % in GBP group and from 7.0 to 7.1 % in medical group.
T2DM medications reduced in 71 % of patients in GBP
group and in 6 % of patients in MT group. Significant
improvements in lipid and metabolic syndrome.
Frenken 16 patients with BMI BPD-DS in 7 pts. 1 year 94 % HbA1c from 8.8 % to 5.6 % for BPD and BPD-DS and from
et al. [142] 26–34.5 and T2DM BPD in 5 pts. 7.8 % to 6.7 % for GBP.
GBP in 4 pts.
Gianos 42 patients with class SG in 24 pts. GBP 14 months 100 % BMI from 33.9 to 26.5. T2DM resolution/improvement in
et al. [143] I obesity in 8 pts. AGB in 68 % of patients. Significant improvements in lipid and
10 pts. metabolic syndrome.
Angrisani 34 patients with class AGB 7 years 100 % %EWL: 70.9 % at 7 years. BMI from 32.6 to 27.4. T2DM
et al. [144] I obesity medications reduced in 100 % of patients at 3 years.
Significant improvements in lipid and metabolic syndrome.

AGB adjustable gastric banding, GBP gastric bypass, BPD-DS biliopancreatic diversion-duodenal switch, BPD biliopancreatic diversion, SG sleeve
gastrectomy

prospective or retrospective, contain however several meth- other studies with small numbers of patients who received
odological deficiencies, related to the different study designs sleeve gastrectomy reported significant weight loss in class I
with different kinds of surgery, number of patients, lengths of obesity patients [128, 139, 143]. Schauer et al. [113] reported
follow-up, primary and secondary endpoints. Weight loss and in their RCT a 88 % EWL at 1 year in patients underwent to
rate of complications were considered in studies with different gastric bypass. Data collected from the BOLD registry includ-
operations. Some studies report lack of control data, propen- ed 109 RYGB patients with BMI < 35 kg/m2 who had 69 %
sity to bias, and lack of information. There was also variability EWL 1 year after surgery [134]. Other studies of weight loss
in the method of weight and co-morbidity reporting. after RYGB in this patient population consistently report BMI
reduction 12–36 months after surgery [116–119, 141–143].
Final Summary Lee et al. [112], in a RCT study, reported 94 % EWL in
patients underwent to mini gastric bypass at 1 year. Two other
The comprehensive evaluation of the randomized control studies with patients who received mini gastric bypass with a
trials, meta-analysis and prospective or retrospective studies follow-up of 12–24 months reported significant weight loss
included in this short overview demonstrated that overall [126, 127]. Finally, three studies reported excellent weight
%EWL was excellent in patients with class I obesity after all loss after biliopancreatic diversion or duodenal switch in
the most established bariatric procedures, with the majority of patients with BMI < 35 kg/m2. Importantly, patients did not
the studies reporting no substantial differences in respect to have excessive weight loss after these procedures, and weight
the weight loss observed in patients with morbid obesity stabilized at a BMI around 25 kg/m2 within a year after
meeting the current BMI criteria. In studies considering gas- surgery [123–125].
tric banding operation, EWL at 2 years was 87.2 % in the RCT In the five RCTs, in systematic reviews and in the obser-
reported by O’Brien et al. [110], Parikh et al. [135] reported a vational studies above considered, resolution or improve-
57.9 % EWL at 1 year and a 53.8 % EWL at 3 years, Sultan ments of type 2 diabetes and various comorbidities were
et al. [136] reported a 69.7 % EWL at 2 years. Finally, the analyzed. The level of detail regarding specific comorbidities
Italian Collaborative study with gastric banding reported a was variable within studies, limiting the generalizability of the
71.9 % EWL at 5 years [140]. Kakoulidis et al. [138] reported results. Several studies focused primarily on the effects of
a 100 % EWL in 23 patients who had reached 6 months of surgery on type 2 diabetes, but evaluated remission rates using
follow-up after sleeve gastrectomy. Schauer et al. [113] re- variable definitions. However, in general, positive effects on
ported in their RCT a 81 % EWL at 1 year after laparoscopic glycemic control and diabetes remission rates were reported,
sleeve gastrectomy and Lee et al. [112] 76 % EWL. Three as previously shown in patients with higher BMI levels. The
508 OBES SURG (2014) 24:487–519

true effects of surgery on glycemic control and diabetes re- age, ethnicity, and disease risk, is not a reliable and fair
mission in patients with type 2 diabetes and class I obesity approach to the denial of surgery to patients for whom this is
may be at present underestimated taking into account the fact the only effective treatment. Instead, this decision should be
that patients with maturity onset diabetes of the young guided, as for other diseases, by the patients’ states of health
(MODY), a group of inherited forms of beta-cell defect, and and the risk/benefits of the operation.
Latent Autoimmune Diabetes in the Adult (LADA), a late-
onset form of type 1 diabetes, have a low probability to remit
after surgery and have not been adequately screened and Special Considerations Regarding Patient Selection
excluded before surgery in some studies [145, 146]. Metabolic
syndrome, when reported, improved after bariatric surgery. Ethnicity
Adverse event’s rate in class I obese patients appears to be
the same than in morbid obesity, with some studies reporting BMI categories have been developed primarily in populations
serious adverse events. In the review of the BOLD database of mainly European ethnicity and often underestimate health
including 109 RYGB and 109 LAGB patients with BMI < risks in other populations. The risk and expression of meta-
35 kg/m2, complications rate was 18 % after gastric bypass bolic syndrome features, and the risk of developing type 2
and 3.3 % after gastric banding (p<0.05). Most complications diabetes, vary with ethnicity [147]. Ethnicity rather than the
were minor (nausea, vomiting), but serious complications, country of residency is important, as often, obesity rates are
including anastomotic leakage, intraabdominal bleeding, and higher for those of high-risk ethnicity when living in devel-
internal hernia, were reported in the gastric bypass group. One oped rather than developing countries. Adjusted BMI action
gastric banding patient developed a band slippage in this cut points for with Asian or other high-risk ethnic groups are
review [134]. In the Italian Collaborative study that retrospec- recommended to be reduced by 2.5 kg/m2 to BMI 27.5, 32.5,
tively considered several class I obese patients treated with and 37.5 kg/m2, respectively [7, 148] (Table 7).
banding operated, one patient died 20 after surgery from
sepsis after gastric perforation in association with a dilated Age
gastric pouch. This study also reported a 8 % of late compli-
cations requiring reoperation for proximal gastric pouch, band Extremes of age present specific challenges when considering
erosion and leakage of the port [140]. bariatric–metabolic surgery in those with a BMI < 35 kg/m2.
Quality-of-life data were seldom reported in the studies Bariatric metabolic surgery is only generally considered suit-
included in this overview. However, quality of life was mea- able for adolescents of developmental and physical maturity
sured using the Short Form Health Survey (SF36) in one RCT who are severely obese. Several position statements from
and the patients treated with gastric banding group had signif- Europe, the US, and Australia have emerged over the last
icant improvements in all the eight domains of the SF-36, with decade and all made similar recommendations for suitable
significantly greater improvement than the nonsurgical group BMI, generally following traditional adult criteria of BMI >
for physical functioning, vitality, and mental health [110]. 40 kg/m2 or BMI > 35 kg/m2 with severe co-morbidities
Kakoulidis et al. [138] also reported good or excellent quality (including type 2 diabetes) [68, 149, 150]. While statements
of life in 22 of 23 patients 6 months after sleeve gastrectomy. have varied in a minor way with youngest age and BMI,
Major limitation of current data on the use of bariatric position statements have yet to recommend lowering the
surgery in patients with class I obesity is the short length of BMI to below 35 kg/m2. This would appear to be in line with
follow-up in most of the studies. All the RCTs and most of the the principle of establishing efficacy, safety, and broad accept-
observational studies were shorter than 2 years in follow-up ability in adults before extending indications in children and
and a more extended follow-up was available only in some adolescents and with lack of data about efficacy and safety of
prospective or retrospective uncontrolled studies. This prob- surgery in class I obese adolescents.
lem limits our knowledge about the long-term risk / benefit There are important considerations with increasing age as
ratio of surgery in this subset of patients. In particular, poten- the effect of obesity on morbidity and mortality is attenuated
tially serious effects of the profound weight loss produced by and the NADIR for the optimal BMI with respect to mortality
surgical procedures on nutritional status and body composi- is in the overweight to class I obese range [17]. The optimal
tion (loss of muscle mass and sarcopenia) cannot be evaluated. weight for lowest mortality appears to be between 25 and
Finally, reliable information about the effects of bariatric 35 kg/m2 for those with an age of 70 years and older [17].
surgery on longevity in patients with class I obesity remains The effect of more severe forms of obesity on mortality after
completely lacking. the age of 65 years is low. These effects are not restricted to the
In conclusion, this review documents the effectiveness of healthy older adults, but are similar in those with diabetes and
bariatric surgery for patients with BMI 30–35. As outlined established cardiovascular disease. Weight loss of 10 % in
above, the BMI, with its failure to account for gender, fitness, obese older patients can reduce functional capacity and
OBES SURG (2014) 24:487–519 509

Table 7 The classification of weight category by BMI

For Asian populations classifications remain the same as the international classification but that public health action points for interventions are set at 23,
27.5, 32.5, and 37.5. We address eligibility and prioritization for bariatric surgery within the different gray shadowing. Source: Adapted from WHO [7, 148].

mobility [151]. Weight maintenance irrespective of BMI and bariatric metabolic services to those of higher BMI, where
improved fitness may be the appropriate focus in older adults. the proven benefits of reduced mortality, improved quality of
If intentional weight loss is desired than modest weight loss in life, and favorable health economic profile are established [14,
association with exercise provides the best functional out- 153]. Those who may be considered prioritized for bariatric
comes [152]. There is no clear guidance regarding intentional metabolic surgery, for example, individuals with a BMI >
weight loss in older adults as it is unclear that benefits out- 50 kg/m2 or type 2 diabetes with a BMI > 40 kg/m2, are not
weigh risks [153]. Weight loss trajectories in older people are provided access to surgery [35, 71]. Surgery is less likely to be
associated with considerable risks of both morbidity and cost-effective in individuals with class I obesity [156].
mortality [154], and while much research is needed into
weight loss and weight gain in the later years of life, major Comorbidity
weight loss in older adults with a BMI < 35 kg/m2 cannot be
currently recommended. Metabolic, mechanical, and psychological comorbidity of
obesity often cluster and are associated with increased risk
Regional, Economic, and Equity Considerations of morbidity and mortality that is poorly related to BMI [100,
101]. Staging systems may provide a useful way of identifying
There are regional variations in access, broad uptake, and type individuals of greatest risk and allow appropriately targeted
of bariatric metabolic surgery performed [155], and there are extension of bariatric metabolic surgery into the BMI<35 kg/
also regional differences in the regulatory and economic con- m2 range; for example, the International Diabetes Federation
ditions that may limit the direction of surgery for patients with has recommended for some circumstances for individuals
a BMI < 35 kg/m2. Economic issues are a particular problem with type 2 diabetes [71]. Caution needs to be considered
in emerging countries where rates of obesity and metabolic when evaluating each individual’s comorbidities and their
disease including diabetes are increasing rapidly and health likely response to bariatric–metabolic surgery in relation to
care resources limited. Lowering the BMI threshold is likely how established therapies treat their conditions. Hypertension
to alter the overall risk to benefit and influence the health and raised LDL cholesterol levels respond well to pharmaco-
economics of bariatric metabolic surgery [156]. National and logical agents and variably to surgery [20], but in combination
regional health services providers need to consider the evi- with other comorbidity, such as type 2 diabetes, nonalcoholic
dence and deliver services that are locally appropriate. steato-hepatitis (NASH), knee osteoarthritis, or obstructive
Issues of equity of access to surgery are strongly influenced sleep apnea, the pendulum may swing to bariatric surgery
by socioeconomic circumstances. In the developed world, being added to, and possibly replacing, conventional thera-
obesity and its related metabolic conditions are more common pies. In many circumstances, we need higher-quality evidence
in the socioeconomically disadvantaged, but the majority of for the effect of bariatric–metabolic surgery on comorbidity
bariatric procedures are performed in the private sector, gen- changes in patients with class I obesity. In those with severe
erating inequity and discriminating against individuals who obesity, the imperative for substantial weight loss with bariatric
are most likely to benefit [71]. National health services pro- surgery has been the major focus, and the range of accompa-
viding for all citizens are struggling to currently provide nying benefits substantial and welcome. But for those with class
510 OBES SURG (2014) 24:487–519

I obesity, the changes in comorbidity and hard health outcomes (laparoscopic access to the abdominal organs), less complex
will take center stage, and bariatric–metabolic surgery will have procedures (adjustable gastric banding and sleeve gastrecto-
to compete with established therapies for each comorbidity my), and a growing body of literature that demonstrates bar-
rather than ride of the coat tails of major weight loss. iatric surgery results in the improvement of several comorbid
conditions in patients with BMI greater than 35 kg/m2 has
Low BMI as a Consequence of Previous Medical or Surgical fueled this interest [158, 159]. There is a slowly growing body
Therapy of literature demonstrating similar benefits for lower BMI
patients [115, 129–132]. Additionally, there is an expanding
As previously specified in the inter-disciplinary European experience with nonconventional procedures in this popula-
guidelines [68], BMI criterion for election to bariatric meta- tion to target diseases such as diabetes [65, 160].
bolic surgery should be the current BMI or a documented Before the widespread of acceptance of conventional and
previous BMI of this severity. This means that weight loss nonconventional operative procedures occurs, ethical due
as a result of intensified treatment before surgery (patients diligence must occur. Unfortunately, that was not always the
who reach a body weight below the required BMI for surgery) is case. Patients have been subject to surgery outside of the
not a contraindication for the planned bariatric surgery and that accepted norm often without Institutional Review Board
surgery is indicated in patients who exhibited a substantial weight (IRB) approval, comprehensive informed consent, and proper
loss in a conservative treatment program but started to gain investigational behavior. More concerning, novel procedures
weight again [68]. Similar considerations should be applied to were performed on human subjects without adequate preclin-
bariatric patents having reached a low BMI after a first interven- ical investigation. Published studies are often the result of
tion, but requiring redo surgery for complications or side effects. small observational investigations with adequately small
study populations, no control or sham groups, and short
follow-up. Due to a multitude of limitations, randomized,
Research Gaps and Priorities prospective, trials are few. The observational trials are often
prone to weak methodologies, subject to investigator bias, or
Introduction conflicts of interest [161]. Risk adjustment to allow more
meaningful outcome analysis has rarely been performed.
The currently accepted thresholds for performing bariatric
surgery were established in 1991 by the U.S. National Insti- Long-Term Outcomes
tutes of Health (NIH) [67]. The NIH assembled an “expert
panel” that reviewed the prevailing literature to make their One of the biggest deficiencies of the prevailing literature
recommendations. The data they assessed were predominately concerning both conventional and nonconventional surgical pro-
published in the 1980s. At the time the only operative proce- cedures for BMI < 35 kg/m2 is the lack of long-term outcome
dures performed were the gastric bypass and the vertical data. Most of the published results reported are 12 months or less.
banded gastroplasty and the only method to perform these This limited follow-up does not prove that the benefits of these
procedures was through a long midline incision (open tech- surgical procedures are durable and does not account for conse-
nique). The NIH Consensus Development Statement conclud- quences of these procedures, such as nutritional deficiencies, that
ed that bariatric surgery should only be considered for patients may occur years after the surgery was performed. It is also highly
with a BMI of 40 kg/m2 (35 kg/m2 if the patient suffered from conceivable that, in some cases, the weight loss and other be-
comorbid conditions such as type 2 diabetes or hypertension) nefits may decline with time. Reis et al. [129] did an extensive
[67]. This proclamation was adopted by private health insur- literature search for published articles that evaluated the effects of
ance providers and society at large and become the rules of bariatric surgery on patients whose BMI < 35 kg/m2 and had type
conduct for performing bariatric surgery. Unfortunately, it 2 diabetes. There were 29 articles selected. Follow-up was as
prevented patients whose BMI was 30–35 kg/m2 from quali- short as 1 month and as long as 60 months. 41 % of the studies
fying for surgery even if they suffered from comorbid condi- had 12 months or less follow-up. Of these, 42 % had
tions. As a consequence, there was little interest at that time in 6 months or less of follow-up. Additionally, only 24 %
pursuing surgery for those patients. of the studies had any patient follow up beyond
However, in the present, the issue of whether it is appro- 36 months. This phenomenon is even more significant
priate to offer patients with BMI less than 35 kg/m2 has in the device and novel procedure literature, where few
developed considerable interest. It is now well understood published papers report follow-up beyond a few months
that patients whose BMI is 30–35 kg/m2 are likely to suffer and rarely past 12 months [65, 160, 162, 163].
from the same comorbid conditions as patients with higher While there is no defined standard for adequate follow-up,
BMIs and also are at risk for premature death [157]. Addi- in bariatric surgery, it should be longer than 12 months. Recent
tionally, the introduction of less invasive techniques studies have suggested that the relapse of type 2 diabetes after
OBES SURG (2014) 24:487–519 511

remission after gastric usually occurs within 5 years of surgery complication profiles. Since dramatic improvements in health
[164]. Similarly, reactive hypoglycemia also occurs 3 years after and well-being can be realized with as little as a mere 10 %
gastric bypass. Ritz and Hanaire [165] reviewed the 89 published body weight loss, new procedures can be deemed successful
cases of severe postoperative hypoglycemia. The time to symp- at significantly less weight loss than the current mainstream
toms varied from 6 to 264 months with a mean of 28.6 months. operations. Additionally, those criteria should be flexible and
Nutritional deficiencies and conditions, such as osteoporosis, able to change as new information is obtained. Lerner et al.
may even take longer. Given these examples, it would be rea- [166] has suggested that if the safety profile exceeds beyond
sonable to suggest that adequate postoperative follow-up for the expected, the efficacy endpoints could be lowered. Converse-
sake of investigational data collection and procedure evaluation ly, if over time, there is an emergence of additional health
should be no less than 3 years and preferably 5 years. consequences, then the acceptable minimum efficacy thresh-
old should be raised accordingly.
How to Assess New Procedures, Devices, and Techniques Traditionally in bariatric surgery, the amount of weight lost
was the sole measure of a procedures success. For decades,
All new procedures, devices, and techniques mandate honest, criteria, such as that developed by Rheinhold (that a successful
thorough, and rigorous assessment before being offered to procedure must result in a 50 % or greater excess weight loss),
patients. Patient safety must be the first priority and risks was used as the litmus test for a procedure’s success [167].
minimized [166]. This includes not only early postoperative However, there are several deficiencies when an outcome mea-
complications but also long-term sequelae. All new surgical sure is primarily focused on weight loss. Firstly, the 50 % excess
interventions must first and foremost demonstrate a favorable weight loss milestone is an arbitrary one at best. There was no
risk/benefit profile. Therefore, for any given degree of risk, the scientific analysis performed. There may be no metabolic differ-
potential for benefit must be on balance, superior. While the ence between one patient who loses 52 % of excess weight and
interpretation of a “favorable risk-to-benefit ratio” is variable, another who loses 48 %. Yet, in this one-dimensional system, the
it should be defined reasonably and free of bias. Any noncon- patient with a 48 % excess weight loss who has improvements in
ventional procedure must be subjected to the appropriate health, ambulatory ability, and quality of life would be classified
scientific analysis and prove to be safe and effective. This as a failure. Secondly, a weight loss onlybased system would
analysis should include a sufficient number of test subjects, a favor patients at lower baseline body weight as they would lose a
sound scientific method, correct use of statistics, adequate greater percentage of their body weight than their much heavier
patient follow-up, and appropriate primary and secondary counterparts. Lastly, a weight loss onlybased system would
endpoints. To minimize harm, new procedures should under- discriminate against less radical procedures. Operations, such
go extensive preclinical investigation. This would mostly as the gastric bypass, can achieve 50 % or greater excess weight
likely require evaluation in a representative animal model. loss because of their extreme restriction to food intake. However,
After the demonstration of efficacy and safety, the procedure some of the novel procedures under development do not rely on
should be rigorously evaluated in clinical human trials. For all radical dietary restriction to achieve results. Some only induce
of these trials, the study design must be carefully conceived to early satiety as the mechanism of weight loss. As eating and
result in the maximal amount of information while minimizing calorie intake is only partly related to actual appetite and is
patient risk. All study protocols should be submitted to the local greatly influenced by other factors such as mood, actively, cul-
Institutional Review Board (IRB), and all patients should agree ture, and the environment, it would be highly unlikely that such a
to participate by signing an IRB approved informed consent procedure will universally achieve such weight loss. However,
[99]. A small open-label feasibility trial in a limited number of procedures that result in modest weight loss and improvements
test subjects should be performed first. If successful, larger- in comorbid conditions with a favorable risk/benefit profile can
scale investigations, possibly multisite, with adequate numbers also be viewed as successful.
of subjects and sufficient follow-up should be undertaken.
Whether a randomized sham-controlled trial is feasible will Reporting Weight Loss Outcomes
depend, in part, on the procedure being evaluated.
It is neither appropriate nor scientifically sound to judge all Throughout the early history of weight loss surgery, there was
new procedures by one set of universal standards such as no consensus as to how to report postoperative weight loss.
weight loss or comorbidity improvements. Each procedure Published papers early on reported weight loss as total pounds
or device will have different safety profiles, degree of com- or kilograms lost, or excess pounds or kilograms lost. How-
plexities, and outcome results. Therefore, each should be ever, these outcome measures are meaningless without taking
judged by its own set of criteria. For example, procedures that the preoperative baseline weight into account. For example, a
are less radical, less complex, and/or less risky for the patient, 50 kg weight loss might be quite significant in a patient whose
can be acceptable even if they result in significantly less preoperative weight was 150 kg but not so impressive if the
benefit than more complex procedures that have higher starting weight was 250 kg. Similarly, reporting weight loss as
512 OBES SURG (2014) 24:487–519

excess kilograms lost maybe equally as meaningless. There is disease state. For example, it is assumed that all morbidly
no standard for how to determine “excess weight.” It has obese patients with adult onset diabetes are therefore “type 2.”
traditionally been calculated from an insurance actuarial table However, that may not necessarily be the case. Some of these
that defines ideal body weight for men and women of all patients may very well have uncommon variants of type 1
heights [168]. However, the determination of “ideal weights” diabetes mellitus such as latent autoimmune diabetes [113,
was derived from actuarial data, not physiologic investigation. 145, 146]. Since type 1 and type 2 patients will respond
It also does not account for heterogeneity of body composi- differently to surgical and medical intervention, it is of utmost
tion. Therefore, its value for assessing procedure outcome is importance for valid clinical research to ensure that all patients
generally considered dubious. in a surgical trial evaluating the effects on type 2 diabetes be
A very popular style for reporting weight loss was as a truly type 2 diabetics. Furthermore, there needs to be unifor-
percentage of the excess weight lost [169]. Unfortunately, it mity in the chemical markers used to label a patient with
too requires the use of an “ideal body weight” determination suffering from a particular disease or not. In the Stampede
to calculate percent excess weight loss and does not take into Trial, the definition of diabetes was a glycated hemoglobin
account body composition [170]. Although clinically inaccu- >7.0 % and remission was defined as a glycated hemoglobin
rate, it is usually more meaningful than total weight lost in that of 6 % or less [113]. Cohen et al. [119] defined type 2 diabetes
it standardizes the outcome across all different degrees of as having 2 fasting serum glucose results greater than or equal
obesity. However, percent excess weight loss is not always to 120 mg/dl. Other studies do not state how the diagnosis of
meaningful. It also is biased to patients of lesser obesity. For diabetes was made [65, 111].
example, a 250 kg patient may lose 80 kg of weight but only Terminology also needs to be clarified for determining the
40 % of excess, while a 150 kg patient may lose 50 kg which severity of the disease. Currently, hyperglycemic conditions
calculates to 60 % of excess weight. The patient with the have been referred to as “prediabetes,” “glucose intolerance,”
lower baseline weight would appear to have had the better “diet-controlled diabetes,” or “poorly controlled diabetes.”
result, but this may not clinically be the case. More recently, it Hypertension is defined as “mild,” “severe,” or “poorly con-
has been suggested that weight loss outcomes be reported as trolled.” Study subjects with well-controlled conditions are
%BMI units lost [171]. It was even chosen as the official often included in study groups alongside patients who are
language for reporting weight loss outcomes in surgical poorly controlled despite multiple medications. However,
journals. While it does not rely on insurance actuarial tables those subjects may actually represent different diseases and
for its calculations, it does utilize a BMI of 25 kg/m2 as their responses to surgical intervention would likely be differ-
“normal weight” and it does not take body composition into ent. For example, insulin-dependent type 2 diabetics are dif-
account. Like using percent excess weight loss, it may be ferent than diet controlled or those patients well managed on
good for generalizations across patient populations, but has oral agents. Therefore, the creation of universal definitions for
not been scientifically validated for individual patients. Final- the disease and its severity, and the effort to group “like”
ly, percent weight loss with standardization of preoperative patients together, should result in richer outcomes data.
baseline weights was advocated [146]. Belle et al. [172] Criteria also need to be established for the various out-
demonstrated that this method was likely to be more accurate, comes after surgery. Preferably, these criteria should be scien-
taking into account that baseline weights are not only diverse tifically based. The difference between “improvement,” “re-
but also will affect outcome results. Further scientific valida- mission,” or “cure” should be uniform across all clinical
tion should occur before this technique is uniformly adopted. investigations. This is particularly important for the lower
Unfortunately, there is still no scientifically validated or BMI patients where surgical intervention will be more focused
even universally accepted method for measuring and record- on disease than weight. As the likelihood of relapse of the
ing weight loss outcomes. Professional medical societies and disease state exists [164], proclamation that a comorbid dis-
medical journals still differ on the preferred method. As the ease has resolved (cured) might require that the patient be
attention turns away from weight loss to other outcome mea- observed for an extended period of time.
sures such as improvements in disease states, reporting of Another area of concern is the variability in the medical
weight loss may become less important. treatment offered to the control subjects. There is currently a
generous use of the terms “best” or “intensive” medical ther-
Measuring and Reporting Comorbidity Outcomes apy, yet no consensus as to what that truly means. In some
studies, the diabetes management for all patients was con-
Like the reporting of weight loss, universal standard defini- trolled by an endocrinologist involved with the trial and by
tions for comorbidity outcomes need to also be instituted protocol [111, 113, 114]. In other trials, the subjects were
across different patient populations, operative procedures, managed by their health care providers independent of the
clinical practice, and research protocols. First and foremost, study. Both suffer from limitations and biases that can ad-
there needs to be uniform acceptance of the definition of each versely affect the study results. In the former where the study
OBES SURG (2014) 24:487–519 513

manages the medical intervention by protocol, there is unifor- Ethics of Surgery for BMI < 35 kg/m2
mity to the intervention, but it may not represent the best
available treatment for every patient. In the latter, the man- The ethical behavior for studying or treating patients whose
agement is diverse and inconsistent. BMI < 35 kg/m2 by surgical interventions should be no less
Lastly, patient compliance with their medical regimen must rigorous (maybe even more rigorous) than that for any other
also be considered when designing outcome trials and patient group. Although the criteria that excluded patients
interpreting the results of such trials. Patient compliance to whose BMI < 35 kg/m2 from consideration of having bariatric
interventions is extremely variable and may also be influenced surgery is over 20 years old (US National Institutes of Health
by the degree of participation the patient has in their own care. 1991) and likely outdated [67], it remains the generally ac-
While, compliance cannot be mandated, researcher can at- cepted criteria [34, 35]. The body of evidence supporting
tempt to select patients more likely to be compliant and surgery for BMI < 35 kg/m2 is growing [115, 129] and even
monitor the compliance of each subject as the study pro- becoming increasingly supported by national medical socie-
gresses. Additionally, study compliance needs to be recorded ties [99]. Therefore, there is growing debate as to whether it
in the publication and its potential influence on the validity of should still be considered investigational and require an Insti-
the data acknowledged. Currently, few, if any, published out- tutional Review Board approval of the research protocol and a
come studies record and document patient compliance. comprehensive informed consent [173]. While there is an
All of the above factors demonstrate the limitations of overwhelming body of evidence that concludes that bariatric
performing outcomes research in human subjects. These fac- surgery is safe and effective for patients whose BMI ≥ 35 kg/
tors may explain the longstanding observations that several m2, it cannot be assumed that the results would be the same for
seemingly similar studies have resulted in different outcomes. patients with BMIs < 35 kg/m2. These patients may be phys-
Therefore, when designing new research studies, these factors iologically different, and therefore, their response to surgical
should be recognized and every attempt made to minimize intervention is currently not well known. Scopinaro et al.
them. Similarly, when reviewing the results of completed [125] reported that the improvements in type 2 diabetes for
investigations, these factors should be considered when ana- patients with lower BMIs undergoing the biliopancreatic bypass
lyzing the results. procedure was not as robust as for patients at higher weights.
Additionally, there is currently no long-term data in lower BMI
Is There a Need for a Large RCT Looking at Hard Outcomes? patients to validate that the observed efficacy will be lasting.
However, there is little debate over the status of novel
In the realm of clinical research, the most highly regarded metabolic operative procedures and devices. They are still
investigation is the randomized control trial (RCT). Random- investigational and must be treated as such. These new ther-
izing patients to different study groups dramatically reduces apies are currently undergoing study. However, the majority
differences, inequalities, and biases between study and control of the published results are from small open-label trials or
subjects. However, while RCTs are relatively common in limited duration. The few RCTs have thus far yielded modest
pharmaceutical trials, RCTs are difficult to conduct in the field results [65, 174, 175]. The majority of the research was also
or bariatric surgery and large long-term RCTs present formi- conducted in patients BMI > 35 kg/m2 not less than 35 kg/m2.
dable challenges. Only a handful can be found in the literature. While there is insufficient data in the higher BMI patients,
Patients rarely will agree to be randomized to a perceived there is even less for the lower BMI subjects.
inferior intervention. Patients in a bariatric surgery program There are often tremendous pressures to advance a novel
recruited to a RCT comparing conventional bariatric surgery procedure or device to practice. These pressures include the
to an endoluminal procedure or medications had generally potential financial gains of the developer or company produc-
entered the program seeking surgery and therefore would ing the device or the academic pressures of the investigators.
not likely be interested in anything else. Additionally, those These pressures create action agendas that can knowingly or
who do agree to participate in such a trial may opt out of the unknowingly result in unethical behavior. It is therefore critical
trial before its completion if they become frustrated with that the standard rules of ethical research apply to these patients:
inferior results. One must also consider that there may be
patient biases for those patients who do enter such trials that (1) Appropriate and sufficient preclinical testing was
could influence the outcomes. performed;
The ethics of such RCTs needs also to be mentioned. Since (2) The trial be well designed and the risks to the patient be
there is currently substantial evidence that anastomotic proce- minimized;
dures are significantly beneficial for treating type 2 diabetes, is (3) The trial design and the informed consent receive IRB
it ethical to randomize patients with type 2 diabetes to less approval;
effective treatments such as purely restrictive procedures or (4) Proper and comprehensive informed consent be given to
novel technologies? the subject;
514 OBES SURG (2014) 24:487–519

(5) Study participants should be selected without pressure or (4) Access to bariatric surgery should not be denied to a
coercion; patient with class I obesity associated to significant
(6) The subject should have all of his/her questions obesity-related co-morbidity simply on the basis of the
answered; BMI level, which is an inaccurate index of adiposity and
(7) Subjects are allowed to withdraw from the trial any time a poor health risk predictor. Patients with class I obesity
without consequence; who are not able to achieve adequate weight loss after a
(8) Subjects understand the anatomic and physiologic reasonable period of nonsurgical therapy should be con-
changes resulting from the operative procedure and ex- sidered for bariatric surgery.
tent in which it can be reversed if the subject chooses to (5) Bariatric surgery should be considered in patients with
withdraw from the trial; class I obesity on an individual basis and after a compre-
(9) The subject is informed of any knowledge of long-term hensive clinical evaluation of the patient’s global health
effects of the surgical procedure; and and a prediction of its future disease risk. The use of
(10) The publication of the results of these trials, favorable or bariatric surgery in patients with class I obesity should be
unfavorable, should be made public. considered only after failure of proper nonsurgical
therapy.
(6) Indication to bariatric surgery in class I obesity should be
Conclusion based more on the comorbidity burden than on BMI
levels. Comorbidities should be evaluated considering
Performing bariatric operative procedures on class I obese their likely response to surgery and in relation to how
patients with significant comorbid conditions is becoming they can be treated by established medical therapies.
increasingly popular. The publish data thus far is supportive (7) The use of bariatric surgery should be avoided in patients
both for low risk and for clinical benefits. However, the pub- with class I obesity and advanced obesity-related or
lished literature on the subject is small and hampered by many obesity-unrelated comorbidities (frailty patients), in
factors related to poor study design, short follow-up, and diver- which intentional weight loss may not have any benefi-
sity of clinical definitions. Better designed research is still cial effect on prognosis or may be harmful.
indicated before wide acceptance particularly with regard to (8) The use of bariatric surgery cannot be currently recom-
novel procedures and new devices. All research endeavors must mended in children/adolescents or in elderly obese pa-
be conducted with the highest levels of ethical behavior. tients with class I obesity.
(9) National and regional health providers need to consider
the current evidences favoring the application of bariatric
Final Recommendations surgery in class I obesity in the context of local health
resources and deliver services that are locally appropriate.
On the basis of the data and considerations on the use of (10) Published literature on bariatric surgery in class I obesity
bariatric surgery in patients with class I obesity (BMI 30– is small and hampered by many factors related to poor
35 kg/m2 down to BMI 27.5 kg/m2 for at risk ethnicities) study design, short follow-up, and diversity of clinical
exposed in this document, the International Federation for the definitions. Accrual of controlled long-term data is strong-
Surgery of Obesity and Metabolic Disorders (IFSO) issues ly advised. The introduction in clinical practice of novel
and endorses the following statements and clinical procedures and new devices should be guided by the
recommendations. results of appropriately designed research protocols con-
ducted with the highest levels of ethical behavior.
(1) The impact on health of class I obesity varies greatly
between subjects. However, the physical, psychological,
and social health burden imposed by class I obesity may
be great at an individual level. References
(2) Nonsurgical therapies may achieve a clinically meaning-
ful weight loss in a significant number of patients with
1. Stevens GA, Singh GM, Lu Y, et al. National, regional, and global
class I obesity, but this weight loss is maintained in the trends in adult overweight and obesity prevalences. Popul Health
long term only in a smaller proportion of them. Metrics. 2012;10:22.
(3) Bariatric surgery is a highly effective weight loss strategy 2. Sturm R. Increases in morbid obesity in the USA: 2000–2005.
in patients with class I obesity at least in the short term. Public Health. 2007;121:492–6.
3. Malik VS, Willett WC, Hu FB. Global obesity: trends, risk factors
Adverse event’s rate in class I obese patients appears to and policy implications. Nat Rev Endocrinol. 2013;9:13–27.
be the same than in morbid obesity, with some studies 4. Flegal KM, Kit BK, Orpana H, et al. Association of all-cause
reporting serious adverse events. mortality with overweight and obesity using standard body mass
OBES SURG (2014) 24:487–519 515

index categories: a systematic review and meta-analysis. JAMA. 26. Renehan AG, Tyson M, Egger M, et al. Body-mass index and
2013;309:71–82. incidence of cancer: a systematic review and meta-analysis of
5. Finkelstein EA, Brown DS, Wrage LA, et al. Individual and aggre- prospective observational studies. Lancet. 2008;371:569–78.
gate years-of-life-lost associated with overweight and obesity. 27. Keating CL, Peeters A, Swinburn BA, et al. Utility-based quality of
Obesity. 2010;18:333–9. life associated with overweight and obesity: the Australian diabetes,
6. Zheng W, McLerran DF, Rolland B, et al. Association between obesity, and lifestyle study. Obesity. 2013;21:652–5.
body-mass index and risk of death in more than 1 million Asians. 28. Kolotkin RL, Crosby RD, Pendleton R, et al. Health-related quality
N Engl J Med. 2011;364:719–29. of life in patients seeking gastric bypass surgery vs non-treatment-
7. WHO. Appropriate body-mass index for Asian populations and its seeking controls. Obes Surg. 2003;13:371–7.
implications for policy and intervention strategies. Lancet. 29. Everson SA, Maty SC, Lynch JW, et al. Epidemiologic evidence for
2004;363:157–63. the relation between socioeconomic status and depression, obesity,
8. Carnethon MR, De Chavez PJ, Biggs ML, et al. Association of and diabetes. J Psychosom Res. 2002;53:891–5.
weight status with mortality in adults with incident diabetes. JAMA. 30. Colles SL, Dixon JB, O’Brien PE. Binge eating disorder: preva-
2012;308:581–90. lence and correlates in sub-groups of the Australian population.
9. Jerant A, Franks P. Body mass index, diabetes, hypertension, and Asia Pac J Clin Nutr. 2005;14:S109.
short-term mortality: a population-based observational study, 2000– 31. McCrea RL, Berger YG, King MB. Body mass index and common
2006. J Am Board Fam Med. 2012;25:422–31. mental disorders: exploring the shape of the association and its mod-
10. Doehner W, Erdmann E, Cairns R, et al. Inverse relation of body eration by age, gender and education. Int J Obes. 2012;36:414–21.
weight and weight change with mortality and morbidity in patients 32. Seidell JC. Societal and personal costs of obesity. Exp Clin
with type 2 diabetes and cardiovascular co-morbidity: an analysis of Endocrinol Diabetes. 1998;106:7–9.
the PROactive study population. Int J Cardiol. 2012;162:20–6. 33. Finkelstein EA, Di Bonaventura M, Burgess SM, et al. The costs of
11. Kokkinos P, Myers J, Faselis C, et al. BMI-mortality paradox and obesity in the workplace. J Occup Environ Med. 2010;52:971–6.
fitness in African American and Caucasian men with type 2 diabe- 34. National Institutes of Health National Institutes of Health. National
tes. Diabetes Care. 2012;35:1021–7. Heart, Lung and Blood Institute: clinical guidelines on the identifi-
12. Tseng CH. Obesity paradox: differential effects on cancer and cation, evaluation and treatment of overweight and obesity in adults.
noncancer mortality in patients with type 2 diabetes mellitus. The Evidence Report, June 1998.
Atherosclerosis. 2013;226:186–92. 35. National Institute for Health and Clinical Excellence. Obesity.
13. The Look AHEAD Research Group. Cardiovascular effects of Guidance on the prevention, identification, assessment and man-
intesive lifestyle intervention in type 2 diabetes. N Engl J Med. agement of overweight and obesity in adults and children. NICE
2013;369:145–54. Clin Guidelines. 2006;43:1–84.
14. Pontiroli AE, Morabito A. Long-term prevention of mortality in 36. Avenell A, Brown TJ, McGee MA, et al. What are the long-term
morbid obesity through bariatric surgery. a systematic review and benefits of weight reducing diets in adults? A systematic review of
meta-analysis of trials performed with gastric banding and gastric randomized controlled trials. J Hum Nutr Diet. 2004;17:317–35.
bypass. Ann Surg. 2011;253:484–7. 37. Tuomilehto J, Lindström D, Eriksson GE, et al. Prevention of type 2
15. Amundson DE, Djurkovic S, Matwiyoff GN. The obesity paradox. diabetes mellitus by changes in lifestyle among subjects with im-
Crit Care Clin. 2010;26:583–96. paired glucose tolerance. N Engl J Med. 2001;344:1343–50.
16. Morse SA, Gulati R, Reisin E. The obesity paradox and cardiovas- 38. Diabetes Prevention Program Research Group. Reduction in the
cular disease. Curr Hypertens Rep. 2010;12:120–6. incidence of type 2 diabetes with lifestyle intervention or metfor-
17. Oreopoulos A, Kalantar-Zadeh K, Sharma AM, et al. The obesity min. N Engl J Med. 2002;346:393–403.
paradox in the elderly: potential mechanisms and clinical implica- 39. Rawal LB, Tapp RJ, Williams ED, et al. Prevention of type 2
tions. Clin Geriatr Med. 2009;25:643–59. diabetes and its complications in developing countries: a review.
18. Colditz GA, Willett WC, Rotnitzky A, et al. Weight gain as a risk Int J Behav Med. 2012;19:121–33.
factor for clinical diabetes mellitus in women. Ann Intern Med. 40. The Look AHEAD Research Group. Reduction in weight and
1995;122:481–6. cardiovascular disease risk factors in individuals with type 2 diabe-
19. Chan JM, Rimm EB, Colditz GA, et al. Obesity, fat distribution, and tes. One-year results of the Look AHEAD trial. Diabetes Care.
weight gain as risk factors for clinical diabetes in men. Diabetes 2007;30:1374–83.
Care. 1994;17:961–9. 41. The Look AHEAD Research Group. Long-term effects of a lifestyle
20. Sjöström L. Review of the key results from the Swedish Obese intervention on weight and cardiovascular risk factors in individual
Subjects (SOS) trial: a prospective controlled intervention study of with type 2 diabetes mellitus. Four-year results of the Look AHEAD
bariatric surgery. J Intern Med. 2013;273:219–34. trial. Arch Intern Med. 2010;170:1566–75.
21. Lindström D, Ilanne-Parikka P, Peltonen M, et al. Sustained reduction 42. Gregg EW, Chen H, Wagenknecht LE, et al. Association of an
in the incidence of type 2 diabetes by lifestyle intervention: follow- intensive lifestyle intervention with remission of type 2 diabetes.
up of the Finnish diabetes prevention study. Lancet. 2006;368: JAMA. 2012;308:2489–96.
1673–9. 43. Heymsfield SB, van Mierlo CA, van der Knaap HC, et al. Weight
22. Diabetes Prevention Program Research Group. 10-year follow-up of management using a meal replacement strategy: meta and pooling
diabetes incidence and weight loss in the diabetes prevention pro- analysis from six studies. Int J Obes Relat Metab Disord. 2003;27:
gram outcomes study. Lancet. 2009;374:1677–86. 537–49.
23. Nguyen NT, Magno CP, Lane KT, et al. Association of hyperten- 44. Chearskul S, Delbridge E, Shulkes A, et al. Effect of weight loss and
sion, diabetes, dyslipidemia, and metabolic syndrome with obesity: ketosis on postprandial cholecystokinin and free fatty acid concen-
findings from the National Health and Nutrition Examination trations. Am J Clin Nutr. 2008;87:1238–46.
Survey, 1999 to 2004. J Am Coll Surg. 2008;207:928–34. 45. Mulholland Y, Nicokavoura E, Broom J, et al. Very-low-energy
24. Livingston EH, Chandalia M, Abate N. Do current body mass index diets and morbidity: a systematic review of longer-term evidence.
criteria for obesity surgery reflect cardiovascular risk? Surg Obes Br J Nutr. 2012;108:832–51.
Relat Dis. 2007;3:577–85. 46. Baker S, Jerums PJ. Effects and clinical potential of very-low-
25. Dixon JB. The effect of obesity on health outcomes. Mol Cell calorie diets (VLCDs) in type 2 diabetes. Diabetes Res Clin Pract.
Endocrinol. 2010;316:104–8. 2009;85:235–42.
516 OBES SURG (2014) 24:487–519

47. Anderson JW, Konz EC, Frederich RC, et al. Long-term weight-loss 66. Sandler BJ, Rumbaut R, Swain CP, et al. Human experience with an
maintenance: a meta-analysis of US studies. Am J Clin Nutr. endoluminal, endoscopic, gastrojejunal bypass sleeve. Surg Endosc.
2001;74:579–84. 2011;25:3028–33.
48. Anderson JW, Vichitbandra S, Qian W, et al. Long-term weight 67. Gastrointestinal surgery for severe obesity. National Institutes of
maintenance after an intensive weight-loss program. J Am Coll Health consensus development conference draft statement. Obes
Nutr. 1999;18:620–7. Surg 1991;1:257–66.
49. Lantz H, Peltonen M, Agren L, et al. Intermittent versus on-demand 68. Fried M, Hainer V, Basdevant A, et al. Inter-disciplinary European
use of a very low calorie diet: a randomized 2-year clinical trial. J guidelines on surgery of severe obesity. Int J Obes. 2007;31:569–77.
Intern Med. 2003;253:463–71. 69. Rubino F, Kaplan LM, Schauer PR, et al. The diabetes surgery
50. Colles SL, Dixon JB, Marks P, et al. Preoperative weight loss with a summit consensus conference recommendations for the evaluation
very-low-energy diet: quantitation of changes in liver and abdomi- and use of gastrointestinal surgery to treat type 2 diabetes mellitus.
nal fat by serial imaging. Am J Clin Nutr. 2006;84:304–11. Ann Surg. 2010;251:399–405.
51. Padwal RS, Majumdar SR. Drug treatments for obesity: orlistat, 70. American Diabetes Association. Summary of revisions for the 2009
sibutramine, and rimonabant. Lancet. 2007;369:71–7. clinical practice recommendations. Diabetes Care. 2009;32:S3–5.
52. Sjöström L, Rissanen A, Andersen T, et al. Randomised placebo- 71. Dixon JB, Zimmet P, Alberti KG, et al. Bariatric surgery: an IDF
controlled trial of orlistat for weight loss and prevention of weight statement for obese type 2 diabetes. Diabet Med. 2011;28:628–42.
regain in obese patients. Lancet. 1998;352:167–73. 72. Australian Government - National Health and Medical Research
53. Torgerson JS, Hauptman J, Boldrin MN, et al. XENical in the Council - Department of Health and Ageing. Clinical practice
prevention of Diabetes in Obese Subjects (XENDOS) study. A guidelines for the management of overweight and obesity in adults,
randomized study of orlistat as an adjunct to lifestyle changes for adolescents and children in Australia, 2013.
the prevention of type 2 diabetes in obese patients. Diabetes Care. 73. Mechanick JI, Youdim A, Jones DB, et al. Clinical practice guide-
2004;27:155–61. lines for the perioperative nutritional, metabolic, and nonsurgical
54. Smith SR, Weissman NJ, Anderson CM, et al. Multicenter, placebo- support of the bariatric surgery patient—2013 update: cosponsored
controlled trial of lorcaserin for weight management. N Engl J Med. by American Association of Clinical Endocrinologists, the Obesity
2010;363:245–56. Society, and American Society for Metabolic and Bariatric Surgery.
55. Gadde KM, Allison DB, Ryan DH, et al. Effects of low-dose, Surg Obes Relat Dis. 2013;9:159–91.
controlled-release, phentermine plus topiramate combination on 74. ASMBS Clinical Issues Committee. Bariatric surgery in class I
weight and associated comorbidities in overweight and obese adults obesity (BMI 30–35 kg/m2). Surg Obes Relat Dis. 2013;9:e1–e10.
(CONQUER): a randomised, placebo controlled, phase 3 trial. 75. Müller MJ, Lagerpusch M, Enderle J, et al. Beyond the body mass
Lancet. 2011;377:1341–52. index: tracking body composition in the pathogenesis of obesity and
56. Garvey WT, Ryan DH, Look M, et al. Two-year sustained weight the metabolic syndrome. Obes Rev. 2012;13:6–13.
loss and metabolic benefits with controlled-release phentermine/ 76. Okorodudu DO, Jumean MF, Montori VM, et al. Diagnostic per-
topiramate in obese and overweight adults (SEQUEL): a random- formance of body mass index to identify obesity as defined by body
ized, placebo-controlled, phase 3 extension study. Am J Clin Nutr. adiposity: a systematic review and meta-analysis. Int J Obes.
2012;95:297–308. 2010;34:791–9.
57. Greenway FL, Fujioka K, Plodkowski RA, et al. Effect of naltrex- 77. Thomas EL, Frost G, Taylor-Robinson SD, et al. Excess body fat in
one plus bupropion on weight loss in overweight and obese adults obese and normal-weight subjects. Nutr Res Rev. 2012;25:150–61.
(COR-I): a multicentre, randomised, double-blind, placebo- 78. Kragelund C, Omland T. A farewell to body-mass index? Lancet.
controlled, phase 3 trial. Lancet. 2010;376:595–605. 2005;366:1589–91.
58. Apovian C, Aronne L, Rubino D, et al. A randomized, phase 3 trial 79. Deurenberg P, Deurenberg-Yap M. Validity of body composition
of Naltrexone SR/Bupropion SR on weight and obesity-related risk methods across ethnic population groups. Forum Nutr. 2003;56:
factors (COR-II). Obesity (Silver Spring). 2013;21:935–43. 299–301.
59. Nauck M, Frid A, Hermansen K, et al. Efficacy and safety compar- 80. Habbu A, Lakkis NM, Dokainish H. The obesity paradox: fact or
ison of liraglutide, glimepiride, and placebo, all in combination with fiction? Am J Cardiol. 2005;98:944–8.
metformin, in type 2 diabetes: the LEAD (liraglutide effect and 81. Fonarow GC, Srikanthan P, Costanzo MR, et al. An obesity paradox
action in diabetes)-2 study. Diabetes Care. 2009;32:84–90. in acute heart failure: analysis of body mass index and in hospital
60. Buse JB, Rosenstock J, Sesti G, et al. Liraglutide once a day versus mortality for 109,927 patients in the acute decompensated heart
exenatide twice a day for type 2 diabetes: a 26-week randomised, failure national registry. Am Heart J. 2007;153:74–81.
parallel-group, multinational, open-label trial (LEAD-6). Lancet. 82. Oreopoulos A, Padwal R, Kalantar-zadeh K, et al. Body mass index
2009;374:39–47. and mortality in heart failure: a meta-analysis. Am Heart J.
61. Astrup A, Rössner S, Van Gaal L, et al. Effects of liraglutide in the 2008;156:13–22.
treatment of obesity: a randomised, double-blind, placebo- 83. Kalantar-Zadeh K, Kopple JD, Kilpatrick RD, et al. Association of
controlled study. Lancet. 2009;374:1606–16. morbid obesity and weight change over time with cardiovascular
62. Astrup A, Carraro R, Finer N, et al. Safety, tolerability and sustained survival in hemodialysis population. Am J Kidney Dis. 2005;46:
weight loss over 2 years with the once-daily human GLP-1 analog, 489–500.
liraglutide. Int J Obes. 2012;36:843–54. 84. Galal W, van Gestel YR, Hoeks SE, et al. The obesity paradox in
63. Genco A, Bruni T, Doldi SB, et al. BioEnterics intragastric balloon: patients with peripheral arterial disease. Chest. 2008;134:925–30.
the Italian experience with 2,515 patients. Obes Surg. 2005;15: 85. Hastie CE, Padmanabhan S, Slack R, et al. Obesity paradox in a
1161–4. cohort of 4880 consecutive patients undergoing percutaneous coro-
64. Verdam FJ, Schouten R, Greve JW, et al. An update on less invasive nary intervention. Eur Heart J. 2010;31:222–6.
and endoscopic techniques mimicking the effect of bariatric surgery. 86. Diercks D, Roe MT, Mulgund J, et al. The obesity paradox in non-
J Obes. 2012;2012:597871. ST-segment elevation acute coronary syndromes: results from the
65. Schouten R, Rijs CS, Bouvy ND, et al. A multicenter, randomized can rapid risk stratification of unstable angina patients suppress
efficacy study of the endobarrier gastrointestinal liner for adverse outcomes with early implementation of the American
presurgical weight loss prior to bariatric surgery. Ann Surg. College of Cardiology/American Heart Association guidelines qual-
2010;251:236–43. ity improvement initiative. Am Heart J. 2006;152:140–8.
OBES SURG (2014) 24:487–519 517

87. Mullen JT, Moorman DW, Davenport DL. The obesity paradox. 110. O’Brien PE, Dixon JB, Laurie C, et al. Treatment of mild to
Ann Surg. 2009;250:166–72. moderate obesity with laparoscopic adjustable gastric banding or
88. Unger RH. Minireview: weapons of lean body mass destruction: the an intensive medical program: a randomized trial. Ann Intern Med.
role of ectopic lipids in the metabolic syndrome. Endocrinology. 2006;144:625–33.
2003;144:5159–65. 111. Dixon JB, O’Brien PE, Playfair J, et al. Adjustable gastric banding
89. Montani JP, Carroll JF, Dwyer TM, et al. Ectopic fat storage in the and conventional therapy for type 2 diabetes: a randomized con-
heart, blood vessels and kidneys in the pathogenesis of cardiovas- trolled trial. JAMA. 2008;299:316–23.
cular diseases. Int J Obes Relat Metab Disord. 2004;28:S58–65. 112. Lee WJ, Chong K, Ser KH, et al. Gastric bypass vs sleeve gastrec-
90. Matsuzawa Y, Nakamura T, Shimomura I, et al. Visceral fat accu- tomy for type 2 diabetes mellitus: a randomized controlled trial.
mulation and cardiovascular disease. Obes Res. 1995;3:S645–7. Arch Surg. 2011;146:143–8.
91. Després JP, Lemieux I. Abdominal obesity and metabolic syn- 113. Schauer PR, Kashyap SR, Wolski K, et al. Bariatric surgery versus
drome. Nature. 2006;444:881–7. intensive medical therapy in obese patients with diabetes. N Engl J
92. Stefan N, Kantartzis K, Machann J, et al. Identification and charac- Med. 2012;26(366):1567–76.
terization of metabolically benign obesity in humans. Arch Intern 114. Ikramuddin S, Korner J, Lee WJ, et al. Roux-en-Y Gastric Bypass
Med. 2008;168:1609–16. vs intensive medical management for the control of type 2 diabetes,
93. Wildman RP, Muntner P, Reynolds K, et al. The obese without hypertension, and hyperlipidemia. The diabetes surgery study ran-
cardiometabolic risk factors clustering and the normal weight with domized clinical trial. JAMA. 2013;309:2240–9.
cardiometabolic risk factor clustering: prevalence and correlates of 2 115. Li Q, Chen L, Yang Z, et al. Metabolic effects of bariatric surgery in
phenotypes among the US population (NHANES 1999–2004). type 2 diabetic patients with body mass index of 35 kg/m2. Diabetes
Arch Intern Med. 2008;168:1617–24. Obes Metab. 2012;14:262–70.
94. International Diabetes Federation. The IDF consensus worldwide 116. Shah SS, Todkar JS, Shah PS, et al. Diabetes remission and reduced
definition of the metabolic syndrome. IDF Publications 2005. cardiovascular risk after gastric bypass in Asian Indians with body
95. Heitmann BL, Lissner L. Hip hip hurrah! Hip size inversely related mass index of 35 kg/m2. Surg Obes Relat Dis. 2010;6:332–8.
to heart disease and total mortality. Obes Rev. 2011;12:478–81. 117. Huang CK, Shabbir A, Lo CH, et al. Laparoscopic Roux-en-Y gastric
96. Dasarathy S, Dasarathy J, Khiyami A, et al. Validity of real time bypass for the treatment of type II diabetes mellitus in Chinese
ultrasound in the diagnosis of hepatic steatosis: a prospective study. patients with body mass index of 25–35. Obes Surg. 2011;21:1344–9.
J Hepatol. 2009;51:1061–7. 118. De Sa VC, Ferraz AA, Campos JM, et al. Gastric bypass in the
97. Fabbrini E, Magkos F, Mohammed BS, et al. Intrahepatic fat, not treatment of type 2 diabetes in patients with a BMI of 30 to 35 kg/
visceral fat, is linked with metabolic complications of obesity. Proc m2. Obes Surg. 2011;21:283–7.
Natl Acad Sci U S A. 2009;106:15430–5. 119. Cohen R, Pinheiro JC, Schiavon CA, et al. Effects of gastric bypass
98. Iacobellis G, Malavazos AE, Corsi MM. Epicardial fat: from the surgery in patients with type 2 diabetes and only mild obesity.
biomolecular aspects to the clinical practice. Int J Biochem Cell Diabetes Care. 2012;35:1420–8.
Biol. 2011;43:1651–4. 120. Ferzli GS, Dominique E, Ciaglia M, et al. Clinical improvement after
99. Pories WJ, Dohm LG, Mansfield CJ. Beyond the BMI: the search duodenojejunal bypass for nonobese type 2 diabetes despite minimal
for better guidelines for bariatric surgery. Obesity. 2010;18:865–71. improvement in glycemic homeostasis. World J Surg. 2009;33:972–9.
100. Sharma AM, Kushner RF. A proposed clinical staging system for 121. Geloneze B, Geloneze SR, Fiori C. Surgery for nonobese type 2
obesity. Int J Obes. 2009;33:289–95. diabetic patients: an interventional study with duodeno-jejunal ex-
101. Padwal RS, Pajewski NM, Allison DB, et al. Using the Edmonton clusion. Obes Surg. 2009;19:1077–83.
obesity staging system to predict mortality in a population- 122. Ramos AC, Galvao MP, de Souza YM. Laparoscopic duodenal–
representative cohort of people with overweight and obesity. jejunal exclusion in the treatment of type 2 diabetes mellitus in
CMAJ. 2011;183:E1059–66. patients with BMI < 30 kg/m2 (LBMI). Obes Surg. 2009;19:307–12.
102. Kuk JL, Ardern CI, Church TS, et al. Edmonton obesity staging 123. Scopinaro N, Papadia F, Marinari G, et al. Long-term control of type
system: association with weight history and mortality risk. Appl 2 diabetes mellitus and other major components of the metabolic
Physiol Nutr Metab. 2011;36:570–6. syndrome after biliopancreatic diversion in patients with BMI <
103. Gill RS, Karmali S, Sharma AM. The potential role of the 35 kg/m2. Obes Surg. 2007;17:185–92.
Edmonton obesity staging system in determining indications for 124. Chiellini C, Rubino F, Castagneto M, et al. The effects of bilio-
bariatric surgery. Obes Surg. 2011;21:1947–9. pancreatic diversion on type 2 diabetes in patients with BMI <
104. Buchwald H, Estok R, Fahrbach K, et al. Trends in mortality in 35 kg/m2. Diabetologia. 2009;52:1027–30.
bariatric surgery: a systematic review and meta-analysis. Surgery. 125. Scopinaro N, Adami GF, Papadia FS, et al. The effects of
2007;142:621–35. biliopancreatic diversion on type 2 diabetes mellitus in patients with
105. Thomas H, Agrawal S. Systematic review of obesity surgery mor- mild obesity (BMI 30–35 kg/m2) and simple overweight (BMI 25–
tality risk score—preoperative risk stratification in bariatric surgery. 30 kg/m2): a prospective controlled study. Obes Surg. 2011;21:880–8.
Obes Surg. 2012;22:1135–40. 126. Lee WJ, Wang W, Lee YC, et al. Effect of laparoscopic mini-gastric
106. The Longitudinal Assessment of Bariatric Surgery (LABS) bypass for type 2 diabetes mellitus: comparison of BMI > 35 kg/m2.
Consortium. Perioperative safety in the longitudinal assessment of J Gastrointest Surg. 2008;12:945–52.
bariatric surgery. N Engl J Med. 2009;361:445–54. 127. Lee WJ, Chong K, Chen CY, et al. Diabetes remission and insulin
107. De Maria EJ, Portenier D, Wolfe L. Obesity surgery mortality risk secretion after gastric bypass in patients with body mass index <
score: proposal for a clinically useful score to predict mortality risk 35 kg/m2. Obes Surg. 2011;21:889–95.
in patients undergoing gastric bypass. Surg Obes Relat Dis. 2007;3: 128. De Paula AL, Macedo ALV, Schraibman V, et al. Hormonal evalu-
134–40. ation following laparoscopic treatment of type 2 diabetes mellitus
108. Blackstone RP, Cortes MC. Metabolic acuity score: effect on major patients with BMI 20–34. Surg Endosc. 2009;23:1724–32.
complications after bariatric surgery. Surg Obes Relat Dis. 2010;6: 129. Reis CEG, Alvarez-Leite JI, Bressan J, et al. Role of bariatric–metabolic
267–73. surgery in the treatment of obese type 2 diabetes with body mass index
109. Turner PL, Saager L, Dalton J, et al. A nomogram for predicting < 35 kg/m2: a literature review. Diab Technol Ther. 2012;14(4):365–72.
surgical complications in bariatric surgery patients. Obes Surg. 130. Cohen R, Caravatto PP, Correa JL, et al. Glycemic control after
2011;21:655–62. stomach-sparing duodenal–jejunal bypass surgery in diabetic
518 OBES SURG (2014) 24:487–519

patients with low body mass index. Surg Obes Relat Dis. 2012;8: 150. Baur LA, Fitzgerald DA. Recommendations for bariatric surgery in
375–80. adolescents in Australia and New Zealand. J Paediatr Child Health.
131. Maglione MA, Gibbons MM, Livhits M, et al. Bariatric surgery and 2010;46:704–7.
nonsurgical therapy in adults with metabolic conditions and a body 151. Bannerman E, Miller MD, Daniels LA, et al. Anthropometric indi-
mass index of 30.0 to 34.9 kg/m2. Rockville: Agency for Healthcare ces predict physical function and mobility in older Australians: the
Research and Quality (US); 2013. Australian longitudinal study of ageing. Public Health Nutr. 2002;5:
132. Parikh M, Issa R, Vieira D, et al. Role of bariatric surgery as 655–62.
treatment for type 2 diabetes in patients who do not meet current 152. Villareal DT, Chode S, Parimi N, et al. Weight loss, exercise, or both
NIH criteria: a systematic review and meta-analysis. J Am Coll and physical function in obese older adults. N Engl J Med.
Surg. 2013;217:527–32. 2011;364:1218–29.
133. Flum DR, Belle SH, et al. Perioperative safety in the longitudinal 153. Houston DK, Nicklas BJ, Zizza CA. Weighty concerns: the growing
assessment of bariatric surgery. N Engl J Med. 2009;361:445–54. prevalence of obesity among older adults. J Am Diet Assoc.
134. De Maria EJ, Winegar DA, Pate VW, et al. Early postoperative 2009;109:1886–95.
outcomes of metabolic surgery to treat diabetes from sites partici- 154. Zajacova A, Ailshire J. Body mass trajectories and mortality among
pating in the ASMBS bariatric surgery center of excellence program older adults: a joint growth mixture-discrete-time survival analysis.
as reported in the bariatric outcomes longitudinal database. Ann Gerontologist 2013 Jan 25 [Epub ahead of print].
Surg. 2010;252:559–66. 155. Buchwald H, Oien DM. Metabolic/bariatric surgery worldwide
135. Parikh M, Duncombe J, Fielding GA. Laparoscopic adjustable 2011. Obes Surg. 2013;23:427–36.
gastric banding for patients with body mass index < 35 kg/m2. 156. Picot J, Jones J, Colquitt JL, et al. Weight loss surgery for mild to
Surg Obes Relat Dis. 2006;2:518–22. moderate obesity: a systematic review and economic evaluation.
136. Sultan S, Parikh M, Youn H, et al. Early U.S. outcomes after Obes Surg. 2012;22:1496–506.
laparoscopic adjustable gastric banding in patients with a body mass 157. Berrington de Gonzalez A, Hartge P, Cerhan JR, et al. Body-mass
index less than 35 kg/m2. Surg Endosc. 2009;23:1569–73. index and mortality among 1.46 million white adults. N Engl J Med.
137. Choi J, Digiorgi M, Milone L, et al. Outcomes of laparoscopic 2010;363:2211–9.
adjustable gastric banding in patients with low body mass index. 158. Buchwald H, Avidor Y, Braunwald E, et al. Bariatric surgery. A
Surg Obes Relat Dis. 2010;6:367–71. systemic review and meta-analysis. JAMA. 2004;292:1724–37.
138. Kakoulidis TP, Karringer A, Gloaguen T, et al. Initial results with 159. Schauer PR, Ikramuddin S, Gourash W, et al. Outcomes after
sleeve gastrectomy for patients with class I obesity (BMI 30–35 kg/ laparoscopic Roux-en-Y gastric bypass for morbid obesity. Ann
m2). Surg Obes Relat Dis. 2009;5:425–8. Surg. 2000;232:515–29.
139. Abbatini F, Capoccia D, Casella G, et al. Type 2 diabetes in obese 160. DePaula AL, Stival A, Halpern A, et al. Thirty-day morbidity and
patients with body mass index of 30–35 kg/m2: sleeve gas- mortality of the laparoscopic ileal interposition associated with
trectomy versus medical treatment. Surg Obes Relat Dis. 2012;8: sleeve gastrectomy for the treatment of type 2 diabetic patients with
20–4. BMI < 35: an analysis of 454 consecutive patients. World J Surg.
140. Angrisani L, Favretti F, Furbetta F, et al. Italian group for lap-band 2011;35:102–8.
system: results of multicenter study on patients with BMI r 35 kg/ 161. McKneally MF, Daar AS. Introducing new technologies: protecting
m2. Obes Surg. 2004;14:415–8. subjects of surgical innovation and research. World J Surg. 2003;27:
141. Serrot FJ, Dorman RB, Miller CJ, et al. Comparative effectiveness 930–5.
of bariatric surgery and nonsurgical therapy in adults with type 2 162. Roman S, Napoleon B, Mion F, et al. Intragastric balloon for “non-
diabetes mellitus and body mass index of 35 kg/m2. Surgery. morbid” obesity: a retrospective evaluation of tolerance and effica-
2011;150:684–91. cy. Obes Surg. 2004;14:539–44.
142. Frenken M, Cho EY. Metabolic intestinal bypass surgery for type 2 163. Champion JK, Williams M, Champion S, et al. Implantable gastric
diabetes in patients with a BMI of 35 kg/m2: comparative analysis of stimulation to achieve weight loss in patients with a low body mass
16 patients undergoing either BPD, BPD-DS, or RYGB. Obes index: early clinical results. Surg Endosc. 2006;20:444–7.
Facts. 2011;4:13–7. 164. Afterburner DE, Bogart A, Sherwood NE, et al. A multisite study of
143. Gianos M, Abdemur A, Fendrich I, et al. Outcomes of bariatric long-term remission and relapse of type 2 diabetes mellitus follow-
surgery in patients with body mass index < 35 kg/m2. Surg Obes ing gastric bypass. Obes Surg. 2013;23:93–102.
Relat Dis. 2012;8:25–30. 165. Ritz P, Hanaire H. Post-bypass hypoglycaemia: a review of the
144. Angrisani L, Cutolo PP, Formisano GP, et al. Long-term outcomes current findings. Diabetes Metab. 2011;37:274–81.
of laparoscopic adjustable silicone gastric banding (LAGB) in mod- 166. Lerner H, Whang J, Nipper R. Benefit–risk paradigm for clinical
erately obese patients with and without co-morbidities. Obes Surg. trial design of obesity devices: FDA proposal. Surg Endosc.
2013;23:897–902. 2013;27:702–7.
145. Deitel M. Slow-progression, autoimmune, type 1 diabetes in adults: 167. Reinhold RB. Critical analysis of long term weight loss following
a cause of failure of resolution of diabetes after bariatric surgery. gastric bypass. Surg Gynecol Obstet. 1982;155:385–94.
Surg Obes Relat Dis. 2009;5:705–6. 168. Metropolitan height and weight tables. 64, 3–9. 1983. New York,
146. Deitel M. Update: why diabetes does not resolve in some patients Metropolitan Life Foundation. Statistical Bulletin
after bariatric surgery. Obes Surg. 2011;21:794–6. 169. Committee on Standards for Reporting Results, American Society
147. Wells JC. Ethnic variability in adiposity, thrifty phenotypes and for Bariatric Surgery. Standards for reporting results. Obes Surg
cardiometabolic risk: addressing the full range of ethnicity, 1994;4:56–65.
including those of mixed ethnicity. Obes Rev. 2012;13 Suppl 170. Junior WS, Campos CS, Nonino CB. Reporting results after bariat-
2:14–29. ric surgery: reproducibility of predicted body mass index. Obes
148. WHO. Obesity: preventing and managing the global epidemic. Surg. 2012;22:519–22.
Report of a who consultation. World Health Organ Tech Rep Ser. 171. Deitel M, Gawdat K, Melissas J. Reporting weight loss. Obes Surg.
2000;894:1–253. 2007;17:565–8.
149. Inge TH, Krebs NF, Garcia VF, et al. Bariatric surgery for severely 172. Belle SH, Berk PD, Courcoulas AP, et al. Reporting weight change:
overweight adolescents: concerns and recommendations. Pediatrics. Standardized reporting accounting for baseline weight. Surg Obes
2004;114:217–23. Relat Dis. 2013;9:782–9.
OBES SURG (2014) 24:487–519 519

173. Sabin J, Fanelli R, Flaherty H, et al. Best practice guidelines on vagal blockade to induce weight loss in morbid obesity. Obes Surg.
informed consent for weight loss surgery patients. Obes Res. 2012;11:1771–82.
2005;13:250–3. 175. Shikora SA, Bergenstal R, Bessler M, et al. Implantable gastric
174. Sarr MG, Billington CJ, Brancatisano R, et al. The EMPOWER stimulation for the treatment of clinically severe obesity: results of
study: randomized, prospective, double-blind, multicenter trail of the SHAPE trial. Surg Obes Relat Dis. 2009;5:31–7.