Bariatric Surgery: Risks and Rewards
Walter J. Pories
Professor of Surgery, Biochemistry, Sport and Exercise Science, Director, Metabolic Institute, East Carolina University, Greenville, North
Context: Over 23 million Americans are afflicted with severe obesity, i.e. their body mass index (in
kilograms per square meter) values exceed 35. Of even greater concern is the association of the
adiposity with comorbidities such as diabetes, hypertension, cardiopulmonary failure, asthma,
pseudotumor cerebri, infertility, and crippling arthritis.
This article examines the effect of surgery on the control of the weight and the comorbidities, as
well as the safety of these operations.
Interventions: Although the article focuses on the outcomes of the three most commonly per-
formed operations, i.e. adjustable gastric banding, the gastric bypass, and the biliopancreatic
bypass with duodenal switch, it aims for perspective with the inclusion of abandoned and current
investigational procedures, a review of the complications, and an emphasis on the appropriate
selection of patients.
2 diabetes in over 80% with salutary effects on the other comorbidities as well with significant
reductions in all-cause mortality. Although the severely obese present with serious surgical risks,
out the United States—similar to the complication rates after cholecystectomy.
Conclusions: Until better approaches become available, bariatric surgery is the therapy of choice
for patients with severe obesity. (J Clin Endocrinol Metab 93: S89–S96, 2008)
chronic diseases? How can one procedure reverse obesity,
diabetes, hypertension, cardiopulmonary failure, polycystic
ovary disease, and pseudotumor cerebri with a reduction in
mortality (Fig. 1)? And, finally, is it really true that the op-
eration, a highly complex abdominal procedure performed in
vulnerable, severely obese patients, is now delivered through-
out the United States with operative mortalities and morbid-
ities that are no greater than the risks for cholecystectomy?
Surprisingly, these claims are well supported.
The intent of this article is to review the history of bariatric
surgery, to compare the various bariatric operations, to list the
current indications for these procedures, to evaluate the out-
comes, and to consider the risks.
retty hard to believe. How can it be that one operation can
produce full and durable remissions of our most resistant
The History of Bariatric Surgery
Bariatric surgery, similar to the other surgical disciplines, was
developed in waves through the contributions of many (1). The
the University of Minnesota, led by Drs. Arnold Kremen and
Richard Varco, that severe obesity was a disease dangerous
enough to warrant surgery. Based on the experience with the
“short gut” syndrome, they developed the intestinal bypass, a
procedure that excludes the majority of the small intestine from
an end-to-end anastomosis between 14 inches (36 cm) of jeju-
num and 4 inches (10 cm) of ileum with the excluded segment
draining into the sigmoid colon provided the most predictable
Printed in U.S.A.
Copyright © 2008 by The Endocrine Society
doi: 10.1210/jc.2008-1641 Received July 28, 2008. Accepted September 15, 2008.
bypass with duodenal switch; GS, gastric sleeve; IGT, impaired glucose tolerance; RYGB,
Roux-en-Y gastric bypass; VBG, vertical banded gastroplasty.
R e v i e w
J Clin Endocrinol Metab, November 2008, 93(11):S89–S96jcem.endojournals.org
it was recognized that although the operations produced signif-
icant weight loss, the complications were unacceptable. Eventu-
ally almost all had to be reversed because they produced acute
severe mineral abnormalities.
The second major breakthrough came with the careful inves-
group, who documented that weight loss could be achieved as
effectively and far more safely through two gastric procedures,
the gastric band, an operation that limited intake with a small
gastric pouch and limited outlet, and the gastric bypass, a pro-
cedure that interfered with digestion as well as intake by exclud-
to be the most widely performed bariatric procedures in the
world today. Gastric banding has been improved with the in-
vention of the adjustable gastric band; the gastric bypass was
extended by Scopinaro et al. (2) with the biliopancreatic bypass
and by Hess et al. (3) with the addition of a duodenal switch.
More recently, the gastric sleeve (GS) (4), the initial step in the
biliopancreatic bypass, is under investigation as another inde-
extending the benefits of the gastric bypass to diabetic patients
bypass. Others are testing ileal transposition, i.e. the translocation
of a segment of ileum close to the Ligament of Treitz, as another
approach to resolving type 2 diabetes without weight loss.
The third advance was the documentation by Pories et al. (6)
and MacDonald et al. (7), with rigorous 95% follow-up of
608 patients for up to 16 yr, that the gastric bypass produced
bidities, even including diabetes; and 3) a decrease in mortality.
The fourth major development was the demonstration in
1994 by Wittgrove and Clark (8) that the gastric bypass, one of
the most difficult abdominal surgical operations, could be per-
formed with the laparoscopic approach
safely and with far less trauma.
basis and the documentation that the oper-
and morbidity in centers with high volume
and experience. Confronted by reports of
disastrous clinical outcomes in hospitals
with limited experience, an explosion of
malpractice suits and unaffordable insur-
ance premiums, the leadership of the Amer-
gery (ASMBS) founded a program for the
certification of Centers of Excellence. The
concept differed from previous attempts at
surgical quality control by requiring stan-
dardization of care paths and focusing not
only on process but primarily on surgical
outcomes. To assure credibility and stake-
holder participation, the Society founded
the Surgical Review Corporation (SRC)
hospitals throughout the United States were certified as ASMBS
Centers of Excellence, delivering bariatric surgery with a 0.14%
hospital and a 0.35% 90-d mortality, similar rates to those re-
ported for cholecystectomy, although the severely obese repre-
sent far greater operative risks (10).
Variations on a Theme
Bariatric operations have traditionally been divided into three
groups: 1) restrictive, i.e. procedures that produce weight loss
digestion and absorption (intestinal bypass); and 3) and mixed,
tric bypass, duodenal switch). Despite this apparently clear clas-
sification, the mechanisms of action remain unclear. For exam-
intake due to the low volume of the tube, the longitudinal gas-
trectomy also discards the source of ghrelin production.
Figure 2 provides a diagrammatic overview of the various
operations previously or currently in use. The list is only partial.
Multiple variations of each of the operations have been per-
formed and discarded over the last 50 yr with variations in the
size of the gastric pouches, length of limbs, type and size of
anastomoses, with or without vagotomy, the addition of con-
stricting rings, and even wrapping the entire stomach in fabric.
Operations No Longer Widely Performed
A listing of the operations that were once popular and are now
no longer performed is important because patients with these
procedures are still encountered in practice.
FIG. 1. Gastric bypass surgery schedule. Dx, Diognosis; OR, operating room.
PoriesBariatric Surgery: Risks and RewardsJ Clin Endocrinol Metab, November 2008, 93(11):S89–S96
small bowel by joining 14 in. (36 cm) of proximal jejunum be-
yond the Ligament of Treitz to 4 in. (10 cm) of terminal ileum.
The excluded segment is drained into the distal colon (11). The
disastrous outcomes of these procedures are noted in the
Vertical banded gastroplasty (VBG)
The VBG was the first successful restrictive procedure. The
operation produced a 30-cc proximal gastric pouch with a ver-
tical staple line from the Angle of His to a circular opening,
measuring about 1 cm, punched out with a circular stapler. This
Marlex, about 1 cm in width, that was tightened to narrow the
outlet to about 1 cm. The operation is still performed by a few
surgeons with excellent results but has been largely replaced
by the adjustable gastric band, an operation that is far easier
to perform, safer, and less likely to fail due to staple line
Minigastric loop bypass
bypass, but it was soon abandoned because of biliary regurgi-
tation with bile gastritis and esophagitis. In
addition, animal studies documenting dys-
plasia raised concerns about potential dan-
gers of esophageal cancer. Although the op-
eration is avoided by most, a few surgeons
insist the procedure is safe and continue to
promote it (12).
Currently Accepted Operations
Roux-en-Y gastric bypass (RYGB)
The RYGB, the most commonly per-
formed bariatric operation in the United
in size, i.e. about the size of a golf ball, by
division. The proximal pouch is drained
with a Roux-en-Y created by dividing the
proximal jejunum about 30 cm below the
Ligament of Treitz, bringing the distal seg-
ment up to form a gastroenterostomy of
about 1 cm in diameter, and joining the
proximal segment to the small bowel about
100 cm below the point of division. Al-
though the procedure is becoming increas-
of a plastic ring at the gastroenterostomy to
limit outflow, length of the Roux-en-Y
limbs, and placement of the small bowel in
front or behind the colon (13).
Adjustable gastric band (AGB)
bracelet-like device placed high in the stomach to produce a
pouch of about 30 cm, similar to that in the RYGB. The band is
allow adjustment of the pouch outflow (14).
Biliopancreatic bypass with duodenal switch (BPDS)
about 2 cm below the pylorus, is reconstituted by a Roux-en- Y
anastomosis to the distal jejunum, excluding significantly more
small bowel than the gastric bypass. The operation is not widely
performed but is gaining adherents due to its effectiveness and
the ease of creating the GS, which appears to offer a reasonable
alternative to the gastric band (13).
Investigational Bariatric Surgical Procedures
Gastric sleeve (GS)
The GS creates a narrow gastric tube through the excision of
most of the stomach. The operation does more than just limit
FIG. 2. Overview of bariatric and metabolic surgical operations.
J Clin Endocrinol Metab, November 2008, 93(11):S89–S96jcem.endojournals.org
intake; it also removes most and perhaps all of the ghrelin-pro-
ducing cells of the gastric mucosa. The procedure, initially in-
troduced as a first stage of the BPDS for use in superobese pa-
tients to reduce risk, appears to be an effective operation on its
be listed as investigational because of the excellent results re-
The duodeno-jejunal bypass stomach-sparing operation was
introduced as a procedure that could induce remission of dia-
betes without weight loss in lean patients with type 2 diabetes
mellitus. The operation is based on the work of Rubino and
colleagues (16) in Goto-Kakizaki genetically diabetic lean rats.
Early human trials are encouraging.
Ileal transposition is still in the early stage of animal and
human trials. It is mentioned here for completeness and to make
the reader aware that there are early reports of satisfactory
weight loss and remission of diabetes, but the evidence remains
scant at this point, even in animal trials (17).
Effect of Bariatric Surgery on Weight
produces dramatic and durable weight loss. Weight loss is most
easily expressed in pounds or kilograms. In our series of 608
patients followed up to 16 yr with a 95% follow-up, the mean
(144.1 to 95.9 kg)? (18).
weight among the operated individuals, many prefer to express
weight loss in terms of “excess weight,” i.e. current weight ?
Metropolitan Life Insurance tables of 1999, a 5-foot 3-in.
woman with a medium frame has an “ideal body weight” of
121–135 lb (55–61 kg). The midpoint of this range is 128 lb
(58.1 kg). At a weight of 300 lb, she has an excess weight of 172
lost 120 lb (54.5 kg) or 69.7% of her excess body weight.
The use of percentage excess weight loss as a measure allows
some comparison between the various bariatric operations.
Buchwald et al. (19) used this metric in a meta-analysis of 2,738
the results of bariatric surgery in 22,094 patients. These data
AGB, 47.5%; VBG, 68.2%; RYGB, 61.6%; and BPDS, 79.1%.
improved stapling devices and bands that are less likely to fail as
well as a sharp move from open to laparoscopic procedures.
The body mass index (BMI; in kilograms per square meter) is
also frequently used as an index of obesity. However, although
the BMI has been adopted widely, it is a badly flawed index for
several reasons. As a unigender measure, it fails to reflect the
differences in muscularity and body composition between the
sexes. Because it only reflects weight and height, it fails to dif-
ferentiate between the well-muscled athlete and the obese indi-
weight of 308 lb (140 kg), numbers that do not reflect the fact
that his body fat represented only 7% of his body weight. The
Asian women suffer similar levels of comorbidities at a BMI of
when the BMI of at least 35 is used as an exclusion index, it denies
surgery. Similarly, it can be argued that some patients currently
undergo surgery based on their BMI, but possibly with a body fat
percentage close to normal range. It is not a sufficiently precise
measure on which to base life and death decisions.
Patients vary in their responses to bariatric operations. Most
patients change their diets (21) with a tendency to avoid beef
products and fibrous vegetables because they require a lot of
chewing and may obstruct the narrow gastric outflow tracts.
Some patients develop the symptoms of “dumping” when they
get drunk much more easily; whereas still others report major
changes in taste and food preference.
Other factors that influence weight loss include age, gen-
level of activity (13). In general, younger patients, females,
Caucasians, muscular and highly motivated individuals who
follow-up, and those who comply with the recommendations
for vitamin/mineral supplements and do not snack will lose
the most weight.
affect the degree of weight loss (22, 23), although the recovery is
significantly faster after minimally invasive approaches.
Most patients reach their maximum weight loss by 2 yr and
experience some increase of weight, perhaps 5–7%, by the fifth
year with a gradual decrease again over the following years.
Weight loss after the insertion of adjustable bands is generally
although there is early evidence that weight loss after banding
may continue into the fifth year.
Failure of bariatric surgery remains to be defined. Failure,
measured by the inadequate loss or the return of lost weight,
varies by procedure and intensity of follow-up, ranging between
5 and 10% with higher rates for adjustable gastric banding.
However, the lack of improvement of the comorbidities such as
diabetes, asthma, stress incontinence, infertility, cardiopulmo-
nary function, and pseudotumor represent more serious failures
than inadequate weight loss. The developing large databases in
the National Institutes of Health (NIH) Longitudinal Assess-
PoriesBariatric Surgery: Risks and Rewards J Clin Endocrinol Metab, November 2008, 93(11):S89–S96
of Excellence for Bariatric Surgery should provide the informa-
tion to develop objective measures.
from such factors as staple line breakdown, revisional surgery is
usually successful. If, on the other hand, the failure is due to
behavioral problems such as patients “out-eating” the opera-
tion, revision is usually not an effective approach.
Effect of Bariatric Surgery on Diabetes
The most remarkable effect of bariatric surgery is the full and
rapid remission of type 2 diabetes mellitus, a disease previously
considered unalterably progressive and minimally unresponsive
to therapies except with a few demanding and unrealistic pro-
to be our most expensive disease. Diabetes, increasing at a faster
rate than any other chronic disease, now affects over 24 million
of heart disease and stroke. The disease accounts for 11–13% of
the long-term outcomes in a series of 608 severely obese indi-
type 2 diabetes and another 165 patients had impaired glucose
tolerance (IGT). Durable resolution of diabetes with a return to
euglycemia and normalization of glycosylated hemoglobin val-
The diabetes clears rapidly, generally in a matter of days, to the
degree that most diabetic bariatric surgical patients are dis-
charged without any antidiabetic medications.
One paper indicates that the gastric bypass also reduces the
umented the reduction in mortality after gastric bypass, Mac-
Donald et al. (26) showed that in diabetics the mortality de-
creased from 4.5 to 1% per year, based on a comparison group.
The finding that six different operations on the intestine can
produce euglycemia has opened new avenues for diabetes re-
peptide, ghrelin, peptide YY3–36, and glucagon-like peptide-1
are only some of the gut hormones under intense investigation.
In fact, exenatide, an analog of glucagon-like peptide-1, is al-
ready enjoying wide and successful clinical application (27).
In the long run, the hope for diabetes lies in the dissection of
the metabolic pathways uncovered by bariatric surgery and ap-
plying the findings to the development of effective medical ther-
apies. We can’t operate on 24 million Americans.
The Effect of Bariatric Surgery on the Other
Comorbidities of Severe Obesity
Severe obesity affects virtually every system of the body with a
broad expression of serious diseases, including pseudotumor
cerebri, hypertension, diabetes, renal failure, immunoincompe-
tence, asthma, gastroesophageal reflux disease, chronic ob-
structive pulmonary disease, cardiac failure, atherosclerosis,
Pickwickian syndrome, arthritis of the weight bearing joints,
infertility, skin breakdown, and an increased prevalence of
cancers, especially colon, prostate, breast, and ovary.
All of these illnesses respond favorably to bariatric surgery,
often with total and permanent remission. It is not unusual for
patients who are restricted to wheelchairs before surgery to re-
turn to the surgeon 3 months later walking, often without even
a cane. Patients diagnosed with asthma and gastroesophageal
the various medications.
Most startling is the reduction in the prevalence of cancer in
patients who have undergone bariatric surgery (28). Is this reduc-
tion, about 80% within 5 yr after the surgery, due to the decrease
in inflammatory cytokines with the reduction of adipocytes? We
can only speculate, but the implications are exciting.
Indications and Contraindications to Bariatric
The original indications and contraindications to bariatric sur-
on the Surgery for Obesity. In 2004, the American Society for
Bariatric Surgery ?ASBS, recently renamed as the American
Society for Metabolic and Bariatric Surgery (ASMBS)? up-
dated that statement with a follow-up Consensus Conference
(29) that reached the following conclusions:
1. Bariatric surgery is the most effective therapy available for
morbid obesity and can result in improvement or complete res-
olution of obesity comorbidities.
2. Types of operative procedures for morbid obesity have in-
creased since 1991 and are continuously evolving. There are
currently four types of procedures that can be used to achieve
sustained weight loss: gastric bypass (standard, long-limb, and
very long-limb Roux), alone or in combination with vertical
banded gastroplasty; laparoscopic adjustable gastric banding;
vertical banded gastroplasty; and biliopancreatic diversion and
3. Both open and laparoscopic bariatric operations are effec-
tive therapies for morbid obesity and represent complementary
4. Bariatric surgery candidates should have attempted to lose
weight by nonoperative means, including self-directed dieting,
nutritional counseling, and commercial and hospital-based
TABLE 1. Full and durable remission of type 2 diabetes
mellitus in 608 patients after gastric bypass with a mean
follow-up of 9.4 yr
Type 2 diabetes
Total no. of patients
No. available for follow-up
Resolution of diabetes150 (99%)
J Clin Endocrinol Metab, November 2008, 93(11):S89–S96jcem.endojournals.org