Article Text

Download PDFPDF

Surgery for morbid obesity
  1. John M H Bennett,
  2. Samir Mehta,
  3. Michael Rhodes
  1. Department of Upper Gastrointestinal Surgery, Norfolk and Norwich University Hospital NHS Trust, Norwich, UK
  1. Correspondence to:
 MrM Rhodes
 Department of Upper Gastrointestinal Surgery, Norfolk and Norwich University Hospital NHS Trust, Colney Lane, Norwich NR4 7UY, UK; mr{at}


The prevalence of morbid obesity in the UK population is rising, bringing with it increased levels of cardiovascular disease, diabetes, arthritis and early mortality. The overall cost to the health service is high, and is set to increase over the coming decades as the overweight population ages. Dietary, lifestyle and pharmacological interventions offer at best reasonable, short-term weight reduction and often fail. Surgical intervention is a safe and effective means of delivering marked long-term weight reduction. This article compares and contrasts the options available for surgical treatment of morbid obesity based on a review of the current literature.

  • BMI, body mass index
  • BPD, biliopancreatic diversion
  • EWL, excess weight loss
  • LAGB, laparoscopic gastric banding
  • NICE, National Institute for Health and Clinical Excellence
  • RYGBP, Roux-en-Y gastric bypass
  • VBG, vertical banded gastroplasty
  • morbid obesity
  • obesity surgery
  • review
View Full Text

Statistics from

The past 20 years has seen a doubling in the worldwide prevalence of morbid obesity.1 In 2002, the National Institute for Health and Clinical Excellence (NICE) estimated that there were 1.2 million morbidly obese people in England and Wales, and that this number would increase by 5% per annum. Morbid obesity is associated with twice the mortality compared with the general population, and with other conditions such as hypertension and type II diabetes, as well as increased levels of psychiatric illness. There is also a considerable strain on healthcare resources and society as a whole, and this cost is set to rise as the general population becomes more obese. The aetiology of morbid obesity is complex, involving the interaction of psychosocial, genetic, endocrine and metabolic factors, making conservative treatment difficult and prone to failure. Surgery is more invasive, but is the only means of achieving considerable and sustained weight loss—a rapidly increasing priority in UK medical practice.

Metabolic effects of morbid obesity

The most common method of assessing obesity is the body mass index (BMI; mass (kg)/height (m2)). Both NICE and the National Institute of Health (Bethesda, Maryland, USA) define morbid obesity as a BMI of ⩾40 kg/m2, or ⩾35 kg/m2, along with one of the comorbidities shown in the box.

Many of the comorbidities listed in the box result from the metabolic effects of obesity. The so-called “metabolic syndrome” becomes more common with rising BMI, and carries a threefold increase in cardiovascular morbidity. Polycystic ovary disease, the most common chronic disorder in women, is also more prevalent in this population. Morbid obesity is also associated with severe venous stasis, degenerative joint disease, gout, cholelithiasis, urinary stress incontinence and pseudotumour cerebri. Further, non-alcoholic fatty liver disease is present in 90% of liver biopsy specimens taken during bariatric surgery, with non-alcoholic steatohepatitis diagnosed in 25% of patients.2

Many of these comorbidities have been shown to improve, and several to resolve completely, after considerable weight reduction. Morbid obesity not only increases the risk of metabolic and physiological abnormalities but is also closely associated with increased levels of depression and anxiety.3,4 When compared with those having chronic medical conditions, obese patients rate their health-related quality of life as considerably worse.

Treatment options for morbid obesity

The three main treatment options for morbid obesity are lifestyle change, pharmacotherapy and surgery. Lifestyle change should involve restriction of calorie intake with increase in exercise so that moderate weight loss can occur in the short term. However, up to 66% of patients regain weight within 24 months and long-term results are poor.5–7 Studies have shown that pharmacotherapy through sibutramine, orlistat or phentermine is more effective, and greater weight loss has been shown to occur compared with lifestyle interventions alone.8,9 Drug treatment is probably most beneficial when used in conjunction with a suitable weight-management programme. Further, these drugs also have other beneficial effects on lipid levels, blood pressure and insulin resistance beyond that expected from their effects on weight loss alone.6 Despite short-term success, weight loss is usually not maintained, and up to 90% regain weight on cessation of treatment. In addition, there are reports of side effects (including hypertension, tachycardia and bowel disorders) that preclude their broad usage.9,10

Bariatric surgery offers the only means of delivering sustained weight loss. Over 50% excess weight loss (EWL) can be achieved after surgical intervention, and these results are sustained at 10–15 year follow-up. The Swedish Obese Subjects study comparing surgical, pharmacological and lifestyle interventions showed a clear benefit for surgery at 10 years, with an average weight loss of 16% in the surgical arm compared with 1.6% weight gain in the non-surgical arm.11–13 In addition, a meta-analysis performed by Buchwald et al14 showed a weight loss of 61.2% of excess weight after surgical intervention.

Box: NICE guidelines—comorbidities associated with morbid obesity

  • Cardiovascular disease

  • Hypertension

  • Type II diabetes mellitus

  • Osteoarthritis

  • Reproductive disorders (infertility)

  • Respiratory disorders (sleep apnoea)

  • Cancers (endometrial/ovarian)

Comorbidity resolution after bariatric surgery occurs in 75–90% of cases, particularly for insulin resistance, hypertension and respiratory disorders.12,14–16 The effects of sustained weight reduction on the lifetime risk of developing obesity-related conditions have been quantified by Cristou et al17 in a large Canadian cohort study. After 5 years, the relative risks of developing cardiovascular, endocrinological (including diabetes), musculoskeletal, genitourinary, psychiatric, respiratory and haematological disorders were markedly reduced. Further, the relative risk of developing cancer in the surgical cohort was 0.21 after surgery compared with non-operated morbidly obese subjects over the 5-year period.

The number of bariatric procedures performed worldwide has increased considerably over the past few years. This review discusses why surgery is now having a greater role in the treatment for morbid obesity. Each of the available procedures is discussed, in turn followed by a systematic comparison of the largest case studies to compare safety and efficacy.


The ideal surgical intervention for morbid obesity should be effective, safe and applicable to all patients. It must achieve considerable weight loss and resolution of comorbidity. Low operative morbidity and mortality is essential, with a short hospital stay and rapid return to normal activity desirable. All patients undergoing a bariatric procedure require meticulous follow-up. A multidisciplinary team incorporating dieticians, psychiatrists and endocrinologists should always be available both for advice and to deal with any problems that arise.

Current surgical practice can be divided by the mechanism of weight reduction: restrictive by decreasing the storage capacity of the stomach, malabsorptive through surgical bypass thus excluding intestinal loops, or a combination of the two.


Vertical banded gastroplasty

Vertical banded gastroplasty (VBG) was originally described by Mason in 1982,18 and involves the placement of a 5 cm band around a gastric pouch created by stapling the gastric fundus (fig 1). The procedure is most commonly performed laparoscopically because of the improved wound, pulmonary and thromboembolic complication rates. It takes around 1½ h to perform, with a hospital stay of approximately 5 days. Satisfactory levels of weight loss can be achieved, with one trial documenting >50% EWL in 74% of patients.19 In most series with 3 to 5-year follow-up, EWL of ⩾50% has been achieved in only 40%.20,21

Figure 1

 Vertical banded gastroplasty.

Mortality is <1% in most studies, with an overall morbidity of 14%.12,15,21–23 Complications include stomal stenosis (20%), staple line disruption (11%), severe oesophagitis (7%) and band migration (1.5%).24 Patients can often complain of intolerance to solid food coupled with persistent vomiting. This leads to a prolonged hospital stay. Occasionally, surgical revision is the only solution.

Vertical banded gastroplasty, open and laparoscopic, has produced good results for both weight loss and comorbidity, but comparable or better results are achieved with either adjustable gastric banding or gastric bypass. Both of these latter procedures also seem to have better side effect profiles, and this has led to a shift towards use of the gastric band as the restrictive procedure of choice.

Adjustable gastric banding

Laparoscopic gastric banding (LAGB) is by far the most popular restrictive bariatric surgical intervention in Europe and Australia, and is rapidly gaining popularity in the US after receiving approval from the Food and Drug Administration in 2001. Developed by Belachew25 in 1992, to date over 250 000 gastric bands have been placed worldwide. The procedure involves the placement of an inflatable band to form a 15–20 ml superior gastric pouch, with band position reinforced by the placement of anterior gastro-gastric sutures. The band connects to a self-sealing reservoir (Portacath) implanted beneath the skin. This allows for adjustment of the stoma diameter to increase or reduce the rate of passage of food from the upper pouch into the body of the stomach (fig 2). The sensation of satiety caused by the distension of the gastric pouch leads to early meal termination, and hence reduced calorie intake. Patients also report prolonged post-meal satiety, paradoxically as they have markedly reduced energy intake, which further decreases their calorie intake. The mechanism by which this occurs is still not understood.26 Band placement takes 30–60 min, with a conversion rate of around 1%. Patients remain in hospital for around 24 h and return to normal activities within 1 week.27–29

Figure 2

 Laparoscopic adjustable gastric banding.

Laparoscopic banding can achieve a mean EWL of approximately 50% in the short term (1–2 years), which continues in the long term.30–34 Both Steffen (n = 824)35 and Zinzindohoue (n = 500)29 have shown an EWL of ⩾50% in 80% of their patients at over 3 years follow-up. After LAGB, the main obesity-related comorbidities (hypertension, diabetes, respiratory disease and osteoarthritis) also improve markedly, with some studies showing complete resolution.29,36,37 Further, banding has been successful in the treatment of the super obese (those with a BMI >50 kg/m2), achieving >50% EWL and 90% comorbidity resolution, with lower postoperative morbidity and mortality than other alternative bariatric procedures.28,36,38,39 Laparoscopic banding is the safest bariatric surgical procedure, with a postoperative mortality of <0.5%.40–42 Many studies have shown a considerably lower incidence of complications after LAGB compared with alternative procedures, and these conclusions are supported by both the Australian Safety and Efficacy Register of New Interventional Procedures-Surgical and NICE.24,28,39,43–45 Further, a single-centre study of the three main surgical interventions has shown a markedly (3.5 times) lower level of overall and severe complications for LAGB compared with either laparoscopic Roux-en-Y bypass or laparoscopic biliopancreatic diversion.46

Procedure specific complications include band slippage, band erosion, oesophageal dilatation and Portacath migration. Slippage (herniation of the stomach superiorly through the band) can cause partial or total gastric obstruction, and often requires surgical correction. An early series reported slippage rates of 12–24%, but by adopting a pars flaccida approach for band placement (dissection outside the lesser sac)46,47 in conjunction with application of anterior gastro-gastric sutures and by delaying the inflation of the band, the incidence has been reduced to 1–2%.47,48 Band erosion prevalence is <2% in a larger series,49 and is commonly caused by gastric microperforation or by oversewing of the band’s buckle. Another problem is oesophageal dilatation/dysmotility, which can often remain asymptomatic. The true incidence is unknown, as regular contrast studies are required to monitor for occurrence.50 However, the condition seems to resolve fully on band deflation.16,51 Portacath-related problems occur in around 7.5% of cases. Studies have shown that complications generally reduce in incidence with increasing experience of the surgeon, and all can be treated safely either by minimally invasive or by local anaesthetic procedures. Technical complications such as band puncture during gastro-gastric suturing may also potentially occur, but become very rare with increasing surgical experience.13,52,53

The disadvantage of LAGB, as with all bariatric surgical procedures, is that it is not effective in everyone. LAGB is therefore best performed in a multidisciplinary unit, so that patients can receive professional support before and after surgery, with access to regular follow-up and band adjustment to ensure optimum weight loss.54 Some patients, however, inevitably require a second procedure. Laparoscopic conversion or reversal of LAGB is safe,40,55 with low operative mortality (0%) and morbidity (14.3%); conversion rates are comparable to those for the primary procedure.56 Weiner’s40 experience with nearly 1000 bands confirms that laparoscopic removal of the band or laparoscopic conversion to an alternative procedure carries a low mortality and morbidity, with duration of hospital stay similar to that for the primary procedure.

LAGB has proved itself fast, effective and safe in the treatment of morbid obesity. It achieves a gradual, controlled and adjustable reduction in weight, with excellent results at 2–3 years. Weight loss is sustained in the long term. It has a short learning curve, and markedly lower mortality/morbidity than alternative procedures. It does fail to achieve sufficient weight loss in a small proportion of cases, and these then require conversion to an alternative, usually malabsorptive, procedure.


Jejunoileal bypass

Jejunoileal bypass was the first widely performed operation for obesity. The procedure entailed excluding most of the small bowel, by leaving just 30 cm of the jejunum and 10 cm of the ileum in continuity. Although successful in causing weight loss, it is now a defunct procedure for the treatment of morbid obesity, owing to a severe side effect profile. The procedure did achieve sustained weight loss in 70% of patients,57 but the long-term complications of liver failure (10%), urolithiasis (29%) and renal failure (9%) have led to its abandonment.58 Many of the 25 000 jejunoileal bypasses performed in the US have since been converted to gastric bypass.

Biliopancreatic bypass with or without duodenal switch procedure

Biliopancreatic diversion is often the procedure of choice for patients with very high BMI (>60 kg/m2) owing to its impressive weight loss profile. Its widespread use is limited by a more severe side effect profile and higher mortality compared with alternative procedures. First reported in 1979,59 BPD involves the formation of a distal gastrectomy with a proximal 200–500 ml gastric pouch and a long Roux-en-Y reconstruction. The ileum is divided 250 cm from the ileocaecal valve and an enteroenterostomy fashioned 50 cm from the ileocaecal valve, thus forming a 200-cm alimentary limb (anastomosed proximally to the stomach) and a 50-cm common channel (fig 3). The operation results in inadequate digestion inside the short common limb.

Figure 3

 Biliopancreatic diversion.

Because of the frequency of stomal ulceration and unpleasant dumping syndrome with biliopancreatic diversion, the duodenal switch procedure, involving a sleeve gastrectomy and preservation of the pyloric sphincter, was developed with marked success (fig 4).60,61 The restrictive component is short lived, so that appetite and eating capacity are restored to preoperative levels within 1 year. Long-term weight loss is achieved mainly through malabsorption.59,62

Figure 4

 Biliopancreatic diversion with duodenal switch.

Operation time for the open procedure can be >3½ h during the learning curve, but averages 2½ h in experienced hands.63 Postoperative hospital stay is longer than for other bariatric procedures, befitting a large open surgical procedure, and there is a considerable recuperation period before return to normal activities. The laparoscopic approach, introduced in 2001, benefits from reduced wound complications, with equivalent non-wound-related morbidity and shorter hospital stay, although operation time is considerably longer.59,64

Scopinaro published two large Italian series of 1356 and 2000 BPDs reporting EWL of 74–78% and 73–78%, respectively. These results have been supported by several smaller series.59,65,66 Follow-up at 15 years has shown little or no weight regain (table 1). Further, weight loss is associated with high rates of comorbidity resolution, particularly for diabetes and hypertension.59,63 Failure to achieve or maintain sufficient weight reduction, up to 15% in some series, requires re-operation to modify limb length or the addition of a greater restrictive component.67,68

Table 1

 List of published studies reporting >500 bariatric procedures

Published mortality for BPD with duodenal switch is around 1%. Early wound-related complications occur in 1.5%, with the incidence of incisional hernia 25% (for open surgery). Respiratory complications occur in 0.4% and stomal ulceration in 8.3% of patients (for patients having BPD without duodenal switch). Further, protein calorie malnutrition can also occur in many and may require a period of parenteral feeding. Despite mineral and vitamin supplementation, persistent anaemia occurs in a considerable proportion of patients and bone demineralisation is also common.59,69,70 Finally, the development of severe metabolic complications may necessitate surgical revision—a technically simple procedure, but one carrying high surgical morbidity and risk of weight gain.71

The poor side effect profile, increased operative time and longer hospital stay compared with other bariatric procedures would seem to restrict its use to patients in whom other interventions fail. Although few centres perform this procedure, BPD may have a role as part of a two-stage procedure, with an initial sleeve gastrectomy and follow-up BPD once a proportion of weight has been lost. This approach may be particularly useful in the super obese (BMI >50 kg/m2), and has recently been shown to be both safe and effective.72


Gastric bypass

This procedure was first described by Mason and Ito in 1967,73 and integrates restriction with altered absorption. The Roux-en-Y gastric bypass (RYGBP) is now the most common bariatric procedure performed in the US. The procedure involves the formation of a 15–20 ml gastric pouch and the fashioning of a Roux-en-Y gastrojejunostomy bypassing the distal stomach, duodenum and a variable length of proximal jejunum (fig 5). This effectively reduces the size of meal that the patient is able to ingest. Additionally, the bypass causes some degree of malabsorption. Operating time for the open procedure is approximately 100 min, but this rises to 140 min if the procedure is performed laparoscopically.74 A hospital stay of 5 days, and several weeks to fully recover, is to be expected after open bypass,43 but both are markedly reduced if the laparoscopic approach is successful.27

Figure 5

 Roux-en-Y gastric bypass procedure.

Gastric bypass achieves excellent initial weight reduction, with a mean EWL of nearly 70% at 1 year.14 A number of case series have shown that after 3 years 60–70% of patients can achieve >50% weight loss.44,57,75–77 Long-term results are good, with an average EWL of 60% at 5 years, but this decreases to around 50% at 8–10 years.78,79 Gastric bypass is also effective in the treatment of “super” obese patients with a BMI of ⩾60 kg/m2.80 As with all bariatric procedures, it carries a risk of failure, with 15% of patients failing to achieve or maintain an EWL of >50%.81 Revisional surgery is complex and carries significant risk, with options including increasing malabsorption (conversion to BPD or increasing limb length) or application of further restrictive procedures such as the lap-band.82

Bypass surgery carries a mortality of approximately 0.5% for both the open and laparoscopic approaches. Wound related and pulmonary complications are comparable to alternative open surgical interventions. Studies have shown that the laparoscopic approach offers a reduction in critical care requirement, postoperative pulmonary complications and incidence of incisional hernia.74,83,84 However, the most serious complication with this operation, anastomotic leak, occurs in 2–5% of open cases,81 but possibly at a higher rate after laparoscopic bypass.85 It is thus important that the surgeon embarking on laparoscopic RYGBP has sufficient expertise to master the technical demands of the procedure. Studies have also shown that gastro-gastric fistula, small bowel obstruction and internal herniation are also more prevalent after laparoscopic bypass.83,84 Marginal ulceration (at the gastrojejunostomy site) has been reported at 3%, and responds to pharmacological treatment in most of the cases.43,86 Long-term vitamin and mineral deficiency, although more amenable to supplementation than in BPD, can still necessitate surgical revision.

Open and laparoscopic gastric bypass can deliver both significant weight loss and resolution of comorbid conditions, although on failure, it requires further intervention, carrying significant risk. The laparoscopic approach has become more popular, with its reduced incidence of postoperative wound, thromboembolic and respiratory complications. These benefits are partially negated by the increased leakage rate and notably longer general anaesthesia. The proven efficacy of gastric bypass has ensured that it remains the most popular intervention in the US.

Mini-gastric bypass

A new development is the laparoscopic mini-gastric bypass procedure, a modification of the older loop gastrojejunostomy. It involves the formation of a long gastric tube approximately 1.5 cm to the left of the lesser curvature of the stomach from the antrum to the angle of His and then a loop gastroenterostomy is formed, about 200 cm from the ligament of Treitz (fig 6). The procedure takes up to 150 min, depending on experience,87,88 and requires conversion in 0.3% of cases. Duration of hospital stay is 2–5 days, with a return to normal activities after 1 week.87–89

Figure 6

 Mini-gastric bypass.

Laparoscopic mini-gastric bypass procedure has been shown to achieve EWL of >70% at 2 years, equivalent to RYGBP,89 but long-term data are not yet available.89,90 In addition, it carries a mortality of <0.1%, anastomotic leak rate of <0.1%, lower risk of thromboembolic and pulmonary complications, and achieves comorbidity resolution in >70% of patients.

It is a simpler and easier laparoscopic procedure to perform than RYGBP; however, long-term data are still needed to determine whether it can match RYGBP in terms of sustained weight reduction, and also whether there is an increased incidence of long-term complications such as biliary reflux, marginal ulceration and reflux oesophagitis.

Concluding comparison of surgical techniques

Morbid obesity is associated with increased prevalence of multiple metabolic, physiological and psychological abnormalities. All of these markedly reduce the lifespan of this increasing section of the population. Surgery is effective not only at reducing weight, but also at resolving associated comorbidities while improving patients’ quality of life.

All of the techniques described above help achieve marked weight loss in morbidly obese patients. Table 1 lists the largest available published case series of bariatric procedures. At present there is considerable interest in LAGB; results in table 1 confirm the relative safety of LAGB, with only a few studies reporting perioperative mortality. Excess weight loss seems to reach a peak in most studies at 2–3 years, and in three studies remains at over 50% after 5 years. The procedure is also easier to perform compared with gastric bypass. Although gastric bypass operations do have the potential to achieve greater weight loss, this benefit is tempered by increased mortality and morbidity. Further, the procedure is technically demanding and surgeons must pass a learning curve before embarking on performing it. In this respect, the laparoscopic mini-gastric bypass may be an easier alternative, although the procedure is still in its infancy, with long-term weight loss data not yet available and the risk of as yet unknown longer-term complications.

Vertical banded gastroplasty is now generally of historical interest, with few centres performing large numbers. This procedure has an unacceptable complication rate, which has led to its widespread disuse. Biliopancreatic diversion, however, may still have a role, particularly in the cohort of patients who have unsatisfactory results from restrictive surgery. This relatively invasive and complex procedure can now be safely executed laparoscopically.91 It can induce substantial weight loss in the long term, but should be considered primarily in the super obese and in those who cannot tolerate food intake restriction but will accept long-term follow-up by a multidisciplinary team.


  1. Morbid obesity is defined as a body mass index of >35 kg/m2

  2. Surgery to treat obesity is recommended only for patients with a BMI >40 kg/m2

  3. Pharmacological treatment combined with lifestyle intervention is a more effective method of producing sustained weight loss than surgery

  4. Surgery to induce weight loss improves or resolves 90% of obesity related comorbidities

  5. Biliopancreatic diversion (BPD) is a restrictive procedure

  6. Roux-en-Y gastric bypass (RYGBP) is a combination of restriction and malabsorption

  7. Protein-calorie malnutrition is a rare complication of BPD

  8. Complication and mortality are ordered (highest to lowest): BPD–RYGBP–VBG–LAGB

  9. Excess weight loss rates are ordered (highest to lowest): VBG–BPD–RYGBP–LAGB

  10. LAGB achieves gradual weight loss reaching parity with RYGBP at ⩾5 years


1. T; 2. F, patients with a BMI >35 kg/m2 with one of the comorbidities listed in table 1, or with a BMI >40 kg/m2 with or without comorbidity; 3. F; 4. T; 5. F; 6. T; 7. F; 8. T; 9. F, BPD>RYGBP>LAGB/VBG; 10. T


View Abstract


  • Competing interests: None.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.