Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Surgery for morbid obesity

Systematic review

Few people can be unaware that being overweight is a bad thing. It contributes to a range of conditions most of us would wish to avoid, including heart disease, stroke, hypertension, diabetes, cancer, arthritis, and asthma, to name but a few. Diet and exercise are the obvious ways forward, with anti-obesity drugs helping in some cases (Bandolier 121).

Morbid obesity has a definition, a BMI of 40 or more, or 35 or more in the presence of significant co-morbidities. A BMI of 40 for someone 1.8 metres (almost six feet) tall would mean a weight of about 130 kg (almost 300 lbs) or more. Here more extreme measures are suggested, including surgical therapy. A new systematic review and meta-analysis of surgical therapy in people massively overweight [1] indicates the results to be impressive, in terms of weight loss and other conditions. Intragastric balloon therapy was not included.

Systematic review

Surgical procedures were grouped into gastric banding, gastric bypass, gastroplasty, biliopancreatic diversion or duodenal switch, and other procedures. Studies of any design were sought as long as they had at least 10 subjects, had at least 30 days of follow up, and were in English.

Extensive searching looked for studies published up to July 2003. Outcomes looked for in addition to weight loss included one or more of diabetes, hypertension, hyperlipidaemia, or obstructive sleep apnoea. Mortality within 30 days was also sought. For some of the co-morbidities the outcome sought was whether the condition had resolved, or had resolved or improved. Resolved here meant the conditions either disappeared or no longer required therapy. For lipid disorders improved meant normalisation of laboratory values or discontinuation of therapy.


There were 136 reports, five randomised studies, 28 nonrandomised controlled trials or series, and 101 uncontrolled case series, with 22,000 patients in total. Patients in the studies had an average age of 39 years, and about 73% were women. The average BMI was 47 (range 32 to 69). These patients had high rates of comorbidity (Figure 1).

Figure 1: Comorbidity rates in people undergoing surgery for morbid obesity

Across all studies with about 10,000 reporting the outcomes, the absolute weight loss averaged 41 kg, and the BMI fell by 14 points. Weight fell by 36% of weight before the operation, and 65% of excess weight was lost. The percentage of excess weight (total preoperative weight minus the ideal weight) lost by each of the four techniques is shown in Figure 2.

Figure 2: Average percentage of excess weight lost with four surgical techniques

Surgery also resulted in high rates of resolution or improvement in comorbid conditions like diabetes, hypertension, sleep apnoea, and lipid disorders (Table 1). In general, there were more substantial benefits with biliopancreatic diversion or duodenal switch, as for diabetes in Figure 3.

Table 1: Overall average resolution or improvement rates for comorbid conditions

Resolved or improved
Obsructive sleep apnoea

Figure 3: Percentage of people with diabetes resolved with four surgical techniques

Laboratory values fell substantially. Total cholesterol fell on average by 0.5 mmol/L across all techniques, but the range for particular techniques was 0.2 mmol/L for gastric banding to 1.8 mmol/L for biliopancreatic diversion or duodenal switch. Fasting blood glucose fell by 4 mmol/L on average, and HbA1 c by 2.7%.

Mortality within 30 days was 0.1% for purely restrictive procedures (gastric banding or gastroplasty, about 3,000 patients), 0.5% in 5,600 undergoing gastric bypass, and 1.1% in 3,000 patients undergoing biliopancreatic diversion or duodenal switch.


Not having thought about it too hard, Bandolier's prejudices were that surgery for obesity was probably a waste of time. This review has shaken up those prejudices, thrown them out, and replaced them with evidence that very large beneficial changes follow surgery for morbid obesity. The weight losses recorded were huge, and the resolution or improvement in a wide range of conditions that too much flesh is heir to was close to spectacular.

A concern is that most of the evidence comes from case series, and not the randomised trials to which we are used. The overall numbers of patients and size of the outcomes are such as to overcome those qualms in most cases, though there were only small numbers of patients in some of the subgroup analyses.

There were no health economic studies included. That is not too unexpected, but one doesn't need a brain the size of a planet to see how the arguments are likely to go: costs of surgery in one column, saving in all those comorbid condition treatments for many years in the other. And finally, this type of surgery is now called bariatic surgery. It's a new word to Bandolier, and we have no idea of its derivation. Perhaps there is an etymological genius out there who can enlighten us?


  1. H Buchwald et al. Bariatric surgery. A systematic review and meta-analysis. JAMA 292: 1724-1737.

previous or next story