Series: Landmark Trials in Endocrinology & Metabolism | Study #16
Category: Obesity & Weight Management | Design: Prospective, matched, controlled intervention study | n: 4,047 | Follow-up: 10.9 years (average)
📋 Summary
Authors: Sjöström L et al., for the Swedish Obese Subjects Study Scientific Group
Journal: N Engl J Med 2007;357:741–752 | DOI: 10.1056/NEJMoa066254
The Swedish Obese Subjects (SOS) study was a prospective, non-randomised, matched controlled intervention study that enrolled 4,047 obese subjects (mean BMI approximately 42 kg/m²) at 25 surgical and 480 primary health care centres across Sweden. Of these, 2,010 underwent bariatric surgery (gastric bypass, vertical-banded gastroplasty, or gastric banding, according to surgeon and patient preference), and 2,037 obese controls received conventional non-surgical obesity treatment, matched for a range of demographic and metabolic variables at baseline. The primary endpoint of the mortality analysis was overall death from any cause. At an average follow-up of 10.9 years, there were 101 deaths in the surgery group and 129 deaths in the control group. The unadjusted overall hazard ratio was 0.76 (p=0.04), and the adjusted hazard ratio was 0.71 (95% CI not reported in abstract; p=0.01). Maximum weight losses in surgical subgroups were observed at 1 to 2 years: 32% with gastric bypass, 25% with vertical-banded gastroplasty, and 20% with gastric banding. After 10 years, sustained weight losses from baseline were 25%, 16%, and 14% respectively, while the control group showed less than 2% weight change throughout. The most common causes of death were myocardial infarction (25 controls vs 13 surgery) and cancer (47 controls vs 29 surgery). SOS was the first prospective controlled study to demonstrate that bariatric surgery is associated with a significant reduction in long-term overall mortality compared with conventional obesity treatment.
📊 Key Findings
| Outcome | Surgery group | Control group | Effect Size |
|---|---|---|---|
| All-cause mortality | 101 deaths | 129 deaths | Adjusted HR 0.71 · p=0.01 |
| Deaths from MI | 13 | 25 | Reduced with surgery |
| Deaths from cancer | 29 | 47 | Reduced with surgery |
| Weight loss at 10 years: gastric bypass | 25% | <2% | Sustained long-term reduction |
| Weight loss at 10 years: VBG | 16% | <2% | Sustained |
| Weight loss at 10 years: banding | 14% | <2% | Sustained |
| T2DM incidence (long-term follow-up papers) | Substantially reduced | — | Reported in subsequent SOS analyses |
💬 Expert Commentary
The SOS study provided the most important long-term evidence on bariatric surgery and mortality available at the time of its 2007 publication. Despite not being a randomised trial — a limitation acknowledged by the investigators and extensively discussed in the literature — SOS filled a critical evidence gap that a conventional RCT could not practically fill in the short to medium term, given the logistical, ethical, and follow-up challenges of randomising severe obesity to surgical versus non-surgical management over a decade or more. The matched cohort design, with meticulous baseline matching and comprehensive long-term follow-up at a rate of 99.9%, provided the most robust observational controlled evidence available for the mortality benefit of weight loss surgery.
The 29% reduction in adjusted overall mortality is a finding of major clinical and public health significance. The causes of death analysis is particularly informative: reductions in both cardiovascular mortality and cancer mortality suggest that the protective effects of sustained weight loss extend beyond the expected haemodynamic and metabolic improvements. The cancer mortality reduction has been further elaborated in subsequent SOS analyses and in independent data, and there are now plausible mechanisms linking obesity-related hormonal and inflammatory pathways to cancer incidence and progression that may explain the surgical benefit. The differential between the surgical techniques, with gastric bypass achieving both the greatest weight loss (25% at 10 years) and the greatest metabolic improvements, is consistent with the later literature demonstrating that gastric bypass produces metabolic effects beyond weight reduction, including incretin-mediated changes in insulin secretion and glucose metabolism that occur before significant weight loss is achieved.
The SOS study is also the appropriate frame of reference for interpreting the weight loss magnitude achieved by tirzepatide (SURMOUNT-1, −20.9%), which for the first time in a pharmacotherapy trial approaches the sustained 10-year weight loss observed with gastric bypass in SOS (−25%). Whether pharmacologically induced weight loss produces the same long-term mortality benefits as surgical weight loss, given the potential differences in metabolic and hormonal mechanisms, remains an open and important question that the SURMOUNT-MMO and OASIS-6 long-term outcome trials will address.
Limitations: The study was not a randomised controlled trial; residual confounding from the matching process cannot be excluded. Surgical technique was not standardised, with gastric bypass, VBG, and banding mixed within the surgery group. The control treatment was conventional care rather than an active comparator lifestyle programme. The population was Swedish and predominantly white, limiting generalisability.
🔑 BOTTOM LINE
The SOS study demonstrated that bariatric surgery is associated with a 29% reduction in adjusted all-cause mortality compared with conventional obesity treatment over an average 10.9 years of follow-up, with reductions in both cardiovascular and cancer deaths, providing the first prospective controlled evidence that sustained surgical weight loss translates into meaningful long-term survival benefit.
⭐ Clinical Impact Rating: ●●●●○ Practice-changing (non-randomised design limits rating)
Next in the series: Study #17 DCCT: Intensive Insulin Therapy and the Long-Term Legacy of Tight Glycaemic Control in Type 1 Diabetes
