This journal, and even this column, has talked a lot about the SOS (Swedish Obese Subjects) study, not surprisingly so given its age and gravitas. It was the first long-term, prospective, controlled trial to provide information on the effects of bariatric surgery on obesity and its related comorbidities, and it was one of the first studies in history to demonstrate that deliberate weight loss in obesity is linked with reduction in mortality (Sjöström, 2008). Since then, this finding has generally only been repeated by bariatric surgery studies. The SOS study has its detractors – a bariatric surgery trial is impossible to randomise and double blind; the surgical procedures utilised in the SOS study are over 20 years old and relatively primitive, even obsolete, and more recent updates from the group have had only a small handful of participants left, allowing very little statistical power.
In comparison, for instance, the latest instalment of the Look AHEAD study of weight loss in over 5000 obese people with diabetes, précised on the opposite page, continues to produce remarkable findings. The latest conclusion is that weight loss by lifestyle alone induces significant increases in adiponectin, which improves insulin sensitivity, alongside anti-apoptotic and anti-inflammatory effects. Adiponectin is produced almost entirely by adipose tissue, but as the adipocyte mass expands, less is produced. Unlike the SOS study, there was no observed reduction in cardiovascular events or mortality after the intervention (Look AHEAD Research Group, 2013).
This instalment of the SOS study, summarised alongside, is important and looks again at the data when the trial was at its height to extract more subtle messages. It attempts to predict the patients who will benefit the most from surgery, and, therefore, who primary care should invest valuable resources in for best outcomes. The complicated trial design divided the original cohort into quartiles of baseline weight and further quintiles of degree of post-operative weight loss. Incidence and remission of diabetes at 2 years was assessed. The main conclusion is that starting BMI is irrelevant to remission of diabetes, i.e. it doesn’t matter how big a person is pre-operatively, it is the degree of weight loss from any starting point that is relevant to diabetes remission. This should prompt a rethink of referral pathways; patients’ size shouldn’t be the main criterion, rather the severity of their comorbidities and their degree of motivation to succeed. This concept fits neatly with the Edmonton Obesity Staging System, which prioritises treatment by degree of medical, psychological or functional comorbidities, rather than size (Sharma, 2009). It also resonates with guidance published by NICE (2014), who have been serially lowering the thresholds for referral for surgery, potentially to as low as a BMI of 30 kg/m2, so that young, moderately obese individuals with diabetes can potentially access life-changing treatment early, when it is most effective.
Of the SOS study, 27% of participants at any baseline BMI and who failed to lose any weight after surgery witnessed their diabetes enter remission. This possibly reflects the specific anti-diabetes effect of gastric bypass, which does not rely on mere weight loss for its effect. For participants who lost significant amounts of weight, a 97.1% remission rate occurred, suggesting that we need to select suitable patients carefully and perhaps open the conversation around bariatric surgery more widely to appropriate patients who may not tip the scales at 200 kg.
To view the summaries of each paper, please download the PDF of this article.
