Dalrymple looks at a study comparing two radical approaches to weight loss:

Gastric bypass proved to have better results all round than gastric banding, except that there were a small number of deaths immediately after surgery…The question that this study did not answer, however, was whether these improvements resulted in a reduced death rate of those who underwent surgery compared with those who did not. For it is always dangerous to suppose that the reduction in the risk factors for something (such as death from heart attacks or strokes) actually results in a reduction of that something itself. What stands to reason in medicine turns out often not to be so reasonable after all.

I happened to read this paper while I was in France. No one can possibly say that the French do not enjoy their food but it is perfectly obvious that the proportion of grossly obese people in the population of France is very much smaller than that in America or Britain. This was an aspect of the question of obesity that the paper did not mention. There is no possibility that surgery will improve the deplorable dietary choices of so many Americans and British.