[derbOac]

Yes, there was at least one meta-analysis several years ago that compared different diets to each each other and to exercise. I think there were actually two meta-analyses that came out about the same time, with slightly different foci, and came to similar conclusions.

What I remember was that diets didn't differ from one another in efficacy once you controlled for caloric restriction, and that dieting was more effective than exercise, although exercise did have a small but quantifiable effect.

I also think one of the meta-analyses compared atkins-style diets to others and concluded that there was some slight evidence for their superiority (in terms of weight loss) in the short term but that after a few weeks they didn't differ from the others.

I've worked in bariatric surgery units and my personal sense from that and reading some of the literature is that a lot of what works best is highly individual-specific due to all sorts of reasons, including personal food preferences, genetics, microbiome, etc etc etc. The standard explanation for diet working better than exercise is that the extra caloric expenditure of exercise gets washed out by excess intake; I think this is undoubtedly true but after having worked clinically in obesity settings, I also think it's really astonishing how little exercise some people get. That is, again, I think some people who start adding a modicum of exercise are likely to see huge gains from it simply because their caloric expenditure is starting so low; conversely, most people who take up exercise probably exercise within a certain range, so it doesn't say a lot about the expectable weight loss of someone who is really exercising intensely.