[glenra]

Hang on, has a counterexample been presented? Maybe we're reading the challenge differently, but I parsed "reputable study on the web where the average person lost 10%+ of their body weight and kept it off for five years" as meaning a study in which a bunch of already-overweight people were selected to try some specific intervention (eg, "follow this diet & exercise program") and as a result of the intervention the median outcome was to lose that much weight and keep it off for 5 years.

A study of only people who succeeded, selected for the study because they succeeded, tells you nothing about how effective their particular strategy is.

For instance, suppose there exists a Grapefuit Diet which has the following effect: 1% of the people who try it lose weight, 2% of the people who try it gain weight, and 97% of the people who try it see no effect. If you take 100 people, tell them to try that diet, that is the result you'll get - a result which tells you the strategy is no good. But if you look at the national weight loss registry you'll only find people who were in the tiny 1% for which that diet worked. In fact, even if there are a LOT of Grapefruit Diet successes in the registry, that just tells you how popular the Grapefruit Diet was, it doesn't tell you if the Grapefruit Diet works.

What we want is an intervention study where a bunch of fat people do some specific thing and that thing is actually effective at producing a clinically significant and stable amount of weight loss. Ideally we'd want them to lose enough weight such that they are no longer "overweight" and then maintain that state.

Does that study exist? Is there any study meeting that set of criteria? My impression is that it does not; there is in fact no non-surgical intervention known to "work". Which explains why people keep grasping at straws to find options that plausibly might work.

[cm2012]

Spot on.