The headline is broadly defensible, but the qualifications matter. Effect sizes vary by population, the strongest claims rest on shorter trials, and credible voices push back on how it's typically framed.
one of them that i particularly like was developed by andrew stokes and his colleagues and that is the maximum attained weight method so instead of using the exposure variable except instead of looking for bmi right now and seeing that how that correlates with mortality risk right now you say what's the heaviest you've ever been and how does that correlate with your health outcomes
Every Sunday: the week’s new conflicts and verdict changes — and nothing else.
Native comments, Twitter mentions, and Reddit threads about this claim — surfaced together so the conversation isn't fragmented across platforms.
Bookmarking — the dossier-vs-overview split is the right call. Most of the time I want overview; sometimes I want receipts.
Would love a "what would change this verdict" RSS feed. Sign me up if it exists.
and when you look at it that way you get a sharpening of the association between bmi and mortality and you find that the nader the lowest point shifts to a lower bmi so essentially whereas before it looked like maybe overweight was the best now you see that in the lean range looks the best and there's a stronger difference a larger difference between being lean and having obesity in terms of mortality so it really sharpens things up