I suspect the causality flows (mostly) in the other direction for that correlation. The set of healthy and active elderly people consists only of those who can

I suspect the causality flows (mostly) in the other direction for that correlation. The set of healthy and active elderly people consists only of those who can be healthy and active, while the set of sedentary / overweight elderly contains both those who choose not to be, and those who are kind of stuck with it due to chronic illness. Which aren't sharp categorical distinctions, mind you, but it virtually guarantees that the latter set will have shorter lifespans and worse end of life outcomes, even though plenty of nonagenarians are overweight (though very few indeed are obese). I stay in the game for what it brings me now. I can think of little so grim as going to the gym with the goal of living longer— but that's my personlity, someone else might find it motivating.