I'm increasingly convinced that studies on correlation without establishing at least some form of concrete causation actually results in more misinformed people than before. For all the speak of 'adjust' or 'control' for factors, it's very rare for studies to significantly cancel out the noise of wealth disparities (among other factors), rare enough that it's borderline impossible to know if any correlational study isn't just "wealthy people live better".
Not that studies like this aren't interesting, or aren't worth doing. But concluding anything more than "huh, interesting, needs much tighter studying" is just not reasonable.