3 ms·

If you’re going to die soon (from the listed conditions), wouldn’t you be less likely to workout (especially with poor sleep)? I read the article and didn’t see

• handrous 5y ago
  You do have to watch out for that. IIRC that effect was behind the widely-reported "one drink a day is measurably better for you than never drinking!" thing, which turned out to be wrong—chronically ill people are more likely to teetotal, for obvious reasons, and also way more likely to die young than a healthy, drinks-sometimes peer—the causality was backwards, healthy people are more likely to drink. If anything, the actual results emphasize just how extremely bad drinking more than very occasionally is for your health, given at what low levels it overwhelms that large selection bias. [EDIT] IIRC it's also the mechanism behind the "people who walk slower have higher mortality rates" thing. It's because people who are already sick or feel bad walk slower, and people who are sick or feel bad are more likely to die, and that wasn't well-controlled-for, not because being a slower walker per se makes you more likely to die, at least not to such a high degree as was reported.