4 ms·

Yup I think they ran straight into a causal reversal.

• cortesoft 5y ago
  Or it could just be a correlation. Same thing causes both.
  • tgb 5y ago
    So the article tries to separate this out by using Mendelian randomization which supposedly fixes this since whether or not you inherit a genetic variant from your parent is a coin toss. However it has to make some other assumptions that are pretty hard to justify when it comes to something as complex as sleep schedules. There are some strategies to at least detect when things go wrong, though.
    • dcolkitt 5y ago
      Except the problem with Mendelian randomization is that genes that control circadian rhythm are also likely to affect mood.
      • tgb 5y ago
        Well, indeed, that is what their main finding is. The question is do they affect mood by means other than affecting circadian rhythm?
• unishark 5y ago
  There have been studies suggesting causality in both directions. The resulting vicious cycles you get from that are common in mental illness.