Hacker News Reader

Note, some of these are studies of humans and some are on mice

Sorry, none of these establish causality, only very mild (probably specious) correlations. One of them is only in cancer patients which is basically worthless when discussing the general population. The rest are in mice, which doesn't translate to humans in general, especially not where psychological issues like anxiety are concerned. There is definitely some possibility for future research though. Even a single replication of any of these studies would be helpful. But so far there's no reason to think a causal link exists.

I agree that causality isn’t established in the human studies. But with the population studies and the mice studies where causation is established that is a strong suggestion for further research as you said. If I were a betting man I would bet a lot of money on this link though, based on my experience and these studies (and there’s more). I think you would too, to a lesser extent :)

And I’m pretty sure one of the human studies does establish causation, although I don’t have time to see which one

Reply on news.ycombinator.com