[throwaway98031]

The study is interesting and worth following up on, but the biggest takeaway from your post is your statement that:

> We don't know which way causality runs.

The study found a correlation, not a causation, and even the authors suggest that reverse causation is actually a very possible explanation.

Note that the study was performed in Sweden, where they do a much better job of rehabilitating offenders than we do in the USA. So good, in fact, that the reoffending rates in Sweden are half of other European countries. Sweden is actually closing prisons lately because their prison population is dropping so much.

I can't find any direct evidence to back this up, but it's not a stretch to think that young violent offenders in Sweden are much more likely than the general population to end up receiving psychiatric treatment, and thus are more likely to be prescribed SSRIs than the general population.

As such, I wouldn't read too much in that study.