200,000 person study. I think we can assume they've taken care of the usual confounding issues. (ok, somebody is going to say "no, we can't" but I do think this

200,000 person study. I think we can assume they've taken care of the usual confounding issues. (ok, somebody is going to say "no, we can't" but I do think this will be more than just correlative.)