... but the effect likely has a low mean, a negative quadratic term, and a high variance. At least beer wasn't lumped into this study warning of increased risk of pancreatic cancer from drinking fizzy drinks.
I'll drink to these studies (but I don't really attach much significance to them).
[h/t JJ]