It's also extremely suspect that a completely irrelevant measure ("hypocampus shape") was selected for this correlation rather than memory function. In the abs

It's also extremely suspect that a completely irrelevant measure ("hypocampus shape") was selected for this correlation rather than memory function. In the absence of a randomized trial, the dose-response relationship is one of the most important tools available for teasing out causality. If you can't show that increased dose strongly correlates with increased response, you likely are looking at population variables, not drug/lifestyle effects. It turns out that almost everything that gets people's hands wringing from fatty foods to chemicals with complicated names have almost flat dose-response curves. The ones that stand out as having strong dose-response curves are alcohol and especially smoking, which meets all of the causal criteria you want to see in a correlative relationship: http://www.skepticalraptor.com/skepticalraptorblog.php/correlation-does-not-imply-causation-except-when-it-does/ http://www.skepticalraptor.com/skepticalraptorblog.php/corre...