big study

and what they don't tell you is that in the experimental group, at least 50% didn't follow the regimen closely or at all. what we know about self-report behavior in sociology, is that people lie or won't tell you the truth. so if 50% actually admitted it, then there were probably more. while there's a large, i would argue that these findings probably are far from conclusive (there might not have been true differences between the experimental and control samples). there were other issues noted in the article, so i won't spend time going over my other methodological concerns. and they say social scientist aren't bad scientists...

(i'm not suggesting that people should/should not take supplements, just stating some methodological concerns)