To get a sense for the dangers of taking large medical studies at face value, check out this comment thread over at Kevin's place. Kevin wrote up a new study that apparently proves Prozac and other antidepressants in the SSRI family are worthless. It's an interesting result, but a bit hard to believe for anyone who's ever seen a friend cycle through antidepressants till they land on the right pill and dosage and suddenly turn back into themselves. Many of Kevin's commenters, who've seen the same thing, begin instantly complicating the research, and by the end, it seems fairly certain that the research could be technically true but its result utterly misleading. It's worth reading through just to get a sense of how skeptically to treat this stuff.
