Unhealthy Science

Whatever you do, don't miss my friend Gary Taubes excellent cover story in the NY Times magazine section today, "Unhealthy Science: Do We Really Know What Makes Us Healthy?".

A careful reading- which it deserves- will illuminate just why the majority of the studies that make headlines (i.e. "antioxidants of no benefit in heart disease", "St. John's Wort useless against depression", "cholesterol causes heart disease") are deeply, deeply suspect. Taubes- one of the best science writers in America- explains the difference between correlation (the fact that two things are found together) and causation (the idea that one causes the other) as well as I've ever seen it explained in a general article.

Bottom line: In Denmark, the more storks the more babies. Diabetes went up during the Clinton presidency. The best spellers in grade school have the biggest feet. These are all examples of correlations- associations- in which the result (babies, diabetes, spelling ability) is clearly not caused by the thing it's associated with. We need to look deeper. Those examples are obvious. The ones Taubes discusses are less so.

Worth reading. And remember it well the next time you hear something utterly and completely counterintuitive, like fruits and vegetables don't make any difference to your health or vitamins don't "do" anything.