Monday, November 19, 2007

Dr. Couch. Mr. Potato.

Just a reminder, I still adhere to the belief that the phrase “scientific study” should be given equal weight with the phrase “movie review”. This is because there is a whole lot of experiment free research being presented as science. With this caveat, the following work of statistical data analysis finds some unconventional wisdom.

The Weight Story No One Wants to Talk About: But in contrast to the cancer report, which received enormous and largely uncritical media attention, a new study about obesity by Katherine Flegal and colleagues from the Centers for Disease Control and the National Cancer Institute made few waves. Yet Flegal's quiet and careful study could do much to calm our growing national hysteria about obesity.

Flegal used data from the National Health and Nutrition Examination Survey, which is a representative sample of the US population, to find the connections between being underweight, overweight and obese and cardiovascular disease (CVD), cancer and many other causes of death. The results are startling since they confound much of the received wisdom about being fat in America.

Flegal discovered that being overweight (BMI's of 25-30) was not responsible for increased mortality. In fact for CVD, cancer and all other causes, being overweight actually increased one's chance of living longer. In total, overweight was associated with a total of 138, 281 fewer deaths. Being overweight is not likely to kill you.

There is a big difference between using experiments to produce reproducible empirical data to exclude alternative explanations, and the statistical manipulation of assembled sets of measurements. These latter calculations yield only correlations and probabilities. So when this Center for Disease Control report finds their collection of recorded numbers to be more of a dull epic than a moralistic horror story, it sounds plausible to me. These conclusions are much more in line with my personal experience than the endless waves of exaggerated hysteria from people actively searching for bad news, because bad news is good for getting headlines and grant funding.