A study in the current Journal of the American Medical Association suggests that overweight people are not really overweight. Analyzing data from the National Health and Nutrition Examination Survey, researchers at the CDC and the National Cancer Institute found that mortality is lowest among people the government considers "overweight" and highest among the extremely obese. People of "normal" weight had a higher death rate than people who were a little fatter.
That is very different from the findings of earlier studies, one of which was described in a front-page New York Times story headlined "Even Moderate Weight Gains Can Be Deadly." The new study also differs sharply with previous estimates of premature deaths due to excessive weight. Other researchers at the CDC had put the number at 400,000 a year, which was later downgraded to 365,000 because of a calculation error. The new study ties about 112,000 deaths to obesity but finds that extra pounds in the "overweight" range prevent about 86,000 deaths a year, leaving a net loss of 26,000 or so–lower than the 34,000 deaths tied to being "underweight" (i.e., below "normal").
"Counting deaths is not an exact science," says the CDC's chief science officer. No kidding.
Since the alleged death toll of several hundred thousand has been repeatedly cited to justify a paternalistic crusade against overeating, I'd like to believe the new numbers are more accurate than the old ones. But I don't really know. The important point is that neither does the government, notwithstanding the confident pronouncements of public health officials.
Policy implications aside, the suggestion that health risks from excess weight don't show up until you're really, really fat should come as a relief to those of us who tend to toggle between "normal" and "overweight." Then again, the possibility that a few extra pounds are protective may force us to choose between health and vanity.