Who lives longer, fat people or skinny people?
The answer, among adults in the United States, is the overweight. You don't believe me? I'm not surprised. We have been conditioned to think that lean is good, fat is bad. You can't be too thin or too rich.
For many years "desirable" weight was defined in tables prepared by numbers crunchers at The Metropolitan Life Insurance Company. Now the standard is BMI (Body Mass Index, weight in kilograms divided by height in meters squared). You can find your own BMI, using pounds and inches, at this link: http://nhlbisupport.com/bmi/ Normal weight is defined as a BMI of 18.5 to 24.9.
The best evidence I have found on the relationship of weight to the risk of death is a study which appeared in April 2005 (http://jama.jamanetwork.com/article.aspx?articleid=200731) with a followup in November 2007 (http://jama.jamanetwork.com/article.aspx?articleid=209359). A team led by Katherine Flegal, a respected and often cited scientist at the National Center for Health Statistics, reported that:
- risk of death was lowest in the overweight group (BMI 25-29.9)
- risk of death in the first category of obesity (BMI 30-34.9) was similar to risk of death in the normal category
- risk of death was highest in the underweight (BMI under 18.5) and most obese (BMI over 35) groups.
Not surprisingly there were cries of outrage. Walter Willett, a prominent researcher at the Harvard School of Public Health, called a press conference the day after the 2005 study was published and called it junk. Willett had drawn very different conclusions from the Nurses Study. His data, however, were drawn from a homogenous group, not a nationally representative sample, and from self-reporting on questionnaires, not directly measured heights and weights. (See Gina Kolata, "Study Tying Longer Life to Extra Pounds Draws Fire," The New York Times, May 27, 2005.)
Over time it has become clear that Flegal's study is not an outlier. Similar studies, with large representative data sets, were done in Australia and Canada, and yielded similar results. Numerous studies have found that within groups with the same diagnoses, heavier patients do better. Other studies have shown that fitness is a more important predictor of mortality than weight.
I don't think this evidence means we should ignore the obesity epidemic. Clearly something is wrong, especially as it affects children. I do think, as we think about our own health and public policy, we need to be aware that there are big cracks in the conventional wisdom.
