Articulatory Loop

Published in today's JAMA, a large study indicates that what we all think we know about fat and body weight and whatnot may be less than useful. Using a methodology developed for breast cancer research, the authors examined BMI and mortality data from three runs of the National Health and Nutrition Examination Survey (NHANES), a large, nationally representative sample. Unlike most (all?) previous studies, this group controlled for age, smoking, race, and gender - all potential confounds. The statistics look robust to me, though I still have qualms about BMI as a particularly useful measure: it is calculated as weight divided by height squared, which could easily place a bodybuilder in the same category as a champion couch potato. The findings are fascinating nonetheless.

They found that being a little bit overweight - with 25<>35 tended to increase it. The New York Times shockingly, blows these data way out of proportion.

Gratifyingly, the relative risks in the 25-59 age group were comparable for the underweight and severely obese (BMI>35) groups.

So: if a BMI between 25 and 30 is optimal for health, can we now stop calling it overweight??? Also, since having a BMI of less than 18.5 is unhealthy, can we please start feeding our supermodels???