Heretofore, the criticism I'd heard about BMI was that it overstated the problem of obesity, which wasn't very believable. Forget statistics and just look around: Only a liar or a fool in denial could go to a mall, or an airport, or a stadium and not see lots of people who are overweight.

The source of the criticism also undercut it: It was the dishonestly named Center for Consumer Freedom (no link provided; please don't go there) that mocked the metric because BMI — body mass index, the calculation of height and weight upon which the government relies when it says that 2 of 3 American adults, and 1 of 3 American children, are obese or overweight — assessed elite athletes as being overweight.

The comment had the veneer of credibility — who could take seriously any measure that would put Lebron James and fat Uncle Morty in the same category? — until I realized that Lebron IS overweight, compared to the typical American, but because of heredity, frame, and sheer hard work, not for being a couch potato. The anomaly might require all the Uncle Mortys to decide which category they were in, but for that, I refer you to paragraph 1 above, about liars and fools.

But now comes a study by Nirav Shah of the NYU School of Medicine and Eric Braverman of Weill-Cornell Medical College that says the BMI understates the problem. (I learned about it at the Downey Obesity Report.)

“BMI significantly underestimates adiposity. A better cutpoint for obesity with BMI is 24 for females and 28 for males. [Currently, overweight is defined as a BMI between 25 and 30. Above 30 is considered obese.] ... Obesity, body fat and increased adiposity are more prevalent than the American public and American physicians are aware of."

Wow.

The authors came to their conclusion by using several other methods to assess the same question, including DXA, a direct measurement of body fat.

So there's your choice: a) BMI overstates overweight and obesity because it gets the freakish bodies of professional athletes wrong, or b) BMI understates overweight and obesity when compared to a direct measurement of body fat.