Further debunking of the BMI

Radley Balko's Morgan Spurlock Watch has some useful things to say about that public agenda tool, the Body Mass Index (BMI):

A huge part of the "ballooning" Spurlock speaks of has nothing to do with overeating. It's due to the fact that in 1998, the U.S. government redefined what it meant to be obese. The Centers for Disease Countrol lowered the bar. One magic night in 1998, then, 29 million Americans went to bed of "normal" weight, and woke up "overweight" — without ever gaining a pound. Millions more went to bed "overweight," and woke up "obese." That's not the fault of McDonalds or Frito lay, or Baskin-Robins. It's the result of an alarmist government moving the goalposts to manufacture hysteria.

[. . .]

Of course, none of these people's risk for these conditions increased overnight. The government merely drew a largely arbitrary line, and announced that one side of that line would now be healthy, and the other side wouldn't.

Of course, all of these statistics flow from the Body Mass Index, or BMI. BMI is by and large a completely arbitrary measure of health. It doesn't account for age, sex, gender, body type, or ethnicity. It also doesn't distinguish between fat tissue and muscle tissue (the latter is more dense). By now, you've probably heard about how big, muscle-bound athletes are classified as "obese" by the government. By BMI standards, more than half the NBA is obese or overweight. But in fact, any person who works out regularly is likely to fall into the "overweight" or "obese" categories. According to the government, for example, Johnny Depp is overweight. And Tom Cruise is obese. If your build is similar to theirs, you're probably obese or overweight, too [. . .]. Should give you an idea of how specious a tool the BMI really is.

Look at it this way: Muscle mass is denser than fat mass. If you've ever started a regular workout regimen after a few months of inactivity, you'll know that your weight tends to go up, not down, after the first few weeks. You're building muscle. Which means if ten people of normal build who don't exercise joined a gym, their collective BMI would go up, not down. But they'd be adding to the overweight-obesity statistics.

I've mumbled bits about the evils of BMI here, here, and here.