For nearly 1 in 5 Americans, BMI may tell the wrong story

Who's healthiest? BMI doesn't always answer that question accurately, say experts. The result: Unhealthy lean patients are often overlooked, while healthy fat ones get harangued.
(Jewel Samad / AFP/Getty Images)

Obese is bad and lean is good. End of story, right?

Wrong, say a pair of University of Pennsylvania physicians and obesity researchers who are calling for better ways to assess individual health prospects than the body mass index, or BMI.

The BMI--a simple calculation that can be done with a scale and a ruler--is cheap, simple and allows comparisons of broad populations across years, cultures and continents. Since it was invented by Belgian mathematician Adolphe Quetelet in 1832, it has been the basis for research that has pointed overwhelmingly in one direction: When the formula for BMI (weight in kilograms divided by height in meters squared) yields the number 30 or higher, the demarcation line for obesity, the risk of cardiovascular disease, Type 2 diabetes and certain cancers goes up--sometimes dramatically--compared with when that number falls below 25.

And that is statistically true for whole populations of people. But as a measure of personal health, a wealth of recent research has underscored that the BMI can be a pretty poor predictor.

For some with a BMI over 30, a welter of recent studies suggests obesity may confer some health benefits--a mystery widely referred to as the “obesity paradox.” Other new studies have shown that when some lean people develop diseases typically associated with obesity--such as Type 2 diabetes or cardiovascular disease--they’re more likely to die than if they had been fat with the same diseases.

Add to that that weight status and metabolic health are imperfectly correlated, and the picture becomes perplexing and surprising: The proportion of obese adults who are metabolically unhealthy is high: 21%, or roughly 50.4 million Americans who should unambiguously worry physicians and public health authorities. But 8% of normal-weight adults are metabolically unhealthy. And if the latest research is borne out, that’s another 19.2 million that physicians and public health authorities should be fretting about (but who currently do not raise alarms).


Meanwhile, research shows that 10% of the nation’s obese adults are metabolically healthy. That may not sound like much. But it does translate into 24 million Americans who are chubby but not in danger of dying early, and yet who are relentlessly harangued by employers, insurers, public health authorities and, yes, health and medicine reporters to lose weight (or else!).

Those numbers, says Dr. Mitchell Lazar, should give actuarial experts (not to mention public health authorities and physicians) pause: They think the BMI tells them who will die early and who will live long and prosper.

But people, Lazar said, may carry excess weight for a variety of reasons--genetic, behavioral, psychological, social--and they may be healthy or unhealthy, die early or live long, for a different constellation of reasons. And there is already growing evidence that certain forms of fatness--that which clusters around the middle rather than around the arms, bottom and legs, for instance--are more predictive of trouble ahead. BMI captures such distinctions quite imperfectly, Lazar said in an interview.

Lazar and Dr. Rexford Ahima, a University of Pennsylvania endocrinologist and obesity expert, write that “there is an urgent need for accurate, practical, and affordable tools for assessing body composition, adipose hormones, myokines, cytokines and other biomarkers as predictive tools” that would let physicians separate the fat-but-fit from those in danger, and to identify the trim ones who look healthy but are actually at risk of illness and early death.