Body Mass Index, the one number that nearly everyone uses to categorize overweight and obesity, is pretty much garbage.
This isn’t actually controversial – scientists and doctors have known it for years – but BMI is convenient, and that means that everyone keeps using it, and every year it gets more entrenched. (Even certain journalists who should know better end up writing about BMI, thereby perpetuating this regrettable state of affairs.)
The claim that BMI is rubbish may not be a shocker, but the actual degree of BMI’s inaccuracy may still be surprising. A BMI of 30 kg/m2 is officially defined as “obese,” but when people are assessed using body fat %, an actually reliable biometric, we learn that BMI misses “more than half of people with excess fat.” While this result suggests that true obesity is even more prevalent than widely understood, merely decreasing the cut-off for the defined obesity doesn’t fix the problem, and would result in millions of people of healthy people now mischaracterized as obese.
To search the recent medical literature for studies on the performance of BMI as a diagnostic is to discover the many narrow ways in which researchers have repeatedly found that BMI, frankly, sucks. So far in 2020 we have learned that BMI “demonstrated relatively poor predictive value with respect to obesity classification”among college students; that it “demonstrated limited diagnostic accuracy for adiposity in Korea;” and even that it “was not a particularly accurate method for classifying weight status among Russian police officers.” One suspects it is similarly inadequate for Finnish hairdressers or Bolivian mechanics.
If there’s a single reason why BMI is so unreliable, it’s the simple fact that it cannot discern the difference between fat and muscle. In most people, these two body tissues have essentially opposite effects on overall health, but in BMI a pound is a pound and they count just the same. The most familiar example of how this can go wrong is the bodybuilder that gets absurdly categorized as “obese.”
If there’s a second reason, it’s that it cannot discern the difference between different types of fat. Visceral fat – the fat that accumulates deep within the belly around one’s vital organs – is extremely unhealthy. Subcutaneous fat – the jiggly fat that accumulates under the skin all around the body – is much less dangerous.
The result is that BMI is, in any one patient, deeply flawed as an estimate of body fat percentage, and even less useful as an indicator of overall health.
Perversely, BMI is actually most useless in assessing the very people that are in most need an accurate measure of metabolic health. After all, it doesn’t take a set of calipers to identify the truly fat and the obviously fit among us. It’s those in the middle that need a more reliable measure, the very people that may or may not have a visceral fat problem that they’re not likely to learn about at the doctor’s office.
There are now millions of Americans with a supposedly healthy body weight (according to BMI) that have visceral obesity – a body type popularly known as “skinny fat.” In fact, visceral obesity is growing even faster than generalized obesity, which is itself rising at an alarming rate. And remarkably, according to a 2019 survey, the combination of high belly fat and regular body weight “has a more profound effect on morbidity and mortality” than does obesity itself:
“… the mortality risk of those with central obesity at normal BMI is similar to or greater than the risk of those with central obesity who are overweight or obese… The implication of this is that normal-weight centrally obese people might be at uniquely high risk of all-cause mortality.”
Unfortunately, this population is difficult to analyze because BMI, the one measure that everyone uses, cannot identify it by definition.
But BMI is probably most useless for the elderly. In old age, we lose both muscle and body fat, and what body fat we retain tends to become more heavily concentrated in the abdomen. In other words, the elderly body metamorphoses inexorably in the direction of “skinny fat.”
That’s enough to make BMI especially untrustworthy for the elderly, but a second factor is the fact that muscle mass and strength actually become increasingly important as we age.
Although few young or middle-aged adults consider low muscle mass and strength to be desirable, it’s also fair to say that it’s not generally seen as an indicator of really poor health. As we get older, though, that changes dramatically. Among the elderly, improved muscle strength is strongly correlated with both longevity and quality of life, not to mention insulin sensitivity.
That the unhealthiest type of fat tends to fly under BMI’s radar, and that BMI mistakes muscle for fat, may help explain the so-called “obesity paradox” – the consistent finding that people that are overweight actually live longer than those at normal weight.
Why does BMI persist? Although its flaws are readily apparent, BMI has proved frustratingly enduring because it is so simple. It’s a snap to measure and track, and any one statistic is easier to share, understand and analyze than a constellation of different measures. It became and will remain the standard for scientific population studies because the data is so ubiquitous. And it is probably at least somewhat useful for those sorts of studies, because over a large enough sample, BMI does tend to correlate with body fat percentage. For any individual purpose, though, it’s practically useless.
Some readers may be amused or appalled to learn that BMI – devised in the 1830s by Lambert Adolphe Jacques Quetelet, a Belgian mathematician – was popularized by Ancel Keys, one of the great bogeymen of the low-carbohydrate and keto communities. Keys was the hugely influential nutritionist that convinced the world that dietary saturated fat caused heart disease, an assertion that has since been discredited.
What’s better than BMI? There are many other options, some requiring expensive imaging equipment, others needing nothing but a tape measure. If you’re trying to assess your own health risks at home, it’s tough to beat waist-to-height ratio, which is easy to analyze and is more accurate than BMI.