BMI isn’t bullshit, but it’s not the whole story either. It’s a rough screening tool that works reasonably well across large populations and poorly for many individuals. The formula, your weight divided by your height squared, was never designed to diagnose anything in a single person. It was invented in 1832 by a Belgian statistician named Adolphe Quetelet who wanted to define the characteristics of the “average man” using bell curves. It was a population-level statistics tool from the start.
Even the physiologist who later popularized the term “body mass index” in the 1970s, Ancel Keys, didn’t intend it for medical use. He described it as a simple, easy-to-obtain measurement useful in research settings. Somehow, it became the default number your doctor uses to decide whether you’re a healthy weight. That gap between what BMI was built for and how it’s actually used is where most of the legitimate criticism lives.
What BMI Gets Right
At the population level, BMI correlates meaningfully with body fat. When researchers compare BMI against actual body fat measurements across thousands of people, the statistical overlap is solid. The relationship between BMI and all-cause mortality follows a U-shaped curve: people at the extremes (very low or very high BMI) face higher health risks, while those in the 22 to 26 range tend to have the lowest mortality rates. For older adults, that sweet spot shifts slightly higher.
If you have a BMI of 45, it’s extremely unlikely that you’re carrying nothing but muscle. At the high and low ends, BMI is telling you something real. The tool also has genuine practical value: it requires no equipment, no blood draw, no body scan. You can calculate it from a bathroom scale and a tape measure, which is why it became so widespread in the first place.
Where BMI Falls Apart
The problems start when BMI is applied to you as an individual rather than averaged across a population. The American Medical Association formally acknowledged this in 2023, adopting a policy stating that BMI “is significantly correlated with the amount of fat mass in the general population but loses predictability when applied on the individual level.” That’s a polite way of saying your personal BMI number can be genuinely misleading.
The most obvious case is athletes and anyone with significant muscle mass. Muscle is denser than fat, so a person who strength-trains seriously can land in the “overweight” or even “obese” BMI category while carrying a perfectly healthy amount of body fat. This isn’t a rare edge case. It affects a meaningful portion of active people.
Then there’s the flip side: people with a “normal” BMI who actually carry too much fat relative to their muscle mass. This is sometimes called “skinny fat” or, in clinical terms, normal-weight obesity. These individuals may fly under the radar on a BMI check while facing real metabolic risks from excess visceral fat packed around their organs.
The Racial Bias Problem
BMI thresholds were built primarily on data from non-Hispanic white populations, and the AMA’s 2023 policy specifically called out “its historical harm” and “its use for racist exclusion.” The issue isn’t abstract. Different ethnic groups develop metabolic disease at very different BMI levels, and using universal cutoffs misses this completely.
A large population study in England found that South Asian populations reach the same type 2 diabetes risk at a BMI of about 24 that white populations reach at a BMI of 30. That’s a six-point gap. For Black populations, the equivalent cutoff was around 28. This means a South Asian person with a BMI of 25, technically just barely “overweight,” already faces obesity-level diabetes risk by the standards that matter. Meanwhile, using the standard cutoff of 30 to flag concern would miss them entirely.
These aren’t small differences. They reflect genuine biological variation in body composition, fat distribution, and metabolic response across populations. A single set of thresholds applied to everyone is going to systematically over-diagnose some groups and under-diagnose others.
Metabolically Healthy Obesity Is Real, but Rare
Some people with a BMI over 30 show no signs of metabolic syndrome: their blood pressure, blood sugar, and cholesterol levels are all normal. This is called metabolically healthy obesity, and it’s a real phenomenon. But it’s less common than you might hope. Among adults with obesity in the U.S., only about 10 to 15 percent qualify as metabolically healthy by strict criteria, and that proportion has only modestly increased over the past two decades.
So while BMI absolutely misclassifies some individuals, the majority of people in the obese BMI range do show at least one metabolic risk factor. Being in that category doesn’t guarantee you’re unhealthy, but the odds aren’t in your favor either.
How BMI Affects Your Insurance and Care
One of the most concrete ways BMI shapes people’s lives is through the healthcare system. Insurance coverage for bariatric surgery, for example, requires specific BMI-based diagnosis codes. Medicare coverage criteria tie approval directly to BMI thresholds of 35 and above (with related conditions) or 40 and above. If your BMI falls just below the cutoff, you may be denied a procedure your doctor recommends, regardless of your actual metabolic health.
The AMA explicitly stated that BMI “should not be used as a sole criterion to deny appropriate insurance reimbursement,” but the billing infrastructure hasn’t caught up. In practice, a single number still gates access to treatment for many patients.
Better Ways to Measure Health Risk
If BMI alone isn’t enough, what should you pay attention to instead? The simplest upgrade is your waist-to-height ratio. The rule is easy: keep your waist circumference below half your height. If you’re 5’8″ (68 inches), your waist should stay under 34 inches. That 0.5 ratio works as a screening threshold across ages, sexes, and ethnic groups, which is something BMI can’t claim. Risk increases noticeably once the ratio passes 0.6, particularly for cardiovascular disease in people with diabetes.
A newer metric called the Body Roundness Index, which factors in waist circumference and height, identifies mortality risk differences across more categories than BMI does. In a large cohort study, BRI flagged meaningful risk variation in four out of five groups, while BMI only caught differences in three. The practical takeaway: measures that account for where you carry fat, not just how much you weigh, tend to perform better.
The AMA now recommends pairing BMI with measures like waist circumference, body composition, and metabolic markers such as blood sugar and cholesterol. No single number captures the full picture. A combination of a tape measure around your waist, basic blood work, and yes, your BMI, gives a far more accurate read than any one of those alone.
The Bottom Line on BMI
BMI is a blunt instrument being used for precision work. It’s not worthless: across large groups, it tracks with body fat and predicts health risks in a general way. But it was never designed to tell any one person whether they’re healthy, and it performs worst for athletes, older adults, and non-white populations. The formula itself isn’t the problem so much as the way it’s been elevated to a definitive verdict on individual health, a role its creator never intended and that the best available evidence doesn’t support. Think of it as one data point, not the answer.