BMI Is a Population Statistic, Not a Personal Diagnosis

Body mass index was devised in the 1830s by the Belgian statistician Adolphe Quetelet, and it was designed specifically to describe the distribution of body size across a large population for statistical and actuarial purposes — never as a diagnostic tool for assessing any individual person's health. That original purpose explains most of the well-documented limitations people run into today when the same formula gets applied to an individual instead of a population: a statistical measure that usefully describes broad patterns across thousands of people can be a genuinely poor fit for many specific individuals within that population, without either fact contradicting the other. Understanding that BMI was never built for individual diagnosis is what makes its modern limitations make sense as a mismatch of purpose, rather than as inexplicable flaws in an otherwise sound diagnostic tool.

The formula uses exactly two inputs — height and weight — and nothing else, which means it has no way whatsoever to distinguish a kilogram of muscle from a kilogram of fat, or to know anything about where on the body any given weight is actually distributed. This is precisely why a muscular athlete or someone who does regular strength training routinely registers as "overweight" or even "obese" by BMI despite carrying very little body fat — dense muscle tissue weighs more than fat, and the formula cannot tell the difference. It is also why two people with an identical BMI, one carrying weight predominantly around the abdomen and the other predominantly elsewhere, can have meaningfully different cardiometabolic health risk despite an identical number, since abdominal fat distribution specifically correlates with different health outcomes than the same total weight carried elsewhere on the body — a distinction BMI has no mechanism to capture at all.

The commonly cited BMI category cutoffs — 18.5, 25, 30 — were derived largely from studies of European populations, and subsequent research has found that health risk actually rises at different BMI thresholds in other populations. Several health authorities across Asia now recommend lower thresholds specifically because cardiometabolic risk factors have been found to increase at a meaningfully lower BMI in those populations than the traditional cutoffs would suggest, which has driven adjusted category boundaries in national health guidelines in those countries. The standard categories also do not apply directly to children, who need age- and sex-specific growth charts entirely, to pregnant people, or to adults over roughly 65, whose body composition changes in ways the standard adult categories were never calibrated to account for.

Someone sitting comfortably inside the "healthy" BMI band can still have poor cardiovascular fitness and concerning metabolic markers, and someone above that band can have excellent ones — BMI is a cheap, easy-to-collect population-screening number, not a clinical diagnosis, and treating it as though it were either guarantees or rules out actual health is a mismatch between what the number is and what it is being asked to do. If a person is genuinely concerned about their weight or overall health, waist measurement, blood pressure, resting heart rate, and blood panel results are all individually more informative than BMI alone, and a doctor interpreting several of those together, in the context of a specific person's history, can say something meaningful in a way that a single population-level statistic never could on its own.