BMI Was Never Meant to Diagnose You

Adolphe Quetelet was a Belgian mathematician and astronomer who, in the 1830s, became obsessed with applying statistics to humans. He wanted to define the average man — l'homme moyen — and one piece of that project was a simple ratio: weight in kilograms divided by height in meters squared. Quetelet wasn't a doctor. He never claimed the number diagnosed anything. He used it to compare groups.

The formula sat largely unused for over a century. Then in 1972, the physiologist Ancel Keys — the same Ancel Keys behind the Mediterranean diet and the K-ration — published a paper comparing several measures of body fatness across thousands of men in five countries. Quetelet's index correlated with skinfold-measured body fat about as well as the more complicated alternatives, and it was much easier to compute. Keys gave it a new name: the body mass index.

The insurance industry and then the WHO seized on it. In 1985 the NIH adopted BMI cutoffs to define overweight and obesity. In 1998 it lowered the overweight threshold from 27.8 to 25, reclassifying about 30 million Americans as overweight overnight. None of them had gained a pound.

The formula's weaknesses are well documented. It treats muscle and fat the same, which is why an NFL running back and a couch-bound office worker can post identical BMIs. It was tuned on European male populations and consistently misclassifies muscular men as overweight and slender South Asians as healthy when their metabolic risk is actually elevated. It says nothing about where fat sits — abdominal fat is more dangerous than the same kilogram on the hips, but BMI cannot see the difference.

None of this is a secret. Researchers know. The formula survives because it is cheap, requires no equipment, and gives clinicians a single number for the chart.