Body mass index is the most widely used health metric in the world, and it’s also one of the worst. It was developed in the 1830s by a Belgian astronomer studying populations, not patients. He explicitly warned that it shouldn’t be used to assess individual health. Two centuries later, your doctor probably still uses it as a stand-in for “are you healthy.”
That gap between what BMI was built for and what it’s used for is the source of nearly every problem with the metric.
It can’t tell muscle from fat
BMI is just weight divided by height squared. It has no idea what the weight is made of. A linebacker with 8% body fat and a sedentary office worker with 35% body fat can have identical BMIs and be classified the same way. The linebacker is metabolically healthy; the office worker likely isn’t. The number is identical.
This isn’t a fringe edge case. Surveys of college athletes routinely show 50% or more classified as overweight or obese by BMI, despite measurable cardiovascular fitness and low body fat. On the other end, the “skinny fat” category — normal BMI but high visceral fat and poor insulin sensitivity — is associated with metabolic outcomes worse than someone with a higher BMI and good muscle mass. The metric flattens body composition into a single misleading number.
It ignores where the fat actually is
Visceral fat — the kind packed around your liver, pancreas, and intestines — drives most of the metabolic risk associated with obesity. Subcutaneous fat, the kind under your skin on your hips and thighs, is mostly metabolically inert. BMI cannot distinguish between them. Two people at a BMI of 31 can have radically different cardiovascular and diabetes risk profiles depending on fat distribution.
Waist circumference and waist-to-height ratio outperform BMI on almost every clinical outcome study. They take ten seconds longer to measure. The reason BMI persists isn’t that it works better; it’s that it’s already in the chart template, the insurance form, and the medical school curriculum. Inertia, not evidence.
Population tools versus individual diagnoses
The most generous defense of BMI is that it’s a useful population-level screen. That’s fair. If you’re studying obesity rates across a country, BMI’s simplicity makes it tractable. The error is treating that same metric as a personal diagnostic.
Better individual measures exist and aren’t expensive: waist-to-hip ratio, fasting glucose, triglyceride-to-HDL ratio, blood pressure, resting heart rate, and cardiorespiratory fitness. Any three of these together predict mortality risk better than BMI alone. Yet BMI gets the headline at every annual checkup, and the rest are buried on page two of the labs.
The bottom line
BMI is a 200-year-old population statistic doing the job of a personal health metric, and it does that job poorly. Use it as a rough screen at most. If your BMI flags as high or low, ask what your body composition, waist measurement, and metabolic markers actually look like. Those numbers tell a story BMI can’t.
Leave a Reply