BMI is a 200-year-old metric — what it actually tells you (and what it doesn't)
BMI was invented in the 1830s by an astronomer to study population averages. Here's where it works, where it fails, and what to use instead.
# The origin
Adolphe Quetelet was a Belgian astronomer and statistician in the 1830s. He wanted to characterise the "average man" — l'homme moyen — and noticed that body weight in adults of the same sex scaled roughly with height squared. He published the ratio that became BMI in 1832, not as a diagnostic tool for individuals but as a tool for studying populations.
For 140 years almost nobody used it. Then in 1972, the American physiologist Ancel Keys ran a study comparing several weight-for-height formulas and picked Quetelet's as the simplest and most strongly correlated with body fat in his sample. He renamed it the "body mass index". The WHO adopted it as a population-level obesity metric in the 1980s. Suddenly a 19th-century statistical curio was being printed on every doctor's chart.
# What BMI is
BMI = weight (kg) / height² (m)
or, in US units:
BMI = 703 × weight (lb) / height² (in)
That's it. No body fat, no muscle mass, no waist measurement, no age. Two numbers in, one number out. The BMI calculator does this conversion and adds the WHO category labels.
# Where BMI works
For large populations — millions of people, broadly similar ages, mixed activity levels — BMI correlates well with cardiovascular risk, type-2 diabetes risk, and obesity-related mortality. Public health agencies use it because it's cheap (no tools needed), reproducible, and the population-level signal is real.
For an individual who's sedentary or moderately active and has an average-for-population body composition, BMI gives a decent first approximation of where they sit on the weight-for-health curve.
# Where it fails
Muscular individuals get flagged "overweight" or "obese" by BMI when they're nothing of the sort. A 95 kg rugby player at 12 % body fat is "obese" by BMI (typically 30+) but in better cardiovascular shape than the average person at BMI 22. The Rock has a BMI of 31. So does Tom Brady at his peak. So do almost all NFL linemen, NBA centres, Olympic weightlifters.
The reverse failure is "skinny-fat" — low BMI, high body fat percentage. A sedentary 65 kg adult at 5'10" has a BMI of 20.5 (within "normal") but might carry 28 % body fat with weak musculature and high visceral fat. Their health risk profile looks worse than a 75 kg person with the same body fat but more muscle.
Other failures:
- Different populations have different healthy ranges. Asian populations show metabolic disease risk at lower BMIs than European-derived populations; many Asian health authorities use a 23 cutoff for "overweight" and 27.5 for "obese", not WHO's 25 and 30.
- Age changes the calculation. Older adults have a higher healthy BMI range (some studies suggest 25-30 is optimal for people over 65) because lean mass declines and slightly elevated BMI protects against frailty.
- Children and teens use a different scale entirely. Adult BMI thresholds don't apply; paediatric BMI uses age-and-sex percentile curves.
- Pregnant and post-partum — BMI doesn't apply meaningfully.
# What to use instead (or alongside)
The single best replacement for BMI as a screening metric is waist-to-height ratio. Stand up straight, measure your waist at the navel, divide by your height. Below 0.5 is the broadly healthy range; above 0.6 is high cardiovascular risk. The math is just as cheap as BMI and the correlation with metabolic disease is stronger, especially across non-European populations.
For a more complete picture:
- Body fat percentage — measured by DEXA scan (gold standard, ~$50-150 in many cities), bioimpedance scales (cheap but ±5 % accuracy), or the US Navy circumference method (free, ±3-4 %).
- Waist circumference alone — > 94 cm (men) or > 80 cm (women) is the WHO cardiovascular risk threshold.
- Strength markers — grip strength, sit-to-stand test. Strong correlation with all-cause mortality, especially as you age.
- Resting heart rate and blood pressure — five-minute checks that say more about cardiovascular fitness than a year of BMI tracking.
# So should I check my BMI at all?
Yes — as one number among several, not as a diagnosis. BMI is a cheap, fast first screen. If yours is in the "normal" range and your waist-to-height ratio is also below 0.5, that's reassuring. If your BMI is high but you're visibly muscular and your waist measurement is fine, BMI is wrong about you. If your BMI is low but you carry visible abdominal fat, BMI is also wrong about you.
The mistake isn't using BMI; it's using only BMI.
# A worked example
A 5'10" (178 cm), 180 lb (82 kg) adult.
- BMI: 82 / (1.78)² = 25.9 → "overweight" by WHO categories
- Body fat (Navy method, 38" waist, 16" neck): ~24 % → moderate, not high
- Waist-to-height: 38" / 70" = 0.54 → slightly elevated, worth addressing
- Resting heart rate: 62 bpm → good
The BMI alone says "overweight". The full picture says "borderline, with abdominal fat as the actionable issue, not body mass in general". The intervention should target waist circumference (cardio, dietary change), not weight loss per se.
# Bottom line
BMI is a 19th-century statistical tool that's still useful in 2026 as a quick screen — but only as the first of several numbers. If your BMI says one thing and waist measurement, body composition, or strength markers say another, trust the other numbers. They have less population-level legibility but more individual accuracy, which is what you actually care about.
# Related tools
Common questions
Frequently asked.
›What's a healthy BMI range?
The WHO defines: underweight < 18.5, normal 18.5-24.9, overweight 25-29.9, obese 30+. These thresholds are population averages, not individual diagnoses — see below for why.
›Why does BMI overestimate body fat for muscular people?
BMI treats all mass as equivalent. Muscle is ~18 % denser than fat, so two people of the same height and weight can have wildly different body compositions. A 95 kg rugby player at 12 % body fat and a 95 kg sedentary office worker at 30 % body fat get the same BMI.
›Should I use waist-to-height ratio instead?
It's often a better single number. Waist-to-height ratio < 0.5 correlates better with cardiovascular risk than BMI, especially for non-European populations. The Lancet Commission recommended it as a primary screening metric in 2023.