What your BMI does, and the bits it quietly leaves out

BMI has to be the most quoted health number that almost nobody can properly explain. Weight over height squared, a threshold at 18.5, another at 25, another at 30, done. GPs use it, insurers use it, gyms print it on a poster by the scales. And yet it was never built to tell an individual person very much at all.

None of that makes it useless. It is cheap, quick, and needs nothing more than scales and a tape measure, which is exactly why it spread everywhere. The trouble starts when it gets treated as a diagnosis rather than a screening question.

What the formula actually measures

The maths is short: weight in kilograms, divided by height in metres, squared. A man who is 1.78m tall and weighs 80kg gets 80 ÷ (1.78 × 1.78), which comes out at 25.2, just over the line into "overweight". Change nothing about him except swap 10kg of fat for 10kg of muscle, and the calculator has no way of noticing. It sees a number on a scale and a number on a tape measure, and nothing else.

That is the whole of it. There is no age term, no sex term, no measure of where the weight sits on the body. A Belgian statistician called Adolphe Quetelet worked out the ratio in the 1830s while trying to describe the "average man" for population statistics, not to size up any one patient. It only picked up the name Body Mass Index and started appearing in individual health checks more than a century later.

Where it breaks down

Three groups get let down by the formula fairly reliably.

Muscular people. Muscle is denser than fat, so it takes up less room per kilogram. Rugby players, powerlifters, and plenty of people who just lift weights seriously can land in the "overweight" or even "obese" band on BMI alone while carrying a body fat percentage a doctor would be pleased with. The number is technically correct and practically misleading.

Older adults. Muscle mass drifts downward with age even when weight on the scale barely moves, replaced quietly by fat. Someone can have the identical BMI at 70 that they had at 30 and be carrying a meaningfully different body composition. BMI cannot see the swap happening.

Different ethnic backgrounds. This one is not a minor footnote. Research consistently shows that people of South Asian heritage tend to carry higher cardiovascular and diabetes risk at a lower BMI than the white European population the standard 25 and 30 thresholds were calibrated on. The NHS now flags overweight for South Asian adults from a BMI of 23 rather than 25, and obese from 27.5 rather than 30, precisely because the standard bands understate risk for that group.

What none of these three have in common with BMI is location. A number on a chart cannot tell you whether extra weight sits around the hips, which is broadly a lower-risk pattern, or around the waist and organs, which is the pattern most strongly linked to heart disease and type 2 diabetes. Two people with an identical BMI can carry genuinely different risk once you look at where the weight actually is.

The categories, and why the lines are fuzzy

The standard adult bands most UK sources use are underweight below 18.5, healthy weight from 18.5 to 24.9, overweight from 25 to 29.9, and obese from 30 upward. They are convenient rather than precise. Someone at 24.8 and someone at 25.1 are, in any meaningful physiological sense, the same person, yet one sits in "healthy" and the other in "overweight" purely because a line was drawn between them.

Treat the categories as a rough zone rather than a verdict. Sitting well inside the healthy band, say a BMI of 22, is a reasonably strong sign things are fine. Sitting right on a boundary tells you almost nothing on its own, and is exactly where it is worth looking at the numbers around it instead of the one number itself.

What actually fills the gap

You do not need a lab to get a fuller picture. A tape measure round the waist, taken at the belly button and not pulled tight, adds real information BMI cannot give you, because waist size tracks visceral fat, the kind packed around the liver and other organs, far better than a height and weight ratio ever could. As a rough marker, health guidance generally flags increased risk above 94cm for men and 80cm for women, and further increased risk above 102cm and 88cm.

Body fat percentage, if you can get a reasonably consistent reading from a decent set of smart scales or a proper caliper measurement, is another useful cross-check, particularly if your BMI sits in the overweight band but your waist measurement and mirror both disagree. And simple trend watching, the same measurements taken the same way every few weeks, tells you more about direction of travel than any single snapshot figure.

None of this is an argument for ignoring BMI. It is quick, it is free, and at a population level the link between higher BMI and higher average health risk is well established and not in serious dispute. The argument is narrower: for one specific person standing on one specific set of scales, it is a first question, not a final answer.

This guide is general information, not medical advice. If you have concerns about your weight or health, speak to a GP or a qualified healthcare professional.