BMI: what it measures and what it does not
The body mass index relates weight to height and yields a value from it. It was developed for population statistics, not for assessing individual people — and that is exactly where both its strengths and its limits come from.
Where it comes from
The formula goes back to the Belgian statistician Adolphe Quetelet, who in the 19th century studied how body measurements are distributed in populations. His aim was not assessing individuals but describing averages. The present name dates from the 1970s.
The index caught on because it needs only two easily obtained figures. For large surveys that is a considerable advantage: weight and height can be determined in thousands of people, body composition cannot.
That a measure from statistics was later used to assess individuals is not a fault of the formula. It is a shift of purpose — and it explains most of the objections raised against BMI today.
What the formula does
The calculation itself takes seconds and needs no further information. BMI divides weight in kilograms by the square of height in metres. The World Health Organization divides adults into ranges based on the result: below 18.5, from 18.5 to under 25, from 25 to under 30, and from 30 up. These boundaries are conventions, not natural constants — they were derived from population data and have been discussed repeatedly.
The result is usually given with one decimal place. That precision is arithmetically correct but says no more about the person behind it than a rounded number would.
What matters is what the formula does not contain: no sex, no age, no indication of what the weight consists of, and no information about distribution on the body. It cannot take these things into account because it does not know them.
Where it is useful
For its actual purpose BMI works well. Across large groups it reflects how body weight is distributed in a population and how it changes over decades. In healthcare it also serves as a quick classification measure, followed by further examination.
For an individual it works as rough orientation — especially when the value is far from the boundaries. A value at the edge of the range says more than one just either side of a line.
It is also useful where it is meant as a first step: it costs nothing, needs no device and can be determined in seconds. A measure with such a low threshold reaches people who would never undergo a more elaborate examination — and that is precisely where its practical value lies.
Where it misleads
Because the formula does not know what the weight consists of, it systematically misclassifies certain people.
- A lot of muscle mass. Anyone who has done strength training for years can have a BMI above 25 without anything following from that about fat percentage. This also affects many sports with a high strength component.
- Little muscle mass. Conversely a value can sit in the middle range although the proportion of muscle tissue is low. With increasing age, composition shifts without BMI having to change.
- Build and height. Because height enters squared while people do not scale uniformly, very tall and very short people tend to be classified differently.
- Distribution. Two people with the same BMI can carry weight distributed very differently. The formula does not distinguish this.
What complements it
Several simple measures provide information BMI does not have. Waist circumference and the waist-to-height ratio capture distribution and can be obtained with a tape measure. The FFMI relates fat-free mass to height and thereby separates what BMI lumps together — though it presupposes an estimate of fat percentage and inherits its uncertainty.
None of these measures replaces the others. Together they give a considerably more complete picture than a single number.
How to place it
BMI is a starting point, not a result. It answers a question roughly and quickly and replaces no closer examination. A value outside the middle range is a reason to look more closely — no more and no less.
Anyone who wants to know what their value means in their own case should discuss it with a doctor. A formula cannot do that, and neither can an app.
For your own trajectory the same applies as for any other measure: the change over months says more than the reading on a single day. And anyone keeping BMI alongside circumferences and training performance sees more quickly whether a change in weight comes from muscle or fat tissue.
The key points
- BMI was developed for population statistics, not for assessing individual people.
- The formula knows neither sex nor age nor what the weight consists of.
- People with a lot of muscle mass are regularly classified too high.
- The cut-off values are conventions from population data, not natural constants.
- Waist circumference and waist-to-height ratio capture the distribution BMI passes over.
- A value outside the middle range is a reason to look closer, not a diagnosis.
Last reviewed: 14.08.2026
Created by AI