BMI and waist size: what they say and what they don't

BMI is weight in kilograms divided by height in metres squared. Example: 80 kg at 1.75 m gives about 26.1.^[2]

This classification applies to adults.

• Fat and muscle: It does not tell fat, muscle and bone apart and says nothing about distribution. Very muscular people can have a high BMI.^[2]
• Accuracy: In 25 studies with 31,968 people, it missed about half of those with raised body fat. A high value, however, was mostly correct.^[3]
• Age: The British NICE advises reading BMI with caution from 65.^[1]
• Background: For people of South Asian, Chinese, Middle Eastern or African background among others, NICE uses lower thresholds, with obesity starting at 27.5.^[1] The German S3 guideline names no values of its own.

Common values are over 94 cm for men and over 80 cm for women for increased risk, and over 102 and 88 cm for substantially increased risk.^[4] They go back to a study in Glasgow, and according to a WHO expert group they are no formal WHO recommendation.^[4]

Measure halfway between the lowest rib and the top of the hip bone, with a tape that does not stretch, after breathing out calmly.^[4]

Divide your waist by your height, in the same unit. NICE classifies it like this: 0.4 to 0.49 no increased risk, 0.5 to 0.59 increased, from 0.6 further increased, in adults with a BMI under 35. The advice is to keep your waist below half your height.^[1]

In reviews with over 300,000 adults, the ratio predicted metabolic risks slightly better than BMI.^[5] Even so, the German guideline only rates measuring it as an optional recommendation because the evidence is low.^[2]

In US data on 40,420 people, almost half of those with overweight and 29% of those with obesity were healthy by metabolic values, and over 30% of those with normal weight were not.^[6] So BMI often misclassifies individuals.

That does not mean free of risk: in 22 studies, healthy obesity went with a 1.45-fold risk of cardiovascular events.^[7] In a large analysis, the lowest risk of death was at a BMI of 20 to 25.^[8] An earlier analysis found no higher risk with overweight.^[9] So research does not fully agree.

Since 2025, a Lancet commission of 58 experts has separated clinical obesity, with impaired organ or daily function, from pre-clinical obesity, with preserved function and raised risk. According to it, BMI is suited to population data and as a screening test, not as a measure of one person's health.^[10] It does not abolish BMI and does not say that obesity without symptoms is harmless.

In Germany, by self-report, 53.5% of adults have overweight and 19.0% obesity, and measured values are higher.^[11] So you are rarely alone with your number.

According to the S3 guideline, these measures are not enough to decide in a single case. Other conditions and how you feel matter more.^[2] If your BMI is under 18.5, or if such numbers occupy you too much, read When to get medical advice.