What BMI actually measures
BMI was devised in the 1830s by the Belgian statistician Adolphe Quetelet, who was studying populations rather than patients. It became a clinical shorthand in the 1970s because it needs only a scale and a tape measure and correlates reasonably well with body fat across large groups.
The categories used worldwide come from the World Health Organization:
- Under 18.5 — underweight
- 18.5 to 24.9 — healthy weight
- 25.0 to 29.9 — overweight
- 30.0 and above — obesity, split into classes I, II and III
Where BMI misleads
BMI is a ratio of mass to height squared. It cannot tell what the mass is made of, and it makes no distinction about where it sits.
- Muscular people are routinely classed as overweight or obese. A lean rugby player or powerlifter can sit above 30 with single-digit body fat.
- Older adults lose muscle and gain fat at the same weight, so a "healthy" BMI can hide a poor body composition.
- Ethnic differences are real. Health risk begins at a lower BMI in South Asian, Chinese and other Asian populations; many guidelines use 23 as the overweight threshold rather than 25.
- Fat distribution matters more than total fat. Visceral fat around the organs carries far more cardiometabolic risk than subcutaneous fat, and BMI is blind to the difference.
A better free screening pair is BMI plus waist-to-height ratio. Keeping your waist under half your height is a simple, well-evidenced target that captures the visceral fat BMI misses.
What the categories actually predict
BMI persists in medicine because it correlates with outcomes across large populations — not because it describes any individual well. Understanding that distinction is the difference between using it and being ruled by it.
At a population level, higher BMI is associated with increased risk of type 2 diabetes, hypertension, cardiovascular disease, several cancers, sleep apnoea and osteoarthritis. Very low BMI carries its own risks: reduced bone density, impaired immune function, fertility problems and poorer recovery from illness. The risk curve is roughly J-shaped, with the flat part sitting across a fairly wide middle.
Three things that curve does not tell you:
- It is an association, not a mechanism. Fitness, diet quality, sleep, smoking and socioeconomic factors travel alongside weight and account for a substantial share of the risk attributed to it.
- Metabolic health varies within every category. A meaningful minority of people in the “overweight” band have entirely normal blood pressure, glucose and lipids, and some in the “healthy” band do not.
- Weight cycling has its own cost. Repeatedly losing and regaining is associated with worse outcomes than a stable higher weight, which is a strong argument against crash dieting.
The most useful follow-up to a BMI reading is not a diet — it is a set of actual measurements: blood pressure, fasting glucose or HbA1c, a lipid panel and a waist measurement. Those describe your health. BMI only describes your shape.
Using the number sensibly
Treat BMI as one input among several — alongside waist measurement, blood pressure, resting heart rate, blood glucose and lipids, activity level and how you actually feel. A single reading is far less informative than a trend over months.
If your BMI falls outside the healthy range, the useful next step is a conversation with a clinician who can look at the rest of the picture, not a crash diet. Weight that comes off in weeks tends to come back, and rapid loss disproportionately costs muscle.