BMI: what this indicator really measures (and what it doesn't)
The Body Mass Index, or BMI, is everywhere: health records, fitness apps, epidemiological studies. Its calculation is disarmingly simple: weight in kilograms divided by height in meters squared. A 70 kg adult of 1.75 m has a BMI of 22.9. But behind this simplicity lie nuances many users ignore, which can lead to erroneous conclusions about one's own health.
A statistical, not medical, origin
BMI — or the Quetelet index — was proposed in 1832 by the Belgian mathematician Adolphe Quetelet, not as a clinical tool, but as a statistical indicator to describe weight distribution in a population. Only in the 20th century, and more precisely after Ancel Keys' 1972 work, did BMI become adopted by the World Health Organization as a reference measure of corpulence at scale.
This origin is essential to understanding its scope: BMI is a good public-health indicator (trend monitoring, cross-country comparison, association studies with diseases), but it is an imperfect tool to evaluate the health of a given individual.
WHO categories
For adults over 18, WHO defines:
- Under 18.5: underweight
- 18.5 to 24.9: normal weight
- 25 to 29.9: overweight
- 30 to 34.9: class I obesity (moderate)
- 35 to 39.9: class II obesity (severe)
- 40 and over: class III obesity (morbid)
These thresholds were set from epidemiological studies showing correlation between BMI and cardiovascular risk, type-2 diabetes, certain cancers and all-cause mortality. A BMI between 22 and 25 is statistically associated with the lowest mortality in European populations.
Main limitations
1. BMI does not distinguish muscle and fat
This is the best-known criticism. A 1.85 m rugby player weighing 100 kg will show a BMI of 29.2, classified "overweight", whereas his body-fat percentage may be below 15%. Conversely, a sedentary person with little muscle and lots of visceral fat can show a BMI of 23 while having an unfavorable metabolic profile (the "skinny fat" phenotype).
2. It does not account for fat distribution
Two people with the same BMI can have very different risks depending on whether their fat is stored abdominally (visceral fat, more dangerous) or on hips and thighs (subcutaneous fat, less risky). Waist circumference and waist-to-hip ratio usefully complement BMI on this point.
3. It is less relevant at age extremes
In children and adolescents, adult thresholds do not apply: age-and-sex-adjusted percentile curves are used. In the elderly (65+), several studies suggest optimal BMI is slightly higher (25-27), as a weight reserve protects against sarcopenia and the consequences of acute illness.
4. It varies by ethnicity
At equal BMI, populations of Asian origin show higher cardiovascular risk. WHO recommends lowered thresholds for these populations: overweight from 23 and obesity from 27.5. Conversely, in populations of African descent, BMI tends to overestimate fat mass.
Alternatives and complements
No single indicator suffices. A serious evaluation combines several measures:
- Waist circumference: risk threshold at 94 cm in men and 80 cm in women; high risk beyond 102 cm and 88 cm respectively.
- Waist-to-hip ratio: from 0.90 (men) or 0.85 (women), metabolic risk increases.
- Body-fat percentage: measured by impedance, skinfold or DEXA. Normal values: 10-20% for men, 18-28% for women.
- Blood panel: fasting glucose, lipid profile, blood pressure. Some people "metabolically obese despite normal weight" present risk even at normal BMI.
Using BMI intelligently
BMI remains a good starting point, fast and equipment-free. Calculated regularly, it provides a useful personal trend — a gradual upward drift signals a body-composition change to investigate. But it should never be interpreted in isolation, especially in borderline categories (between 18.5 and 25, between 25 and 30).
Our BMI calculator displays your index and its category. For a health decision, discuss with your doctor who will combine this indicator with your history, lifestyle and complementary measures.