BMI Explained: What the Number Means and What It Misses

A plain guide to Body Mass Index: the official WHO categories, the different thresholds used for Asian populations, what large-scale mortality research actually shows, and the situations where BMI gives a misleading answer.

Abin Kark

By Abin Kark

·

BMI is probably the most widely used health number in the world, and one of the most widely misunderstood. It gets treated as a verdict when it was designed as a filter: a quick way to sort large populations into groups worth looking at more closely.

This guide covers what the number is, where the category boundaries come from, what the research actually shows about them, and the specific situations where BMI will give you an answer you should ignore.

Calculate your BMI with a visual comparison →

The Calculation

BMI is your weight in kilograms divided by your height in metres squared.

BMI = weight (kg) ÷ height (m)²

Someone who is 175 cm and weighs 72 kg: 72 ÷ (1.75 × 1.75) = 72 ÷ 3.0625 = 23.5.

In imperial units, multiply weight in pounds by 703, then divide by height in inches squared.

That is the whole formula. Notice what is not in it: age, sex, body composition, fat distribution, ethnicity, fitness, or bone density. BMI knows two facts about you and nothing else. Most of its limitations follow directly from that.

The Official Categories

The World Health Organization set these boundaries in a 1995 expert consultation, with refinements published in a 2000 technical report. They remain the international standard.

BMI RangeCategory
Below 18.5Underweight
18.5 – 24.9Normal weight
25.0 – 29.9Overweight
30.0 – 34.9Obese, Class I
35.0 – 39.9Obese, Class II
40.0 and upObese, Class III

These cutoffs are round numbers chosen for practical use, not sharp biological thresholds. Nothing measurable changes about a person's health between a BMI of 24.9 and 25.1. Treat the boundaries as signposts on a continuum.

Different Thresholds for Asian Populations

This is the part most BMI calculators leave out, and it matters for a very large share of the world's population.

A 2004 WHO expert consultation concluded that people of Asian descent tend to carry a higher proportion of visceral fat, the fat stored around internal organs, at any given BMI. Because visceral fat drives metabolic risk more strongly than subcutaneous fat, health risk appears at a lower BMI in these populations.

The revised threshold: overweight begins at a BMI of 23.0 rather than 25.0.

A person of South Asian or East Asian descent with a BMI of 24 sits comfortably inside the "normal" band on the standard chart while already carrying elevated risk under the population-specific guidance. Several countries, including India, China and Japan, use adjusted national thresholds for this reason.

What the Mortality Research Actually Shows

The largest study on this question is the Global BMI Mortality Collaboration, published in The Lancet in 2016, which pooled data from roughly 10.6 million participants.

Its finding is more specific than the WHO chart suggests. Lowest all-cause mortality clustered at a BMI between 20 and 25, a narrower band than the official 18.5 to 24.9 "normal" range. Risk rose steadily above 25, and it also rose at the lower end, below about 20.

Two things follow. First, the bottom of the "normal" range is not a target to aim for. Second, the relationship is a curve with risk increasing in both directions, not a cliff edge at 25.

See where your weight sits on the distribution →

Five Situations Where BMI Misleads

Muscular people. Muscle is denser than fat. A trained athlete can carry a BMI of 27 or 28 with very low body fat and excellent metabolic health. This is the best-known failure mode and the reason many strength athletes and rugby players register as "obese."

People with low muscle mass. The reverse case gets less attention and is arguably more common. Someone sedentary with little muscle can sit at a BMI of 23, inside the normal range, while carrying substantial visceral fat. This is sometimes described as being metabolically unhealthy at a normal weight.

Older adults. Muscle mass declines with age while fat proportion increases, so an unchanged BMI across decades can conceal a meaningful shift in body composition. Height loss in later life also nudges BMI upward for reasons unrelated to weight.

Very tall and very short people. Because the formula divides by height squared rather than cubed, it slightly overestimates BMI in tall people and underestimates it in short people. The effect is modest but real at the extremes.

Pregnancy and children. Standard adult BMI does not apply during pregnancy. For children and teenagers, BMI is interpreted against age-and-sex-specific percentile charts rather than fixed cutoffs, in the same way height is.

What to Measure Alongside It

BMI is more useful as one input than as a conclusion. The measures that add the most context:

  • Waist circumference, which captures central fat distribution that BMI cannot see. This is often the single most informative addition.
  • Waist-to-height ratio, where a common rule of thumb is keeping your waist under half your height.
  • Blood pressure, blood glucose and lipids, which measure what actually drives cardiovascular risk.
  • Fitness level, which independently predicts health outcomes even when weight does not change.

Two people with an identical BMI can have very different risk profiles. The extra measurements are what separate them.

Tools

Frequently Asked Questions

1. What is a healthy BMI?

The WHO classifies 18.5 to 24.9 as normal weight. Research from the Global BMI Mortality Collaboration found the lowest mortality risk in a somewhat narrower band, roughly 20 to 25.

2. How do I calculate BMI?

Divide your weight in kilograms by your height in metres squared. In imperial units, multiply weight in pounds by 703 and divide by height in inches squared.

3. Is BMI accurate?

It is accurate as a population-level screening tool and unreliable as an individual diagnosis. It cannot distinguish muscle from fat, and it says nothing about where fat is stored.

4. Why do Asian populations have different BMI thresholds?

A 2004 WHO expert consultation found that people of Asian descent tend to carry more visceral fat at a given BMI, so health risk appears earlier. The overweight threshold is commonly set at 23.0 instead of 25.0.

5. Can an athlete have an unhealthy BMI reading?

Yes, and it is common. Muscle is denser than fat, so a lean, heavily muscled person can register as overweight or obese while having very low body fat.

6. Does BMI work for children?

Not with adult cutoffs. Children and teenagers are assessed against age-and-sex-specific percentile charts, because body composition changes substantially through growth.

7. What should I measure alongside BMI?

Waist circumference is the most valuable addition, since it captures central fat distribution. Blood pressure, blood glucose, lipid levels and cardiorespiratory fitness all add context that BMI cannot provide.

8. Does being tall affect my BMI reading?

Slightly. The formula divides by height squared, which tends to overestimate BMI in very tall people and underestimate it in very short people. The effect is small for most of the population.

In Short

BMI answers one narrow question well: is this person's weight unusual for their height? That is genuinely useful information, and it is cheap to obtain, which is why it became a global standard.

What it cannot tell you is why, or whether it matters for you specifically. For that you need a waist measurement, a blood test, and ideally a conversation with a doctor. Use the number as a starting point and treat any confident conclusion drawn from it alone with suspicion.

Calculate and visualise your BMI →


Sources: WHO BMI classification, the 2004 WHO expert consultation on Asian populations, the Global BMI Mortality Collaboration (The Lancet, 2016), and the Cleveland Clinic guidance on BMI limitations.

Loading analytics...