BMI explained — what it measures, and the better numbers to track

Body Mass Index is probably the most-cited health number in the world, and also one of the most misunderstood. People treat it as a verdict on their health when it was never designed to be one. The truth is more nuanced and more useful: BMI is a fast, cheap screening tool that does one job reasonably well and several other jobs badly. Knowing which is which means you can use the number without being ruled by it — and know which other numbers are worth tracking instead.

What BMI actually is

BMI is just a ratio of your weight to your height, squared. The formula is:

BMI = weight (kg) / height (m)²

So if you weigh 70 kg and stand 1.75 m tall:

1.75 × 1.75 = 3.0625
70 / 3.0625 = 22.9

A BMI of 22.9. In imperial units, the formula carries a conversion factor:

BMI = 703 × weight (lb) / height (in)²

That's all there is to it. No body-fat measurement, no muscle assessment, no consideration of where weight sits on your frame. It's height and weight, nothing more. The Belgian mathematician Adolphe Quetelet devised the ratio in the 1830s — as a tool for describing populations, explicitly not for diagnosing individuals. That original purpose is the key to understanding everything about its strengths and limits.

The WHO categories

The World Health Organization defines standard adult categories based on the BMI value:

• Below 18.5 — underweight
• 18.5 to 24.9 — normal / healthy weight
• 25.0 to 29.9 — overweight
• 30.0 and above — obese (often subdivided into class I at 30–34.9, class II at 35–39.9, and class III at 40+)

These cutoffs aren't arbitrary. They were chosen because, across large populations, the risk of weight-related conditions — type 2 diabetes, cardiovascular disease, certain cancers — tends to rise as BMI climbs into the overweight and obese ranges. The bands are a convenient way to flag elevated statistical risk. You can compute your own figure with the BMI Calculator and see where it lands.

What BMI gets right

At the scale of thousands or millions of people, BMI is genuinely useful, for a few solid reasons.

It's cheap and universal. You need a scale and a tape measure — no calipers, no scanners, no lab. That means public-health researchers can collect it everywhere and compare across countries and decades.

It correlates with body fat across a population. For the average person who isn't an athlete, a higher BMI usually does mean more body fat, and tracking BMI distributions across a population catches real trends in obesity prevalence.

It's a reasonable first screen. For a clinician seeing a patient for the first time, BMI is a fast flag that says "this might be worth a closer look" — in either direction. As a starting question, not a final answer, it earns its place.

Where BMI gets it wrong for individuals

The problems begin the moment you apply a population statistic to one specific human being. BMI has no idea what your weight is made of or where it sits.

It can't tell muscle from fat. Muscle is denser than fat, so a muscular person weighs more for their height. A lean, heavily trained athlete can post a BMI in the "overweight" or even "obese" range while carrying very little body fat. The number sees only mass. This is the most famous BMI failure, and it's real — though worth noting most people in the overweight BMI range are not elite athletes, so for the general population the muscle objection applies to a minority.

It ignores body shape and fat distribution. Two people with identical BMIs can have very different risk profiles depending on where their fat is. Fat stored around the abdomen (visceral fat, packed around the organs) is far more metabolically harmful than the same amount stored on the hips and thighs. BMI is completely blind to this distinction, yet it may be the single most important factor for individual risk.

It can mislead at the extremes of height. Because the formula divides by height squared, it tends to slightly overstate the BMI of taller people and understate it for shorter people relative to their actual body composition. The squaring is a mathematical convenience, not a biological law.

It shifts with age and sex. Older adults lose muscle and bone density, so a "normal" BMI can mask a higher body-fat percentage. Women naturally carry more body fat than men at the same BMI. One set of cutoffs can't capture all of that.

Ethnicity matters: the Asian thresholds

Here's a point that often gets missed. The standard WHO cutoffs were derived largely from populations of European descent, and they don't transfer cleanly to everyone.

Research has consistently shown that people of South Asian, East Asian, and Southeast Asian descent tend to develop elevated risk of type 2 diabetes and cardiovascular disease at lower BMI values than the standard thresholds suggest — often carrying more visceral fat at a given BMI. In response, the WHO and many national health bodies recommend lower trigger points for these populations. A common adjusted guideline treats:

• 23 and above as overweight / increased risk
• 27.5 and above as the point of substantially higher risk

So someone of Asian descent with a BMI of 24 — comfortably "normal" on the standard chart — may already be in an elevated-risk zone. If this applies to you, the standard category your BMI lands in may understate the picture, and it's worth discussing the adjusted thresholds with a clinician. This isn't a flaw to panic over; it's exactly the kind of context that turns a blunt number into a useful one.

A better personal number: waist-to-height ratio

If BMI's biggest individual weakness is that it ignores where fat sits, the obvious fix is to measure that directly. The simplest, most accessible option is waist-to-height ratio (WHtR), and a growing body of evidence suggests it predicts cardiometabolic risk better than BMI for individuals.

The measurement is easy. Take a tape measure around your waist — roughly at the level of your navel, after a normal breath out, without sucking in. Then divide that by your height, using the same units for both:

WHtR = waist measurement / height

A 81 cm (32 in) waist on someone 180 cm (71 in) tall gives 81 / 180 = 0.45.

The widely cited rule of thumb is memorably simple: keep your waist to less than half your height. A WHtR under about 0.5 is generally considered healthy for most adults; values climbing above that suggest accumulating central fat — the metabolically risky kind BMI can't see.

Why is this often better than BMI for an individual?

• It targets abdominal fat directly, the strongest single indicator of metabolic risk.
• It partially cancels out the muscle problem, because a muscular athlete with a lean midsection scores well even when their BMI flags "overweight."
• It travels across populations more robustly, since the "half your height" guideline holds up reasonably well across different ethnic groups, where fixed BMI cutoffs don't.
• It needs only a tape measure — as cheap and accessible as stepping on a scale.

None of this makes BMI useless. The most informative approach is to look at both: BMI for a quick overall sense of where your weight sits, and waist-to-height ratio to check whether that weight is distributed in a way that raises risk. Where the two disagree — a borderline BMI but a healthy waist, or a "normal" BMI with a thick midsection — that disagreement is itself the useful signal.

Using the number sensibly

BMI is a screening tool, not a diagnosis. Treat it as one input among several. A single value tells you very little; trends over time, alongside your waist measurement, activity level, blood pressure, and how you actually feel, tell you a great deal more. If your BMI lands somewhere that concerns you, the right next step is a conversation with a healthcare professional who can look at the whole picture — not a crash diet driven by a number that was never meant to judge you individually.

Run the figure with the BMI Calculator to know where you stand, then grab a tape measure and check your waist-to-height ratio too. Two cheap numbers, read together with a clear head, will tell you far more than either one shouted in isolation.