metaboliq
BMI is accurate at one thing: dividing your weight by your height squared. As a measure of body fat or of health, it is rough. When it says you carry too much fat, it is usually right. When it says your weight is fine, it misses about half the people who do carry excess fat, according to a review of 25 studies. In Indians the miss is larger, because many carry more fat, and more of it around the belly, at the same BMI as Europeans.
Use BMI as a first screen and then look further. Add a waist measurement, judged against half your height, and blood tests for sugar, insulin and fats. A high BMI is a good reason to check more closely, while a normal one is only part of the picture.
What does BMI actually measure?
Body mass index is weight in kilograms divided by height in metres, squared. A woman who weighs 64 kg and stands 1.58 m has a BMI of about 25.6. The number knows nothing else about her: not how much of the 64 kg is fat, where that fat sits, or how her blood sugar is doing.
It became the standard because it is cheap and needs only a scale and a tape. For comparing whole populations it works well, since across thousands of people a higher average BMI tracks a higher average body fat. The trouble starts when a population tool is used to judge one person in a clinic or a gym.
Abel Romero-Corral's team at the Mayo Clinic showed where it breaks down. Using data from 13,601 American adults, they found BMI tracked lean mass more closely than body fat in men. In the middle band of 25–29.9, it could not tell fat from muscle in either sex.
Sources: [2]
How accurate is BMI at spotting excess body fat?
Okorodudu and colleagues pooled 25 studies covering about 32,000 adults whose body fat had been measured properly. At the usual cut-offs, BMI correctly flagged about half of the people with excess body fat. Its specificity was 90%, so a high BMI was rarely a false alarm.
At the stricter obesity line of 30, BMI caught only 42% of adults with excess fat. Romero-Corral's American data said much the same: a BMI of 30 or more missed more than half of the people whose body fat was in the obese range. Accuracy also slipped with age, as muscle shrinks and fat replaces it while the weight barely changes.
So the error runs mostly one way. A high BMI is a fairly reliable sign of too much fat, apart from the rare heavily muscled athlete. A normal BMI is much weaker evidence of a healthy amount of fat, and that is the half of the question most people care about.
Why is BMI less accurate for Indians?
In 2004 a WHO expert consultation looked at the evidence from Asian countries. It found a substantial share of Asians at high risk of type 2 diabetes and heart disease at BMIs below 25, the international line for overweight. Depending on the population, the extra risk began somewhere between 22 and 25.
Indian guidance followed. As Anoop Misra summarised in 2015, the Indian consensus counts a BMI of 23 as overweight and 25 as obese. That shift alone moves an extra 10–15% of the population into the overweight or obese range. The lower lines exist because the standard ones, drawn from white populations, missed risk that was already there.
A study from KEM Hospital in Pune shows the size of the gap. Researchers measured body fat in 141 men with deuterium dilution, a lab method. Three in four of the urban middle-class men had more than 25% body fat. By the standard BMI lines, only about a quarter of them were overweight, and none were obese. Chittaranjan Yajnik, who led the Pune unit, has called this pattern the thin-fat Indian.
Can a normal BMI hide poor metabolic health?
Often, yes. Tomiyama's group at UCLA took 40,420 American adults from a national survey and checked six markers: blood pressure, triglycerides, cholesterol, glucose, insulin resistance and an inflammation marker. More than 30% of those with a normal BMI were cardiometabolically unhealthy on those markers. Nearly half of those in the overweight band were healthy.
Indian bodies tend to add to the first group. Fat stored around the liver and abdominal organs is linked more closely with insulin resistance than fat under the skin of the hips. A man with a BMI of 23 and a firm, rounded belly may have a fasting sugar and insulin that look worse than a heavier colleague's.
This is why a weight check alone so often gives false comfort. The useful question is how the body is handling sugar and fat, and only blood tests answer that.
Sources: [6]
Is the BMI machine at a gym or chemist accurate?
For BMI itself, yes, as long as it was given the right height. The machine is doing the same sum you would do on a phone. The extra numbers on the slip are another story. Body fat percentage, muscle mass and a visceral fat level come from bioelectrical impedance. A weak current runs between the metal plates under your feet or in your hands, and the machine turns how easily it flows into an estimate of body water, and from that, fat.
How good is the estimate? In a study at KEM Hospital in Pune, impedance read about 1.2 kg more fat than deuterium dilution, the lab reference, though the two moved closely together across 141 men. Xu's team in Beijing tested two common home devices against DXA and MRI scans in 200 adults. Body fat came out close on average. The visceral fat level, the number people worry about most, had wide margins for any one person.
The reading also moves with the body's water. A big lunch, a sweaty workout or the days before a period can all shift it. If you use one of these machines, stick to the same one, early in the day before food, and judge it by its direction over two or three months.
What should you measure alongside BMI?
A tape measure, first. Margaret Ashwell, a British nutrition scientist, pooled 31 studies covering more than 300,000 adults from several ethnic groups. Waist-to-height ratio picked out people with diabetes, high blood pressure and heart risk better than BMI did, in men and in women. The rule of thumb that came out of that work is to keep your waist under half your height. At 158 cm that means under 79 cm, measured at the navel after a normal breath out.
In 2025 a Lancet commission of 58 experts, led by the surgeon Francesco Rubino, went further. It said BMI should be used to screen people and never on its own to diagnose them. Excess fat, it said, must be confirmed with a body fat measurement or at least one waist-based measure. It also split obesity into two states: excess fat already harming organs, and excess fat that is not yet doing so. Telling them apart takes blood results and symptoms.
In practice the useful set is short. Weight and BMI, the waist against half your height, and a fasting blood panel covering sugar, HbA1c, insulin, lipids and thyroid. Take a 34-year-old man of 170 cm with a BMI of 24 and a 92 cm waist. His waist is well past half his height, and his blood work will show how far that matters.
Why does BMI become even less useful once you start losing weight?
Because a kilo of muscle lowers BMI exactly as much as a kilo of fat, and diets cost some of both. Textbooks long repeated a rule that about a quarter of the weight lost on a diet is lean tissue. Steven Heymsfield's team at the Pennington Biomedical Research Center went back through the evidence for it in 2014. The share turned out to vary with age and activity, and exercise during the diet shifted it most.
Crash diets are where this bites. Someone living on two small meals a day, with no strength work, can watch their BMI fall month after month while their legs get weaker on the stairs. In a Maastricht trial, dieters who lost a bigger share of lean mass regained more weight over the following nine months. Daily walks, two strength sessions a week and dal, curd, paneer or eggs at each meal tilt the loss back towards fat.
Metaboliq's doctor-led plans are measured the same way: losing up to 20% of body weight while keeping muscle, with the tape and strength tracked beside the scale. They start with a home blood test and a doctor's review, since BMI cannot show whether thyroid, sugar or insulin is shaping the weight.
When is BMI good enough on its own?
At the extremes it rarely misleads. The Rubino commission said that above a BMI of 40, excess fat can simply be assumed without further checks. At the other end, a BMI under 18.5 is a clear prompt to look for undernutrition, thyroid trouble or another illness, particularly if the weight fell without trying.
It also earns its place as a way of tracking your own change, provided the tape measure comes along. A BMI that eases from 29 down near 26, while the waist shrinks and you can still climb three floors without stopping, most likely reflects lost fat. When BMI drops and the waist barely moves, it is worth asking what else is going on. Muscle loss, dehydration from a crash diet or illness can all lower the number without improving health.
Sources: [9]