Asian BMI cut-offs, explained without the muddle

The evidence that forced the question

Through the 1990s, body-composition studies kept finding the same pattern: at any given BMI, many East, South and Southeast Asian populations carry several percentage points more body fat than European reference groups, with more of it stored viscerally, around the organs, where it does metabolic damage. Type 2 diabetes and cardiovascular disease showed up at BMI values the standard table called healthy. A cut-off system calibrated on European cohorts was systematically under-warning a large fraction of the world.

Deurenberg and colleagues put numbers on the gap in 2002. Pooling body-composition data across populations, they found that for the same BMI, body fat percentage in Asian groups ran three to five percentage points higher than in Caucasians, and that for the same body fat percentage, BMI ran three to four units lower. Differences in body build explain part of it (trunk-to-leg-length ratio and slenderness), and differences in muscularity explain more. Crucially, they also found Asian populations differ substantially from each other, which is why no single replacement number was ever going to satisfy everyone.

It is worth saying plainly what this is and is not. It is a statement about the arithmetic relationship between a height-weight ratio and body composition in different populations. It is not a claim that anyone's body is wrong, and it is not a licence to treat ancestry as a diagnosis. The whole point of recalibrating a screening threshold is to make the screen equally informative for everyone it is applied to.

WPRO 2000: the Asia-Pacific classification

In 2000, the WHO's Western Pacific regional office, with the International Association for the Study of Obesity and the International Obesity Task Force, published The Asia-Pacific Perspective: Redefining Obesity and its Treatment. It proposed a full alternative category table: normal 18.5–22.9, overweight (at risk) 23–24.9, obese class I 25–29.9, obese class II from 30. This is the classification many Asian health systems adopted, and it is what our Asian BMI calculator shows.

Note what WPRO did structurally: it kept the shape of the WHO table and moved every line down, so 'obese class I' begins at 25 rather than 30. That is why a WPRO reading of an ordinary BMI can sound so much more alarming than the international one for the same number. It is the same distance along the same curve, relabelled to where the risk in these populations actually sits.

WHO 2004: action points, not replacement

Four years later a WHO Expert Consultation reviewed the accumulated evidence for a global decision, and made a deliberately conservative one. It retained the standard international cut-offs for worldwide comparability, and added public-health action points at 23.0, 27.5, 32.5 and 37.5 as trigger levels where screening and intervention should kick in for Asian populations. The Consultation explicitly noted risk varies across Asian populations, which is why it published trigger points rather than one new table.

So there are three honest ways to read a BMI of 24 for someone of Asian ancestry: 'normal' (WHO standard), 'overweight, at risk' (WPRO 2000), and 'past the first action point' (WHO 2004). All three are correct within their systems. A site that merges them into one table has stopped being able to cite any of them, which is precisely the mistake we built this site to avoid.

The same BMI read three ways

Set the three systems side by side and the disagreement becomes legible rather than confusing. The left-hand column is one number; the three columns to its right are what three different bodies would say about it. Nothing about the person changes across a row, only the ruler.

The 23 row is where most of the practical difference lives, because that is the first point at which the three systems diverge. The 27.5 row is the second, where the WHO 2004 scheme escalates from 'increased risk' to 'high risk' while the international table is still saying 'pre-obese'.

Which cut-offs national bodies actually use

Because the WHO published trigger points rather than a replacement table, individual countries made their own decisions, and they did not all land in the same place. The table below sets out what the relevant national bodies adopted. This is the layer most calculator sites skip, and it is the layer that determines what a clinician in a given country will actually say to you.

Japan is the outlier worth understanding. The Japan Society for the Study of Obesity defines obesity at a BMI of 25 with no intermediate overweight band at all, and reserves 35 and above for 'high-degree obesity'. It then adds a second concept on top: obesity counts as 'obesity disease' (the thing warranting treatment) only when it comes with one of eleven specified obesity-related health disorders, or with a visceral fat area of at least 100 cm² measured on imaging. That two-stage logic anticipated the 2025 Lancet Commission framework by more than a decade.

India revised its own position recently. The 2009 consensus statement led by Misra set overweight at 23 and obesity at 25 with abdominal obesity above 90 cm in men and 80 cm in women. A 2025 revision replaced the single BMI threshold with two stages: stage 1 is BMI above 23 with no effect on organ function or daily activities; stage 2 requires BMI above 23 plus a raised waist circumference or waist-to-height ratio, together with an obesity-related condition or a limitation of daily activity.

The waist thresholds move too

Arguing about which BMI table to use is the less useful half of this subject, because the measure that carries the population-specific signal best is the waist. The International Diabetes Federation's worldwide definition of the metabolic syndrome sets central obesity at 90 cm for South Asian and Chinese men against 94 cm for Europid men, with the women's threshold at 80 cm in all three groups. The Working Group on Obesity in China settled on 85 cm for men and 80 cm for women in its own data, and India's consensus statements use 90 cm and 80 cm.

Japan is again the exception, and a genuinely odd-looking one: the values in use there are 85 cm for men and 90 cm for women, the only case in the IDF scheme where the threshold for women is higher than for men. They were derived from visceral fat area measured on CT rather than from outcome data in the same way as the others, which is the usual explanation offered, and they remain the most debated numbers in the table.

Where no population-specific data existed, the IDF said so rather than inventing a number. South and Central American populations are told to use the South Asian values; Sub-Saharan African populations and Eastern Mediterranean and Middle Eastern (Arab) populations are told to use the European ones, explicitly as an interim measure pending better data. That is the right way to handle a gap, and it is worth noticing that a major international body was willing to publish 'we do not know yet'.

Which system applies to you

The honest answer is that it depends on who is asking and why, and that no single chart is 'yours'. If you are being screened within a national health system, that system's cut-offs are the ones that will be applied: a clinic in Beijing will use 24 and 28, one in Singapore 23 and 27.5, one in London 25 and 30. If you are reading a research paper, check which classification it used before comparing its prevalence figures with anything else.

For personal use, the practical rule is simple: if you have East, South or Southeast Asian ancestry, treat 23 rather than 25 as the number that deserves your attention, and do not wait until 25 to start paying attention to the other measures. That is the common ground between WPRO 2000, WHO 2004 and every national body in the table above, which is about as close to consensus as this subject gets.

Mixed ancestry has no published answer, because the studies were not designed to produce one. The sensible response is to stop trying to pick a BMI table and measure your waist instead, since waist-to-height ratio at 0.5 behaves consistently across populations and sidesteps the whole argument. A ratio is a ratio; it does not need a chart.

What to do with this

If you have Asian ancestry, treat 23 rather than 25 as the number worth noticing. That means: measure your waist (the ratio to your height adds the fat-distribution signal), and fold BMI 23+ into routine health checks, blood glucose and lipids especially, since that is where the elevated risk shows first. None of this is a judgement about bodies; it is calibration, and better calibration is simply better information.

Two things follow for anyone using this site. Our Asian BMI calculator shows the WPRO 2000 classification and the WHO 2004 action points separately, labelled, rather than blending them, so you can see which system is saying what. And if you only take one habit away, take the waist measurement: it costs thirty seconds, applies under every system in this guide, and is the measure the 2025 clinical criteria now require alongside BMI regardless of ancestry.

Frequently asked questions

Put it into practice

Run your own numbers through the asian BMI calculator, the BMI calculator and the waist-to-height calculator. Related reading: BMI explained, 2025 obesity definition and Measure your waist.

Sources

1. WHO/IASO/IOTF. The Asia-Pacific Perspective: Redefining Obesity and its Treatment. WPRO, 2000. iris.who.int/handle/10665/206936
2. WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 2004;363:157–63. doi.org/10.1016/S0140-6736(03)15268-3
3. Deurenberg P, et al. Asians are different from Caucasians and from each other in their body mass index/body fat per cent relationship. Obes Rev 2002;3:141–6. doi.org/10.1046/j.1467-789X.2002.00065.x
4. Misra A, et al. Consensus statement for diagnosis of obesity, abdominal obesity and the metabolic syndrome for Asian Indians. J Assoc Physicians India 2009;57:163–70. pubmed.ncbi.nlm.nih.gov/19582986/
5. Misra A, et al. Revised definition of obesity in Asian Indians living in India. Diabetes Metab Syndr 2025;19:102989. doi.org/10.1016/j.dsx.2024.102989
6. Zhou BF; Cooperative Meta-Analysis Group of the Working Group on Obesity in China. Predictive values of body mass index and waist circumference for risk factors of certain related diseases in Chinese adults. Biomed Environ Sci 2002;15:83–96. pubmed.ncbi.nlm.nih.gov/12046553/
7. Ogawa W, et al.; JASSO. Definition, criteria, and core concepts of guidelines for the management of obesity disease in Japan. Endocr J 2024;71:223–31. doi.org/10.1507/endocrj.EJ23-0593
8. Lee YS, et al. Health Promotion Board–Ministry of Health Clinical Practice Guidelines: Obesity. Singapore Med J 2016;57:292–300. doi.org/10.11622/smedj.2016103
9. Alberti KG, Zimmet P, Shaw J. Metabolic syndrome: a new world-wide definition. A consensus statement from the International Diabetes Federation. Diabet Med 2006;23:469–80. doi.org/10.1111/j.1464-5491.2006.01858.x
10. NICE. BMI: preventing ill health and premature death in black, Asian and other minority ethnic groups. Public health guideline PH46, 2013. www.nice.org.uk/guidance/ph46