BMI and Type 2 Diabetes Risk: Why the Risk Gradient Differs by Population
Type 2 diabetes risk rises with BMI in all populations, but the gradient begins at a lower BMI in Asian groups due to greater visceral fat and insulin resistance at the same weight.
BMI and Type 2 Diabetes Risk: Why the Risk Gradient Differs by Population
The BMI type 2 diabetes risk gradient population relationship shifts leftward for some groups. Risk of developing the disease rises at a lower BMI in specific populations, particularly those of Asian descent. The evidence is settled: your diabetes risk at a BMI of 23 is not the same for everyone.
The core reason is that at the same BMI, some populations carry more fat stored around the organs and have higher levels of insulin resistance. This metabolic profile makes the standard BMI cut-offs, set by the WHO in 1995, a poor screening tool for those groups. The WHO itself recognised this and, in 2004, published separate, lower public health action points for Asian populations. The WHO Western Pacific Regional Office recommended that a BMI of 23.0 kg/m² signals increased risk, and 27.5 kg/m² signals high risk, compared to the global thresholds of 25 and 30.
If you are of South Asian, East Asian, or Aboriginal and Torres Strait Islander descent and rely solely on your BMI number from a standard chart, you may be told you are healthy when your actual type 2 diabetes risk is already elevated. You need a different measurement approach.
Asian BMI Diabetes Risk Lower Threshold: The Evidence That Shifted the Guidelines
The evidence that forced a change came from prospective cohort studies. In 1994, Chan et al. studied 51,529 male health professionals in the USA and found that a BMI of 25-26.9 was associated with a relative risk of type 2 diabetes of 2.2 compared to those with a BMI under 23. In 1995, Colditz et al. studied 114,281 female nurses and reported that a BMI of 25-28.9 carried a relative risk of 9.9 compared to those with a BMI under 22. These were white-majority populations.
When similar studies were conducted in Asian populations, the same relative risks appeared at lower BMIs, in the 23-27 kg/m² range. The 2004 WHO Expert Consultation, published in The Lancet, reviewed this data and concluded that the standard cut-offs were inadequate. The consultation proposed lower action points for public health screening: 23 kg/m² for overweight and 27.5 kg/m² for obesity. The International Diabetes Federation has since incorporated ethnicity-specific thresholds into its definition of the metabolic syndrome.
If a healthcare professional uses only the global BMI bands, insist on an assessment that uses the lower, population-specific thresholds. Ask directly: 'Are you using the WHO Asian BMI cut-offs for my ethnic group?'
BMI Insulin Resistance Association: Why Fat Distribution Changes the Story
How Fat Location Drives Insulin Resistance
The BMI insulin resistance association is not a simple one-to-one relationship because BMI does not measure where fat is stored. The critical link is between fat stored around the organs and insulin resistance. This internal fat, unlike subcutaneous fat, releases free fatty acids and inflammatory markers directly into the portal vein, impairing insulin signalling in the liver and muscle tissue. That drives beta-cell dysfunction over time, culminating in type 2 diabetes mellitus.
At any given BMI, individuals of Asian descent tend to have a higher body fat percentage and a greater proportion of that fat as central adiposity, stored in the abdomen, compared to white individuals. This is why the risk gradient shifts. A BMI of 23 in a South Asian person can represent the same metabolic risk as a BMI of 27 in a white person. The 2004 WHO cut-offs were an attempt to capture this biological reality at a population screening level.
The Normal-Weight Trap
Relying on BMI alone misses elevated insulin resistance in a person with a normal BMI but a high waist circumference. This is the 'normal-weight obesity' phenomenon, and it is most common in older adults and certain ethnic groups.
| Measurement | Standard Threshold (Global) | Threshold for Risk in Asian Populations |
|---|---|---|
| BMI (kg/m²) - Overweight | 25.0 | 23.0 |
| BMI (kg/m²) - Obese | 30.0 | 27.5 |
| Waist circumference (men) | 102 cm (NHLBI) | 90 cm (IDF for Asian men) |
| Waist circumference (women) | 88 cm (NHLBI) | 80 cm (IDF for Asian women) |
| Waist-to-height ratio | 0.5 (NICE CG189) | 0.5 (NICE CG189, applies to all) |
Waist Circumference Diabetes Risk: Capturing What BMI Misses
The Evidence for Waist Measurement
The waist circumference diabetes risk is independent of BMI. In 2005, Wang et al. studied 27,270 men from the Health Professionals Follow-up Study and found that a waist circumference of 102 cm or more was associated with a relative risk of type 2 diabetes of 5.0, even after adjusting for BMI. Two men with the same BMI but different waist sizes have very different risk profiles.
The standard US thresholds, issued by the NHLBI in 1998, are 102 cm for men and 88 cm for women. These were derived from white populations. The International Diabetes Federation, in its 2006 definition of the metabolic syndrome, adopted lower thresholds for Europid men (94 cm) and women (80 cm), and recommended even lower values for Asian populations: 90 cm for men and 80 cm for women.
How to Measure Yourself
For practical screening, the NICE CG189 guideline, published in 2022, recommends using waist-to-height ratio instead of waist circumference alone, because it adjusts for stature. The threshold is 0.5. Measure your waist at the midpoint between the lowest rib and the iliac crest, at the end of a normal expiration. Keep the tape snug but do not compress the skin.
The Biological Mechanism That Links Internal Fat to Diabetes
Why Internal Fat Is Metabolically Active
The mechanism is well understood. Fat stored around the organs is metabolically active. It secretes pro-inflammatory cytokines, such as interleukin-6 and tumour necrosis factor-alpha, and reduces the production of adiponectin, a hormone that improves insulin sensitivity. This inflammatory environment promotes insulin resistance in peripheral tissues. Over time, the pancreas cannot compensate, leading to beta-cell dysfunction and the onset of type 2 diabetes.
This is why two people with the same total body fat percentage can have vastly different diabetes risk. A person with high subcutaneous fat on the hips and thighs has a lower risk than a person with the same amount of fat stored internally in the abdomen. BMI cannot distinguish between these two patterns. Neither can a standard bathroom scale. Only a waist measurement, or a CT or MRI scan for internal fat area, can reveal the distribution.
What to Do With a Normal BMI
If you have a 'normal' BMI but a waist-to-height ratio above 0.5, you likely have excess internal fat and elevated metabolic risk. Do not accept reassurance from a low BMI number alone.
Common Questions
What is the single most actionable measurement to screen for type 2 diabetes risk?
Your waist-to-height ratio. Keep it below 0.5. Measure your waist at the midpoint between the lowest rib and the iliac crest, at the end of a normal expiration. This is the recommendation of NICE guideline CG189 (2022).
I have a BMI of 23. Am I at risk?
It depends on your ethnicity. If you are of South Asian, East Asian, or Aboriginal and Torres Strait Islander descent, a BMI of 23 already puts you in the 'increased risk' category according to the WHO 2004 cut-offs. If you are white European, the global threshold of 25 applies.
Why do the WHO Asian cut-offs from 2004 still apply?
They remain the most authoritative evidence-based thresholds for public health screening in Asian populations. No more recent global expert consultation has superseded them. The principle that BMI-risk gradients differ by population is settled and durable.
Can a thin person have type 2 diabetes?
Yes. Normal-weight obesity, having a normal BMI but a high body fat percentage and excess internal fat, is a recognised phenotype. It is most common in older adults and certain ethnic groups. A low BMI does not guarantee metabolic health.