Body mass index (BMI) and waist circumference (WC) be measured concurrently to assess obesity-related health risk in populations with a predisposition to abdominal obesity, especially in Asians. However, there is no existing prospective study to support a specified level of WC and the clinical utility of the combined use of BMI and WC to indicate the risk of T2DM risk in Asians with a lower degree of obesity.


We conducted a prospective assessment of 21,122,422 subjects who were free of diabetes at baseline examination, and we used 2009-2015 health checkup data from a large-scale population-based cohort from The National Health Insurance System. The subjects were divided into 8 BMI subgroups based on the following levels; 18.5, 23, 25, 27.5, 30, 32.5 and 35. We also stratified subgroups with 20 sex-specific quantiles of WC. A specified level of WC was defined based on the level at which subjects’ risk of type 2 diabetes doubled compared to the risk for the low-risk group.


During a mean follow-up of 5.3 years, 633,825 (3.0%) patients developed type 2 diabetes. The HRs for type 2 diabetes were increased by 19.1% per 1 kg/m2 increment of BMI. The multivariate-adjusted HRs across BMI categories were 0.747, 1.0, 1.642, 2.369, 3.407, 4.814, 6.653 and 9.136, respectively. The specified levels of WC (HR, 2.079; 95% CI, 2.023 to 2.136) were determined at the level of WC ≥ 88 cm for men and ≥ 80 cm for women. Based on this result, the prevalence of central obesity was 28.11% in men and 31.67% in women. Across BMI and WC categories, the HRs for type 2 diabetes were significantly increased for those with a WC above a specified level for a given BMI category.


Additional measurement of WC on the basis of BMI is a useful approach for refining type 2 diabetes risk. Our findings suggest that action levels based on BMI could be increased by one level if the WC exceeds a specified level in Asian populations.