Aims: To study the within ethnic subgroup variations in diabetes and central obesity among South Asians.
Methods: Data from 9442 individuals age > or = 15 years from the National Health Survey of Pakistan (NHSP) (1990-1994) were analysed. Diabetes was defined as non-fasting blood glucose > or = 7.8 mmol/l, or known history of diabetes. Central obesity was measured at the waist circumference. Distinct ethnic subgroups Muhajir, Punjabi, Sindhi, Pashtun, and Baluchi were defined by mother tongue.
Results: The age-standardized prevalence of diabetes varied among ethnic subgroups (P = 0.002), being highest among the Muhajirs (men 5.7%, women 7.9%), then Punjabis (men 4.6%, women 7.2%), Sindhis (men 5.1%, women 4.8%), Pashtuns (men 3.0%, women 3.8%), and lowest among the Baluchis (men 2.9%, women 2.6%). While diabetes was more prevalent in urban vs. rural dwellers [odds ratio (OR) 1.50, 95% confidence interval (CI) 1.24, 1.82], this difference was no longer significant after adjusting for central obesity (OR 1.15, 95% CI 0.95, 1.42). However, the ethnic differences persisted after adjusting for major sociodemographic risk factors (unadjusted OR for Pashtun vs. Punjabi 0.59, 95% CI 0.42, 0.84, adjusted OR 0.54, 95% CI 0.37, 0.78). Ethnic variation was also observed in central obesity, which varied with gender, and did not necessarily track with ethnic differences in diabetes.
Conclusions: Unmeasured environmental or genetic factors account for ethnic variations in diabetes and central obesity, and deserve further study.