Aims/hypothesis

Multiple genetic variants are associated with type 2 diabetes-related traits in Europeans, but their role in South Asian populations needs further study. We hypothesised that genetic variants associated with diabetes-related traits in Europeans would explain a similar proportion of phenotypic variance in a Pakistani population and could be used in Mendelian randomisation analyses.

Methods

We used data from 2,131 individuals from the Control of Blood Pressure and Risk Attenuation Trial (COBRA) in Karachi, Pakistan. Individuals were aged 40 years or older.

Results

Combining information from multiple genetic variants showed that fasting glucose, BMI, triacylglycerol, and systolic and diastolic blood pressure variants explained 2.9%, 0.7%, 5.5%, 1.2% and 1.8% of the variance in those traits respectively. Genetic risk scores of fasting glucose, triacylglycerol, BMI, systolic blood pressure and diastolic blood pressure variants were associated with these traits, with per allele SD effects of 0.057 (95% CI 0.041, 0.074), p=3.44 × 10(-12), 0.130 (95% CI 0.105, 0.155), p=2.9 × 10(-21), 0.04 (95% CI 0.014, 0.072), p=0.004, 0.031 (95% CI 0.016, 0.047), p=7.9 × 10(-5), 0.028 (95% CI 0.015, 0.042), p = 5.5 × 10(-5), respectively. These effects are consistent with those observed in Europeans, except that the effect of triacylglycerol variants in South Asians was slightly lower. Mendelian randomisation provided evidence that genetically influenced, raised triacylglycerol levels do not causally affect type 2 diabetes risk to the extent predicted from observational data (p=0.0003 for difference between observed and instrumental variables correlations).

Conclusions/interpretation

Genetic variants identified in Europeans are associated with type 2 diabetes-related traits in Pakistanis, with comparable effect sizes. Larger studies are needed to perform adequately powered Mendelian randomisation and help dissect the relationships between type 2 diabetes-related traits in diverse South Asian subgroups.