Causal relationships between lipid and glycemic levels in an Indian population: A bidirectional Mendelian randomization approach

Background Dyslipidemia and abnormal glycemic traits are leading causes of morbidity and mortality. Although the association between the two traits is well established, there still exists a gap in the evidence for the direction of causality. Objective This study aimed to examine the direction of the causal relationship between lipids and glycemic traits in an Indian population using bidirectional Mendelian randomization (BMR). Methods The BMR analysis was conducted on 4900 individuals (2450 sib-pairs) from the Indian Migration Study. Instrument variables were generated for each lipid and glycemic trait (fasting insulin, fasting glucose, HOMA-IR, HOMA-beta, LDL-cholesterol, HDL-cholesterol, total cholesterol and triglycerides) to examine the causal relationship by applying two-stage least squares (2SLS) regression in both directions. Results Lipid and glycemic traits were found to be associated observationally, however, results from 2SLS showed that only triglycerides, defined by weighted genetic risk score (wGRS) of 3 SNPs (rs662799 at APOAV, rs780094 at GCKR and rs4420638 at APOE/C1/C4), were observed to be causally effecting 1.15% variation in HOMA-IR (SE = 0.22, P = 0.010), 1.53% in HOMA-beta (SE = 0.21, P = 0.001) and 1.18% in fasting insulin (SE = 0.23, P = 0.009). No evidence for a causal effect was observed in the reverse direction or between any other lipid and glycemic traits. Conclusion The study findings suggest that triglycerides may causally impact various glycemic traits. However, the findings need to be replicated in larger studies.