Causal Association of Overall Obesity and Abdominal Obesity with Type 2 Diabetes: A Mendelian Randomization Analysis.

OBJECTIVE: This study aimed to compare the causal effect of overall obesity and abdominal obesity on type 2 diabetes among Chinese Han individuals.

METHODS: The causal relationship of BMI and waist-to-hip ratio (WHR) with the risk of glucose deterioration and glycemic traits was compared using two different genetic instruments based on 30 BMI loci and 6 WHR loci with Mendelian randomization (MR) in three prospective cohorts (n = 6,476).

RESULTS: Each 1-SD genetically instrumented higher WHR was associated with a 65.7% higher risk of glucose deterioration (95% CI = 1.069-2.569, P = 0.024), whereas no significant association of BMI with glucose deterioration was observed. Furthermore, a causal relationship was found only between BMI and homeostatic model assessment β-cell function (HOMA-B) (β = 0.143, P = 0.001), and there was a nominal association with Stumvoll second-phase insulin secretion traits (β = 0.074, P = 0.022). The significance level did not persist in sensitivity analyses, except in the causal estimate of WHR on the Gutt index in MR-Egger (β = -0.379, P = 0.022) and the causal estimate of BMI on homeostatic model assessment β-cell function in weighted median MR (β = 0.128, P = 0.017).

CONCLUSIONS: The data from this study support the potential causal relationship between abdominal obesity and hyperglycemia, which may be driven by aggravated insulin resistance, in contrast with the potential causal relationship between overall obesity and insulin secretion.