Combined effects of bisphenol A and diabetes genetic risk score on incident type 2 diabetes: A nested case-control study

Observational studies reported inconsistent results on the association between bisphenol A (BPA) and type 2 diabetes (T2D) risk. Whether genetic factors modified the association remains unclear. The present nested case-control study prospectively investigated the association of BPA with T2D risk, and the interaction and combined effects of diabetes genetic risk score (GRS) and serum BPA on T2D risk. Based on the Dongfeng-Tongji cohort study, 995 incident diabetes cases and 1:1 age- and gender-matched controls were included. T2D was diagnosed based on the American Diabetes Association criteria. Serum BPA concentration was measured at baseline. Diabetes GRS was constructed by 88 diabetes-related SNPs selected from large-scale GWASs. A U-shaped association was observed between serum BPA levels and T2D risk, with the lowest odds of T2D at the serum BPA levels of 1.00 ng/mL (P = 0.001 for nonlinearity). Compared with the middle group, the multivariate-adjusted ORs of T2D in the lowest group and the highest group of serum BPA were 1.52 (95% CI: 1.04, 2.22) and 1.40 (95% CI: 1.08, 1.81), respectively. Both serum BPA levels (β = 0.107, P = 0.001) and weighted-GRS (w-GRS) (β = 0.072, P = 0.02) were significantly associated with baseline FPG levels. Participants with both highest w-GRS and serum BPA levels had highest risk of T2D (OR = 2.53, 95%CI: 1.49, 4.31, P = 0.001) and higher baseline FPG levels (β = 0.218, P = 0.01), compared with those with both lowest w-GRS and serum BPA levels. Non modified effects of serum BPA levels and w-GRS on T2D, baseline FPG levels, and 5-y changes of FPG levels were detected (All Pinteraction > 0.05). Our results suggested a U-shaped association between serum BPA levels and T2D risk. Participants with higher serum BPA levels and diabetes genetic risk had higher FPG levels and higher risk of T2D.