Synergistic effect of non-alcoholic fatty liver disease and history of gestational diabetes to increase risk of type 2 diabetes.

Abstract

Whether non-alcoholic fatty liver disease (NAFLD) improves risk prediction for subsequent type 2 diabetes mellitus (T2DM) in women with prior gestational diabetes mellitus (pGDM) is uncertain. We examined the combined effects of NAFLD and pGDM on risk prediction for incident T2DM. This retrospective cohort study included 97,347 Korean parous women without diabetes mellitus at baseline whose mean (SD) age was 39.0 (7.8) years. Cox proportional hazards models were used to estimate hazard ratios (HRs) for incident T2DM according to self-reported pGDM and ultrasound-diagnosed NAFLD at baseline. When combined with conventional diabetes risk factors, the incremental predictive ability of NAFLD and pGDM to identify incident T2DM was assessed. During a median follow-up of 3.9 years, 1,515 cases of incident T2DM occurred. Multivariable-adjusted HRs (95% confidence intervals [CIs]) for incident T2DM comparing pGDM alone, NAFLD alone, and both NAFLD and pGDM to the reference (neither NAFLD nor pGDM) were 2.61 (2.06-3.31), 2.26 (1.96-2.59), and 6.45 (5.19-8.00), respectively (relative excess risk of interaction = 2.58 [95% CI, 1.21-3.95]). These associations were maintained after adjusting for insulin resistance, body mass index, and other potential confounders as time-dependent covariates. The combination of NAFLD and pGDM improved risk prediction for incident T2DM (based on Harrell's C-index, also known as the concordance index, and net reclassification improvement) compared to conventional diabetes risk factors. In conclusion, NAFLD synergistically increases the risk of subsequent T2DM in women with pGDM. The combination of NAFLD and pGDM improves risk prediction for T2DM.