Persistent organic pollutants and risk of type 2 diabetes: A prospective investigation among middle-aged women in Nurses' Health Study II

Abstract

BackgroundExposure to persistent organic pollutants (POPs) may predispose to the development of type 2 diabetes (T2D), but prospective human evidence is scarce. ObjectivesWe investigated the association between plasma-POP concentrations in the late 1990s and incident T2D over 11 years of follow-up in the Nurses' Health Study II. DiscussionThree organochlorine pesticides and 20 polychlorinated biphenyls (PCBs) were measured in banked plasma from 793 case–control pairs of T2D. In a multiviarate-adjusted model, T2D ORs (95%CIs) comparing extreme POP tertiles (high vs. low) were 1.67 (1.24, 2.23; Ptrend < 0.001) for hexachlorobenzene (HCB), 3.62 (2.57, 5.11; Ptrend < 0.001) for β-hexachlorocyclohexane (β-HCH), 1.55 (1.13, 2.13; Ptrend = 0.05) for p,p′-dichlorodiphenyldichloroethylene (p,p′-DDE), and 1.95 (1.42, 2.69; Ptrend < 0.001) for total dioxin-like PCBs (DL-PCBs) which included 5 mono-ortho congeners, PCB-105, 118, 156, 157, and 167. Adjustment for previous weight change and body mass index (BMI) at blood draw attenuated these associations, but that for DL-PCBs remained (OR[95% CI] = 1.78[1.14, 2.76]; Ptrend = 0.006). Age, breastfeeding history, previous weight change and BMI at blood draw were significant predictors of plasma POP concentrations. In addition, we found significant interactions of POPs and weight change before blood draw on T2D risk. ORs (95%CIs) of T2D comparing extreme (high vs. low) POP groups were 2.00 (1.02, 3.92; Ptrend = 0.01) for HCB, 2.69 (1.34, 5.40; Ptrend < 0.001) for β-HCH, and 2.41 (1.22, 4.77; Ptrend < 0.001) for DL-PCBs in the lowest weight gain group, whereas these values were 1.29 (0.73, 2.28; Ptrend = 0.46; Pinteraction = 0.04) for HCB, 1.41 (0.77, 2.60; Ptrend = 0.24; Pinteraction = 0.003) for β-HCH, and 0.90 (0.50, 1.63; Ptrend = 0.61; Pinteraction = 0.01) for DL-PCBs in the highest weight-gain group. ConclusionsOur findings suggest that elevated POP exposure may have diabetogenic potential. These data also highlight the impact of lifestyle factors, especially history of weight gain, on circulating POP concentrations and their associations with subsequent T2D risk.