Combining plasma measurements and mechanistic modeling to explore the effect of POPs on type 2 diabetes mellitus in Norwegian women

Abstract

The number of studies on persistent organic pollutants (POPs) and type 2 diabetes mellitus (T2DM) is growing steadily. Although concentrations of many POPs in humans have decreased substantially, only some studies consider temporal and inter-individual changes in POP concentrations when assessing exposure. Here we combined plasma measurements with mechanistic modeling to generate complementary exposure measures to our single blood draw after disease diagnosis.Blood was collected between 2003–2006 from 106 subjects with T2DM and 106 age-matched controls, and POP concentrations were compared after adjustment for relevant risk factors and multiple testing. Area under the curve (AUC) of PCB-153 from birth until age 18, representing early-life exposure, and AUC from birth until time of diagnosis were generated as well as examples of life-time exposure trajectories using a mechanistic exposure model. The rank sum of polychlorinated biphenyls (PCBs) and organochlorine pesticides (OCPs, OR=16.9 (95% CI: 3.05–93.6)) as well as β-hexachlorocyclohexane (β-HCH, OR=203.8 (95% CI: 11.5–3620)) and 1, 1-dichloro-2,2-bis(p-chlorophenyl) ethylene (p,p′-DDE, OR=11.3 (95% CI: 2.55–49.9)) were associated with T2DM. Neither of the AUCs reflecting early life exposure and total life-time exposure at the time of diagnosis were associated with the disease. The predicted life course trajectories display clear differences within and between individuals in the past and suggest that a single blood draw provide limited information on POP exposure earlier in life. The predicted AUCs for PCB-153 did not support the positive association between T2DM and measured blood concentration of certain POPs. This may suggest that the model is either too simplistic and/or that strength of the association may vary through life and with time to/past diagnosis.