Associations of dioxins, furans and dioxin-like PCBs with diabetes and pre-diabetes: is the toxic equivalency approach useful?

Toxic equivalency factors for dioxins and dioxin-like compounds have been established by the World Health Organization. Toxic equivalency (TEQ) was derived using 6 chlorinated dibenzo-p-dioxins, 9 chlorinated dibenzofurans and 8 polychlorinated biphenyls, in blood, from the 1999-2004 National Health and Nutrition Examination Survey. Relationships of 8 individual chemicals, the number of compounds elevated, and TEQ with pre-diabetes and total diabetes (diagnosed and undiagnosed) were investigated using logistic regressions. For the 8 chemicals analyzed separately, values above the 75th percentile were considered elevated, whereas for the other 15 compounds, values above the maximum limit of detection were considered elevated. Pre-diabetes with glycohemoglobin (A1c) 5.9-6.4% was associated with PCB 126, PCB 118 and having one or more compounds elevated (odds ratio 2.47, 95% CI 1.51-4.06). Pre-diabetes with A1c 5.7-5.8% was not associated with any individual chemical or the number of compounds elevated. Total diabetes was associated with 6 of the 8 individual compounds tested, and was associated with having 4 or more compounds elevated. Toxic equivalency ≥81.58 TEQ fg/g was associated with total diabetes (odds ratio 3.08, 95% CI 1.20-7.90), but was not associated with A1c 5.9-6.4%. Having multiple compounds elevated appeared to be important for total diabetes, whereas for pre-diabetes with A1c 5.9-6.4%, having a single compound elevated appeared most important. Diabetes plus A1c ≥5.9% was associated with 34.16-81.57 TEQ fg/g (odds ratio 2.00, 95% CI 1.06-3.77) and with ≥81.58 TEQ fg/g (odds ratio 2.48, 95% CI 1.21-5.11), indicating that half the population has elevated risk for this combination of conditions.