To Örebro University

Background and aims: In previous studies on environmental contaminants and diabetes only a limited number of contaminants have been evaluated. The aim with the present study is therefore to obtain a comprehensive picture of the relationships between a large number of environmental contaminants and prevalent diabetes.

Materials and methods: In 10 examination cycles in The National Health and Nutrition Examination Survey (NHANES) (1999-2017) altogether 116 different environmental contaminants were evaluated in the circulation or urine in relation to prevalent diabetes. Similar analyses were also performed in the Prospective Investigation of Vasculature in Uppsala Seniors (PIVUS) study (n=1016, all aged 70 years, 50% women, 42 contaminants). The logistic regression models were adjusted for age, sex, race, education, BMI, alcohol intake, smoking and lipids.

Results: In a meta-analysis of the 10 NHANES examinations, thirty-six contaminants were significantly related to prevalent diabetes. Those contaminants represent a number of different classes, such as metals, PCBs, dioxins, furans, pesticides, PFAS, phthalates, and phenols. Some of these relationships were inverse, such as lead, cadmium, mercury, barium, cesium and strontium, some PFAS and furans, as well as benzo-phenone-3. In the smaller PIVUS study (119 prevalent diabetes cases), significant relationships similar to NHANES were seen for some PCBs, p,p´-DDE and lead. PFHxS and PFOA, showed inverse relationships also in PIVUS. The same was seen for mercury and cadmium.

Conclusion: Many environmental contaminants were related to diabetes in the NHANES study. Some of these relationships, mainly some metals and PFAS, were negative. Many of these results were similar in the smaller PIVUS study. These findings need be evaluated in prospective studies.