Polychlorinated biphenyls (PCBs) and risk of dementia and Parkinson disease: A population-based cohort study in a North Italian highly polluted area

Background: Serum polychlorinated biphenyls (PCBs) have been associated with some chronic diseases, but little evidence exists on their possible relationship with neurodegenerative diseases. We aimed to assess the relationship between PCB exposure and the occurrence of dementia and Parkinson disease in a prospective cohort study in a highly polluted area (Brescia-Caffaro). Methods: PCB exposure was assessed by measuring serum levels of 24 congeners. Data on the onset of dementia and Parkinson disease were retrieved by the Brescia Health Protection Agency Database. We used Poisson regression models adjusted for possible confounders to calculate rate ratios (RRs). A mediation analysis was performed to evaluate the mediatory role of cardiovascular diseases. Results: 699 subjects without neurologic diseases at baseline were enrolled (48.1% males, 63.2 years of mean age) in 2001–2013 and followed up to 2018. During a mean follow-up of 8.8 years, 36 and 20 subjects developed dementia and Parkinson disease. Subjects in the 2nd and 3rd tertiles of the total PCBs distribution, compared with those in the 1st tertile, had a higher risk of dementia (RR = 2.30 and RR = 4.35). The estimates for Parkinson disease included the null value with wide confidence intervals. In the mediation analysis, the association between PCB exposure and dementia was dominated by the direct pathway and not by the hypertension-mediated pathway. Conclusions: We observed a positive association between total PCBs serum levels and the onset of dementia not mediated by hypertension. For Parkinson, the unstable risk estimates did not allow to draw a conclusion on a possible association.