Background: Bisphenol A (BPA), an endocrine disrupting chemical (EDC), may pose a risk to human health, including diabetes. However, epidemiological studies from the U.S., China and South Korea showed inconsistent results.
Objective: To examine the associations between BPA and indicators of diabetes in the general Canadian population.
Methods: The analysis was based on cross-sectional data from the Canadian Health Measures Survey (CHMS) Cycle 2 (2009-2011). We included 1915 participants with glycated hemoglobin (HbA1c) measurement and 2405 participants with serum glucose (SG) measurement aged 3-79 years, respectively. Multiple linear regression analyses were performed to model HbA1c and log-transformed SG levels associated with quartiles of urinary BPA concentrations controlling for potential confounders. Further, in adults (age≥18 years), logistic regression analysis was conducted to evaluate the association between BPA and physician-diagnosed diabetes mellitus (DM). Bootstrap weights were applied to all the analyses to account for the complex survey design.
Results: The geometric mean of urinary BPA was 1.21 (±0.05)µg/L. Overall, a positive association was observed between urinary BPA quartiles and HbA1c levels in men after controlling for potential confounders (P<0.05), but not in women and children. Similar patterns were found for the associations of BPA with log-transformed SG levels and doctor-diagnosed DM.
Conclusions: Higher urinary BPA levels were associated with adverse glucose homeostasis in Canadian men, independent of major covariates. Prospective studies with longitudinal design are needed to further investigate the causality.
Keywords: Bisphenol A; Canada; Diabetes; Glycated hemoglobin; Serum glucose; Survey.
Copyright © 2016 Elsevier Inc. All rights reserved.