Associations between exposure to brominated flame retardants and metabolic syndrome and its components in U.S. adults

Abstract

Background: Humans are simultaneously exposed to numerous of environmental brominated flame retardants (BFRs). We aim to explore the overall associations of BFRs mixture on metabolic syndrome (MetS) and its components and further identify significant chemicals.

Methods: This study included 4641 adults from the National Health and Nutrition Examination Survey (NHANES) in 2007-2016. The weighted logistic regression was conducted to estimate the association of a single BFR exposure with MetS and its components. Meanwhile, the weighted quantile sum (WQS) regression and Bayesian kernel machine regression (BKMR) were adopted to evaluate the overall associations of BFRs mixture on MetS and its components, and to identify significant chemicals. We also evaluated potential associations modified by sex.

Results: In the weighted logistic regression model, PBB153 were positively associated with MetS in a dose-dependent manner (P_trend < 0.05). For its components, increasing quartiles of most BFRs were positively associated with abdominal obesity, hypertriglyceridemia, and low HDL. However, we found no statistically significant associations between BFRs and hypertension and hyperglycemia. WQS analyses found that BFRs mixture was positively associated with MetS (OR: 1.30; 95%CI:1.14, 1.46), abdominal obesity (OR: 1.15; 95%CI:1.03, 1.27), hypertriglyceridemia (OR:1.43; 95%CI:1.19, 1.67), and low HDL (OR: 1.15; 95%CI:1.01, 1.29). BKMR showed associations in a similar direction as WQS for BFRs mixture. For MetS, hypertriglyceridemia and Low HDL, PBB153, PBDE28 and PBDE209 were the most heavily weighting chemicals and had the highest the posterior inclusion probabilities in the WQS and BKMR, respectively. BFRs showed stronger associations of MetS and its components in males than in females.

Conclusions: The present study suggested exposure to BFRs mixture was positively associated with MetS and its components in adults, and PBB153, PBDE28 and PBDE209 were the significant chemicals. However, prospective cohort studies are still needed to confirm the causal effect between BFRs mixture and MetS.

Keywords: Bayesian kernel machine regression; Brominated flame retardants; Chemical mixture exposure; Metabolic syndrome; Weighted quantile sum regression.