Although many studies have confirmed metabolic syndrome (MetS) is correlated with metal exposures, few studies have elucidated the associations of multiple metals with MetS risk. We aim to explore the relationship between serum 22 metals and MetS. We determined serum 22 metals using ICP-MS and used LASSO regression to select metals independently related with MetS to construct multiple-metals model. We further explored the dose-response relationship between positive metals and MetS by the restricted cubic spline regression. After screening by LASSO regression, serum 11 metals were selected to construct multiple-metals model in cross-sectional analysis, while 5 metals in longitudinal analysis. In the 11-metal model, only tin and zinc were associated with MetS in cross-sectional analysis (ORtin = 2.22, 95% CI:1.43, 3.45; ORzinc = 2.17, 95% CI: 1.42, 3.32; both Ptrend < 0.05). Besides, the same results were found in the 5-metal model in longitudinal analysis (HRtin = 1.66, 95% CI: 0.87, 3.17; HRzinc = 1.83, 95% CI: 1.07, 3.14; both Ptrend < 0.05). Moreover, there were positive linear relationships between serum tin and zinc concentrations and the increasing risk of MetS (both Poverall < 0.05, Pnon-linearity > 0.05). Furthermore, the interaction between high tin and high zinc was also associated with increasing MetS risk (Pinteraction < 0.05). We found that serum tin and zinc were independently and interactively associated with MetS in the southern Chinese men. Our results suggested that high tin and zinc may be the risk factors of MetS.
Keywords: Co-exposure; Cross-sectional analysis; Longitudinal cohort analysis; Metabolic syndrome.