Combined exposure to multiple metals on cardiovascular disease in NHANES under five statistical models
Xianwei Guo,Ning Li,Hao Wang,Wanying Su,Qiuxia Song,Qiwei Liang,Mingming Liang,Chenyu Sun,Yaru Li,Scott Lowe,Rachel Bentley,Evelyn J Song,Qin Zhou,Xiuxiu Ding,Yehuan Sun
DOI: https://doi.org/10.1016/j.envres.2022.114435
Abstract:Background: It is well-documented that heavy metals are associated with cardiovascular disease (CVD). However, there is few studies exploring effect of metal mixture on CVD. Therefore, the primary objective of present study was to investigate the joint effect of heavy metals on CVD and to identify the most influential metals in the mixture. Methods: Original data for study subjects were obtained from the National Health and Nutrition Examination Survey. In this study, adults with complete data on 12 kinds of urinary metals (antimony, arsenic, barium, cadmium, cobalt, cesium, molybdenum, mercury, lead, thallium, tungsten, and uranium), cardiovascular disease, and core covariates were enrolled. We applied five different statistical strategies to examine the CVD risk with metal exposure, including multivariate logistic regression, adaptive elastic net combined with Environmental Risk Score, Quantile g-computation, Weighted Quantile Sum regression, and Bayesian kernel machine regression. Results: Higher levels of cadmium, tungsten, cobalt, and antimony were significantly associated with Increased risk of CVD when covariates were adjusted for multivariate logistic regression. The results from multi-pollutant strategies all indicated that metal mixture was positively associated with the risk of CVD. Based on the results of multiple statistical strategies, it was determined that cadmium, tungsten, cobalt, and antimony exhibited the strongest positive correlations, whereas barium, lead, molybdenum, and thallium were most associated with negative correlations. Conclusion: Overall, our study demonstrates that exposure to heavy metal mixture is linked to a higher risk of CVD. Meanwhile, this association may be driven primarily by cadmium, tungsten, cobalt, and antimony. Further prospective studies are warranted to validate or refute our primary findings as well as to identify other important heavy metals linked with CVD.