Abstract
Although the association between pollution exposure and chronic kidney disease (CKD) has been explored, previous studies have focused on specific effects observed via in vitro or animal experiments. We first conducted a priority screening of pollutants for population CKD risk by using machine learning approaches. We then used the National Health and Nutrition Examination Survey (NHANES) 2007-2016 data from 2415 adults aged 40 years and over to study the joint effects of low-concentration metal exposure and the mediating effects of α-klotho by using Bayesian kernel machine regression (BKMR) and mediation analyses. Priority screening revealed that cadmium (Cd), mercury (Hg), lead (Pb), and thallium (Tl) were associated with the highest risk of developing CKD. The BKMR model revealed a negative joint effect of mixed-metal exposure on CKD risk. Tl presented the highest posterior inclusion probability (PIP) of 1.0000, followed by Pb, with a PIP of 0.6080. Significant mediating effects of α-klotho on Hg-CKD associations were observed. Mendelian randomization demonstrated that a high level of α-klotho is associated with a decreased risk of developing CKD. This is the first study to reveal the risk prioritization of various pollutants in CKD patients, as well as the coexposure effects of metals. Our study also provides insight into the potential mechanisms underlying the association between metal exposure and CKD risk.