Abstract:Background Exposure to multiple metals may cause adverse effects, particularly in the kidneys. However, studies on the combined and interaction effects of metal mixtures on human health remain limited. Objective The study aims to evaluate the interaction between metals and assess the combined effects of exposure to metal mixtures on tumor necrosis factor-alpha (TNF-α) levels and kidney function Methods Particular emphasis has been placed on the impact of various metals, including arsenic (As), cadmium (Cd), lead (Pb), as well as essential trace elements, such as cobalt (Co), copper (Cu), selenium (Se), and zinc (Zn), on human health and their potential collective influence on both TNF-α and kidney function. This cross-sectional study analyzed the data of 421 adults who underwent a health examination. Generalized linear model (GLM), Bayesian kernel machine regression (BKMR), and quantile-based G-computation (qgcomp) were used to evaluate the association and joint effects between the metals and TNF-α, as well as kidney function. Results Increased concentrations of As (β = 0.11, 95% CI = 0.05, 0.17) and Pb (β = 0.30, 95% CI = 0.23, 0.37) in the blood were associated with elevated levels of TNF-α, while elevated Cu (β = -0.42, 95% CI = -0.77, -0.07) levels were linked to a significant reduction in TNF-α. The overall effect of metals mixture showed a significant association with a decline in eGFR and an increase TNF-α in the BKMR model. Qgcomp analysis of the metals mixture (β = -0.06, 95% CI = -0.07, -0.05) indicated that As, Pb, and Zn were the primary contributors to the reduction in eGFR, while As and Pb were the major contributors in metals mixture (β = 0.12, 95% CI = 0.08, 0.15) to the elevation of TNF-α levels. Conclusion Exposure to multiple metals could have joint association with the TNF-α levels and kidney function. Furthermore, TNF-α could act as a mediator between metal mixtures and eGFR.

Unveiling the Interaction and Combined Effects of Multiple Metals/Metalloids Exposure to TNF-α and Kidney Function in Adults using Bayesian Kernel Machine Regression and Quantile-based G-computation