Effects of chemical mixtures on liver function biomarkers in the Korean adult population: thresholds and molecular mechanisms for non-alcoholic fatty liver disease involved.

Abstract

There is a scarcity of research on the effects of a mixture of chemicals on liver function biomarkers and non-alcoholic fatty liver disease (NAFLD) indices, including FSI, HIS, and FBI-4. Thus, we aimed to explore whether there is an association between chemical mixtures, including 26 chemicals found in blood and urine, liver function biomarkers, and non-alcoholic fatty liver disease (NAFLD) indices in Korean adults. The effects of exposure to chemical mixtures on liver function biomarkers and NAFLD indices were investigated using linear regression models, weighted quantile sum (WQS) regression, quantile g-computation (qgcomp), and Bayesian kernel machine regression (BKMR) among 3669 adults. In silico toxicogenomic data-mining, we evaluated molecular mechanisms associated with NAFLD, including pathways, diseases, genes, miRNAs, and biological processes. The linear regression models showed blood or urine Hg levels were the most important factors associated with AST, ALT, GGT, FSI, and HSI levels, and significant trends were observed for these chemical quartiles (p < 0.01). The WQS index was significantly associated with ALT, GGT, FSI, and HSI. The qgcomp index also found an association between chemicals and AST, ALT, GGT, and FSI. In the BKMR model, the overall effect of the mixture was significantly related to ALT, GGT, FSI, and HSI. In silico analysis, we found mixed chemicals interacted with the CYP1A2 gene and were associated with NAFLD. Seventy-eight percent of interactions were identified as physical interactions in the CYP1A2 gene related to NAFLD. Transcription factor regulation in adipogenesis and lipid metabolic processes are fundamental molecular mechanisms that could be influenced by NAFLD-related mixed chemicals. Cutoff thresholds for chemical exposure levels associated with liver function indicators and NAFLD indices were also reported. The strongest interactions and expression of miRNAs involved in NAFLD development were also identified.