Associations of metal mixtures with insulin resistance and glucose homeostasis in U.S. adults from the NHANES 2011-2018.

Research on the association of metal mixtures with glucose-insulin homeostasis is limited, and previous studies have typically focused on single metals. This study utilized data from 3110 adult subjects in the NHANES survey (2011-2018). Generalized linear models (GLM), logistic regression (LR), and restricted cubic splines (RCS) were employed to assess the associations of blood and urine metals with insulin resistance (IR) and glucose-insulin homeostasis. The Bayesian kernel machine regression (BKMR) and Bayesian weighted quantile sum (BWQS) models were further used to explore the independent and combined effects of metal exposures. In single-metal analyses, manganese (Mn) was positively correlated with insulin resistance (IR); cadmium(Cd), lead(Pb), mercury(Hg), and arsenic (As) were negatively correlated with the homeostasis model assessment of beta-cell function (HOMA-β); manganese and selenium (Se) were positively correlated with fasting plasma insulin (FPI); Se and cobalt (Co )were positively correlated with fasting plasma glucose (FPG); molybdenum (Mo) was positively correlated with HbA1c. In addition, both BWQS and BKMR models consistently showed that overall metal co-exposure had a positive effect on insulin resistance in the general population. Manganese was the most heavily weighted metal across all subgroups, with this association being more pronounced in males and individuals over 60 years of age. A negative association of metal mixtures with HOMA-β was observed in BWQS models. Furthermore, the analysis of BKMR models revealed possible interactions between insulin resistance and some components of metal mixtures in glucose homeostasis. The RCS model also identified nonlinear relationships between urinary Mo and HOMA-β, as well as between Co and both FPG and HbA1c. Our results suggest that metal mixtures may have adverse individual or combined effects on insulin resistance and glucose homeostasis in different population subgroups. These findings highlight the need for targeted interventions to mitigate the adverse effects of metal exposure on insulin-glucose homeostasis, which may provide new ideas for preventing and controlling the risk of type 2 diabetes due to metal exposure.