Association between metal co-exposure and frailty: exploring the potential explanatory pathway of inflammation

Frailty is a critical geriatric syndrome associated with disability and mortality, yet the associations and mechanisms between environmental metal co-exposure with frailty remain unclear. This cross-sectional study included 4484 adults aged ≥ 60 years from China. Metal concentrations were measured by inductively coupled plasma mass spectrometry (ICP-MS), and frailty was assessed using a frailty index (FI). Single-metal analyses used generalized linear models (GLMs). For multi-metal analyses, we applied machine-learning-based variable selection, followed by Bayesian kernel machine regression (BKMR) and quantile g-computation (qg-comp) model. Potential explanatory pathways involving inflammatory markers, as well as interactions with age, sex, and lifestyle, were also explored. Positive associations were observed for magnesium (Mg), molybdenum (Mo), and vanadium (V) with FI, whereas rubidium (Rb) showed an inverse association (P < 0.001). Significant effect modification by age, sex, and lifestyle was identified. Positive metal‑FI associations were stronger in older females, while the inverse Rb‑FI correlation was more evident in participants with unhealthy lifestyles (P for interaction < 0.05). Multi-metal models showed overall positive co-exposure effects on frailty, with Mo (positive) and Rb (negative) contributing the largest weights. In exploratory mediation analyses, the neutrophil-to-high-density lipoprotein ratio (NHR) accounted for 2.09% of the Rb-FI association and 1.81% of the Mo-FI association. The platelet-to-albumin ratio (PAR) accounted for 1.52% of the Mo-FI association and 5.30% of the Mg-FI association (P < 0.05). These findings suggest that metal co-exposure is positively associated with frailty, with inflammation as a potential explanatory pathway. Metal‑frailty associations vary by age, sex and lifestyle, with distinct patterns for different metals. Although longitudinal studies are needed to establish temporality and causal pathways, these findings identify high-risk subgroups for future investigation.