Multi-level evidence for Pb-Cd co-exposure nephrotoxicity: An integrated analysis of NHANES, outcome-oriented network toxicology, and cell-based validation.

BACKGROUND: Lead (Pb) and cadmium (Cd) remain widespread environmental toxicants, and the kidney is a major target organ. However, reliance on a single biomarker matrix or renal endpoint may incompletely characterize Pb-Cd-associated renal abnormalities. Whether blood- and urine-based Pb/Cd biomarkers exhibit differential association profiles across renal outcomes, and whether outcome-oriented networks converge on shared candidate pathways, remains unclear. METHODS: We analyzed NHANES 2011-2018 data from 5049 adults. Blood and urinary Pb and Cd biomarkers were evaluated in relation to estimated glomerular filtration rate (eGFR), albuminuria [log(UACR)], and a CKD-compatible phenotype using survey-weighted regression, whereas Pb-Cd mixtures were examined using case-weighted quantile g-computation and generalized weighted quantile sum regression. Effect modification by age, sex, BMI, diabetes, and hypertension was evaluated using formal interaction tests with false-discovery-rate correction. An outcome-oriented network toxicology framework was constructed using CKD-, albuminuria-, and broad renal injury-oriented disease modules. HEK293T cells were used for targeted cellular assessment of cytotoxicity, intracellular metal accumulation, redox and inflammatory responses, and transcriptional changes in genes prioritized by network analysis following Pb, Cd, or combined exposure. Exploratory molecular docking was performed to examine potential metal-protein interaction sites. RESULTS: Individual-biomarker and mixture analyses showed that the observed associations varied across biomarker matrices and renal outcomes. Urinary Pb was positively associated with eGFR; urinary Cd and blood Pb were associated with higher log(UACR); and blood Pb and blood Cd were associated with higher odds of the CKD-compatible phenotype. The urinary Pb-Cd mixture was positively associated with eGFR and log(UACR), whereas the blood mixture was associated with higher log(UACR) and higher odds of the CKD-compatible phenotype. Formal interaction analyses identified age-related heterogeneity in selected blood-based associations, while inverse associations of blood Pb, blood Cd, and urinary Cd with eGFR were stronger among participants with diabetes. Outcome-oriented network analysis consistently prioritized IL1B and IL6 as shared inflammatory hubs, while peripheral network architecture differed across renal modules. In HEK293T cells, combined Pb-Cd exposure increased intracellular metal burden and cytotoxicity, reduced antioxidant capacity, and altered the expression of IL1B, IL6, and selected stress-response genes. CONCLUSION: Blood- and urine-based Pb and Cd biomarkers exhibited distinct association profiles across renal outcomes, indicating that a single biomarker matrix or endpoint may incompletely characterize Pb-Cd-associated renal abnormalities. Outcome-oriented networks converged on IL1B and IL6 as shared inflammatory hubs within distinct peripheral architectures, with targeted cellular evidence supporting the responsiveness of prioritized targets. These findings establish an integrative, multi-matrix and multi-outcome framework for the mechanistic interpretation and risk assessment of Pb-Cd co-exposure.