Integrated multi-omics reveals roles and interactions of plasma metals, metabolites, and blood clinical indicators in cervical cancer

This study applied integrated multi-omics to explore the role of plasma elements, blood clinical indicators, and metabolome in cervical cancer (CC). Based on data from 436 subjects, untargeted/targeted metabolomics and ICP-MS analyses revealed significant plasma metabolome reprogramming in CC. This was characterized by upregulated levels of estrone glucuronide and cortisol, and downregulation of androstenedione. A random forest model constructed with four metabolites (androstenedione, calcitriol, paraxanthine, theobromine) achieved an AUC of 0.884, superior to conventional markers (CEA, CA125). Among nine metals linked to CC risk, Cd, Cr, Pb, Ba, and Li increased risk, while Mn, Cu, and Sn were protective; mixed metal exposure elevated risk dose-dependently. CC patients also showed abnormal blood parameters, including reduced RBC and lymphocyte counts, and higher TG, GLU, CEA, and CA125. Multi-omics network analysis indicated co-clustering of Li, Cr, and Co with estrone glucuronide, 5-HIAA, GLU, PCT, and CA125; Pb and Ba with L-valine; Mn with paraxanthine, androstenedione, A/G, and GLOB; cortisol, caffeine, and theobromine with RBC and LYM#. These integrated network highlights complex interactions among metals, clinical indicators, and metabolites, providing new insights for non-invasive diagnosis and environmental health interventions in CC. • Large-scale UHPLC–Orbitrap–HRMS plasma profiling revealed profound metabolome reprogramming in cervical cancer, with marked disruption of steroid hormone pathways underpinning evaluation of endocrine-disrupting chemical effects. • Discovery and targeted validation of HPO/HPA–linked plasma markers showed that higher levels of caffeine metabolites (paraxanthine, theobromine) and calcitriol (1,25-dihydroxyvitamin D₃) associate with reduced cervical cancer risk; LC–MS/MS assays and a machine-learning model outperformed CEA, CA125, and their combination for diagnosis. • ICP–MS–based plasma elemental profiling combined with single-metal and mixture models (including BKMR) demonstrated that metal co-exposure is strongly associated with increased cervical cancer risk and exhibits interaction effects. • Integrative multi-omics networking of plasma metabolites, blood clinical indices, and elemental profiles linked exposure, metabolism, and disease, prioritizing candidate biomarkers and targets for intervention.