Liver- and Immune-Enriched Molecular Signatures Associated With Mortality in Older Adults.
Key claim: A 20-protein plasma mortality score enriched for liver- and immune-related proteins predicted all-cause mortality better than metabolite scores or conventional clinical risk factors in older adults, with directionally consistent external validation.
Abstract
Quantifying biological aging requires biomarkers that capture multisystem physiological decline beyond chronological age. We aimed to compare the prognostic performance of plasma proteomics, metabolomics, and conventional clinical risk factors for all-cause mortality, and to characterize molecular pathways associated with mortality risk and age-related physiological decline. Untargeted plasma proteomics and metabolomics were profiled in 848 community-dwelling adults from the I-Lan Longitudinal Aging Study (ILAS), followed for a mean of 8.5 years, during which 92 deaths occurred. Cox proportional hazards models were applied to identify mortality-associated molecular features and their enriched biological pathways. We identified 79 proteins associated with all-cause mortality (FDR < 0.1), predominantly liver-derived and immune-related. Pathway analysis revealed coordinated dysregulation across coagulation cascades, complement activation, oxidative stress responses, glucose metabolism, and bile acid metabolism. Elastic net regression was subsequently used to construct omics-based mortality risk scores, which were further evaluated in an independent validation cohort from the Longitudinal Aging Study of Taipei (LAST). A 20-protein mortality score achieved strong discrimination (C-index 0.81), outperforming metabolite-based models (C-index 0.77) and clinical risk factors (C-index 0.73) in the discovery cohort. In the external validation cohort, both proteomic- and metabolomic-derived scores showed directionally consistent associations with mortality risk. Together, these findings identify liver- and immune-enriched molecular signatures associated with mortality risk in older adults and support the utility of plasma multi-omics profiling as a scalable precision tool for biological age assessment. The identified proteomic and metabolomic pathways may help inform future interventions targeting systemic aging and age-related functional decline.
Why this matters
Most proteomic aging signatures are reported in isolation; this finding is one of the rarer head-to-head comparisons showing a curated plasma protein score beats both metabolite scores and the standard clinical risk panel for mortality prediction in older adults, with external validation. The liver- and immune-enriched composition points to specific organ-system biology driving late-life mortality risk rather than a diffuse ‘aging’ signal, sharpening what a next-generation clinical panel might need to measure.