Multimodal Ageing Biomarkers and Plasma Proteomic Signatures Associated with All-Cause Mortality

Area of Science:

• Biogerontology
• Proteomics
• Molecular Epidemiology

Background:

• Biological age estimation using plasma proteins for eleven organs has emerged as a novel predictor of mortality.
• Systematic benchmarking of these proteomic organ ages against established aging biomarkers is currently lacking.

Purpose of the Study:

• To benchmark proteomic organ ages against multimodal aging biomarkers for predicting all-cause mortality.
• To identify plasma proteomic signatures associated with mortality risk.

Main Methods:

• Cox regression analysis was used to assess associations between proteomic organ ages and established aging biomarkers (epigenetic age, telomere length, neuroimaging, cognitive, and physical function) with mortality in the Lothian Birth Cohort 1936 (LBC1936).
• Survival analyses identified plasma proteins associated with mortality using SomaScan 11K data.

Main Results:

• Accelerated liver, immune, and heart ageing were strongly associated with higher mortality risk.
• Established biomarkers including epigenetic age (GrimAge2), total brain volume (TBV), cognitive function (g), and walk time showed stronger mortality prediction than most proteomic organ ages.
• In a multivariable model, TBV, white matter hyperintensity volume, g, and walk time remained significant predictors of mortality.
• Survival analysis identified 202 plasma proteins linked to mortality, with GDF15, CST3, and COL18A1 showing the strongest effects and enrichment in liver and immune pathways.

Conclusions:

• Proteomic organ ages provide insights into mortality risk, but established biomarkers often exhibit superior predictive capacity.
• Plasma proteomic signatures, particularly GDF15, CST3, and COL18A1, are significant mortality predictors and highlight the role of liver and immune processes in aging.
• This study offers a systematic cross-modal benchmarking of novel proteomic aging measures against established ones.