Proteomic organ-specific signatures, ageing traits, disease risks and genetic architecture in an East Asian population

BACKGROUND: Protein-based ageing clocks can be used to help identify individuals at high risk of death or morbidity. There is a lack of evidence in non-European populations on clocks derived from combining different proteomic platforms and their relationship with age-related traits, diseases and genetic architecture. METHODS: The prospective China Kadoorie Biobank assayed proteins via Olink and SomaScan in ∼4000 individuals with a mean age of 58 years. We used organ-enriched plasma proteins identified from the Genotype-Tissue Expression Project and trained Light Gradient Boosting models on chronological age (ChronAge) to derive protein organ age and age gaps (ProtAgeGap) for 18 organs. We then investigated their relationship with age-related traits and incident diseases after adjustment for multiple testing. Moreover, we conducted genome-wide association studies to identify genetic variants for overall and organ-specific ProtAgeGaps. FINDINGS: The overall proteomic and organ-specific ageing clocks for each platform combined for all participants were strongly related with ChronAge. The organ-specific and overall proteomic age were associated with age-related traits and a range of incident diseases independent of ChronAge. In particular, the kidney organ-specific ageing clock was associated with higher risk of stroke (1.03; 1.02-1.05), chronic liver disease (CLD, 1.11; 1.05-1.18); chronic kidney disease (CKD, 1.19; 1.11-1.27) per 1-year higher ProtAgeGap. GWAS of overall proteomic age identified AFF3 and IL1RAPL1 as associated with biological ageing. INTERPRETATION: Advanced proteomic ageing, in particular kidney organ ageing, was associated with age-related traits and diseases and, pending further validation, may have utility as a potential biomarker for clinical trials. FUNDING: British Heart Foundation.