Biological age uncertainty reveals discordant aging and predicts human healthspan

Biological aging clocks typically summarize complex biological profiles as a single age estimate, yet how uncertain these estimates are and whether this uncertainty represents a distinct, biologically meaningful dimension of human aging remain unclear. Here, we develop a probabilistic framework across nuclear magnetic resonance (NMR) metabolomic, plasma proteomic and routine blood test (RBT) profiles in more than 450,000 UK Biobank participants to derive biological age gap (BAG) and biological age uncertainty (BAU) from the mean and width of predicted-age distributions. BAU was weakly correlated with BAG and increased with age-signal disagreement across individual molecules and organs, linking uncertainty to discordant aging across biological systems. Higher BAU predicted mortality, loss of healthspan(definition) and multimorbidity independently of BAG, with mortality hazard ratios of 1.24 (95% CI, 1.23-1.26) and 1.40 (95% CI, 1.37-1.43) per standard deviation for NMR- and proteomic-derived BAU, respectively. Adding BAU to conventional clinical models improved 10-year risk discrimination; moreover, longitudinal increases in NMR-derived BAU tracked subsequent health decline, and mortality associations were replicated using RBT-derived BAU across three independent external cohorts. Multi-omics and physiological analyses linked BAU to inflammation, metabolic dysregulation and tissue remodeling beyond BAG. Genetic analyses further revealed overlapping but distinct architectures of BAU and BAG, with BAU showing stronger genetic correlations with frailty and cardiometabolic diseases, and tissue and functional mapping implicating hepatic, immune and lipid-metabolic processes. This study establishes a framework for characterizing biological age through both its mean deviation and uncertainty, defining BAU as an independent dimension of human aging that links discordant aging to future healthspan.