A blood aging clock defines disease-induced deviation from healthy molecular aging

Abstract

Aging shapes disease susceptibility and recovery, yet it remains unclear whether disease-induced molecular perturbations can be quantitatively mapped onto healthy aging trajectories. Here, we integrated 10 753 whole-blood transcriptomic profiles from 23 cohorts and developed a blood aging predictor (BAP) using 5808 healthy individuals to define a reference trajectory of blood molecular aging. Applying BAP to COVID-19 revealed marked transcriptomic age acceleration (AgeAccel), with patients showing an average AgeAccel of 9.34 years after cohort-specific baseline correction. This aging-referenced deviation occurred across age groups and was associated with interferon signaling, immune-inflammatory activation, cytokine and neutrophil-related programs, DNA damage responses, proteostasis stress, and impaired cellular quality-control pathways. Longitudinal analysis showed that AgeAccel increased during acute infection but declined during recovery, indicating a potentially reversible aging-like molecular state at the cohort level rather than irreversible aging. Extension to sepsis, influenza, and major depressive disorder further demonstrated that elevated AgeAccel captures disease-associated molecular burden across distinct pathological contexts. Together, our study redefines transcriptomic AgeAccel as an interpretable measure of disease-induced deviation from healthy molecular homeostasis and provides a quantitative framework for assessing systemic disease burden and recovery dynamics from peripheral blood transcriptomes.