A novel framework to estimate cardiovascular biological age, termed HeartAge, and its deviation from chronological age (HeartAge-gap) was evaluated in 31,784 participants from the UK Biobank. The study, presented at ESC 2025, used cardiovascular magnetic resonance data to generate HeartAge through machine learning and validated findings in 897 participants from the Multi-Ethnic Study of Atherosclerosis.

Positive HeartAge-gap values, indicating an older cardiovascular system, were independently associated with hypertension, diabetes, and ischemic heart disease in both sexes. Over nearly six years of follow-up, higherHeartAge-gap predicted composite cardiovascular outcomes in females (hazard ratio 1.022) and males (hazard ratio 1.017), independently of chronological age and other confounders such as body mass index, preexisting cardiovascular disease, and diabetes. Notably, in females, advanced cardiovascular ageing also predicted all-cause mortality regardless of chronological age.

These results suggest that HeartAge-gap provides an individualized, biologically informed metric of cardiovascular risk and could complement traditional risk stratification, offering new opportunities for prevention and early intervention.