Overview

• Using whole-body and cardiac MRIs from 21,241 UK Biobank participants, a machine-learning model derived an age-delta reflecting vascular stiffness, cardiac motion, and myocardial fibrosis.
• Visceral, liver, and muscle-infiltrated fat were among the strongest predictors of accelerated cardiovascular ageing in both sexes, outperforming body mass index as a risk indicator.
• Sex-specific patterns emerged, with android and abdominal subcutaneous fat linked to higher age-delta in men, while gynoid fat was protective in premenopausal women and higher estradiol correlated with slower ageing.
• Physical activity modestly attenuated the adverse association of visceral fat with cardiovascular ageing, diabetes amplified it, and metformin use was associated with a somewhat reduced effect.
• The analysis is cross-sectional and drawn from a cohort with limited diversity, prompting calls for longitudinal studies and trials testing fat-targeting interventions such as GLP-1 therapies.