Abstract Body (Do not enter title and authors here): Background
Electrocardiogram-derived age is a novel biomarker generated using artificial intelligence that reflects physiological rather than chronological aging. A higher electrocardiogram-derived age relative to chronological age, or “delta age,” has been linked to increased cardiovascular risk. While behavioral changes have been shown to improve delta age, the effect of cardiometabolic drug therapy on this biomarker remains unclear.
Methods
We conducted a retrospective cohort study using electronic health records mapped to a common data model. Adults with at least two electrocardiograms within 24 months who initiated therapy with a glucagon-like peptide-1 receptor agonist, sodium-glucose cotransporter 2 inhibitor, or proprotein convertase subtilisin/kexin type 9 inhibitor were included. Delta age was defined as electrocardiogram age minus chronological age. Patients were categorized as having accelerated (>5 years), normal (±5 years), or decelerated (<−5 years) aging. We calculated time-weighted mean delta age before and after drug initiation to account for irregular electrocardiogram timing. The primary outcome was the change in time-weighted mean delta age and transition in aging category. Odds ratios were used to quantify the likelihood of improvement.
Results
Among 1,178 patients with adequate data, 78.6 percent experienced a reduction in delta age following treatment. Of those with accelerated aging at baseline, 56 percent transitioned to normal or decelerated aging. Favorable category transitions were supported by strong odds ratios: 14.03, 9.88, and 10.57, calculated from contingency tables. These findings suggest that cardiometabolic therapy is strongly associated with improvement in biological aging classification based on electrocardiogram signals.
Conclusion
Cardiometabolic drug initiation was associated with significant reductions in electrocardiogram-derived biological age. Patients with accelerated aging were the most likely to benefit, with over half shifting to a healthier classification. These results support electrocardiogram-derived age as a dynamic, treatment-responsive biomarker with potential clinical utility in monitoring therapeutic response and guiding preventive cardiovascular care.