Key result
Multimodal deep learning reveals 38 genomic loci for cardiovascular aging, with only ~1% multi-system involvement.
Why the study?
Individuals of the same chronological age can differ markedly in the organs and biological pathways through which cardiovascular vulnerability emerges.
Cohort (n=101,621)
Yes
Multimodal deep learning demonstrates that cardiovascular biological aging is not a single systemic clock but rather decomposes into four genetically and clinically distinct modules (electrical, myocardial structural, macrovascular, and microvascular) that progress independently and confer domain-specific cardiometabolic disease risks.
Answers are grounded in this paper and cite their sources.
Supports modular cardiovascular aging with distinct genetics; leaves open clinical utility of multi-modal age gaps pending validation.
Age is the dominant risk factor for cardiovascular disease, yet individuals of the same chronological age can differ markedly in the organs and biological pathways through which cardiovascular vulnerability emerges. We used deep learning to estimate biological age from four cardiovascular data streams in more than 100,000 UK Biobank participants - 12-lead electrocardiograms, cardiac magnetic resonance imaging, carotid ultrasound, and retinal fundus photographs - representing electrical, structural, macrovascular, and microvascular domains. The resulting biological age gaps showed limited overlap across modalities in participants with complete phenotyping, and phenome-wide analyses linked each axis to distinct disease profiles. Genome-wide association analyses identified 38 independent loci, with largely modality-specific signals involving electrophysiologic, myocardial structural, vascular regulatory, and ocular/microvascular pathways. Cross-trait LD score regression and polygenic risk scores further supported partial genetic separation of the four axes, with modality-specific associations for atrial fibrillation, heart failure, peripheral arterial disease, hypertension, diabetes, and diabetic retinopathy in UK Biobank and broadly concordant patterns in All of Us. Integration with myocardial single-cell transcriptomic data nominated distinct cellular contexts for these genetic signals. These findings suggest that AI-derived cardiovascular age is not a single biomarker of systemic senescence, but a family of related phenotypes that decompose cardiovascular aging into measurable electrical, myocardial, macrovascular, and microvascular modules.
No takes yet. Share an insight, caveat, or question.
Choi et al. (2026) conducted a cohort in Cardiovascular aging (n=101,621). Deep learning-derived modality-specific biological age gaps (BAG) and polygenic risk scores (PRS) across ECG, CMR, carotid ultrasound, and retinal imaging vs. Chronological age and cross-modality aging comparisons was evaluated on Modality-specific biological age gaps (BAG), independent genomic risk loci on GWAS, and polygenic risk score associations with cardiometabolic outcomes (e.g., atrial fibrillation with ΔCMR-age PRS) (OR 1.10, p=<5e-8). Deep learning biological age gaps across 4 cardiovascular modalities identified 38 distinct genomic risk loci, with only 1.0% of participants showing accelerated aging across all systems.
Synapse has enriched 10 related papers. Consider them for comparative context: