Integrating Multimodal Data and Genomics for a Comprehensive Assessment of Cardiovascular Aging and Its Impact

Area of Science:

• Cardiovascular Medicine
• Biomarkers
• Genetics

Background:

• Age-related changes in the heart and blood vessels increase cardiovascular disease (CVD) risk.
• Comprehensive assessments using multimodal imaging and genetic data are limited.
• Quantifying cardiovascular aging is crucial for understanding CVD development.

Purpose of the Study:

• To quantify cardiovascular aging using multimodal biomarkers.
• To evaluate the genetic architecture and lifestyle determinants of cardiovascular aging.
• To assess the prognostic relevance of a novel cardiovascular aging metric.

Main Methods:

• Integrated cardiovascular magnetic resonance (CMR), electrocardiogram, arterial stiffness, and carotid ultrasound biomarkers from UK Biobank data.
• Developed and validated machine learning (ML) models to calculate cardiovascular age gap (CardioAG).
• Conducted whole-genome sequencing (WGS) and regression analyses to identify genetic variants and lifestyle associations with CardioAG.

Main Results:

• A CatBoost ML model best predicted cardiovascular age (Pearson r = 0.75).
• CardioAG was independently associated with increased risk of hypertension, stroke, atrial fibrillation, coronary artery disease, and major adverse cardiovascular events (MACE).
• Multimodal integration improved predictive accuracy and prognostic value; novel genetic variants (e.g., RN7SKP155, SVIL, CBFA2T3) and lifestyle factors were linked to CardioAG.

Conclusions:

• Developed a unified multimodal cardiovascular aging metric integrating cardiac, vascular, and electrophysiological features.
• This metric provides incremental prognostic utility for CVD prediction.
• Broadened the understanding of the biological, genetic, and lifestyle underpinnings of cardiovascular aging.