Meta-analysis of genome-wide associations and polygenic risk prediction for atrial fibrillation in more than 180,000 cases.

Summary

By meta-analyzing AF GWAS (>180,000 cases), the study identifies over 350 loci and functionally annotates 139 through chromatin assays in atrial cardiomyocytes. An updated AF polygenic risk score outperforms both the CHARGE-AF clinical score and prior PRS, strengthening genomic risk prediction.

Key Findings

• Meta-analysis identifies >350 AF-associated loci, doubling known risk architecture.
• At 139 loci, candidate genes tied to muscle contractility, cardiac development, and cell-cell communication are prioritized.
• Chromatin accessibility (ATAC-seq) and H3K4me3 support sentinel variant activity in atrial cardiomyocytes.
• An updated AF PRS outperforms CHARGE-AF and prior PRS in risk prediction.

Clinical Implications

Improved AF polygenic risk scoring may augment clinical risk tools (e.g., CHA2DS2-VASc context) and identify high-risk individuals for monitoring, prevention, or early rhythm-control strategies.

Limitations

• Summary-level meta-analysis limits fine-grained phenotype resolution and gene-environment interaction assessment
• PRS transferability across diverse ancestries requires further validation and optimization

Future Directions

Implement PRS in prospective cohorts and health systems, test ancestry-specific/transfer learning approaches, and integrate PRS with rare variants for composite AF genetic risk tools.