Among 403,974 UK Biobank participants followed for 13+ years, high-risk liver fibrosis scores (NFS hazard ratio 1.54; FIB-4 HR 1.53; MRI-based cT1 HR 1.41) independently predicted incident atrial fibrillation across 26,677 cases. Integrating 249 NMR metabolomic features with cardiac phenomapping, researchers identified two distinct mechanistic tracks: one linking systemic inflammation and lipotoxicity to electrophysiological changes without chamber enlargement, and another connecting fatty-acid imbalance to structural atrial dilation. A metabolomic risk score (MRS) improved AF prediction beyond established clinical scores, with particularly notable gains over C2HEST (ΔAUC +0.042; cNRI 32.2%), and amplified risk across all genetic susceptibility strata.
This preprint — not yet peer-reviewed — represents one of the most ambitious attempts to mechanistically dissect the liver-heart axis at population scale. The finding that a "lean, normolipidemic" fibro-inflammatory phenotype carries substantial residual AF risk is clinically provocative: it implies conventional lipid panels and BMI measurements miss a meaningful at-risk subgroup. If validated, this repositions liver fibrosis assessment — already accessible via FIB-4 blood tests — as a practical AF screening adjunct. Limitations include the observational design, predominantly European UK Biobank ancestry limiting generalizability, and the modest absolute AUC gains over existing scores in some comparisons. The MRS also requires clinical translation before deployment. Still, the dual-track mechanistic framework and the gene-environment interaction data are genuinely novel contributions, elevating this beyond confirmatory work into potentially paradigm-shifting territory for AF prevention strategy.
