Accelerated electrocardiographic aging of the heart as a risk determinant for atrial fibrillation: A Mendelian randomization study

Abstract

Background: Artificial intelligence (AI)-derived electrocardiographic aging (ECG-aging) has been found to be associated with the risk of atrial fibrillation (AF). Objectives: We aimed to assess the causal association between the discrepancy in AI-predicted electrocardiographic age and chronological age (AI-ECG age gap) and AF risk. Methods We analyzed 12-lead ECGs in UK Biobank to derive the AI-ECG age gap using our latest AI-based age prediction model. Associations between measured ECG-aging and genetically predicted ECG-aging based on a genetic risk score (GRS) were evaluated in relation to AF using multivariable regression and Mendelian randomization (MR). MR analyses incorporated GRS-based individual-level data and summary-level genome-wide association study (GWAS) data from large external consortia, with causal estimates obtained using inverse-variance weighting and complementary sensitivity methods. Mediation MR was additionally performed to assess intermediary causal pathways. Results: Each 1-SD increase in the AI-ECG age gap was associated with an increased risk of AF in observational analyses (HR 1.43 [95% CI, 1.29-1.59]) and GRS-based MR (OR 1.06 [95% CI, 1.01-1.12]). Two-sample MR analyses in two large, independent GWAS datasets replicated these findings (OR 1.13 [95% CI, 1.02-1.25] in both), with multiple sensitivity analyses confirming robustness. Bidirectional MR suggested potential reverse causation from AF to ECG-aging. Non-linear MR demonstrated a consistent positive association across the AI-ECG age gap spectrum. Mediation MR further identified heart failure as a key intermediary, accounting for approximately 70% of the total causal effect of ECG-aging on AF. Conclusion: This study provides genetic evidence supporting a causal association between ECG-aging and the risk of AF. These findings highlight ECG-aging as a causally relevant and clinically informative biomarker for AF that may help identify individuals at increased risk.