Abstract 4364268: Structural Cardiac Abnormalities Modify Atrial Fibrillation–Associated Heart Failure Risk and Identify a High-Risk Phenotype in Older Adults: The ARIC Study

Abstract

Background: The combined impact of atrial fibrillation (AF) and structural cardiac abnormalities on heart failure (HF) risk remains incompletely understood in older adults. We examined whether echocardiographic parameters and cardiac biomarkers modify the association between AF and risk of incident HF. Methods: Among 5,572 participants aged ≥65 years without baseline HF from the ARIC study, we investigated the association between prevalent AF and incident HF over median follow up of 9 years. We tested interactions between AF and cardiac biomarkers (hs-cTnT, hs-cTnI, and NT-proBNP as continuous variables) and echocardiographic parameters including left atrial volume index (LAVI), average E/e', left ventricular ejection fraction (LVEF), longitudinal strain, left ventricular mass index (LVMI), and mitral regurgitation jet area as continuous variables using Cox regression. Significant interactions were further studied using stratified analyses with clinically relevant cutpoints. Results: Among participants (mean age 76, standard deviation: 5 years, 60% female, 21% Black), 359 (6.4%) had prevalent AF. Baseline AF was significantly associated with HF risk (hazard ratio [HR]: 2.52, 95% confidence interval [CI]: 1.95-3.25). No significant interactions were found with cardiac biomarker analyses. Significant interactions were observed between AF and LAVI (p=0.006), E/e' (p=0.043), and mitral regurgitation jet area (p=0.015), with attenuation of the effect of AF on incident HF risk in the presence of abnormal echocardiographic parameters. Participants with concomitant AF and abnormal echocardiographic findings had the highest absolute risk of incident HF: those with AF and LAVI ≥35 ml/m 2 had an event rate of 50.1 per 1000 person-years (HR: 3.80, 95% CI: 2.71-5.34 vs. no AF + LAVI <35 ml/m 2 ), while those with AF and E/e' >14 had an event rate of 49.6 per 1000 person-years (HR: 3.68, 95% CI: 2.37-5.71 vs. no AF + E/e' < 14). Conclusion: In older adults, the association between AF and HF varies by cardiac structural parameters, with attenuated effects in those with pre-existing structural abnormalities. Despite effect attenuation, absolute risk remains highest in those with both abnormal echocardiographic findings and prevalent AF, identifying a high-risk phenotype that can inform approaches to risk stratification in older adults and may warrant intensified prevention strategies.