Abstract 14987: Association of ECG Abnormalities and Incident Heart Failure Events

Abstract

Objectives We assessed in elderly persons the association between ECG abnormalities and incident heart failure (HF) events and the predictive value of ECG when added to a HF risk score.

Background Limited prospective data exist regarding the association between ECG abnormalities and incident HF events in elderly adults.

Methods We followed 2915 men and women enrolled in the Health ABC study, aged 70-79 and without preexisting HF at baseline, for a median period of 11.4 years. We used the Minnesota Code ECG Classification to define minor (minor ST-T changes) and major (major ST-T changes, Q wave, left ventricular hypertrophy, atrial fibrillation, atrio-ventricular bloc) abnormalities at baseline and at 4 years. Using Cox models, we assessed the value of adding ECG abnormalities to the Health ABC HF Risk Score to predict documented HF events. Results At baseline, 380 participants (13.0%) had minor and 620 (21.3%) major ECG abnormalities. During follow-up, 485 (16.6%) developed incident HF events. After adjusting for the variables of the HF Risk Score, the hazard ratio (HR) was 1.27 (95% confidence interval [CI] 0.96-1.67) for minor and 1.94 (CI 1.56-2.40) for major abnormalities compared to no ECG abnormalities (P for trend <0.01). At 4 years, 263 participants developed new and 549 had persistent ECG abnormalities: adjusted HR was 1.98 (CI 1.32-2.97) for new and 2.24 (CI 1.62-3.08) for persistent abnormalities compared to no ECG abnormalities (P for trend <0.01). ECG abnormalities correctly reclassified 10.6% (CI 3.2%-18.1%) of all participants across the categories of the Health ABC HF Risk Score.

Conclusion Baseline ECG abnormalities, as well as new and persistent abnormalities at 4 years, were 1) associated with an increased risk of HF events and 2) improved HF prediction beyond traditional risk factors.