Abstract 14269: The Role of Heart Rate Variability in Predicting Sudden and Non-Sudden Cardiac Death in Ischemic Heart Disease
Background: Depressed left ventricular ejection fraction (i.e., LVEF ≤ 35%) remains the only clinically used tool to predict sudden cardiac arrest (SCA). However, this approach has been found insufficient to identify all patients at greater risk of SCA. Alternatively, altered sympathetic and vagal control mechanisms are known to play a significant role in the pathogenesis of SCA, and evidence suggests that heart rate variability (HRV) strongly correlates with such modulation mechanisms. The analysis of HRV can, therefore, predict SCA independently from LVEF, allowing for the targeting of implantable cardioverter defibrillator (ICD) therapy.
Methods: Subjects were recruited from the Prediction of ARrhythmic Events with Positron Emission Tomography (PAREPET) study. Twenty-four hour Holter ECG recordings were obtained at the start of study, and follow up of subjects’ condition was conducted by phone at 3-month intervals. Patients with persistent pacing, atrial fibrillation, or low-fidelity recordings were excluded. Eight measures of time (e.g., SDNN and RMSSD), frequency (e.g., power, very low, low, high, and low/high ratio), and non-linear (e.g., approximate entropy) trends in HRV were automatically analyzed. Endpoints were evaluated by an adjudication committee. ICD discharges due to ventricular fibrillation or tachycardia (i.e., > 240 beats/min) were considered equivalent to SCA. Other non-arrhythmic cardiovascular deaths were classified as cardiac non-sudden death (C/NS).
Results: The sample was mainly composed of older men (n = 127, age 67 ± 12, 92% male, LVEF 28 ± 9%). After a median follow up of 4 years, there were 22 (i.e., 17%) SCA cases and 18 (i.e., 14%) C/NS cases. In univariate analyses, the low frequency power component of HRV, a marker of myocardial sympathetic denervation, could differentiate cases from survivors. In multivariate cox regression, reduced low frequency power (i.e., < 26%) was a significant predictor of SCA (HR = 6.3 [95% CI 1.5-26.3]) and C/NS (HR = 4.9 [95% CI 1.2-19.9]) independent of age, sex, and LVEF.
Conclusions: Frequency analysis of HRV can identify not only patients at greater risk for SCA for optimal targeting of early ICD therapy, but also those at greater risk for C/NS death for more aggressive medical therapy.
Author Disclosures: S.S. Al-Zaiti: None. M.G. Carey: Research Grant; Modest; NIH/NINR. J.M. Canty: None. J.A. Fallavollita: None.
- © 2014 by American Heart Association, Inc.