Reduced ventricular response irregularity is associated with increased mortality in patients with chronic atrial...

BACKGROUND-Variations in the ventricular response interval (VRI) during atrial fibrillation (AF) may be reduced in patients with adverse clinical outcomes. The properties of VRI dynamics associated with prognosis remain undetermined. METHODS AND RESULTS-In 107 patients with chronic AF (age, 64+/-9 years), we analyzed a 24-hour ambulatory ECG for VRI variability (SD, SD of successive differences, and SD of 5-minute averages) and VRI irregularity (Shannon entropy of histogram, symbolic dynamics, and approximate entropy of beat-to-beat and minute-to-minute fluctuations [ApEn(b-b) and ApEn(m-m)]). During a follow-up period of 33+/-16 months, 18 patients died (17%), 9 from cardiac causes, 7 from fatal strokes, and 2 from malignancies. Reductions in all VRI variability and irregularity measures were associated with an increased risk for cardiac death but not for fatal stroke. A significant association with cardiac death was also found for ejection fraction (relative risk, 1.10; 95% confidence interval [CI], 1.04 to 1.17, per 1% decrement) and ischemic AF (relative risk, 6.52; 95% CI, 1.62 to 26. 3). After adjustment for these clinical variables, all irregularity measures except symbolic dynamics had predictive value (relative risks [95% CIs] per 1SD decrement: Shannon entropy of histogram, 2. 03 [1.14 to 3.61]; ApEn(b-b), 1.72 [1.14 to 2.60]; and ApEn(m-m), 1. 90 [1.03 to 3.52]); however, the predictive power of variability measures was no longer significant. When the patients were stratified with the 33rd and 67th percentile values of ApEn(b-b) (1. 83 and 1.94, respectively), the 5-year cardiac mortality rates for the upper, middle, and lower tertiles were 0%, 13%, and 43%, respectively (log-rank test, P=0.04). CONCLUSIONS-Reduced VRI irregularity in a 24-hour ambulatory ECG has an independent prognostic value for cardiac mortality during long-term follow-up in patients with chronic AF.