Fractal analysis and time- and frequency-domain measures of heart rate variability as predictors of mortality in patients with heart failure.

Time-domain measures of heart rate (HR) variability provide prognostic information among patients with congestive heart failure (CHF). The prognostic power of spectral and fractal analytic methods of HR variability has not been studied in the patients with chronic CHF. The aim of this study was to assess whether traditional and fractal analytic methods of HR variability predict mortality among a population of patients with CHF. The standard deviation of RR intervals, HR variability index, frequency-domain indexes, and the short-term fractal scaling exponent of RR intervals were studied from 24-hour Holter recordings in 499 patients with CHF and left ventricular ejection fraction < or =35%. During a mean follow-up of 665 +/- 374 days, 210 deaths (42%) occurred in this population. Conventional and fractal HR variability indexes predicted mortality by univariate analysis. For example, a short-term fractal scaling exponent <0.90 had a risk ratio (RR) of 1.9 (95% confidence interval [CI] 1.4 to 2.5) and the SD of all RR intervals <80 ms had an RR of 1.7 (95% CI 1.2 to 2.1). After adjusting for age, functional class, medication, and left ventricular ejection fraction in the multivariate proportional-hazards analysis, the reduced short-term fractal exponent remained the independent predictor of mortality, RR 1.4 (95% CI 1.0 to 1.9; p <0.05). All HR variability indexes were more significant univariate predictors of mortality in functional class II than in class III or IV. Among patients with moderate heart failure, HR variability measurements provide prognostic information, but all HR variability indexes fail to provide independent prognostic information in patients with the most severe functional impairment.