Depressed low frequency power of heart rate variability as an independent predictor of sudden death in chronic heart failure

AIMS: Identification of patients with chronic heart failure at risk for sudden death remains difficult. We sought to assess the prognostic value for all-cause and sudden death of time and frequency domain measures of heart rate variability in chronic heart failure. METHODS AND RESULTS: We prospectively enrolled 190 patients with chronic heart failure in sinus rhythm, mean age 61+/-12 years, 109 (57.4%) in NYHA class II and 81 (42.6%) in classes III or IV, mean cardiothoracic ratio 57.6+/-6.4% and mean left ventricular ejection fraction 28.2+/-8.8%, 85 (45%) with ischaemic and 105 (55%) with idiopathic dilated cardiomyopathy. Time and frequency domain measures of heart rate variability were obtained from 24 h Holter ECG recordings, spectral measures were averaged for calculation of daytime (1000h-1900h) and night-time (2300h-0600h) values. During follow-up (22+/-18 months), 55 patients died, 21 of them suddenly and two presented with a syncopal spontaneous sustained ventricular tachycardia. In multivariate analysis, independent predictors for all-cause mortality were: ischaemic heart disease, cardiothoracic ratio > or =60% and standard deviation of all normal RR intervals <67 ms (RR = 2.5, 95% CI 1.5-4.2). Independent predictors of sudden death were: ischaemic heart disease and daytime low frequency power <3.3 ln (ms(2)) (RR = 2.8, 95% CI 1.2-8.6). CONCLUSION: Depressed heart rate variability has independent prognostic value in patients with chronic heart failure; spectral analysis identifies an increased risk for sudden death in these patients.