Objectives: We sought to assess the prognostic value of heart rate variability measures, including Poincaré plots, in patients with mild to moderate chronic heart failure.
Background: Mortality is high in patients with heart failure, and many of them die suddenly. However, identification of high risk patients, particularly those with an increased risk for sudden death, has remained difficult.
Methods: We studied 95 patients with heart failure (mean [+/- SD] age 60 +/- 8 years, left ventricular ejection fraction 0.29 +/- 0.09, New York Heart Association functional class II [81%] and III [19%]) during up to 4 years of follow-up. Heart rate variability measures and Poincaré plots were obtained from 24-h Holter recordings.
Results: During follow-up, 17 (18%) of the 95 patients died. In 15 patients, death was cardiac related (11 patients experienced sudden death). None of the conventional time and frequency domain measures of heart rate variability were related to survival. In contrast, abnormal Poincaré plots identified a significantly higher risk for all-cause cardiac death (Cox proportional hazards ratio 5.7, 95% confidence interval [CI] 1.6 to 20.6, univariate analysis) and for sudden cardiac death (hazards ratio 6.8, 95% CI 1.5 to 31.4) compared with those with normal Poincaré plots. Patients with abnormal Poincaré plots were shown to have a lower left ventricular ejection fraction (0.26 +/- 0.10 vs. 0.31 +/- 0.08, p < 0.05) and higher plasma norepinephrine concentrations (506 +/- 207 pg/ml vs. 411 +/- 175 pg/ml, p < 0.05). In multivariate analysis, abnormal Poincaré plots still had independent prognostic value, both for all-cause cardiac mortality and for sudden cardiac death (hazards ratio 5.3, 95% CI 1.2 to 17.1, hazards ratio 4.5, 95% CI 1.0 to 27.5, respectively.
Conclusions: Heart rate variability analysis, as assessed by Poincaré plots, has independent prognostic value in patients with mild to moderate chronic heart failure and identifies an increased risk for all-cause and sudden cardiac death in these patients.