Background: Morbidity and mortality due to chronic heart failure remain unacceptably high despite effective drug therapies, and the search for a better risk predictor is ongoing. Statistics derived from beat-to-beat fluctuations in heart rate or heart rate variability (HRV) have been used for this purpose, but the current predictability level is low or moderate at best.
Objective: The purpose of this study was to evaluate whether a recently proposed non-Gaussian index of HRV is a significant and independent mortality predictor in patients with congestive heart failure (CHF).
Methods: Twenty-four-hour Holter ECGs from 108 CHF patients were evaluated. Thirty-nine (36.1%) of the patients died during the follow-up period of 33 +/- 17 months. Cox proportional hazards regression analysis was performed to determine factors related to all-cause mortality. The factors evaluated derived from clinical information, including plasma brain natriuretic peptide, conventional time- and frequency-domain and fractal HRV measures, and a recently proposed non-Gaussian index lambda of HRV.
Results: The short-term (<40 beats) non-Gaussian index lambda(40) (hazard ratio per increment of unit standard deviation 1.64, 95% confidence interval [1.23, 2.18], P <.001) and the long-term (<1,000 beats) index lambda(1000) (hazard ratio 1.42, 95% confidence interval [1.07, 2.18], P <.02), together with brain natriuretic peptide (hazard ratio 2.26, 95% confidence interval [1.45, 3.53], P <.001), are significant univariate risk predictors of mortality. In a multivariate model, lambda(40) (1.49, [1.13, 1.96], P <.005) and brain natriuretic peptide (2.39, [1.53, 3.75], P <.001) are independent predictors of the survival statistics of patients. None of the conventional HRV measures have predicted the mortality of patients in a significant and independent manner.
Conclusion: The results of this study indicate the usefulness of the short-term non-Gaussian index of HRV for risk prediction in patients with CHF.