Fractal Analysis of Heart Rate Dynamics as a Predictor of Mortality in Patients With Depressed Left Ventricular Function After Acute Myocardial Infarction

A number of new methods have been recently developed to quantify complex heart rate (HR) dynamics based on nonlinear and fractal analysis, but their value in risk stratification has not been evaluated. This study was designed to determine whether selected new dynamic analysis methods of HR variability predict mortality in patients with depressed left ventricular (LV) function after acute myocardial infarction (AMI). Traditional timeand frequency-domain HR variability indexes along with short-term fractal-like correlation properties of RR intervals (exponent a) and power-law scaling (exponent b) were studied in 159 patients with depressed LV function (ejection fraction <35%) after an AMI. By the end of 4-year follow-up, 72 patients (45%) had died and 87 (55%) were still alive. Short-term scaling exponent a (1.07 6 0.26 vs 0.90 6 0.26, p <0.001) and power-law slope b (21.35 6 0.23 vs 21.44 6 0.25, p <0.05) differed between survivors and those who died, but none of the traditional HR variability measures differed between these groups. Among all analyzed variables, reduced scaling exponent a (<0.85) was the best univariable predictor of mortality (relative risk 3.17, 95% confidence interval 1.96 to 5.15, p <0.0001), with positive and negative predictive accuracies of 65% and 86%, respectively. In the multivariable Cox proportional hazards analysis, mortality was independently predicted by the reduced exponent a (p <0.001) after adjustment for several clinical variables and LV function. A short-term fractal-like scaling exponent was the most powerful HR variability index in predicting mortality in patients with depressed LV function. Reduction in fractal correlation properties implies more random short-term HR dynamics in patients with increased risk of death after AMI. Q1999 by Excerpta Medica, Inc. (Am J Cardiol 1999;83:836–839)