Linear measures of heart rate variability (HRV) during ECG stress test has been widely questioned due to the high signal non-stationary. Such limitations can be overcome by nonlinear measures of HRV based on typical properties of chaotic systems and deterministic fractal. Very few paper addressed such issue and aim of the paper is to describe fractal behavior of HRV during exercise. A fractal analysis by Higuchi's algorithm (FD) has been performed on 26 cardiac patients during resting, stress, early and late recovery phases of ECG stress test. Results showed a significant FD increasing values from resting to stress phase that was not recovered at all immediately after the exercise, and it was slightly recovered both during early and late recovery phase. The performance of fractal analysis of HRV during and after high intensity exercise suggests that it could be a useful index assessing relevant information about underlying physiological recovery.
