Correlation dimension of heart rate variability: a new index of human autonomic function.

Correlation dimension analysis (CDA) of spontaneous heart rate variability in humans was performed with the use of autonomic blocking agents (propranolol and atropine) and postural change, and was compared with heart rate power spectral analysis (HRPSA). The correlation dimension was increased by the suppression of sympathetic activity and decreased by the suppression of parasympathetic activity. Atropine abolished the correlation dimension and the low- and high-frequency components of HRPSA. With a postural change from supine to standing, the low- to high-frequency power ratio was significantly increased, indicating the augmentation of sympathetic activity or the reduction of parasympathetic activity, or both. Thus, it was suggested that a postural change from supine to standing causes the predominance of sympathetic activity. As this postural change decreased the correlation dimension, it was also indicated that the predominance of sympathetic activity may decrease the correlation dimension. CDA was superior to HRPSA in that the former method provided a measure of the effect of propranolol on the heart in the supine position, which could not be evaluated by the latter method. Thus, CDA of heart rate variability is a useful non-invasive tool for evaluating the autonomic balance.