Vegetative background of sleep: Spectral analysis of the heart rate variability

Abstract

Phasic events during sleep, either arousal or REM-burst related, and the associated transient cardiovascular responses have been the subject of intensive research in previous studies. However, nontransient (stationary) fluctuations in heart rate have been studied less extensively in the past. They allow a differentiation of the sympathetic and parasympathetic activation, which are related to a low-frequency (LF) and a high-frequency (HF) component of the heart rate variability (HRV) signal, respectively. The resulting LF/HF ratio is a quantitative index of the sympatho-vagal balance. Sleep polygrams from 20 healthy volunteers were recorded in a sleep laboratory. Standard vegetative tests (orthostatic and Valsalva tests) were evaluated. A segmentation procedure, performed on the HRV signal, separated 70-130 transients during the night from the records of stationary heart rate fluctuations. From these periods the sympatho-vagal balance, quantified by the LF/HF, was computed by means of spectral analysis. The more synchronized the sleep was, the more the LF/HF decreased, whereas the LF/HF was significantly increased during REM sleep,indicating a sympathetic predominance during this period. Such an increase was also evident during the last 15 min before REM sleep onset. Results suggest that spectral analysis of the HRV provides additional information of the ultradian rhythmic behavior of the autonomic nervous system function beyond the traditional cardiovascular measurements (mean heart rate, blood pressure, etc.). In contrast to these measurements, which generally show a continuously decreasing cardiovascular activity, as the night proceeds, the results of this study reveal a high sympathetic peak activity during the later REM sleep periods, which is comparable in magnitude to that found in the upright position in wakefulness. This activity may be associated to the well-known incidence peak of ischemic events in the early morning hours. (C) 1997 Elsevier Science Inc.