The description of structural and temporal characteristics of tonic electrophysiological activity during sleep.

Abstract

The structural and temporal characteristics of tonic variation among electrophysiological measures during sleep are examined from the point of view of sleep stage scoring procedures and the digital quantification and statistical analysis of these measures. Sleep stage scoring procedures are nominally scaled and therefore less able to index structural and temporal characteristics than are measures derived from digital methods of quantification having the properties of interval or ratio scales. These arguments are illustrated using principal components analysis to describe the structure of tonic covariation among 36 variables derived from the computer analysis of EEG, EMG, and EOG. Three components resulting from the analysis are tentatively named 'slow-fast', 'hemispheric-shift' and 'eyes-activation'. They appear across variations in experimental manipulations such as night time awakenings and also across differences in age and species. The periodic characteristics of the vectors associated with the three components are specified by means of the Fast Fourier transform in combination with digital low pass and band pass filtering. Cycles are found which are both slower and faster than the paradigmatic 90 min ultradian rhythm. These results indicate that currently available techniques of digital and statistical analysis provide new possibilities for research on the electrophysiology and psychophysiology of sleep.