PREDICTION OF THE RESPONSE TO TOTAL SLEEP -DEPRIVATION (TSD) IN ENDOGENOUS DEPRESSED-PATIENTS BY QUANTITATIVE EEG PARAMETER

The aim of the present study was to examine specific EEG parameters wi th regard to their qualification for predictive statements in depresse d subjects concerning their response to sleep deprivation and to analy se the differences between responders and non-responders before sleep deprivation. 26 patients (ICD 9:296.1:20 patients; 296,3:6 patients) w ere included in this study. Responders to sleep-deprivation were defin ed by an improvement of 30% in the daily mean of the Hamilton-Depressi on-Rating-Scale. EEG data of all patients were recorded the day before sleep deprivation and day 1 and day 2 after sleep deprivation. The ev aluation of the EEG data was performed by Fourier-Analysis, in which t he following EEC parameters were included: centroid, power, frequency deviation and coherence. On the day before sleep deprivation significa nt (p < 0.05) differences between responders and non-responders were f ound for the following EEG parameters: centroid alpha, total frequency deviation and coherence. Especially in the alpha spectrum the degree of response (relative change of the mean Hamilton day score) to sleep deprivation was significantly negative correlated with the centroid fr equency before sleep deprivation in the frontal area. In relation to t he frequency deviation in the whole spectrum there could be stated als o a significantly negative correlation to response before sleep depriv ation in the frontal areas. Regarding the EEG parameter coherence it w as possible to detect a significantly positive correlation at frontal sites to the degree of response to sleep deprivation in the alpha and beta band. These results show, especially in combination, a prospectiv ly strong instrument of prediction for sleep deprivation response in d epressive illness.