An array of variables have been assessed as potential early predictors of a
ntidepressant response in depressed patients. This exploratory study examin
ed the relationship of clinical outcome, following pharmacotherapeutic trea
tment, with quantitative electroencephalographic (EEG) features assessed pr
ior to treatment onset. In 70 major affective disorder patients, pre-treatm
ent spectrum-analysed topographic EEG indices (absolute power, relative pow
er, mean frequency, inter-hemispheric power asymmetry and coherence for 4 f
requency bands) were assessed in relation to baseline HAM-D ratings and HAM
-D rating changes following 6 weeks of open-label paroxetine treatment. EEG
slow wave (theta) activities were positively correlated with depression ra
tings prior to treatment. Of the patients (n = 51) completing treatment, 80
% evidenced a > 50 % reduction in HAM-D ratings. improved rating changes i
n general were found to be negatively related to slow (delta and theta) wav
e activity and positively related to fast (beta) activity at frontal record
ing sites. Findings are discussed in relation to the neurochemistry and neu
robiology of depressive disorders.