Up until now, no subclassification of affective psychoses has been validated biologically. This follows unavoidably from a research practice of defining diagnostic subtypes in consensus conferences and only thereafter allowing their validation. There is evidence that electroencephalograms (EEG) may be a useful tool in psychiatry, provided that the relevant information is extracted.

Our EEG quantification procedure aims at an assessment of both the amount and range of variation of spontaneous changes of topographical alpha-power distribution, developing within a certain period of recording under resting conditions. Our measures were designed to characterize the dynamic organization of the EEG. This is quite obviously an eyeball evaluation but it has nevertheless been neglected in research. The study design was done retrospectively. Included were inpatients with a primary depressive disorder. Main exclusion criteria were an age older than 62 years and psychotropic drugs other than antidepressants. The psychopathology and other clinical data were routinely assessed within three days after admission by the AMDP documentation. An EEG was also routinely performed at admission. We made use of robust, generally known non-parametric statistics.

Those patients who exhibited a dynamically rigid EEG are especially prone to recurrences, have a relative late onset of their illness, and show an acute symptomatology characterized by organic-like features.

The findings lend support to our contention that the quantitative assessment of the dynamics of the EEG-Gestalt allows the delimitation of a clinically important subtype that is characterized both cross-sectionally and in long-term respects.