Quantitative spectral analysis of EEG in psychiatry revisited: drawing signs out of numbers in a clinical setting.

Abstract

To evaluate the incidence, sensitivity and specificity of abnormal quantitative EEG (QEEG) measures in normal subjects and patients with mental disorders. Normalized QEEG measures were blindly assessed in 67 normal human beings and 340 psychiatric patients. QEEG results were correlated to subject condition or diagnosis and magnetic resonance imaging (MRI) findings. QEEG was abnormal in 83% of patients, and 12% of normal subjects. The most frequent abnormality was a decrease in slow (delta and/or theta) bands, either alone, with beta increase, or with alpha decrease, followed by increase in beta band. No normal subject showed delta and/or theta decrease. Slow band decrease was more frequent in depression and mental disorders due to general medical condition, alcohol and drug dependence. However, no pattern was specific of any entity, and patients within the same diagnostic may present different patterns. Delta-theta decrease was correlated with cortical atrophy as seen in MRI. Beta increase was correlated with psychoactive medication. No association was found between any other QEEG pattern and MRI abnormalities, or medication. Decrease in the delta and theta bands of the QEEG can be regarded as a specific sign of brain dysfunction, and is correlated with cortical atrophy. However, this sign, as other QEEG abnormal patterns, can be found in many different disorders and none of them can be considered as pathognomonic of any specific disorder. This work attempted to circumvent the alleged lack of Class I evidence of QEEG utility in the study of psychiatric patients by means of a prospective, blinded study, searching for specific signs of physiopathology in individual patients.