REM sleep EEG spectral analysis in patients with first-episode schizophrenia

J Psychiatr Res. 2008 Oct;42(13):1086-93. doi: 10.1016/j.jpsychires.2008.01.003. Epub 2008 Feb 15.

Abstract

The pathophysiology of schizophrenia includes abnormalities in subcortical-cortical transfer of information that can be studied using REM sleep EEG spectral analysis, a measure that reflects spontaneous and endogenous thalamocortical activity. We recorded 10 patients with first-episode schizophrenia and 30 healthy controls for two consecutive nights in a sleep laboratory, using a 10-electrode EEG montage. Sixty seconds of REM sleep EEG without artifact were analyzed using FFT spectral analysis. Absolute and relative spectral amplitudes of five frequency bands (delta, theta, alpha, beta1 and beta2) were extracted and compared between the two groups. Frequency bands with significant differences were correlated with BPRS positive and negative symptoms scores. Patients with schizophrenia showed lower relative alpha and higher relative beta2 spectral amplitudes compared to healthy controls over the averaged total scalp. Analysis using cortical regions showed lower relative alpha over frontal, central and temporal regions and higher relative beta2 over the occipital region. Absolute spectral amplitude was not different between groups for any given EEG band. However, absolute alpha activity correlated negatively with BPRS positive symptoms scores and correlated positively with negative symptoms scores. Since similar results have been reported following EEG spectral analysis during the waking state, we conclude that abnormalities of subcortical-cortical transfer of information in schizophrenia could be generated by mechanisms common to REM sleep and waking.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Electroencephalography* / methods
  • Female
  • Humans
  • Male
  • Middle Aged
  • Schizophrenia / physiopathology*
  • Sleep, REM / physiology*
  • Spectrum Analysis*
  • Statistics, Nonparametric