Electroencephalography microstates imbalance across the spectrum of early psychosis, autism, and mood disorders

Abstract

Background: Electroencephalography (EEG) microstates translate resting-state temporal dynamics of neuronal networks throughout the brain and could constitute possible markers of psychiatric disorders. We tested the hypothesis of an increased imbalance between a predominant self-referential mode (microstate C) and a decreased attentional mode (microstate D) in psychosis, mood, and autism spectrum disorders.

Methods: We retrospectively included 135 subjects from an early psychosis outpatient unit, with available eyes-closed resting-state 19 electrodes EEG. Individual-level then group-level modified K-means clustering in controls provided four microstate maps that were then backfitted to all groups. Differences between microstate parameters (occurrence, coverage, and mean duration) were computed between controls and each group, and between disease groups.

Results: Microstate class D parameters were systematically decreased in disease groups compared with controls, with an effect size increasing along the psychosis spectrum, but also in autism. There was no difference in class C. C/D ratios of mean duration were increased only in SCZ compared with controls.

Conclusions: The decrease in microstate class D may be a marker of stage of psychosis, but it is not specific to it and may rather reflect a shared dimension along the schizophrenia-autism spectrum. C/D microstate imbalance may be more specific to schizophrenia.

Keywords: EEG microstates; Electroencephalography; autism spectrul disorder; mood disorders; psychosis; transdiagnostic approaches.

Figure 1.

Mean spectral power densities for each group after preprocessing. ASD, autism spectrum disorder; BIP, bipolar disorder; CTRL, controls; FEP, first-episode psychosis; MDD, major depressive disorder; SCZ, schizophrenia; UHR, ultra-high-risk.

Figure 2.

Spatial correlations between the four clusters of each disease group and the controls.

Figure 3.

Posterior estimation of the effect-size of the difference in microstate D parameters (occurrence, coverage, mean duration) between each disease group and the controls. The red dotted interval represents the region of practical equivalence. HDI: 90% highest density interval of the posterior estimation.