Resting-state electroencephalography microstates in antipsychotic-naïve individuals across the psychosis spectrum

Aberrant microstate features of resting-state electroencephalography (rsEEG) have been proposed as potential endophenotypic markers for schizophrenia. However, longitudinal investigations across the psychosis continuum remain limited, particularly regarding treatment effects and illness-related alterations over time. We examined 47 antipsychotic-naive, first-episode patients with psychosis (FEP), 32 individuals at ultra-high risk for psychosis (UHR), and 94 matched healthy controls (HC). All participants under-went rsEEG at baseline, FEP patients and HC were reassessed after six weeks and two years, during which the FEP-patients received antipsychotic treatment. We analyzed microstate features: Duration, Occurrence, Coverage, and Sample Entropy, and investigated group differences at baseline, as well as within group and between group effects over time. Additionally, we explored the effect of medication, and associations with symptom level using a linear model. Post hoc correlation analyses were performed to investigate the stability of microstate features at an individual level over time. At baseline, no differences were observed among HC, UHR, and FEP groups. The main linear mixed models including HC and FEP, as well as all microstates A, B, C, and D at the 3 timepoints indicated an overall effect of time between baseline and two years for Occurrence (p=0.044), and Coverage (p=0.036), but not for Duration (p=0.986). Post-hoc tests for each microstate showed a significant effect of time within FEP between baseline and two years (Occurrence p=0.047, Duration C p=0.011, and Coverage C p=0.007). Furthermore, a group effect emerged between HC and FEP at two years for Coverage C (p=0.039), p-values uncorrected. No other effects of time or group were observed. Occurrence of Microstate D was negatively correlated with general symptom level at baseline (p=0.008, corrected), but not at any follow-up. No associations were found between microstates and antipsychotic medication at six weeks. These findings indicate that Microstate C increases in antipsychotic naive patients the first two years after first episode psychosis, contrasted with the temporal stability in controls. This highlights the need for further research disentangling longitudinal effects of pharmacological and pathophysiological modulation of EEG microstates in large samples.