Schizophrenia and other psychotic disorders exhibit significant clinical and cognitive heterogeneity, challenging diagnosis and treatment. Given indications of widespread cortical hyperactivity and dysregulation of neural oscillations in schizophrenia, investigating resting-state activity is highly relevant. This study examined resting-state EEG alterations across previously defined cognitive subtypes within the psychosis spectrum. We analyzed resting-state EEG data from 141 psychosis patients (64 chronic schizophrenia, 40 first-episode schizophrenia, 37 bipolar disorder) and 80 healthy controls. Patients were a priori classified into two distinct cognitive subgroups: Cluster 1 (severe impairment, n = 47) and Cluster 2 (moderate impairment, n = 94). Both patient clusters exhibited increased spectral power across most frequency bands compared to healthy controls. Notably, the more severely impaired Cluster 1 showed significantly higher power in the gamma-1 band (30–45 Hz) compared to Cluster 2. Furthermore, in Cluster 1, a significant positive correlation was found between resting-state gamma-1 power and positive symptom scores. These results support our hypothesis of distinctive basal hyperactivation linked to the cognitive profile, suggesting that altered intrinsic brain activity, particularly gamma-1 hyperactivation, may underlie cognitive heterogeneity in psychosis. This also suggests that gamma-1 band hyperactivation at rest serves as a distinct neurophysiological marker differentiating both subgroups. Our findings highlight the importance of subdivision approaches to identify more homogeneous patient subgroups and emphasize the potential of resting-state gamma activity as a precise biomarker for specific symptom dimensions and personalized treatment strategies. • Resting-state gamma-1 power distinguishes psychosis subtypes. • Severe cognitive impairment links to higher gamma-1 activity. • Gamma-1 hyperactivation correlates with positive symptoms. • Highlights potential for precise biomarkers in psychosis.
Osorio-Iriarte et al. (Tue,) studied this question.