EEG Resting-State Beta Band Dynamic Functional Connectivity Reveals State-Dependent Disruption of Language Network Lateralisation in First-Episode Schizophrenia.
TL;DR
Findings demonstrate that language dysfunction in early psychosis is expressed as state-dependent disruption of brain network stability and language lateralisation, detectable only through dynamic connectivity analysis.
Abstract
Language disorganisation is a core feature of schizophrenia with significant functional consequences. However, its neurophysiological substrates remain poorly characterised, particularly in the early stages of the illness. This study investigated whether first-episode schizophrenia (FES) is associated with disruption in the organisation of language-network lateralisation, using both static and dynamic functional connectivity metrics in the EEG beta band. Thirty-one FES patients and 32 age-matched healthy volunteer subjects (HVS) underwent resting-state 64-channel EEG. Source-level connectivity was estimated using the weighted phase lag index with sliding windows, yielding 979 connectivity matrices per participant. Brain network states (BNS) were identified via k-means clustering. Static connectivity was obtained by time-averaging. Temporal features, graph-theoretical metrics, and language-network laterality indices were extracted for both whole-network and language-specific analyses. Group comparisons used permutation-based tests with family-wise error correction. Static connectivity yielded no significant group differences in any metric. Dynamic analysis identified six BNS. FES showed a significant increase in the variability of the temporal dynamics and graph-theory features of BNS5, a state characterised by parieto-occipital dominance. The core language network laterality index was significantly reduced in FES within BNS1, a state characterised by auditory-visual dominance. BNS1 occurrence correlated negatively with language disorganisation scores. Dynamic BNS-specific language network metrics correlated with verbal fluency performance exclusively in FES. These findings demonstrate that language dysfunction in early psychosis is expressed as state-dependent disruption of brain network stability and language lateralisation, detectable only through dynamic connectivity analysis.