Altered dynamic functional connectivity states and cognitive dysfunctions in first-episode schizophrenia: evidence from leading eigenvector dynamics analysis

Abnormal temporal coordination of large-scale brain networks may contribute to cognitive impairment in schizophrenia. However, the dynamic characteristics of functional connectivity in first-episode schizophrenia (FES) remain unclear. Resting-state fMRI data from FES patients and healthy controls were analyzed using Leading Eigenvector Dynamics Analysis (LEiDA) to identify recurrent dynamic functional connectivity (dFC) states. Group differences in occurrence, dwell time, transition probability, and metastability were compared. Cognitive performance was assessed with the MATRICS Consensus Cognitive Battery, and correlations between altered dFC metrics and cognitive domains were examined. Four dFC states were identified. Compared with controls, FES patients showed reduced occurrence and transition probability of a distributed control–default mode–attention state (Mode 3) and increased metastability in global and cognitive-control-related states (Modes 1 and 2). Transition frequency from the global to perceptual-attentional state was elevated. The occurrence of the integration state was positively correlated with processing speed, verbal learning, and visual learning, whereas transition frequency was negatively correlated with visual learning performance. FES patients exhibit alterations in the temporal organization of large-scale networks. Dynamic connectivity abnormalities, particularly in frontoparietal and default mode systems, were associated with cognitive dysfunction. These findings support further investigation of dynamic functional connectivity measures for early schizophrenia. ChiCTR2000041106 (Date of Registration 20201217). Illustration of the neural mechanisms summarizing associations between dFC abnormalities and cognitive dysfunction in patients with FES