Dynamic Meta-Networking Identifies Distinct Network Correlates of Positive and Negative Formal Thought Disorder in Schizophrenia

Formal thought disorder (FTD) is a core symptom of schizophrenia, yet the neural network mechanisms underlying this phenotype remain poorly understood. In this study, we applied a dynamic meta-networking framework, which captures temporally recurring functional network states, to investigate alterations in the language and executive control networks and their associations with positive and negative FTD. Resting-state fMRI data were collected from three independent cohorts: a discovery cohort comprising 150 first-episode, drug-naive patients with schizophrenia and 175 healthy controls (HCs); a replication cohort including 183 first-episode, drug-naive patients and 109 HCs; and a third cohort consisting of 71 patients who had received two weeks of antipsychotic treatment and 71 HCs. Meta-networking analysis identified four distinct resting-state meta-states within both the language and executive control networks. Connectivity-behavior correlation analyses and machine-learning regression models revealed that positive FTD was associated with aberrant connectivity across specific meta-states in both networks. In contrast, negative FTD was linked exclusively to dysfunction within two meta-states of the executive control network. Notably, these polarity-specific, multi-state connectivity disruptions normalized following short-term antipsychotic treatment, highlighting their potential as clinically relevant neuroimaging biomarkers.