Disentangling the topological symptom structure of schizophrenia: A network analysis.

Abstract

Background and hypothesis: The Positive and Negative Syndrome Scale (PANSS), comprehensively assesses schizophrenia severity. While network analyses of schizophrenic symptoms have yielded inconsistent results, components of disorganized thought consistently rank high in centrality. The present study aims to explore the centrality of disorganized thought across patient subgroups and its potential as a treatment target. We hypothesize that disorganized thought will emerge as a central feature in the symptom network across different patient populations.

Study design: We conducted a network psychometric analysis on data from 1435 schizophrenia patients, stratified into four groups based on family history and sex. Local and global network properties, including centrality, clustering coefficient, degree, density, and community detection, were investigated. Network comparisons were performed across groups, and results were validated using an independent dataset.

Study results: Disorganized thought emerged as the most central factor in Marder 5-factor model, maintaining stability across family history and sex differences. While family history did not significantly impact symptom structures (Females: M = 0.2, P = 0.4; S = 0.4, P = 0.7; Males: M = 0.2, P = 0.7; S = 0.1, P = 0.9), significant differences were observed between male and female symptom structures (Positive family history: M = 0.3, P < 0.05; Negative family history: M = 0.3, P < 0.01). The centrality and high stability of disorganized thought were further confirmed in the validation dataset.

Conclusions: The consistent centrality of disorganized thought across different patient subgroups suggests its potential as a key treatment target for schizophrenia.