Distributed gray-white matter structural covariance alterations and interindividual heterogeneity in adolescents with functional/dissociative seizures.

OBJECTIVE: To characterize distributed structural brain alterations and interindividual structural variability in adolescents with functional/dissociative seizures (FDS) and examine their relationship with dissociative symptom severity. METHODS: Seventy adolescents with FDS and 70 healthy controls underwent high-resolution structural magnetic resonance imaging. Gray- and white-matter features were integrated using multiset canonical correlation analysis followed by joint independent component analysis to identify latent gray-white matter structural covariance components. Surface-based morphometric measures were extracted from 68 cortical regions. Cross-validated Elastic Net and Random Forest models evaluated whether structural features predicted dissociative symptom severity. Normative reference-based deviation analysis quantified individualized departures from healthy structural organization. RESULTS: Eight multimodal structural components were identified, of which five demonstrated moderate-to-strong gray-white matter coupling (r = 0.52-0.70) and significant between-group differences involving sensorimotor projection, callosal-periventricular, posterior association, subcortical-cerebellar, and global deep white-matter systems. These differences remained significant after adjustment for anxiety and depression. Within the FDS group, multimodal component loadings showed no significant associations with dissociation, anxiety, or depression after adjustment for age (all p > 0.05). Although structural features differentiated adolescents with FDS from healthy controls, they did not reliably predict dissociative symptom severity. Normative reference-based deviation analysis demonstrated greater interindividual structural heterogeneity in FDS than in healthy controls. CONCLUSIONS: Adolescents with FDS exhibit distributed gray-white matter structural covariance alterations and marked interindividual heterogeneity rather than focal structural abnormalities. These findings support a distributed structural framework in which latent gray-white matter covariance patterns reflect vulnerability-related structural organization rather than current symptom severity.