Longitudinal white matter network reorganization after rTMS in first-episode schizophrenia: associations with sleep improvement

Abstract

Sleep disturbances are highly prevalent in first-episode schizophrenia and are associated with poor clinical and functional outcomes. Although repetitive transcranial magnetic stimulation (rTMS) has shown therapeutic potential for insomnia and schizophrenia, its efficacy for sleep disturbances in first-episode schizophrenia remains insufficiently characterized, and the underlying white matter network mechanisms remain unclear. This study aimed to determine whether active rTMS improves sleep quality compared with sham stimulation and whether longitudinal white matter network reorganization is associated with treatment-related sleep improvement. One hundred patients with first-episode schizophrenia were included in the final analysis and randomly assigned to receive active rTMS or sham stimulation over an 8-week intervention period. Subjective sleep quality was assessed using the Pittsburgh Sleep Quality Index (PSQI), and objective sleep parameters were measured using wrist actigraphy. Diffusion tensor imaging was acquired at baseline and after treatment to construct individual white matter structural networks. Graph-theoretical metrics, nodal efficiency, and rich-club organization were examined. Linear mixed-effects models were used to evaluate group-by-time interaction effects, and correlation analyses were performed to examine associations between network changes and sleep improvement. Compared with sham stimulation, active rTMS produced greater improvements in subjective and objective sleep outcomes, including reduced PSQI scores, increased sleep efficiency, and decreased wake after sleep onset and sleep latency. Active rTMS was also associated with greater longitudinal increases in global efficiency, local efficiency, clustering coefficient, and nodal efficiency in prefrontal, default-mode, temporal, and salience-related regions. Rich-club connectivity increased more strongly after active rTMS than after sham stimulation, whereas feeder connectivity, normalized rich-club coefficient, and local connectivity did not show significant group-by-time interaction effects after FDR correction. In the active rTMS group, increases in rich-club connectivity were positively associated with improvements in actigraphy-derived sleep efficiency ( r = 0.43, P = 0.002). Active rTMS improved sleep quality more effectively than sham stimulation in patients with first-episode schizophrenia and was accompanied by longitudinal changes in DTI-derived white matter network organization. The association between increased rich-club connectivity and improved sleep efficiency suggests that hub-related network reorganization may represent a network-level correlate of rTMS-related sleep improvement.