Multidimensional reorganization of static and dynamic brain network architecture following rTMS in young children with autism spectrum disorder and co-occurring intellectual disability.

AIMS: Repetitive transcranial magnetic stimulation (rTMS) has emerged as a promising intervention for autism spectrum disorder (ASD). However, the neurophysiological mechanisms driving its therapeutic effects, particularly regarding the complex reorganization of brain network dynamics, remain largely uncharacterized. This study aimed to elucidate the multidimensional network reconfiguration induced by low-frequency rTMS and identify specific electrophysiological predictors of clinical response in young children with ASD.

METHODS

In a randomized, sham-controlled trial, 44 young children with ASD were assigned to receive either active (1 Hz, left dorsolateral prefrontal cortex) or sham rTMS for 20 sessions. Resting-state EEG data were acquired at baseline and post-intervention. We employed a comprehensive analytical framework integrating static Low- and High-Order Functional Connectivity (LOFC/HOFC), graph theoretical topology, dynamic Network State Entropy (NSE), and between-frequency coupling to assess network reconfiguration. Clinical efficacy was evaluated using the Autism Behavior Checklist (ABC).

RESULTS

Active rTMS yielded significant improvements in sensory behavior and total behavioral scores compared to sham stimulation. Neurophysiologically, rTMS induced a "restorative" reorganization characterized by: (1) the suppression of pathological hyper-connectivity in the delta band and the enhancement of long-range synchronization in the alpha band; (2) the optimization of network topology, evidenced by increased global and local efficiency in the alpha band; (3) the modulation of dynamic flexibility, manifested as stabilized low-frequency dynamics and enhanced high-frequency metastability; and (4) the decoupling of aberrant cross-frequency interactions, specifically reducing pathological δ-β coupling. Crucially, the increase in delta network segregation and the reduction in pathological delta-beta coupling were significantly correlated with the magnitude of clinical symptom alleviation.

CONCLUSIONS

Low frequency rTMS promotes a systemic reorganization of brain networks in children with ASD, shifting the neural architecture from a disorganized, hyper-connected state towards an optimized, hierarchically structured system. Despite the limited spatial resolution of portable EEG, the rigorous multidimensional analysis revealed robust temporal dynamic changes. The direct link between network segregation, cross-frequency decoupling, and sensory-behavioral gains highlights multidimensional EEG metrics as potential biomarkers for monitoring therapeutic responses in early ASD intervention.