Structure-function decoupling of the default mode network in autism spectrum disorder.
Abstract
Autism spectrum disorder (ASD) is widely associated with abnormalities in large-scale brain connectivity. However, most studies have examined structural and functional connectivity independently, leaving unclear whether functional disturbances in ASD reflect a disruption of the structural architecture that normally constrains functional communication in the human brain. To address this question, we quantified structure-function coupling (SFC) using multimodal MRI data from the Autism Brain Imaging Data Exchange II dataset. This framework measures the correspondence between anatomical connectivity and functional interactions across the brain. We further evaluated the diagnostic and clinical relevance of SFC alterations and integrated cortical gene-expression data to explore potential molecular mechanisms. Our results showed that individuals with ASD exhibited a tendency toward lower global SFC, indicating a widespread weakening of the relationship between structural and functional brain organization. These effects were particularly pronounced within the default mode network (DMN), with selective decoupling within and between the DMN-A and DMN-B subsystems. Importantly, SFC features reliably distinguished individuals with ASD from controls and significantly captured cross-diagnostic variation in social responsiveness, demonstrating clinical relevance. Transcriptomic analyses further revealed that regions showing prominent SFC alterations were associated with gene sets enriched for neurodevelopmental and immune-related biological processes. These findings indicate that ASD is characterized by a breakdown of the normal correspondence between structural architecture and functional brain dynamics, particularly within default mode network circuitry. Disrupted SFC may therefore represent a systems-level mechanism linking atypical brain organization to social-cognitive impairment in ASD.