Disrupted network segregation of the default mode network in autism spectrum disorder

BackgroundChanges in the brains default mode network (DMN) in the resting state are closely related to autism spectrum disorder (ASD). Module segmentation can effectively elucidate the neural mechanism of ASD and explore the intra- and internetwork connections by means of the participation coefficient (PC). MethodsResting-state fMRI data from 269 ASD patients and 340 healthy controls (HCs) were used in the current study. The PC of brain network modules was calculated and compared between ASD subjects and HCs. In addition, we further explored the features according to different age groups and different subtypes of ASD. Intra- and internetwork differences were further calculated to find the potential mechanism underlying the results. ResultsASD subjects showed significantly higher PC of the DMN than HC subjects. This difference was caused by lower intramodule connections within the DMN and higher internetwork connections between the DMN and networks. When the subjects were split into age groups, the results were verified in the 7-12 and 12-18 age groups but not in the adult group (18-25). When the subjects were divided according to different subtypes of ASD, the results were also observed in the classic autism and pervasive developmental disorder groups, but not in the Asperger disorder group. In addition, compared with the HC group, the ASD group showed significantly increased intranetwork connections between the DMN and the frontoparietal network. ConclusionsLess developed network segregation in the DMN could be a valid biomarker for ASD, and this feature was validated with different measures. The current results provide new insights into the neural underpinnings of ASD and provide targets for potential interventions using brain modulation and behavioural training.