ABSTRACT Background The prevalence of autism spectrum disorder (ASD) is significantly higher in males than in females; although the underlying etiology remains unclear. This study aimed to investigate the multi‐scale reorganization of brain networks in preschool‐aged boys with ASD and their impact on clinical symptoms. Methods A total of 54 children with ASD (40 boys and 14 girls) and 44 typically developing (TD) children (28 boys and 16 girls), aged between 2 and 6 years, were recruited for this study. Functional gradient analysis and dynamic functional connectivity were used to examine differences in the hierarchical organization of brain functional networks between preschool‐aged boys and girls with ASD compared to their corresponding typically developing peers. Subsequently, multiple machine learning models were applied to evaluate the classification performance of the identified abnormal features in distinguishing ASD from TD. Results The results showed that the ASD group exhibited functional gradient abnormalities in multiple brain networks: (1) In boys with ASD, primary gradient abnormalities were identified in the dorsal attention network, limbic network, ventral attention network, and default mode network (DMN), whereas in girls with ASD, primary gradient abnormalities were only found in the DMN; (2) Secondary gradient abnormalities in boys with ASD were found in the sensorimotor network (SMN), ventral attention network, and DMN, while in girls with ASD, secondary gradient abnormalities were restricted to the DMN and SMN; (3) Third gradient abnormalities in boys with ASD were observed only in the visual network, whereas in girls with ASD, abnormalities were present in the limbic network, SMN, and visual network; (4) Enhanced dynamic functional connectivity was detected in boys with ASD only in state 1. Conclusion Preschool‐aged boys and girls with ASD exhibit significant sex differences in functional gradients and dynamic functional connectivity, underscoring the complexity and heterogeneity of ASD. These findings provide a novel theoretical framework for understanding the neuroimaging mechanisms underlying ASD.
Wu et al. (Fri,) studied this question.