Altered spatial and temporal autocorrelations in autism spectrum disorder reflect dysregulated brain network topology

Autism spectrum disorder (ASD) is characterized by atypical functional brain network topology, yet how spatiotemporal properties of neural activity contribute to these alterations remains poorly understood. This study investigated the spatial autocorrelation (SA) and temporal autocorrelation (TA) patterns in resting-state functional magnetic resonance imaging data obtained from the Autism Brain Imaging Data Exchange database, including 349 individuals with ASD and 459 demographically matched typical controls (TC), with particular emphasis on sex-specific differences in these spatiotemporal properties. A spatiotemporal surrogate time series model parameterized by SA and TA was further constructed to assess the potential contribution of altered SA and TA to the aberrant brain network topology in ASD. Significant sex-related heterogeneity was observed in SA and TA at the whole brain, network and regional levels in ASD, especially in the default mode network. Furthermore, a high similarity in altered network topology in ASD was found between empirically observed functional brain networks and model-generated networks. In addition, altered network-level spatiotemporal autocorrelations can predict sex-divergent symptom profiles in ASD. These findings suggest that altered SA and TA may underlie the aberrant network topology observed in ASD, providing a parsimonious yet powerful framework for understanding the complex network heterogeneity in ASD.