Frequency-dependent brain state dynamic alterations in autism spectrum disorder: A co-activation pattern analysis

Autism spectrum disorder (ASD) is a complex neurodevelopmental disorder that affects normal brain development and results in impaired brain function. Most studies have focused on connectivity changes within the traditional low-frequency range, whereas the frequency-dependent nature of brain dynamics remains largely unexplored. This study employed whole-brain co-activation pattern (CAP) analysis to investigate the frequency-dependent spatiotemporal dynamics of spontaneous brain activity in ASD across three frequency bands: LFO (0.01-0.1 Hz), slow-5 (0.01-0.027 Hz), and slow-4 (0.027-0.073 Hz). Resting-state fMRI data were obtained from the NYU site of the ABIDE I database, comprising 52 individuals with ASD and 52 typical controls (TCs). Six CAPs were identified within each frequency band using k-means clustering. We then calculated and compared CAP dynamics, including the appearance frequency, duration, entry rate, and transition probability. Our results revealed that (1) CAPs across different frequency bands exhibited overall similar spatial patterns but showed significant differences in temporal evolution, with the slow-5 band demonstrating lower dynamic variability; (2) compared to the TC group, individuals with ASD exhibited abnormal brain dynamics in both the LFO and slow-4 bands, whereas no significant differences were observed in the slow-5 band; and (3) significant correlations were found between the dynamic metrics of CAPs in the LFO and slow-5 bands and the severity of restricted and repetitive behaviors (RRB) in individuals with ASD. These findings reveal frequency-specific abnormalities in brain dynamics in ASD, providing new insights into its time-varying neural mechanisms.