Description
Autism spectrum disorder (ASD) is a heterogeneous neurodevelopmental disorder associated with repetitive behavior (RB), which may impair daily functioning at a lowerorder sensorimotor level (e.g. rocking back and forth) and a higher-order cognitive level (e.g. adherence to a routine). Evidence for brain abnormalities involved in RB is shown in volumetric studies indicating basal ganglia enlargement and functional studies indicating altered corticostriatal circuitry. Connectivity differences in the default mode network (DMN), involved in self-referential, baseline 'resting-state,' and the salience network (SN), involved in inhibitory functioning and directing attentional resources, may also have a role in RB symptoms. To investigate patterns of connectivity underlying RB in ASD, the current study examined two regions of interest (ROIs) in the striatum (dorsal caudate nucleus, dorsal caudal putamen), two in the SN (anterior insula, dorsal anterior cingulate cortex), and two in the DMN (ventral medial prefrontal and posterior cingulate cortex). Archival resting state functional magnetic resonance imaging (fMRI) data and repetitive behavior scores were obtained for overall 66 participants (35 ASD, 31 typically developing [TD]). Imaging data were motion-corrected and preprocessed using spatial smoothing, band-pass filtering and removal of nuisance regressors, including 6 rigid-body motion parameters and their derivatives, white matter and ventricular signals and their derivatives, and physiological time series. Average blood oxygen level dependent (BOLD) time series were extracted for each region of interest in each participant and correlated with all other voxels in the brain in a whole-brain analysis. The resulting Pearson's correlation values were standardized using Fisher's r-to-z transformation. One- and two-sample t-tests were then used to examine functional connectivity within and between groups. Results showed similar patterns of within-group functional connectivity; however, in between-group comparisons, mixed findings emerged with the ASD group showing mixed patterns of both over- and underconnectivity for each seed, except for the SN and dorsal caudate nucleus, for which only underconnectivity was detected. Clusters showing significant group differences in connectivity (mean z' converted from BOLD signal correlations) were then correlated with cumulative total RB scores, using the Repetitive Behavior Scale-Revised along with separate cognitive and motor RB subscales, and several areas in which underconnectivity was associated with higher RB scores were found. Finally, network connectivity was assessed by calculating the correlation between time series for each ROI pair in each participant, separately for each group. Only the putamen and caudate nucleus ROI pair emerged as significant, with overconnectivity in the ASD group. Analyses conducted within the ASD group showed increased connectivity of the putamen, primarily with the DMN, correlating with fewer RB symptoms. Furthermore, increased connectivity between the SN and DMN was associated with decreased symptom severity on the cognitive subscore of the Social Responsiveness Scale, a measure highly correlated with the RBS-R. Overall, results suggest that functional connectivity of the DMN, putamen, and - to a lesser extent - the salience network are associated with symptom severity and repetitive behaviors in ASD. Patterns of connectivity were largely decreased within networks and increased between networks, emphasizing the need for further examination of the segregation and interaction of separate functional networks in relation to characteristics of ASD.
