Cerebro-Cerebellar Resting State Functional Connectivity for Motor and Prefrontal Networks in Adolescents with Autism Spectrum Disorders

Background: The cerebellum is considered to play an important role in motor, somatosensory, and executive functioning. A number of studies have reported abnormalities in neuronal integrity, volume, metabolism, and activity in the cerebellum in autism spectrum disorders (ASD). A previous functional connectivity magnetic resonance imaging (fcMRI) study reported cerebro-cerebellar underconnectivity during motor performance in ASD. Resting-state (RS) fcMRI identifies spontaneous low-frequency blood oxygenation level dependent (BOLD) signal fluctuations allowing for the examination of intrinsic network connectivity. Impairments of motor and executive functions have been reported in numerous studies of ASD. Such impairments may be associated with abnormal cerebellar connectivity with prefrontal and motor cortices. However, a focused investigation of intrinsic cerebro-cerebellar connectivity for motor and prefrontal networks in ASD is currently unavailable.

Objectives: To examine cerebro-cerebellar intrinsic functional connectivity in motor and prefrontal networks in adolescents with ASD.

Methods: RS functional MRI data were acquired for 6:10 minutes on a 3T GE scanner for 18 adolescents with ASD and 22 age, sex, IQ, and motion-matched typically developing (TD) adolescents. Participants were required to remain still and fixate a crosshair in the center of a projection screen. Data were preprocessed using AFNI and included motion and field map correction, cardiac and respiratory regression, spatial smoothing, isolation of low frequency fluctuations (.01<f<.1), and normalization to Talairach space. The Talairach-Tournoux Daemon atlas in AFNI was used to obtain masks for the cerebellum as well as the motor and prefrontal cortices. Partial correlation analyses were performed between each voxel in the cerebellar mask and the time series extracted from motor and prefrontal regions of interest (ROIs), respectively. Pearson’s correlation analyses were conducted for fcMRI effects (motor, prefrontal) with diagnostic (ADOS, ADI), Social Responsiveness Scale (SRS), and Visuo-Motor Integration (VMI) scores.

Results: Connectivity results in TD adolescents were consistent with previous findings by O’Reilly and colleagues (2010), with motor connectivity in anterior cerebellum and prefrontal connectivity in posteriolateral neocerebellum. Between-group comparisons revealed atypically increased cerebellar connectivity for the motor ROI, but reduced connectivity for the prefrontal ROI in the ASD group. Across both groups, significant negative correlations for motor-cerebellar connectivity and VMI scores, r = -.761, p < .004, and prefrontal-cerebellar connectivity and SRS scores r = -.377, p < .05. Increased motor-cerebellar connectivity correlated with reduced VMI scores in the TD group (r = -.905, p < .005), whereas increased motor-cerebellar connectivity correlated with reduced ADOS-social subtest scores (r = -.576, p < .05) in the ASD group.  

Conclusions: Our findings suggest reduced intrinsic functional connectivity between prefrontal cortex and cerebellum in ASD, accompanied by increased connectivity with cortical motor regions. Our latter finding for intrinsic connectivity contrasts with a previous report of cerebro-cerebellar underconnectivity during motor performance (Mostofsky et al., 2009). The pattern of group differences may indicate an expansion of motor-related cerebro-cerebellar connectivity at the expense of connectivity subserving executive functions, consistent with reports of executive impairment in ASD. Correlations of functional connectivity measures with social responsiveness and visuo- motor abilities support the behavioral relevance of cerebro-cerebellar connectivity.