Atypical Ventral Premotor Cortex Activity during Motor Imitation in Children and Adolescents with Autism

Background: Deficits in imitation have been widely reported in children and adolescents with autism spectrum disorders (ASD) (Rogers & Pennington, 1991; Williams et al., 2001). Recent evidence from neuroimaging research on imitation in autism points to the dysfunction of mirror neurons (Williams, 2008; Bernier and Dawson, 2009), although some studies report normal movement selectivity with no mirror neuron dysfunction in autism (Dinstein et al., 2010). While activity of the mirror neuron system (MNS) has been studied relatively more, the connectivity among the different nodes of MNS has been less explored (Kana, Wadsworth, & Travers, 2011). Thus, this fMRI study examined the MNS responses underlying motor imitation in children with autism.

Objectives: The primary goal of this project was to examine the nature and extent of MNS response in high-functioning children and adolescents with autism during motor imitation.

Methods: Participants included 12 high-functioning children and adolescents with autism (HFA) and 15 age-and-IQ-matched typically developing children (TD) (right-handed, 8-17 years of age).  Participants took part in 2 separate sessions: 1) Assessment; and 2) MRI scanning. During the first session, the participants completed a series of neuropsychological tests, such as attention, and IQ. In the fMRI scanner, participants were asked to imitate a series of 20 high-resolution snapshots of hand actions performed by a model. These images were displayed at the center of the screen, one at a time for a period of 4000ms with an inter-stimulus interval of 6000ms in an event-related design.  Participants were instructed to use their right hand to imitate the same hand action displayed in the picture.  The fMRI data collected from a Siemens 3 tesla scanner were analyzed using SPM8.   

Results: Both participants with an ASD and TD controls showed activation in the core regions of the MNS (posterior parietal, and ventral premotor cortex) while performing the motor imitation task. Between-group analysis of brain activation revealed significantly less activity in participants with autism, relative to TD controls, in the ventral premotor aspect of the MNS, but not in the posterior parietal area (p=0.001, 10 voxel extent threshold).  In the assessment scores, we found participants with autism performing worse than TD controls in the Beery Visual-Motor Integration (VMI) task [t(25)=2.06; p = 0.05]. In addition, the nonverbal cognitive scores (as measured by the WASI-2), which was higher for the ASD group, did not have a significant impact on brain activity as evident from a multiple regression analysis.     

Conclusions: The results from this study are consistent with recent findings suggesting that the MNS as a whole system may not impaired in individuals with ASD (Cook & Bird, 2012).  As such, these results suggest further examination of the specific role of each node of this system in the imitation deficits found in individuals with ASD. This may provide specific avenues to target for intervention.  In these lines, future analyses for this study include an examination of the functional and effective connectivity, and examining brain-behavior relationships.