Mirror Neuron System Response to Action Simulation in Children with Autism

Background:  While deficits in the ability to imitate have been reported widely in individuals with autism spectrum disorders (ASD) (Williams et al., 2004), not all forms of imitation are equally impaired (Hobson & Hobson, 2008). At the neural level, dysfunction of the mirror neuron system (MNS) has been proposed to explain imitation deficits in autism (e.g., Williams et al., 2001; Dapretto et al., 2006). However, several recent studies have criticized this view finding evidence for robust activation and modulation of activation in MNS in ASD individuals (Dinstein et al., 2010).  Previous research indicates that action simulation may provide an important resource for developing a better understanding of MNS functioning in ASD by allowing analysis of the functioning of this system independent of motor production (Wadsworth & Kana, 2011).

Objectives:  To examine the role of MNS in mediating action simulation (mental imitation of an action) in children and adolescents with ASD.  

Methods:  Fourteen high-functioning children and adolescents with ASD (8-17 years old) and 15 age and IQ-matched typically developing (TD) controls participated in this functional MRI study.  Participants watched 20 cartoon images, developed by Heilman and colleagues (Mozaz et al., 2006), of people performing everyday actions (10 transitive actions, 10 intransitive actions) but with the hands missing.  Participants were asked to identify which hand orientation (of three answer choices presented) would best fill in the gap.

Results:  The main results of this study are: 1) Action simulation elicited robust activity in the MNS in ASD and TD participants; 2) Significantly reduced activity in ASD participants, relative to TD, while simulating transitive actions in right inferior parietal lobule (RIPL) and increased activity in right middle temporal cortex for intransitive actions (p<0.005, k = 80 contiguous 2mm³ voxels); 3) Significantly increased percent signal change (psc) was found in ASD participants in the left inferior frontal gyrus (LIFG) and regions related to motor planning and control (right precentral and postcentral gyrus, right cerebellum) during action simulation (p<0.05); 4) PSC in the LIFG was negatively correlated with scores on a measure of visual-motor integration (r = -0.53, p<0.05); and 5) No significant group differences on performance accuracy, but longer reaction time in ASD during action simulation.

Conclusions:  Overall, participants with ASD showed intact performance in action simulation, and we did not find any evidence for a global MNS dysfunction in these participants. However, when transitive actions were simulated, ASD children showed reduced RIPL activity. This may be associated with limited processing of object affordances necessary for designing an appropriate motor plan (Maranesi et al., 2014) and also may be associated with less mirror neuron activity (Williams et al., 2006). In addition, ASD participants showed increased activation of the LIFG, potentially related to increased task difficulty (Martineau et al., 2010).  Thus, it is possible that altered MNS response in ASD may be manifested only in complex action simulation.  More research, manipulating task demand and contexts, is needed to generate a consensus on the role of MNS in ASD symptomatology.