Preserved Mimicry in Children with Autism; Enhanced Mimicry in Children with Williams Syndrome

Background: Upon seeing an emotional facial expression typically developing individuals rapidly mimic the expression.  These responses typically occur within the first second after seeing the expression (Moody, McIntosh, Mann, & Weisser, 2007). Deficits in rapid mimicry have been found in adults and children with Autism Spectrum Disorder (ASD) (Beall, Moody, McIntosh, Hepburn, & Reed, 2008; McIntosh, Reichmann-Decker, Winkielman, & Wilbarger, 2006), and are theorized to influence social functioning (Moody & McIntosh, 2006).  Recent research has found that children with ASD may be able to spontaneously mimic under some task demands, but that the response may be delayed relative to typically developing children (Oberman, Winkielman, & Ramachandran, 2009). It is still unclear what features lead to spontaneous mimicry in those with autism, and whether mimicry in those with other social disorders, such as Williams Syndrome, is preserved relative to those with ASD.

Objectives: To determine if children with ASD spontaneously mimic dynamic expressions of expressions and other actions, and to compare the levels of mimicry to typically developing controls and children with Williams Syndrome. 

Methods: 28 children with ASD, 21 typically developing children and 7 children with Williams Syndrome were shown 3000 ms videos of actors making happy and angry expressions, and arm wrestling with instructions to “just watch.”  Activity over their cheek (zygomaticus major), brow (corrugator supercili) and forearm (forearm flexor) was monitored with electromyography (EMG) while they watched these videos.  Maximal reactivity occurred 2000 to 3000 ms post stimulus onset; the pattern of activity during this period was analyzed using 100 ms windows, as described below.  

Results: Separate MANOVAs were run for each Stimulus (happy, anger and arm wrestling) with Muscle (corrugator, zygomaticus, forearm flexor) entered at a within-subjects factor, Diagnosis (ASD, typical, Williams Syndrome) entered as a between-subjects factor and Time (100 ms windows between 2000 to 3000 ms post stimulus onset) as a repeated measure. All groups demonstrated muscle specific mimicry to emotional expressions (i.e., increased zygomaticus to happy; corrugator to angry expressions), and combined flexor and corrugator activity to Arm Wrestling. There was a significant Time by Diagnosis by Muscle interaction for responses to Angry videos, F (2, 40) = 1.50, p = .03. No other three-way interactions were significant. In all models those with Williams Syndrome had greater absolute levels of mimicry than those with ASD and typically developing controls. Those with ASD did not have different levels of mimicry than typically developing children.

Conclusions: Children with Williams Syndrome may have greater levels of mimicry to dynamic expressions relative to children with ASD and typically developing children. Children with ASD showed similar levels of mimicry as those who are typically developing. These findings suggest that those with ASD may be able to spontaneously mimic dynamic facial expressions of emotion under minimal task demands. Future research should work to establish which situations those with ASD are likely to mimic spontaneously and the nature of mimicry in those with Williams Syndrome.