Distinctive Abnormalities in Imitative Response to Socially Engaging Versus Neutral Partners Are Specific to ASD and Predict Treatment Outcomes

Background:   While imitation difficulties are often documented in preschoolers with ASD, it is not yet clear whether such abnormalities reflect differences of a social, attentional, cognitive and/or motor nature. As early educational intervention programs often require children to imitate actions and behaviors that are demonstrated to them during social routines, it is plausible that differences in imitation will impact on response to educational programs in this population. Therefore, a fine-grained understanding of the social, attentional, motor and cognitive processes underlying imitation abnormalities in ASD has the potential to advance knowledge on both the mechanisms of impairment and the mechanisms of treatment response in this population.  

Objectives:  We investigated multiple processes underlying imitation in 35 preschoolers with ASD and 20 peers with Williams syndrome (WS), matched for age, cognitive, verbal and motor functioning, with the aim to identify whether putative imitation differences between the groups were (1) specifically linked to differences in sociability observed in these two disorders, or associated with motor and cognitive difficulties shared by ASD and WS, and (2) linked to learning outcomes in response to early intervention in ASD.

Methods: We tested participants’ spontaneous propensity to imitate others and the accuracy of imitation performance in response to a series of novel eye-tracking-based imitation tasks, in which the following factors were manipulated 1) motor demands, 2) social engagement of the model demonstrating the action, 3) presence/absence of clear mean-goals structure, and 4) attentional demands. In the ASD group, we also examined the extent to which individual differences in imitative learning were correlated to intensive early intervention gains occurring in the 12 months following the test. 

Results:   We found that children with WS increased their attention to the model and their propensity to imitate the demonstrated action in response to a socially engaging versus a neutral partner. In contrast, preschoolers with ASD showed a similarly reduced propensity to look at the model and to imitate her actions regardless to whether the model was socially engaging or neutral (Group by Condition Interaction F (2, 52)=5.5; p<0.05, η2= 0.95). We also found that children with WS imitated actions that were not relevant to the actions’ outcomes (overimitation) while this was not the case in children with ASD (Mann Whitney U = 290, p<.05).  In contrast, the two groups showed similar strengths in the instrumental aspects of imitation (e.g. understanding of the mean-goals articulation of the demonstrated actions) and a similar decrease in imitation accuracy as the motor complexity of the demonstrated actions increased. Imitative response to a socially engaging but not to a neutral model in ASD was a robust predictor to early intervention outcomes (r=.7, p<.005).

Conclusions:   Social motivational factors underlying imitation appear to be distinctively impaired in young children with ASD and to be linked to early intervention outcomes.