Imitation and Joint Attention As Predictors of Language Outcome in Infants at High and Low Risk for ASD

Background:  Language development requires the coordination of many early abilities. Motor imitation and responding to joint attention (RJA) are two early emerging behaviors that have been found to predict later expressive language ability in both typically developing (TD) children and those with ASD (Charman et al., 2000; 2003; Morales et al., 2000; Stone et al., 2001). Theory suggests that imitation promotes increased RJA behaviors because imitative play encourages infants to follow the attention of others for further imitation opportunities (Ingersoll et al., 2008; McDuffie et al., 2007). In turn, RJA skills are thought to facilitate language learning by increasing the infant’s ability to map language onto another person’s referent (Akhtar et al., 1996; Baldwin, 1995). In this study we propose and test a mediational model of language development in which the longitudinal relation between imitation and expressive language is mediated by infants’ ability to respond to joint attention bids.  To examine this pathway while these behaviors are first emerging, we study infants at elevated risk for ASD (i.e., infant siblings of children with ASD; HR infants) and their low-risk peers (LR infants). We predict that the directionality of the mediational pathway will be similar for the two groups. In light of previous studies finding stronger concurrent associations between these abilities for ASD than TD children (Mundy et al., 1990), we also predict that the mediation effect will be stronger for HR infants.

Objectives:  To examine a mediational model of the longitudinal relations between imitation, RJA, and expressive language in both HR infants and LR infants.

Methods:  The sample comprised 63 infants (HR n=38; LR n=25) who were assessed at 12, 15, 24, and 36 months of age through a longitudinal multisite study.  Motor imitation was measured at 12 months using the Screening Tool for Autism in Toddlers (STAT; Stone et al., 2008), RJA was measured at 15 months using the Early Social Communication Scales (ESCS; Mundy et al., 2003), and expressive language was measured at 24 months using the Mullen Scales of Early Learning (MSEL; Mullen, 1995). 

Results:  To conduct the mediation analysis, simple and conditional indirect effects were assessed using the PROCESS macro with bias-corrected, bootstrap-generated confidence intervals (10,000 samples; Hayes, 2013). As hypothesized, RJA mediated the relation between imitation and language such that: (1) 12-month imitation predicted 15-month RJA (R²=.09, F(1,61)=5.90, b=1.07, p<.02); (2) 15-month RJA predicted 24-month expressive language controlling for 12-month imitation (R²=.18, F(2,60)=6.49, p<.01); and (3) the indirect effect was significant, as the confidence interval did not include zero (c’=2.25, 95% CI [.35, 5.44]). Contrary to prediction, the strength of the mediation pathway was comparable for HR and LR infants. 

Conclusions:  These findings support the idea of a developmental sequence in which infants’ early imitation ability leads to higher levels of later expressive language in part through their intermediate RJA behaviors. The strength of the mediation is similar for both risk groups, which suggests that infants share a common developmental pathway.  Further analyses will be conducted as coding of all behaviors and diagnostic visits continue.