Relationships Between Engagement States and Early Functioning in Children with Autism and Typical Development

Background:

In both Typically Developing (TD) children and children with ASD, different engagement types have been found to influence language development (1). For example, Joint Attention (JA) and Supported Joint Engagement (SJE) have been shown to predict later vocabulary (3). More specifically, Bottema-Beutel et al. (2014) reported that after dividing SJE into ‘Higher’ and ‘Lower’ only HSJE predicted later social communication and expressive language in children with ASD. However, it is not clear which aspects of SJE are most relevant for language growth, nor whether participation in SJE is uniquely facilitative for children with ASD, as Bottema-Beutel et al., (2014) did not compare SJE and JA, nor did they compare children with ASD and TD children.

Objectives:

We extend the paradigm of Bottema-Beutel et al., (2014) to a new sample of children with ASD, compare them to TD children, and investigate how each of the different engagement states are related to the children’s early language scores.

Methods:

Participants include 15 children with ASD (ASD_MAge=34.93 months, 12 males), and 15 TD children (TD_MAage=19.82 months, 13 males), recorded at 3 visits, 4 months apart, and initially matched on language level (raw scores) (ASD_MullenEL=16.44(6.22), TD_MullenEL=20.35(5.70)). Children and caregivers engaged in 30-minute play sessions, which were coded for episode types in two separate waves. First, three types of JA were coded, designated as when the child and caregiver alternated between looking at each other and an object, with the child either responding (RJA) or initiating (IJA), or JA being mutually established (JA). Second, two types of SJE were coded, designated as when the caregiver influenced the child's object play, but the child engaged with the caregiver without visually referencing him/her (1): Higher (HSJE) occurred when the child responded by reciprocating and collaborating with the caregiver and Lower (LSJE) occurred when the child responded without reciprocal or collaborative exchange with the caregiver. A seventh episode type, Passive Attention (PA) was coded when caregivers followed the child’s attention with no behavioral response. Correlations were performed between children’s durations of engaging in each episode type and their Mullen scores.

Results:

In the TD group, only Expressive language correlated significantly with joint activity; children with higher scores also engaged in longer episodes of JA (r=.5419, p=.0369). In contrast, in the ASD group four significant correlations were obtained: Children with better Receptive language engaged in longer RJA episodes (r=.5697, p=.0335) and shorter LSJE episodes (r=-.5593, p=.0376), and children with better Visual Reception engaged in longer HSJE episodes (r=.6618, p=.0099) and longer RJA episodes (r=.6233, p=.0172).

Conclusions:

These findings demonstrate different patterns of correlations in children with ASD and TD children, with joint activities being related to cognition as well as language in the ASD group but only language in the TD group. We next investigate whether the aspects that predict children’s later vocabulary growth will also be different in the 2 groups, and compare which engagement states are predictive of later language development.