Object Play as a Moderator of Intervention Effects on Responding to Joint Attention In Children with ASD

Background:  Children with autism spectrum disorders (ASD) appear to demonstrate deficits in responding to bids for joint attention.  Importantly, the ability to allocate attention more flexibly and consistently may allow children to map adult linguistic input more accurately to objects or events to which their attention has been drawn (Bono et al., 2004).  Previous findings suggest children with ASD who are more competent responders to adult bids for joint attention may make larger language gains over time (Siller & Sigman, 2002).  Development of spoken language has been linked to better outcomes in individuals with ASD.  Therefore, early prelinguistic skills such as responding to joint attention (RJA) are important targets for early intervention in young children with ASD.  Many early intervention and instructional strategies are implemented during routines or play involving objects.  Logically, a child’s knowledge of objects used during such routines may influence a child’s ability to engage successfully in triadic interactions with adults and access the content of interventions.  Additionally, as children become competent in their knowledge of objects, they may direct more cognitive resources to new skills, such as following another’s attention. 

Objectives:  The current study examined the following research question: Does object knowledge (the number of differentiated actions performed on toys) moderate the effects of a parent-mediated social-communication intervention (More Than Words [MTW]) on frequency of RJA acts in young children with ASD?

Methods:  A multi-site, randomized controlled trial was used to address the research question.  Sixty-three participants were randomized to the treatment or control group.  Forty-five children had analyzable data.  Twenty-three children were randomized to the control group, and 22 were randomized to the MTW intervention.  Object knowledge was measured pretreatment (Time 1) using the Developmental Play Assessment (Lifter, 2000).  Responding to joint attention was measured at the five month posttreatment follow up period (Time 3) using the Early Social Communication Scales (ESCS; Mundy et al., 2003).  The ESCS was administered by an unfamiliar examiner using different materials, setting, and interactional style, providing a measure of far generalization of RJA skills.    

Results:  Hierarchical linear regression was used to examine object knowledge as a moderator of effects of treatment on RJA.  A significant interaction was found, where treatment led to changes in RJA conditional upon Time 1 object knowledge (t(45) = -2.60, p = .013).  Children with lower levels of object knowledge at Time 1 showed greater gains in RJA when assigned to the MTW treatment; children with higher levels of object knowledge at Time 1 showed greater gains in RJA when assigned to the control condition.

Conclusions:  Children with lower levels of object knowledge may have experienced increases in object knowledge due to the MTW treatment, which allowed them greater access to intervention strategies targeting social-communication skills (e.g., RJA).  It is not yet understood why children with more play skills showed greater gains in RJA when assigned to the control condition.  These findings highlight the importance of continued research into child and family characteristics that influence effects of treatment.