Examining Mediators of an Adaptive Communication Intervention for Young Children with Autism Spectrum Disorders

Background: As many as 30% of children diagnosed with ASD are classified as nonverbal or minimally verbal at age 5 despite access to early intervention (Tager-Fluberg & Kasari, 2013). Persistent minimal verbal status is associated with poor long-term prognoses for social and adaptive functioning (DeMyer et al., 1973; Liss et al., 2001). In order to better address the skill deficits of young, preverbal preschoolers who are at risk for becoming minimally verbal children, an intervention combining two evidence-based treatments, J-EMT (Kasari et al., 2014) and DTT (Smith, 2001) was developed and manualized for implementation in a RCT. The intervention was designed to specifically target foundational skills for communication that individual children did not display and to modify instruction based on their progress during intervention. These foundational skills included: joint attention, receptive language, imitation, and proficiency in using a speech-generating device (SGD). These skills have shown to be predictive of long-term outcomes in language and communication (e.g. Charman, 2003; Luyster, et al., 2007; Gernsbacher, 2008; Kasari, et al., 2014).

Objectives: The objective of the current study is to examine how changes in foundational communication skills mediated the outcomes of an adaptive communication intervention by asking the following question: Do joint attention initiations, receptive language ability, imitation skill, and SGD proficiency mediate the effect of group assignment on social communication outcomes?

Methods: Data for this analysis are from a RCT (R40MC27707). Seventy-three children with ASD were randomly assigned to either a 36-session intervention combining DTT and J-EMT including an SGD or a business as usual control group. The dependent variable, defined as the total number of social communicative utterances, was measured in a 20-min language sample with an unfamiliar assessor. The putative mediators were measured as follows: initiating joint attention (IJA) was measured from the Early Social Communication Scales (ESCS; Mundy et al., 2003); receptive language was measured from the auditory comprehension subscale of PLS-5 (Zimmerman, et al., 2011); verbal imitation was measured as the number of attempted consonants during the Profiles of Early Expressive Phonology (PEEPS; Williams & Stoel-Gammon, 2014); SGD proficiency was measured as the size of visual field from which the child is able to identify an object on the SGD. Mediators were measured at posttest; the dependent variable was measured 2-months following posttest.

Results: Baseline characteristics are presented in Table 1. There were no significant between-group differences on relevant variables at pretest. Coding and analysis for all time-points are currently 80% complete. Mediation analyses will be used to examine the underlying mechanism driving the relationship between the independent variable (group assignment) and dependent variable (social communicative utterances; Hayes, 2009).

Conclusions: Findings of this study will identify foundational skills that mediate language and communication outcomes for preschooler-aged children with ASD who are at risk for becoming minimally verbal. These findings may lay the foundation for developing and testing a decision-making framework for early intervention targets, including establishing criterion levels of behavior change on specific malleable factors required to benefit from interventions based in naturalistic principles.