Longitudinal Indirect Effect of Parent Responsiveness and Vocal Complexity on Expressive Language Outcome in Toddlers at Risk for ASD

Background: Up to 40% of later-born siblings of children with ASD are diagnosed with ASD (7-19%) or other language and/or cognitive delays (14-20%) by age three. Vocal complexity, a measure of the quality and maturity of vocal communication, is a strong predictor of later expressive language in typically developing children and children with ASD, but no studies have examined this relation for toddlers at high familial risk (HR) for ASD.

Parent responsivity to toddlers’ vocalizations may facilitate children’s complex vocal communication, which in turn may provide the basis for expressive language. ImPACT is a parent-implemented, naturalistic developmental behavioral intervention (NDBI) that teaches parents to respond contingently to toddlers’ vocalizations, which could facilitate complex vocal communication and expressive language.

Objectives: (1) To examine whether parents’ verbal responsiveness (PVR) indirectly predicted expressive language through vocal complexity in an ongoing longitudinal RCT of the ImPACT intervention; and (2) To explore longitudinal indirect effects of ImPACT on expressive language through PVR or HR toddlers’ vocal complexity.

Methods: Participants were 54 HR toddlers and parents randomly assigned to receive 3 months of ImPACT (n=28) or to business-as-usual (n=26). Toddlers were 12-18 months old at study entry (M=14.32 months; SD=2.03 months). PVR was measured at Time 2 (immediately post-ImPACT). Vocal complexity was measured at Time 3 (3 months post-ImPACT). Expressive language was measured at Time 4 (6 months post-ImPACT).

Parents’ verbal responsiveness (PVR) to children’s vocally complex communication acts was coded from a parent-child free play and snack time task using observational coding definitions from previous research. Vocal complexity, an aggregate of canonical syllable frequency and consonant inventory in communication acts, was coded from the Communication and Symbolic Behavior Scales (CSBS-DP) and the Brief Observation of Social Communication Change (BOSCC). Expressive language was an aggregate of T4 MCDI expressive vocabulary, number of word roots used during the CSBS-DP, and MSEL expressive language subscale. Indirect effects were assessed using bias-corrected bootstrapping.

Results: For toddlers in both treatment conditions, there was a significant indirect effect: Time 2 PVR significantly predicted T4 expressive language, in part, through Time 3 vocal complexity, ab=1.74, 95%CI [0.64, 3.19] (Table 1; Figure 1). This indirect association did not vary by treatment group. Participation in the ImPACT intervention did not significantly affect Time 4 expressive language indirectly through Time 2 PVR or Time 3 vocal complexity.

Conclusions: To our knowledge, this was the first study to assess whether the relation between parental verbal responsiveness and children’s later expressive language is partly due to intermediate increases in the maturity of children’s vocalizations. Although this preliminary analysis did not find a significant effect of the ImPACT intervention on PVR or vocal complexity, it did find that regardless of treatment condition, PVR predicted an increase in the proportion of HR toddlers’ communication acts that were vocally complex, which in turn predicted later expressive language ability. Parents’ verbal responsiveness and children’s vocal production clearly influence each other in a transactional manner over time, a phenomenon that could have cascading implications for HR children.