26038
Early Developmental Trajectories of Social Contingency Predict Language Outcome in Toddlers with ASD

Friday, May 12, 2017: 1:57 PM
Yerba Buena 9 (Marriott Marquis Hotel)
R. M. Fleurissaint, J. Bailey, S. Ghai and G. Ramsay, Marcus Autism Center, Children's Healthcare of Atlanta & Emory University School of Medicine, Atlanta, GA
Background: Although deficits in speech and language are no longer part of the definition of autism, most children with ASD are delayed in acquiring spoken language, and many of those children never learn to speak. A significant body of research has recently demonstrated the importance of contingent interaction between infant and caregiver in stimulating vocal development, and it is possible that early deficits in social interaction, which are a core feature of the syndrome, are responsible for later deficits in speech acquisition. If this is correct, early biomarkers quantifying the development and derailment of social engagement in infancy should be useful in predicting later language, and identifying developmental pathways that can be targeted by early intervention.

Objectives: The goal of this study was to test whether early developmental trajectories of social contingency between infant and caregiver over 0-24 months predict later language outcome at two years of age.

Methods: As part of an NIH Autism Center of Excellence, we tracked vocal development in 44 high-risk infant siblings and 30 low-risk controls. Using a recording device (LENA) worn by each child, we made audio recordings of each child's language environment at monthly intervals from 0-24 months. Using automatic speech recognition technology developed in our laboratory, we counted the number of vocalizations per hour for child and adult, as well as the rate of contingent interactions based on timing statistics. Using Functional Data Analysis, we calculated developmental trajectories for each child as well as mean trajectories for each group. We found significant differences in all three trajectories between risk groups, but these differences emerged at different time points. Starting at around 12 months, low-risk infants show a significant increase in the number of vocalizations they make, the number of times adults vocalize back to them, and the rate of conversational turns between adult and child. In contrast, high-risk infants show no increase in contingent interaction, which may be an early sign of autism. Three months later, adult vocalizations begin to decline. After a further three months, high-risk infants lag behind their low-risk peers. The disruption of the natural process of vocal interaction between infant and caregiver clearly demonstrates the developmental cascade of derailment that unfolds from early deficits in social engagement. To determine the impact of this process on later language outcome, we examined the correlation between receptive and expressive language scores from the Mullen Scales of Early Learning for children who completed clinical assessments at 24 months, and the slope and intercept of a two-parameter exponential growth curve fitted to every child’s data.

Results: We found significant correlation (r=0.399, P<0.05) between the expressive language score and the slope of the trajectory for conversational turns; surprisingly, no other correlations were significant.

Conclusions: Infants at risk of ASD undergo a developmental cascade of derailments that begins with an early decline in contingent interaction that predicts lower expressive language, consistent with the hypothesis that core deficits in social engagement associated with ASD are responsible for comorbid deficits in spoken language.