Vocal-Gesture Profiling at Age 2: Diminished Gesturing Is Associated with Greater Distress in Toddlers with ASD

Poster Presentation
Friday, May 11, 2018: 11:30 AM-1:30 PM
Hall Grote Zaal (de Doelen ICC Rotterdam)
S. Plate1, N. Libster2, J. Brown1, A. Faggen1, R. F. Slomowitz3, J. Wood1, J. Maldarelli3, J. Pandey1, R. T. Schultz1, J. Parish-Morris1 and .. The IBIS Network4, (1)Center for Autism Research, Children's Hospital of Philadelphia, Philadelphia, PA, (2)UCLA Center for Autism Research and Treatment, Los Angeles, CA, (3)Center for Autism Research, The Children's Hospital of Philadelphia, Philadelphia, PA, (4)University of North Carolina, Chapel Hill, NC
Background: Autism Spectrum Disorder is defined by early and persistent deficits in social communication, as well as restricted interests and repetitive behaviors. Previous research has shown that gesture and language are both atypical in infants who ultimately receive an ASD diagnosis (Colgan et al., 2006; Lazenby et al., 2016; Zwaigenbaum et al., 2005). In this study, we use a system of high-dimensional behavioral coding to explore 2-year-olds’ relative use of vocalization and gesture during a structured clinical evaluation. Vocal-gesture profiles generated using this method could help identify homogeneous clusters of children that respond comparably to targeted treatments, or who follow similar developmental trajectories.

Objectives: Examine relative gesture and vocalization use in 2-year-olds at high- or low- risk of ASD that did or did not receive a diagnosis.

Methods: Forty-four 2-year-olds were administered the Communication and Symbolic Behavior Scales (CSBS, Wetherby & Prizant, 2003) as part of the Infant Brain Imaging Study (IBIS; Estes et al., 2015). Participants were classified as LR- (no family history of ASD; N=22, 6 female), HR-neg (high familial risk for ASD, no ASD themselves; N=11, 4 female), or HR-ASD (high familial risk for ASD, ASD themselves; N=11, 1 female). Videos of the CSBS were coded for vocalizations and communicative gestures. Vocalizations were categorized as speech (e.g., words, babbles) or non-speech (distress, delight, and atypical sounds like squealing or growling). To explore relative reliance on vocalization and gestures, we calculated the amount of time each child spent vocalizing and/or gesturing divided by their own total coded behavior duration.

Results: A linear mixed effects model with participant ID as a random effect revealed a significant interaction between group (LR-, HR-neg, and HR-ASD) and behavior type (speech, non-speech, gesture). Compared to the LR- group, the HR-ASD group produced a higher ratio of nonspeech vocalizations vs. communicative gestures (t=2.10, p=.04; Figure). Speech ratios did not differ by group. To determine what type of nonspeech vocalizations the HR-ASD group produced in lieu of gestures, an additional mixed effects model compared nonspeech categories in the HR-ASD and LR- groups. Results revealed that the HR-ASD group produced a significantly larger proportion of distress vocalizations (crying) vs. delight vocalizations (laughing) than the LR- group (t=2.56, p=.01).

Conclusions: Granular profiles of vocal and gestural behavior could help identify homogeneous subgroups of children that respond better to one treatment vs. another, or follow similar developmental pathways. The results of this preliminary study showed that reduced gesturing and increased nonspeech vocalization differentiated the HR-ASD and LR- groups, and that diminished gestural communication was associated with greater distress vocalizations despite similar relative amounts of speech. This suggests that when communicative gestures are diminished, children with ASD do not compensate by adding more communicative speech – rather, they produce more non-speech vocalizations that express distress. Our next steps include tracing the developmental trajectory of these vocal-gesture profiles by coding vocalizations and gestures in the same children at 6 and 12 months, and exploring ways that children use speech and gesture separately vs. in combination (Winder et al., 2013).