Changes in Social Referencing in Response to Pivotal Response Treatment in ASD: A Preliminary Eye Tracking Study

Poster Presentation
Friday, May 11, 2018: 5:30 PM-7:00 PM
Hall Grote Zaal (de Doelen ICC Rotterdam)
M. Kim1, C. Foster2, Q. Wang3, C. A. Wall4, B. Li5, E. Barney6, Y. A. Ahn7, L. L. Booth8, M. Lyons3, C. A. Paisley9, C. C. Kautz8, F. Shic6 and P. E. Ventola10, (1)Seattle Children's Research Institute, Seattle, WA, (2)Binghamton University, Binghamton, NY, (3)Child Study Center, Yale University School of Medicine, New Haven, CT, (4)Department of Psychology, University of South Carolina, Columbia, SC, (5)Computer Science and Engineering, University of Washington, Seattle, WA, (6)Center for Child Health, Behavior and Development, Seattle Children's Research Institute, Seattle, WA, (7)University of Miami, Miami, FL, (8)Yale Child Study Center, Yale School of Medicine, New Haven, CT, (9)The University of Alabama, Tuscaloosa, AL, (10)Yale Child Study Center, Yale University School of Medicine, New Haven, CT
Background: Social referencing behavior is exhibited less in children with ASD than those with typical development (TD; Dawson et al., 2004). Pivotal Response Training (PRT) addresses this by targeting social motivation and social use of language. In an ongoing study, we used eye tracking paradigms designed to capture intervention outcomes. Here we report preliminary results from a Social Referencing paradigm that examines the monitoring of nonverbal information.

Objectives: To explore PRT treatment effects by comparing eye tracking data at Baseline and Endpoint (16 weeks later) between PRT and Waitlist Control (WLC) groups and examine the utility of eye tracking in capturing phenotypic changes moderated by PRT.

Methods: 4-to-8 year olds with ASD randomized to either PRT (n=16) or Waitlist Control (WLC, n=13) watched videos of an actor involved in an escalation of stressful activities (e.g. stacking a tower) that comes to a resolution (e.g. tower falls down). A linear mixed model, controlling for age, was used to examine eye tracking variables of valid onscreen looking time (ValidLooking%) and looking percentages to actor/scene regions (Face%, Activity%, Eyes%, and Mouth%) throughout the different stages of the activity (Escalation and Resolution). Participants re-watched the videos at 8-week Midpoint (PRT n=14, WLCl n=9) and 16-week Endpoint (PRT n=12, WLC n=9). Relationships between changes in clinical variables and eye tracking variables were examined using Pearson’s correlations.

Results: Between-group differences were only detected for Mouth% at Endpoint during the Resolution (p=.048), with lower mouth looking in the PRT group (M=10.1, SD=1.10) than WLC (M=6.60, SD=1.00). Within the PRT group, there was a negative correlation between time point changes in Face% during the Resolution and time point changes in SRS T-scores in Autistic Mannerism (r(10)=-.725, p=.018), suggesting face looking increases were associated with reduced severity in autism mannerisms. However time point changes in Face% during Resolution was also negatively correlated with time point changes in behavioral coding of reciprocal turns in conversations (r(12)=-.657, p=.020), suggesting that increased face monitoring during Resolution was not a prerequisite for improving social reciprocity.

Conclusions: Preliminary results suggest that PRT is associated with Endpoint outcomes of decreased attention to the mouth compared to WLC. Previous studies have suggested that mouth scanning can supplement communicative cues, especially in uncertain situations (Klin et al., 2002). However when the situation is easier to understand, attention to the mouth may be less necessary (Norbury et al., 2009). By treatment Endpoint, PRT participants may have developed efficient processing skills and begun to rely less on mouth scanning. Caution is warranted, however, as correlational results suggest heterogeneity within the PRT group in treatment and eye tracking results. Furthermore, improvements in PRT targets may not be adequately captured by time point increases in looking at actor/scene regions of the Social Referencing paradigm. Future analyses will examine the complicated and nuanced relationships between eye tracking data and behavioral coding of PRT intervention goals.