Broad Autism Phenotype in Typically Developing Children Predicts Performance on An Eye-Tracking Measure of Joint Attention

Background: The current study constitutes the second step in a program of research that aims to develop, refine and validate a candidate endophenotype measure that has the potential to enhance our understanding of the etiology of Autism Spectrum Disorder (ASD). Children with this disorder often show deficits in gaze following. The proposed measure takes advantage of modern eye-tracking technology and evaluates how individuals allocate their attention when viewing social video vignettes that display an adult model (Face) who is gazing at a series of targets (Target) that appear and disappear in the four corners of the screen (congruent condition). Gaze allocation in the experimental condition is compared to a set of control stimuli where the model’s gaze is not directed at the targets (incongruent condition). Data from our previous research (Swanson et al., under review) on neurotypical adults (N = 44) revealed two major findings. First, gaze allocation of adults differed significantly between the congruent and incongruent conditions. Second, individual differences in gaze allocation were significantly predicted by a self-report measure evaluating features of the broad autism phenotype (BAP).

Objectives: The current study adds to this program of research in two ways. First, we evaluate the feasibility of administering our experimental paradigm to a population of typically developing children. Second, we investigate how scores on a parent-report measure of the BAP predict gaze allocation during our experimental paradigm (SRS, Social Responsiveness Scale; Constantino et al., 2003).

Methods: Fifty typically developing children between the ages of 3 and 9 years participated in a series of standardized assessments to evaluate their verbal and non-verbal cognitive abilities. Parents completed the SRS, background, and medical questionnaires. Four children (8%) were excluded due to poor eye tracking data. The remaining 46 children (21 girls, 25 boys) were on average 6 years, 4 months old (SD=19.6 months).

Results: Typically developing children allocated less attention to the Face in the congruent (estimated marginal mean = 28.35%, SE = 2.17) than in the incongruent (estimated marginal mean = 32.41%, SE = 2.18) condition, F(1,1221) = 10.77, p<.01.  Conversely, participants allocated more attention to the Target in the congruent (estimated marginal mean = 66.72%, SE = 2.31) than in the incongruent (estimated marginal mean = 62.54%, SE = 2.33) condition, F(1,1221) = 10.59, p<.01. We also fit a model with gaze time to the Target as the dependent variable, experimental condition (congruent or incongruent) and SRS t scores as main effects, and one interaction term (experimental condition*SRS t scores).  With regards to the Target, results showed a significant condition*SRS t scores interaction effect, F(1,1221) = 7.73, p<.01. These results indicate that differences in gaze time allocation between the congruent and incongruent condition are more pronounced for participants with low SRS scores and less pronounced for individuals with high SRS scores.

Conclusions:  

This program of research enhances our knowledge of basic science in regards to underlying neuropsychological processes that represent a potential ASD endophenotype. We are currently extending this research to include 22 language-matched children with ASD.