Engagement and Disengagement of Visual Attention in Children with Autism Spectrum Disorder

Background: Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder commonly characterized by abnormalities in social interaction, including decreased time spent looking at others, as well as abnormal reciprocal gaze and eye-contact. Previous studies using eyetracking technologies have identified general deficits in engagement and disengagement of visual attention in children with ASD. These findings suggest that abnormalities in social interaction may be, at least in part, the product of visual attention deficits. Such studies, however, have yet to investigate whether engagement of visual attention in children with ASD is differentially impacted by social and non-social stimuli.

Objectives: We sought to investigate whether visual attention in children with ASD is differentially engaged and disengaged by social and non-social stimuli.

Methods: Eyetracking data were recorded from 11 individuals with ASD and 10 typically-developing (TD) individuals, between the ages of 9 and 16. Participants fixated on a central stimulus for 1s, at which point a peripheral stimulus appeared to one side, either while the central stimulus remained on screen (“overlap” condition) or 250ms after the central stimulus disappeared (“gap” condition). Stimuli were presented in a randomized order, in grayscale, for both social (20 faces of children) and non-social (20 objects, including tennis ball, the Earth, cartwheel, etc.) conditions. Latency to saccade to the peripheral stimulus was recorded, and the gap effect (difference between average “overlap” and “gap” saccade latencies) for each participant was calculated. Reduced gap effect reflects deficits in engagement of visual attention; elevated gap effect indexes deficits in disengagement of visual attention.

Results: There was no significant difference in overall gap effect between ASD and TD participants (p=.77). For trials where both central and peripheral stimuli were of the same kind, there was no significant difference in gap effect between ASD and TD children (social:social p=.34; nonsocial:nonsocial p=.73). On the other hand, for trials with a social central stimulus and a nonsocial peripheral stimulus, there was a gap effect in TD children (p<.01) but none in children with ASD (p=.39). For trials with a nonsocial central stimulus and a social peripheral stimulus, there was a gap effect in children with ASD (p<.01), but none in TD children (p=.30).

Conclusions: An overall reduction in engagement of visual attention was not observed in ASD children compared to TD children, as there was no significant difference in gap effect between groups overall, or for trials in which both central and peripheral stimuli were of the same kind. However, ASD children displayed comparatively reduced engagement with social stimuli when presented with peripheral nonsocial stimuli, as well as increased engagement with nonsocial stimuli when presented with peripheral social stimuli. This finding suggests that children with ASD may have reduced engagement of visual attention that is specific to social stimuli, rather than a general deficit in visual engagement. It may provide a potential explanation for some of the abnormalities in social interaction observed in children with ASD.