Distinguishing Gaze Aversion From Gaze Indifference in Two-Year-Olds with Autism

Background: Atypical eye contact is among the most prominent and early-emerging symptoms of autism spectrum disorders (ASD). Eye-tracking studies of individuals with ASD during viewing of dynamic social stimuli and some instances of static stimuli have found reduced fixation on the eyes as well as increased fixation on both the mouth and background objects relative to typically developing (TD) controls. However, the underlying cause of this difference in visual attention remains controversial. One hypothesis is that children with ASD purposefully shift visual attention away from the eyes because the eyes have negative emotional saliency (e.g., leading to hyperarousal). Alternately, children with ASD may show reduced fixation on the eyes because they are indifferent to the significant social cues conveyed by the eyes. In other words, one hypothesis suggests an active avoidance with implicit awareness of the social significance of eye contact, while the other suggests insensitivity to the underlying social signal.

Objectives: The current study seeks to determine whether reduced visual fixation on eyes in toddlers with ASD reflects an aversion or indifference to direct gaze.

Methods: Eye-tracking data were collected while two-year-olds with ASD and matched TD peers viewed video clips of actresses portraying caregivers engaged in naturalistic, child-directed interaction. During interstimulus intervals between each video scene, visual attention was directed towards a centering cue so that all viewers began viewing of the scenes by fixating at the same locations.  The position of the actresses within the video frame was systematically adjusted between scenes so that the location of the centering cue would prime a viewer to fixate on either the actress’ eyes or mouth when the cue disappeared and the video began. Measures of the directionality of initial saccades and duration of fixations were compared between conditions. In addition, a novel computational method of quantifying visual scanning patterns through time was used to determine periods within the clips when the eyes were most salient to TD viewers. If children with ASD find eye gaze aversive, they would be more likely to shift attention away from the eyes during these time periods; on the other hand, if children with ASD were insensitive to social cues from the eyes, their tendency to look towards or away from the eyes would be unchanged during these periods.

Results: Preliminary analyses suggest that when primed to orient towards the eyes, children with ASD exhibit increased attention to the eyes as compared to cases when they are primed to orient towards the mouth. Across conditions, children with ASD did not specifically redirect their scanning away from the eyes when the eyes conveyed the most salient social cues.

Conclusions: The reduced eye fixations observed in the viewing patterns of children with ASD indicate insensitivity to social cues conveyed by another’s eyes rather than an aversion to gaze.