Sensitivity and Response to Direct and Averted Gaze in Toddlers with Autism Spectrum Disorders

Background: Individuals use direct and averted gaze for a variety of social and communicative purposes, including conveying emotion and directing attention. Research has repeatedly found altered gaze monitoring in individuals with autism spectrum disorders (ASD). Atypicalities include diminished sensitivity to direct eye contact and a lack of differential response to attentional cues originating from eye gaze rather than from non-social cues.  Much of the research has either relied on general observational measures (e.g., parent or clinician ratings of how a child responds to gaze cues) or has used fine-grained measures at the expense of naturalistic presentation (e.g., relying on static or otherwise simplified experimental paradigms), thus reducing ecological validity.  It is unclear how this work translates into sustained naturalistic interaction, which requires an ongoing and fine-scaled temporal sensitivity to direct and averted gaze. Examining these processes in young children will help illuminate the developmental path of gaze monitoring in individuals with ASD.  To that end, the present study investigates the reactions of toddlers to gaze shifts while watching naturalistic video of an actress.

Objectives: This study uses eye-tracking technology to measure the sensitivity and response of toddlers with ASD to direct and averted gaze during a natural viewing task.

Methods: Two-year-olds with ASD and typically developing (TD) two-year-olds, matched on age and nonverbal functioning, watched a three-minute video of an actress interacting with hand puppets.  Throughout the video, the actress shifted her gaze back and forth from the viewer to the puppets.  Her gaze shifts were spontaneous and contingent upon her social actions, resulting in variable durations of averted and direct gaze.  Each toddler’s visual scanning patterns were examined within a fixed temporal window following each gaze shift to measure temporal sensitivity and response to change in gaze.  Two aggregated variables were also analyzed for each child, one of which combined responses for all instances of direct gaze, and the other which collapsed across instances of averted, puppet-directed gaze.  All measures were then evaluated at the group level, comparing toddlers with ASD to TD toddlers.

Results: Preliminary analyses indicate that TD toddlers and toddlers with ASD respond differently to gaze shifts by the actress.  TD toddlers appear more sensitive to instances of direct gaze, as they were quicker to look at the actress’s face when she re-established eye contact. A similar bias toward the face was found during averted, puppet-directed gaze.  In these situations, TD children were slower to shift attention away from her face to the referenced object.

Conclusions: Toddlers with ASD differ from their TD peers both in their sensitivity to re-establishment of direct gaze and in their response to gaze shifts directed to a particular object.  This early divergence from the typical response to natural gaze has implications for both social and communicative development.