Patterns of Attention to Non-Social Stimuli in Children with Autism Spectrum Disorder

Background: Aberrant attention patterns in Autism Spectrum Disorder (ASD) have been widely reported. Eye-tracking offers a non-invasive way to examine visual attention; however, most ASD studies have focused on attention to social stimuli (e.g. Jones & Klin, 2013) with very few studies examining attention to non-social stimuli. The impact of low-level sensory features (e.g. motion, modality) on attention patterns in ASD has not been comprehensively studied.

Objectives: In this eye tracking study, we aimed to test how the sensory properties of non-social stimuli impacted the allocation of attention in ASD and how patterns of attention associate with clinical sensory profiles.

Methods: Forty children [15 ASD, 15 Typically Developing children (TD) and 10 Developmental Delay excluding ASD (DD)] ages 4 to 13 years were included. Parents provided data on child sensory response patterns via the Sensory Experiences Questionnaire, Version 3.0 (SEQ-3; Ausderau et al. 2014). Children completed a passive-viewing eye-tracking task designed to measure attention to non-social stimuli; six novel, objects with interesting visual and auditory qualities were presented in the center of the screen in either unimodal (static or dynamic) or multimodal (dynamic+auditory) format. Stimuli were shown on the screen one at a time for 15 seconds in a random order.

Results: There were no group (ASD/TD/DD) differences in total attention to stimuli (F(2, 79)=1.48, p=.23). However, there was a condition effect: all subjects spent more time attending to dynamic and dynamic+auditory stimuli compared to static (F(2, 79)=8.94, p<.001). There were no group or condition effects for time to first fixation, suggesting initial orientation was similar across groups and conditions. However, first fixations were longer to both dynamic and dynamic+auditory stimuli compared to static (F(2, 79)=23.43, p<.001). There was no condition effect for the number of individual fixations, however children in the ASD group fixated fewer times overall than DD and TD controls (F(2, 79)=9.89, p<.001). Associations between clinical sensory response patterns and attention were found in the ASD group only; Hyperresponsivness correlated positively with both total fixation duration (r=.57 to77, all p’s<.05) and mean fixation duration (r=.61 to.79, all p’s <.01) across all conditions. Time to first fixation for dynamic and dynamic+auditory stimuli correlated negatively with hyperresponsiveness scores (both p’s<.05).

Conclusions: Children in all three groups had comparable rates of overall attention and initial orientation to non-social stimuli in our study, with more attention to dynamic and dynamic+auditory versus static stimuli. Thus, sensory attributes (e.g. motion) appeared similarly salient to children regardless of diagnosis. Our findings using non-social stimuli are partially consistent with Chevalier et al. (2015) showing no group differences in static and dynamic stimuli when presented independent of social interaction. We conclude that overall attention and initial orientation to non-social stimuli is comparable across groups, with some sensory qualities, such as motion, producing a facilitatory effect on visual attention for all children. However, we found enhanced attention (i.e. looking more quickly and for longer periods to non-social stimuli) was only associated with clinical measures of hyperresponsiveness in children with ASD, suggesting the need for future mechanistic research.