Components of Limited Activity Monitoring in Toddlers and Children with ASD

Background: Our previous work with eye tracking has shown that 20 month old toddlers with ASD monitor the activities of others to a lesser extent than both developmentally delayed and typically developing peers. However, it is unclear whether the gaze cues of others, the presence of distractors, or motion cues were responsible for the differences between groups. Furthermore, it is unclear whether diminished activity monitoring is only present in the toddler years, resolving as the children grow older, or whether these deficits persist.

Objectives:  To use eye tracking to examine activity monitoring in toddlers and children with ASD and to decompose factors that impact activity monitoring.

Methods:  Toddlers with ASD (N=10; Age: M=23, SD=3 months) and typical development (TD; N=23; Age: M=21, SD=3 months) and children with ASD (N=17; Age: M=37, SD=1 month) and TD (N=9; Age: M=38, SD=3 months) were shown 16 20s video clips and 16 10s static images depicting two female adults interacting over a shared activity. Stimuli varied along 3 dimensions: (1) Gaze: mutual towards each other or towards the activity; (2) Distractors: many distractors or no distractors, where distractors were colorful toys; and (3) Motion: static image or video clip. Stimuli were counterbalanced across and within participants and eye tracking was used to evaluate patterns of attention. A 2nd order linear mixed model approach was used to examine attention to the scene, activities, people, and background elements.

Results:  Decreased looking at the scene overall was associated with ASD (p<.01), fewer distractors (p<.05), and no motion (p<.01). TD participants looked more at the scene when motion was present (Group x motion interaction, p<.01). Decreased looking at activities was associated with ASD (p<.01), being in the younger age group (p<.01), the presence of more distractors (p<.01), and the lack of motion (p<.01). Older TD children looked more at activities than other groups (Group x age interaction, p<.01). Increased looking at the background was associated with ASD (p<.01), being younger (p<.05), more distractors (p<.01), and no motion (p<.01). Decreased looking at the people in the scene was associated with ASD (p<.01), being older (p<.01), more distractors (p<.01), and presence of motion (p<.01). Younger TD toddlers looked more at people than all other groups (Group x age interaction, p<.01). Eye tracking outcome measure associations with clinical characterization in ASD replicated previously observed findings.

Conclusions: Our results suggest that toddlers and children with ASD show a general pattern of diminished attention towards people and their activities. In typical development but not ASD, transitions consistent with a sharpening of attention towards the activities of others appear between 2 and 3 years of age. Interestingly, an effect of gaze direction was not present in the results of any outcome measures, suggesting that dynamic and complexity cues may play a greater role in shaping attention to scene-relevant context at these ages.