Eye Avoidance in Young Children with Autism Spectrum Disorder When Scanning Emotional Faces

Background: Abnormal face scanning patterns have been consistently found in individuals with autism spectrum disorders (ASD) (see Falck-Ytter & von Hofsten, 2011, for a review), which may be associated with their face processing deficits. Particularly, some studies have found that participants with ASD spent less time looking at the eye region when scanning faces compared to typically developing (TD) counterparts (e.g., Pelphrey et al., 2002; Yi et al., 2013). Tanaka and Sung (2013) proposed an “eye avoidance’’ hypothesis of autism in face processing, which suggested that individuals with ASD avoid the eye region because it is perceived as socially threatening. In line with this hypothesis, individuals with ASD may show stronger tendency to avoid looking at eyes of faces with threatening expressions (e.g., anger) than other expressions.

Objectives: The current study aimed to further examine the “eye avoidance” hypothesis by comparing the eye looking time of children with ASD to that of TD children when viewing faces with different expressions.

Methods: We showed faces with different emotional expressions (joy, anger, sadness, and neutral) to 2- to 5-year-old ASD (n = 28) and age-matched TD (n= 31) children for 5 seconds when their eye movements were recorded using a Tobii T60 eye tracker. Sad facial expressions were included to avoid a possible confound between threat-relatedness and negativity of the displayed emotion. We used temporal course measures to examine the proportional fixation time on the eyes in three continuous phases: the 5-second looking time (300 sample data in total) was divided equally into three phases (early, middle, and late phases, each of which was about 1667 ms) and we wanted to determine whether and when effects of different scanning patterns to the eyes appeared. This analysis allowed us to test whether the patterns observed in the current study are stable or more dynamic across viewing time and thus it may provide a more complete picture into face scanning.

Results: We found that: (1) children with ASD fixated less on the eyes than TD children only for angry and neutral faces, but the two group scanned similarly for happy and sad faces; (2) children with ASD scanned less on the eyes of angry faces than that of happy and sad faces, while TD children scanned more on the eyes of angry and sad faces than that of happy faces; (3) temporal course analysis revealed that: for neutral faces, children with ASD showed an attention avoidance toward eye region than TD children in the early, middle phases but not in the late phase; whereas for angry faces, children with ASD showed more avoidance to the eyes than TD children in all the three phases.

Conclusions: Our study demonstrates that facial expressions moderate eye avoidance pattern in children with ASD when scanning faces. These results suggest that young children with ASD do not have a general eye avoidance pattern in face scanning, rather, this pattern is limited to the threat-related expressions.