Downcast Gaze and Hypersensitivity to Direct Gaze in Young Children with ASD

Background:

The use of eccentric viewing has been found to be common in ASD and it is likely to allow the filtering of details during visual exploration (Mottron et al., 2007; Noris et al., 2012). Specifically, a subset of eccentric glances (EG), downcast gazes (DG), is believed to occur in order to reduce sensory overload (Bogdashina, 2003; Simmons et al., 2009). In Noris et al. (2012) children with ASD during a social interaction were more likely to look downwards compared to matched TD children. These findings left open the question whether DG) are linked to the hypersensitivity to high-frequency visual information or they result from a hypersensitivity to social stimuli. Indeed some individuals with ASD are likely to show gaze aversion (Tottenham et al., 2013; Dalton et al., 2005) and atypical face scanning (Chawarska & Shic, 2009).

Objectives:  

Our study involving a sample of young children with ASD investigated  whether the patterns of attention to pictures of human faces with direct gaze, recorded by means of an eye-tracker, vary in relation to the frequency of EG during a social interaction episode with an experimenter.

Methods:  

22 preschoolers with ASD mean age  4.8 years (3.7 to 6.4 years, SD= 10 months), mean non verbal IQ 111 (Leiter-VR range 90-133, SD = 13.5) were included in the study. Subjects were diagnosed by expert clinicians according to DSM-IV criteria for ASD and standardized diagnostic instruments (ADOS and ADI-R).
The children were shown a series of 6 images of neutral faces with direct gaze. Each image lasted 5 seconds on the screen of the eye-tracker Tobii T60, frequency and duration of fixations on eyes, mouth and the whole face were recorded and computed as a ratio to gaze-time on screen. Participants were video-recorded during an interaction with an adult administering selected items from the ESCS (Mundy et al., 2003). Mean frequency of EG per minute was coded from the videos and the median of this measure was used to divide the sample in two groups: low frequency (n=11, group LFreq), high frequency (n=11, group HFreq).

Results:  

Children in the HFreq group produced a significantly higher mean frequency of DG during the ESCS interaction (HFreq=3.07 [SD=1.33]; LFreq=1.63 [SD=1.08]; t=2.765, df=20, p=0.012). Independent sample T tests revealed that children in the HFreq group showed significantly reduced attention to the eyes area of the pictures, in terms of frequency (HFreq=0.43 [SD=0.15]; LFreq=0.60 [SD=1.12]; t=2.975, df=20, p=0.007) and duration (HFreq=0.48 [SD=0.21]; LFreq=0.68 [SD=1.15]; t=2.539, df=20, p=0.02) of fixations, while no significant difference resulted for the mouth area, or for the whole face in itself.

Conclusions:  

Our results seem to suggest a relationship between DG and hypersensitivity to direct gaze. Children with ASD might look downwards in order to regulate excessive arousal caused by the direct eye gaze of a social partner.