Preliminary Findings in Adolescents with ASD: Pupil Diameter As a Proxy for Cognitive Load during Passive Viewing of Facial Expression Stimuli

Background:  Pupillometry, a method for indirectly measuring neural firing within the locus-coeruleus norepinephrine (LC-NE) system, has long been used as a proxy for measuring arousal and visual attention. In recent years, it has been established that changes in pupil dilation are associated with cognitive demand in typical populations. Pupil diameter increases as demands on visual attention increases, and as general cognitive tasks increase in complexity. As cognitive load increases, pupil diameter will increase regardless of accuracy of the participant’s response. Within ASD, it is well known children and adolescents demonstrate diminished attention to faces. It is not clear, however, whether there is a difference in cognitive demand between typically developing individuals and those with ASD while visually processing facial stimuli depicting emotion.

Objectives:  This study examined the pupillary changes in two groups of adolescents, typically developing (TD), and those with ASD without language impairment, upon passive viewing of facial expressions. The aim of the study was to demonstrate differences in pupil diameter between groups, serving as a proxy for measuring the cognitive demand associated with passive visual processing of facial expression stimuli. We hypothesized that the ASD group would demonstrate greater average pupil diameter while passively viewing facial expressions when compared to average pupil diameter within the typically developing group.

Methods:  Participants sat at a stationary eye-tracking station with chin positioned in a chin rest. NimStim photos (Tottenham, 2009) were presented on a computer monitor for 4000 ms each, within 2 randomized blocks. Participants were told to look at the face on the monitor. Binocular infrared eye tracking recorded the changing pupil diameter in response to passive viewing of the facial expression stimuli.

Results:  A significant difference in average pupil diameter was observed between groups. We ran a two-tailed, bi-directional t-test, which suggested the mean pupil size was smaller for the TD group (M=.0043, SD=.00040) than for the ASD group (M=.0052, SD=.00051) (t(38)=6.408, p<.000001). When considered in the context of cognitive demand and pupillary dynamics, this observation suggests the likelihood of greater allocation of cognitive resources by the adolescents in the ASD group during visual processing of facial expressions. The ASD group demonstrated greater average pupil diameter, suggesting the increase in pupil dilation may be associated with greater demands of attentional resources while examining the human face.

Conclusions:  The relative increase in average pupil diameter within the ASD group suggests measurement of changes in pupil diameter may serve as a proxy for cognitive load when viewing visual stimuli. This population may use increased cognitive effort to process common facial expressions and the pupil response provides a physiologic measure of visual processing in relation to cognitive demand. These results have implications for treatment targeting non-verbal language use. These findings also support further analysis using an “index of cognitive ability” algorithm (Bartels & Marshall, 2012) to measure changing pupil diameter as a covariate when examining visual processing of facial expressions within the ASD population.