Looking at the Ceiling: Eye-Tracking Data Indicate a Restricted Range of Reactivity to Threat in Autism

Background: High levels of anxiety in autism have been linked to a number of mechanisms including the idea of reduced emotional awareness described as “alexithymia”. In a previous fMRI study of classical fear conditioning, we found less differentiation of threat and safe contexts in autism. We suggested one possible reason for this could be a “ceiling effect” in which autistic participants were already so physiological elevated (whether from everyday life or the context of the experiment, or both) that further reactivity to our threat task is limited. For this experiment, a similar fMRI study of fear conditioning with a less variable reinforcement schedule, we additionally collected eye tracking data as fast (millisecond response, as opposed to a multiple second time course for fMRI BOLD data) measure of threat reactivity, and we also added a non-task baseline condition to examine initial tonic pupil size.

Objectives: We explored non-task “baseline” data for evidence of hyperarousal in autism before the onset of our fear conditioning experiment, and subsequent changes in pupil size during the experiment. We explicitly explored the possibility of a “ceiling effect” that influences the possible range of response in autism.

Methods: Sixty young adults (28 autism, 32 neurotypical) with average- to above-average IQ completed a classical fear conditioning task that used visual cues as conditioned (threat versus safe) cues and a burst of air on the neck as the unconditioned threat stimulus. Using Eyelink 1000 Plus eye-tracking system we collected tonic pupil size while participants stared at a fixation cross in the center of the screen for 60 seconds, and change in pupil size following the onset of each cue (threat or safe) during the fear conditioning task.

Results: The autism group demonstrated significantly elevated tonic pupil size compared to matched neurotypical participants, before the test even started. During the fear conditioning task, the autism group demonstrated a restricted range of response, including decreased responses to both safe and threat conditions and a decreased difference between safe and threat in the autism group.

Conclusions: While there are several possible explanations for this reduced reactivity during the fear conditioning task, we suggest that a “ceiling effect” arising from the elevated baseline arousal is parsimonious. The experimental threat of the air burst to the neck was not enough in thautism group to overcome existing arousal, which could be due to pressure or anxiety from the MRI environment and/or from everyday arousal. this latter possibility suggest the need to study physiological activity and response over the course of days or weeks using ecologically-valid methods. In the meantime, treatment for anxiety in autism should acknowledge difficulties latching on to “safe” contexts and possibly an elevated everyday physiological arousal which could interfere with relationships, work/school performance, and therapy.