Others’ emotions teach, but not in autism: an eye-tracking pupillometry study

To examine this process, we used a novel eye-tracking pupillometry paradigm in which we showed 20 preschoolers with autism and 20 matched typically developing preschoolers’ videos of an actor opening a box and reacting to the occluded object inside, with fear or happiness. We expected preschoolers to come to perceive the box as containing a positive or threatening stimulus through emotionally calibrating to the actor’s emotional expressions. Children’s mean pupil diameter (indicating emotional reactivity) was measured whilst viewing an up-close, visually identical image of the box before and then after the scene, and this difference was taken as an index of social-emotional calibration and compared between groups.

Whilst the typically developing preschoolers responded more emotionally to the box after, compared to before the scene (as indexed by an increase in pupil size), those with autism did not, suggesting their reaction to the object was not affected by the actor’s emotional expressions. The groups did not differ in looking duration to the emotional expressions; thus, the pupil dilation findings cannot be explained by differences in visual attention. More social-emotional calibration on the happy condition was associated with less severe autism symptoms.

Through the measurement of physiological reactivity, findings suggest social-emotional calibration is diminished in children with autism, with calibration to others’ positive emotions as particularly important. This study highlights a possible mechanism by which individuals with autism develop idiosyncratic reactions to features of their environment, which is likely to impact their active and harmonious participation on social and cultural practices from infancy, throughout the lifespan. More research is needed to examine the mediators and developmental sequence of this tendency to emotionally calibrate to others’ feelings about the world.