Pupillary Responses Indicate Working Memory Processing Differences: Implications For Healthy And Clinical Populations

Abstract

Task-evoked pupillary responses (PR) are established psychophysiological measures for neural resource allocation. During working memory tasks, pupils dilate proportional to increasing cognitive demands and constrict when demands exceed resource availability. PRs have demonstrated greater sensitivity to load-dependent processing differences in clinical populations, beyond performance accuracy, under high cognitive demands. Few studies have examined concussion history, sex, and performance accuracy effects on PRs across varying cognitive load levels. PURPOSE: To examine effects of concussion history, sex, and performance accuracy on PR in healthy individuals during a digit-span task. METHODS: Participants self-reported sex (female vs. male) and concussion history (yes vs. no), and completed a backwards digit-span task in a single testing session. A virtual reality headset with 60Hz infrared eye tracking displayed the task and recorded pupil size fluctuations. Pupil size (diameter in mm) was recorded before each trial (baseline=3sec) and following randomly presented digit sequences between 4 and 14 digits long (retention=2sec). PR was calculated as the mean size during retention, normalized to mean baseline. Accuracy was calculated as the proportion of correctly recalled digits by serial position. A mixed effects model examined concussion history, sex, and accuracy effects on PR across sequence-lengths (a priori α=0.05). RESULTS: 40 participants were included [age=21.9±2.1years; males=50%; 17 (43%) with concussion history]. There were significant effects of sex (F1,36=15.66, p<0.001) and accuracy (F1,2588=4.70, p<0.03) on overall relative PR controlling for all other model predictors and interactions. Specifically, females exhibited smaller mean PRs compared to males. Smaller mean PRs were also associated with higher average task accuracy. CONCLUSION: In our study, females exhibited smaller overall pupillary responses during a digit-span working memory task compared to males, indicating potential sex-dependent processing differences. The association between better task accuracy and smaller PRs may further support PR meauress to better inform neurocognitive processing differences in healthy and clinical populations, when demands exceed cognitive resource availability.