Multimodal stress assessment: Connecting task-related changes in self-reported stress, salivary biomarkers, heart rate, and facial expressions in the context of the stress response to the Trier Social Stress Test.

Abstract

When we are stressed, do we show it? Although the answer to this question may feel intuitive, prior work on the topic does not provide a clear answer. To address this gap, we present the results of the first study that examined physiological and psychological stress responses and facial expressions using a validated acute stress task and control condition in a within-subjects design. Participants (N = 105; 59 women; 22.36 ± 3.52 years of age) underwent the Trier Social Stress Test (TSST) and the friendly TSST (f-TSST) on consecutive days and provided self-reports via the Short Stress State Questionnaire in German (SSSQ-G), saliva samples, and heart rate. Participants were further filmed during both conditions, allowing us to examine their observable emotional displays using their facial muscle movement data (Action Units, AUs). As expected, the TSST elicited higher SSSQ-G scores and greater cortisol and heart rate increases than the f-TSST. Additionally, the trajectory of cortisol was influenced by the order in which the conditions were presented. The total score, along with the Self-evaluation and Worry subscale scores of the SSSQ-G, correlated with the cortisol maximum increase in response to the TSST, as did heart rate. We found no evidence for displays of common emotions during the manipulation, but we did observe more friendly expressions in the f-TSST (compared to the TSST). Individual AUs neither predicted physiological outcomes nor self-reported stress state scores. Together, these results highlight both the complexity of the stress response in relation to observable emotions and the importance of multimodal stress assessment.