A meta-analysis of cortisol reactivity to the Trier Social Stress Test in virtual environments

Background: Maladaptive responses to stressors can lead to poor physical and psychological health outcomes. Laboratory studies of stress induction commonly use the Trier Social Stress Test (TSST). The TSST has been shown to reliably induce a stress response, most commonly measured via cortisol reactivity. Recently, researchers have used virtual environment versions of the TSST (V-TSST) in place of the traditional TSST. The V-TSST has many advantages over the traditional TSST, including increased standardization and use of fewer resources, but V-TSST has yet to be quantitatively reviewed and compared to the traditional TSST. This review aims to quantifying the effectiveness of V-TSST with a meta-analysis of cortisol response effects and identify potential moderating variables that are more likely to induce a cortisol response with V-TSST. Methods: Literature searches were conducted including the key words Trier Social Stress Test, TSST, and virtual reality. Thirteen studies were included in this meta-analysis after meeting the inclusion criteria of utilizing a V-TSST and having cortisol measurements at baseline and peak stress to assess cortisol reactivity. The standardized mean gain effect size was used. Results and discussion: There was a medium average effect size (ES_sg = 0.65) across all studies for increase in cortisol from baseline to peak measurement. Significant moderating effects were seen for participant age, sex, and level of immersivity of the virtual environment. Studies in which participants were under 25 years old, or all male, showed greater effect sizes for cortisol reactivity. Virtual environments that were more immersive also evidenced greater effect sizes. Although the V-TSST is effective at inducing psychosocial stress, the magnitude of this response is less than the traditional TSST. Based on these results, recommendations for future research are provided.