Meta-analytical assessment of the effects of protocol variations on cortisol responses to the Trier Social Stress Test

Background: The Trier Social Stress Test (TSST) is one of the most widely used laboratory stress tests. Exposure to this psychosocial stressor has been shown to stimulate an acute cortisol stress response in the majority of healthy individuals, while deviations from the typical pattern, i.e., cortisol reactivity dysfunctions have been linked to an ever-increasing number of negative health outcomes. However, significant variability between labs exists in strength of observed cortisol responses in healthy individuals. This variability raises the question of how to distinguish across labs between cortisol stress response patterns that reflect health risk from those that are due to methodological differences. Thus, we propose a systematic review and meta-analysis that aims at quantifying the effects of methodological variation in study and TSST protocol elements on cortisol stress responses in healthy individuals. Methods: Literature searches were conducted using standard databases for English language with key words including Trier Social Stress Test, TSST, Cortisol, and Laboratory Stressor among others. 186 studies met our inclusion criteria of healthy human participants without systemic immunological or endocrine dysfunction and provided sufficient information to compute a total of 237 sub-sample effect sizes. Results and discussion: With regard to study protocol variations that may risk confounding baseline cortisol values and thus influence subsequent reactivity measures, meta-analytical examination revealed that acclimation periods pre-TSST below 30 or perhaps even 15 min may suffice, at least as long as no interfering activities, i.e., questionnaires, are taking place during that timeframe. Assessing the effects of TSST protocol variations on cortisol response strength, several observations are noteworthy. First, shortening speech preparation time did not change cortisol responses in any way, nor did including questionnaires during that period show an effect. As such, our findings suggest that speech preparation time is one TSST element that can be used to reduce the burden for participants as well as laboratory logistics. Secondly, having an all female panel and instructing panel members to show negative instead of neutral behavior towards the participants both were associated with considerably reduced cortisol stress response strengths. Thirdly, several variables of interest, such as content of the speech task or gender match between active panel member and participant, were problematic to evaluate due to the large number of studies not reporting those details. This calls for future studies to report more details regarding potentially relevant protocol specifications. (C) 2017 Elsevier Ltd. All rights reserved.