Evidence that the Biomarker-Behavior Relationship Depends on Negative Evaluation Level: Development and Initial Application of a Scale for Behavioral Engagement During the TSST

Abstract

Behavioral engagement is integral to mental health, but little is known about the relationship of lab-based stress and behavioral engagement. This talk will report on the development and validation of an observer-rated measure of behavioral engagement for lab-based stress inductions, and on the relationship of stress-responsive biomarkers and affect to behavioral engagement during the Trier Social Stress Test (TSST). Young adults (N=109) completed one of three TSST versions—a non-stressful Control, an ambiguously evaluative Intermediate, or the explicitly negative-evaluative Challenge condition. They provided self-reports of positive and negative affect and saliva samples for cortisol and salivary alpha-amylase (sAA) at four timepoints; research assistants rated behavior. Psychometric inspection and EFA of the behavioral engagement items yielded an 8-item measure with good interrater reliability and well-fitting 2-factor structure, capturing Persistence (4 items; loadings=.41-.89), and Quality of Speech (4 items; loadings=.53-.92). Growth curve modeling indicated that as negative evaluation level increased, behavioral engagement became more tightly associated with relative preservation of positive affect, in the context of positive affect decline for most participants across the protocol. For both cortisol and sAA, the relationship between biomarker baseline levels—but not biomarker reactivity—and behavioral engagement varied by condition. Under milder conditions (Control for sAA, Intermediate for cortisol), elevated biomarkers predicted greater engagement, but under explicit negative evaluation, they predicted less engagement, that is, behavioral withdrawal. Findings support the critical role of context—especially negative evaluation—in the relationship of biomarkers to behavioral engagement.