Unpacking cortisol stress reactivity: Associations with psychological state and trait data in a pooled data sample

Prior research suggests that cortisol responses to acute stress induction paradigms (i.e., "cortisol stress reactivity") likely arise from complex interactions between experiment-level, biological, and psychological factors. Here, we investigated the extent to which cortisol stress reactivity can be explained by psychological state and trait data alone when biological and experiment-level factors are strictly controlled. In a pooled dataset comprising five studies (total n = 443), healthy participants with no history of mental disorders were exposed to either the Maastricht Acute Stress Test experimental (MAST_EXP; n = 297) or the no-stress control (MAST_PLC; n = 146) condition. We calculated common cortisol stress reactivity metrics (peak reactivity, total cortisol turnover, and two binary responder classifications) and investigated associations with stress-induced changes in psychological state (momentary affect; PANAS), psychological traits (approach and avoidance tendencies; BIS/BAS), vital signs, and other contextual variables using LASSO generalized linear and logistic regression. Stress-induced changes in momentary states, particularly decreases in positive affect, were modestly associated with cortisol reactivity metrics, although the robustness of these associations was dependent on the metric employed. In contrast, negative affect was significantly associated with only one cortisol reactivity metric, while trait-level approach and avoidance tendencies showed no significant associations. Our results underscore the subtle nature of cortisol stress reactivity-psychological state/ trait data associations, lending further credence to the idea that cortisol stress reactivity, in reality, may likely arise from small contributions across diverse (experimental, biological, psychological) variables, rather than a single dominant psychological factor.