Is there a female-male self-selection bias in TSST-based reactive stress research?

A selection bias occurs when a given sample of participants only represents a subset of the population under study, which may subsequently limit the generalizability of findings. While previous studies have noticed a potential female-male selection bias in human stress research, with female participants often being over-represented, no prior research has directly addressed this issue in the context of stress reactivity. This exploratory study aimed to systematically examine this observation. A total of 120 scientific articles (N = 10,103) published from 2014 to 2023 on the topic of human stress reactivity retrieved from PUBMED and PsycINFO were examined to compile sex ratios by study location (United States, Germany, China, Canada, Israel, United Kingdom). The meta-analysis and meta-regression results indicated that females participate in reactive stress studies more frequently than males, although the observed difference is small. Moreover, there is no significant discrepancy regarding male and female participation rates between the countries examined. This result supports a higher female representation level in stress research samples. The findings provide leads for future studies aiming to further investigate the underlying antecedents of selection bias in human stress research. A better understanding of the phenomenon could lead researchers to optimize recruitment methods to obtain more representative samples. • Key Finding: This meta-analysis demonstrates a small but consistent overrepresentation of women in human stress reactivity studies; • Methodology: We analyzed 166 studies, encompassing 14,324 participants across six countries, using meta-analysis and meta-regression techniques to evaluate sex ratios; • Implication: The observed male-female selection bias suggests addressing recruitment methods to reduce sex bias and ensure more representative participant samples in stress research.