Meta-Analysis Finds No Link Between Testosterone Levels and Risk-Taking

Testosterone, an androgen produced primarily in the testes of males and in smaller amounts by the ovaries and adrenal glands of females, regulates sexual development, muscle mass, bone density, and other physiological functions. Because men, on average, display higher risk-taking than women, researchers have long hypothesized that testosterone may drive this gender gap. Two prominent biological accounts have been proposed: a direct effect of testosterone on risk preferences and the “dual-hormone hypothesis,” which posits that testosterone’s influence depends on concurrent cortisol levels.

Psypost

A meta-analysis titled “No relationship between testosterone and risk aversion: A meta-analytic review” was published in *Neuroscience and Biobehavioral Reviews* by Irene Sánchez Rodríguez, Luca Bailo, Folco Panizza, Emiliano Ricciardi, and Francesco Bossi. The authors searched Google Scholar, PubMed, and Scopus for human studies using the terms “risk seeking,” “risk attitude,” and “risk aversion.” Inclusion required a reported statistical association between testosterone (measured, administered, or inferred) and a behavioral or self-report risk measure, and sufficient data to compute an effect size. The final dataset comprised 52 studies, representing 17 340 participants, with risk tasks ranging from gambling games and balloon-popping tasks to self-report questionnaires.

Psypost

Aggregating the 52 studies yielded an overall effect size that was “practically zero,” indicating that testosterone levels did not reliably predict risk-taking. Individual study results were highly heterogeneous: some reported positive, others negative, associations. Sub-analyses revealed that only lottery-based economic tasks produced a modest positive link; all other task types showed no association. Moreover, the null relationship held for both males and females, suggesting no sex-specific effect. Studies employing indirect morphological proxies (e.g., digit ratios) sometimes hinted at a link, whereas those using direct hormone assays or experimental administration did not.

Psypost

The authors concluded that the accumulated evidence undermines the view that testosterone serves as a general hormonal basis for human risk preferences. They argue that risk-taking is better understood within a biopsychosocial framework, where task demands, cognitive-affective processes, and situational context interact, and endocrine influences are narrow and context-dependent. The paper also notes a paucity of studies suitable for testing the dual-hormone hypothesis, limiting conclusions about cortisol-testosterone interactions.

Psypost

Prior research has frequently cited testosterone as a driver of risk-related behavior, and the dual-hormone hypothesis remains influential despite limited empirical support. Critics of the meta-analysis might point to the methodological diversity of included studies or argue that the exclusion of non-English, non-Spanish, and non-Italian works could bias the sample. Nonetheless, the authors’ systematic approach and large pooled sample provide a robust counterpoint to earlier single-study claims.

Psypost

The heterogeneity of individual findings—positive, negative, or null associations—highlights unresolved variability in measurement methods and task designs. The analysis could not assess the dual-hormone hypothesis due to an insufficient number of studies reporting concurrent cortisol data, leaving a notable gap in the literature.

Psypost

• “Overall, the evidence challenges the notion that testosterone provides a general hormonal basis for human risk preferences,” — Irene Sánchez Rodríguez, lead author, *Neuroscience and Biobehavioral Reviews*

Psypost

• “Instead, findings support a biopsychosocial framework in which ‘risk taking’ reflects the interaction of task demands, cognitive–affective processes, and situational context, with endocrine effects appearing narrow, context-dependent, and method-specific.” — Irene Sánchez Rodríguez, lead author

Psypost

• “Google News Preferences Add PsyPost to your preferred sources When the researchers aggregated the data, the results showed that the overall association between testosterone levels and risk-taking across all 52 studies was practically zero.” — Irene Sánchez Rodríguez, lead author

Psypost

• “For example, only studies that used lottery-based economic tasks to measure risk-taking showed a modest positive association, while studies measuring risk-taking via other methods (like impulsive games or self-reporting) did not.” — Irene Sánchez Rodríguez, lead author

Psypost

• “Importantly, the lack of an association between testosterone levels and risk-taking did not depend on sex, meaning the relationship (or lack thereof) was no different in males than in females.” — Irene Sánchez Rodríguez, lead author

Psypost

Future investigations should prioritize designs that simultaneously assess testosterone and cortisol to evaluate the dual-hormone hypothesis. More uniform risk-taking paradigms, especially those that isolate economic decision-making, could clarify the modest effect observed in lottery tasks. Expanding the evidence base beyond the current language and publication constraints will further test the robustness of the present findings.

Psypost