The Association of Cortisol and Testosterone Interaction With Inpatient Violence: Examining the Dual-Hormone Hypothesis in a Psychiatric Setting

Abstract

Psychiatric inpatient aggression is a concern as it poses a threat to safety of both patients and staff. While psychosocial and behavioral approaches are often put forward, the role of biological factors remains underexplored in a clinical context such as psychiatric hospitals. The dual-hormone hypothesis (DHH) posits that low levels of cortisol combined with high levels of testosterone promote status-seeking behaviors with some differences between sexes. This has yet to be studied among psychiatric inpatients. To explore the joint association of the DHH (cortisol and testosterone) and sex with psychiatric inpatient aggression. The sample included 375 psychiatric inpatients (206 women) from the Signature Biobank in Canada. Following their admission in a psychiatric hospital, participants provided hair and saliva for cortisol and testosterone analysis, respectively. Aggressive behaviors from the clinical files were reviewed from admission to discharge. Men with high salivary testosterone combined with low hair cortisol had higher odds of displaying aggression compared to men with high salivary testosterone and high hair cortisol. Men with low salivary testosterone and low hair cortisol had lower odds to perpetrate aggression compared to men with low salivary testosterone and high hair cortisol levels. The cortisol and testosterone interaction was not significant in women. Findings are consistent with the DHH for men. Given that the context hospitalization may trigger status-seeking behaviors, actions could be taken such as identifying specific hormonal profiles at the time of admission to identify patients at risk of aggression, allowing for tailored care protocols.

Keywords: aggression; cortisol; dual‐hormone hypothesis; psychiatry; sex differences; testosterone.

Figure 1

Probability of aggression by levels of salivary testosterone and hair cortisol for each sex. −1 refers to (mean − 1 SD) and +1 to (mean + 1 SD). Standard error bars (vertical bars) indicate how much the population mean is likely to differ from the sample mean. [Color figure can be viewed at wileyonlinelibrary.com]