Testosterone to cortisol ratio and aggression toward one’s partner: Evidence for moderation by provocation

•Greater testosterone to cortisol ratio predicted greater aggression towards one’s partner.•The association between T/C ratio and aggression was weaker when aggression was provoked.•Predictions of the dual-hormone hypothesis may be limited to unprovoked aggression. Consistent with the dual-hormone hypothesis, the combination of high testosterone levels and low cortisol levels has been linked to increased dominant and aggressive behaviors. However, recent research indicates that this association is weaker or even reversed following provocation. It is also unclear whether the association between testosterone/cortisol and aggression is similar for men and women and for those with and without attention-deficit/hyperactivity disorder (ADHD). Using data from a larger project examining ADHD in the context of romantic relationships, the current study tested the dual-hormone hypothesis in 32 heterosexual young adult couples who engaged in a conflict discussion and a competitive reaction time task in the laboratory. Aggressive behavior was indexed by greater noise blast intensity toward one’s partner during the competitive reaction time task. Two potential sources of provocation were examined: 1) affective responses to a conflict discussion task preceding the competitive reaction time task, and 2) whether participants received/did not receive a noise blast before the first two trials of the competitive reaction time task. Salivary testosterone and cortisol were assessed three times throughout the laboratory session, and the ratio of testosterone to cortisol output across the session (T/C AUCg ratio) was calculated. Consistent with the dual-hormone hypothesis, greater AUCg T/C ratios were associated with greater aggression. Further, T/C ratio–aggression associations were weaker under provoked conditions but did not differ as a function of sex or ADHD status. Results provide support for the dual-hormone hypothesis and suggest that provocation may be an important moderator of the T/C–aggression relationship.