How testosterone administration affects learning to avoid harm in healthy men: A double-blind, placebo-controlled study

Preventing harm to others is a foundational principle of human morality that relies on distinct learning processes when avoiding harm to others versus oneself. Using a behavioral neuroendocrinology perspective, we investigated the effects of a single dose of testosterone administration on harm-avoidance learning. 120 healthy men randomly received either testosterone or placebo. Three hours later, participants completed a harm-avoidance task, learning over time to choose the option with a lower probability of electric shock to avoid harm either for themselves (Self condition) or for a stranger (Other condition). Behavioral analyses revealed that testosterone administration prolonged recipient differences across trials compared to placebo. Computational modeling quantified these recipient differences which were likely caused by testosterone administration. A reinforcement learning model with dual learning rates for positive and negative prediction errors (PEs) best accounted for individuals’ choices. Analysis showed that, compared to placebo, testosterone enhanced learning from negative outcomes but suppressed learning from positive outcomes in the Self condition, without affecting learning rates in the Other condition. Additionally, testosterone administration slowed prosocial learning from negative outcomes compared to placebo and reversed its positive relationship with trait anxiety. Collectively, our findings offer precise computational insights into the role of testosterone administration in harm-avoidance learning. The results indicate that testosterone influences harm-avoidance learning, potentially in a self-protective and strategic prosocial manner. This study aims to enhance the comprehensive understanding of hormonal effects.