Oxytocine en de ontwikkeling van gehechtheid: een blik voorbij de verwachting?

Oxytocin is a hypothalamic neuropeptide that was initially described as a prosocial hormone due to its positive effects on social behavior. Subsequent research, however, demonstrated that its effects are highly context- and person-dependent. More recent theories converge on the view that oxytocin does not directly promote prosociality, but instead modulates underlying mechanisms of both social and non-social behavior, such as social salience and prediction error processing. Prediction error refers to the discrepancy between expected and actual outcomes that drives learning. The current dissertation investigated these mechanisms using the social transmission of food preference (STFP) paradigm, in which rodents learn about food safety by interacting with a conspecific. We conceptualized STFP acquisition as a functional parallel to human epistemic trust, or trust in communicated knowledge that enables humans to use social information to guide subsequent decision-making (Chapter 2). Within this framework, Chapter 3 showed that oxytocin enhanced trust-based learning but blocked updating after trust violation (prediction error) in male mice, but not in females. Chapter 4 extended these findings by demonstrating that oxytocin effects partly depended on dopamine signaling and were particularly evident in males for whom social information was less salient. Building on the proposed parallel between rodent STFP acquisition and human epistemic trust, Chapter 5 synthesized evidence from oxytocin administration studies in both species. Meta-analyses revealed no uniform overall effect of oxytocin in rodents or humans, but substantial heterogeneity across studies. Narrative syntheses indicated that oxytocin can either enhance or reduce trust in social information depending on informant and recipient characteristics, as well as broader contextual factors. These findings are consistent with the allostatic model of oxytocin, which posits that oxytocin modulates information processing, learning, and prediction to support adaptive behavior. Finally, Chapter 6 applied these mechanistic insights to human attachment development by examining whether children's oxytocin response to anticipatory stress moderated the association between parental support and changes in attachment across one year. Results showed that child-reported parental support predicted increases in trust, but this association was stronger for children with a higher oxytocin response to stress. Taken together, this dissertation moves beyond the outdated view of oxytocin as a uniformly prosocial love hormone and advances a mechanistic account grounded in the allostatic model of oxytocin. By integrating the epistemic trust framework with the STFP paradigm, it establishes a translational approach to probe allostatic oxytocin effects and underlying neural mechanisms. This integration provides valuable tools to resolve longstanding paradoxes in oxytocin research and to develop precise accounts of how this ancient neuropeptide supports adaptive behavior. Ultimately, such insights can guide personalized and effective interventions in clinical and developmental contexts.