Despite the sometimes heated debate about the validity of human oxytocin studies, experimental oxytocin research with intranasal administration is a growing field with promising preliminary findings. The effects of intranasally administered oxytocin compared to placebo on brain neural activity have been supported in animal studies and in human studies of neural resting state. In several studies, oxytocin sniffs have been shown to lead to down-regulation of amygdala activation in response to infant attachment vocalisations. Meta-analytic evidence shows that oxytocin enhances the salience of (emotional) stimuli, lowers stress and arousal, and elevates empathic concern and tender care, in particular for offspring and in-group members. Less firm evidence points at the amnestic effects of oxytocin. We also note that the average effect sizes of oxytocin experiments are small to modest, and that most studies include a small number of subjects and thus are seriously underpowered, which implies a high risk for publication bias and nonreplicability. Nevertheless, we argue that the power of within-subjects experiments with oxytocin has been underestimated. Much more work is needed, however, to create a firm knowledge base of the neural and behavioural effects of oxytocin. Human oxytocin research is still taking place in the context of discovery, in which bold conjectures are being generated. In the context of justification, these conjectures should subsequently be subjected to stringent attempts at refutations before we jump to theoretical or clinical conclusions. For this context of justification, we propose a multisite multiple replications project on the social stimuli salience enhancing effect of oxytocin. Clinical application of oxytocin is premature. Meta-analytically, the use of oxytocin in clinical groups tends to show only effectiveness in changing symptomatology in individuals with autism spectrum disorders but, even then, it is not yet a validated therapy and its use is premature because safety and long-term side-effects have not been sufficiently studied, in particular in children.
- multisite multiple replications
- statistical power