Functional connectivity between the amygdala and the medial prefrontal cortex (mPFC) has been identified as a neural substrate of emotion regulation that undergoes changes throughout development, with a mature profile typically emerging at 10 years of age. Maternal bonding in childhood has been shown to buffer amygdala reactivity and to influence the trajectory of amygdala–mPFC coupling. The oxytocinergic system is critical in the development of social behavior and maternal bonding. Early-life parental care influences the methylation status of the oxytocin receptor (OXTRm) in animal models and humans, and higher OXTRm is associated with lower amygdala–PFC functional connectivity in adults. Using a neuroimaging-epigenetic approach, we investigated saliva-derived OXTRm as a biological marker of structural and functional connectivity maturation in 57 typically developing children (P < 0.05). We utilized seed-based connectivity analysis during a novel abstract movie paradigm and find that higher levels of OXTRm are associated with a more adult-like functional connectivity profile. Concurrently, more adult-like functional connectivity was associated with higher reported self-control and more diffusion streamlines between the amygdala and mPFC. OXTRm mediates the association between structural and functional connectivity with higher levels of OXTRm being associated with more streamlines. Lastly, we also find that lower OXTRm blunts the association between amygdala–mPFC connectivity and future internalizing behaviors in early adolescence. These findings implicate OXTRm as a biological marker at the interface of the social environment and amygdala–mPFC connectivity in emotional and behavioral regulation. Ultimately, identification of neurobiological markers may lead to earlier detection of children at risk for socio-emotional dysfunction.

Social attention involves selectively attending to and encoding socially relevant information. We investigated the neural systems underlying the wide range of variability in both social attention ability and social experience in a neurotypical sample. Participants performed a selective social attention task while undergoing fMRI and completed self-report measures of social functioning. Using connectome-based predictive modeling, we demonstrated that individual differences in whole-brain functional connectivity patterns during selective attention to faces predicted task performance. Individuals with more cerebellar-occipital connectivity performed better on the social attention task, suggesting more efficient social information processing. Then, we estimated latent communities of autistic and socially anxious traits using exploratory graph analysis to decompose heterogeneity in social functioning between individuals. Connectivity strength within the identified social attention network was associated with social skills, such that more temporal-parietal connectivity predicted fewer challenges with social communication and interaction. These findings demonstrate that individual differences in functional connectivity strength during a selective social attention task are related to varying levels of self-reported social skill.