Neural Reactivity to Social Punishment Predicts Future Engagement in Nonsuicidal Self-injury Among Peer-Rejected Adolescents

Abstract

BackgroundRates of nonsuicidal self-injury (NSSI) increase dramatically in adolescence. Affective reactivity and adverse social experiences have been linked to NSSI, but less is known about whether these factors may separately or interactively predict NSSI, especially longitudinally. This study combined functional magnetic resonance imaging and a sociometric measure to test whether a combination of neural (e.g., amygdala) reactivity to social punishment and peer-nominated peer acceptance/rejection predicts NSSI longitudinally in adolescence. Amygdala reactivity was examined as a potential neural marker of affective reactivity to social punishment, which may heighten NSSI risk in contexts of social adversity. MethodsOne hundred twenty-five adolescents (63 female) completed a social incentive delay task during neuroimaging and school-based peer nominations to measure peer acceptance/rejection. NSSI engagement was assessed at baseline and 1-year follow-up. ResultsGreater amygdala reactivity to social punishment predicted greater NSSI engagement 1 year later among adolescents with high peer rejection. This effect for the amygdala was specific to social punishment (vs. reward) and held when controlling for biological sex and pubertal development. Exploratory analyses found that ventral striatum reactivity to social reward and punishment similarly interacted with peer rejection to predict NSSI but that amygdala connectivity with salience network regions did not. ConclusionsAmygdala reactivity to social punishment, in combination with high peer rejection, may increase NSSI risk in adolescence, possibly via heightened affective reactivity to adverse social experiences. Objective measures of neurobiological and social risk factors may improve prediction of NSSI, while therapeutic approaches that target affective reactivity and increase prosocial skills may protect against NSSI in adolescence.