Altered resting-state networks in adolescent non-suicidal self-injury—a graph theory analysis

Non-suicidal self-injury (NSSI) is a highly prevalent transdiagnostic symptom and risk marker for mental health problems among adolescents. Research on the neurobiological mechanisms underlying NSSI is needed to clarify the neural correlates associated with the behavior. We examined resting-state functional connectivity in n = 33 female adolescents aged 12–17 years engaging in NSSI, and in n = 29 age-matched healthy controls using graph theory. Mixed linear models were evaluated with the Bayes Factor to determine group differences on global and regional network measures and associations between network measures and clinical characteristics in patients. Adolescents engaging in NSSI demonstrated longer average characteristic path lengths and a smaller number of weighted hubs globally. Regional measures indicated lower efficiency and worse integration in (orbito)frontal regions and higher weighted coreness in the pericalcarine gyrus. In patients, higher orbitofrontal weighted local efficiency was associated with NSSI during the past month while lower pericalcarine nodal efficiency was associated with suicidal thoughts in the past year. Higher right but lower left pericalcarine weighted hubness was associated with more suicide attempts during the past year. Using a graph-based technique to identify functional connectivity networks, this study adds to the growing understanding of the neurobiology of NSSI.