Unveiling Convergent and Divergent Intrinsic Brain Network Alternations in Depressed Adolescents Engaged Non-Suicidal Self-Injurious Behaviors with and without Suicidal Attempts

Abstract Nonsuicidal self-injury (NSSI) and suicidal attempt (SA) frequently occur during depressed adolescents, but the underlying neurobiological mechanisms are poorly understood. The maturation of network connectivity is a central feature of brain development during adolescence. However, few researchers have used resting-state functional magnetic resonance imaging (fMRI) to investigate the similarities and differences in the intrinsic brain networks between NSSI with NSSI+SA in depressed adolescent populations. In this study, we collected self-reported questionnaires and resting-state (fMRI data from a sample of depressed adolescents (N = 114), categorized into three groups: clinical control (non-self-harm), NSSI-only, and NSSI+SA based on their self-harm history. The alternations of FC patterns were identified through support vector machine-based classification, as machine learning approaches can help to overcome the multiple-comparison problems of their multivariate nature. Referring to the convergent alternations between adolescent NSSI with or without a history of SA, the alternations of intra-/inter-network FCs among the Control, SalVAtten, Limbic, and Default networks. Besides, divergent alternations were also observed for depressed adolescents with a history of SA, involving the Limbic, DorsAttn, Visual, and Subcortical networks. Besides, the severity of depressive symptoms only showed a significant correlation with altered FCs in Limbic-DorAttn and Limbic-Visual networks, strengthening the fact that these observed alternations of FC could not all be explained by increased depression severity. The emphasis of this study on the network basis could extend the existing evidence at a circuit level and the observed convergent alternations may explain an underlying neurobiological mechanism pertaining to the continuum of self-destructiveness in adolescents. In addition, our discovery of divergent alternations could help to identify biomarkers that will aid in differentiating those at risk for SA among those engaging in NSSI and contribute to the development of neurobiologically informed interventions.