Predictors and Correlates of Adolescent Non-Suicidal Self-Injury

Non-suicidal self-injury (NSSI) is a dangerous and common behaviour, particularly among adolescents. Childhood trauma, insecure child-parent attachment, psychological distress, and impulsivity are some of the risk factors for NSSI that have been previously identified. However, the pathways from distal risk factors to NSSI and the ways in which these correlated risk factors interact with each other remains unclear. Identifying these pathways will provide valuable insight into the aetiology of NSSI and potentially highlight targets for treatment and intervention. In this dissertation I examine data from multiple large samples of young people, looking at multiple risk and protective factors together, and examining moderation and mediation pathways between risk factors. Using longitudinal data from 933 adolescents with no prior history of NSSI I demonstrated that the association between childhood family adversity before age 5 and new onset of NSSI between the ages of 14 and 17 was mediated by age 14 family functioning and possibly mental illness. Next, I validated a new measure of child perceptions of positive parenting, which I used to demonstrate the uni-directional prospective association between positive parenting and lower rates of NSSI amongst 1489 adolescents (ages 14-25). I then used this new measure of positive parenting to demonstrate that the prospective parenting-NSSI association was mediated by psychological distress. This is also one of the first prospective studies to show that impulsivity is independently predictive of NSSI. Using data I collected myself from a sample of 596 adolescents (ages 16-19) I validated a much needed measure of childhood trauma, with which I then demonstrated that the trauma-NSSI association was mediated by attachment and distress. Using data from this sample I was also able to reaffirm my previous findings that the attachment-NSSI association was mediated by psychological distress, and that impulsivity was uniquely associated with NSSI. Finally, using data from a sample of 559 Flemish 13 year-olds, I demonstrated that behavioural problems were more salient to NSSI than emotional problems among young adolescents, and that the attachment-NSSI association might be mediated by hyperactivity and conduct problems. Together, these findings reaffirm that childhood trauma, insecure child-parent attachment, psychological distress, and impulsivity are robust risk factors for NSSI and potential targets for treatment and intervention. Moreover, both distress and child-parent attachment may be viable targets for interventions aimed at attenuating the impact of early childhood trauma after it has occurred. Future research should use randomised controlled trails to test the efficacy of NSSI treatments aimed at these risk factors.Matthew Cassels' doctoral studies were funded by a Gates Cambridge scholarship