Deliberate self-harm : characteristics, clinical correlates and interventions

Abstract

Background: Deliberate self-harm (DSH) and nonsuicidal self-injury (NSSI) are serious global health problems prevalent among both community and clinical samples. Several studies have suggested that aggression towards others may be a clinically relevant correlate in self-harming individuals, but the association is still unclear. Current research suggests that NSSI primarily serves an emotion regulating function, but there are few treatments directly targeting NSSI available. Thus, there is currently a need for research on the characteristics, clinical correlates and treatments for DSH and NSSI.

Aims: The general aim of this thesis was to study DSH and NSSI and its association with violence towards others and to evaluate the utility of a short adjunctive treatment for NSSI. The specific aims of the individual studies were: • to compare suicidal severity and use of interpersonal violence in a group of suicide attempters with and without a history of NSSI (Study I). • to investigate the association between DSH and perpetration of violence towards others (Study II). • to evaluate the utility of a short, group-based behavioural treatment targeting emotion dysregulation and NSSI (Study III). • to examine what patient characteristics predicted outcomes in the treatment evaluated in Study III (Study IV)

Methods: A cross sectional study was conducted examining the impact of comorbid NSSI in a group of suicide attempters (Study I). The unique association between DSH and violent crime was investigated through Cox proportional hazards regressions in a survival analysis based on register data (Study II). The utility and feasibility of a short, adjunctive treatment for NSSI (emotion regulation group therapy, ERGT) was evaluated in an open, uncontrolled pilot study. Within-group effects and interaction effects of predictors on outcome at post treatment and six-month follow-up was analysed using multilevel generalized estimated equations negative binomial and linear mixed regression models (Study III and IV).

Results: Suicide attempters with comorbid NSSI engaged in more violent and frequent suicide attempts, and had used interpersonal violence towards others to a significantly higher extent than suicide attempters without NSSI. A significant and unique association between DSH and violent crime was found suggesting a shared vulnerability between violence towards oneself and others. ERGT was associated with significant reductions in NSSI, emotion dysregulation and other psychiatric symptoms at post-treatment and six-month follow-up. Analysing predictors of outcome revealed that ERGT may also be useful for individuals with high frequency of NSSI, and that comorbidity was negatively associated with the maintenance of treatment gains.

Conclusions: Engagement in deliberate self-harm and violence towards others are associated, although what characterizes this shared vulnerability to violence against one self and others is yet unclear. ERGT may be a useful and transportable treatment for individuals with NSSI.