By their own young hand: Non-suicidal self-harm and suicide attempts in young people with psychotic experiences

Abstract

Background: Non-suicidal self-harm (NSSH) and suicide attempts, which for this thesis are grouped under the umbrella term of ‘self-harm’, account for a considerable portion of disability burden and are robust predictors of suicide death in young people. Psychotic experiences (PEs), such as hallucinatory and delusional experiences, have become increasingly of interest as potential markers of youth self-harm, although it remains unclear as to why there is an association. The purpose of this thesis was to investigate the PE–self harm association in adolescents (aged 12–17 years), as well as bio-psychosocial mechanisms underlying the relationship. This involved investigating (1) the role of third variables (confounders, mediators) in the PE–self harm association; (2) how the PE–self harm association fits in with self-injurious and suicidal behaviour theories; (3) the direction and dynamic nature of the relationship; and (4) whether the association varies by PE subtype.Method: A systematic review of the literature was undertaken, which critically examined the role of confounders and mediators in the PE–self harm association. Relevant variables, as identified in the review, were included and examined in an epidemiological analysis of PEs and self-harm, using a nationally representative sample of Australian adolescents (the Young Minds Matter survey). Structural equation modelling was then conducted to investigate proposed mediating pathways of the association, across three waves of prospective adolescent cohort data (the HEALing study). Using the same Australian adolescent cohort study, the role of the Interpersonal Theory of Suicide (IPTS), a well-established suicidal behaviour theory, was examined; where hallucinatory experiences and IPTS constructs were explored in the suicidal thought-to-attempt transition. This was conducted to address gaps in previous epidemiological studies which have been largely atheoretical and data driven. Alternative pathways which challenge broad assumptions about directionality were also investigated. Multivariable logistic regression analyses were used to examine potential reverse causal pathways, including relationships between baseline self-harm and incident PEs, across three waves of prospective adolescent data (the HEALing study).Results: The systematic review demonstrated the PE–self harm association has been well-replicated, but that almost 30% of all studies did not control or adjust for confounders or other third variables of interest. In the majority of studies where adjustment did occur, the association persisted, but its strength was attenuated. Common mental disorders such as depression, as well as various psychosocial factors (including psychological distress and negative environmental exposures such as bullying) explained a substantial amount of the variance, whereas most demographic and family history variables were of less importance. Nationally representative analyses found PEs are common in Australian adolescents (3.3–14.0% 12-month prevalence) and that their clinical relevance varies by subtype. Auditory hallucinatory experiences, in particular, were strongly and consistently associated with NSSH (OR=2.44; 95%CI=1.48–4.03) and suicide attempts (OR=3.35; 95%CI=1.61–6.97), after adjustment and consideration for an extensive set of bio-psychosocial confounders and mediators. Variables such as depression, psychological distress, self-esteem, and being bullied were identified as potential key explanatory variables and were included in subsequent longitudinal mediation analyses. Auditory hallucinatory experiences were found to be indirectly associated with future NSSH and suicide attempts via recent traumatic life events (e.g. bullying, sexual assault), high psychological distress, and low self-esteem, across three waves of prospective cohort data. Other PE subtypes were mostly not associated with incident NSSH and suicide attempts at 1- and 2-year follow-ups, either directly or indirectly. Regarding theoretical frameworks, the IPTS theoretical model provided limited explanatory value to the PE–self harm association (albeit statistical power and measurement concerns). Alternative pathways of reverse causality were also considered, however, there was little evidence of an association between NSSH, suicide attempts, and subsequent PEs.Conclusion: PEs are common in adolescents, even among young people who do not have a mental health condition. These experiences are indicative of increased risk across a wide range of poor mental health and psychosocial outcomes, including NSSH and suicide attempts. Overall, the collection of evidence presented in this thesis suggested the PE–self harm association is not direct but instead explained by shared risk factors, including recent negative life events and trauma, as well as affective and negative self-evaluation processes. In addition to identifying important explanatory and modifiable factors, this thesis highlighted that “not all PEs are created equal”, where the phenomenological qualities of PEs are key considerations to understanding and predicting self-harm and suicide risk in this group. Adolescents reporting persistent auditory hallucinatory experiences, over several years, are at particularly high risk for engaging in NSSH and suicide attempts. Despite measurement and sampling limitations, this thesis has provided a strong foundation for future research and can be used to guide ongoing self-harm/suicide prevention and early intervention efforts among young people with PEs. Such efforts would likely benefit most from targeting mediating and modifiable factors of acute distress, low self-esteem, and exposure to traumatic events with interpersonal harm.