The Etiology of Suicidal and Non-Suicidal Self-Injury in a Population-Based Sample

The overarching research aim of the current study is to identify factors that are differentially associated with risk for suicide attempt versus non-suicidal self-injury (NSSI). NSSI has been identified as a risk factor for suicide attempts, though its predictive utility is inconsistent. Previous research has frequently approached the relationship between NSSI and suicide attempt as a pathway, in which NSSI facilitates progression to suicide attempts; or as a continuum, in which risk factors are shared, with NSSI representing a less severe manifestation of self-harm than suicide attempts. However, there is evidence that supports the hypothesis that NSSI and suicide attempts, while phenotypically correlated, have at least partially distinct etiologies. 
In the current study, we will test this hypothesis using data from UK Biobank, a large, population-based cohort for which data are available on NSSI, suicide attempts, and a vast array of factors previously associated with NSSI and suicide attempt, including depression, substance use, demographic factors, and childhood maltreatment. Critically, prior studies support a moderate to strong genetic component underlying suicidal behavior and NSSI. However, efforts have largely focused on genetic liability shared with depression, which offers only an incomplete portrait of genetic risk for these outcomes. Genetic data are available for UK Biobank participants, which we will use to generate polygenic risk scores to investigate differential associations between outcomes (suicide attempt versus NSSI). To clarify the contributions of genetic factors to etiologic differences in suicide attempt versus NSSI, we will derive polygenic risk scores for UK Biobank participants for phenotypes as wide-ranging as alcohol problems, cognitive function, major depression, and impulsivity. We will use a series of regression models to determine which predictors — phenotypic and genetic — are differentially associated with suicide attempts and NSSI, thereby improving our understanding of the etiology of both outcomes.

The proposed analyses stand to impact the field in numerous capacities. First and foremost, we will approach suicide attempt and NSSI as unique outcomes rather than imposing a pathway/continuum model, with the goal of identifying factors that distinguish these outcomes from one another. Second, we will employ data from a large, population-based cohort (N=157,359 individuals completed a mental health questionnaire that included items about self-harm), in contrast to many prior studies that utilized selected samples. This sample consists of older adults (aged 40-69 upon joining the study) rather than the adolescent or young adult samples used in much of the extant literature. Finally, we will substantively contribute to the understanding of genetic factors underlying risk for suicide attempt and NSSI by expanding beyond a focus on genetic liability to depression. We will ensure scientific rigor through the use of internal and external replication samples. Findings from this proposal can be used to guide risk assessment and the design of targeted genetic analyses.