Identifying the Longitudinal Outcomes of Suicide Loss in a Population-Based Cohort

Mental and Physical Health Consequences of Suicide or Accidental Death among First Degree Relatives and Co-Habitants: A Longitudinal Population-based Study PI: Gradus (Boston University) UK PI: Cerel Keywords: Suicide; Bereavement; Epidemiology; Suicide Exposure Suicide represents a significant public health problem – it is among the top 10 causes of death in the US and on a short-list of causes of death that have not decreased in the last few decades. Research focused on persons at risk for suicide behaviors and suicide death has increased in recent years, appropriately commiserate with this growing concern. And yet, the full extent of the suicide public health crisis is underestimated and understudied. This is because there is a relative dearth of research on health outcomes among suicide loss survivors. One suicide can impact up to 135 additional people. There is evidence that loss survivors are at increased risk for a range of adverse mental and physical health outcomes, as well as early mortality. However, much of this literature is limited by a focus on specific familial relations (e.g., spouses), a lack of appropriate reference groups, and a small number of a priori specified outcomes with no differentiation between short- and long-term impact. There are a broad range of negative physical and mental that can occur suicide loss, and yet there is a notable lack of comprehensive epidemiologic information on the population of suicide loss survivors, hindering our ability to understand the full extent of the suicide public health crisis. Recently, the field of suicide epidemiology has benefitted from the use of electronic medical record “big data,” and the application of novel data science techniques, to develop suicide prediction models and further our understanding of suicide risk among individuals. Currently missing from this landscape are studies that apply data-driven methods to understanding the impact of suicide on the short and long-term health of loss survivors. This is likely because there are few sources of longitudinal population-level medical record data that can be used to link suicide decedents to tho. In turn, the absence of easily linked, population-based, longitudinal relationship and healthcare data leaves little ability to use statistical discovery tools (i.e., machine learning) that have been used in suicide prediction research to elucidate the health effects of suicide loss. In a 2015 report, the National Action Alliance for Suicide Prevention highlighted knowledge gaps and strategic directions for this field of research including (1) designing studies of suicide loss survivors using appropriate methods, (2) establishing valid and reliable estimates of the number of people exposed to suicide and the immediate and longer-term impact of exposure and (3) identifying common and unique impacts of suicide bereavement as well as individual difference variables that function as risk factors or buffers to such effects. This project will address these and other critical knowledge gaps by examining thousands of short- and long-term physical and mental health outcomes in a full population of suicide loss survivors, with up to 25 years of follow-up data. This work builds on two previously funded R01 studies (PI: Gradus) of suicide and trauma using Danish national healthcare and social registries to 1. determine the highest risk psychiatric and physical health outcomes of suicide loss, stratified by sex, from among all possible diagnoses in the Danish registries using a novel discovery method called TreeScan; and 2. Use traditional rigorous epidemiologic analyses to quantify the associations and determine if associations are robust to adjustment for confounding.