CE25-029 - Pathways Between Child Maltreatment and Self-Directed Violence: A Longitudinal, Population-Based Study Using Machine Learning Approaches

This K01 award application is for Dr. Lindsey Palmer, a PhD-trained social worker whose overarching career goal is to become an independent violence prevention scientist, focused on promoting child health and well- being by advancing data-driven, evidence-based strategies to prevent maltreatment and its long-term consequences. This K01 will support three key areas of career development: 1) the application of machine learning approaches on violence prevention research, 2) cross-cutting violence prevention strategies, and 3) professional development and leadership. Dr. Palmer has assembled an interdisciplinary mentoring team comprised of Kristine Campbell, MD, MSc, a nationally recognized expert in pediatric child maltreatment with extensive experience collaborating with public agencies to develop cross-system prevention efforts; Fernando Wilson, PhD, an expert in the application of machine learning techniques on large-scale databases to examine health services and policy; Brooks Keeshin, MD, an internationally recognized expert in trauma assessment and suicide prevention; and Angela Fagerlin, PhD an expert in faculty enhancement, leadership and representation. Over the past decade, rates of self-directed violence (SDV) have risen sharply, particularly among 10- to 17-year-olds, with children and adolescents who have experienced maltreatment being at particularly heightened risk. A staggering 57% of children and adolescents who die by SDV have a history of alleged child maltreatment, which encompasses physical abuse, sexual abuse, emotional abuse, physical neglect, and exposure to intimate partner violence. These youths often face the compounded challenges of trauma, family dysfunction, and mental health issues. While child welfare system (CWS) involvement frequently signals heightened vulnerability, the pathways linking child maltreatment to SDV remain poorly understood. Contributing factors such as parental mental illness, substance use, overlapping forms of maltreatment, family instability are not well defined or understood. Additionally, there is limited evidence on the effectiveness of CWS interventions in reducing the risk of SDV for these children. This study’s Specific Aims include: 1) Determine the relationship between child maltreatment and SDV, specifically: Establish how the timing, type, and frequency of child maltreatment indicators are associated with SDV; and characterize the association between child maltreatment intervention and SDV; and 2) Leverage machine learning based approaches to identify direct and indirect pathways between child maltreatment and SDV, focusing on the progression of suicidal thoughts and behaviors over time. This study is significant and innovative because it will clarify the relationship between child maltreatment and SDV, identify high-risk subgroups, and examine if existing CWS interventions mitigate or exacerbate SDV risk, providing critical insights into the strengths and limitations of current maltreatment practices in reducing other forms of violence.