Childhood Adversity and Delinquency Development in Youth: Single Adversity Approach, Latent Class Analysis and Cumulative Impacts

Authors: Yuhui Liu, Lydia Speyer, Jasmin Wertz, Ingrid Obsuth

Abstract:

Childhood adversities involve situations that threaten a child's basic sense of safety and stability, necessitating significant adaptation. This construct is multifaceted, requiring consideration of different conceptual dimensions, including type, accumulation, chronicity, and timing. Existing fields, such as poly-victimization and adverse childhood experiences, primarily focus on limited types but overlook some adversities associated with the development of different problems like chronic illness and injuries. According to theories like attachment theory, ecological-transactional model, and general strain theory, adversities with different chronicity and timing impact outcomes differently. The single adversity approach, cumulative adversity score, and latent class analysis (LCA) should be combined to comprehensively address the multifaceted dimensions of childhood adversities. Covering 14 types of adversities from birth to adolescence, the study is the first one to examine the association between childhood adversities and delinquency, with comprehensive consideration of dimensions of childhood adversities. The study employs the Millennium Cohort Study, which tracks the trajectories of approximately 19,000 individuals who were born between 2000 and 2002 in England, Scotland, Wales, and Northern Ireland. For the purpose of the present study, the final sample is restricted to 9,980 individuals. An LCA was conducted to capture patterns of exposure to 14 adversities over 14 years, and linear regression models were applied to examine associations between different early adversity assessment methods (single adversity, cumulative scores as an average and categorical score, LCA) and delinquency. To compare cumulative adversity scores and LCA, two ANOVA tests and effect size calculations were conducted to compare delinquency levels between cumulative adversity categories and latent classes. When controlling for covariates and other single adversities, parental recreational drug use and various forms of direct violence, including bullying/online bullying, criminal victimization, and sexual victimization, were significantly associated with delinquency. Cumulative adversity scores also showed significant associations with delinquency after adjusting for covariates. The LCA identified nine classes, with three (low adversities, especially in childhood, low adolescent and household health adversity, and persistent poverty with low outside violence during adolescence) negatively associated with delinquency and three (high parental substance use, unstable households, and high adversity) positively associated. After adjusting for all covariates, only the persistent poverty and high adversity classes remained significant. Cumulative adversity scores had a stronger impact on delinquency compared to LCA, accounting for more variance. The single adversity approach and LCA yielded different insights. In conclusion, the study demonstrates that three approaches each provide unique insights into the link between early adversities and delinquency. This suggests that combining these methods is crucial for a comprehensive understanding of how early adversities affect outcomes. Additionally, the study, covering a wide range of adversities over extensive periods, highlights the co-occurrence and dynamic nature of adversities. It shows how different dimensions of adversity influence their impact on delinquency.

Keywords: childhood adversity, delinquency, latent class analysis, cumulative adversity score