Polytraumatization as a Predictor of Behavioral Outcomes among At-Risk Adolescents

• In Event: 1-108 - Toward Understanding Heterogeneity in Adolescent Trauma Experiences and Later Adjustment Outcomes

Integrative Statement

Research has shown that 68% of children have experienced a traumatic event prior to age 16 and 37% have experienced more than one (Copeland et al., 2007). In addition, exposure to multiple traumas has been shown to lead to a variety of adverse outcomes such as depression, PTSD, anxiety, and externalizing behaviors in adults and adolescents (e.g., Agorastos et al., 2014; Stimmel, Cruise, Ford, & Weiss, 2014). While it is important to understand how cumulative trauma affects individuals, it is also important to consider whether the experience of multiple types of trauma (i.e., polytraumatization) affects behavioral outcomes.
Therefore, the current study had two primary aims. First, the study sought to take a person-centered approach by identifying meaningful groups of individuals among a sample of at-risk youth, based on their polytraumatization profiles. Next, the study examined whether these unique profiles of polytraumatization differentially predicted internalizing and externalizing behaviors. We hypothesized that children who had experienced numerous types of trauma would display elevated levels of internalizing and externalizing behaviors.
Data were collected from a sample of 304 at-risk youth ranging from 6th to 8th grade whose rate of aggressive behaviors were in the top 25% of students, according to teacher ratings. Children retrospectively reported on their experience of lifetime trauma on the UCLA PTSD Index. One year later, parents rated children’s behavior on the BASC-2 and the Reactive and Proactive Aggression Questionnaire.
Analyses were conducted in Mplus and SPSS. Latent class analysis (LCA), a type of mixture modeling, was used to identify subgroups among the sample characterized by different profiles of polytraumatization, based on endorsement of traumatic experiences on the UCLA PTSD Index. 2-class and 5-class models were chosen based on the Bayesian Information Criterion (BIC) and entropy values. Within the 2-class model, high polytraumatization and low polytraumatization groups emerged. Within the 5-class model, a low polytraumatization group emerged, along with four other groups characterized by high polytraumatization and each of the following “primary” traumas: community violence (CV) victimization, CV witness, physical abuse, and medical treatment/hospitalization.
An ANOVA was used to compare levels of externalizing behavior, internalizing behavior, and reactive and proactive aggression across polytraumatization subgroups defined in the two models. Using the 2-class model, no differences emerged between the high and low polytraumatization groups on any of the outcomes variables. Using the 5-class model, the CV victimization group exhibited significantly higher levels of externalizing behaviors than the low polytraumatization and medical treatment/hospitalization groups. The CV victimization group also exhibited significantly higher levels of reactive aggression than the low polytraumatization, CV witness, and medical treatment/hospitalization groups.
Results suggest that unique patterns of polytraumatization differentially predict externalizing, but not internalizing outcomes. These findings also indicate that characterizing polytraumatization by “primary” traumas provides better predictive power than simply examining high versus low levels of polytraumatization. Finally, these results demonstrate that there is heterogeneity among adolescents with high polytraumatization, which may inform future research seeking to identify youth who may benefit most from preventive interventions.

Authors

• Francesca Kassing, University of Alabama
  Presenting Author

• John E. Lochman, University of Alabama
  Non-Presenting Author

• Kristina McDonald, University of Alabama
  Non-Presenting Author

• Eric M. Vernberg, University of Kansas
  Non-Presenting Author