Latent Classes of Adolescent Trauma Exposure, Posttraumatic Stress Disorder Symptoms, and Substance Use Predict Clinical Diagnoses at 12-Month Follow-Up.

Abstract

Background: Trauma exposure, posttraumatic stress disorder (PTSD), and substance use commonly co-occur among youth. Identifying specific subgroups of youth based on unique constellations across these domains may provide a novel way to identify and target youth at prospective risk for specific types of negative clinical outcomes.

Methods: Trauma exposed youth completed structured clinical assessments as part of a longitudinal study (N = 1826; ages 13-21). Latent class analyses identified distinct subgroups of youth based on lifetime trauma histories and current PTSD symptom and substance use inventories collected at the baseline study visit. Logistic regression analyses determined if the latent classes were associated with elevated risk for PTSD or substance use disorder (SUD) diagnoses as the 12-month follow-up study visit (n = 1029). Logistic regression models controlled for baseline clinical characteristics and demographic factors in a stepwise fashion to elucidate if latent classes carried conferred risk beyond established risk factors. Sensitivity analyses included latent profile analyses and predictive modeling with an alternative number of latent classes.

Results: Four latent classes were identified which differentiated participants based on the type of trauma exposure, the number of PTSD symptoms endorsed, and the propensity to be engaged in polysubstance use. Latent classes which were characterized by exposure to interpersonal violence at the baseline study visit had an elevated risk of PTSD 12 months later, relative to the latent class which was principally exposed to incidental trauma (odds ratios ranged from 4.11-5.88). Likewise, a distinct latent class which was characterized by poly-substance use at the baseline study visit had an elevated risk of SUD diagnoses at the 12-month follow-up (odds ratio = 2.48). The findings were robust to sensitivity analyses.

Conclusion: These results highlight nuanced patterns of co-occurrences between trauma exposure, PTSD symptomatology, and substance use that differentiate unique sub-groups of youth at varying degrees of risk for negative clinical outcomes one year later. Evaluating the co-expression of trauma and psychopathology inventories, as opposed to only assessing the summative epidemiological indices of these constructs, may help identify adolescents who are most at risk for sustaining deleterious health outcomes.