Developmental Patterns of Mental Health Problems Among Youth Exposed to Violence: A Latent Class Analysis

Background and Purpose: An estimated 61% of youth in the United States have been either direct victims or have witnessed violence in the past year, with 35 to 65% of youth reporting exposure to violence in multiple settings. Exposure to violence (EV) has a significant impact on youth development across multiple domains including social functioning, perpetration and experience of dating violence, and mental health. Despite evidence that the impact of EV may vary depending on developmental timing, existing studies on the impact of EV are limited to a single developmental period and only focus on cross-sectional outcomes. The current study addressed this gap through the following aims: 1) Identify latent subgroups of youth EV based on developmental timing and context (i.e., family-based vs. community-based violence), 2) Identify distinct adolescent mental and behavioral health trajectories based on subgroup membership.

Methods: This secondary analysis utilized the LONGSCAN dataset. The total sample at baseline consisted of 1345 youth. Variables were measured at different developmental periods constructed around school transitions: Birth to age five (“pre-elementary”), six to 11 (“elementary”), 12 – 13 (“middle school”) and 14 – 18 (“high school”). Five forms of EV (maltreatment, intimate partner violence, community violence, school violence, and non-parental victimization) were measured using administrative data, caregiver, youth, and teacher report. Youth mental health was measured using youth and caregiver report, and perpetration of violence was measured using youth self-report at age 12, 14, 16, and 18. A latent class analysis was used to identify latent subgroups of EV during each developmental period. Multi-level growth mixture models were used to assess the effect of patterns of EV on future mental and behavioral health symptomology.

Results: Three latent classes emerged within the pre-elementary, elementary, and middle school developmental periods defined by primary exposure to maltreatment, primary exposure to IPV, and lower exposure to maltreatment and IPV. Two latent classes emerged for the high school developmental period, defined by low exposure to maltreatment and IPV, and high exposure to maltreatment and IPV. Youth who were in the exposure to IPV class during the pre-elementary period had higher rates of anger, anxiety, depression, trauma symptoms, and dissociation than the other two groups by age 12. In contrast, youth who were in the high child maltreatment group during the elementary period had higher rates of anger, anxiety, depression, and trauma symptoms by age 16. Youth who were in the high child maltreatment group in the pre-elementary and elementary developmental periods were both more likely to meet diagnostic criteria for oppositional defiant and conduct disorders by age 14.

Conclusions and Implications: These results suggest distinct patterns of EV experienced by youth defined by developmental period that become less distinct in later adolescence. Results further suggest that the effect of EV on future mental health symptomology differs based on the form of EV and the developmental period in which the EV occurred. Therefore, interventions that aim to address the potential impact of EV should be tailored to specific forms of EV and the child’s developmental period.