Latent class analysis of violence against adolescents and psychosocial outcomes in refugee settings in Uganda and Rwanda

Abstract

Little is known about violence against children in refugee camps and settlements, and the evidence-base concerning mental health outcomes of youth in refugee settings in low and middle-income countries is similarly small. Evidence is needed to understand patterns of violence against children in refugee camps, and associations with adverse mental health outcomes. Surveys were conducted with adolescent refugees (aged 13-17) in two refugee contexts - Kiziba Camp, Rwanda (n = 129) (refugees from Democratic Republic of Congo) and Adjumani and Kiryandongo refugee settlements, Uganda (n = 471) (refugees from South Sudan). Latent Class Analysis was utilized to identify classes of violence exposure (including exposure to witnessing household violence, verbal abuse, physical violence and sexual violence). Logistic regressions explored the association between latent class of violence exposure and symptoms of depression and anxiety. In Rwanda, a two-class solution was identified, with Class 1 (n = 33) representing high levels of exposure to violence and Class 2 (n = 96) representing low levels of exposure. In Uganda, a three-class solution was identified: Class 1 (high violence; n = 53), Class 2 (low violence, n = 100) and Class 3 (no violence, n = 317). Logistic regression analyses indicated that latent violence class was associated with increased odds of high anxiety symptoms in Rwanda (AOR 3.56, 95% CI 1.16-0.95), and high v. no violence class was associated with depression (AOR 3.97, 95% CI 1.07-7.61) and anxiety symptoms (AOR 2.04, 95% CI 1.05-3.96) in Uganda. The present results support the existing evidence-base concerning the association between violence and adverse mental health outcomes, while identifying differences in patterns and associations between refugee youth in two different contexts.