Sexual violence and general functioning among formerly abducted girls in Northern Uganda: the mediating roles of stigma and community relations--the WAYS study

Abstract

Background: Although sexual violence in war is associated with long-term mental health problems, little is known about its association with general functioning and the factors that explain this association. This study aims to illuminate the path from sexual violence to poor functioning. The prevalence of sexual violence among formerly abducted girls in Northern Uganda was assessed as well as the extent to which stigma and community relations explain the association between sexual violence and general functioning.

Method: In a cross-sectional analysis using data from the WAYS study (N = 210, baseline age 22.06, SD = 2.06, minimum-maximum 18-25), the extent of mediation of the association between sexual violence and general functioning was assessed in multiple regression models.

Results: Sexual violence was found to be associated with increased stigma, poor community relations, and poor general functioning. The association between sexual violence and general functioning was mediated by stigma and community relations. The bootstrap results indicated significant mediation by stigma of 47 % (95 % confidence interval [CI] 35 to 78 % and by community relations of 67 % (95 % CI: 52 to 78 %) in the association between sexual violence and general functioning.

Conclusion: Thus, poor functioning among formerly abducted girls is largely mediated by stigma and poor community relations. However, due to the relatively small effect sizes of the associations, targeted interventions to prevent impaired functioning may have only modest benefits to the formerly abducted girls. Interventions to alleviate the toxic effects of sexual violence in formerly abducted girls would benefit from a holistic approach that targets stigma and poor relationships within communities.

Fig. 1

Mediation model

Fig. 2

Mediation by stigma/discrimination of the relations between reported sexual violence and general functioning. Total effect: β = .15(95 % CI: .02, .30). Total indirect effect: β = .07(95 % CI). Total direct effect: β = .08(95 % CI: −.07, .22). The β below the continuous line from reported sexual violence to functioning represents the total effect of reported sexual violence on general functioning while the β above the dotted line represent the effect of reported sexual violence after stigma was added to the model as a mediator. Approximately 47 % of the effect of reported sexual violence on functioning is mediated through stigma. The direct effect of reported sexual violence on stigma reduced and became insignificant (β = .08 (95 % CI: −.07, .22)). All analyses were adjusted for age, duration in captivity and clustering by districts

Fig. 3

Mediation by community relations of the relations between reported sexual violence and general functioning. Total effect: β = .15(95 % CI: .02, .30). Total indirect effect: β = .10 (95 % CI: 02, .21). Total direct effect: β = .05 (95 % CI: −.08, .19). The β below the continuous line from reported sexual violence to general functioning represents the total effect of reported sexual violence on general functioning while the β above the dotted line represent the effect of reported sexual violence after community relation was added to the model as a mediator. Approximately 67 % of the effects of reported sexual violence on general functioning is mediated through community relations. The direct effect of reported sexual violence on functioning attenuated and became insignificant (β = .05 (95 % CI: −.08, .19)). All analyses were adjusted for age, duration in captivity and clustering by districts