Social Ecological Factors Associated with Multi-Dimensional Experiences of Violence Among Urban Refugee Adolescent Girls and Young Women in Kampala, Uganda

Background and Purpose: Nearly 60 percent of the world’s 22 million refugees live in urban areas rather than camp settings. Little is known about the needs of urban refugee adolescent girls and young women. Urban refugees may experience increased vulnerability to exploitation, violence, and poverty. Uganda hosts over 1 million refugees—most women and girls—and nearly 100,000 live in the capital city Kampala. Urban refugees receive limited social services and humanitarian support after leaving camps. The study objective was to examine the prevalence and correlates of multi-dimensional experiences of violence among urban refugee adolescent girls and young women (AGYW) in Kampala, Uganda.

Methods: We conducted a community-based study involving a cross-sectional tablet-based survey with urban refugee AGYW aged 16-24 in Kampala. We used peer driven sampling to recruit participants from 5 slums (Nsambya, Katwe, Kasanga, Kabalangala, Rubanga) in Kampala. We conducted multivariable logistic regression to examine social ecological factors associated with experiencing sexual and physical violence.

Results: Participants (n=333; mean age: 19.3, SD: 2.56) were born in: Democratic Republic of the Congo (45.9%), Burundi (33.3%), South Sudan (9.0%), Rwanda (5.7%) and others (6.0%). Two-thirds (67.3%; n=224) reported a physical violence history and one-quarter (23.7%; n=79) a sexual violence history. One-fifth reported experiencing both physical and sexual violence (polyvictimization) (21.9%; n=73). Over two-thirds reported food insecurity (71.4%; n=238) and one-quarter engaged in past year transactional sex (26.5%; n=86). In multivariable logistic regression analyses, the odds of experiencing physical violence were higher among participants who were currently dating (AOR: 8.74; 95 % CI 1.89–40.34); engaged in transactional sex (AOR: 2.99; 95 % CI 1.10–8.09); reported depressive symptoms (AOR: 1.09; 95 % CI 1.02–1.18); and perceived their community as unsafe (AOR: 3.16; 95 % CI 1.42–6.99). Higher social support was associated with lower odds of physical violence (AOR: 0.86; 95 % CI 0.79–0.95). Having children (AOR: 3.08; 95 % CI 1.17–8.10) and transactional sex engagement (AOR: 4.54; 95 % CI 2.07–9.96) were associated with increased odds of experiencing sexual violence. Transactional sex engagement (AOR: 7.19; 95 % CI 2.91–17.76) and lower social support (AOR: 0.92; 95 % CI 0.85–0.99) were associated with increased odds of polyvictimization.

Conclusion and Implications: Refugee AGYW living in Kampala’s urban slums experience alarming rates of physical and sexual violence. Social ecological factors across community (perceived neighbourhood safety, social support), interpersonal (relationship status, having children), and intrapersonal (transactional sex involvement, depression) domains were associated with experiencing violence. Multi-level interventions with urban refugee AGYW can address social and structural factors, such as improved neighbourhood safety through community-based slum development; economic security and empowerment; and support the needs and priorities of sex work involved refugees. International social work can address the vulnerabilities of refugee AGYW experiencing intimate partner violence, young parents, and those experiencing depression. Social support is a protective factor that can be leveraged through social support groups. There is an urgent need to reduce violence exposure and promote human rights of urban refugee adolescent girls and young women in Uganda.