Caregiver Support as a Mediator of Sexual Risk Taking Among Orphaned Adolescents in Rural Uganda

Purpose: In sub-Saharan Africa 47 million children are orphans (UNICEF,2008). Orphaned children are primarily cared for by extended family members and approximately 1 of every 4 households in Uganda are providing care to an orphaned child (Cheng,2004). Orphaned children in sub-Sahara Africa are at an elevated risk for early onset of sexual relations and sexual exploitation, which increases their risks of HIV, STIs, and early pregnancies (Thurman,2006; Gregson,2005). The Suubi (Hope) family-based economic empowerment intervention demonstrated a positive effect on reducing adolescents' attitudes to engage in sexual risk behaviors (Ssewamala et al,2009). To understand mechanisms of change and connect intervention components and intervention outcomes, it is crucial to examine variables mediating behavioral change (Kraemer,2002). Family resilience framework suggests that family support can serve as a buffer balancing hardships and adversities that would push adolescents towards risky sexual behaviors (Fergus & Zimmerman,2005; Walsh,2003). Thus, this study examines the following research questions: 1) Does caregiver support change as a result of a family-based economic empowerment intervention (the Suubi intervention)? 2) Does caregiver support play a mediating role in changing attitudes toward sexual risk taking, as a primary outcome of Suubi intervention?

Method: This study uses data from the NIMN-funded Suubi (Hope) Research Project on a family-based economic empowerment intervention promoting life options for orphaned children. The study utilized a group-randomized experimental design and included 284 orphaned children (ages 11-17) from fifteen comparable schools in Rakai District of Uganda. A 90-minute individual interview was conducted at baseline, 10-month post-intervention and 20-month follow-up. First, the effects of Suubi intervention on perceived caregiver support and child-caregiver communication were tested using mixed effects regression models adjusting for nesting within schools. Second, using mediation analysis (Sobel,1982; MacKinnon,2002) we examined if changes in caregiver support and communication mediated the change in attitudes toward sexual risk taking.

Results: Compared to adolescents from the control group (not receiving the economic empowerment intervention), at 10-month follow-up, adolescents in the treatment group report higher levels of perceived instrumental support (B=.20, 95%CI=.03, .38, p<.05) and emotional support from caregivers (B=.25, 95%CI= .07, .43, p<.01), are more willing to talk to caregivers about their problems (B=.23, 95%CI=.01, .45, p<.05), and feel more comfortable talking about risk taking behaviors with their caregivers (B=.90, 95%CI=.58, 1.23, p<.001). The effects are sustained at 20-month follow-up. Mediation analysis demonstrates that the improvement in emotional support accounts for 15.3% of the total effect of Suubi intervention on reduction in attitudes toward sexual risk taking among adolescents. Sobel's test demonstrates that the mediation effect of emotional support from caregivers is statistically significant (z=-1.97, p<.05). Implications: Family-based economic empowerment intervention has a potential of strengthening family stability and increasing the sense of closeness. This has important programming and practice implications, especially given the increasing numbers of orphaned children in sub-Saharan Africa and the need for best practices in supporting and caring for such children. Interventions strengthening existing social networks and improving connectedness with surviving family members are critical in preventing sexual risk taking behaviors among orphaned adolescents in low resource communities.