Social and Behavioral Predictors of Adherence to Oral Antiretroviral Pre-Exposure Prophylaxis for HIV Prevention among Adolescent Girls and Young Women in Southern Africa

Adolescent girls and young women (AGYW) in sub-Saharan Africa represent a key population at risk for acquiring HIV infection. Oral antiretroviral pre-exposure prophylaxis, or PrEP, is an efficacious biomedical HIV prevention strategy that holds enormous potential to reduce HIV acquisition in key populations if taken consistently. Research is needed to identify barriers and facilitators to PrEP adherence among AGYW to inform implementation efforts. The aims of this dissertation were to examine how disclosure of PrEP use and intimate partner violence (IPV) affect adherence to daily oral PrEP among AGYW in southern Africa. We used data from two prospective cohort studies evaluating PrEP uptake and adherence among sexually-active HIV-uninfected women aged 16-25 years. Participants who initiated once daily oral PrEP were included in analyses. We first conducted descriptive analysis of PrEP use disclosure and IPV exposure variables. Participants self-reported disclosure of their PrEP use during follow-up separately for parents, partners, and peers. IPV included any self-reported physical, emotional, sexual, or psychological partner violence. Then, for both aims, we estimated the relative prevalence of high PrEP adherence at months 3 and 6, defined as TFV-DP ≥700 fmol/punch at a given visit, using univariate and multivariable (adjusted) modified Poisson or log-binomial regression. Effects were estimated in the full study population and within younger and older AGYW age group strata (dichotomized at the sample median). We did not observe a strong effect of disclosure or IPV on adherence overall; however, the effect of disclosure to a parent and IPV were significantly modified by age (α=.15). For the parent disclosure-adherence relationship, there was evidence of effect modification by age at month 6 (p=.062): among younger AGYW (≤18 years), those who disclosed to a parent were 6.8 times as likely to have high adherence at month 6 than those who did not (95% CI 1.02, 45.56); among older AGYW (>18 years), there was no significant effect of disclosure to a parent (aPR=0.69, 95% CI 0.17, 2.90). For the IPV-adherence relationship, there was evidence of effect modification by age at month 3 (p<.001): among younger AGYW (