PROJECT SUMMARY
Roughly one-fifth of girls in East and Southern Africa (ESA) have given birth by age 20. Social and economic
conditions accompanying the transition to motherhood increase HIV risk for adolescent mothers (AM). Social
protection programs, such as social asset building, educational subsidies, and economic strengthening, have
received growing attention and investment across ESA, and seek to reduce adolescent girls’ HIV risk by
increasing their empowerment, educational attainment, and economic stability. While AMs’ social and
economic realities may introduce unique barriers to participating in such programs, no studies have assessed
the extent to which AMs participate in – or benefit from –these programs. There are also gaps in understanding
which AMs are most likely to participate in social protection programs, which programs AMs are most likely to
use, and which programs are best positioned to decrease sexual risk behavior and increase HIV testing for
AMs after childbirth (when risk continues). Such information is needed to refine or develop programs that reach
AMs, who are highly vulnerable to HIV yet often overlooked in HIV prevention efforts. Therefore, our study
objective is to understand whether AMs participate, which AMs participate and which programs to prioritize to
optimize impacts of HIV prevention initiatives for AMs across countries with high HIV-burden among girls. In
this study, we will leverage publicly available, nationally representative survey data collected between 2019-
2022 from over 14,000 adolescent girls (15-19 years old), of whom 2,600 are mothers, as part of the second
round of the Population-based HIV Impact Assessment (PHIA) Projects which occurred in 8 countries across
East and Southern Africa to execute the following aims: (1a) Characterize AMs and their access to social
protection within and across countries; (1b) Examine whether adolescent motherhood (versus adolescent non-
motherhood) is associated with lower participation in social protection programs; (1c) Explore whether the
association between social protection programs and HIV outcomes differs for AMs compared to adolescent
non-mothers. (2) Using causal inference approaches, identify which social protection programs would have the
largest impact on decreasing sexual risk behavior and increasing HIV testing among AMs. For Aim 1, we will
use descriptive statistics and multilevel mixed effects regression models to examine differences in participation
in social protection programs among AMs within and across countries. For Aim 2, we will use a simulation-
based approach (the parametric g-formula) to quantify the change in sexual risk behavior/HIV testing that
would result from increasing AMs’ participation in existing programs. In ESA, where adolescent girls’ HIV
prevalence is 3-5X greater than boys’, and where adolescent motherhood increased during the COVID-19
pandemic, refining or developing new social protection programs to reduce HIV risk for AMs is key to meeting
UNAIDS’ 95-95-95 targets by 2030.
