Integrating spatial and non-spatial data to examine multilevel drivers of HIV risk among adolescent mothers in sub-Saharan Africa

点击翻译按钮获取中文摘要

PROJECT SUMMARY/ABSTRACT HIV transmission in sub-Saharan Africa (SSA) occurs predominantly among young people (aged 15-24 years old) and within micro-epidemics (high HIV prevalence areas). Among adolescents, girls account for 80% of all adolescent HIV infections (15-19 years old) and adolescent mothers (AMs) are at higher HIV risk than non-parenting girls. However, as AMs have been largely overlooked in HIV prevention research, gaps in knowledge regarding where micro-epidemics occur among AMs and how structural (e.g., urbanization) and relational HIV risk factors impact adolescent mothers’ HIV risk remain. Further, despite differences in AMs’ relationship configurations, no studies have examined how the co-occurrence of relational factors impact AMs’ vulnerability to HIV and unprotected sex using a typological approach. Moreover, most HIV studies on AMs in SSA use a single-level approach to describe HIV risk as opposed to a multilevel framework that incorporates spatial (i.e., structural factors) and non-spatial (i.e., relationship typologies, individual and familial factors) risk factors. This exploratory study will use a modified social ecological framework to identify micro-epidemics, and to characterize relationship typologies and multilevel factors that impact AMs’ HIV risk. The specific aims are to: 1) characterize HIV micro-epidemics among AMs in SSA, 2) identify typologies of AMs’ sexual relationships, and 3) assess whether multilevel factors are associated with HIV risk and unprotected sex among AMs. To address these aims, this study will use (1) secondary data from the Population-based HIV Impact Assessment (PHIA), which includes 2,879 AMs (aged 15-19 years old); and (2) publicly available structural-level data, from various sources which will be linked to PHIA cluster-level data. Multiple quantitative analytic approaches will be used to execute aims, including spatial and machine learning techniques (Aim 1), latent class analysis (Aim 2), and mixed effects modeling (Aim 3). Study findings may inform targeted HIV prevention interventions for AMs. This study aligns with the NICHD’s research priorities by examining structural risk factors of HIV in a high-risk setting and the UNAIDS 95-95-95 goal. The training plan developed by the PI, sponsor Dr. Félice Lê- Scherban, and co-sponsors Drs. Allison Groves and Alex Ezeh, supports the proposed research and the PI’s training goals, which are to: 1) gain expertise in theoretical frameworks of social epidemiological and adolescent health research, 2) develop methodological skills in machine learning and latent class analysis, and 3) refine research dissemination, communication, and scholarly writing skills. The PI will leverage resources at Drexel University Dornsife School of Public Health, a collaborative and multidisciplinary institution that prioritizes health disparities and global research. The proposed research, training, institution, and mentorship team will support the PI’s pursuit of becoming an independent social epidemiologist focusing on advancing health equity among marginalized adolescent populations.