Share
 
Title
Multilevel exposures and mediators of the relationship between schooling and risk of incident HSV-2 infection in young women enrolled in HPTN 068
Presenter
Marie C.D. Stoner
Authors
M.C.D. Stoner1, T.B. Neilands2, K.B. Rucinski3, A. Selin3, J.P. Hughes4,5, F.X. Gómez-Olivé6,7, C. MacPhail6,8,9, K. Kahn6,7,10, A. Pettifor3, HPTN 068
Institutions
1University of North Carolina, Carolina Population Center, Chapel Hill, United States, 2University of California, San Fransisco, United States, 3University of North Carolina, Chapel Hill, United States, 4University of Washington, Department of Biostatistics, Seattle, United States, 5Fred Hutchinson Cancer Research Center, Seattle, United States, 6MRC/Wits Rural Public Health and Health Transitions Research Unit (Agincourt), School of Public Health, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa, 7INDEPTH Network, Accra, Ghana, 8School of Health and Society, University of Wollongong, NSW, Australia, 9Wits Reproductive Health and HIV Research Institute, University of the Witwatersrand, Johannesburg, South Africa, 10Umeå University, Umeå, Sweden

Background: Our prior research from South Africa found that partner selection mediates the relationships between school attendance and incident HIV and HSV-2 infection, but little is known about other pathways. This study explores whether certain individual-, household-, and community-level factors account for the protective effect of schooling on incident Herpes Simplex Virus Type 2 (HSV-2) infection in young women.
Methods: We use longitudinal data from a randomized controlled trial of young women in rural South Africa (HPTN 068) collected during the main trial. We used discrete time survival models to estimate the three-year hazard ratio for the effect of distinct measures of schooling on incident HSV-2 infection, overall and through different pathways. Exposures of schooling included low attendance in school (≤ 80% of school days), household SES (quintiles of assets), parental education and mean community years of education. The natural indirect effect and natural direct effect were estimated as measures of mediation. Potential mediators included HIV knowledge, hope for the future, sex in the last 12 months, unprotected sex in the last 3 months and age-disparate relationship (>=5 years). Additionally, we explored interactions between the individual-, household-, and community-level exposures of schooling.
Results: Low attendance in school (≤ 80% of school days) was associated with incident HSV-2 infection (HR 3.45; 95% CI: 1.93, .18) while other household and community exposures of schooling were not. We did not find interactions between any of the multi-level exposures of schooling. The mediators age-disparate relationship, hope for the future and sex in the past 12 months were associated with HSV-2 infection but HIV knowledge and unprotected sex were not. Indirect effects were significant for sex in the last 12 months and age-disparate relationship.
Conclusions: Low attendance in school was associated with a higher hazard of incident HSV-2 infection and partnership level factors were the most important mediators. The relationship between school attendance and HSV-2 was mediated primarily by having an age-disparate relationship, and sex in the past 12 months. Given the protective effect of individual school attendance, focus should be on interventions to keep girls in school and influence behaviors related to exposure to risky partners.