Share
 
Title
HIV incidence trends by marital status in mature HIV epidemics: Evidence from two community cohort studies in Uganda
Presenter
Georges Reniers
Authors
G. Reniers1,2, K. Risher1, E. Sully3, E. Slaymaker1, T. Lutalo4, R. Newton5,6, D. Nabukula4, J. Seeley6,7, B. Zaba1
Institutions
1London School of Hygiene and Tropical Medicine, Department of Population Health, London, United Kingdom, 2University of the Witwatersrand, School of Public Health, Johannesburg, South Africa, 3Guttmacher Institute, New York, United States, 4Uganda Virus Research Institute, Rakai Health Sciences Program, Rakai, Uganda, 5University of York, Department of Health Sciences, York, United Kingdom, 6Medical Research Council / Uganda Virus Research Institute, Uganda Research Unit on AIDS, Entebbe, Uganda, 7London School of Hygiene and Tropical Medicine, Department of Global Health and Development, London, United Kingdom

Background: The association between marital status and HIV incidence has at times been heavily contested in the literature. In this study, we revisit the evidence against the background of increasingly expansive combination HIV prevention programmes, including medical male circumcision and antiretroviral treatment for prevention.
Methods: Data for this study come from two population-based HIV surveillance sites Uganda: the Kyamulibwa General Population Cohort in the Masaka District and the Rakai Community Cohort Study. Each of these have conducted repeated serological surveys, which allow for the direct measurement of HIV incidence. Our dataset spans the years 1999-2015 (to be updated by the time of the IAS conference), and is used to describe trends in HIV incidence rates and the distribution of seroconverters by marital status. Poisson regression is used to test for changes in the relationship between marital status and HIV incidence over time. All analyses are gender disaggregated and control for age.
Results: Known HIV negative adults (aged 15-50) in Rakai and Masaka respectively contribute 68,132 and 66,667 person-years of exposure to the analysis. HIV incidence rates are typically highest among never married or formerly married men and women, but married individuals still account for around 40% of new infections because Ugandans spend most of their sexually active lives in a formal union. We find evidence for a decline in HIV incidence rates in young married men in Kyamulibwa (Incidence Rate Ratio: 0.38, 95%-CI: 0.17-0.84) and young married women in Rakai (0.70, 95%-CI:0.49-0.99, Figure 1). There is no evidence of a decline in HIV incidence among married individuals above age 30, but their incidence rates have always been relatively low. HIV incidence among never married or formerly married men and women stagnated, or in the case of older women in Masaka, increased.
Conclusions: We find support for recent studies that documented declines in HIV incidence in the period that HIV prevention efforts expanded. However, the possible impact of these programs is confined to young married men and women. Other prevention efforts are needed to reach never or formerly married men and women.


Figure 1: HIV incidence rate ratios 2006-?14/15 versus 1999-?03 by sex, study site, marital status and age
[Figure 1: HIV incidence rate ratios 2006-?14/15 versus 1999-?03 by sex, study site, marital status and age]