Declines in HIV incidence among men and women in a South African population-based cohort
Southern Africa is the region with the world’s highest rate of new HIV infections. Over the last decade, there has been a multi-million-dollar investment in antiretroviral therapy to prevent the onward transmission of HIV and bring the epidemic under control.
As a post-doctoral researcher, I held several assumptions about how the unprecedented scale-up of antiretroviral therapy (ART) would impact the incidence of HIV in a hyperendemic South African setting (Figs. 1 & 2). These scale-up efforts included novel strategies to empower women and remove the structural barriers to ART uptake, barriers such as lack of education, poverty, sexual violence, and gender inequality.1 When I began my research in 2016, I therefore assumed that the earliest and largest declines in HIV incidence would occur among women. To the study team’s surprise, we observed the opposite.
We found that ART uptake was indeed higher among women, but this translated into larger declines in HIV incidence among men. In fact, male incidence was the first to decline between 2012 and 2017, from 2.5 to 1.0 new infections per 100 person-years—a reduction of 59%. This reduction was markedly higher than the 37% decline among women, from 4.9 to 3.1 new infections per 100 person-years between 2014 and 2017 (Fig. 3). How can these unequal declines in HIV incidence be explained?
One explanation is that once women achieved higher ART coverage and hence lower levels of detectable viremia, they were less likely to transmit HIV to their uninfected male partners. In the study area, as well in the broader sub-Saharan African region, the predominant mode of HIV transmission is through heterosexual intercourse. Plausibly, male HIV incidence declined first because women were the first to achieve a critical threshold of ART coverage in 2012 (Fig. 3).
Another explanation is that HIV-uninfected men benefitted from the scale-up of voluntary medical male circumcision services. Between 2009 and 2016, there was a steady increase in male circumcision coverage, from 3% to 33% (Fig. 3), mainly due to a decree by the Zulu King, in consultation with the local Department of Health, that all Zulu men in the study area be circumcised.2,3 Circumcision uptake, which can reduce male HIV risk during sexual intercourse,4 is likely to have contributed to the earlier and larger reductions in HIV incidence among men.
Our study findings reminded me of a paraphrased quote from my PhD dissertation: “what happens inside the body may be a biological matter, but HIV’s journey between bodies in blood and semen is both biological and social.”5 Biologically speaking, the unequal declines in HIV incidence are consistent with what we already know about the preventative benefits of ART and male circumcision. But it is the social aspects—the program strategies to increase ART uptake among women and a Zulu King’s decree for male circumcision—that led to our unexpected results. Getting more men onto consistent, suppressive ART needs to be part of the broader public health response, so that new HIV infections can be reduced among women.
1 See for example: PEPFAR (President's Emergency Plan for AIDS Relief) and DREAMS (Determined, Resilient, Empowered, AIDS-free, Mentored and Safe). https://www.usaid.gov/global-health/health-areas/hiv-and-aids/technical-areas/dreams
2 Sithole B, Mbhele L, van Rooyen H, Khumalo-Sakutukwa G, Richter L. Only skin deep: limitations of public health understanding of male circumcision in South Africa. South African Med J 2009; 99: 647.
3 KZN Department of cooperative governance and traditional affairs. Zulu Monarch calls for more men to undergo circumcision. 2018. http://www.kzncogta.gov.za/zulu-monarch-calls-for-more-men-to-undergo-circumcision/.
4 Kong X, Kigozi G, Ssekasanvu J, et al. Association of medical male circumcision and antiretroviral therapy scale-up with community HIV incidence in Rakai, Uganda. JAMA 2016; 316: 182–90.
5 Heimer, C. Old inequalities, new disease: HIV/AIDS in sub-Saharan Africa. Annu. Rev. Sociol. 2007, 33:551–577.