The real-world association between male circumcision and risk of HIV infection in sub-Saharan Africa: a household fixed-effects analysis of 279,351 men from 29 countries

IntroductionWhile voluntary medical male circumcision (VMMC) reduces the individual- level risk of HIV acquisition by approximately 60% in randomised-controlled trials, little is known about the real-world long-term effect of medical and traditional male circumcision on the cumulative risk of HIV infection. We estimate the association between these for the first time using a quasi-experimental study design--a household fixed-effects analysis--for sub- Saharan Africa, the global region with the largest HIV burden. MethodsWe pooled individual-level cross-sectional data from the nationally-representative Demographic and Health Surveys and AIDS Indicator Surveys across all sub-Saharan African countries in which the surveys included data on both male circumcision and HIV status. We estimated the association between male circumcision and HIV status using modified Poisson regression models with household fixed-effects--which control for unobserved and observed confounding shared by men living in the same household--and included additional individual- level controls for demographic characteristics, socio-economic factors, and sexual behaviour. ResultsWe included individual data from 279,351 male participants in 48 nationally- representative surveys conducted in 29 countries between 2003-2018. The mean survey-level prevalence of male circumcision was 65.9% (median 84.5%, IQR 28.8%-68.1%) and HIV was 5.6% (median 2.5%, IQR 1.2%-10.2%). We estimated that male circumcision was significantly associated with a nearly one-fifth reduction in the cumulative risk of HIV infection (adjusted risk ratio 0.81, 95% CI 0.73-0.89). ConclusionsMale circumcision was associated with a significant reduction in the risk of HIV infection in sub-Saharan Africa over the past two decades. Increased political and financial commitment to VMMC could likely lead to further reductions in HIV prevalence, especially when rolled out as a HIV prevention option in combination with other interventions.