Abstract Body

Traditional male circumcision (TMC) is widespread across different cultures and religions, and is highly prevalent in sub-Saharan Africa. Given variability between cultures in the traditional procedure in terms of foreskin removal, it is unknown if TMC is always as protective against HIV as voluntary male medical circumcision (VMMC) regardless of the traditional procedure used.

HPTN 071 (PopART) is a large-scale combination prevention trial underway in South Africa and Zambia, for which a population cohort of adults aged 18-44 has been randomly recruited. We use logistic regression on data from the baseline survey of the population cohort to examine associations between measured HIV serostatus and self-reported circumcision status, adjusting for potential covariates. We divide individuals into those who have not been circumcised, and those who have undergone one of the following: TMC in Zambia; TMC in South Africa; VMMC; and those who report being circumcised by a practitioner other than a traditional practitioner or government health worker, or report being circumcised but do not specify by who. We differentiate TMC by country to reflect that the traditional procedure used may differ. 

5,301 out of the 11,231 men in the population cohort reported being circumcised. 60.6% of those circumcised reported being circumcised by a traditional practitioner, and 28.5% reported VMMC. 83.2% of circumcised men in South Africa (mostly Xhosa) were circumcised by a traditional practitioner, while only 14.0% of circumcised men in Zambia had undergone TMC. Figure 1 shows that in both countries HIV prevalence was higher amongst men who have undergone TMC (16.1% in South Africa; 16.7% in Zambia) than those who had undergone VMMC (6.3% and 7.8% respectively). After adjusting for age, education, marital status and number of lifetime partners, HIV prevalence remained significantly higher in men who underwent TMC in both South Africa (adjusted odds ratio=1.67, 95% confidence interval 1.32-2.14, P<0.001) and in Zambia (1.55, 95% confidence interval 1.00-2.36, P=0.046) compared to those who underwent VMMC.

Traditional male circumcision, as practiced by ethnic groups in South Africa and Zambia, is associated with higher HIV prevalence than voluntary male medical circumcision. Further work, designed to examine the extent to which traditional circumcision, as practiced in this and other regions, is protective against HIV acquisition, should be carried out.