Circumcision is not associated with a reduced risk of HIV transmission from infected men to their uninfected female sex partners, according to results of a randomized trial conducted in the Rakai District of Uganda.1 The cumulative 24-month probability of acquiring HIV did not differ significantly between the female partners of infected men who underwent circumcision and those of infected men who did not (22% vs. 13%). The findings were similar after other factors potentially affecting the risk of infection were taken into account. Moreover, at a six-month follow-up, the rate of HIV infection among female partners in couples who had resumed having intercourse before the surgical wound was fully healed was 3.5 times that among female partners of men who did not have the surgery.

Men were eligible for the trial if they were aged 15–49, uncircumcised, infected with HIV and showed no signs of immunosuppression. The men were assigned to be circumcised either immediately (within two weeks of enrollment) or after a delay of 24 months. Those having immediate surgery were instructed to abstain from sex until medical staff had certified that the surgical wound was completely healed; during visits in the first six weeks after surgery, the wound was examined and the men were asked whether they had resumed intercourse. If a man had a wife or permanent consensual partner who was uninfected, he was asked to invite her to enroll, and those who accepted were given the same instructions regarding resumption of intercourse. In total, 922 men and 163 women participated in the trial. At enrollment and at follow-up visits six, 12 and 24 months later, blood samples were collected from both partners and tested for HIV. At each visit, participants were offered their HIV results, and were told that the effects of circumcision on HIV transmission were unknown and that adherence to safe sexual practices was essential. In addition, at each visit, participants were interviewed and asked about social, demographic, behavioral and health factors; were given intensive education about STI/HIV prevention; and were offered free condoms and voluntary HIV counseling and testing.

The study, conducted in parallel with a larger trial that examined whether circumcision reduces the risk of male infection, began in 2003 and was stopped early four years later because an interim analysis suggested that circumcision was highly unlikely to reduce the risk of male-to-female HIV transmission. Descriptive analyses were based on 93 couples in which the HIV-infected man was assigned to immediate circumcision and 70 couples in which the HIV-infected man was assigned to delayed circumcision. Efficacy analyses were restricted to the 92 couples in the former group and 67 couples in the latter group in which the female partner had had at least one follow-up visit.

Most of the men in the two groups (71–81%) and 31–51% of their female partners were 30–49 years old. Only 20–21% of the men and 6% of the women had more than a primary education. About four-fifths of participants were monogamous. Nearly all of the men—98% of those assigned to immediate circumcision and 94% of those assigned to delayed circumcision—opted to receive the results of their HIV test when they enrolled in the trial; somewhat smaller proportions of their female partners—69% and 74%, respectively—did so, but an additional 16% in each group already knew their HIV status. Roughly half of the men and more than two-thirds of the women said they had not used condoms at all in the past year.

During 12–24 months of follow-up, 18% of the female partners of men assigned to immediate circumcision and 12% of the partners of men assigned to delayed circumcision became infected with HIV. The cumulative 24-month probability of infection was 22% in the former group and 13% in the latter—a nonsignificant difference. The risk of acquiring HIV remained statistically indistinguishable between groups after the researchers adjusted for baseline differences in factors such as female age and condom use.

An unplanned secondary analysis revealed that in the immediate circumcision group, female partners' rate of HIV infection within six months of starting the trial was markedly higher among couples who resumed intercourse more than five days before doctors certified that the man's surgical wound had completely healed than among couples in the delayed circumcision group (rate ratio, 3.5). By contrast, female partners of men in the immediate circumcision group were not at higher risk if the couple had resumed intercourse within five days prior to or any time after certification of wound healing.

At follow-up assessments, female partners in the immediate circumcision group did not differ from those in the delayed circumcision group with respect to number of sexual partners, condom use or alcohol use with sex. In addition, similar proportions of female partners in each group developed genital ulcers (16%), vaginal discharge (32–36%) and painful urination (15–16%). The incidence of bacterial vaginosis was statistically indistinguishable between groups as well (52–56%), although female partners in the immediate circumcision group had a slightly lower incidence of trichomoniasis than did those in the delayed circumcision group (7% vs. 15%), a difference of borderline statistical significance.

The trial's findings suggest that circumcision does not reduce the risk of HIV transmission from infected men to uninfected women in the short term, according to the investigators. The findings contrast sharply with those of trials showing that male circumcision roughly halves the risk of HIV transmission from infected women to uninfected men. Nonetheless, the investigators recommend that circumcision be offered to all men regardless of their HIV status, to minimize any stigma associated with being circumcised or uncircumcised and to reduce the risks of genital ulcers and human papillomavirus infection in HIV-infected men, among other reasons. They stress, however, that it is imperative that men who undergo this surgery strictly abstain from sex during wound healing and consistently use condoms thereafter. In the broader context, the investigators note, although male circumcision does not appear to directly reduce women's risk of HIV acquisition, it may ultimately decrease women's exposure to infected men by reducing male infection. "Male circumcision programmes are thus likely to confer an overall benefit to women," they conclude.—S. London


1. Wawer MJ et al., Circumcision in HIV-infected men and its effect on HIV transmission to female partners in Rakai, Uganda: a randomised controlled trial, Lancet, 2009, 374(9685):229–237.