Africa is home to just 10% of the world’s population but has a startling 67% of the world’s cases of
HIV/AIDS. And the rates of new infection continue to rise—2.8 million new cases of HIV were reported in sub-Saharan Africa in 2006, up from 2.6 million in 2004 (in North America, the rate of new cases was 43,000 in both 2004 and 2006).
In the 1980s, researchers noticed that HIV was less common in African cultures where men were routinely circumcised, while men living in the regions hardest hit by HIV were not circumcised. Further observations suggested that circumcision reduces the risk of HIV infection. This theory was recently tested in three randomized, controlled trials conducted in Uganda, South Africa, and Kenya. All three reported significant drops in rates of HIV infection among men who were circumcised, compared with men who were not. The findings were so dramatic that the trials were stopped early so the men in the control groups could be offered circumcision.
The largest of these trials, reported in the February 24, 2007
involved 5,000 uncircumcised men in Uganda. Half of the men were circumcised, while the other half served as controls. After two years of follow-up, the rate of HIV infection was 51%-60% lower among the men who were circumcised.
Scientists from Johns Hopkins University teamed up with researchers at several sites in Uganda to recruit uncircumcised, HIV-negative, heterosexual men, aged 15-49, for this study. Half of the men (2,474) were circumcised at the start of the study, while the other half (2,522) remained uncircumcised as the control group. The initial screening included a physical exam, HIV test, and questions about HIV-related behaviors. Exams and interviews were repeated six weeks into the study, as well as 6, 12, and 24 months later. Researchers compared the number of men who contracted HIV in the two groups.
After 24 months, the rate of HIV infection was significantly lower among circumcised men regardless of marital status, education, age, or risky sexual behaviors. The reductions in incidence ranged from 51%-60%, depending on the statistical analysis used. Eight percent of men who were circumcised reported side effects, most of which were mild and resolved with treatment.
In the South African study, men aged 18-24 were randomly assigned to be circumcised (1,546 men) or to be in the control group (1,582 men). During 18 months of follow-up, there were 20 cases of HIV infection in the circumcised group and 49 cases in the control group, which translates to a 60% reduction in HIV infection associated with circumcision.
A similar design was used in the Kenyan trial, in which men aged 18-24 were assigned to either circumcision (1,391 men) or control (1,393 men). Over 24 months of follow-up, circumcision was linked to a 53%-60% reduction in risk of HIV infection. There were 22 cases of HIV in the circumcised group and 47 cases in the control group. Twenty of the circumcised men reported side effects, all of which were mild to moderate in severity and resolved quickly.
One caveat to these findings is the chance that stopping a trial early could overestimate the benefit of the intervention.
These three clinical trials in three different populations revealed profound benefits of circumcision in the reduction of HIV infection among heterosexual men. Assuming the 60% reduction in risk can be applied to the general population, an estimated 2 million new HIV infections and 300,000 deaths could be avoided over the next ten years if all men in sub-Saharan Africa are circumcised. Women will likely benefit as well; with fewer infected men, the rates of male-to-female transmission of HIV will decline.
Men who are circumcised should not interpret these findings as license to be less vigilant with other HIV-prevention measures. Circumcision does not provide complete protection and should be just one facet of a low-risk lifestyle that includes the following:
- Condom use
- HIV testing and counseling
- Prompt diagnosis and treatment of sexually transmitted infections
- Avoidance of intravenous (IV) drug use