Share this article on:

Response to Ronald Gray et al., Male circumcision and HIV acquisition and transmission: cohort studies in Rakai, Uganda (AIDS 2000, 14:2371–2381).

Halperin, Daniel T.a; Weiss, Helen A.b; Hayes, Richardb; Auvert, Bertranc; Bailey, Robert C.d; Caldwell, Jacke; Coates, Thomasa; Padian, Nancya; Potts, Malcolmf; Ronald, Allang; Short, Rogerh; Williams, Briani; Klausner, Jeffreya


aUniversity of California, San Francisco, CA, USA (e-mail Dr Halperin at; bLondon School of Hygiene and Tropical Medicine, London, UK; cUniversity of Paris, Paris, France; dUniversity of Illinois, Chicago, IL, USA; eAustralian National University, Australia; fUniversity of California, Berkeley, CA, USA; gUniversity of Manitoba, Canada; hUniversity of Melbourne, Melbourne, Australia; and iCouncil of Scientific and Industrial Research, South Africa.

Received: 1 June 2001;

revised: 3 September 2001; accepted: 1 November 2001.

The renewed interest in the potential role of male circumcision for HIV prevention in regions severely affected by heterosexual HIV transmission [1–14] was further stimulated by recent findings from Uganda. The Rakai Project researchers, who had previously reported that none out of 50 HIV-negative circumcised men in serodiscordant relationships seroconverted after nearly 2 years of exposure (versus 40 seroconversions among 137 uncircumcised men) [15], now conclude that the protective effect of circumcision may be lower or negligible among certain sub-populations such as non-Muslims, adolescents, and never-married men. We are concerned that this conclusion is based on analysis of small sub-groups containing insufficient numbers to be informative, and should not be used to negate the more than 35 studies from 10 other countries, including six previous cohort investigations [1–4,8,9,12,14], which found a protective effect of circumcision among mainly non-Muslim men.

Gray et al. found overall a statistically significant protective effect of circumcision on HIV incidence in a population of rural Ugandan men [adjusted rate ratio (RR) 0.53, 95% confidence interval (CI) 0.33–0.87]. Circumcision status was, however, highly associated with religion in this population (99% of Muslims were circumcised compared with 3.7% of non-Muslims), and we agree with the authors that other characteristics associated with Muslim religion (i.e. post-coital cleansing, abstinence from alcohol, polygamous marriage) may influence the risk of HIV. Yet their conclusion that the latest Rakai findings ‘suggest that the protective effect of circumcision may be lower or negligible among … non-Muslims’ has several problems: (i) The observed difference in incidence between Muslim and non-Muslim circumcised men was, in fact, not significant (RR 0.63, 95% CI 0.21–2.07), and could have occurred by chance. (ii) Whereas virtually all Muslims were circumcised for religious/cultural reasons, 75% of non-Muslims who were circumcised post-pubertally reported that the procedure was performed for health reasons, typically involving sexually transmitted diseases (STD) or other genital/foreskin infections. This fundamental difference in the reasons for undergoing circumcision clearly confounds the interpretation of data on men circumcised post-pubertally, because such infections increase the risk of HIV acquisition, and are also more frequent among higher-risk individuals. If analysis is restricted to men who were circumcised pre-pubertally, there would be only two incident HIV cases among circumcised non-Muslims, resulting in an RR of 0.71 (95% CI 0.18–2.85) compared with uncircumcised men, but with wide confidence intervals. (iii) Among the HIV-discordant couples analysed in the study, there was also a very strong and significant protective effect of circumcision on the risk of HIV acquisition among the 14 circumcised non-Muslim men – regardless of their partners’ viral load levels (RR 0, 95% CI 0–0.88). (iv) The confounding between religion and circumcision postulated in Rakai has not been observed in other studies of circumcision and HIV infection. In most other regions where a strong association between the lack of circumcision and HIV risk has been identified, Muslims account for only a minority of circumcised men. For example, contrary to the suggestion of Gray et al. that circumcision in Kenya is ‘only practised in selected ethnic groups’ that may have certain cultural practices that differentially affect the risk of HIV, in fact it is performed by all major ethnic groups except the Luo, who have experienced the country's most severe AIDS epidemic [2,3,6,9,11]. A recent study [3,6] has shown that even among the Luo, circumcision is associated with a strongly significant reduction in the risk of HIV (odds ratio 0.26, 95% CI 0.12–0.56). Fewer than 1% of men in this population were Muslim. In a Kenyan prospective study [12], which found that uncircumcised truck drivers were four times more likely to become HIV infected, just 24% of circumcised men were Muslims, and in a Tanzanian study [16], in which 18% of circumcised men were Muslims, a significant association between the lack of circumcision and HIV infection remained after results were re-analysed excluding Muslims.

Furthermore, there is mounting biological evidence for the protective effect of male circumcision on the risk of HIV. Researchers have found that the highly vascularized inner foreskin tissue contains a much greater density of Langerhans cells than vaginal, rectal, or cervical mucosa, as well as large numbers of CD4 T cells, macrophages, and other HIV immune system targets [7–9,17,18]. A recent in-vitro investigation [9] found the uptake of HIV virus in the inner foreskin mucosa to be many times more efficient than in a control sample of cervical tissue. In addition, a lack of circumcision may increase HIV transmission indirectly through increasing the risk of ulcerative sexually transmitted infections such as chancroid and syphilis [2,4,8,9,14,19], which further exacerbate the risk of HIV infection.

Gray et al. conclude by emphasizing the need for randomized controlled trials, and the potential difficulties associated with these. We agree with the authors and others [1,4,5,8,9] that clinical trials would be useful for better establishing the causal relationship between circumcision and HIV infection, but are less pessimistic regarding their feasibility. The primary biological mechanism for the protective effect of male circumcision is now presumed to be the removal of Langerhans and other HIV target cells, and there is no other persuasive biological or epidemiological evidence that circumcision will be more effective among men circumicised pre-pubertally [1,8]. There is thus no reason why clinical trials could not be conducted among consenting men in the general population in their late teens or early-to-mid 20s, which is usually before many African men are infected by HIV (unlike the case in young women) [3,6], thereby avoiding the long follow-up and possible ethical problems involving research on minors raised by Gray et al.

It is critical to assess carefully a range of issues including sexual behaviour change, the age of circumcision, complication rates, community support, and cost-effectiveness before and during such studies. Feasibility research and training programmes for introducing safe male circumcision services have been carried out in western Kenya [8,9,11], and preliminary acceptability studies have begun in South Africa [20,21], Botswana [21] (R. Shapiro, Harvard AIDS Institute, unpublished data), Uganda (R. Gray, personal communication), and Zimbabwe (D. Halperin et al., unpublished data). Although most men in these surveys report that they would elect the procedure if it were safely and affordably available, few believe that circumcision can actually prevent AIDS. (The primary reasons for wanting circumcision are related to hygiene and infection/STD prevention, and some also say, e.g. to make condom use easier [22].) This suggests the possibility that if the practice becomes more common in such regions, widespread behavioural ‘disinhibition’ is unlikely. The first clinical trial of male circumcision, in Kisumu, Kenya, has been approved by ethical review boards and began enrolling participants, aged 18–24 years, in February 2002 [8].

We feel strongly, however, that the improvement and evaluation of male circumcision services must not wait until we know, several years and perhaps millions of new infections hence [22], the results from this or other intervention trials. Growing numbers of young men in many of the eastern and southern African countries most severely affected by HIV/AIDS (including some of the most economically developed nations in the region [23]) are seeking access to safer circumcision services [2,4,5,8–11,19–21,23,24]. The responsible dissemination of information on the risks and benefits of male circumcision and the development of affordable services for safe, voluntary circumcision should be integrated, on a pilot basis in appropriate locales, with existing HIV/STD and reproductive health programmes. Such efforts must commence now, lest the opportunity for a potentially important prevention measure be further delayed.

Daniel T. Halperina

Helen A. Weissb

Richard Hayesb

Bertran Auvertc

Robert C. Baileyd

Jack Caldwelle

Thomas Coatesa

Nancy Padiana

Malcolm Pottsf

Allan Ronaldg

Roger Shorth

Brian Williamsi

Jeffrey Klausnera

Back to Top | Article Outline


The authors would like to thank Esther Hudes, PhD, Stephen Moses, MD, MPH, Phillip Nieburg, MD, MPH, George Rutherford, MD, and two anonymous reviewers for their useful comments.

Back to Top | Article Outline


1. Weiss H, Quigley M, Hayes R. Male circumcision and risk of HIV infection in sub-Saharan Africa: a systematic review and meta-analysis. AIDS 2000, 14: 2361–2370.
2. Halperin DT, Bailey RC. Male circumcision and HIV infection: 10 years and counting. Lancet 1999, 354: 1813–1815.
3. Auvert B, Buvé A, Lagarde E. et al. Male circumcision and HIV infection in four cities in sub-Saharan Africa. AIDS 2001, 15: S31–S40.
4. van Dam J, Anastasi MC (editors). Male circumcision and HIV prevention: directions for future research. Population Council/Horizons Project published report. Washington, DC, June 2000.
5. Schenker I, Farley T. Report of World Health Organization consultation meeting on male circumcision and HIV/AIDS. Durban, South Africa, 14 July 2000.
6. Auvert B, Buvé A, Ferry B. et al. Ecological and individual level analysis of risk factors for HIV infection in four urban populations in sub-Saharan Africa with different levels of HIV infection. AIDS 2001, 15: S15–S30.
7. Szabo R, Short RV. How does male circumcision protect against HIV infection? BMJ 2000, 320: 1592–1594.
8. Bailey RC, Plummer FA, Moses S. Male circumcision and HIV prevention: current knowledge and future research directions. Lancet Infect Dis 2001, 1: 223–231.
9. Bailey RC. Male circumcision as a possible HIV prevention strategy: current evidence.Paper presented at the 8th Conference on Retroviruses and Opportunistic Infections. Chicago, 4–8 February 2001. ( cfm?catID=90).
10. Nnko S, Washija R, Urassa M, Boerma T. Dynamics of male circumcision practices in northwest Tanzania. Sex Transm Dis 2001, 28: 214–218.
11. Bailey RC, Muga R, Poulussen R, Abicht H. The acceptability of male circumcision to reduce HIV infections in Nyanza province, Kenya. AIDS Care 2002, 14: 27–40.
12. Lavreys L, Rakwar JP, Thompson ML. et al. Effect of circumcision on incidence of HIV-1 and other sexually transmitted diseases: a prospective cohort study of trucking company employees in Kenya. J Infect Dis 1999, 180: 330–336.
13. Moench TR. Neglected opportunities. J Infect Dis 2000, 181: 804–805.
14. Moses S, Bailey RC, Ronald AR. Male circumcision: assessment of health risks and benefits. Sex Transm Infect 1998, 74: 368–373.
15. Quinn TC, Wawer MJ, Sewankambo N. et al. Viral load and heterosexual transmission of human immunodeficiency virus type 1. N Engl J Med 2000, 342: 921–929.
16. Quigley M, Munguti K, Grosskurth H. et al. Sexual behavior patterns and other risk factors for HIV infection in rural Tanzania: a case–control study. AIDS 1997, 11: 237–248.
17. Hussain LA, Lehner T. Comparative investigation of Langerhans’ cells and potential receptors for HIV in oral, genitourinary and rectal epithelia. Immunology 1995, 85: 475–484.
18. Miller CJ, Vogel P, Alexander NJ, Dandekar S, Hendrickx AG, Marx PA. Pathology and localization of simian immunodeficiency virus in the reproductive tract of chronically infected male rhesus macaques. Lab Invest 1994, 70: 255–262.
19. Caldwell J, Caldwell P. The African AIDS epidemic. Scientific American 1996, 274: 62–68.
20. Taljaard R, Taljaard D, Auvert B. Cutting it fine: male circumcision practices and the transmission of STDs in Carletonville, South Africa.Paper presented at World AIDS Conference. Durban, 10 July 2000 (Abstract MoOrC195).
21. Halperin DT. Cut and dry: feasibility and acceptability of reviving male circumcision and introducing dry sex prevention in Botswana and South Africa.Published abstracts of the American Anthropological Association 99th Annual Meeting. San Francisco, 18 November 2000.
22. Potts M. Circumcision and HIV [Letter]. Lancet 2000, 355: 926–927.
23. Halperin DT, Allen A. Is poverty the root cause of African AIDS?AIDS Analysis Africa 2000, 11:1, 3, 15.
24. Green EC, Zokwe B, Dupree JD. Indigenous African healers promote male circumcision for prevention of sexually transmitted diseases. Trop Doct 1993, 23: 182–183.
© 2002 Lippincott Williams & Wilkins, Inc.