See also this less activist and more mainstream article at CBCNews by Stephen Strauss: http://www.cbc.ca/news/viewpoint/vp_strauss/20051121.html

[A] study in Africa that was examining whether circumcision reduces the risk of men catching AIDS was stopped early, because it was so clearly true:

French and South African AIDS researchers have called an early halt to a study of adult male circumcision to reduce HIV infection after initial results reportedly showed that men who had the procedure dramatically lowered their risk of contracting the virus.

The study’s preliminary results, disclosed Tuesday by the Wall Street Journal, showed that circumcision reduced the risk of contracting HIV by 70 percent — a level of protection far better than the 30 percent risk reduction set as a target for an AIDS vaccine.
…
There is a reason to suspect a connection between AIDS transmission and circumcision:

Laboratory studies have found that the foreskin is rich in white blood cells, which are favored targets of HIV, the virus that causes AIDS. So the theory is that men who are uncircumcised are much more likely to contract the virus during sex with an infected woman, and that the epidemic spreads when these newly infected men have sex with other women within their network of sexual partners.

As I was reading the article, I found myself wondering if there might be some other factors involved in the disparity in AIDS rates, since circumcision in much of the world is associated with particular religious beliefs. Sure enough, at the end of the article is a discussion of why this might be a factor:

Although the apparent protective effect of circumcision has been noted for more than 20 years, doubts linger as to whether circumcision itself is protective, or whether the lower risk may be the result of cultural practices among those who circumcise. HIV rates are low in Muslim communities, for example, which practice male circumcision but also engage in ritual washing before sex and frown on promiscuity.

How well did this study distinguish circumcision from lower promiscuity rates?

Like the girls, some would and some wouldn’t. Maybe the proportions would be higher for the guys, but surely not 100%.

1. You state that the researchers essentially asked female virgins to lose their virginity. If so, they didn’t do a very good job of it. Certainly the NEJM paper only includes data on the 82 students who had vaginal intercourse during (or immediately before) the years they spent in the study, but as you note yourself, 65 more students went through the study without ever having vaginal sex. The virgin students weren’t included in the statistical analyses, but this is for the very good reason that their data wouldn’t shed any light on the question of whether condoms reduce the risk of HPV infection. (Excluding their data from analysis does not, by the way, mean that the virgin students wouldn’t receive a study stipend. In an observational study like this, a stipend is typically given for a subject’s time and trouble in filling out forms and making clinic visits, and the virgin students would have visited the clinic just as often as the sexually active students.)

2. You state that the researchers failed to inform participants of all the risks of participating in the study – that is, all the risks of having sex in the first place. However, the students weren’t required to have sex to participate in the study – the study only tracked the HPV rates of those who did. Look at it this way: if I were performing an observational study of the cholesterol levels of college students, does this mean that I’d be somehow encouraging them to eat unhealthy foods? Or that I’d be responsible for informing them of the dangers of emotional eating?

3. You state that the study didn’t obtain consent from more than half the participants being studied – that is, the women’s male partners. It seems to me unreasonable to consider them study participants in any sense. The data collected on the men included number of sexual encounters, whether they were circumcised, whether they used condoms, and whether they reported having had prior partners. There is no indication that the data collected for the study included any information that could be used to actually identify the men (and this would be unusual); in fact, the study diaries kept by the women were undoubtedly far more discreet that their emails, their blogs, or their chats with friends. Even if having sex with a 19 year-old were an activity that carried a reasonable expectation of complete discretion, there’s no indication that their privacy was invaded.

4. You state that the researchers obviously couldn’t be encouraging condom use, as this “would ruin their research.” Given that the entire reason the study was published in the most prestigious medical journal in the country is precisely that they DID find that condom use might lower HPV infection rates, this statement seems definitely incorrect.

Thanks for giving me an opportunity to comment.