The following post was inspired by the seventy-first suggestion in No One Cares What You Had for Lunch: 100 Ideas for Your Blog, which was randomly selected by Alison Headley of bluishorange.
I used to work in the field of HIV and AIDS research. So when a friend of mine recently discovered that he was to be the father of a baby boy, he sent me the following email:
What do you know about connection between circumcision and reduced chance of acquiring HIV? We hadn't even considered circumcision until we heard about the study, but now we're wondering about it
To which I responded:
I don't want to discount the HIV transmission thing, but, in your case, I don't know that I'd put a lot of stock in it either. There was a lot of talk about this study back when I was working at the lab, and I don't dispute the findings. But bear in mind that these trials were conducted in areas where HIV was prevalent, and where the participants were engaging in "high risk" behavior (multiple partners, unprotected sex, etc).
If you educate your kid to take precautions against HIV, and he lives in an area where HIV isn't rampant, and he's monogamous (or just-a-fewgamous), his being circumcised might only decrease his overall chance of infection from "pretty low" to "pretty low minus a smidge" (as opposed to, say, a Kenyan trucker who has lots of sex with multiple, concurrent partners, and whose circumcision is his only form of "protection.").
But, before hitting send, it occurred to me that I might not know what I was talking about. I mean, yes, I used to work in the field of HIV and AIDS research, but only as a programmer — not as one of the genius who actually design the clinical trials or analyze the results.
So I sent my friend’s question, and my reply, to M, a statistician I know who still works there. Here’s what she had to say:
To assess individual risk, one would need to account for many characteristics and behaviors on the individual level. In most clinical studies, such as those conducted regarding male circumcision, data is collected on risk factors associated with the outcome (HIV infection in this case) and the exposure (circumcision, say). In stat. analyses, we adjust for these factors so that we can come up with a reasonable estimate for the risk of the exposure accounting for all the other potential "confounding" characteristics/behaviors on a population level.
For the individual, however, his/her risk is highly dependent on his/her individual profile of risk. Circumcision, for example, is only one characteristic of a man that might put him at risk for HIV infection. There exist a plethora of others such as HIV status of his sex partners, numbers of sex partners, alcohol/ drug use, injection drug use, condom use, etc.
Your points are well made regarding the differences between heterosexual risk in the US and that in high HIV prevalent areas such as sub-Saharan Africa in that in the US, among heterosexual males, the risk of transmission is much lower since the prevalence of HIV among all people is lower. For example,
"In 2004, men who have sex with men (MSM) (47%) and persons exposed through heterosexual contact (33%) accounted for an estimated 80% of all HIV/AIDS cases diagnosed in areas in the U.S. with confidential name-based reporting. Blacks accounted for 49% of cases and Hispanics for 18%. Infection rates in both groups were several-fold higher than that in whites. An overall prevalence of about less than 0.5% was estimated for the general population ."
See http://www.cdc.gov/hiv/resources/factsheets/circumcision.htm for more details.
In observational studies in the US, however, the relative risk of HIV infection among non-circumcised men was typically two-fold that of circumcised men. So, whether you are a man in Africa driving a truck and having sex with many women, or you are a man in the US having sex with one woman, if you are having sex with an HIV+ woman and you are not circumcised, you are pretty much at the same level of risk for HIV, with all other characteristics being equal. One issue that has not been determined, though, is whether or not the different clades of HIV strains could have an impact on the susceptibility of acquisition. If there were a difference, the geographic location (i.e. who you were having sex with and thus, what strain of virus they have), could have an impact on acquisition.
In other words, your risk is highly dependent on your own personal behaviors, rather than the population's behaviors. We use the population stats to help us understand, in general, what behaviors on the individual level will put us at higher risk than other behaviors. But, we cannot quantify an individual's risk based on population numbers unless we design a study in such a way to make these calculations possible.
There is, of course, many other considerations regarding male circumcision for babies, such as risk of infection, pain, etc. to go through the operation. This should be weighed with the benefit (as you pointed out to your friend) of other harms (such as HIV, as well as other sexually transmitted diseases such as Human Papillomavirus, Gonorrhea, Herpes, etc.). Also safe sex practices (using a condom!), in general, will most likely outweigh any risk of not being circumcised if this baby boy grows up to be a real swinger (either with men or women)!!
Hope that helps. Also see http://www3.niaid.nih.gov/news/QA/AMC12_QA.htm for another website to consult, and the press release on the trial in Kenya and Uganda.
M also asked that I add the following disclaimer: “This was written by an anonymous, somewhat crazy biostatistician-woman who happens to have some extended experience researching HIV/AIDS, among mostly & ironically, Men who have Sex with Men (MSM). Please take her words regarding the male genitalia, and what should be done with it, with a grain of salt!”
Me, I wouldn’t think that circumcision and salt would go together but, like I said, I’m just a code-jockey.