Risk and the Rate of Spreading HIV and AIDS

Why are homosexuals at such higher risk of contracting HIV and AIDS than are sexually active heterosexuals? There are a two main reasons.

First, gays tend to have more sex partners than straights.

... the GSS data indicates that the median number of sexual partners since age 18 for heterosexual men is 6, and 10 for homosexual men. The averages, however, are 15 and 37, respectively. Now, it's true that the averages are a better way of figuring out what a majority is doing--but the median gives us a pretty good that the high end of the promiscuity curve for homosexuals could be pretty large. The median for homosexual men is 1.66 times that of heterosexual men; the average, however, is 2.46 times. This suggests that some significant minority of homosexual men have a lot of different sexual partners--although by itself, this doesn't tell us how large or how promiscuous this crowd is.

A small number of gay men have a huge number of sex partners, and these men act like HIV seeds that drop the virus into the community at many distant points. Picturing the gay community as a graph of nodes and vectors, the mean number of sexual partners is more important than the median for determining how quickly the virus will spread because the virus does not lose any power as it jumps from person to person.

A small number of straights have a huge number of sex partners also: they're called prostitutes. The difference between gay men and straight prostitutes brings us to the second reason straights have less of a risk of getting HIV: vaginal and oral sex are much less likely to spread the virus, particularly if the carrier is the woman. Some risk statistics:

Studies of HIV transmission have calculated the risks of infection. The studies came up with very different rates. For example, one study reported the risk for infection from one episode of unprotected receptive anal intercourse with an HIV-infected partner at 1 in 3,333. Another study said 1 in 50 episodes.

For regular partners who were active in anal sex, the risk for transmission was 1 in 10. The risk for the insertive partner (the "top") is believed to be about 10 times less than for the receptive partner (the "bottom").

The risk of HIV infection during vaginal intercourse is believed to be much less. One estimate was 1 in 200,000 for transmission from infected women to men and 1 in 100,000 for transmission from infected men to women.

Because it's hard for HIV to spread from women to men, the virus spreads much less quickly in the heterosexual community -- every second step in the infection chain is a woman. So, for instance, a man M1 who has sex once with a woman W1 who had sex once with a man M2 who has HIV will have a cumulative 1 in 20 billion chance of contracting HIV. A man M3 who has sex once with a man M4 who had sex once with a man M5 who has HIV will have a cumulative chance of contracting HIV between 1 in 11 million and 1 in 1.1 billion (depending on how they had sex).

So, as you can see, even if gays had the same number of sex partners as straights, HIV would spread approximately 200 times as quickly through the gay population. Since gays have around 2.5 times as many partners, HIV will spread approximately 500 times faster among homosexuals than among sexually active non-monogamous heterosexuals. ("Approximately" because it's not as simple as multiplying 200 by 2.5, but that's close enough.) Of course, people who are monogamous or not sexually active will not get HIV in this way; there are probably more chaste straights than gays, which further reduces the relative number of HIV infections per capita within the straight community.

It's good that many gays are waking up to these numbers in the wake of the recent so-called super HIV strain, but unless people are willing to reduce their participation in risky behaviors they won't make much progress. Fear is a powerful motivator, and if you can't take pleasure in doing right, you can at least be afraid of what will happen when you do wrong.

"You can't have a core group of people having sex with large numbers of people without amplifying any sexually transmitted disease that enters the system," said Gabriel Rotello, author of "Sexual Ecology: AIDS and the Destiny of Gay Men." "I don't have any doubt that a resurgent H.I.V. epidemic will hit the gay population in the near future," he said. ...

Until people really believe an unstoppable virus is out there, he said, they will continue to indulge in unsafe sexual practices. "People are not going to modify their sexual habits in ways that are difficult or unpleasant until they see their friends dying again," he said. "And to me that's just an unbelievably depressing thought."

Update:
David Orland asks why men and women in sub-Saharan Africa tend to have nearly equal infection rates, and links to tihs page about women and AIDS which says (not about Africans specifically):

This may already be happening - data from the CDC in America shows that among teens, girls accounted for more than half of new HIV infections reported in 2001. Globally, women make up 60% of the 15 – 24 year olds who are HIV+.

But note that in a heterosexual community, both genders will get infected at the same rate even though it's easier to spread the virus from male to female than from female to male. As long as people only have sex with those of the opposite gender, every second link in the infection chain will be a man, and every second link will be a woman. If there's a significant difference in the number of partners between men and women (median, this time) there may be a slightly different rate of infection for each gender, but ultimately an equilibrium rate will be established that will be the same for both genders.

Since many HIV infections in Africa are now spread from mother to child at birth, it's likely that the infection rate in Africa has reached an equilibrium with between one-third and one-half infected with the virus. That's an astonishing infection rate, and it probably represents nearly the entire sexually active population, plus lots of children.