Relationships that maximize HIV transmissions

A couple of months ago, I came across this review of Helen Epstein's book, The Invisible Cure: Africa, the West, and the Fight Against AIDS. The reviewer William Easterly, a well known development economist at New York University, points to an important insight in Epstein's book on why HIV transmission rates are so high in Africa.

The conventional view was that promiscuity contributed to the rapid spread of AIDS in the continent. The World Health Organization and The World Bank suggested in their 1997 report that the infections were because of “core transmitters such as prostitutes and the truck drivers and migrant workers who patronized them”. These views understandably produced a defensive reaction, and explain “the astonishing wall of silence surrounding AIDS that Epstein documents and that still prevails among both the African public and politicians.”

But were promiscuity and the sex trade really the main reasons why so many were infected? Epstein posits a more nuanced theory based on the studies of independent researchers. Its premise is that transmission rates are high because of the prevalence of concurrent long-term relationships in Africa. (In the West too, people tend to have multiple long term partners over the course of a lifetime but these relationships are typically not concurrent.)

In African countries the situation is exacerbated by the fact that both men and women tend to have more than one long term partner at the same time. This differs significantly from other places where women’s monogamy is strictly enforced, often due to religious conservatism. It is tragic and a strange irony that greater sexual freedom for women should aggravate the AIDS problem.

Concurrent relationships in certain regions of Africa are a product of work lifestyles. Men work long durations away from home, hence opening up opportunities for both men and women to have partners on the side:

In southern Africa (where the epidemic is concentrated), one of the few opportunities for gainful work open to men is to become long-distance migrants to the mines. Both husbands and wives may have other long-term partners during the months when they are separated. The African tradition of polygamy (described by historians like John Iliffe as a cultural response to maximize fertility in what used to be a lightly settled continent) has given way to modern relationships between older, well-to-do, gift-bestowing men and multiple young girlfriends. This is not so different from the successive trophy wives of American fat cats, but much more widespread since Africa's poverty often makes it a matter of survival for African young women to have a rich (older) boyfriend. The desire of young women for young boyfriends can be accommodated on the side.

And here’s the crux of the problem - why concurrent relationships are more likely to transmit the virus than one-night stands or serial monogamy:

For many reasons, concurrent, long-term sexual relationships are much more dangerous for the spread of AIDS than serial monogamy. When both men and women have concurrent relationships, they are part of a huge web of sexual partners by which the HIV virus moves through the population. Long-term relationships are much more likely to spread AIDS than one-night stands because of the low probability of a single sex act spreading the virus. Since the HIV-positive are most contagious soon after they themselves become infected, a long-term partner who has just become infected in another relationship poses much more risk than a prostitute who has been infected for a long time. Serial monogamy in the West kept the virus largely trapped within single relationships, a fact Epstein nicely illustrates with some clever graphs. Her explanation based on concurrent relationships has gained broad acceptance and has been confirmed by mathematical modeling and by surveys of sexual habits in various countries; but one still wishes the evidence was a little more extensive for such a critical issue.

From the operations research perspective, there is an intriguing question here: What types of relationships/sexual mores are more likely to cause a greater number of infections in a population? If a comparative simulation modeling exercise were to be done in isolation to analyze this, we might get some insights, though the model will likely ignore such complexities as cultural factors and the prevalence of AIDS education. From the excerpt above, it's clear that mathematical modeling has already been attempted. I’ll have to look the references in the book up to see what's been proposed so far.