The rate of infection in some parts of the continent is 100 times higher than in the United States, yet sexual activity is similar. Epidemiologists, forced to reconsider their theories of how the disease spreads, have come up with surprising new insights.
Botswana seems an unlikely place for an AIDS epidemic. Vast and underpopulated, it is largely free of the teeming slums, war zones, and inner-city drug cultures that epidemiologists say are typical niches for the human immunodeficiency virus. Botswana is an African paradise. Shortly after gaining its independence from Britain in 1966, large diamond reserves were discovered, and the economy has since grown faster—and for longer—than that of virtually any other nation in the world. Education is free, corruption is rare, crime rates are low, and the nation has never been at war. Citizens are loyal: A visitor quickly learns that even mild criticism of anything related to Botswana is considered impolite. Yet this country, with all these advantages, has the highest HIV-infection rate in the world. Check out NYTimes article on AIDS progress.
The virus has spread extremely rapidly in Botswana. Two decades ago, virtually no one there was HIV-positive. By 1992 an estimated 20 percent of sexually active adults were infected. By 1995 that proportion had reached one-third, and today it is roughly 40 percent. In Francistown, Botswana’s second largest city, nearly half of all pregnant women in the main hospital test positive for HIV. The picture in the rest of sub-Saharan Africa is nearly as dire. AIDS has killed Zulu nurses in South Africa, Masai teachers in Tanzania, Kikuyu housewives in Kenya, Pygmy elders in Uganda. HIV infection rates range from around 6 percent in Uganda to 39 percent in Swaziland.
Such numbers are astronomical compared with most of the world. In the United States, less than 1 percent of the population is infected; in Russia and India the figure hovers around 1 percent. Even in Thailand, with its thriving sex and drug trades, the proportion of infected barely exceeds 2 percent.
The high rates come despite efforts in many communities to stem the HIV epidemic through educational programs, condom distribution, and treatment for such sexually transmitted diseases as gonorrhea and syphilis, which create genital sores and ulcers that make it easier for the virus to spread. In most cases these programs have had little effect. The growing disaster has forced AIDS experts to reconsider old theories about how HIV spreads in Africa.
Outside of sub-Saharan Africa, many HIV-positive people are injecting drug users, prostitutes, and highly promiscuous homosexual men who may have hundreds of different sexual partners every year. But most Africans with HIV claim never to use drugs, engage in prostitution, or have large numbers of sexual partners. To explain the high infection rates, scientists have advanced theories ranging from nutritional deficiencies to more virulent HIV strains to different sexual customs. In the 1980s Australian demographer John Caldwell insisted that the virus was spreading rapidly in Africa simply because people there tended to have more sexual partners than people elsewhere. He pointed to the cultural desire for many children, the tradition of polygamy, and other aspects of African society that contributed to a greater tolerance of promiscuous behavior than in the West. Caldwell’s views sparked controversy and for years received little attention. Recently, though, some experts, including epidemiologist James Chin of the University of California at Berkeley, have revisited the theory. Chin believes it’s the only possible explanation: “People tell me not to say it, but I strongly believe it.”
Some studies do show that Africans have more—but not vastly more—sexual partners, on average, than people in Western countries. For example, a study of sexual behavior in Zimbabwe, where roughly 33 percent of adults are HIV-positive, found that in a single year, most people have between one and three sexual partners. Of course, prostitutes in Zimbabwe may have more than 100 partners a year, just as prostitutes elsewhere in the world do, but most HIV-positive Zimbabweans are not prostitutes.
In the early 1990s, Martina Morris, then a member of the sociology and public-health departments at Columbia University (and now a professor of sociology and statistics at the University of Washington in Seattle), tried to solve the mystery of HIV in Africa mathematically. She had helped devise a computer program to predict the spread of HIV in a given population based on such factors as the number of sexual partners people had and the duration of those relationships. At the time, Uganda had one of the highest HIV-infection rates in the world, so she flew there in 1993 to gather data on sexual behavior.
“Just after I arrived in Uganda, I had to give a lecture to Ugandan doctors at the medical school in Kampala, telling them what I planned to do,” she recalls. “At the time there was talk in Uganda about helicopter scientists—whites from the United States and Europe who just parachuted in, took data, and didn’t work with local African experts. I was the only American woman in the room, and it was a tough audience. The HIV rate was estimated to be 18 percent at the time, and here I was trying to explain how mathematical models were going to help. They listened, and then at the end, one man raised his hand and asked, ‘Could your model handle more than one partner at a time?’ I said, ‘No.’ The man walked out. The others sat down with me and said I had to include concurrent partnerships in my model. Otherwise it would be irrelevant.”
The idea that long-term simultaneous partnerships might increase the spread of HIV was first proposed by British epidemiologists Robert May and Charlotte Watts in 1992. But Morris had not seen their article when she set out for Uganda in 1993, and her mathematical tools were not up to the complicated task of modeling multiple long-term partnerships anyway.
An added difficulty was that Morris would be asking Ugandans to answer intimate questions about their sexual behavior. So she replaced the impersonal language of standard questionnaires with a structured conversation. She asked the respondent whom he or she last had sex with, how the couple met, how long they had been together, whether they were still together, and so on. Then she asked about the respondent’s previous sexual partner. “Respondents love it, because it’s really like gossip,” Morris explains. “In a way, people are telling the story of their lives.”
Morris then conducted similar surveys in Thailand and the United States—with fascinating results. She found that the average Ugandan and the average American claimed roughly the same number of sexual partners in their lives. About 25 percent of people (of both sexes) in both countries said they had more than 10 partners in their lives. But similar rates of promiscuity did not result in similar rates of infection. The HIV rate in Uganda peaked at 18 percent in the early to mid-1990s but never exceeded 1 percent in the United States. And in Thailand, where many more men—65 percent—reported 10 or more partners, the HIV rate barely rose above 2 percent.
A key difference between Uganda and Thailand, Morris found, is that men in Uganda often maintained two or more long-term sexual relationships at once. In Thailand, most men had only one long-term sexual relationship—with their wives. Half the Thai men in Morris’s survey said that they had sex with prostitutes but rarely the same one twice. On average they saw five prostitutes each year. Although many Thai prostitutes are HIV-positive, the men’s risk of infection was relatively low because Thai men generally had sex with each one only once.
The likelihood of contracting the virus during a single sexual act is believed to be quite low, between 1 in 100 and 1 in 1,000. So if an HIV-positive man has sex once with hundreds of different uninfected people, chances are he will infect only one of them. Generating an HIV epidemic such as Uganda’s probably requires that people be exposed to the virus repeatedly. As Morris discovered, Ugandan men and women had sex many times over many years with each of their partners. If one of those partners was HIV-positive, the relationship would prove very risky over time.
In the United States, Morris found a different pattern. Heterosexual Americans, like Ugandans, tend to have several long-term relationships, but they usually have them sequentially, not at the same time. If an American contracts HIV, she probably won’t pass it on right away, and if she eventually does, her new partner probably won’t pass it on right away either.
In 1993 Morris teamed up with mathematician Mirjam Kretzschmar of the National Institute of Public Health in Holland to develop a new computer program that could model simultaneous partnerships. So far she has run the program with data from Uganda, Thailand, and the United States, and the simulations reproduce the same prevalence of HIV observed in those countries in the early 1990s, when the data were collected.
Morris contends that Africans in ordinary heterosexual relationships are linked, not only to each other but also to the partners of their partners’ partners—and to the partners of those partners—via a web of sexual relationships extending across huge regions. If one member contracts HIV, then everyone else may too. Anti-AIDS campaigns warn against contact with prostitutes, but Morris says simultaneous long-term relationships are far more dangerous.