The decision by pharmaceutical companies to withdraw their patent suit in South Africa removed an important obstacle to access to antiretroviral treatment of HIV/AIDS in the developing world. Drug copiers, such as Cipla Ltd. of India, have offered to provide a three-drug combination for about $350 instead of the $10,000 to $15,000 paid in the United States for one year's treatment at patent-protected prices. Even at Cipla's price, however, treatment of HIV/AIDS raises serious questions of feasibility for very poor countries.

But treatment of people with HIV/AIDS is not only the compassionate solution, it is also cost-effective. The cost of not treating AIDS includes the burden of opportunistic infections in, and early death of, 25-35 percent of the work force in the next decade. The additional impact on the economy and society of a generation of 6-year-olds raised by 11-year-olds cannot be adequately addressed in macroeconomic models of the cost of HIV/AIDS. Access to treatment is a necessity for Africa.

What might be overlooked, however, as life-sustaining drugs become available, is the fact that prevention is still by far the more compassionate and more cost-effective answer. Prevention does not replace treatment, but it does reduce the number of people whose lives will depend on expensive drugs with significant side effects. The key to a good prevention program is understanding the dynamics of disease transmission. But most prevention programs have been extremely narrow in scope. International AIDS policy derives from an erroneous preconception that the high rate of HIV transmission in Africa is primarily an issue of sexual behavior that can be addressed by behavior modification. That policy has met with some success in the United States and other rich countries. In Africa and in similarly impoverished populations, however, biological factors that result from poverty play a determining role in the high rates of HIV transmission. Failure to recognize those biological factors and to integrate them into a model of transmission for poor populations has led to prevention strategies that have failed and to the perpetuation of racial stereotypes.

Sexual behavior is obviously an important factor in the transmission of sexually transmitted diseases. Education programs have been very successful in Uganda in reducing incidence (new cases) in recent years. But sexual behavior alone cannot explain HIV prevalence as high as 25 percent of the adult population in some African countries and less than 1 percent in the United States. The presumption that high rates of partner change explain the very high rates of HIV is generally not explicit. But it is the implicit assumption behind an AIDS-prevention policy that consists of behavior modification and condom provision, with some recent attention to sexually transmitted diseases (STDs). There has been very little analysis of the myriad factors that influence HIV transmission in a profoundly poor region.

The question is straightforward: How is AIDS different in Africa? Instead of being addressed with mainstream methods of scientific inquiry, the AIDS-in-Africa debate was hijacked in the 1980s by an extreme behavioralist explanation that transformed the question into: How are Africans different? So ingrained in Western thought is the notion of racial difference that the distortion of the question went unnoticed, and the behavioral approach has dominated both research and policy for more than a decade.

The justification for a behavioral explanation of African AIDS was found not in careful empirical studies but in hypothetical arguments based on a powerful racial metaphor that portrays Africans as a special case. The literature that formed the basis of AIDS policy is characterized by sweeping generalizations about an imagined pan-African culture of sex as a commodity and fertility as a duty to one's ancestors. Unlike most scientific and social science work, the literature on African sexuality relies on suggestive language, double-entendre and innuendo. It also resembles a pyramid scheme, since almost all works cite the same few anthropological studies of the 1980s that designate Africa as the "domain of Homo Ancestralis," a hypothetical early stage of human development in which sexual relations were unencumbered by Eurasian patriarchy. The social science literature on "African sexuality" exhibits much lower standards of proof than would be required for publications on European or North American subjects. The momentum of Western stereotypes clothed in social science jargon compensates for the empirical weaknesses of the works. The use of species terminology in the designation of Africans as Homo Ancestralis is a metaphor that inescapably carries the intellectual baggage of nineteenth-century racial science, which maintained that Africans were a genetically distinct species from Europeans. The ease with which the image of hypersexualized Africans was incorporated into AIDS policy indicates the persistent force of this nineteenth-century racial view. The emphasis on racial difference is evident in the terminology used to describe multipartnered sex in Africa and in the West. Africans are described as being polygamous, which lends an exotic air and marks Africans as the social "other," whereas Westerners simply have multiple partners.

By the almost exclusive emphasis on behavior modification, AIDS research and policy for sub-Saharan Africa implicitly incorporate the assumption that behavior explains the differences in HIV prevalence between African countries and rich countries without examining its implausible conclusions. How much sex are we talking about that would produce, in the absence of other factors, prevalence of HIV in Botswana that is over fifty times that of the United States, eighty times that of France and 1,000 times that of Cuba? It was not until 1999 that the central tenet was tested; a study published by UNAIDS showed no correlation between rates of sexual behavior and prevalence of HIV.

There are significant levels of unprotected multipartnered sex in the United States and Europe, as evidenced by serious epidemics of other STDs, such as herpes-2 and chlamydia. In spite of the level of unprotected sex, there has not been a heterosexual epidemic of AIDS in the rich countries. Among otherwise healthy, well-nourished people in industrialized countries, heterosexual transmission of HIV is relatively rare–about one in 1,000 contacts between an HIV-positive female and an HIV-negative male, and about one in 300 contacts between an HIV-positive male and an HIV-negative female.

Since the African epidemic is heterosexual, it is clearly different from that of the West, a fact that was noticed by South African President Thabo Mbeki. He had the temerity, as some Western scientists viewed it, to ask how conditions of poverty in Africa affect the development of HIV/AIDS. It was a very conventional epidemiological question, well within the bounds of standard research. His question was treated, however, as heresy by at least a vocal minority of mainstream scientists, who seemed to regard any inquiry as beyond his prerogative. Spurned by the mainstream, Mbeki turned to scientists on the fringe of AIDS research who were eager for an audience. South Africa lost valuable time in addressing the AIDS crisis as the government vacillated and explored dead ends.

Mbeki's question, however–how does poverty influence AIDS in Africa?–is neither inappropriate nor unscientific. The environment in which any infection is transmitted in poor countries is very different from that of the United States and Europe and is strongly influenced by poverty, malnutrition, bad water and poor access to preventive and curative care. The standard epidemiological approach to understanding disease is characterized by Louis Pasteur's comment: "The microbe is nothing, the terrain everything." In other words, pre-existing health conditions play a key role in susceptibility to disease. We should expect HIV/AIDS to develop differently in rich and poor countries, just as do tuberculosis, pneumonia, measles and nearly all other infectious diseases.

Ironically, mainstream biological science has the answers to Mbeki's questions, but the specialized and conservative nature of biomedical research inhibits any one scientist from coming forth with the solution, or perhaps even acknowledging it. Most social science has failed to incorporate biomedical data into its analysis, in spite of the obvious fact that HIV and AIDS are biological conditions. What has been missing is an interdisciplinary approach that incorporates biological and social data into an analysis of the social context of HIV disease in Africa.

Sex tends to be distracting, for researchers as much as for the general public. The sexual transmission of HIV diverted attention from the broader epidemiological environment in which a heterosexual epidemic developed in sub-Saharan Africa. Both rich and poor countries are characterized by high rates of unprotected multipartnered sexual activity. Populations in poverty are also characterized by malnutrition, parasite infection and lack of access to medical care and antibiotics for bacterial STDs, which are important co-factors for transmission of HIV. To acknowledge the synergistic relationship among malnutrition, parasite infestation and infectious disease is not to say that AIDS itself is a nutritional disease. Nor does it deny that HIV is sexually transmitted in Africa and causes AIDS. It merely subjects STDs, including HIV/AIDS, to the same methodology employed in the study of other infectious diseases, however transmitted.

Even a brief survey of economic conditions in sub-Saharan Africa in the years in which the AIDS epidemic began reveals an extremely compromised health environment. From 1970 to 1997, sub-Saharan Africa was the only world region to experience a decrease in food production, calorie supply and protein supply per capita. In ten countries (including Zimbabwe, Kenya, Uganda, Zambia and Malawi), protein supply fell by more than 15 percent. Eighteen of the nineteen famines worldwide from 1975 to 1998 were in Africa, and 30 percent of the total population of the region was malnourished. Refugees from internal and external conflicts crowded into unsanitary camps where food rations were deficient in necessary nutrients. Sub-Saharan Africa is not the only region in which malnutrition is associated with HIV/AIDS. Among all low- and middle-income countries, HIV prevalence is strongly correlated with falling protein consumption, falling calorie consumption, unequal distribution of national income and, to a lesser extent, labor migration. Almost all of sub-Saharan Africa is tropical, with a very high prevalence of parasite infection, including malaria, schistosomiasis and various intestinal and skin ailments.

There is a large body of mainstream biomedical literature that documents the mechanisms by which malnutrition and parasite infection undermine the body's specific and nonspecific immune response. Protein-energy malnutrition (general calorie deficit) and specific micronutrient deficiencies, such as vitamin-A deficiency, weaken every part of the body's immune system, including the skin and mucous membranes, which are particularly important in protecting from STDs, including HIV. Parasite infestation plays a dual role in suppressing immune response. It aggravates malnutrition by robbing the body of essential nutrients and increasing calorie demand. Moreover, the presence of parasites chronically triggers the immune system, impairing its ability to fight infection from other pathogens.

Poverty not only creates the biological conditions for greater susceptibility to infectious diseases, it also limits the options for treating disease. Infection with other STDs is an important co-factor for transmission of HIV; genital ulcer diseases in particular, such as chancroid, provide an entry point for HIV. Such painful bacterial STDs are relatively uncommon in rich countries because of the availability of antibiotics. In Africa, South Asia and Latin America, however, even when poor people have access to healthcare, the clinics may have no antibiotics to treat bacterial STDs that act as co-factors for AIDS. These are among the conditions we have to consider in poor countries, and they are standard variables in epidemiology.

Treating African AIDS as a special case caused by a hypersexualized culture obviously reinforces racist stereotypes and pessimism over Africa's future. It also pushes AIDS policy to an almost exclusive reliance on behavior modification and condom use and away from general health and nutrition, and it gives us little preparation for similar epidemics that are now incubating in South Asia and Latin America. Some parts of India already have substantial epidemics; one out of every nine HIV-positive people in the world is Indian. Nascent or concentrated epidemics throughout Latin America, in combination with deteriorating economic conditions, produce conditions as propitious to HIV as those in Africa ten years ago. By 1997, after several years of high unemployment, the prevalence of HIV in Buenos Aires was already what it had been in Zambia ten years earlier. At least a quarter of Latin Americans and Caribbeans cannot obtain 80 percent of their minimum calorie requirement even if they were to spend their entire income on food, and almost half the population of the region lives in poverty.

Social conditions in Latin America aggravate the population's vulnerability to HIV/AIDS. As in Southern Africa, highly concentrated land ownership forces millions of workers to migrate for work, internally or internationally, increasing risk of HIV and other STDs through new sexual liaisons, including prostitution. Forty million children in Latin America live on the street; they eat from garbage cans, and many of them sell sex to survive. Sex tourism has shifted from Asia to the Americas, with children as the primary targets of an Internet-based industry. Considering the extent of poverty and the immunological effects of that deprivation, an AIDS epidemic of African magnitude is possible in parts of Latin America. Throughout the region, the preponderance of new cases has shifted from upper to lower income and from men to women and children. Based on the demographics of those recently infected, Latin American AIDS is already "Africanized."

The policy implications of understanding the broader causes of the AIDS epidemic in Africa are reason for both optimism and pessimism. Reducing HIV transmission requires health education, availability of condoms and also a broad assault on malnutrition, diarrheal diseases and parasitic diseases, including malaria and schistosomiasis. To treat those already infected will require upgrading the health services infrastructure and expanding health education. The steps that are necessary for both prevention and treatment of HIV/AIDS are the same as for addressing the other health and development needs of poor countries.

Identifying those needs is relatively easy, but they have not been adequately addressed in the past–not by the governments of poor countries or by their bilateral and multilateral aid partners. Even that dismal acknowledgment, however, should be balanced by a recognition of how relatively inexpensive some solutions are. Vitamin A supplementation sufficient to prevent blindness and other deficiency diseases for one person for an entire year costs less than one condom. A year's supplementation with vitamin A, iron and iodine costs less than ten condoms. To fortify the food supply with iron to prevent anemia and increase disease resistance and work capacity costs about 20 cents per person per year. Attending to broader health concerns is not as expensive, or as hopeless, as it might seem. There are also serious weaknesses in a prevention plan that relies exclusively on provision of condoms, even with health education. It does not address women's lack of power in sexual relationships, nor the irrelevance of condoms to most people after a few beers. Strengthening immune systems will help to protect people from some of the consequences of unsafe sex and from other infectious diseases as well. What it will take to prevent HIV transmission and to treat people with HIV/AIDS is no less, but no more, than what has been needed all along in sub-Saharan Africa and other poor regions. It would have been cheaper to provide the infrastructure, the nutrition, the education and the medicines before HIV/AIDS, but it is still a bargain calculated in both compassionate and cost-effective terms.