Pervasive public health messages warning about HIV from sex can make it hard for people who know of one or more unexplained infections to get neighbors, reporters, clergy, and others to believe the infections did not come from sex, and that their source should be investigated. This chapter provides evidence and references that anyone can use to challenge the widely believed myth that sexual transmission accounts for almost all HIV infections among adults in Africa.

As reviewed here, the best evidence indicates that far less than half of HIV infections in Africa come from sex. Of course, sex is a risk for HIV, but that says nothing about how much comes from sex. Similarly, getting hit by a car is a risk, but that does not mean cars kill everyone. In both cases, people should take care to protect themselves even if the risk in question accounts for a minority of infections or deaths.

How the myth got started

Racial stereotypes of sexual behavior played a major role in early speculation about Africa’s HIV epidemics, even within medical journals. For example, a 1986 editorial in the Journal of the Royal Society of Medicine attributed differences between HIV epidemics in the US and Europe vs. Africa to “the much lower contact rate… among North American and European heterosexuals.”[1] A 1987 paper in the Review of Epidemiology states: “[M]ost traditional African societies are promiscuous by Western standards. Promiscuity occurs both premaritally and postmaritally” and “seems to be the most important cultural factor contributing to the transmission of HIV in Africa.”[2] In 1988, the head of WHO’s Global Programme on AIDS along with the future head of UNAIDS and others authored an article in Science claiming: “[S]tudies in Africa have demonstrated that HIV-1 is primarily a sexually transmitted disease and that the main risk factor for acquisition is the degree of sexual activity with multiple partners, not sexual orientation.”[3]

In 1988, WHO experts estimated, without presenting any supporting evidence, that roughly 90% of HIV infections in adults in Africa came from heterosexual sex.[4] That estimate disagreed with evidence. As of 1988, the substantial amount of evidence already available about risks for HIV infections in Africa suggested bloodborne risks may have been causing more infections than sex.[5] That evidence includes: children with unexplained infections, both new and old HIV infections more common in people with recent medical injections, and infections not concentrated in sexually more active adults.

Soon after international experts committed to the myth, information from surveys about sexual behavior in countries around the world disagreed with racial stereotypes. A 2006 summary of information from such surveys reported a “comparatively high prevalence of multiple partnerships in developed countries, compared with parts of the world with far higher rates of sexually transmitted infections and HIV, such as African countries…”[6] Nevertheless, the myth lives on. Many people continue to think that differences in sexual behavior in Africa vs. the rest of the world explain Africa’s HIV epidemics.

15 experts can’t find evidence linking most HIV to sexual risks

During 2002-3, articles in medical journals along with presentations at a closed-door meeting at WHO challenged the idea that sex accounts for almost all HIV in Africa (see Chapter 3). In response, WHO and UNAIDS staff led a team of 15 authors publishing a rebuttal in The Lancet medical journal in 2004. The authors claimed: “epidemiological evidence indicates that sexual transmission continues to be by far the major mode of spread of HIV-1 in the [Africa] region.”[7]

What evidence? The only evidence they presented linking HIV to sexual behavior was the age distribution of HIV-positive Africans: “the prevalence [percent infected] in children aged 5-14 years… was much lower than the prevalence in adolescents and adults aged 15 years or older.” Women, especially, get more HIV infections beginning in their late teens. But does that show HIV comes from sex? Other risks change with age. From their late teens, women get more blood exposures from reproduction-related health care and maybe also from cosmetic services.

If the ages of those who get HIV is sufficient to show infections come from sex, then by that same argument, tuberculosis and even parking tickets are sexually transmitted.

The meager evidence presented to link HIV to sexual behavior is telling. The 15 authors had between them been involved in dozens of studies in Africa looking at sexual behavior as a risk for HIV infection. The problem for the 15 authors was that the huge body of available evidence on sexual behavior vs. HIV infection disagreed with their claim – adults reporting no partner, one HIV-negative partner, or 100% condom use were getting HIV.[8] As an excuse for ignoring disagreeable evidence, the authors explained “data on sexual behavior are notoriously imprecise.” Of course, many studies show some infections come from sex, but that is not at all the same as showing a majority of infections – much less almost all infections – come from sex.

Trace, test, sequence

In the years after WHO’s and UNAIDS’ 2004 defense of the view that sex explains almost all HIV in Africa, a lot of new and unsupportive evidence has come from studies that traced and tested sexual partners. At least three different study designs contribute useful evidence.

Sequencing HIV from a community

One of the best ways to see how much sex contributes to Africa’s epidemics is to collect blood samples from a large percentage of HIV- positive people in a community, sequence collected HIV (determine the order of each HIV’s constituent parts), and then look for similarities among sequences. If HIV sequences from two people are similar, one infected the other directly or indirectly through one or more other people. If people with similar sequences are or have been sexual partners, that is evidence one infected the other through sex.

Three recent studies sequenced HIV collected from large percentages of HIV-positive people in study communities in Africa, identified clusters (pairs or larger groups) of sequences that were similar, and compared clusters with information about sexual partnerships. What they found does not at all fit the view that sex explains almost all HIV infections in adults.

Specifically, the three studies identified sexual partnerships that could explain only 1.8% to 6.6% of the HIV infections they sampled and sequenced from each community (Figure 6.1). All identified sexual partners with similar sequences were spouses or long-term partners living together. Studies did not ask about and identify short-term partners. Of course, some HIV infections come from short-term partners, but if sex accounted for most HIV infections in adults, such partners would have to infect many times more people than long-term partners, which is absurd.

One of the three studies looked at HIV collected from a community in Botswana, and two examined HIV from communities in Uganda. Here are some details.

The study in Botswana (Figure 6.1) sequenced HIV samples collected in 2010-13 from 833 adults in northeast Mochudi, a town about 50 kilometers north of Gaborone, the capital. These HIV samples represent almost half of all HIV-positive adults aged 16-64 years in the community.[9,12] Among the 833 HIV sequences, 511 were not similar to any other sequence from the community, whereas 322 were similar to one or more other sequences. In other words, 322 HIV infections were linked directly or indirectly (through other infections) to one or more infections in the community.

Thirty of these 322 sequences were in 15 pairs linking a man and a woman in a household. The study does not say if these household pairs were from spouses, but I assume so. Assuming the 15 pairs linked spouses, the study identified a sexual explanation for 15 infections – one spouse likely infected the other – but says nothing about how the first spouse got HIV. In other words, the study identified a sexual source for 1.8% (= 15/833) of HIV samples collected and sequenced. Aside from these 15 pairs, the study reports no sexual partnerships to explain why 292 (= 322 – 30) sequences from Mochudi were similar to other Mochudi sequences.

Two studies in Uganda (Figure 6.1) sequenced HIV samples collected from adults aged 15-49 years in Rakai District in southern Uganda. The first study sequenced HIV collected in 2008-9 from 1,099 adults, an estimated 42% of HIV-positive adults in 46 selected communities across Rakai District.[10] Among the 1,099 HIV sequences, 890 were not similar to any other sequence from the community, while 209 were similar to one or more other sequences. Fifty-one husband-wife pairs had similar sequences, providing a sexual explanation for 4.6% (= 51/1,099) of HIV infections with sequences in the study. Aside from 51 spouse pairs with 102 sequences, the study reported no sexual link for 107 (= 209 – 102) HIV infections with sequences similar to one or more other sequences.

The second study of HIV sequences from Rakai District used “deep sequencing,” a more advanced technique that gives a more reliable picture of which HIV infections have direct or near-direct transmission links.[11] The study deep sequenced HIV samples collected during 2011-15 from 2,652 adults, an estimated 35% of HIV-positive adults in 40 selected communities across Rakai. Comparing sequences, the study assigned 1,334 sequences into clusters with one or more similar sequences. These clusters included 176 pairs linking couples. This provides a sexual explanation for 6.6% (= 176/2,652) of HIV infections with sampled sequences. There were no known sexual links to explain any of the other HIV sequence clusters.

Moreover, two of the three studies – the two from Rakai – reported percentages of married couples with similar or dissimilar HIV sequences. In both studies, when husbands and wives were both HIV-positive and their HIV was sequenced, only about half of such couples had similar HIV and therefore linked infections.[10,11] The study with deep sequencing reported the highest percentage: 53% of couples with HIV sequences from both partners had similar HIV. In other words, almost half of such couples had dissimilar HIV, showing that husbands and wives had gotten their HIV infections from different sex or blood risks.

Following people to see who gets HIV

Beginning in 1987, dozens of studies tested various ways to protect adults in Africa from HIV infections, for example, warning them about sexual risks or treating other sexual infections. These studies followed HIV-negative men and women, meeting them at intervals to see who had gotten a new HIV infection and to ask about risks.

As reported through August 2011, 44 such studies followed more than 120,000 adults for an average of almost two years each and saw more than 4,000 new HIV infections. They traced only 9.8% (393) of more than 4,000 new infections to HIV-positive sex partners.[13] To confirm new infections came from sex partners, some studies sequenced HIV from both partners. After sequencing, studies found similar HIV in sex partners for only 4.6% (186) of more than 4,000 new infections recognized in the 44 studies.

Tracing and testing contacts

For diseases such as tuberculosis and syphilis, tracing and testing is a time-honored strategy for public health agencies to find and treat new cases and to protect people who are not infected. In recent years, the strategy has been applied to HIV in Africa.

In the eight countries with the worst HIV epidemics – with more than 10% of adults infected – HIV testing programs during 2016-18 traced and tested more than 400,000 adult contacts of people who tested HIV-positive (mostly sex partners, but also needle-sharing contacts).[14] In five of these eight countries, the percentages of contacts found to be HIV-positive was less than the estimated percentages of all adults in the country who were HIV-positive in 2019 (Figure 6.2). Such results do not support the myth that almost all infected adults got HIV through sex.

Sexual risks cannot explain so many new infections in young women

Another way to assess the contribution of sex to Africa’s HIV epidemics is to look at groups getting a lot of new infections for which there is good information about sexual risks. In six of the eight countries in Southern Africa with the worst HIV epidemics, national surveys during 2016-17 reported women aged 15-24 years getting HIV from all causes at rates ranging from 0.46% to 1.67% per year (Figure 6.3). For those countries and years, there is sufficient information about sexual risks to estimate how much HIV they got from sex.

In South Africa, for example, in a 2016 national survey, 33% of young women aged 15-24 years said they had sex at least once in the last four weeks.[25] Although the specific women who are sexually active changes from month-to-month, considering all young women together and all months, the sexual risks that go with this behavior would be similar to 33% of all young women having sex regularly (column A Table 6.1).

What percentage of their partners were risks to transmit? First, consider partners’ ages. In repeat national surveys during 2002-12, 31%- 40% of sexually active women aged 15-24 years reported a recent partner at least five years older than themselves.[28] To make things simple, and to err on the high side, I assume all partners were aged 25-29 years. Second, the same 2017 South African national survey that reported new infections in women reported 12.4% of men aged 25-29 years were HIV- positive, of which 58.5% were not virally suppressed (had at least 1,000 HIV per milliliter of blood, or at least 50 HIV per drop) and were therefore a threat to transmit sexually (columns B and C in Table 6.1).[19]

For women with regular sexual exposure to HIV, how many got HIV in a year? From five studies in Africa that followed couples in which most or many wives did not know their husbands were infected, wives at risk got HIV at the rate of 11.1% per year (see Table A2.1).

From these data and estimates, an estimated 0.27% (= 33% x 12.4% x 58.5% x 11.1%) of young women in South Africa got HIV from sex in 2017 (column D in Table 6.1). Table 6.1 provides similar data and calculations for all six countries.

Figure 6.3 compares the estimated rates women got HIV from sex during 2015-17 (from Table 6.1) to the rates they got HIV from all risks as observed and reported by national surveys during the same years. In South Africa, for example, the estimated rate women got HIV from sex (0.27% per year) was much less than the observed 1.51% per year rate they got HIV from all risks. Across all six countries, the estimated rates women got HIV from sex were much less than the observed rates at which they were getting HIV from all risks.

If most HIV infections in young women did, indeed, come from sex, one or more of the component estimates I have used to estimate women’s infections from sex must be far too low. But these estimates may also be too high. For example, I ignored condom use. Among young women who reported at least one sex partner in the previous year, from 44% to 59% said they used a condom at last sex (data from the same sources used in Table 6.1[22-27]). Did women under-report their sexual activity? Considering that more than half of young women were aged 15- 19 years, there is not a lot of room for the percentages of women having sex regularly to be much greater than reported in Table 6.1.

Slow transmission through sex

Slow HIV transmission through heterosexual sex can be seen in several types of evidence from sub-Saharan Africa as well as from non-African countries.

Evidence from national surveys

In countries where low percentages of adults are HIV-positive, a husband or wife with an HIV-positive spouse has little risk to get HIV from any other source. In such countries, how fast does an HIV-positive spouse infect his or her partner? Before 2010 (before antiretroviral treatment was sufficiently common to make a big difference in sexual transmission within couples) seven national surveys in mainland sub- Saharan Africa found not more than 1.5% of adults HIV-positive. In these seven surveys, among couples in which at least one spouse was infected, the percentages of couples with both spouses infected (that is: one likely infected the other) ranged from 10% in Benin in 2006 to 19% in Guinea in 2005 (Figure 6.4).

For example, in DRC in 2007, the husband only was infected in 0.6% of couples, the wife only in 1.1% of couples, and both in 0.2% of couples. Assuming one infected the other, initially HIV-positive husbands and wives infected only 11% (= 0.2/[0.6 + 1.1 + 0.2]) of their initially HIV-negative partners.[30]

These data do not show who brought HIV into the home or how long couples lived together before one infected the other. However, we can conclude from these data that only about 15% of men and women who brought HIV into the home infected their partners (15% is the unweighted average from seven studies of the percentage of couples with both partners infected divided by the percentage with at least one infected).

If couples were living together on average for even three years after one was infected, the rate of transmission from the initially HIV-positive partners to infect 15% of their spouses would be less than 6% per year. (Even if one makes the absurd assumption that all transmission came from wives, or that all came from husbands, the rate of transmission in both cases would be less than 10% per year to reach observed percentages of both couples infected within three years.)

Couples unaware of their infections or risks

Combining results from five studies in Africa during 1989-98 in which many or most spouses did not know they were infected or at risk, initially HIV-negative men and women with HIV-positive spouses got HIV at an average rate of almost 10% per year (see Table A2.1). Some new infections that contributed to that rate may have come from sources other than HIV-positive spouses. Later studies that sequenced HIV in Rakai, Uganda, found that when both spouses were infected, HIV sequences from about half of such couples were dissimilar – husbands and wives had gotten HIV from different sources.[10,11]

Too slow!

If HIV-positive adults live 10 years without treatment, infecting sex partners at the rate of 10% per year, that could maintain a steady number of infections over time. But that would require all HIV-positive adults to be sexually active all the time with HIV-negative partners, which is not what one sees from national surveys. Many adults who are HIV-positive have HIV-positive spouses, and many others report no sex partner in the previous year. In other words, someone who is HIV-positive will on average die before he or she infects anyone through heterosexual sex.

Hence, HIV transmission through heterosexual sex is too slow to even maintain steady numbers of HIV infections in a community, much less drive an expanding epidemic. Fortunately, antiretroviral treatment (ART) extends lives. At the same time, ART reduces the risk someone with HIV will transmit through sex. With or without ART there is no heterosexual epidemic.

Similarly slow HIV transmission through heterosexual sex has been observed and reported outside Africa in countries with good information about sources of infections. In Latvia, for example, “it is difficult to observe a sustained heterosexual epidemic without continued input from an ‘outside’ reservoir.”[36] A study in Switzerland estimated that HIV- positive heterosexuals infected an average of only 0.44 persons though sex in their lifetime, “far below the epidemic threshold.”[37] In layman’s terms: HIV transmission through heterosexual sex was too slow to even maintain the same number of infections over time.

Not much more HIV in sexually more active adults

In seven of the eight countries with the worst epidemics (comparable data are not available for Botswana), 6.2%-15.6% of adults aged 15-49 years reported more than one sex partner in the previous year (light grey bars in Figure 6.5). The average across the seven countries was less than 10%.

These sexually more active adults did not have a lot more HIV than other adults. The percentages of all HIV infections in adults that were in those who reported more than one sex partner last year (dark grey bars in Table 6.5) were similar to the percentages of adults reporting more than one partner. In Namibia and South Africa, the percentages of HIV in sexually more active adults were less than their percentage of all adults: they were less likely to be HIV-positive! And in Zimbabwe, they were equally likely to be HIV-positive.

If sex accounts for most HIV infections, HIV should concentrate in those who are more sexually active, as one sees with common sexually transmitted disease such as syphilis. But that is not what happens. Furthermore, considering the low percentages of adults reporting more than one sex partner per year, a lot of people getting HIV from sex will be staying with the partner who infected them, so their HIV has no chance to get to anyone else through sex.

What do models show?

Some proponents of the view that almost all HIV infections among adults in Africa come from sex have proposed models to show how that could be so. But to generate anything that looks like Africa’s epidemics, model-builders have to assume sexual behaviors and/or rates of HIV transmission though sex that do not agree with evidence.[39] For example, when a team of experts tried to model Zambia’s HIV epidemic using sexual behavior data from a national survey and the rate of HIV transmission per sex act from a study in Africa, they failed: “no epidemic could be generated.”[40] To get the result they wanted they made up what they needed: “After exploring a wide range of parameters [sexual behaviors and transmission per sex act], we found a combination giving a good fit to the age and sex patterns of prevalence in 2001.” To make their model generate something that looked like Zambia’s epidemic, they assumed men had more commercial sex than they reported and a higher than observed rate of HIV transmission per sex act. But what did that prove? Adjusting assumptions to reach desired conclusions can “prove” almost anything: cows fly, pigs lay eggs, all HIV in adults comes from heterosexual sex, etc.

Why has the myth lasted so long?

In countries where a lot of adults are HIV-positive, HIV prevention programs should warn people to avoid HIV from sex. But to make that point stick, such programs do not have to say all HIV comes from sex. Similarly, clergy, parents, and others who want to discourage pre-marital or extra-marital sex may cite HIV as a danger for those who “misbehave.” But to discourage sex outside marriage, clergy and parents do not have to say all HIV comes from sex, just that some comes from sex.

So why has the myth lasted so long? Health professionals appear to be the only group delivering HIV prevention messages that has any reason to promote the myth. Blaming sex distracts attention from unexplained infections and bloodborne risks. That allows health aid agencies and governments to extend healthcare programs – immunizations, antenatal care, hospital deliveries – without having to take the effort to make health care reliably safe and without having to be accountable to patients.

The myth that almost all HIV-infections among adults in Africa comes from sex emerged in the face of contradictory evidence and continues despite more such evidence. John Potterat, a veteran of years of debates in medical journals about HIV from sex and bloodborne risks in the US and Africa has written a short, detailed criticism of bad science defending the myth.[41]

References

1. Pinching A. AIDS and Africa: lessons for us all. J Roy Soc Med 1986; 79: 501-503.

2. Hrdy DB. Cultural practices contributing to the transmission of human immunodeficiency virus in Africa. Rev Infect Dis 1987; 9: 1109-1119.

3. Piot P, Plummer FA, Mhalue FS, et.al. AIDS: an international perspective. Science 1988; 239: 573-579.

4. Chin J, Sato PA, Mann JM. Projections of HIV infections and AIDS cases to the year 2000. Bull World Health Org 1990;68:1–11.

5. Gisselquist D, Potterat JJ, Brody S, Vachon F. Let it be sexual: how health care transmission of AIDS in Africa was ignored. Int J STD AIDS 2003; 14: 148-161.

6. Wellings K, Collumbien M, Slaymaker E, et al. Sexual behaviour in context: a global perspective. Lancet 2006; 368: 1706-1728.

7. Schmid GP, Buve A, Mugyenyi P, et al. Transmission of HIV-infection in sub-Saharan Africa and effect of elimination of unsafe injections. Lancet 2004; 363: 482-488.

8. Gisselquist D, Potterat JJ, Heterosexual transmission of HIV in Africa: an empiric estimate. Int J STD AIDS 2003; 14: 162-173.

9. Novitsky V, Bussmann H, Okui L, et al. Estimated age and gender profile of individuals missed by a home-based HIV testing and counseling campaign in a Botswana community. J Int AIDS Soc 2015; 18: 19918.

10. Grabowski MK, Lessler J, Redd AD, et al. The role of viral introductions in sustaining community-based HIV epidemics in rural Uganda: evidence from spatial clustering, phylogenetics, and egocentric transmission models. PLoS 2014; 11: e1001610.

11. Ratmann O, Grabowski MK, Hall M, et al. Inferring HIV-1 transmission networks and sources of epidemic spread in Africa with deep-sequence phylogenetic analysis. Nat Commun 2019; 10: 1411.

12. Novitsky V, Kuhnert K, Moyo S, et al. Phylodynamic analysis of HIV sub-epidemics in Mochudi, Botswana. Epidemics 2015; 13: 44-55.

13. Gisselquist D. Randomized controlled trials for HIV/AIDS prevention among men and women in Africa: untraced infections, unasked questions, and unreported data. Social Science Research Network [internet] 9 October 2011. Available at: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1940999 (accessed 9 June 2020).

14. Lasry A, Medley A, Behel S, et al. Scaling up testing for human immunodeficiency virus infection among contacts of index patients -- 20 Countries, 2016–2018. MMWR Morb Mort Wkly Rep 2019; 68: 474-477.

15. UNAIDS. HIV estimates with uncertainty bounds 1990-2019. Geneva: UNAIDS, 2020.

16. ICAP. Swaziland HIV incidence measurement survey 2 (SHIMS2) 2016-2017: Final Report. New York (NY): ICAP, Columbia University, 2019.

17. ICAP. Lesotho population-based HIV impact assessment (LePHIA) 2016-2017: Final report. New York (NY): ICAP, Columbia University, 2019.

18. ICAP. Namibia Population-based HIV Impact Assessment (NAMPHIA) 2017: Final Report. New York (NY): ICAP, Columbia University, 2019.

19. Simbayi LC, Zuma K, Zungu N, et al. South African National HIV Prevalence, Incidence, Behaviour and Communications Survey 2017. Cape Town: HSRC; 2019.

20. ICAP. Zambia Population-based HIV Impact Assessment (ZAMPHIA) 2016: Final Report. New York (NY): ICAP, Columbia University, 2019.

21. ICAP. Zimbabwe Population-based HIV Impact Assessment (ZIMPHIA) 2015- 2016: Final Report. New York (NY): ICAP, Columbia University, 2019.

22. Macro International. Swaziland Demographic and Health Survey 2006-07. Calverton (MD): Macro International; 2008.

23. ICF International. Lesotho Demographic and Health Survey 2014. Rockville (MD): ICF; 2016.

24. ICF International. The Namibia demographic and health survey 2013. Rockville (MD): ICF International, 2014.

25. ICF. South Africa Demographic and Health Survey 2016. Rockville 1 (MD): ICF; 2019.

26. ICF. Zambia Demographic and Health Survey 2018. Rockville (MD): ICF, 2019.

27. ICF International. Zimbabwe Demographic and Health Survey 2015: Final Report. Rockville (MD): ICF International, 2016.

28. Evans M, Risher K, Zungu N, et al. Age-disparate sex and HIV risk for young women from 2002 to 2012 in South Africa. J Int AIDS Soc 2016; 19: 21310.

29. Macro International. Enquête Démographique et de Santé (EDSB-III) - Bénin 2006. Calverton, (MD): Macro International. 2007.

30. Macro International. Enquête Démographique et de Santé, République Démocratique du Congo 2007. Calverton (MD): Macro International, 2008.

31. OCR Macro. Ethiopia Demographic and Health Survey 2005. Calverton (MD): OCR Macro, 2006.

32. ORC Macro. Enquête Démographique et de Santé, Guinée 2005. Calverton (MD): ORC Macro, 2006.

33. Macro International. Liberia Demographic and Health Survey 2007. Calverton (MD): Macro International, 2008.

34. Macro International. Enquête Démographique et de Santé at a Indicateurs Multiples du Niger 2006. Calverton (MD): Macro International, 2007.

35. ICF Macro. Sierra Leone Demographic and Health Survey 2008. Calverton (MD): ICF Macro, 2009.

36. Graw F, Leitner T, Ribeiro RM. Agent-based and phylogenetic analyses reveal how HIV-1 moves between risk groups: injecting drug users sustain the heterosexual epidemic in Latvia. Epidemics 2012; 4: 104-116.

37. Turk T, Bachmann N, Kadelka C, et al. Assessing the danger of self- sustained HIV epidemics in heterosexuals by population based phylogenetic cluster analysis. eLife [internet] 2017; 6: e28721.

38. ICF. Inquérito de Indicadores de Imunização, Malária e HIV/SIDA em Moçambique 2015. Rockville (MD); ICF, 2018.

39. Deuchert E, Brody S. Plausible and implausible parameters for mathematical modeling of nominal heterosexual HIV transmission. Ann Epidemiol 2007; 17: 237-244.

40. Leclerc PM, Matthews AP, Garenne ML. Fitting the HIV epidemic in Zambia: a two-sex micro-simulation model. PLoS One 2009; 4: e5439.

41. Potterat JJ. Why Africa? the puzzle of intense HIV transmission in heterosexuals. In: Potterat JJ. Seeking the positives: a life spent on the cutting edge of public health. North Charleston (SC): Createspace, 2015. pp 175-229.