Numerous studies have now shown the ease with which HIV infection can be transmitted during unprotected vaginal intercourse. A number of factors unrelated to condom use seem to increase the chance of infection through vaginal intercourse. These are the following:

The level of HIV in semen or vaginal fluid

Levels of HIV vary in both semen and vaginal fluid, partly as a result of disease progression, with higher viral load observed later in HIV infection; in a 1992 study only 4% of those with CD4 counts over 200 produced semen from which it was possible to culture HIV, compared with 24% of those with CD4 counts under 200. In a 1994 study HIV isolation was most frequent from the semen of men whose CD4 count was below 100 (87% vs. 40% of men with CD4 count of 101–200) (Anderson; Vernazza).

However, virus levels in semen appear to be unpredictable, and some studies have reported no correlation between virus levels in blood and virus levels in semen (see under Viral load in semen above).

This may be partially explained by the presence of undiagnosed or asymptomatic genitourinary infections in individuals with high viral load in semen but low plasma viral load (see The mechanisms of HIV transmission in this section).

The European Study Group on Heterosexual Transmission of HIV noted in 1994 that partners of men or women who had a CD4 count below 200 at the beginning of the study were much more likely to have seroconverted after 20 months than partners of HIV–positive men or women who remained asymptomatic throughout the 20 month follow–up period (De Vincenzi 1994).

A study in Uganda has confirmed that viral load is the most important factor in determining whether HIV is transmitted. But it is unclear how these findings can be translated into a Western setting where people are using antiretroviral drugs, experts warn.

415 couples in whom one partner was HIV-positive were identified in the Rakai district of Uganda. The couples were identified by confidential HIV testing. All couples were offered counselling in condom use and the opportunity to learn their HIV status at the beginning of the study, and access to free condoms throughout the study. People diagnosed HIV-positive were left to decide whether or not to inform their partners, a decision criticised in an accompanying editorial in the New England Journal of Medicine (Quinn).

Sexually transmitted infection treatment was offered at regular intervals as part of another study to half the participants. The other half were required to seek free treatment if they experienced any symptoms. HIV antibody status and exposure to sexually transmitted infections was tested every ten months.

Viral load prior to transmission was determined by measuring the viral load of the HIV-positive partner at the check-up prior to seroconversion of his or her partner. To estimate the relative risk, this viral load result was matched with one from another individual of similar age and sex who had not seroconverted.

In this study, every year of follow-up for an individual counts as a person year. In this study 415 couples were enrolled and followed for a median of just under two years. This means that the study followed people for approximately 800 person years.

The key findings of the study were:

• 22% of all partners seroconverted during the follow-up period.
• Men were just as likely to become infected as women at any given level of viral load.
• Circumcision appeared to be protective: none of the circumcised male partners of HIV-positive women became infected, whereas 40 out of 197 uncircumcised men became infected.
• Symptomatic sexually transmitted infections did not affect an individual's likelihood of catching HIV, but a history of genital discharge in the HIV-positive partner was associated with an increased risk of HIV infection (p<0.05).
• Viral load above 50,000 copies in the HIV-positive partner was most strongly associated with the risk of transmission, at a rate of 23 infections per 100 person years.
• 5.6% of all transmission occurred in couples where the HIV-positive partner had a viral load between 400 and 3,499 copies, indicating that transmission can take place even from individuals considered to be at very low risk of disease progression. This translates into 2.2 cases per 100 person years, a tenfold lower risk than seen in couples where the HIV-positive partner had viral load greater than 50,000 copies.
• No infections occurred in couples where the HIV-positive partner had viral load below 1500 copies.

In an editorial accompanying the publication of the study, Dr Myron Cohen of the University of North Carolina noted that the findings did not necessarily support the view that lowering viral load with antiretroviral treatment would reduce HIV transmission rates:

HIV-1 can still be cultured from the genital secretions of some patients who are receiving antiretroviral therapy and who have undetectable levels of HIV-1 RNA in blood, a finding that means that one cannot reassure patients that they are not contagious. Indeed, if the use of such therapy increased the likelihood that HIV-1-infected patients would practice unsafe sex in the mistaken belief they were unable to transmit the virus, it could offset the benefit of viral suppression, he wrote.

The levels of HIV in vaginal fluid tend to vary too. Levels of HIV are likely to be highest around the time of menstruation when HIV–bearing cells shed from the cervix are most likely to be found in the vaginal fluid, along with blood (Mostad).

However, avoiding sexual intercourse during menstruation is likely to provide little protection to male partners of HIV–positive women: the 1992 report of the European Study Group showed that almost half the men infected by female partners in the study were men who avoided intercourse during menstruation.

One study (Lennox) has found that subclinical inflammation within the female genital tract might also increase cervicovaginal shedding of HIV.

Putting these studies together, the clinical relevance becomes clear: High blood plasma viral load means that genital tract viral load is also likely to be high, as the two appear to be reasonably correlated in most studies.

However, clinical or asymptomatic genital inflammation will further increase genital HIV shedding, and there appears to be a small number of patients with no apparent genital tract infections who still may have exceptionally high genital tract viral loads and only moderately elevated plasma viral loads.

The evidence concerning the effects of anti-HIV treatment on the levels of HIV in semen and vaginal fluid is discussed in detail in Treating HIV in genital fluids in Anti-HIV-therapy: Choosing your treatment strategy in the HIV & AIDS Treatments Directory or at aidsmap.com.

Sexually transmitted infections

Ulcers caused by sexually transmitted infections may provide a route of infection for HIV. This is due to damage to the mucous membranes and the increased presence of the white blood cells which HIV infects. Bleeding from these ulcers may also increase the chance of infection (Padian). Non–ulcerative sexually transmitted infections such as gonorrhoea, chlamydia, bacterial vaginosis and trichomoniasis also appear to assist HIV infection according to epidemiological evidence, although the precise mechanism is unclear. It is impossible to say at present whether genital warts play any role in assisting HIV infection (Laga).

In 1994 the European Study Group on Heterosexual Transmission of HIV reported that those with sexually transmitted infections were three or four times as likely to be infected after 20 months of follow–up than those who reported no sexually transmitted infections (De Vincenzi 1994).

In 1999 a study of Thai women with HIV-positive partners showed that women whose partners had a history of sexually transmitted infections were twice as likely to be HIV-positive as women whose partners did not. In this study women with other possible exposure routes were excluded, and CD4 counts were also measured. Women almost always had higher CD4 counts than male partners, suggesting later exposure to HIV (Nelson).

Studies in Malawi and Cote d'Ivoire amongst men and women respectively (Cohen; Ghys) have demonstrated that HIV levels in sexual fluids are reduced by treatment of urethritis and gonorrhoea.

If an HIV–infected person is also concurrently suffering a sexually transmissible infection, this is likely to increase the infectivity of his semen or her vaginal fluid. HIV infection is also likely to lead to an increased incidence of ulcerative sexually transmitted diseases, such as herpes and cancroid. Sexually transmitted diseases play a key role in HIV infections amongst heterosexuals. Thrush (candida infection) also makes women more vulnerable to HIV infection, for the same reason, and women with a history of thrush infection who are HIV–positive seem more likely to pass HIV to their male partners.

One study estimates that the presence of an ulcerative genital infection on a man's penis increases the per–sex act transmission risk from a woman to a man 50–300 fold (Hayes).

In 1995 a study conducted in the Mwanza district of Tanzania showed that at a population level a reduction in the prevalence of sexually transmitted infections and a syndromic treatment programme for STIs were associated with a reduction in the incidence of new HIV infections. These findings suggest that the control of sexually transmitted infections can play a major part in HIV prevention (Grosskurth). (Syndromic treatment is a simplified, rule-based approach for settings where diagnostic facilities are limited.)

However, in 1999 another group working in the Rakai district of Uganda reported that in a region with higher HIV prevalence (15.9% vs 4%), a randomised study of mass STI treatment did not find any difference in HIV incidence between those who received STI treatment and those who did not. The authors suggested that in a more mature epidemic, STIs may not play the same role as in an early epidemic, where they might be amplifying transmission opportunities amongst individuals recently infected with HIV and their partners. In the early stages of an epidemic, these individuals are likely to be those with the largest number of sexual contacts with other individuals who also have a high number of sexual contacts. In a more mature epidemic, other host factors (such as degree of natural immunity) and the number of sex acts with each partner will play a more important role. An initial core group of HIV-infected individuals might just as easily infect 100 people if they have repeated sex with those individuals as if they had single sexual contacts with thousands of people, irrespective of STD incidence (Hitchcock; Wawer).

In a 2004 study of gay men (Buchacz), syphilis infection in HIV-positive men was associated with an increase in viral load and reduction in CD4 cell count. The study investigators believed that their findings indicate that HIV-positive men with syphilis were potentially more infectious and called for integrated public health campaigns to prevent the spread of both HIV and syphilis.

There has been considerable debate about the role of STI treatment programmes within HIV prevention programmes. One of the key points to remember is that the two studies discussed above are not entirely comparable, not only because the epidemic may have been more mature in the latter study, but also because the nature of the treatment programmes differed. In the latter study a one-off programme of treatment took place, but experts point out that single dose antimicrobial treatment of the kind used in the Rakai study would have no impact on conditions such as herpes, syphilis and bacterial vaginosis, all of which may be implicated in HIV transmission (Grosskurth 2000). Herpes simplex in particular was a much more common condition in the Rakai district; 43% of diagnosed genital ulcers were caused by HSV-2, compared with less than 10% in the Mwanza district.

The role of genital herpes in HIV transmission is receiving attention. Herpes simplex virus II (HSV-2) is the most common cause of genital ulcers worldwide and it has long been suggested that by disruption of the epithelial barrier and general inflammation, it may increase the chances of infection with HIV.

In a South African study 400 female sex workers at truck stops between Durban and Johannesburg were screened for HIV. Only 198 were found to be HIV-negative and they were monitored monthly over three years. Of the women who seroconverted during the study, all but six became HSV-2 positive before they became HIV-positive, suggesting to the investigators that “immediately after infection with HSV-2, the risk of contracting HIV-1 increases significantly.”

However, investigators also found that women who were already HSV-2-positive at the start of the study had a lower risk of HIV seroconversion than those who were HSV-2-negative at the start of the study. The investigators speculate that the women already infected with HSV-2 were protected from reinfection with HSV-2, thus reducing the chance of acquiring HSV-2 ulcerative lesions which are a risk factor for HIV.

HIV subtype

Research from Thailand shows that HIV–1 subtype E is more frequently transmitted through vaginal intercourse than HIV–1 subtype B (the subtype predominant amongst drug injectors and gay men in the West). When all other factors were controlled for, female partners of men infected with HIV–1 subtype E were 3.1 times more likely to be infected with the same strain than female partners of men carrying HIV–1 subtype B (Kunanusont). This finding may explain why epidemiologists have calculated that the risk of heterosexual transmission per episode of sexual intercourse in the Thai population was 31–fold to 56–fold higher than estimates for the US population (Mastro). Subtype E has also been shown to reproduce more efficiently than subtype B in immune cells called Langerhans cells, which are numerous in the vaginal wall.

However, many researchers are still sceptical about differences in transmission rates between HIV subtypes (see Physical co–factors which encourage transmission earlier in this chapter). A Thai/American group found that viral load was significantly lower (approximately 0.5 log) in the semen of 87 Thai men infected with subtype compared to 70 men infected with subtype B (Coombs).

Individual cases studies also suggest that HIV–1 is a great deal more easily transmitted than HIV–2. Austrian doctors reported in 1994 that a woman infected with HIV–1 by a partner co–infected with HIV–1 and HIV–2 did not seroconvert for HIV–2 despite nearly 14 months of follow–up (Most).

Damage to the tissue of the genitals

More violent sexual penetration of the type often associated with rape may lead to vaginal or penile abrasions which may increase the chances of HIV infection. This is most likely to be caused by rough sex and/or dryness in the vagina, and will create small abrasions which make it more easy for the virus to get into the bloodstream.

Bleeding from damaged tissue on the penis may also facilitate infection to the female partner.

A number of studies have also suggested a possibly increased risk of HIV infection when the hymen is broken on the first occasion of sexual intercourse.

Reduction of vaginal lubrication becomes more common with age, and some researchers say that this is one reason why some studies show higher rates of HIV infection amongst post–menopausal women. Younger women in early puberty may also produce less vaginal and cervical secretions, perhaps increasing their vulnerability to HIV infection and contributing to the disproportionate prevalence of HIV amongst adolescent women. Hysterectomy in older women has also been reported as a factor which may pre–dispose to the risks of HIV infection in older women (Holmberg).

Vulnerability of the cervix

Cells particularly vulnerable to HIV infection are located beneath the surface of the cervix (Pomerantz). These cells are more vulnerable to infection in adolescence and during a woman's first pregnancy, perhaps accounting for the higher prevalence of HIV amongst women under 30 than amongst men under 30 in Africa.

These cervical cells may also become vulnerable as a consequence of changes in the cervix caused by human papilloma virus and chlamydia. Sexual health programmes which seek to reduce the incidence of unprotected sex amongst adolescent women may therefore be particularly important.

Circumcision

The first ever randomized controlled trial (RCT) of malecircumcision as an HIV prevention measure, presented at the 3rd IAS Conference in 2005 (Auvert), produced such strong evidence of a protective effect that the trial was halted early and all participants offered circumcision.

There were only 35% as many infections in the circumcision arm as opposed to the control arm, implying that circumcision can prevent at least six out of ten female-to-male HIV transmissions.

However, when the results were analysed according to true circumcision status rather than by intervention group, the protective effect went up to 75%. This is because there were ‘crossovers’ between the intervention and control arms in that some men randomized to be circumcised were not, and some in the control arm were.

The trial, the first of four RCTs of circumcision being conducted in Africa, randomised 3,273 men aged 16 to 24 to be circumcised at the start of the trial or to be offered circumcision at the end of it 21 months later.

The men lived in the Orange Farm township near Johannesburg, South Africa. A previous acceptability study had found that 70% of the local male population said they were willing to be circumcised if it could prevent HIV.

The HIV prevalence in the area is high, at 31.6% of the adult population. In the trial population 90% of men were sexually active by the start of the study, with a mean age of sexual debut of 16.6.

Circumcisions in the intervention arm were carried out by a surgeon under local anaesthesia and with post-operative pain relief given.

Researcher Bertrand Auvert of the French HIV Research Institute INSERM stressed the safety of the procedure and said that there had been no deaths or permanent adverse effects in any participant. Thirty-one per cent complained of pain and 15% initially had problems with the changed appearance of their penis.

HIV incidence was measured at three and twelve months into the trial and finally at 21 months though the average follow-up period was in fact 20 months due to the premature termination of the trial.

Although all participants received intensive safer sex counselling and condoms, there were 51 HIV seroconversions in the control arm versus 18 in the circumcision arm. This translates as HIV incidences of 2.2% and 0.77% a year respectively.

In the control arm there were nine, 15 and 27 new infections at three, 12 and 21 months and in the circumcision arm two, seven and nine.

He added that as a short-term study it could not predict the long term effect of circumcision, but that its compelling results now demanded discussions on the use of circumcision as a public health measure.

Circumcision may also substantially cut the male to female transmission rate, the conference heard. Ronald Gray, one of the investigators behind the long-standing Rakai prevention cohort in Uganda, said that studies at Rakai and other centres had indicated that HIV-negative women with circumcised HIV-positive partners had only 0.41 as many seroconversions as partners of uncircumcised men.

Previous nonrandomised studies had already suggested that men who are not circumcised stand a greater chance of HIV infection. This is thought to be due to the vulnerability to abrasion of the tissue beneath the foreskin, and the increased likelihood that an uncircumcised man will contract sexually transmitted diseases, and pass these on to his partners. For example, bacterial vaginosis is more common in women with uncircumcised sexual partners. Cells vulnerable to HIV infection may be more common in the foreskin too (Moses).

Several studies in Africa have detected a substantially greater risk of HIV infection amongst uncircumcised men. A review of 6994 men in the Rakai district of Uganda found that men circumcised before puberty were 60% less likely to be HIV-positive. Circumcision between the ages of 13 and 20 had some protective effect. However, circumcision after the age of 20 did not appear to have a protective effect. These findings suggest that circumcision of adult males will have no effect on HIV incidence if it is pursued as an HIV prevention policy (Kelly).

Another study conducted in Uganda found nil incidence of HIV infection in 50 circumcised HIV-negative men in HIV-discordant sexual relationships during the the follow-up period, compared with an incidence of 16.7 infections per 100 person years of follow-up in the uncircumcised males (Gray). Among the partners of HIV-positive males, the risk of HIV acquisition was greater where the male partner was uncircumcised if he also had a viral load below 50,000 copies. At higher levels of viral load the protective effect of circumcision disappeared.

A prospective study amongst 746 male trucking company employees in Kenya found a similar protective effect for circumcision, although the study did not look at the age at which circumcision took place. Uncircumcised men were more than four times as likely to become infected with HIV during the four years of follow-up (Lavreys). Uncircumcised men were also more likely to report genital ulceration during the follow-up period, an additional risk factor for HIV infection. However, only 13% of men in the cohort were uncircumcised, indicating a wide statistical margin of error.

However, a meta-analysis of these studies found that while the association between circumcision and reduced risk of HIV transmission held true when data were pooled from 33 studies, a much more rigorous analysis of the data is still needed to tease out the role of confounding factors such as sexually transmitted infections, age of first sexual intercourse in women and HIV prevalence (O'Farrell).

Why are uncircumcised men more likely to be infected with HIV? A recent US study (Patterson) suggests that it is because the cells on the inside of the foreskin are more susceptible to HIV

Investigators examined the foreskin tissue from eight children and six adults. This was compared to cervical tissue. They found that the cells under the foreskin contain higher concentrations of CD4+ T cells, and other cells including macrophages and Langerhans’ Cells which are specifically targeted by HIV, than tissue from the cervix. In addition, the cells were founded in the greatest densities in the foreskin tissue men with a history of sexually transmitted infections (STIs), a finding consistent with studies which suggest that men with STIs are more vulnerable to HIV.

When the tissue cultures from both the foreskin and cervix was exposed to HIV, the CD4 and other cells from the foreskin were infiltrated by HIV much more rapidly than the cervical cells.

The keratin layer found on the inner mucosal surface of the foreskin predisposed uncircumcised men to HIV infection.

A report on research needed in order to establish the value of circumcision as an HIV prevention intervention is available at http://hivinsite.ucsf.edu/topics/circumcision/2098.477c.html.

Female genital mutilation (female circumcision)

In women, the practice of female genital mutilation, the cutting of the genitals, is speculated to assist HIV infection, but a comparison of areas of high HIV prevalence in Africa with areas in which female genital mutilation is commonly practised has shown no clear correspondence.

A study from Tanzania presented at the 3^rd IAS conference found that female genital mutilation actually had a negative effect on HIV transmission (Stallings).

Number of exposures

One sexual contact is enough for a man to be infected: 8.2% of 73 men followed for an average of 13 weeks seroconverted after one sexual contact with a woman (all had genital ulcer disease, five were uncircumcised) (Johnson).

One exposure to infected semen is enough for a woman to be infected; this is demonstrated by the 1990 report of HIV infection following artificial insemination (MMWR). A 1995 review of 199 women at five US fertility clinics exposed to HIV-infected semen from a single donor found that seven women were infected from a single exposure (Araneta).

Seroconversion is dependent on sexual intercourse with an infectious partner; 100 sexual contacts with a partner who has a very low level of infectivity is probably far less risky than sexual contacts with 100 different HIV–positive partners. This is because there is a high probability that at least one of these partners is highly infectious.

There is no evidence that seroconversion occurs after a cumulative process of infection – in other words, the more times one has sex with an infected person, the more virus one becomes infected with, until a critical mass is achieved and seroconversion occurs. This belief has arisen in part because of lay beliefs about health and illness, but also from the observation that many partners of HIV infected people have gone for long periods without being infected. This is due to the fact that conditions which permit infection may not exist all the time.

Immune function

Although there is some evidence that despite repeated exposure to HIV, some people do not become infected, the significance of these observations is not fully understood. We do not know if this means that some people can resist HIV infection, and what the factors might be which confer some kind of resistance to infection.

There are several possible explanations for this phenomenon:

Genetic susceptibility to HIV is a hypothesis supported by a number of observations, notably the identification of a common genetic marker on the immune cells of some people who appear to have resisted infection, correlated with the identification of a number of genetic markers on the immune cells of people who have developed AIDS.

Resistance to infection within a population develops through generations of interactions between the host immune system and infectious organisms, and develops by a process of natural selection; those with the best resistance will survive the infection and pass their genetically conferred resistance onto some of their descendants. This is the way in which populations adapt to some conditions; an example of this is the adaptation of Europeans to various forms of pox virus (e.g. smallpox), and the lack of resistance shown by inhabitants of the Americas when Europeans first introduced these viruses. Resistance to HIV infection is discussed in Resistance to infection earlier in this chapter.

Stronger cellular immunity: Some people have been observed to have consistently higher CD4 counts, higher levels of CD8 cells and lymphocytes throughout periods of study than comparable individuals who seroconvert during the course of the study (Visscher). It may be that the overall state of the immune system plays an important role in preventing or allowing infection, but the mechanism is still unknown. Some research has suggested that exposure to very small amounts of HIV may have an immunising effect by stimulating cell–mediated rather than antibody responses to HIV, but this view remains to be proven.

Defective form of HIV: An alternative suggestion is that individuals who have apparently resisted HIV infection may simply have encountered a defective or less virulent form of HIV. In 1995 considerable publicity was given to the case of a child who had cleared HIV infection after perinatal infection. Researchers subsequently demonstrated that the child had been infected with a defective strain of HIV and had been able to clear the virus and revert to HIV–negative by twelve months of age (Bryson).

For more on immune function and HIV, see Cytotoxic T cell protection against HIV above.

Nutritional deficiency

Deficiency in blood levels of the mineral selenium has been associated with increased rates of vaginal HIV shedding in a study conducted among women in Kenya. HIV-positive women with selenium deficiency were almost three times more likely to have evidence of HIV DNA in their vaginal fluids when sampled (once only) (Baeten).

Although the authors of this study have suggested a potential role for selenium supplementation in reducing transmission, the confounding effect of viral load not taken into account in this study. While the authors controlled for the effects of CD4 count, they did not test for plasma viral load, and it is possible that an independent association might exist between plasma viral load levels and selenium deficiency.

References

Anderson DJ et al: Effects of disease stage and zidovudine therapy on the detection of human immunodeficiency virus type 1 in semen, JAMA 267: 2769–2774, 1992.

Araneta MRG et al: HIV transmission through donor artificial insemination, JAMA 273(11): 854–858, 1995.

Auvert B et al. Impact of male circumcision on the female-to-male transmission of HIV. IAS Conference on HIV Pathogenesis and treatment, Rio de Janeiro, abstract TuOa0402, 2005

Baeten J et al. Selenium deficiency is associated with shedding of HIV-1 infected cells in the female genital tract. XIII International AIDS Conference, Durban, abstract MoOrA226, 2000.

Bryson YJ et al: Clearance of HIV infection in a  perinatally infected infant, NEJM 332(13): 833–838, 1995.

Buchacz K et al. Syphilis increases HIV viral load and decreases CD4 cell counts in HIV-infected patients with new syphilis infections. AIDS 18: 2075-2079, 2004.

Cohen M. Reduction of concentration of HIV–1 in semen after treatment of urethritis: implications for prevention of sexual transmission of HIV–1. Lancet 349: 1868–1873, 1997.

Coombs RW et al. Does viral sub-type determine HIV-1 shedding in semen? Annual Meeting of the American Urological Association, abstract 1111, 1999.

De Vincenzi I et al: A longitudinal study of human immunodeficiency virus transmission by heterosexual partners, NEJM 331(6): 341–346, 1994.

Ghys A. HIV shedding, STD and immunosuppression. AIDS 11(12): F85–F93, 1997.

Gray R et al. HIV incidence associated with male circumcision in a population-based cohort, and HIV acquisition/transmission associated with circumcision and viral load in discordant couples: Rakai, Uganda. XIII International AIDS Conference, Durban, abstract MoOrC193, 2000.

Grosskurth H et al: Impact of improved treatment of sexually transmitted diseases on HIV infection in rural Tanzania: randomised controlled trial, Lancet 436: 530–536, 1995.

Grosskurth H et al. Control of sexually transmitted diseases for HIV-1 prevention: understanding the implications of the Mwanza and Rakai trials. The Lancet 355: 1981-1987, 2000.

Hayes RJ et al: The cofactor effect of genital ulcers on the per–exposure risk of HIV transmission in sub–Saharan Africa, J Trop Med Hyg 98: 1–8, 1995.

Hitchcock P, Fransen L. Preventing HIV infection: lessons from Mwanza and Rakai. The Lancet 353: 9152, 1999.

Holmberg SD, Horsburgh CR: Biologic factors in the sexual transmission of human immunodeficiency virus, J Inf Dis 160(1): 116–125.

Ilaria G et al: Detection of HIV–1 DNA sequences in pre–ejaculatory fluid, Lancet 340(8833): 1469, 1992.

Johnson AM et al: Heterosexual transmission of HIV, AIDS(S1): 549–556, 1988.

Kelly R et al. Age of male circumcision and risk of prevalent HIV infection in rural Uganda. AIDS 13: 399-405, 1999.

Kunanusont C et al: HIV–1 subtypes and male to female transmission in Thailand, Lancet 345: 1079–1083, 1995.

Laga M: Interactions between STDs and HIV infection, Transcript 28, Aug 1992.

Lavreys L et al. Effect of circumcision on incidence of human immundeficiency virus type 1 and other sexually transmitted diseases: a prospective cohort study of trucking company employees in Kenya. Journal of Infectious Diseases 180: 300-336, 1999.

Lennox J et al. Subclinical inflammation in the female genital tract is strongly associated with vaginal viral shedding independent of plasma viral load. Program and abstracts of the tenth Conference on Retroviruses and Opportunistic Infections, Boston, Massachusetts, abstract 101, 2003. 

Mastro TD et al: Probability of female–to–male transmission of HIV–1 in Thailand, Lancet 343: 204–207, 1994.

Moses S et al: Male circumcision and risk for HIV infection: A review of epidemiological studies conducted to date, Ninth International Conference on AIDS Po C11–2838 1993.

Most J: Differences in heterosexual transmission of HIV–1 and HIV–2, Lancet 344: 1572–1573, 1994.

Mostad S: Shedding of HIV–1 in the genital tract, AIDS 10: 1305–1315, 1996.

Nelson KE et al. High rates of transmission of subtype E human immunodeficiency virus type 1 among heterosexual couples in Northern Thailand. Journal of Infectious Diseases 180: 337-343, 1999.  

O'Farrell N et al. Circumcision in men and the prevention of HIV infection: a meta analysis re-visited. Int J STD AIDS 11(3): 137-142, 2000.

Padian N et al: Female to male transmission of Human Immunodeficiency Virus, JAMA 266(12): 664 –667, 1987.

Patterson B et al.  Susceptibility to Human Immunodeficiency Virus-1 Infection of Human Foreskin and Cervical Tissue Grown in Explant Culture. 161: 867-873, 2002

Pomerantz RJ et al: Human immunodeficiency virus infection of the uterine cervix, Ann Int Med 108(3): 321–327, 1988.

Ponce de Leon, S et al: Gossypol suppresses p24 antigen in semen, Eighth International Conference on AIDS, PoC 4538, 1992.

Quinn CT et al. Viral load and heterosexual transmission of human immunodeficiency virus type 1. New England Journal of Medicine 342(12): 921-929, 2000.

Ramjee G et al. Herpes simplex virus type II infection is a risk factor for HIV seroconversion. South African MRC policy brief 3, August 2002.

Stallings RY. Female circumcision and HIV infection in Tanzania: for better or for worse? Third IAS Conference, Rio de Janeiro, abstract TuOa0401, 2005.

Vernazza PL et al: Detection and biologic characterization of infectious HIV–1 in semen of seropositive men, AIDS 8(9): 1325–1329, 1994.

Visscher B et al: Levels of immune cells in HIV–resistant men.  Eighth International Conference On AIDS, abstract WeA 1048, 1992.

Wawer MJ et al. Control of sexually transmitted diseases for AIDS prevention in Uganda: a randomised community trial. The Lancet 353: 9152, 1999.