Factors which may increase the risk of infection if exposed to HIV include:

Sexually transmitted infections

Having other sexually transmitted infections at the time, such as herpes, syphilis, especially, but not only, if they cause genital ulceration, may increase the risk of transmitting HIV. Genital ulceration provides an enhanced route of infection. In Thai women with HIV-positive partners, a history of syphilis, herpes, gonorrhoea or chlamydia all increased the risk of infection, and women were twice as likely to be HIV-positive if their partners had a history of sexually transmitted infections since seroconversion (Nelson).

Sexually transmitted infections also increase the quantity of HIV–bearing cells present in the vaginal fluid or semen, although some infections appear to increase levels of viral shedding more than others. A study amongst commercial sex workers in Kenya found that active gonorrhoeal infection was much more strongly associated with viral shedding than chlamydial infection (Mostad). A study amongst female sex workers in Cote d'Ivoire found that the presence of cervical or vaginal ulcers, gonorrhoea or chlamydia were most strongly associated with HIV–1 shedding in cervicovaginal fluid (Ghys). Similarly, gay men studied in the US had a threefold greater likelihood of viral shedding in their semen if they also had detectable cytomegalovirus infection in their semen (Dyer).

The genital shedding of herpes simplex virus (HSV) and HIV is closely correlated, meaning that a person shedding HSV is more infectious for HIV.

Investigators screened women for HIV who were receiving hormonal contraceptives at the Coast Provincial General Hospital in Mombassa, Kenya between May 1996 and September 1999. In total 210 women tested HIV-positive and were eligible for the study, which involved baseline tests including taking blood samples, plasma viral load, and CD4 cell counts, and the taking of cervical swabs to detect the shedding of HSV and HIV. Follow-up was one and two months later.

The lower limit of detection for the plasma HIV viral load tests was 50 cells/mL and 125 cells m/L for the cervical swabs. Detectable cervical HIV was present in 66.7% of women (28/42) with detectable cervical HSV and in 65.8% of women (98/149) without, leading the investigators to comment that “the detection of cervical HSV DNA was not significantly associated with HIV-1 levels greater than 125 copies/swab or with the detection of HIV-1 infected cells.”

However, the investigators add that although there was no relationship between detection of the two viruses, there was a significant relationship between the quantities of HSV and HIV found in cervical secretions of HSV shedding women. A 10-fold (one log) increase in the amount of cervical HSV was associated with a 1.36-fold increase in cervical levels of HIV. This relationship remained unchanged when differences in CD4 count and plasma viral load were adjusted for. No significant relationship was found between the shedding of cervical HSV and plasma HIV viral load.

The investigators speculate that “local interactions between HSV and HIV-1 may provide a plausible explanation for the association between the genital shedding of the two viruses”. They conclude that “HSV suppression could be an important means of decreasing” the spread of HIV, and given the high levels of HSV infection in their study population call for “additional prospective studies of the relationship between HSV and HIV-shedding.”

The relationship between HSV and HIV infectivity and disease progression is the subject of intensive investigation. A study conducted in the US before the introduction of HAART found that treatment for HSV resulted in a modest reduction in HIV viral load, and a recent South African study found that recent HSV infection was associated with HIV seroconversion.

The health status or disease progression of the infected partner

It has been theorised that individuals are most infectious at two points during HIV infection: during primary infection (the period before antibodies appear) and the symptomatic phase (after symptoms appear). However, research to test this theory has produced contradictory results. As discussed above, levels of HIV in semen and vaginal fluid are unpredictable and not always closely associated with virus levels in blood (see The presence and concentration of HIV above). For example, one study found no difference in seminal viral load between individuals tested during primary infection and others who had more advanced HIV infection (Dyer). But these results need to be treated with caution because they refer to 3 individuals from a study of 101 men.

A large study in Thailand which investigated HIV transmission in 467 couples where the male partner had been identified as HIV-positive found that even when the man's CD4 count was above 500, 39% of women were already infected. These findings suggest that viral load is high even during early disease, although it was not possible in this study to distinguish between infection in primary infection and infection after an immune response had reduced viral load (Nelson).

Another study in Thailand could find no difference in infection rates between women exposed to partners during their primary infection phase and women who began sexual relationships with men after their partners had become infected (Duerr).

An unequivocal link between recent HIV-infection and high risk of HIV transmission was reported at the tenth Conference on Retroviruses and Opportunistic Infections in Boston in February 2003.

The researchers calculated the incidence of HIV transmission per monogamous, vaginal heterosexual act as thus: in the five months immediately following seroconversion, rate of transmission was ten times greater than during chronic infection at 8.2 per 1000. In late stage infection (five to 15 months before death) it was 4.5 per 1000. During chronic infection, once the viral load set point had been reached - after 5 months and before late stage HIV disease - the transmission rate was 0.8 per 1000.

Although it has been understood for some time that there was an association between the very high viral loads seen during the primary infection and a greater likelihood of infecting others, this report from the Johns Hopkins University Rakai Project Team suggests that much HIV prevention work would benefit greatly by focusing on those currently untested or very recently infected.

The strain of virus

There have been recent claims by some scientists that different strains of HIV have differing infectivity, and that strains which are predominant in developing countries could be more easily transmitted through vaginal intercourse and from mother-to-baby.

The evidence that HIV subtypes have differing levels of infectiousness is conflicting. The main evidence comes from test tube experiments which tested the ability of different HIV subtypes to replicate in immune cells called Langerhans cells, which are plentiful in the wall of the vagina. Researchers found that subtype E, which is more common in Asia and Africa, could replicate faster in these cells than subtype B. They suggested that this showed subtype E was likely to be transmitted more easily through vaginal intercourse than subtype B (Soto–Ramirez).

Thai researchers found that when all other factors were controlled for, female partners of men infected with subtype E were 3.1 times more likely to be infected with the same strain than female partners of men carrying subtype B (Kunanusont).

These findings have raised the fear that subtypes such as E may spread rapidly in Europe and North America. However, these subtypes have been transmitted in European countries such as Belgium and the UK without showing a more pronounced transmission rate (although this may be a function of sexual mixing patterns and lower levels of STDs in those countries).

A review by the US Centers for Disease Control and Prevention found little epidemiological evidence that some subtypes are more infectious (Mastro). They made a number of observations:

• Transmission rates of different subtypes are similar from mother to baby in countries such as Thailand where several different subtypes are present in the population.
• Different transmission rates of subtypes amongst the adult population can be explained by the fact that subtypes which first appear in injecting drug users are likely to spread faster because a greater number of high risk contacts take place in this group.
• Where HIV subtypes appear to be confined to particular population groups, this may be a consequence of the route by which HIV entered that population group, not a consequence of its ease of transmission by a particular route. For example, although subtype B HIV is predominant amongst drug injectors in Thailand, whilst subtype E is predominant amongst those who acquired HIV through sexual intercourse, this is almost certainly a consequence of the fact that HIV was introduced into the drug injecting population by drug injectors from the US or Europe, where subtype B predominates.

The question of varying infectiousness remains unanswered, and will require close surveillance in years to come.

Damaged tissue

Damage to tissue in the throat or mouth may facilitate infection; in reported cases of infection through oral sex, inflammation of the throat because of allergy and frequently bleeding gums have been cited along with pharyngeal gonorrhoea as co–factors for infection. In women, oral contraception can sometimes cause cervical ectopy, in which cells in the cervix are exposed on the surface where they can easily be broken, allowing for easier HIV transmission.

Immune system activation

Generalised immune system activation may play some role, as yet unclear, in facilitating HIV infection, in the view of some researchers. They point to increases in viral load when other active infections, such as tuberculosis, are present. However, a study in Tanzania failed to find any evidence that people with a previous history of tuberculosis were more likely to acquire HIV during a five year follow up period (Borgdorff).

However a 1996 study (Goletti) found that levels of HIV in the bloodstream increased from five- to 160-fold in HIV-infected people who develop active tuberculosis. The findings may help explain why HIV-infected people with active TB have a poorer prognosis than HIV-infected people without TB.

References

Borgdorff MW et al: Tuberculosis infection and HIV incidence in Africa, Lancet 348: 1743, 1996.

Druett A et al: Heterosexual transmission during the seroconversion versus the post–seroconversion period, Eleventh International AIDS Conference, abstract Mo.C.571, 1996.

Geijtenbeek, T.B.H. and others. DC-SIGN, a dendritic cell-specific HIV-1-binding protein that enhances trans-infection of T cells. Cell 100(5): 587-597. March 3, 2000.

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

Goletti D, et al. Effect of Mycobacterium tuberculosis on HIV replication: role of immune activation. Journal of Immunology 157: 1271-1278, 1996.

Mastro TD et al: Why do HIV–1 subtypes segregate amongst persons with different risk behaviours in South Africa and Thailand? AIDS 11(1): 1997.

McCutchan, F: HIV Genetic Diversity, Eleventh International AIDS Conference, Abstract No. M02, 1996.

McClelland RS et al. Association between cervical shedding of herpes simplex virus and HIV-1. AIDS 16: 2425-2430, 2002.

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

Nelson K et al. High rates of transmission of subtype E human immunodeficiency virus type 1 among heterosexual couples in northern Thailand: role of sexually transmitted diseases and immune compromise. Journal of Infectious Diseases 180: 337-343, 1999.

Wawer M J et al. HIV-1 Transmission per coital act, by stage of HIV infection in the HIV+ index partner, in discordant couples, Rakai, Uganda. Tenth Conference on Retroviruses and Opportunistic Infections, Boston, abstract 40, 2003.

Gray R H et al. Probability of HIV-1 transmission per coital act in monogamous, heterosexual, HIV-1-discordant couples in Rakai, Uganda. Lancet 2 Apr 14; 357(9263): 1149-1153, 2001.