HIV risk levels for the insertive and receptive partner in different types of sexual intercourse

Summarised from Boily MC et al. Heterosexual risk of HIV-1 infection per sexual act: systematic review and meta-analysis of observational studies. Lancet Infect Dis 9: 118-29, 2009.

A combined analysis of all studies of HIV-transmission risk undertaken to date in high-income countries has found that unprotected vaginal intercourse with an HIV-positive man who is chronically infected and not on antiretroviral treatment (ART) carries an estimated 0.08% risk of infection for the woman per sexual act. In other words, an average of one transmission is expected to occur for every 1250 instances of exposure. Combined data from studies undertaken in low-income countries have yielded a considerably higher estimated risk level of 0.30% for the woman per sexual act: here, an average of one transmission is expected to occur for every 333 instances of exposure.²

In the combined analysis for high-income countries, the risk for a man who has unprotected vaginal intercourse with an HIV-positive woman who is chronically infected and not on antiretroviral treatment (ART) is estimated to be 0.04% per sexual act. In other words, an average of one transmission is expected to occur for every 2500 instances of exposure. Combined data from studies in low-income countries indicate a much higher risk level, 0.38%: here, an average of one transmission is expected to occur for every 263 instances of exposure.²

Women are thought to be at a greater risk than men of acquiring HIV during vaginal intercourse for the following reasons:

• The exposed surface area of the vagina is larger than that of the penis

• Vaginal intercourse causes minute tears in the vaginal lining

• The tissue of the vaginal lining contains certain types of cells that HIV can easily enter

• Semen remains in the vagina for a prolonged period.

HIV can also pass through the skin of the penis, and certain factors (see Important factors affecting risk, below) can increase the likelihood of this happening during vaginal intercourse.

Researchers are uncertain about how to account for the differences observed between the risks of vaginal transmission in high-income versus low-income countries. One possibility is that less accurate estimates came from low-income countries (e.g., due to the under-reporting of high-risk behaviours in the studies that took place there). However, since there are many factors affecting risk levels, they also suggest that any number of as-yet-undetermined viral, biological and genetic factors may be implicated in these differences.²

Source: Jin F et al. Per-contact probability of HIV transmission in homosexual men in Sydney in the era of HAART. AIDS, published online ahead of print, 2010.

Although the risk for the insertive partner in anal intercourse has previously been considered to be much lower than the risk for the receptive partner, this study suggests that for a man who is not circumcised, the risks of insertive intercourse are broadly similar to the risks of receptive intercourse without ejaculation.

Few studies have examined the risk of heterosexual unprotected anal sex. For an HIV-negative woman the per-act risk of anal intercourse is thought to be around ten times higher than the per-act risk for vaginal intercourse.⁸ There are no reliable data regarding the risk for the insertive HIV-negative man of anal intercourse between a man and a woman, but it is likely that the per-act risk is similar to sex between men.²

Oral sex can mean fellatio (mouth-penis sex); cunnilingus (mouth-vagina sex); or anilingus (mouth-anus sex). Fellatio is the only type of oral sex that carries more than a theoretical risk of HIV transmission, although the receptive partner in fellatio (the person who takes the partner’s penis into his or her mouth) is still much less likely to acquire HIV than the receptive partner in anal or vaginal intercourse. The risk for the insertive partner in fellatio is thought to be virtually nonexistent.

Researchers disagree, however, about the receptive partner’s exact risk level, with estimates ranging from zero risk (based on epidemiological studies amongst heterosexuals) to a 1-in-2500 risk of HIV infection (based on case reports amongst men who have sex with men). The uncertainty partly reflects the fact that most people who perform fellatio also engage in other higher-risk sexual activities; HIV transmission in those situations cannot therefore be attributed to fellatio in particular.

The receptive partner’s HIV risk level in oral sex is generally thought to be higher if he or she has bleeding gums or has cuts, sores or other abrasions inside the mouth, including those caused by dental procedures. However, there is no definitive evidence regarding such factors.

Researchers have identified several likely reasons for the relatively much lower risk of acquiring HIV from fellatio than from other forms of sex. An enzyme in saliva inhibits HIV; the low salt concentration of saliva may also inhibit HIV transmission; and tissues in the mouth and throat appear to be less susceptible to HIV than are genital or anal tissues. It is also probable that even if semen is swallowed, digestive enzymes can destroy HIV in the stomach.