When it comes to HIV, the term "safe sex" is really a misnomer. The risk of transmission exists along a continuum. At one extreme is total abstinence; at the other, multiple acts of passive anal sex. Everything else falls into some vast gray area, and the relative safety of kissing, heterosexual intercourse, fellatio, cunnilingus and other practices has been endlessly debated, discussed and fretted over by researchers, doctors, politicians, health educators, AIDS activists and anyone having sex.
Some things are clear. Touching, holding hands and rubbing against each other contain no risks. Kissing is a theoretical risk but not a real-world one, unless both partners have open sores or lesions in their mouths, or blood from diseased gums, dental work or some other reason. Condoms greatly reduce the risk of intercourse but don't eliminate it; they can break or fall off inside the other person. For many years it was thought that the nonoxynol-9, a spermicide found in many lubricants, could help prevent infection; now the Centers for Disease Control says it might actually facilitate transmission.
In the United States, unprotected anal sex is by far the most effective means of sexual transmission of HIV. It involves greater risk than vaginal intercourse because the lining of the rectum is more likely than the vaginal walls to suffer small fissures or tears that can increase the likelihood of infection. Despite the risk, studies indicate that the rate of transmission to the receptive partner during a single incident of anal sex is quite low -- up to about 3 percent for the receptive partner and extremely tiny, but still possible, for the active partner. In heterosexual intercourse, the woman is at significantly lower risk than the receptive partner in anal sex but at about twice the risk of the male in a two-gender coupling.
However, numerous factors -- such as the presence of blood in the rectum or vagina and high viral load levels in the semen of the HIV-positive person -- can drastically increase the likelihood of infection. The presence of another sexually transmitted disease in the HIV-positive person may also increase his or her infectiousness.
Oral sex is clearly far less risky than anal or vaginal sex. There have been about 30 documented cases of transmission to the receptive partner in fellatio, and the receptive partner in cunnilingus also faces a tiny risk, particularly if the woman is menstruating. But gathering epidemiological data on oral sex transmission is complicated because most people combine it with other sexual activities. Pre-seminal fluid has been found to contain virus but in smaller quantities than semen, so swallowing semen carries greater risk. A few cases of transmission from the receptive to the inserting partner in fellatio have been documented.
Protease inhibitors and effective drug "cocktails" have affected HIV transmission rates in ways that are not fully understood. Because the medications suppress the viral load, scientists generally believe that transmission is less likely to occur. However, it is clear that the new drugs have led many to conclude that AIDS no longer poses the threat it once did; as a result, some people have started engaging in behaviors they might have avoided earlier. But because the virus becomes resistant to the drugs in some of the people taking them, AIDS educators warn that those who take chances may get infected with viral strains that don't respond well to treatment.