Safer sex is a gay, grass-roots invention. In fact, just about every behavioral technique used in the U.S. to stop or slow the spread of HIV was invented and first practiced by gay men. These techniques were considered risky by the public health establishment when they were first promoted. It was only later, as their effectiveness was gradually revealed, that public health authorities accepted them and turned them into public health policy. Now, a recent trend in HIV prevention among gay men, called serosorting, is getting a lot of attention.
But how effective is serosorting? Serosorting is the practice of selecting partners according to whether they are HIV positive or HIV negative, having sex only with someone of the same HIV status, or always using condoms only with people of a different HIV status. It implies having unprotected anal intercourse when both partners are already HIV positive, or when both are HIV negative.
Other behavioral strategies that aim to prevent HIV infection without using condoms fall within the larger category of "seroadaptation". These include "strategic positioning", in which the insertive (top) and receptive (bottom) sexual roles are chosen according to HIV status, relative risk, or viral load, with the uninfected partner typically taking the insertive role. Another seroadaptive behavior is the use of viral load as a basis for decision-making. When a person with HIV has an undetectable viral load, some gay men feel it is safe to have sex with an uninfected partner without using condoms, especially if the man with HIV is the bottom. But while an undetectable viral load lowers the risk of HIV transmission, it does not eliminate it.
What serosorting and other seroadaptive behaviors have in common is the search for ways to have unprotected or condomless sex while still avoiding HIV infection. Since all these practices feature sex without condoms, public health officials and HIV advocates are concerned that this new "grass-roots" method of HIV-prevention may turn out to be risky and lead to an increase of new infections among gay men. Are they right?
There are real issues with risk-reduction practices that try to do away with condoms. To be effective, they need to be used carefully, intelligently, consistently, and with awareness of the larger context. Serosorting is certainly safer for HIV-positive men than it is for HIV-negative men. First, while people with HIV can be superinfected with a different strain of HIV, experts are divided on how common this is and its consequences. Second, men who have tested positive know their HIV status -- someone who says he has HIV is probably not lying. Finally, sex between HIV-positive men cannot transmit HIV to uninfected men -- a lesser impact on public health.
For HIV-negative men, however, strategic positioning, partner serosorting, and unprotected receptive anal sex with an HIV-positive partner who has an undetectable viral load are all subject to misuse. In 2008, 63% of sexually active HIV-positive men in U.S. urban areas between the ages of 18 and 29 were unaware they had HIV, according to the latest figures from the CDC's National HIV Behavioral Surveillance report. This is quite worrying, because a person who has been recently infected is extremely infectious. In fact, approximately half of new sexually acquired HIV infections in the U.S. are transmitted by the 21% of HIV-positive people who are unaware of their HIV status. This means that having unprotected sex with a guy who says, even believes, that he is uninfected may be very risky for someone who is HIV negative.
One study that ran from 2001 to 2007 observed serosorting trends among HIV-positive and negative gay men who received services from a clinic in Seattle, Washington. The study showed that the practice of serosorting increased over the six years, especially among HIV-positive gay men. Of those men who became HIV positive over the course of the study, 32% were gay men who practiced serosorting. Perhaps those men had taken other, greater risks in their sexual practices. Still, those findings suggest that serosorting does not eliminate the risk of contracting HIV. The level of protection that comes from serosorting is uncertain. But what is certain is that there are many ways serosorting by HIV-negative men can fail to provide protection from HIV.
Recent studies have shown that HIV treatment can substantially lower the rate of transmission from a positive but treated partner to a negative partner, especially when the positive partner has an undetectable viral load. But this belief may have particularly dangerous results for gay men. First, new research suggests that only 19% of HIV-positive people in the U.S. actually have an undetectable viral load. The CDC has a more optimistic estimate of 28%, based on more recent data and a different method of calculation, but it still seems reasonable to assume that only about a quarter of all Americans currently living with HIV have their viral load down to an undetectable level. Second, viral load fluctuates significantly, even in people who succeed in getting it down to undetectable levels. Third, even if the virus cannot be detected in your blood, it may be present in your semen (which is not tested).
A recent study of 101 HIV-positive men who have sex with men found this to be true. The Fenway Institute in Boston discovered that roughly a quarter of the men with undetectable HIV in their blood had detectable HIV in their semen. The study also showed, however, that men with undetectable viral loads in their blood had significantly lower viral loads in their semen compared to men with detectable HIV in their blood. How exactly that translates to infectiousness remains to be seen. The study did not look at whether those with low-but-detectable levels of HIV in their semen were more likely to transmit the virus than those whose semen had undetectable viral loads. Researchers are also concerned that the HIV present in semen when a person's viral load is undetectable in the blood could be resistant to a particular HIV medication, in which case exposure to that semen could lead to the spread of a treatment-resistant virus.
The most striking evidence for an undetectable viral load reducing HIV infection comes from a recent study of serodiscordant heterosexual couples (one partner HIV positive and the other HIV negative). It showed a 96% reduction in HIV transmission when the HIV-positive partner started HIV treatment early. It may be assumed, however, that much if not all of the sex that occurred among the heterosexual couples was vaginal sex. Unfortunately, the risk of HIV transmission through anal sex is much higher than through vaginal sex. So gay men should not throw out their condoms because of the results of this study. More research is needed to understand if an undetectable viral load in the blood reduces HIV transmission in people who practice anal sex.
There are also other unwelcome consequences of trying to move beyond condom use while still trying to avoid HIV. The technique of partner sorting rather than condom use leads some gay men to practice racial sorting, because the prevalence of HIV is known to be higher among black men and Latinos than among white men. Sorting partners by age is another misguided attempt to stay HIV negative without using condoms. These techniques only serve to continue misinformation about contracting HIV, while promoting racist and ageist behavior.
Although there is much to be concerned about regarding serosorting, we should also remember that the riskreduction practices that are currently approved as safe were not considered safe when they were first invented. What is now called safer sex (using condoms) is a case in point. Nowadays, safer sex is routinely contrasted with risk-taking, barebacking, and other supposedly irresponsible forms of unprotected sex. The terms "safe" and "unsafe" sex are pitted against each other as polar opposites. But the use of condoms for anal sex is not now, nor has it ever been, absolutely safe. Condoms can break, especially if misused. People have to know how to put them on properly, use a water-based lubricant, store them properly, and not use them past their expiration date.
When the concept of safer sex was introduced, no one knew whether it would prove effective. When AIDS first appeared, its cause was a mystery. People blamed everything from poppers to hot tubs to leather. Truly safe sex in those days meant only one thing -- no sex at all. Many voices were raised, both within and outside the gay community, to demand that gay men stop having sex. The response of certain gay men back in 1983 was to say, "No, we will not stop having sex -- we just created a vibrant, open, public sexual culture for the first time in history! What is killing us is not some divine punishment or evil fate, but a virus or bacterium that we don't understand yet. So we'll try to protect ourselves, and our partners, by using condoms for anal sex."
Scientists did not prove condoms could prevent the spread of AIDS until 1985. Prior to this, AIDS prevention campaigns warned gay men that the use of condoms for anal sex was "possibly unsafe". The use of condoms was a gamble by gay men that involved a willingness to accept a reasonable but unknown risk, in order both to defend their lives from a lethal epidemic and to preserve their erotic pleasures.
The use of condoms for anal intercourse, the practice of oral sex without ejaculation, and other kinds of risk-reduction techniques have proven practical and effective in slowing or reducing the spread of HIV, despite having once been considered risky. This is important because it is unreasonable and impractical to expect people to take no risks at all. A policy of total risk elimination can be difficult to sustain over a person's entire lifetime. Too strict an insistence on eliminating all risk can backfire, leading to periodic and highly risky lapses, which can be quite dangerous. The safety of any specific prevention technique should be measured not just by how well it protects people from infection in any particular act of sex, but how well it protects them over the course of a lifetime.
Once an infectious disease has taken hold in a population, it is very difficult to get rid of it by purely behavioral means. It's all very well to ask people to alter their behavior, but no one expects behavioral interventions to be completely effective. During a flu outbreak, we may try to wash our hands and cough into our sleeves, but we don't think it's up to us as individuals to stop the epidemic by altering our behavior. We place our hopes in a vaccine. Until there is a vaccine or a cure for AIDS, behavioral interventions remain our best weapon (though we are gradually finding biomedical options, too -- see Can a Pill a Day... Keep HIV Away?). Meanwhile, some of us are simply going to get sick. That is not necessarily our fault; there's an epidemic out there!
It is clear why the safer sex techniques that gay men invented could not have been embraced at the outset by public health authorities; they were not absolutely safe. They involved risks, sometimes quite significant risks, depending on how skillfully they were practiced, how intelligently and consistently they were put into operation, and how the larger social context of their use evolved. Risk-reduction practices have to be monitored so that they do not lead to risk increase.
What these historical examples imply is that "safety" and "risk" are not black-andwhite categories. A lot of what passes today for safer sex is the result of earlier spontaneous improvisation by gay men with varying degrees of risk. This collective effort continues, as gay men experiment with their own bodies to find compromises and practical solutions that may prevent the spread of HIV while not destroying their unique cultures of pleasure.
Serosorting is the latest of these attempts. But when HIV-negative men use it instead of condoms, especially with casual partners or with those whose HIV status is unknown (which is often what "HIV-negative" really means), they are taking significant risks -- both for themselves and for the gay community as a whole.
"Negotiated safety" is a concept that might be called "serosorting plus". It involves an agreement between two HIV-negative partners in a stable relationship, who after being together for a while, repeatedly testing negative, and talking through the issue, decide to give up using condoms with each other. They also agree either to be monogamous or always to have safer sex with other partners. Some agreements include a decision not to have anal sex with casual partners.
"Negotiated safety" was coined in 1993 by the Australian researcher Susan Kippax, who discovered this prevention technique by studying how some HIV-negative gay male couples actually behaved. She argued that not all condomless sex should be considered unsafe -- it depended on the context. The same thing is true of serosorting in general. In the case of negotiated safety, Kippax found that couples who had an agreement not to have anal sex with casual partners turned out to have less chance of being infected.
The effectiveness of negotiated safety as a prevention technique can vary widely. To make it work, it should involve:
Without the final element, an agreement could be very risky, since few of us would likely admit to our primary partners that we had had unsafe sex with someone else if we thought that such an admission would mean the end of our primary relationship.
The same thing applies to monogamy as a prevention technique. After all, being in a monogamous relationship with a partner who is unfaithful but afraid to admit it does not afford any protection. That is one of the reasons why love itself can be a risk factor for HIV infection.
Does negotiated safety really work? In 2005, a U.S. survey of 340 HIV-negative gay men in San Francisco found that 38 of them were in long-term seroconcordant HIV-negative relationships and had some form of negotiated safety. Of those men, 11 had broken their own rules in the preceding three months.
Three-quarters of the men also had a rule that they must tell their primary partners if they had broken their agreement. This did seem to help -- only 18% of those with an "always tell" rule had strayed outside their agreement, while the 60% of those who had no such rule had in fact broken their agreement.
Finding ways to push safer sex beyond condom use and exploring ways to have safer sex without condoms could prevent the "condom fatigue" that is now so common in the gay community. And it should also remind us that the condom is our friend, insofar as using it gives us more sexual freedom than we can safely have without it. It also sends the message that even though HIV prevalence is higher in the gay community than among heterosexuals, it is still possible for both HIV-positive and HIV-negative gay male couples to enjoy the intimacy of condom-free sex, as their straight counterparts routinely do.
Gay men have a bad rap about sexual risktaking. But we shouldn't lose sight of the fact that the majority of gay men in the U.S. are not HIV positive, that they consistently practice safer sex, and that most will likely remain uninfected. Until there is a vaccine or a cure for AIDS, behavioral interventions remain our best weapon. As we wait for better information on the effectiveness of serosorting and other grass-roots HIV prevention techniques, efforts should be made to remind gay men that we were the first to adopt the condom as a weapon against HIV -- and as our ticket to an expansive and nondiscriminatory sexual culture.
David Halperin is the W. H. Auden Distinguished University Professor of the History and Theory of Sexuality at the University of Michigan. Robert Valadéz is a policy analyst at GMHC.