Multiple and overlapping sexual relationships, commonly referred to as sexual concurrency, are believed by some to account for the rapid spread of HIV infection in sub-Saharan Africa [1]. Research shows that concurrent sex partners are prevalent in several populations hit hardest by AIDS including gay communities, commercial sex workers, and some countries of sub-Saharan Africa. Recent research shows sexual concurrency may be playing a role in resurgent HIV infections among gay and bisexual men in North America [2] and emerging HIV epidemics among men who have sex with men in Asia [3]. In addition to epidemiological trends, the potential impact of concurrent sex partners on the spread of HIV transmission is biologically grounded. Sexual concurrency is thought to afford the rapid turnover of HIV when multiple partners are exposed to the virus during the brief and highly infectious period of acute infection [4]. Mathematical models suggest that concurrent sex partners during acute HIV infection may be a driving force in heterosexually transmitted HIV epidemics of southern Africa [5, 6]. Other known and unknown co-factors for HIV transmission also interact with sexual concurrency to propel HIV epidemics.

While stimulating great interest, the existing empirical research on the role of sexual concurrency in HIV epidemics is not definitive. In the February 2010 issue of AIDS and Behavior (volume 14, Number 1) Lurie and Rosenthal [7] pointed out that sexual concurrency has not yet been empirically shown to increase HIV transmission beyond what would be expected from multiple sex partnerships that do not overlap in time. They also note that epidemiological evidence is mixed as to whether sexual concurrency is spreading HIV in southern Africa and that there is even evidence that polygamy, certainly an example of concurrency, can protect against HIV transmission [7–10]. At the heart of the controversy surrounding the role of sexual concurrency in HIV epidemics is the question of whether limited HIV prevention resources should be directed at interventions to target sexual concurrency [1, 11]. Furthermore, even if interventions are designed to specifically target sexual concurrency, it is not clear how they would differ from programs that aim to reduce not-necessarily concurrent multiple sex partners. Sending a strong message against concurrent partnerships may even have adverse outcomes if people believe they are safe by having serial multiple partners.

The jury may be out on whether sexual concurrency is necessary for the rapid spread of HIV, but there is no disputing that multiple sex partners, whether concurrent or serial, are important in HIV epidemics. Recognizing the need for behavioral interventions that reduce numbers of sex partners regardless of their temporal sequencing is not new. Throughout the 1980s and 1990s social marketing campaigns for HIV prevention in US gay communities commonly promoted reducing numbers of sex partners. Indeed, many of the early HIV prevention successes in gay communities and countries like Uganda are attributed to aggressive efforts aimed at reducing numbers of sex partners [12, 13]. Mass public health message campaigns result in behavior change for only a segment of a population. Thus, more focused behavioral interventions are needed to reduce high-risk behaviors, including reducing numbers of sex partners, in populations with high HIV prevalence.

Several controlled intervention trials have demonstrated significant reductions in numbers of sex partners. Table 1 summarizes the findings from 15 selected prevention trials that report decreased numbers of sex partners over time. All of these studies found evidence for reductions in numbers of sex partners, with eleven trials demonstrating an experimental intervention that reduced numbers of partners to a significantly greater degree than a control condition. The interventions varied in their content and duration, with some lasting several hours and conducted over multiple sessions and others having only one brief session. All of the interventions included a heavy dose of interactive training for communication and preventive behavioral skills. The interventions were tested in various settings and with a wide range of populations including men who have sex with men, women, substance users, adolescents, and patients attending sexually transmitted infection clinics. The magnitude of partner reduction varied, with some studies showing more than a three-fold reduction in numbers of sex partners over time.

Table 1 Partner reduction outcomes from selected behavioral HIV prevention interventions
Full size table

Examining the outcomes reported in these trials suggests that reducing numbers of sex partners may not be any more difficult to achieve than changing other sexual behaviors, such as increasing condom use. However, reductions in sexually transmitted infections (STI) has not been directly linked to partner reductions relative to other changes in behavior. Research is needed to better understand motivations for maintaining multiple sex partners and how the meaning of multiple partners differs by gender, sexual orientation, and culture. The importance of multiple sex partners in facilitating the spread of HIV is indisputable. The importance of acute HIV infection in HIV transmission also applies to multiple partners even if non-overlapping and should therefore remain a focus in HIV prevention. Interventions that have shown promise in reducing numbers of sex partners are available and should be implemented in places with high-HIV prevalence and high-rates of multiple partners, concurrent or not. Given that the role of sexual concurrency in HIV epidemics is not known, limited HIV prevention resources should be concentrated on known risks and evidence-based interventions.