Rapid turnover of sexual partners in brief intervals of time (e.g., multiple recent partners) as well as repeated sexual encounters with multiple partners that overlap in time (e.g., sexual concurrency) are critical factors in establishing and maintaining sexually transmitted human immunodeficiency virus (HIV) epidemics. Evidence from mathematical modeling and field research shows that multiple sexual partnerships may amplify sexually transmitted infections (STI), including HIV.1–3 A national study of Native American men who have sex with men found 34% HIV prevalence and 55% cumulative prevalence of concurrency, indicating an alarmingly high risk for HIV transmission.4 Multiple recent sex partners can increase exposures proximal to the acute phase of HIV infection when individuals are most infectious.2,5,6 In addition, STI themselves cause HIV shedding in genital secretions, further elevating HIV infectiousness.7 The combination of multiple recent sex partners and co-occurring STI is particularly alarming for the potential spread of HIV infection. Senn et al.8 found that 2 of 3 STI clinic patients in steady relationships reported at least 1 additional concurrent sex partner. High prevalence of multiple recent sex partners in combination with STI has been reported in other studies.9–11
The rapid spread of HIV in southern Africa has been attributed, at least in part, to multiple sex partners.3,12 Of particular importance are overlapping sex partners, a subset of multiple partners referred to as concurrent. For example, a household survey in Botswana found that 23% of sexually active persons reported concurrent sex partners.13 A study of South African youth found that 13% of 16- to 26-year olds living in Cape Town reported concurrent sex partners.14 In addition, reductions in HIV incidence observed in Uganda during the 1990s are believed to have occurred after reductions in concurrent sex partnerships.15 Men are more likely to have concurrent sex partners and are, therefore, more likely to expose multiple partners to HIV during acute infection and during episodes of co-occurring STI.16
Multiple sex partners are also known to co-occur with other behavioral risks for STI. Senn et al.8 found that men in the Northeastern United States who were recently diagnosed with an STI commonly reported multiple sex partners and drank alcohol before sex. A national study in the United States also found that men who had concurrent partners were more than twice as likely to use drugs or drink alcohol before sex.10 These findings have been replicated in other studies of men to suggest a link between multiple and concurrent sex partners and substance use.17 In South Africa, a country with among the world's fastest growing HIV epidemics, alcohol use is a known risk factor for STI and/or HIV.3,14,18 For example, Townsend et al.19 found that problem drinking was associated with relationship violence among men who reported multiple sex partners. The study did not, however, investigate whether alcohol use was tied to having multiple recent sex partners. To our knowledge, no study has examined alcohol use as a contributing factor to unprotected sex in relation to multiple recent sex partners in southern Africa.
The current study examined the association between alcohol use and multiple recent sex partnerships among men and women seeking treatment for an identified STI in South Africa. We hypothesized that alcohol use would be associated with multiple recent sex partners and that alcohol use would moderate the relationship between multiple recent partners and frequency of unprotected sex. We also examined a population estimate of sexual concurrency among men and women seeking treatment for an identified STI in Cape Town.
Participants and Setting
The clinic that served as the site for this study is located in an urban neighborhood adjacent to Cape Town's downtown business district. The clinic serves over 1800 patients per month, more than 90% of whom are seeking STI diagnostic and treatment services. Approximately half of all patients seen at the clinic have previously received STI services. The estimated HIV prevalence at the clinic is 25% among persons opting for HIV testing. A total of 529 men and 210 women received STI services. The majority (60%) had previously received STI treatment services from the clinic.
Demographic and Health Characteristics.
Participants reported their age, race (black/Xhosa and all other races), years of formal education, whether they were employed, and their marital status. Participants also indicated whether they had ever traded sex for money, a place to stay, or material goods. The survey included HIV testing history and experience of STI symptoms and diagnoses.
Sexual Behaviors and Multiple Recent Partners.
Participants reported their number of male and female sex partners and frequency of vaginal or anal intercourse in the previous 2 months. Sexual behavior assessments were constructed using partner-level measures such that participants were instructed to report their frequency of intercourse occasions, with and without condoms, and the frequency of intercourse occasions that included alcohol use before or during sex for each of up to 3 sex partners in the previous 2 months. For each partner, we also asked the partner's age, their relationship status, the time that they last had sex with that person, and whether they believed the partner was having sex with other partners. After the first partner assessment, the questions were repeated for a second partner and then again for a third partner. Participants with only 1 partner did not complete additional measures for partners 2 and 3, and participants with 2 partners did not complete measures for partner 3. Age differences between participants and their partners were also calculated (participant age—-partner age). For participants with 2 or more partners, we calculated the participant–partner age difference using the most recent nonprimary partner. We also asked participants to report the number of sexual partners they believe that most of their current friends have.
Participants reported their frequency of alcohol use in the previous 2 months and whether they had consumed 6 or more drinks in a single drinking episode during that period. Measures were adapted from the frequency and quantity items of the Alcohol Use Disorders Identification Test.20 In addition, participants reported whether they used cannabis (dagga), the most common nonalcohol substance in South Africa, in the previous 2 months.
Participants were referred to the study after their consultation with a nurse for STI diagnosis and treatment. All potential participants present in the clinic were referred to the project recruiter, and 87% of patients approached consented to the study and completed the survey. Less than 10% of participants could not complete the surveys on their own and required interviewer administration. All measures were administered in English and Xhosa (an indigenous African language), the 2 languages spoken by nearly all clinic patients. Participants received 15 South African Rand (approximately US$22) for completing the survey.
For all analyses, participants were defined as either not having or having multiple recent sex partners. Participants who reported only 1 partner in the previous 2 months were defined as having 1 partner and participants with 2 or more partners in the past 2 months, including those with concurrent and nonconcurrent partners, were defined as having multiple recent sex partners. Restricting multiple partnerships to a 2-month timeframe increases the likelihood of partner exposures during acute HIV infection.3,12 Multiple recent sex partners were further described by examining the timeframes within which the specific sexual partnerships occurred. We also calculated the concurrency index developed by Ktrezschmar and Morris.21,22 The concurrency index (k) is the estimated fraction of sexual partnerships that are concurrent at any given point in time; k = σ2/μ + μ − 1. Values of k range from 0 to 1, with higher values indicating greater concurrency in the population. Initial analyses determined that only 16 of the 210 women (9%) reported multiple recent sex partners. We, therefore, conducted the main study analyses comparing participants with 1 partner to those with multiple recent partners for men only.
Sexual behaviors were examined using frequencies of both unprotected and protected sex acts. Frequencies of sexual behaviors were summed across partners. We also calculated the proportion of sexual occasions that were protected by condoms and the proportion of sexual occasions that involved alcohol use, including when the participant drank and when their partner drank. The proportions were calculated across partners and therefore allowed comparisons between participants with varying number of partners. Group comparisons were made between participants who did and did not have multiple recent sex partners using bivariate logistic regressions, reporting unadjusted odds ratios (OR) and 95% confidence intervals (CI). We also conducted separate logistic regressions to compare men who reported drinking (drinkers) and those who did not drink (nondrinkers) on sexual behaviors. Finally, to test for moderating effects of alcohol on sexual risks in relation to multiple recent sex partners, separate logistic regression models were performed comparing multiple partnership groups whereas adjusting for alcohol use in sexual contexts. For all analyses, a list-wise deletion procedure was used for missing values and statistical significance was defined as P < 0.05.
Among the 844 STI clinic patients invited to participate in the study, 739 (87%) consented and completed the measures. The distributions of sex partners for men and women are shown in Table 1. The mean number of sex partners was 1.45 (standard deviation [SD] = 0.83; range, 1–12) in the previous 2 months. A total of 264 (31%) participants reported 2 or more sex partners in the previous 2 months. However, less than 1% of persons reporting multiple recent sex partners were women. Using the mean (μ = 1.45) and variance (σ2 = 0.69) as a population estimate, we calculated a concurrency index k of 0.92, indicating a high level of sexual concurrency in the entire sample. Among men, the mean number of sex partners was 1.58 and σ2 = 0.61, yielding k = 0.96. For women, the mean number of sex partners was 1.11, σ2 = 0.76, k = 0.79. Thus, the fractions of concurrency were high among men and moderate among women.
Rapid turnover of sex partners was common among men with multiple recent sex partners; 31 (12%) men had sex with at least 2 of their partners less than a week apart, 87 (35%) had 2 partners about a week apart, and 97 (39%) had sex with 2 or more partners within a month of each other. The remaining 33 (13%) men with multiple partners indicated sexual contacts beyond 1 month but less than 2 months apart. These overlapping timeframes also support substantial concurrency in sexual relationships.
Risk Factors Associated With Multiple Recent Sex Partners
The mean age of the sample was 29.8 (SD = 0.78) and the mean years of education was 10.1 (SD = 2.3). Nearly all participants (96%) were black South African of Xhosa heritage. Less than 1 in 4 (23%) participants were married and 66% were employed. A total of 77% had been tested for HIV and 7.5% of those tested received an HIV-positive test result. The only demographic characteristics associated with multiple recent sex partners were gender (OR = 0.09, P < 0.01, 95%CI = 0.06–0.16) and employment status (OR = 1.70, P < 0.01, 95%CI = 1.21–2.39); men and employed participants were significantly more likely to report multiple recent partners. Individuals who reported multiple recent sex partners were also less likely to have been tested for HIV (OR = 0.65, P < 0.01, 95%CI = 0.45–0.92) and less likely to have tested HIV-positive (OR = 0.28, P < 0.01, 95%CI = 0.12–0.69). Multiple recent sex partners were also significantly related to perceptions that their friends have more sex partners (OR = 1.34, P < 0.01, 95%CI = 1.17–1.51).
Among men who had multiple recent sex partners, we found a greater likelihood of being unmarried, slightly younger age, and having greater years of education (Table 2). Men who had multiple recent sex partners also reported knowing more men and women who they believed had multiple partners. In addition, multiple partners were related to greater age disparity between themselves and their sex partners; men with multiple partners were significantly older than their partners relative to men who reported 1 recent partner.
Men with multiple recent sex partners were more likely to have experienced genital pain and more likely to have paid for sex than men who reported 1 partner (Table 3). There were no differences between the sexual partner groups for having had previous STI diagnoses, experienced genital ulcers, or genital discharge.
Multiple Recent Sex Partners and Substance Use
Participants who reported multiple recent partners also drank more frequently, drank in greater quantities, and were more likely to use cannabis (Table 3). Men with multiple recent partners were also significantly more likely to drink alcohol and report partners drinking before or during sexual encounters. Because frequencies of alcohol use in sexual contexts may have been weighted by number of sex partners, we also examined the proportion of sexual occasions that involved the participant or their partner drinking. Results showed that men with multiple recent sex partners had a significantly greater proportion of sexual encounters in which they drank before sex. Men with multiple recent sex partners indicated a significantly greater proportion of sexual encounters that involved any alcohol use by them or their partners. The difference between groups for the proportion of sexual encounters in which their partners drank before sex was not significant.
Table 4 shows sexual behaviors, proportion of sexual behaviors protected by condoms, and perceptions of partner concurrency comparing men who had 1 partner and men with multiple recent partners stratified for current alcohol use. Results showed that multiple recent sex partners were associated with higher rates of unprotected and protected sexual intercourse (Table 4, first column OR). In addition, men with multiple partners were more likely to believe that their sex partners also had concurrent partners.
In a second analysis, comparisons of men who reported and who did not report alcohol use showed that drinkers reported more sex partners in the past 2 months than men who did not drink (Table 4 second column OR). However, global alcohol use was not associated with rates of unprotected sex or proportion of sexual behaviors protected by condoms. Finally, comparison of recent multiple partner groups adjusting for alcohol use indicated that drinking did not moderate the association between multiple recent sex partners and unprotected intercourse (Table 4, third column OR).
A growing body of evidence suggests that rapid turnover of sexual partnerships promotes the spread of HIV.20 The current study found that nearly half of men seeking treatment for an identified STI had 2 or more sex partners in 2 months, well within the timeframe of acute HIV infection. The rate of sexual concurrency among men exceeded that observed in women as well as in that reported in nonclinic samples in southern Africa.3 For example, research conducted in general populations throughout southern Africa has found a considerable range of sexual concurrency, with fractions of relationships that were concurrent (k index) ranging from 0.98 in Yaounde' Cameroon to 0.44 in Kisumu Kenya, and 0.18 in Dakar Senegal.22 We found that1 in 5 men with multiple recent sex partners was married, and men with multiple partners had more years of education, were more likely to believe that the people they know have concurrent sex partners, and were significantly older than their partners. These factors replicate previous research to suggest that characteristics associated with multiple recent sex partners are reliable across vastly different samples.3,12,14,15
Our study provides among the first analyses of substance use in relation to multiple recent sex partners in southern Africa. Men receiving STI clinic services who had multiple recent sex partners reported greater alcohol use than their counterparts who reported 1 recent sex partner. One in 4 sexual occasions among men with multiple sex partners involved their using alcohol and nearly 40% of their sexual occasions involved either they themselves or their sex partner drinking. Alcohol is a known correlate of sexual risk behaviors in southern Africa and these data further bolster the importance of alcohol's role in sexual risks.16 Although drinkers reported a greater frequency of recent sex partners, we did not find that alcohol use moderated the association between multiple recent partners and unprotected sex acts. Specifically, higher rates of unprotected and protected sex among men with multiple recent sex partners remained significantly different from men who did not have multiple recent sex partners after controlling for alcohol use in sexual contexts. These findings, therefore, demonstrate that both multiple partners and alcohol use are critical, albeit independent, risk factors for HIV transmission in South Africa.
The current findings should be interpreted in light of their methodological limitations. The study was conducted in a single STI clinic in Cape Town. Because the clinic was dedicated to STI services, it may represent a different population of individuals who seek treatment for an identified STI. In addition, there are substantial cultural differences throughout South Africa that cannot be assumed similar to our mostly Xhosa sample. We, therefore, suggest caution when generalizing these results to other regions within South Africa. We also relied solely on self-report for behavioral assessments. Although we used anonymous survey methods, our behavioral measures are subject to underreporting biases and should be taken as lower-bound estimates of sexual risk and substance use behaviors. The surprisingly low rate of multiple recent partners among women in the STI clinic setting may have resulted from women underreporting number of partners. We were also unable to link our behavioral data to new or recurrent STI and HIV infection. Future research should be conducted using prospective designs to confirm associations between multiple recent sex partners, behavioral risk factors, and HIV infection. Despite these limitations, we believe that our findings have implications for STI and/or HIV prevention interventions for people seeking STI treatment.
Multiple recent sex partners and sexual concurrency are common in populations with high-rates of HIV infection.4 Fortunately, behavioral interventions have demonstrated significant effects on reducing number of sex partners.21 Although no behavioral intervention has yet specifically targeted reducing concurrent partners, reductions in number of sex partners will, by definition, reduce concurrent partners. There have also been reductions in number of partners when public policies provide adequate resources for behavioral interventions and culturally relevant messages.13 The urgent need to address multiple and concurrent sex partners for HIV prevention should look toward existing behavioral interventions and past public health messages that have demonstrated efficacy in reducing number of sex partners.
1. Doherty IA, Minnis A, Auerswald CL, et al. Concurrent partnerships among adolescents in a Latino community: The Mission District of San Francisco, California. Sex Transm Dis 2007; 34:437–443.
2. Morris M, Kurth AE, Hamilton DT, et al. Concurrent partnerships and HIV prevalence disparities by race: Linking science and public health practice. Am J Public Health 2009; 99:1023–1031.
3. Mah TL, Halperin DT. Concurrent sexual partnerships and the HIV epidemics in Africa: Evidence to move forward. AIDS Behav 2010; 14:17–24.
4. Cassels S, Pearson CR, Walters K, et al. Sexual partner concurrency and sexual risk among gay, lesbian, bisexual, and transgender American Indian/Alaska Natives. Sex Transm Dis 2010;37:272–278.
5. Morris M, Kretzschmar M. Concurrent partnerships and the spread of HIV. AIDS 1997; 11:641–648.
6. Pilcher CD, Tien HC, Eron JJ Jr, et al. Brief but efficient: Acute HIV infection and the sexual transmission of HIV. J Infect Dis 2004; 189:1785–1792.
7. Johnson LF, Lewis DA. The effect of genital tract infections on HIV-1 shedding in the genital tract: A systematic review and meta-analysis. Sex Transm Dis 2008; 35:946–959.
8. Senn TE, Carey MP, Vanable PA, et al. Sexual partner concurrency among STI clinic patients with a steady partner: Correlates and associations with condom use. Sex Transm Infect 2009; 85:343–347.
9. Nelson SJ, Manhart LE, Gorbach PM, et al. Measuring sex partner concurrency: It's what's missing that counts. Sex Transm Dis 2007; 34:801–807.
10. Adimora AA, Schoenbach VJ, Doherty IA. Concurrent sexual partnerships among men in the United States. Am J Public Health 2007; 97:2230–2237.
11. Gorbach PM, Stoner BP, Aral SO, et al. “It takes a village”: Understanding concurrent sexual partnerships in Seattle, Washington. Sex Transm Dis 2002; 29:453–462.
12. Halperin DT, Epstein H. Concurrent sexual partnerships help to explain Africa's high HIV prevalence: Implications for prevention. Lancet 2004; 364:4–6.
13. Carter MW, Kraft JM, Koppenhaver T, et al. “A bull cannot be contained in a single kraal”: Concurrent sexual partnerships in Botswana. AIDS Behav 2007; 11:822–830.
14. Mah TL. Prevalence and correlates of concurrent sexual partnerships among young people in South Africa. Sex Transm Dis 2010; 37:105–108.
15. Slutkin G, Okware S, Naamara W, et al. How Uganda reversed its HIV epidemic. AIDS Behav 2006; 10:351–360.
16. Gregson S, Nyamukapa CA, Garnett GP, et al. Sexual mixing patterns and sex-differentials in teenage exposure to HIV infection in rural Zimbabwe. Lancet 2002; 359:1896–903.
17. Operario D, Smith CD, Arnold E, et al. Sexual risk and substance use behaviors among African American men who have sex with men and women. AIDS Behav. In press.
18. Kalichman SC, Simbayi LC, Kaufman M, et al. Alcohol use and sexual risks for HIV/AIDS in sub-Saharan Africa: Systematic review of empirical findings. Prev Sci 2007; 8:141–151.
19. Townsend L, Jewkes R, Mathews C, et al. HIV risk behaviours and their relationship to intimate partner violence (IPV) among men who have multiple female sexual partners in Cape Town, South Africa. AIDS Behav. In press.
20. Saunders JB, Aasland OG, Babor TF, et al. Development of the alcohol use disorders identification test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption—II. Addiction 1993; 88:791–804.
21. Kretzschmar M, Morris M. Measures of concurrency in networks and the spread of infectious disease. Math Biosci 1996; 133:165–195.
22. Lagarde E, Auvert B, Caraël M, et al. Concurrent sexual partnerships and HIV prevalence in five urban communities of sub-Saharan Africa. AIDS 2001; 15:877–884.