Concurrent sexual partnerships are widely believed to play a significant role in the transmission of sexually transmitted diseases (STDs), including human immunodeficiency virus (HIV), and there is growing advocacy for concurrency reduction as a key prevention tool.1–3 These views are not, however, unanimously held and opponents argue that the empirical evidence is not strong enough to support this hypothesis and justify such advocacy.4,5
Mathematical models have demonstrated that STDs can spread more rapidly in sexual networks that include concurrent sexual partnerships.6,7 Some empirical studies have found that individual concurrency—having more than one partner oneself—is associated with a higher STD risk8 and with HIV status.9 On the other hand, other empirical studies have found no association between individual concurrency and STD risk10 and HIV status.11,12
These inconsistent findings can be explained by the fact that individual concurrency increases the risk of transmitting an STD, not of acquiring it,13,14 and thus measures of individual concurrency must act as a proxy for some other correlated factor that is omitted from analyses. Consistent with the mechanism by which concurrency could influence STD transmission, individual concurrency was associated with increased odds of transmitting an STD in the United States,15 and partner concurrency—partners with other partners—was associated with increased odds of having an STD in Russia16 and having HIV among monogamous women in Tanzania.17 However, Landman et al (2008) found no relationship between HIV and partner concurrency among men or the full sample of sexually active women in Tanzania17; neither did Jewkes et al (2010) or Tanser et al (2011) among South African women.18,19
One potential factor that could help explain these inconsistent findings is that concurrency studies generally use binary indictors of concurrency (yes/no) that combine all concurrent sexual partnerships into one category. Concurrent sexual partnerships are defined by UNAIDS as overlapping sexual partnerships in which sexual intercourse with one partner occurs between two acts of intercourse with another partner.20 However, different forms of concurrency should represent different degrees of risk for disease transmission. The duration of overlap and coital frequency may contribute as much or more to transmission dynamics than the mere presence of such partnerships.1,4,12,21,22 The number of concurrent partnerships a person has may also be important.22
Dichotomous measures of concurrent partnerships may therefore conflate high and low risk partnerships and obscure the statistical relationships between STDs and concurrency. It is therefore necessary to expand the concept of risk from a focus on the individual to one on partnerships to better understand sexual behavior and disease transmission.23
This article examines variation in both individual and partner concurrency among Xhosa speaking men in Cape Town, South Africa, and how different forms of partner concurrency are associated with self-reported STDs. Longitudinal data from sexual partner history tables are used to assess the type of relationships (main vs. nonmain) and condom use associated with concurrency. A unique set of self-administered questions is also used to examine the number of partners men have had concurrently and the duration of concurrent partnerships.
MATERIALS AND METHODS
The data for this article come from the Cape Area Panel Study (CAPS). The first wave of CAPS (in 2002), surveyed a representative sample of 4752 young adults aged 14 to 22 living in Cape Town. A 2-stage sample was used, stratified by the 3 main population groups (African, colored, and white). In the first stage, clusters were selected, categorized by predominant population group, and in the second stage, households were randomly selected from clusters to achieve a representative sample. Respondents were reinterviewed up to 4 more times, most recently in 2009 (wave 5), with the cohort then aged 20 to 30 years. The African male sample initially comprised 930 men, and 582 were reinterviewed in 2009. The estimation sample for this article (n = 439) consists of African men interviewed in waves 3 and 5 with complete data on all dependent and independent variables. Ethical approval was granted by the University of Cape Town and University of Michigan.
Wave 3 (2005) and 5 (2009) included a sexual partners history table. Sexual behavior questions were also included in a self-administered section in wave 5 (see Document, Supplemental Digital Content, online only, [available at: http://links.lww.com/OLQ/A37], which displays the relevant sexual behavior questions asked in CAPS).
Individual concurrency data were collected in the sexual partner history tables with the question “Did you have any other sexual partners during the time that you and [partner] were having a sexual relationship?” Partner concurrency was measured via “As far as you know, did [partner] have any other sexual partners during the time that you and he/she were having a sexual relationship?” This question involves perceptions of another person's sexual behavior and therefore must be regarded with a healthy degree of skepticism.
A wide variety of definitions of concurrency have been used in previous studies, and this has lead to different measurement rates.4 It is therefore important to clarify what is being measured here. The UNAIDS recommended method to measure the cumulative prevalence of concurrent partnerships—more than one overlapping sexual partnership at any point in the past year20—was not applied because exact start and end dates of partnerships were not collected in CAPS. The direct method of measuring concurrency used in CAPS captured concurrency at any time during a partnership. The prevalence of concurrency measured in CAPS will therefore be greater if the UNAIDS method had been used because concurrent sexual partnerships that occurred more than 12 months before the CAPS interviews can be included in the CAPS measure.
Concurrency was categorized by different types of partnerships (main vs. nonmain) and condom use (consistent vs. inconsistent). A partner was considered a “main partner” in wave 3 if the relationship lasted longer than 5 months (a cut off dictated by the response options) and involved a coital frequency greater than 10 (i.e., the couple had sex >10 times). For wave 5, data on relationship length and coital frequency were not available. Partnerships reported as “main partners” were considered as such and all others taken to be “nonmain partners.” In terms of condom use, a partnership was considered to involve consistent condom use if participants reported that condoms were always used with a specific partner.
The wave 5 self-administered module included 3 unique individual concurrency questions. The first asked “Have you ever been in a sexual relationship with someone and had sex with somebody else? This includes main partners, side-partners, roll-ons, and one night stands.” “Roll-on” is a common term used in Cape Town for a concurrent partner.24 The second question, “What is the largest number of sexual partners you have had at the same time for a week or more?” was used to separated respondents with 2 partners concurrently from those who reported 3 or more partners concurrently. Finally, the longest duration of any concurrent partnership was assessed using the question “What is the longest period of time during which you have had more than one sexual partner?” with partnerships separated into shorter or longer than 6 months.
The wave 5 (2009) self-administered module asked 2 STD-related questions: “Have you ever had problems with your penis or vagina such as pain when you pee, sores or unusual fluids?” and “Have you ever had a sexually transmitted disease (STD) that is not HIV?” The STD measure used in this article was a binary indicator created for respondents who reported ever having an STD or a history of dysuria, genital discharge, or sores.
Covariates of STDs
Measures of educational attainment (0–12 years of schooling, with any tertiary education recorded as 13) and socioeconomic status (logged monthly per capita household income from wave 1) were included as controls in the models, given their importance in predicting HIV/STD status in other studies.25,26 Circumcision status (self-reported) was used because of the negative association found between being circumcised and the risk of contracting herpes simplex virus type 2 and human papillomavirus,27 and HIV.28
Regression models also included a set of sexual behavior indicators. A binary measure reflecting the number of lifetime sexual partners (1 = 5 or more, and 0, otherwise) was created to control for multiple partners, whether serially or concurrently. A continuous measure of years sexually active was also constructed to control for increases in STD risk that accrue from additional years of sexual exposure. Third, age at first sex was included as a baseline for years sexually active and because younger initiation of sex has been associated with greater STD risk.29 Finally, a binary measure of whether respondents used contraception the first time they had sex was added as a control for attitudes/preferences toward unprotected sex and/or access to protective contraceptives.
Descriptive statistics for the study sample are presented first. Tabulations of the prevalence and variations in individual and partner concurrency are then provided. Finally, 4 probit regression models on STD history were estimated with partner concurrency as the key independent variable. The first 2 models assessed the relationship between STD history and partner concurrency using the partnership type (nonmain vs. main) variable. Two individual concurrency variables were included in the models as potential proxies for unobserved factors, such as being in a highly connected sexual network. The first model controlled for number of partners men have had concurrently, and the second for the duration of concurrent partnerships. The third and forth models controlled for the same individual concurrency variables but replaced partnership type with condom use (always vs. inconsistent) for the partner concurrency variable.
Marginal effects are presented for all models instead of probit coefficients, as marginal effects are more easily interpretable (for a continuous variable the coefficient reflects the percentage point increase in the probability of observing the dependent variable for a 1 unit change in the independent variable; for binary variables, it reflects a similar change in the dependent variable from moving from 0–1 on the independent variable of interest). All standard errors were corrected for heteroskedasticity.
Descriptive statistics for the estimation sample are presented in Table 1. Almost a third (28%) reported an STD. The average respondent was just <25 years of age, had completed 10.5 years of schooling, and was from a household with an average monthly per capita income of R404 ($39 based on 1 July 2002 exchange rate). Almost all men reported Xhosa as their preferred language and the majority (92%) was circumcised. Regarding sexual behavior, everyone in the estimation sample had reported having sex, with an average age at first sex of 15.5 years and 9.2 years sexually active, and significant variation was evident in number of sexual partners and the use of contraceptives at first sex.
The prevalence and variation in individual and partner concurrency are displayed in Table 2. The majority of men reported individual concurrency (66%) and almost half reported partner concurrency (46%) in the sexual partner history tables. Individual and partner concurrency were positively correlated: 61% of individuals reported both individual and partner concurrency (results not shown). In terms of individual concurrency, men were almost 3 times more likely (74%) to report having sex with someone else while with partners classified as main rather than nonmain partners and 44% reported consistent condom use with the concurrent partner listed in the partnership tables. In contrast, the majority of men (61%) who reported partner concurrency indicated that these were nonmain partners perceived to have additional partners rather than main partners; and 59% (i.e., 27%/46%) reported always using condoms with partners perceived to have partners.
Table 2-a. Prevalenc...Image Tools
The second panel in Table 2 presents data from the self-administered questions. Significant variation was evident in both the greatest number of partners respondents had ever had concurrently and the length of their longest concurrent partnership. Nearly a third (31%) of respondents who reported a concurrent partnership indicated having had 3 or more partners concurrently, and more men reported a concurrent partnership of longer than 6 months (52%) than shorter than 6 months (43%).
Table 2-b. Prevalenc...Image Tools
Table 3 displays probit marginal effects for models with STD history as the dependent variable. Models 3.1 and 3.2 indicate that, in terms of partner concurrency, men with main partners perceived to have other partners were 16% to 17% (P < 0.05) more likely to have reported an STD than those with no partner concurrency reported. Men with nonmain partners perceived to have other partners were also more likely to report an STD, but the associated risk was lower compared with main partners: 9 to 10%. Models 3.3 and 3.4 also indicate that variation in partner concurrency mattered for STD risk with partner concurrency involving inconsistent condom use associated with a greater probability (18%, P < 0.001) of reporting an STD than partnerships in which condoms were always used (8%–10%, P < 0.1).
In terms of the individual concurrency measures, models 3.1 and 3.3 indicate that reporting 3 or more sexual partners concurrently was associated with a significantly greater probability of reporting an STD than having had 2 partners concurrently: 34% versus 12%. In models 3.2 and 3.4, respondents who reported concurrent partnerships of longer than 6 months were twice as likely to report an STD than respondents in shorter concurrent partnerships, when compared with their counterfactuals.
Substantial differences between concurrent sexual partnerships were observed, and these variations were associated with different disease risk in the models predicting STD history, suggesting that concurrent partnership dynamics must be taken into account in studies assessing the role of concurrency in STD transmission.
The majority of young adult African men in this study reported having had overlapping sexual partnerships. These partnerships often involved a main partner and inconsistent condom use. Measures of ever engaging in individual concurrency created from self-administered questions assessed 2 aspects of concurrency rarely measured: (1) the largest number of partners someone has concurrently had and (2) the longest duration of concurrency. A significant proportion of men (18%) reported having had 3 or more partners concurrently. Concurrency in this sample comprised longer term partnerships that lasted 6 months or more and partnerships of shorter than 6 months, with slightly more men reporting “long-term” concurrency.
Men who reported having had more partners concurrently and those who reported longer concurrent sexual partnerships were significantly more likely to have reported an STD than their counterparts. As individual concurrency increases the risk of transmitting an STD, not of acquiring it,13,14 these variables would have to have acted as a proxy for other factors. Individual concurrency could simply have acted as a proxy for number of sexual partners. However, as an indicator for having many partners was included in the analysis and there was not a perfect or near perfect match between this variable and individual concurrency (64% of individuals who reported having a concurrent partnership also reported having five or more lifetime sexual partners) this explanation is unsatisfactory. Another possibility is that they acted as a proxy for being in a highly connected sexual network—a network in which high rates of concurrency or very rapid rates of partner acquisition establish the connectivity needed to sustain STD transmission. Further research is required to explore the mechanism(s) underlying these findings.
Individuals who perceived any of their sexual partners to have a concurrent partner were associated with greater STD risk. However, the degree of this risk was not similar for all partner concurrency: main partners who had other partners represented greater risk than other partners; and partner concurrency in relationships that involved inconsistent condom use represented approximately twice the risk compared with those when condoms were always used. Almost a fifth of the study sample reported partner concurrency with these higher risk characteristics, indicating their relevance for a substantial number of men. As partnership type and condom use are associated—condom use is less consistent with main partners23—STD risk is thus greatest with long-term partners who have other partners, presumably because sex with these partners is relatively frequent and unprotected.
There are several limitations to this study, many of which motivate further research. First, STD history was self-reported. Second, the precise timing of STDs and concurrent partnerships is unknown. Causal inference is thus complicated by the fact that some men may have had an STD before engaging in concurrency (leading to an upward bias) or may not have engaged in concurrency because they had an STD (leading to a downward bias). Third, as mentioned previously, perceptions of another person's sexual behavior almost certainly involve measurement error, and measures of partner concurrency must therefore be viewed with a healthy degree of skepticism. Fourth, and potentially most important, no information was available on the characteristics of partnerships making up partner concurrency. Risk differentials via number of partners' partners, length of partnerships, coital frequency, and condom use are unknown. Finally, our results are for a specific population group and it is unclear whether they generalize to other populations.
The variations in concurrency highlighted in this article and that these variations were found to matter for STD risk, despite the study limitations, indicates that future research on concurrency should include more nuanced analysis using measures that incorporate such variation. The debate about the relationship between concurrency and HIV, in particular, is driven by inconsistent empirical evidence and it is clear that more and better empirical evidence is required. Morris (2010) suggests that an improved research design to assess this question would involve a prospective longitudinal study of incident HIV-infection among couples, where both are HIV-negative on enrollment and both are enrolled in the study.14 The findings in this article indicate that such studies, and other attempts to assess this question, should include measures of concurrency that capture variation in partnership type, condom use, number of partners involved in concurrent partnerships and duration of concurrent partnerships.
1. Mah TL, Halperin DT. Concurrent sexual partnerships and the HIV epidemics in Africa: Evidence to move forward. AIDS Behav 2010; 14:11–16.
2. Shelton JD. Why multiple sexual partners? Lancet 2009; 374:367–369.
3. Halperin D, Epstein H. Why is HIV prevalence so severe in southern Africa? The role of multiple concurrent partnerships and lack of male circumcision: Implications for AIDS prevention. South Afr J HIV Med 2007; 8:19–25.
4. Lurie MN, Rosenthal S. Concurrent partnerships as a driver of the HIV epidemic in Sub-Saharan Africa? The evidence is limited. AIDS Behav 2010; 14:17–24.
5. Sawers L, Stillwaggon E. Concurrent sexual partnerships do not explain the HIV epidemics in Africa: A systematic review of the evidence. J Int AIDS Soc 2010; 13:1–23.
6. Doherty IA, Shiboski S, Ellen JM, et al.. Sexual bridging socially and over time: A simulation model exploring the relative effects of mixing and concurrency on viral sexually transmitted infection transmission. Sex Transm Dis 2006; 33:368–373.
7. Morris M, Kretzschmar M. A microsimulation study of the effect of concurrent partnerships on the spread of HIV in Uganda. Math Popul Stud 2000; 8:109–133.
8. Rothenberg M, Gurvey J, Nancy A, et al.. Concurrent sex partners and risk for sexually transmitted diseases among adolescents. Sex Transm Dis 1999; 26:208–212.
9. Jewkes R, Dunkle K, Nduna M, et al.. Intimate partner violence, relationship power inequity, and incidence of HIV infection in young women in South Africa: A cohort study. Lancet 2010; 376:41–48.
10. Kelley S, Borawski E, Flocke S, et al.. The role of sequential and concurrent sexual relationships in the risk of sexually transmitted diseases among adolescents. Adolesc Health 2003; 32:296–305.
11. Jewkes R, Dunkle K, Nduna M, et al.. Factors associated with HIV sero-positivity in young, rural South African men. Int J Epidemiol 2006; 35:1455–1460.
12. Mattson C, Bailey R, Agot K, et al.. A nested case-control study of sexual practices and risk factors for prevalent HIV-1 infection among young men in Kisumu, Kenya. Sex Transm Dis 2007; 34:731–736.
13. Morris M. Concurrent partnerships and syphilis persistence: New thoughts on an old puzzle. Sex Transm Dis 2001; 28:504–507.
14. Morris M. Barking up the wrong evidence tree. Comment on Lurie & Rosenthal, “Concurrent Partnerships as a Driver of the HIV Epidemic in Sub-Saharan Africa? The Evidence is Limited.” AIDS Behav 2010; 14:31–33.
15. Koumans E, Farley T, Gibson J, et al.. Characteristics of persons with syphilis in areas of persisting syphilis in the United States. Sustained transmission associated with concurrent partnerships. Sex Transm Dis 2001; 28:497–503.
16. Zhan W, Krasnoselskikh T, Niccolai L, et al.. Concurrent sexual partnerships and sexually transmitted diseases in Russia. Sex Transm Dis 2011; 38:543–547.
17. Landman K, Ostermann J, Crump J, et al.. Gender Differences in the risk of HIV infection among persons reporting abstinence, monogamy, and multiple sexual partners in Northern Tanzania. PLoS One 2008; 3:1–6.
18. Tanser F, Bärnighausen T, Hund L, et al.. Effect of concurrent sexual partnerships on rate of new HIV infections in a high-prevalence, rural South African population: A cohort study. Lancet 2011; 378:247–255.
19. Jewkes R, Dunkle K. Intimate-partner violence and HIV in South African women—Authors' reply. Lancet 2010; 376:1219–1220.
20. UNAIDS. HIV: consensus indicators are needed for concurrency. Lancet 2010; 375:621–622.
21. Morris M, Epstein H, Wawer M. Timing is everything: International variations in historical sexual partnership concurrency and HIV prevalence. PLoS One 2010; 5:e14092.
22. Aral S. Partner concurrency and the STD/HIV epidemic. Curr Infect Dis Rep 2010; 12:134–139.
23. Gorbach P, Holmes K. Transmission of STIs/HIV at the partnership level: Beyond individual-level analyses. J Urban Health 2003; 80:15–25.
24. Mah T, Maughan-Brown B. Social and cultural contexts of concurrency in Khayelitsha. Centre for Social Science Research Working Paper 251, University of Cape Town. 2009:1–19.
25. Bärnighausen T, Hosegood V, Timaeus I, et al.. The socioeconomic determinants of HIV incidence: Evidence from a longitudinal, population-based study in rural South Africa. AIDS 2007; 21(suppl 7):S29–S38.
26. Krieger N, Waterman P, Chen JT, et al.. Monitoring socioeconomic inequalities in sexually transmitted infections, tuberculosis, and violence: Geocoding and choice of area-based socioeconomic measures. The Public Health Disparities Geocoding Project (US). Public Health Rep 2003; 118:240–258.
27. Tobian AA, Serwadda D, Quinn TC, et al.. Male circumcision for the prevention of HSV-2 and HPV infections and syphilis. N Engl J Med 2009; 360:1298–1309.
28. Siegfried N, Muller M, Deeks JJ, et al.. Male Circumcision for Prevention of Heterosexual Acquisition of HIV in Men. Cochrane Database Syst Rev. 2009:CD003362. DOI: 003310.001002/14651858.CD14003362.pub14651852.
29. Kaestle C, Halpern C, Miller WC, et al.. Young age at first sexual intercourse and sexually transmitted infections in adolescents and young adults. Am J Epidemio 2005; 161:774–780.