Kasamba, Ivan MSc*; Sully, Elizabeth BA†; Weiss, Helen A PhD‡; Baisley, Kathy MSc‡; Maher, Dermot DM*‡
An estimated 33 million people worldwide are HIV infected, of whom two-thirds live in sub-Saharan Africa where most new infections occur from unprotected heterosexual intercourse.1 HIV prevalence is greater among married individuals than never-married counterparts.2,3 There is heterogeneity in HIV risk among married individuals, determined in part by the age differences between partners,3 and each individual's duration between sexual debut and marriage.4 HIV within marriage may result from infection by one's spousal partner or due to nonspousal exposure. Introduction of HIV into a partnership from outside accounted for 1/3 of transmissions to the HIV-negative partner among initially serodiscordant couples in a recent study in Eastern and Southern Africa.5 This suggests that extraspousal partnerships are an important pathway in introducing HIV infection into a marriage if both spouses are HIV negative or if they are serodiscordant.
Extraspousal partnerships are a form of sexual concurrency, defined as overlap between 2 or more sexual partners, with intercourse with 1 partner occurring between 2 acts of intercourse with another partner.6 This definition encompasses many partner types and duration of partnerships.6,7 Measurement of concurrent partnerships should make a distinction, where possible, between a single long-term partnership with intermittent short-term sexual partners, and 2 or more sustained, overlapping sexual partnerships. Concurrency is hypothesized to contribute to the spread of HIV through increasing the probability of transmitting HIV to one's partners. Having 2 sequential versus 2 concurrent partners caries the same risk for the index individual: the added risk of concurrency is through the increased probability that the index person transmits the virus to 1 or both their partners.8-10 High viral load and infectiousness during recent HIV infection contributes to increased risk of HIV transmission to any partner.11,12 There is debate as to the importance of concurrency as a driving factor in the HIV epidemic in sub-Saharan Africa.7-9,13-16 Although mathematical modelling suggests an important role,11,17 the assumptions of the models are contested in the literature,18 and there remains limited empirical data.
A population-based cohort established in rural southwest Uganda in 1989 to investigate HIV epidemiology19 provided the opportunity to conduct an empirical examination of the association between concurrency and HIV transmission.
The study cohort comprises approximately 20,000 residents of 25 neighbouring villages in southwestern Uganda. There are no tarmac roads and access may be difficult during the rainy season. The main income-earning activities are growing bananas, coffee, and beans and trading fish.20 The community is mostly from the Baganda tribe (73%), with 15% being of Rwandese origin. Religious affiliation is predominantly Christian, with a significant Muslim minority (28%). HIV prevalence was 6.2% in 2000 and 7.7% in 2005.21
Annual General Population Cohort Survey
Details of the study design have been published previously.21-23 In brief, household surveys of sociodemographic and behavioral characteristics and HIV serostatus of consenting participants aged ≥13 years have been conducted annually since 1989, with all residents eligible for inclusion. The average annual participation rate is about 60%-65%, although an estimated 84% have ever participated. All households are visited and consenting residents are interviewed at home in the local language by trained survey staff and provide a blood sample for HIV testing.
Collection of Data on Extraspousal Partnerships
Because marriage in Baganda society can take many forms, including customary, religious, civil, and informal,24 participants were asked if they had ever been married and if they considered themselves to be currently married or to have someone they considered a spouse. Data on sexual behavior included the characteristics of current and most recent sexual partners, including type of partnership and condom use. Assuming that married couples are engaging in continued coitus, we define concurrent extraspousal partnerships as the occurrence of sexual intercourse with a nonspousal partner within the last 12 months. Our measure assumes that extraspousal partnerships overlapped with sexual intercourse with a spousal partner, although this may not have been the case. Concurrent partnerships could not be directly determined because the questions on partnership included the time of occurrence of last sex with an extraspousal partner but not the duration of the partnership. Partners identified as extraspousal were classified as “ex-spouse”, “regular”, or “casual”. For polygynous men, extraspousal partnerships were defined as sexual partners who were not considered to be wives.
Survey participants eligible for this study were aged 15 years or older, who reported a main (spousal) partner with whom sex had occurred in the past year, and for whom we confirmed the spouse's unique identification number. The unique identification number of each survey participant allowed us to determine the HIV-status concordance of the marital union. Spousal identification numbers were also used to link up characteristics of women whose husbands engage in extraspousal partnerships.
We tabulated the prevalence of extraspousal partnerships, distribution of participant characteristics, and responses to questions on HIV testing by sex. We calculated median age at first sex (AFS) and age at first marriage (AFM) for those who had ever been married. AFS and AFM were left censored at 10 years of age. Logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for the association of demographic and behavioral characteristics with the prevalence of extraspousal partnerships in 2008, the survey with the most detailed sexual behavior data collected. Variables were then adjusted by other variables associated with extraspousal partnerships at P < 0.15 in univariate analyses. The reported prevalence of extraspousal partners was too low among women to permit multivariable analysis. Among men who were not in polygynous marriage, we investigated the association between the prevalence of extraspousal partnerships and the characteristics of their wives. Polygynous unions were excluded from this analysis due to difficulties in linking up all spousal identification numbers for each wife. The examination of linked partner data allows us to assess whether an women whose partner is engaged in an extraspousal partnership is at an increased risk of HIV infection, as the concurrency hypothesis would suggest.
Table 1 shows characteristics of the adult population in 2008. Approximately, 45% of men and 51% of women reported being married, and further analyses focus on these 879 married males and 1380 married females. Data from 2002, 2004, and 2006 survey rounds indicate similar demographic and behavioral characteristics of married participants.
Married men and women had similar median AFS (18 years for men; 17 for women), but median AFM was 22 years for men and 18 years for women. Polygynous marriages were reported by 15% of men and 22% of women. Data from 2006 (not presented) found a median of 5 lifetime partners for men and 2 for women. HIV prevalence among married respondents was 7.0% for men and 8.0% for women, compared with an overall adult prevalence of 5.3% for men and 8.2% for women in 2008.
Overall, about 17% of men and 2% of women reported having had an extraspousal partnership within the past year, which were most often characterized as “regular” rather than “casual” partnerships.
Extraspousal partnerships were more common among younger and polygynous men and those who first had sex below age 18; these associations persisted after adjusting for age, condom use within marriage and knowledge of spouse's HIV status (Table 2). Men reporting extraspousal partnerships were also more likely to report ever having used a condom with a spouse [adjusted odds ratio (aOR) = 1.85; 95% CI: 1.24 to 2.75]. There was no evidence that men reporting extraspousal relationships were associated with increased risk of HIV (aOR = 0.98; 95% CI: 0.48 to 2.01), although they were more likely to have unknown HIV status (ie, not participated in the serosurvey) (aOR = 1.76; 95% CI: 0.97 to 3.18) and to not know their partner's HIV status (aOR = 1.63; 95% CI: 1.10 to 2.42).
Among women, reporting not knowing their partner's HIV-status was the only variable found to be strongly associated with reporting an extraspousal partnership (OR = 2.97; 95% CI: 1.04 to 8.54). However, the power to detect associations between women's characteristics and behaviors with reporting an extraspousal partnership was limited by the low numbers of women reporting extraspousal partners.
Men not in polygynous unions who participated in the 2008 survey were linked with their wives using spousal identification numbers; 734 (98%) wives also participated in 2008, and 645 (87%) had known HIV status. Table 3 shows the relationship between women's characteristics and their husbands reporting an extraspousal partnership. There was no evidence of an association between a women's HIV status and her husband reporting an extraspousal partnership, with 14% of HIV-positive and 17% of HIV-negative women having husbands who reported extraspousal partnerships (OR = 0.82; 95% CI: 0.36 to 1.87). Women who reported not knowing their husband's HIV status were more likely to have husbands reporting an extraspousal partnership (aOR = 1.76; 95% CI: 1.13 to 2.75).
In a 2008 survey, extraspousal partnerships within the past year were reported by 17% of married men, in contrast to 2% of married women. The majority of extraspousal partnerships were with a “regular” partner, indicating that they are likely to be concurrent with marital relationships. Direct comparisons of concurrency prevalence are difficult due to discrepancies in measurements of concurrency, data sources, and our sample being limited to only married individuals. However, our findings of prevalence of extraspousal partnerships among men is similar to concurrency rates found in Demographic Health Survey's in sub-Saharan Africa, but below prevalence estimates from Asia and the United States.18 Women's prevalence, on the other hand, is slightly higher than other concurrency estimates from sub-Saharan Africa, but smaller than concurrency estimates from the United States and Europe.18
Married people have reported higher rates of concurrency than unmarried people in Zambia and Tanzania.5,25,26 A study from the same population-cohort data in this article found that when both partners are HIV negative at the time of marriage, men were twice as likely as women to be the partner bringing HIV into the marriage.22 Our results indicate, however, that there is no evidence of an association between men's HIV-status and reporting an extraspousal partnership.
In this study, polygynous unions were associated with increased odds of reporting an extraspousal partnership. There are limitations in determining extraspousal (rather than concurrent) partnerships for polygynous men using cross-sectional data, as it is unknown whether the partners later become spouses. Making a distinction between spousal and nonspousal partners may therefore be an artefact based on the time of observation, suggesting that simple classification schemes of concurrency may sometimes be too crude to address the reality of concurrency and nonspousal partners in polygynous unions.
Previous studies examining concurrency have had limited ability to look at partner data. Research in this area has only been able to look at characteristics of those who assume their partners are having concurrent partnerships.8,27,28 This study adds to the literature by using spousal identification numbers to link characteristics of women to those whose husbands report extraspousal partnerships. Despite the hypothesis that concurrency would lead to a greater likelihood of transmission to sexual partners, we do not find evidence that women whose husbands report having extraspousal partnerships are at greater risk of HIV infection.
Individual-level protective behaviors may be counterbalancing the potential increased risk associated with concurrency, leading to the observed lack of association between women's HIV-status and their husbands reporting extraspousal partnerships.10 For example, reported condom use with spousal partners was greater among men who had engaged in an extraspousal partnership. It is possible that individuals perceive their extraspousal partnerships as risky, and therefore use condoms to protect themselves or their spouse. There is some evidence of consistent condom use with extraspousal partners, with 41% of men and 50% of women reporting condom use at last sex with an extraspousal partner. The observed differential condom use with spouses and extraspousal partners could have a protective effect that is offsetting the potential risk associated with concurrency. Data from a nationally representative cross-sectional household survey in Uganda found that using condoms with a partner outside of marriage is protective, with a similar odds of HIV as among those who don't have any extraspousal partners (adjusted OR: 1.0; 95% CI: 0.3 to 2.7).29
Knowing one's partner's HIV status was associated with a lower odds of engaging in extraspousal partnerships. Relationships where partners communicate about HIV status are also partnerships where extraspousal partners are less likely. This finding supports previous research on HIV prevention among married couples and the importance of developing open communication between partners and increasing uptake of couples counseling and testing.30
Our study had several strengths. First, we provide estimates of the prevalence of extraspousal partnerships from a large population cohort in a rural community in sub-Saharan Africa. For participants with a spouse resident in the study area, we were able to obtain linked information on their spouse's HIV-status using a unique spousal identification number. Few studies have been able to explore index case concurrency and their partner's HIV status because of the difficulty of enrolling both partners in a sexual relationship.
There are limits to fully assessing concurrency among married people in this study. The definition we used of extraspousal partners assumes an overlap between marital and other partners if one is currently married and had an extraspousal partner in the last 12 months. This does not take into account that a spouse may have been away from their marital partner when they were with their other partner. As well, respondents who were not married for the full duration of the year before survey could overestimate concurrency. Although a broader definition of concurrency among polygynous men would include overlapping sex with more than one wife, we limited our definition to partnerships occurring with a nonspousal partner, which may underestimate overall concurrency levels. Furthermore, we did not assess coital frequency, or the duration of extraspousal partnerships. In addition, the classification of extraspousal partnerships into “regular” and “casual” may not fully capture the distinction respondents make between these 2 partnership types. The findings presented are also constrained by lack of information on multiple extraspousal partners; it is assumed that reports of behavior with extraspousal partners refer to either the same partner across questions or that behavior is consistent across different partners.
An additional limitation to our study is that the average annual survey participation is 60%-65%. Maintaining a high participation rate is a challenge in long established community-based annual HIV serosurveys.20 Some participants may decline participation in the serosurvey because they have previously had a blood test for HIV, know their result, and decide against a repeat test. Finally, all of our findings are based on cross-sectional data, which constrains any analysis between time of extraspousal partnership and time of HIV infection. Further analysis should assess previous extraspousal relationships as a determinant of current HIV incidence.
Among married men, we found no evidence of an association of reported extraspousal partnerships and HIV status. Furthermore, using linked partner data, we found no evidence of an association between men's extraspousal partnerships and the HIV status of their wives. We did find that men's extraspousal partnerships were associated with early age of sexual debut, ever using a condom within marriage, and not knowing their partner's HIV status. Although this study was able to use unique linked partner data, our measurements of concurrency among married individuals were not without limitation. Future studies should attempt more precise measurements of concurrency that include exact time overlap and coital frequency to expand upon our findings. Future research should be directed towards understanding how individual behaviors could be offsetting potential risks associated with concurrency.
We would like to thank the MRC/UVRI General Population Cohort study survey team, data and statistics team, and participants. This study was supported by the Medical Research Council (MRC) of the UK.
1. UNAIDS. UNAIDS Report on the Global AIDS Epidemic. Geneva, Switzerland: UNAIDS; 2010.
2. Lowndes C, Alary M, Belleau M, et al. West Africa HIV/AIDS Epidemiology and Response Synthesis: Implications for Prevention. Washington, DC:World Bank; 2008.
3. Clark S. Early Marriage and HIV Risks in Sub-Saharan Africa. Stud Fam Plann. 2004;35:149-160.
4. Bongaarts J. Late marriage and the HIV-epidemic in sub-Saharan Africa. Popul Stud. 2007;61:73-83.
5. Celum C, Wald A, Lingappa JR, et al. Acyclovir and transmission of HIV-1 from persons infected with HIV-1 and HSV-2. N Engl J Med. 2010;362:427-439.
6. UNAIDS Reference Group on Estimates Modelling and Projections: Working Group on Measuring Concurrent Sexual Partnerships. HIV: consensus indicators are needed for concurrency. Lancet. 2010;375:621-622.
7. Lurie MN, Rosenthal S. Concurrent partnerships as a driver of the HIV epidemic in sub-Saharan Africa? The evidence is limited. AIDS Behav. 2009;14:17-24.
8. Mah TL, Halperin DT. The evidence for the role of concurrent partnerships in Africa's HIV epidemics: a response to Lurie and Rosenthal. AIDS Behav. 2009;14:25-28.
9. 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. 2009;14:31-33.
10. Reniers G, Tfaily R. An inquiry into the mechanisms linking polygyny, partnership concurrency and HIV transmission in sub-Saharan Africa. Demography. 2011; In press.
11. Morris M, Kretzschmar M. Concurrent partnerships and the spread of HIV. AIDS. 1997;11:641-648.
12. Green EC, Mah TL, Ruark A, et al. A framework of sexual partnerships: risks and implications for HIV prevention in Africa. Stud Fam Plann. 2009;40:63-70.
13. Halperin DT, Epstein H. Concurrent sexual partnerships help to explain Africa's high HIV prevalence: implications for prevention. Lancet. 2004;364:4-6.
14. 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.
15. Mah TL, Halperin DT. Concurrent sexual partnerships and the HIV epidemics in Africa: evidence to move forward. AIDS Behav. 2010;14:11-16.
16. Lurie MN, Rosenthal S. The concurrency hypothesis in sub-Saharan Africa: convincing empirical evidence is still lacking. Response to Mah and Halperin, Epstein, and Morris. AIDS Behav. 2009;14:34-37.
17. Kretzschmar M, Morris M. Measures of concurrency in networks and the spread of infectious disease. Math Biosci. 1996;133:165-195.
18. 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;13:34-34.
19. Mbulaiteye SM, Mahe C, Whitworth JAG, et al. Declining HIV-1 incidence and associated prevalence over 10 years in a rural population in south-west Uganda: a cohort study. Lancet. 2002;360:41-46.
20. Nakibinge S, Maher D, Katende J, et al. Community engagement in health research: two decades of experience from a research project on HIV in rural Uganda. Trop Med Int Health. 2009;14:190-195.
21. Shafer LA, Biraro S, Nakiyingi-Miiro J, et al. HIV prevalence and incidence are no longer falling in southwest Uganda: evidence from a rural population cohort 1989-2005. AIDS. 2008;22:1641-1649.
22. Carpenter LM, Kamali A, Ruberantwari A, et al. Rates of HIV-1 transmission within marriage in rural Uganda in relation to the HIV sero-status of the partners. AIDS. 1999;13:1083-1089.
23. Nunn AJ, Mulder DW, Kamali A, et al. Mortality associated with HIV-1 infection over five years in a rural Ugandan population: cohort study. BMJ. 1997;315:767-771.
24. Nabaitu J, Bachengana C, Seeley J. Marital instability in a rural population in South-West Uganda: implications for the spread of HIV-1 infection. Africa (Lond). 1994;64:243-251.
25. Sandøy IF, Dzekedzeke K, Fylkesnes K. Prevalence and correlates of concurrent sexual partnerships in Zambia. AIDS Behav. 2008;14:59-71.
26. Boerma JT, Urassa M, Nnko S, et al. Sociodemographic context of the AIDS epidemic in a rural area in Tanzania with a focus on people's mobility and marriage. Sex Transm Infect. 2002;78(suppl 1):i97-i105.
27. Guwatudde D, Wabwire-Mangen F, Eller LA, et al. Relatively low HIV infection rates in rural Uganda, but with high potential for a rise: a cohort study in Kayunga District, Uganda. PLoS One. 2009;4:e4145.
28. 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;359:1896-1903.
29. Mermin J, Musinguzi J, Opio A, et al. Risk factors for recent HIV infection in Uganda. JAMA. 2008;300:540-549.
30. Zulu EM, Chepngeno G. Spousal communication about the risk of contracting HIV/AIDS in rural Malawi. Demographic Res. 2003;1:247-278.
© 2011 Lippincott Williams & Wilkins, Inc.