Abstract
Objectives
The study aimed to test if there was an association between bacterial vaginosis and women reporting that their partner had other partners at the same time (partner concurrency). This association has not been assessed in a longitudinal cohort.
Methods
The Longitudinal Study of Vaginal Flora recruited a cohort of 3620 non pregnant women aged 15–44 years who presented for routine primary health care at 12 clinics in Birmingham, Alabama. Behavioural questionnaires and vaginal smears were obtained quarterly for a year and BV was defined by a Nugent score 7 or higher as well as Amsel criteria. Mixed effects logistic regression was used to assess the relationship between prevalent BV and reporting that one’s partner had had other partners in the preceding 3 to 6 month time interval.
Results
Nugent-score BV was associated with both reporting that one’s partner definitely (adjusted Odds ratio [aOR] 1.5; 95% confidence interval [CI], 1.2–1.8) and possibly (aOR 1.5; 95% CI, 1.2–1.8) engaged in partner concurrency in the preceding 3 to 6 month time period. BV diagnosed by Amsel criteria were similar.
Conclusion
A diagnosis of BV was associated with reporting that one’s partner possibly or definitely engaged in partner concurrency.
Keywords: Partner concurrency, bacterial vaginosis, vaginal microbiota, sexual network, Nugent score
Background
Despite bacterial vaginosis (BV) being the most common vaginal infection in women of reproductive age, both its pathogenesis and the reasons for the large differences in BV prevalence between different populations are incompletely understood 1. Sexual activity appears to play an important role. BV has, for example, been associated with a number of individual level sexual risk factors such as number of partners, type of sex and inconsistent condom use 1,2. In keeping with a number of STIs, individual level risk factors have not been able to explain the large differences in BV prevalence between ethnic/racial groups. In the USA, for example, being black remained a residual risk for BV after controlling for a range of individual and partner level variables 1,2. A number of authors have speculated that racial differences in the composition of the vaginal microbiota may explain differences in BV prevalence 3. Others have however hypothesized that sexual network factors may be responsible 2,4. Studies with other STIs such as chlamydia, HIV, HSV-2 and syphilis have found that differences in network structure, including differences in partner concurrency, can explain a large part of the ethnic differences in STI prevalence 5.
Sexual partner concurrency can be defined as one’s partner having sex with someone else during one’s sexual relationship with them 5. It is thought to enhance STI transmission via a number of mechanisms that include enhancing network-connectivity and removing the protective effect of partner-sequencing 5.
Evidence has been found of a positive association between the prevalence of male concurrency and BV at inter country and inter ethnic group comparisons but these analyses involved small sample sizes 4. This association has also been found at an individual level. A study from rural Peru found an association between prevalent BV and reporting that one’s partner had another partner in the past year 6. However, this study was cross sectional and controlled for a very limited number of potential confounders (and no other sexual behavioural variables). In this study, we assess, for the first time, the association between partner concurrency and prevalent BV in a longitudinal cohort. In a supplementary analysis, we also looked at incident BV.
Methods
This study involves a secondary data analysis of the Longitudinal Study of Vaginal Flora. The study recruited 3620 15 to 44 year old women between 1999 and 2003 when they presented for a routine health check up at any one of 12 clinics in the Birmingham, Alabama area. A total of 3,228 women contributed 10,067 visits to the multivariable analysis. Women were excluded if they were pregnant, had significant gynaecological or medical conditions, or were planning to leave the area in the upcoming 12 months. All participants provided informed consent and the study was approved by local Institutional Review Boards (FLORA, 1999–2002) 7.
Participants were seen at an initial visit and then every three months over a one-year timeframe for a total of 5 visits. At each visit they followed a structured interview performed in private by a female researcher. They were asked a range of questions including those pertaining to their sexual behavior and that of their partners since their last visit. For the baseline and follow up visits, these questions typically referred to the 6 and 3 month time period preceding the interviews, respectively. It was made explicit that women could decline/refuse answering any questions they so chose.
Partner concurrency (PC)
Women were asked ‘Do you think that any of your sex partners had sex with anyone other than you in the past 6 months (baseline visit)/30 days (visits 2–5)?’ They were asked to choose between the following options which were read out to them: ‘Yes, definitely’, ’Yes, possibly’, ’Do not think so’, ’Definitely not’. If they were unable to answer the interviewer could classify their answer to this question ‘Unknown’ or ‘Refused to answer.’ Only on one occasion did a person refuse to answer and we recoded this event as missing.
BV
At each study visit, each woman had a vaginal swab collected at the time of pelvic examinations. This was used to diagnose BV via the Nugent method. BV was defined as a score of 7 or more 8. Amsel criteria were also used to diagnose BV and analyses using Amsel criteria are presented in the supplementary file.
Race/ethnicity
Previous analyses of this cohort have found that women’s self-reported race was associated with incident BV 9. Only 51 women in the cohort were not white or black. To simplify the analyses we limited the analyses to woman who reported themselves to be black or white.
Mixed effects logistic regression was used to assess the odds ratio for a diagnosis of prevalent BV by self-reported partner concurrency in the preceding time period. Each woman could be diagnosed with prevalent BV at each study visit and thus up to 5 times.
Secondary analyses
A number of secondary analyses were conducted (see Supplementary file Box 1).
Potential confounders
The pathways for BV transmission are moderated by a number of factors, including the timing, frequency and type of sex, the use of douching, smoking, race, education attained, hormonal contraception and condom use. These may also be related to reported PC and were thus controlled for in each of the models. These confounders were selected based on a literature review and biological plausibility.
Results
The 3620 participants in the study were young (mean age 25.3, Standard Deviation 6.95) and predominantly African American (79.9%). Participant characteristics are displayed by reported partner concurrency status at baseline visit in the Supplementary file Table 1. Respondents reporting partner concurrency were more likely to be black, smoke, report douching and always using condoms. They reported a higher mean number of sex partners, and were more likely to report rectal and oral sex. A lower proportion reported vaginal sex four or more times per week or using hormonal contraception. Prevalent BV was more common in those reporting definite/possible partner concurrency (PC) (46.4%/55.4%) than those reporting no PC (34.9% - P < 0.001 for both comparisons).
Association between BV and partner concurrency
A diagnosis of prevalent BV was associated with both reporting that one’s partner definitely (adjusted Odds ratio [aOR] 1.5; 95% Confidence Interval [CI], 1.2–1.8) and possibly (aOR 1.5; 95% CI, 1.2–1.8) engaged in partner concurrency in the preceding time period (Table 1). In various secondary analyses the diagnosis of BV remained associated with PC (Supplementary files Box 1 & Table 2).
Table 1.
Multivariable model of the association between prevalent bacterial vaginosis and reported partner concurrency (adjusted Odds Ratios [95% Confidence Intervals] P-values*).
| Prevalent BV | |
|---|---|
| Partner concurrency | |
| None | Ref |
| Yes | 1.5 (1.2–1.8) ** |
| Possibly | 1.5 (1.2–1.8) ** |
| Don’t think so | 1.1 (1.0–1.3) |
| Unknown | 1.1 (0.7–1.1) |
| Age | |
| 15–25 | Ref |
| 26–35 | 1.0 (0.9–1.2) |
| 36–46 | 0.9 (0.7–1.1) |
| No. sex partners | 1.0 (0.9–1.1) |
| Race | |
| White | Ref |
| Black | 4.5 (3.7–5.5) ** |
| Douche | |
| No | Ref |
| Yes | 1.4 (1.3–1.6) ** |
| Rectal sex | |
| No | Ref |
| Yes | 0.7 (0.6–1.0) |
| Oral Sex (receptive) | |
| No | Ref |
| Yes | 1.0 (0.9–1.1) |
| Frequency of vaginal sex/week | |
| 0–1 | Ref |
| 2–3 | 1.2 (1.1–1.3) ** |
| ≥4 | 1.4 (1.2–1.7) ** |
| Hormonal contraception | |
| No | Ref |
| Yes | 0.6 (0.5–0.7) ** |
| Condom use | |
| Never | Ref |
| Seldom | 1.0 (0.9–1.1) |
| Always | 0.9 (0.8–1.0) |
| Education completed | |
| <12 years | Ref |
| 12 years | 0.8 (0.7–1.0) * |
| >12years | 0.6 (0.5–0.7) ** |
| Smoked | |
| No | Ref |
| Yes | 1.4 (1.2–1.6) ** |
P<0.05,
P<0.005
Discussion
We found that women who reported possible or definite PC were more likely to be diagnosed with prevalent BV. The strengths of this analysis include the fact that it includes a large cohort of woman with detailed information about their behavior and that of their partners followed prospectively for one year. Previous analyses of this cohort have established the association between BV and vaginal douching 7, partners’ race 9 and incident STIs 2. No other analysis of this or any other cohort that we are aware of has investigated the relation between BV and PC in a prospective cohort. We used a number of models to investigate the relationship between BV and PC and they all provided similar results. There are however a number of limitations to the study. We do not know how accurate our respondents’ assessments of PC was. Our definition of PC may also be subject to two non-differential misclassification biases. It may have misclassified serial monogamy as PC and if a woman reported multiple partners we could not ascertain which of these partners had engaged in PC.
Our analyses controlled for a wide range of factors but we cannot exclude the possibility that there may be some other attributes of women who report PC or an attribute of their partners, rather than the PC itself, that is responsible for the association between BV and PC. For example, men who engage in PC may have had more partners in the preceding year than other men and it may be this higher number of partners, rather than the PC, which is responsible for the association between PC and BV.
BV is not only very prevalent but it also is associated with enhanced susceptibility to a number of STIs 2. If PC is found to play a role in generating higher BV prevalences then this would have important implications for the modeling and control of STIs. In populations with high PC rates this may increase the prevalence of BV and other STIs directly. PC may also indirectly increase the prevalence of other STIs indirectly via its susceptibility-enhancing effect through BV. Furthermore, the main way that PC enhances STI transmission is via its effect on enhancing network connectivity. This effect operates at a network level and is difficult to accurately capture at the individual or partner level 5. The accurate delineation of these effects will require prospective studies that include accurate descriptions of sexual networks and regular sampling/characterization of the rectal/oropharyngeal/vaginal/penile microbiotas of women and their partners. This would enable us to assess changes in the vaginal microbiota in relation to that of changes in the partners’ microbiotas and possible extra dyadic sex. Ultimately, the link between PC and BV may provide additional motivation to interventions which have shown to be effective at reducing PC, such as Uganda’s ‘Zero Grazing’ campaign and a Social-Cognitive-Therapy-based health promotion intervention 10.
Supplementary Material
Key messages.
Previous studies have found an ecological association between partner concurrency and bacterial vaginosis.
In a longitudinal cohort, we found that reporting partner concurrency was an independent risk factor for the acquisition of bacterial vaginosis.
Interventions to reduce partner concurrency should be investigated as a means to reduce the prevalence of BV.
Acknowledgments
We would like to thank the participants and original study team of the Longitudinal Study of Vaginal Flora for allowing us to conduct this secondary data analysis.
Contributor Information
Chris R Kenyon, Email: ckenyon@itg.be, Professor of Sexually Transmitted Infections, HIV/STI Unit, Institute of Tropical Medicine, Antwerp, Belgium., Nationalestraat 155, Antwerpen, 2000, Tel 0032 (0)32476666, Fax 0032 (0)32161431.
Jozefien Buyze, Email: jbuyze@itg.be, HIV/STD Unit, Institute of Tropical Medicine, Antwerp, Belgium., Nationalestraat 155, Antwerpen, 2000, Tel 0032 (0)32476666, Fax 0032 (0)32161431.
Mark Klebanoff, Email: Mark.Klebanoff@NationwideChildrens.org, The Research Institute at Nationwide Children’s Hospital, Professor of Epidemiology, The Ohio State University College of Public Health, Professor of Pediatrics and Obstetrics and Gynecology, The Ohio State University College of Medicine, 700 Children’s Drive WB5231, Columbus, Ohio 43205, Tel 011- (614) 355-6628.
Rebecca M. Brotman, Email: rbrotman@som.umaryland.edu, Associate Professor, Department of Epidemiology and Public Health, Institute for Genome Sciences, University of Maryland School of Medicine, BioPark Building II, 801 West Baltimore Street, Room Number 627, Baltimore, MD 21201, Tel 011- (410) 706-6767, Fax 011- (410) 706-1482
References
- 1.Fethers KA, Fairley CK, Hocking JS, et al. Sexual risk factors and bacterial vaginosis: a systematic review and meta-analysis. Clin Infect Dis. 2008;47(11):1426. doi: 10.1086/592974. [DOI] [PubMed] [Google Scholar]
- 2.Brotman RM, Klebanoff MA, Nansel TR, et al. Bacterial vaginosis assessed by gram stain and diminished colonization resistance to incident gonococcal, chlamydial, and trichomonal genital infection. J Infect Dis. 2010;202(12):1907–15. doi: 10.1086/657320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Buvé A, Jespers V, Crucitti T, et al. The vaginal microbiota and susceptibility to HIV. AIDS. 2014;28(16):2333–44. doi: 10.1097/qad.0000000000000432. [DOI] [PubMed] [Google Scholar]
- 4.Kenyon CR, Colebunders R. Strong association between the prevalence of bacterial vaginosis and male point-concurrency. Eur J Obstet Gynecol Reprod Biol. 2014;172:93–6. doi: 10.1016/j.ejogrb.2013.10.011. [DOI] [PubMed] [Google Scholar]
- 5.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(6):1023–31. doi: 10.2105/AJPH.2008.147835. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Garcia PJ, Chavez S, Feringa B, et al. Reproductive tract infections in rural women from the highlands, jungle, and coastal regions of Peru. Bull World Health Organ. 2004;82(7):483–92. [PMC free article] [PubMed] [Google Scholar]
- 7.Brotman RM, Klebanoff MA, Nansel TR, et al. A longitudinal study of vaginal douching and bacterial vaginosis, a marginal structural modeling analysis. Am J Epidemiol. 2008;168(2):188–96. doi: 10.1093/aje/kwn103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. J Clin Microbiol. 1991;29(2):297–301. doi: 10.1128/jcm.29.2.297-301.1991. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Klebanoff MA, Andrews WW, Zhang J, et al. Race of Male Sex Partners and Occurrence of Bacterial Vaginosis. Sex Transm Dis. 2010;37(3):184. doi: 10.1097/OLQ.0b013e3181c04865. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wingood GM, Diclemente RJ, Robinson-Simpson L, et al. Efficacy of an HIV intervention in reducing high-risk human papillomavirus, nonviral sexually transmitted infections, and concurrency among African American women: a randomized-controlled trial. J Acquir Immune Defic Syndr. 2013;63(Suppl 1):S36–43. doi: 10.1097/QAI.0b013e3182920031. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.