Secondary Logo

Journal Logo


A Delicate Balance: Risk Factors for Acquisition of Bacterial Vaginosis Include Sexual Activity, Absence of Hydrogen Peroxide-Producing Lactobacilli, Black Race, and Positive Herpes Simplex Virus Type 2 Serology

Cherpes, Thomas L. MD*†; Hillier, Sharon L. PhD*†; Meyn, Leslie A. MS*; Busch, James L. BS; Krohn, Marijane A. PhD*†

Author Information
Sexually Transmitted Diseases: January 2008 - Volume 35 - Issue 1 - p 78-83
doi: 10.1097/OLQ.0b013e318156a5d0
  • Free

ALTHOUGH BACTERIAL VAGINOSIS (BV) IS unambiguously characterized by decreased concentrations of the normally hegemonic lactobacilli and increased concentrations of Gardnerella vaginalis, Mycoplasma hominis, Mobiluncus species, Prevotella species, and other anaerobic bacteria, its etiology remains poorly understood. Cross-sectional studies have demonstrated that BV is associated with higher number of sexual partners, higher rates of partner change, and a history of another bacterial sexually transmitted infection; all characteristics of a sexually transmitted infection.1–3 BV, however, is also frequently detected in sexually inexperienced women,4,5 and the prevalence of BV is higher at earlier stages of the menstrual cycle and is inversely related to hormonal contraceptive use.6 These latter results suggest that estrogen and progesterone levels may also contribute to BV pathogenesis. On the other hand, longitudinal studies have showed BV acquisition to be associated with the presence of a new sexual partner, a higher number of sexual partners, and the lack of hydrogen peroxide (H2O2)-producing vaginal lactobacilli.7–9

Irrespective of its etiology, BV has been associated with many adverse health outcomes. For example, published data has suggested a causal relationship between BV and premature labor and delivery.10,11 Furthermore, it is likely that the presence of BV is associated with an increased risk of postprocedural gynecologic infections,12 postpartum endometritis,13 and pelvic inflammatory disease.14 Longitudinal studies have also suggested that BV increases susceptibility for the acquisition of human immunodeficiency virus type 115 and herpes simplex virus type 2 (HSV-2).16

Because of the numerous adverse sequelae associated with BV, the Centers for Disease Control and Prevention has recommended that further research be performed to identify possible primary and secondary prevention strategies.17 Their position is bolstered by evidence that suggests antimicrobial therapies for BV are initially 80% effective, but that within 1 month after cessation of therapy, relapse or recurrence of disease is quite common.18 Therefore, the objectives of this prospective, longitudinal observational investigation of young, nonpregnant women were to identify alterable risk factors associated with the acquisition of BV, as well as to improve our understanding of this complex disease entity.

Materials and Methods

Study Participants

Sexually active, nonpregnant women between the ages of 18 and 30 years were originally recruited from 3 different Pittsburgh-area health sites [the University of Pittsburgh Student Health Clinic, the Allegheny County Health Department Sexually Transmitted Diseases (STD) Clinic, and the Family Health Council of Aliquippa] into a longitudinal cohort investigation of risk factors associated with vaginal acquisition of group B Streptococcus. Eligibility criteria for enrollment included a willingness to return for an additional 3 visits (at 4-month intervals) over the next 12 months, whereas exclusion criteria included known pregnancy, vaginal bleeding, current use of systemic antimicrobials, or the use of douches, spermicides, or antifungal vaginal products in the 24 hours before enrollment. Informed consent was obtained from all study participants, and the Magee-Womens Hospital Institutional Review Board of the University of Pittsburgh Medical Center approved the study protocol used in this investigation.

Between 1998 and 2001 inclusive, a total of 1248 women were enrolled. Of the 1089 women who returned for at least 1 follow-up visit, 773 did not have BV at enrollment. It was this subset of women that was used to investigate risk factors associated with the acquisition of BV. At each visit, demographic and behavioral interview data, a vaginal swab, a vaginal smear, and a serum sample were collected from all study participants.

Laboratory Methods

The laboratory methods used in this investigation have been previously described.16 Vaginal smears were Gram stained and evaluated for BV using standardized criteria (lactobacilli predominant specimens received a score of 0–3; intermediate flora, 4–6; and BV, 7–10.19 Vaginal swab specimens were used for identification of H2O2-producing and H2O2-nonproducing lactobacilli, Group B Streptococci, and vaginal yeast.20 Also as previously described, serum samples were tested for type-specific antibodies to herpes simplex virus types 1 and 2 (HSV-1 and HSV-2) using commercially available enzyme-linked immunoassays (Focus Technologies, Cypress, CA).16

Statistical Analyses

Acquisition of BV was defined as a change in the vaginal morphology from a score for normal (0–3) or intermediate (4–6) at enrollment to BV (7–10) at the 4-, 8-, or 12-month follow-up visits. Loss of H2O2-producing vaginal lactobacilli was similarly defined as the presence of these organisms (as detected by bacterial culture) at enrollment and their absence at any one of the follow-up visits.

Cox proportional hazards models with time-varying covariates were used to evaluate variables associated with the acquisition of BV or loss of H2O2-producing lactobacilli. Models were developed using Stata statistical software, release 9.2 (Stata Corporation, College Station, TX), and included use of the vaginal culture and HSV serologic results from the visit before BV acquisition. Efron’s method was used for handling tied failure times,21 and the variance estimates were calculated using the method of Lin and Wei.22 Variables were considered for inclusion in these models if the P value from the log-rank test for equality of survivor functions was less than 0.1. Forward stepwise regression was employed, and variables were retained in the model if the P value from the Wald χ2 test statistic was 0.05 or less.


Of the 773 women who did not have BV at enrollment, 538 (70%) described themselves as white, 203 (26%) as black, and 32 (4%) as Hispanic, Asian, Native American, or multiethnic. The latter group of women was combined in a single racial category, referred to as “other” for purposes of analysis. The number of women from each site included in this analysis was similar: 306 (39%) from the University of Pittsburgh Student Health Clinic, 243 (31%) from the Family Health Council of Aliquippa, and 224 (29%) from the Allegheny County Health Department STD Clinic.

As determined by Gram stain analysis of vaginal smears there were 223 acquisitions of BV. The 773 study participants returned to clinic for a total of 1763 follow-up visits, from which 619 woman-years of follow-up were accumulated. This resulted in an overall BV incidence rate in our investigation of 36 cases/100 woman-years. On univariate analysis, BV acquisition was found to be associated with enrollment from either the Family Health Council of Aliquippa or the Allegheny County Health Department STD clinic (Table 1). Other demographic and behavioral characteristics associated with the acquisition of BV included black race, ≤12 years of education, and cigarette smoking or use of douche preparations in the previous 4 months (Table 1).

Demographic and Behavioral Characteristics of Women Who Acquired Bacterial Vaginosis by Gram Stain Evaluation of Vaginal Smears

As shown in Table 2, univariate analysis of selective sexual behavioral characteristics demonstrated that the rate of BV acquisition was significantly higher among women who had reported being sexually active during the past 4 months. Similarly, the rate of BV acquisition appeared to be directly correlated with the frequency of vaginal intercourse. As previously demonstrated in other longitudinal investigations,7,8 we found that the number of sex partners in the prior 4 months was directly related to the risk of BV acquisition. However, in contrast to another previously completed longitudinal investigation,9 we found that the risk for the acquisition of BV associated with a new male sex partner did not reach statistical significance [34.0 cases/100 woman-years among women who did not report a new male sex partner vs. 45.3 cases/100 woman-years among women who did; hazard ratio (HR), 1.3; 95% confidence interval (CI), 1.0–1.8; P = 0.09]. As shown in Table 2, receptive oral sex, unprotected vaginal intercourse, sexually activity with another woman, and vaginal intercourse during the 4 days before a follow-up visit also did not appear to be associated with a significantly increased risk for the acquisition of BV. Interestingly, vaginal intercourse immediately after receptive anal intercourse (HR, 2.0; 95% CI, 1.0–4.2; P = 0.05) and sex with an uncircumcised male partner (HR, 1.9; 95% CI, 1.0–3.5; P = 0.04) were both associated with a significantly increased risk for the acquisition of BV (Table 2).

Sexual Behavioral Characteristics of Women Who Acquired Bacterial Vaginosis by Gram Stain Evaluation of Vaginal Smears

As shown in Table 3, women with intermediate vaginal flora morphotypes were much more likely to acquire BV than were women with normal vaginal flora (HR, 3.2; 95% CI, 2.5–4.1; P <0.001). As previously demonstrated,9 lack of H2O2-producing vaginal lactobacilli also significantly increased the risk for the acquisition of BV (HR, 2.3; 95% CI, 1.7–3.0; P <0.001). By univariate analysis, the identification of vaginal Group B Streptococci or vaginal yeast in the study visit preceding the acquisition of BV was associated with BV acquisition. In addition, women who had serum antibodies to either HSV-1 or HSV-2 at the antecedent study visit were more likely to acquire BV than were women who were seronegative for these viruses (HR, 1.4; 95% CI, 1.1–1.8; P = 0.02, and HR, 2.7; 95% CI, 2.1–3.6; P <0.001, respectively) (Table 3).

Vaginal Microbiologic and Herpes Simplex Virus Types 1 and 2 Serologic Results of Women Who Acquired Bacterial Vaginosis by Gram Stain Evaluation of Vaginal Smears

Cox proportional hazards analysis demonstrated that the association between BV acquisition and HSV-2, but not HSV-1, antibodies persisted after adjusting for potential confounders (HR, 1.7; 95% CI, 1.3–2.3; P <0.001) (Table 4). The demographic and behavioral characteristics found to be independently associated with the acquisition of BV were black race, cigarette smoking, sexual activity, vaginal intercourse immediately after anal intercourse, and vaginal intercourse with an uncircumcised male partner. In addition, we found that those women who had vaginal smears consistent with intermediate vaginal flora or whose vaginal cultures did not contain H2O2-producing vaginal lactobacilli at their antecedent study visit were much more likely to acquire BV than were women with a normal vaginal Gram stain or H2O2-producing vaginal lactobacilli (Table 4).

Summary of the Multivariable Logistic Regression Analyses of the Factors Associated With Acquisition of Bacterial Vaginosis by Gram Stain Evaluation of Vaginal Smears


Our study shows that a number of quite diverse risk factors increase the likelihood that a woman will acquire BV: black race, sexual activity, vaginal intercourse after anal sex, intercourse with an uncircumcised male partner, lack of H2O2-producing vaginal lactobacilli, intermediate (reduced lactobacilli) vaginal flora, cigarette smoking, and HSV-2 infection may all, at various times and in different women, play a role in BV pathogenesis. Because our study delineates such a varied and multifactorial etiology for BV, it also suggests that this condition will likely frustrate those who seek to identify specific microorganisms wholly causative for this condition.

The microbial flora present in the normal vagina is illustrative of the permutability of biologic ecosystems in that a number of diverse system perturbations can produce the same deleterious effect. Specifically, the commonality of the risk factors for BV acquisition identified in this study may be that they all adversely affect survivability or maintenance of H2O2-producing lactobacilli. Vallor et al. reported that vaginal intercourse, at a weekly frequency ≥1, was the only behavioral variable associated with loss of H2O2-producing vaginal lactobacilli.23 As the alkaline buffering action of semen abolishes vaginal acidity for many hours after intercourse,24 it is possible that this temporary loss of acidity is detrimental to the maintenance of H2O2-producing lactobacilli or permissive for anaerobic bacterial overgrowth. Previous research has shown that intermediate (reduced lactobacilli) vaginal flora is unstable—nearly two-thirds of women with this pattern will quickly shift to either normal flora or acquire BV.25 As shown in Table 3, women with no vaginal lactobacilli have similar rates of BV acquisition when compared with those women who lack H2O2-producing lactobacilli; this suggests that it is the absence of H2O2-producing vaginal lactobacilli, rather than an absolute deficiency of lactobacilli, that is more closely associated with increased acquisition of BV. Our results corroborate those reported by Hawes et al., who found that BV was independently associated with the lack of vaginal H2O2-producing lactobacilli.9

Cross-sectional analyses demonstrate that BV is associated with increased frequencies of Neisseria gonorrhoeae, Chlamydia trachomatis, and HSV-2 infections,26,27 whereas a longitudinal analysis shows that BV can increase susceptibility for HSV-2 acquisition.16 However, these findings do not exclude the possibility that bacterial and viral infections of the lower genital tract can also increase susceptibility for the acquisition of BV. In our longitudinal cohort investigation, acquisition of BV is independently associated with the prior detection of HSV-2 serum antibodies. Comparable results to these were reported in a longitudinal study of 273 women conducted in Burkina-Faso.28 Additional multivariable analyses in our study cohort demonstrate that HSV-2 serum antibodies are also independently associated with the loss of H2O2-producing vaginal lactobacilli (HR, 1.6; 95% CI, 1.2–2.1; P = 0.002) (data not shown). These results suggest that altered local physiological and immunologic responses induced by genital tract infections such as genital herpes may produce a vaginal environment more likely to promote the overgrowth of anaerobic Gram-negative bacteria or decrease the survival of H2O2-producing lactobacilli.

Although establishing consistently causal relationships between specific microorganisms and BV acquisition may prove difficult, it is likely that the introduction of bacteria not typically present in the normal vaginal ecosystem can contribute to the pathogenesis of this disease. Our current study demonstrates that both vaginal sex after anal intercourse and a recent history of an uncircumcised male sex partner increase the likelihood for the acquisition of BV. Although previously published cross-sectional investigations did not detect a significant association between circumcision status of the male partner and BV,29,30 we do know that, in comparison with circumcised males, higher proportions of Gram-negative rods are found in the subpreputial space of uncircumcised males.31 Further work is needed to delineate the mechanisms by which particular bacterial infections increase the incidence of BV. However, it is possible that receptive anal sex before vaginal intercourse or vaginal intercourse with an uncircumcised male partner allows certain microorganisms and microbial products to disrupt normal vaginal homeostasis or induce local inflammatory responses that increase the likelihood for the acquisition of BV.

Although the current investigation is one of the largest studies of BV acquisition to date, our conclusions may be limited by the small number of woman-years of follow-up among women who acquired BV and reported a recent history of vaginal intercourse after receptive anal intercourse or vaginal intercourse with an uncircumcised male partner. It is also likely that some cases of BV acquisition were missed between the 3-month follow-up visits. Our investigation does strongly suggest that multiple behavioral and biologic variables are associated with BV acquisition, and that H2O2-producing lactobacilli appear to be a necessary component of a stable vaginal ecosystem. Diverse perturbations, including those identified by this study, may work separately or in combination to disrupt the homeostatic balance of this biologic ecosystem, thereby increasing susceptibility for the acquisition of BV.


1. Hart G. Factors associated with trichomoniasis, candidiasis, and bacterial vaginosis. Int J STD AIDS 1993; 4:21–25.
2. Nilsson U, Hellberg D, Shoubnikova M, et al. Sexual behavior risk factors associated with bacterial vaginosis and Chlamydia trachomatis infection. Sex Transm Dis 1997; 24:241–246.
3. Smart S, Singal A, Mindel A. Social and sexual risk factors for bacterial vaginosis. Sex Transm Infect 2004; 80: 58–62.
4. Bump RC, Buesching WJ. Bacterial vaginosis in virginal and sexually active adolescent females: Evidence against exclusive sexual transmission. Am J Obstet Gynecol 1988; 158:935–939.
5. Yen S, Shafer MA, Moncada J, et al. Bacterial vaginosis in sexually experienced and non-sexually experienced young women entering the military. Obstet Gynecol 2003; 102:927–933.
6. Calzolari E, Masciangelo R, Milite V, et al. Bacterial vaginosis and contraceptive methods. Int J Gynecol Obstet 2000; 70:341–346.
7. Barbone F, Austin H, Louv WC, et al. A follow-up study of methods of contraception, sexual activity, and rates of trichomoniasis, candidiasis, and bacterial vaginosis. Am J Obstet Gynecol 1990; 163:510–514.
8. Avonts D, Sercu M, Heyerick P, et al. Incidence of uncomplicated genital infections in women using oral contraception or an intrauterine device: a prospective study. Sex Transm Dis 1990; 17:23–29.
9. Hawes SE, Hillier SL, Benedetti J, et al. Hydrogen peroxide-producing lactobacilli and acquisition of vaginal infections. J Infect Dis 1996; 174:1058–1063.
10. Eschenbach DA, Gravett MG, Chen KC, et al. Independent associations of bacterial vaginosis and Chlamydia trachomatis infection with adverse pregnancy outcome. JAMA 1986; 256:1899–1903.
11. Hillier SL, Nugent RP, Eschenbach DA, et al. Association between bacterial vaginosis and preterm delivery of a low-birth-weight infant. The Vaginal Infections and Prematurity Study Group. N Engl J Med 1995; 333:1737–1742.
12. Larsson P, Platz-Christensen JJ, Thejls H, et al. Incidence of pelvic inflammatory disease after first trimester legal abortion in women with bacterial vaginosis after treatment with metronidazole: A double-blind, randomized study. Am J Obstet Gynecol 1992; 166:100–103.
13. Watts DH, Krohn MA, Hillier SL, et al. Bacterial vaginosis as a risk factor for post cesarean endometritis. Obstet Gynecol 1990; 75:52–58.
14. Paavonen J, Teisala K, Heinonen PK, et al. Microbiological and histopathological findings in acute pelvic inflammatory disease. Br J Obstet Gynaecol 1987; 94:454–460.
15. Martin HL, Richardson BA, Nyange PM, et al. Vaginal lactobacilli, microbial flora, and risk of human immunodeficiency virus type 1 and sexually transmitted disease acquisition. J Infect Dis 1999; 180:1863–1868.
16. Cherpes TL, Meyn LA, Krohn MA, et al. Association between acquisition of herpes simplex virus type 2 in women and bacterial vaginosis. Clin Infect Dis 2003; 37:319–325.
17. Koumans EH, Kendrick J; CDC Bacterial Vaginosis Working Group. Preventing adverse sequelae of bacterial vaginosis: A public health program and research agenda. Sex Transm Dis 2001; 28:292–297.
18. Clinical Effectiveness Group (Association of American and the Medical Society for the Study of Venereal Diseases). National guideline for the treatment of bacterial vaginosis. Sex Transm Infect 1999; 75(1S):16S–18S.
19. 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; 29:297–301.
20. Meyn LA, Moore DM, Hillier SL, et al. Association of sexual activity with colonization and vaginal acquisition of Group B Streptococcus in nonpregnant women. Am J Epidemiol 2002; 155:949–957.
21. Efron B. The efficiency of Cox’s likelihood function for censored data. J Am Stat Assoc 1977; 72:557–565.
22. Lin DY, Wei LJ. The robust inference for the Cox proportional hazards model. J Am Stat Assoc 1989; 84:1074–1078.
23. Vallor AC, Antonio MAD, Hawes SE, et al. Factors associated with acquisition of, or persistent colonization by, vaginal lactobacilli: Role of hydrogen peroxide production. J Infect Dis 2001; 184:1431–1436.
24. Boskey ER, Telsch KM, Whaley KJ, et al. Acid production by vaginal flora in vitro is consistent with the rate and extent of vaginal acidification. Infect Immun 1999; 67:5170–5175.
25. Hillier SL, Krohn MJ, Nugent RP, et al. Characteristics of three vaginal flora patterns assessed by gram stain among pregnant women. Vaginal Infections and Prematurity Study Group. Am J Obstet Gynecol 1992; 166:938–944
26. Wiesenfeld HC, Hillier SL, Krohn MJ, et al. Bacterial vaginosis is a strong predictor of Neisseria gonorrhoeae and Chlamydia trachomatis infection. Clin Infect Dis 2003; 36:663–668.
27. Cherpes TL, Meyn LA, Krohn M, et al. Risk factors for herpes simplex virus type 2: Role of smoking, douching, uncircumcised males, and vaginal flora. Sex Transm Dis 2003; 30:405–410.
28. Nagot N, Ouedraogo A, Defer MC, et al. Association between bacterial vaginosis and herpes simplex virus type-2 infection: Implications for HIV acquisition. Sex Transm Infect 2007; 83:365–368.
29. Zenilman JM, Fresia A, Berger B, et al. Bacterial vaginosis is not associated with circumcision status of the current male partner. Sex Transm Infect 1999; 75:347–348.
30. Bukusi EA, Cohen CR, Meier AS, et al. Bacterial vaginosis: Risk factors among Kenyan women and their male partners. Sex Transm Dis 2006; 33:361–367.
31. Serour F, Samra Z, Kushel Z, et al. Comparative periuretheral bacteriology of uncircumcised and circumcised males. Genitourin Med 1997; 73:288–290.
© Copyright 2008 American Sexually Transmitted Diseases Association