点击显示 收起
Department of Obstetrics and Gynecology, MetroHealth Medical Center and Case Western Reserve University, Cleveland, Ohio
Department of Obstetrics, Gynecology and Reproductive Sciences, Magee-Womens Hospital, University of Pittsburgh
Magee-Womens Research Institute, Pittsburgh, Pennsylvania
Background.
Women with bacterial vaginosis have different microbiological profiles. Our objective was to identify risk factors for an absence of H2O2-producing lactobacilli among women with bacterial vaginosis.
Methods.
We performed a retrospective analysis of 947 women with bacterial vaginosis who were enrolled in prospective studies investigating vaginal colonization and genital-tract infections.
Results.
Women were categorized into 2 groups: those with H2O2-producing lactobacilli present (n = 191; 20.2%) and those with H2O2-producing lactobacilli absent (n = 756; 79.8%). Multivariate logistic regression demonstrated that douching 2 times during the past month (odds ratio [OR], 2.5; 95% confidence interval [CI], 1.16.0) and having 3 sex partners during the past year (OR, 4.5; 95% CI, 1.811.7) were predictive of an absence of H2O2-producing lactobacilli.
Conclusion.
Among women with bacterial vaginosis, H2O2-producing lactobacillus colonization is influenced by sexual activity and douching habits. These findings may have important implications for response to treatment, relapse rate, and risk for sexually-transmitted-disease acquisition among women with bacterial vaginosis.
Bacterial vaginosis is the most prevalent vaginal infection in reproductive-age women [1, 2]. A healthy vaginal environment is characterized by an abundance of lactobacilli that produce hydrogen peroxide (H2O2) and a corresponding acidic vaginal pH. In contrast, bacterial vaginosis is characterized by a shift in vaginal flora, from one characterized by lactobacilli dominance to one with a relative decrease or complete absence of lactobacilli and a corresponding increase in both types and numbers of bacteria, including Gardnerella vaginalis, anaerobic gram-negative rods, anaerobic gram-positive cocci, Mycoplasma hominis, and Mobiluncus species. However, it is increasingly recognized that bacterial vaginosis is microbiologically heterogeneous, with 1 in 4 women with bacterial vaginosis having H2O2-producing lactobacilli present [3].
Numerous recent reports have heightened the interest in and awareness of bacterial vaginosis and its importance in reproductive health. Important associations have been demonstrated between bacterial vaginosis and preterm delivery and intrauterine infection [4, 5], postpartum endomyometritis [6], pelvic inflammatory disease [7, 8], and postprocedural gynecological infections [9, 10]. Moreover, there is evidence linking bacterial vaginosis to an increased susceptibility to acquisition of other sexually transmitted diseases (STDs), including HIV infection [1114]. H2O2 is toxic to many microorganisms at concentrations that are readily achievable in the vaginal fluid and, thus, provides an intrinsic protective mechanism in the vaginal compartment [15, 16]. Most women with bacterial vaginosis have, at the very least, a relative decrease in lactobacilli concentration, and some have a complete absence of lactobacilli [3, 15]. One theory proposed to explain the enhanced susceptibility to STDs and HIV infection among women with bacterial vaginosis relates to decreased amounts of H2O2-producing lactobacilli in women with bacterial vaginosis [2, 15, 16].
Epidemiological investigations thus far have focused on risk factors for and predictors of altered vaginal flora and/or bacterial vaginosis among healthy women. Given the central role played by lactobacilli in the maintenance of vaginal health, we sought to determine the factors that predict presence and/or absence of lactobacilli (including H2O2-producing lactobacilli) among women with bacterial vaginosis. These different microbiological characteristics among women with bacterial vaginosis may have important implications for response to treatment and future vaginal susceptibility to other infections.
SUBJECTS AND METHODS
A cohort of women with bacterial vaginosis was constructed from 4 separate studies performed at Magee-Womens Hospital from 1998 to 2002. Informed consent was obtained from all patients included in this evaluation, and all protocols were approved by and followed the guidelines of the Magee-Womens Hospital Institutional Review Board of the University of Pittsburgh Medical Center. Two of the studies were bacterial vaginosis treatment studies, 1 was a longitudinal cohort study of vaginal colonization, and the 1 was a study evaluating women at risk for STDs and pelvic inflammatory disease. Only those women noted to have bacterial vaginosis at their enrollment visit, as determined by a vaginal-smear Nugent score of 710, were included in the cohort [17]. All women enrolled in these 4 studies underwent the same procedures (interviews, physical exams, and specimen collection) performed by trained research personnel from the same research group.
Interviews consisted of closed-format questions addressing demographics, medical and reproductive history, sexual history and practices, recent medication use, social history, and current/recent genital-tract symptoms. In the creation of the data set, study number and patient name were checked to ensure that no woman was included twice in the data set. For some of the demographic and behavioral variables, information was missing because some women did not answer all questions.
Women underwent a thorough gynecologic examination with documentation of physical findings. Vaginal swabs were obtained, in a standardized fashion, for culture and Gram stain evaluation of microbiological evidence for bacterial vaginosis [17]. The microbiological analyses were performed in the laboratory of one of the coauthors (S.L.H.) and were standardized for all 4 studies. Vaginal swabs were inoculated onto Rogosa agar for the isolation of Lactobacillus species. The agar was incubated anaerobically for 4 days at 37°C. After 4 days of incubation, the plates were examined for Lactobacillus, which was identified by assessment of colony and Gram stain morphology. Lactobacilli were tested for H2O2 production, by use of a qualitative assay on a tetramethylbenzidine agar plate. These agar plates were subsequently incubated in an anaerobic glove box for 2 days at 37°C and then were exposed to ambient air. The H2O2 that was produced reacted with the horseradish peroxidase in the agar to oxidize the tetramethylbenzidine, which turns the Lactobacillus colonies blue [18].
Data were analyzed using SPSS statistical software (version 10.1.4). Two microbiological categories of lactobacilli colonization were created among these women: (1) those with H2O2-producing lactobacilli present and (2) those with either nonH2O2-producing lactobacilli present or no lactobacilli present. Women with no Lactobacillus species and women with nonH2O2-producing Lactobacillus species were combined into 1 group because there was no difference in demographic and behavioral variables between these groups (data not shown). The comparison of patient demographic data (table 1) was performed using generalized 2 test statistics. The resultant dichotomous outcome was evaluated for adjusted point estimates of odds ratios determined by binary logistic regression. Multiple variables of interest were tested for univariate relationships to the outcome, and only those factors found to be significantly associated (at the 0.1 level) were included in the multivariate logistic regression model. Models were generated using forward stepwise regression based on the likelihood-ratio test, and variables were retained in the model if the Wald test statistic had a P value of .05. Adjusted point estimates of risk for having no H2O2-producing lactobacilli are presented with their 95% confidence intervals (CIs).
RESULTS
A total of 947 women with bacterial vaginosis were available for analysis. H2O2-producing lactobacilli were identified in 191 women (20.2%), whereas 222 women (23.4%) had nonH2O2-producing lactobacilli present, and the remaining 534 women (56.4%) had no lactobacilli. Combining the last 2 categories resulted in a group of 756 women (79.8%) who lacked H2O2-producing lactobacilli.
Demographic and behavioral data for the 2 groups are shown in table 1; these data were statistically similar between the 2 outcome groups, with the exception of marital status, employment, cigarette use, douching habits, numbers of sex partners in the past year, frequency of sexual intercourse, history of bacterial vaginosis, and recent antibiotic use. The racial mix in the cohort was 340 (36%) white and 607 (64%) African American. Most of the women (72%) were 25 years of age, and 46% of the population had a high school education or less. The vast majority of the women (90%) were not married. Forty percent of the women reported cigarette use, and 45% reported alcohol use.
In terms of sexual history, we found that women with H2O2-producing lactobacilli had statistically significantly fewer sex partners during the past year and reported less sexual intercourse during the past month, compared with women lacking H2O2-producing lactobacilli. However, numbers of sex partners during the past month did not vary significantly between these 2 groups. Of 946 women, 99 reported antibiotic use during the past month. Three categories of antibiotic use were created for the regression. One category consisted of women reporting metronidazole use during the past month (N = 23), another category consisted of women reporting use of all other classes of antibiotics during the past month (N = 76), and the last category consisted of women who reported no antibiotic use during the past month (N = 847).
Nearly 20% of the women douched at least 1 time per month, and an additional 5% douched 2 times per month. In addition, 9% of the women reported that they had douched within 7 days of the interview. Factors associated with a history of frequent douching (2 times in the past month) were then evaluated. Nonwhite race (P < .001), single marital status (P = .02), cigarette use (P = .02), and having 1 sex partners during the past month (P = .01) were related to douching frequency. Number of sex partners during the past year and frequency of sexual intercourse during the past month did not show a significant relationship with douching frequency.
Multivariate logistic regression was performed using the variables for which a statistically significant difference existed between the 2 microbiological groups, as well the 2 variables hypothesized a priori to be related to lactobacilli colonization but that did not reach statistical significance in univariate analyses (i.e., race and number of sex partners during the past month). Douching 2 times during the past month (OR, 2.5; 95% CI, 1.16.0) and having 3 sex partners during the past year (OR, 4.5; 95% CI, 1.811.7) independently predicted an absence of H2O2-producing lactobacilli (table 2). Women who used metronidazole during the past month demonstrated a nonsignificant trend toward more H2O2-producing lactobacilli colonization, compared with women without recent antibiotic use (OR, 0.4; 95% CI, 0.21.0). In addition, use of antibiotics other than metronidazole was statistically significantly associated with the presence of H2O2-producing lactobacilli, compared with no use of antibiotics during the past month (OR, 0.6; 95% CI, 0.30.9). Given that this was an analysis generated from 4 different studies, we analyzed the data stratified by study in which the patients were recruited and found no differential effect on our findings.
DISCUSSION
The present study delineates risk factors for an absence of H2O2-producing lactobacilli among women with bacterial vaginosis. Douching 2 times during the past month more than doubled the odds of lacking H2O2-producing lactobacilli, and having 3 sex partners during the past year more than quadrupled the odds. This study differs from others that have examined risk factors for disrupted vaginal flora, because it focuses solely on women with bacterial vaginosis and not on the general population.
Lactobacilli and the H2O2 they produce are increasingly recognized as essential components of a healthy vaginal environment. H2O2 is toxic to many microorganisms at concentrations that are typical in the vaginal fluid and, thus, provides an intrinsic protective mechanism in the vaginal compartment [15, 16]. Others have investigated the risk factors for acquisition of bacterial vaginosis and/or the loss of vaginal lactobacilli colonization in the general population, with similar findings. It is recognized that sexual and vaginal hygienic practices play a role in vaginal ecology. Barbone et al. demonstrated that bacterial vaginosis is associated with increased numbers of sex partners in the month preceding diagnosis [19]. Douching (particularly frequent and recent) is associated with a 2-fold increase in the risk for bacterial vaginosis and for presence of bacterial vaginosisassociated microflora [20, 21]. These studies are consistent with our data on the role of sexual activity and douching among women who currently have bacterial vaginosis.
Given the relationship between douching and vaginal flora that has been found in this investigation and in others, we evaluated factors related to douching. This was also done because douching was hypothesized to be a factor that is potentially modifiable through education. We found that some of the factors related to douching are nonmodifiable (with the exception of cigarette use and number of sex partners). It is unclear whether changing these nonmodifiable factors would change douching habits. Moreover, whether a change in douching habits would produce an improvement in vaginal flora among women who subsequently get bacterial vaginosis remains to be determined.
In another recent prospective cohort study [22], Vallor et al. investigated the factors associated with acquisition and loss of vaginal lactobacilli. They reported that more frequent sexual intercourse (>1 sex act per week) and recent use of antibiotics predicted a loss of colonization of H2O2-producing lactobacilli. They also demonstrated that women colonized with H2O2-producing lactobacilli at the baseline visit were significantly more likely to be persistently colonized throughout the entire 8-month study period. The results of our study are consistent with the findings from these other investigations, in that we found that sexual habits are important factors in predicting an absence of H2O2-producing lactobacilli. The exact mechanism for the association is unclear. Sexual intercourse may cause a disruption of the local vaginal microbiological and/or inflammatory milieu; therefore, more frequent and/or new partners may accentuate these proposed changes.
Multiple investigators have demonstrated increased risks for acquisition of other STDs in patients with bacterial vaginosis and/or disrupted vaginal flora. Martin et al. published results from a prospective cohort study assessing acquisition of HIV-1 and other STDs in Kenyan female sex workers [11]. Compared with women with H2O2-producing lactobacilli, women without lactobacilli had a significantly increased hazard ratio (HR, 2.8) for acquisition of HIV-1. The investigators also demonstrated a nonsignificant trend for acquisition of HIV-1 in women with nonH2O2-producing lactobacilli (HR, 1.7). The stepwise increase in susceptibility to HIV acquisition gives strength to the biological plausibility of the association. In a separate prospective longitudinal study of pregnant and postpartum women, Taha et al. reported that ORs for HIV seroconversion increased as the level of vaginal flora disturbance increased (from normal to bacterial vaginosis) and that the increase reached statistical significance for linear trend in pregnant and postpartum women [12]. Wiesenfeld et al. demonstrated that women with bacterial vaginosis who were exposed to partners with diagnoses of gonococcal, chlamydial, or nongonoccocal urethritis were more likely to test positive for Neisseria gonorrhoeae (OR, 4.1) and Chlamydia trachomatis (OR, 3.4) [13]. Women colonized by H2O2-producing lactobacilli were then compared with women with no lactobacilli colonization. They found that, when H2O2-producing lactobacilli were present, women were significantly less likely (OR, 0.4; CI, 0.20.8) to test positive for C. trachomatis, and a nonsignificant trend for less N. gonorrhoeae was also found (OR, 0.6; CI, 0.21.3). Most recently, Cherpes et al. used a longitudinal cohort design (with subjects followed for 1 year at 4-month intervals) of sexually active women aged 1830 years to demonstrate that women with bacterial vaginosis at baseline were twice as likely (HR, 2.1) to experience seroconversion to HSV-2, compared with those without bacterial vaginosis [14]. These studies, taken together, demonstrate the importance of disrupted vaginal flora in susceptibility to STDs. The reestablishment of an H2O2-producing, lactobacillus-predominant vaginal flora among women with bacterial vaginosis may be essential as a host defense against the acquisition of HIV and other STDs.
The finding of a nonsignificant trend for metronidazole improving the flora in the vagina is intriguing. Likewise, the observation that the use of all other classes of antimicrobials was associated with an increased likelihood of H2O2-producing lactobacilli being present also deserves notice. Agnew et al. evaluated the effect that different antimicrobials used for vaginitis and cervicitis had on vaginal lactobacilli and demonstrated findings similar to ours. In their evaluation, women who used either oral or intravaginal metronidazole had a net increase in vaginal lactobacilli (H2O2-producing and nonH2O2-producing strains) colonization after posttreatment periods of 1 week and 1 month [3]. These results may be explained by metronidazole's potent antianaerobic spectrum of action, which suppresses anaerobic organisms and allows for recolonization by aerobic lactobacilli. Further, since it lacks activity against gram-positive bacteria, there are few adverse effects on lactobacillus growth. The explanation for antibiotic use in general producing a modest (40%) increase in H2O2-producing lactobacilli is not easily explained. It is possible that the use of antimicrobials is a proxy for health-care utilization in general, affording women an opportunity for evaluation and treatment of lower-genitourinary-tract infections. It is also possible that our findings are spurious, because of the relatively small numbers of women reporting antibiotic use, and that larger numbers may have altered the findings.
The results of the present analysis suggest (on the basis of univariate analyses only) that a higher percentage of women with H2O2-producing lactobacilli had 3 episodes of bacterial vaginosis in the past year. At first, this seems counterintuitive, since one would speculate that women with more frequent infections may have worse microbiological profiles. However, as noted above, women treated with metronidazole may demonstrate higher rates of lactobacilli colonization after treatment. We hypothesize that the women in this study who had recently recognized and treated bacterial vaginosis infections may have benefited from treatment with metronidazole.
Some limitations of our study are worth noting. Although enrollment numbers are large, this cohort of women was constructed from 4 separate studies and, thus, may have an amount of unrecognized heterogeneity. The impact of this nonquantifiable factor is likely to be small, given the standardization of our microbiological techniques and the assignment of the same research personnel to perform the patient interviews and specimen collection. Moreover, stratification of the analysis by study origin failed to change our results in any appreciable manner. Another potential limitation is the fact that the historical component of the interview session was based entirely on patient recall and was thus susceptible to recall bias and other potential unidentified biases. In addition, the cross-sectional nature of this investigation did not allow for the evaluation of the clinical importance of the findings. Along those lines, an additional potential limitation was our inability to assess the acuity of the infection and to evaluate how long the women had had bacterial vaginosis. It is possible that lactobacilli colonization fluctuates during the course of an infection, which would make timing important in terms of lactobacilli recovery. Despite these factors, the large numbers of patients (undergoing the same procedures and laboratory methods) and the strength of associations in our study provide clear insight into the risk factors for an absence of H2O2-producing lactobacilli among women with bacterial vaginosis.
In summary, our data indicate that douching frequently and having multiple sex partners increases the risk for an absence of protective H2O2-producing lactobacilli among women with bacterial vaginosis. Recent use of metronidazole may promote the presence of H2O2-producing lactobacilli in the vaginal microenvironment. In light of the central role that lactobacilli and hydrogen peroxide play in vaginal health, these findings may have important implications for response to treatment of bacterial vaginosis and future risk for acquisition of STDs, including HIV.
References
1. Sobel JD. Vaginitis in adult women. Obstet Gynecol Clin North Am 1990; 17:85179. First citation in article
2. Hillier SL, Holmes KK. Bacterial vaginosis. In: Holmes KK, Sparling PF, Mrdh P-A, et al., eds. 3rd ed. Sexually transmitted diseases. New York: McGraw-Hill, 1999:56386. First citation in article
3. Agnew KJ, Hillier SL. The effect of treatment regimens for vaginitis and cervicitis on vaginal colonization by lactobacilli. Sex Transm Dis 1995; 22:26973. First citation in article
4. Hillier SL, Nugent RP, Eschenbach DA, et al. Association between bacterial vaginosis and preterm delivery of a low birth weight infant. N Engl J Med 1995; 333:173742. First citation in article
5. Hillier SL, Martius J, Krohn MA, et al. A case-control study of chorioamnionic infection and histologic chorioamnionitis in prematurity. N Engl J Med 1988; 319:9728. First citation in article
6. Watts DH, Krohn MA, Hillier SJ, Eschenbach DA. Bacterial vaginosis as a risk factor for post cesarean endometritis. Obstet Gynecol 1990; 75:528. First citation in article
7. Paavonen J, Teisala K, Heinonen PK, et al. Microbiological and histopathological findings in acute pelvic inflammatory disease. Br J Obstet Gynaecol 1987; 94:45460. First citation in article
8. Hillier SL, Kiviat NB, Hawes SE, et al. Role of bacterial vaginosis-associated microorganisms in endometritis. Am J Obstet Gynecol 1996; 175:43541. First citation in article
9. 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; 166:1003. First citation in article
10. Soper DE, Bump RC, Hurt WG. Bacterial vaginosis and trichomoniasis vaginitis are risk factors for cuff cellulitis after abdominal hysterectomy. Am J Obstet Gynecol 1990; 163:101621. First citation in article
11. Martin HL, Richardson BA, Nyange PM, et al. Vaginal lactobacilli, microbial flora, and risk of human immunodeficiency virus type I and sexually transmitted disease acquisition. J Infect Dis 1999; 180:18638. First citation in article
12. Taha TE, Hoover DR, Dallabetta GA, et al. Bacterial vaginosis and disturbances of vaginal flora: association with increased acquisition of HIV. AIDS 1998; 12:1699706. First citation in article
13. Wiesenfeld HC, Hillier SL, Krohn MA, Landers DV, Sweet RL. Bacterial vaginosis is a strong predictor of Neisseria gonorrhoeae and Chlamydia trachomatis infection. Clin Infect Dis 2003; 36:6638. First citation in article
14. Cherpes TL, Meyn LA, Krohn MA, Lurie JG, Hillier SL. Association between acquisition of herpes simplex virus type 2 in women and bacterial vaginosis. Clin Infect Dis 2003; 37:31925. First citation in article
15. Hillier SL, Holmes KK. Normal vaginal flora. In: Holmes KK, Sparling PF, Mrdh P-A, et al., eds. 3rd ed. Sexually transmitted diseases. New York: McGraw-Hill, 1999:191203. First citation in article
16. Eschenbach DA, Davick PR, Williams BL, et al. Prevalence of hydrogen peroxide-producing Lactobacillus species in normal women and women with bacterial vaginosis. J Clin Microbiol 1989; 27:2516. First citation in article
17. 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; 29:297301. First citation in article
18. Rabe LK, Hillier SL. Optimization of media for detection of hydrogen peroxide production by Lactobacillus species. J Clin Microbiol 2003; 41:32604. First citation in article
19. Barbone F, Austin H, Louv WC, Alexander WJ. A follow-up study of methods of contraception, sexual activity, and rates of trichomoniasis, candidiasis, and bacterial vaginosis. Am J Obstet Gynecol 1990; 163:5104. First citation in article
20. Ness RB, Hillier SL, Richter HE, et al. Douching in relation to bacterial vaginosis, lactobacilli, and facultative bacteria in the vagina. Obstet Gynecol 2002; 100:76572. First citation in article
21. Hawes SE, Hilier SL, Benedetti J, et al. Hydrogen peroxide-producing lactobacilli and acquisition of vaginal infections. J Infect Dis 1996; 174:105863. First citation in article
22. Vallor AC, Antonio MA, Hawes SE, Hillier SL. Factors associated with acquisition of, or persistent colonization by, vaginal lactobacilli: role of hydrogen peroxide production. J Infect Dis 2001; 184:14316. First citation in article