Secondary Logo

Journal Logo

Hormonal contraception decreases bacterial vaginosis but oral contraception may increase candidiasis: implications for HIV transmission

van de Wijgert, Janneke H.H.M.a,b; Verwijs, Marijn C.b,c; Turner, Abigail Norrisd; Morrison, Charles S.e

doi: 10.1097/QAD.0b013e32836290b6
Epidemiology and Social
Free

Objective: A 2012 WHO consultation concluded that combined oral contraception (COC) does not increase HIV acquisition in women, but the evidence for depot medroxyprogesterone acetate (DMPA) is conflicting. We evaluated the effect of COC and DMPA use on the vaginal microbiome because current evidence suggests that any deviation from a ‘healthy’ vaginal microbiome increases women's susceptibility to HIV.

Methods: We conducted a systematic review and reanalysed the Hormonal Contraception and HIV Acquisition (HC-HIV) study. Vaginal microbiome outcomes included bacterial vaginosis by Nugent scoring, vaginal candidiasis by culture or KOH wet mount and microbiome compositions as characterized by molecular techniques.

Results: Our review of 36 eligible studies found that COC and DMPA use reduce bacterial vaginosis by 10–20 and 18–30%, respectively. The HC-HIV data showed that COC and DMPA use also reduce intermediate microbiota (Nugent score of 4–6) by 11% each. In contrast, COC use (but not DMPA use) may increase vaginal candidiasis. Molecular vaginal microbiome studies (n = 4) confirm that high oestrogen levels favour a vaginal microbiome composition dominated by ‘healthy’ Lactobacillus species; the effects of progesterone are less clear and not well studied.

Conclusion: DMPA use does not increase HIV risk by increasing bacterial vaginosis or vaginal candidiasis. COC use may predispose for vaginal candidiasis, but is not believed to be associated with increased HIV acquisition. However, the potential role of Candida species, and vaginal microbiome imbalances other than bacterial vaginosis or Candida species, in HIV transmission cannot yet be ruled out. Further in-depth molecular studies are needed.

aInstitute of Infection and Global Health, University of Liverpool, Liverpool, UK

bAmsterdam Institute for Global Health and Development (AIGHD) and Department of Global Health, Academic Medical Center, Amsterdam

cDepartment of Medical Microbiology, University Medical Center Utrecht, Utrecht, The Netherlands

dDivision of Infectious Diseases, The Ohio State University, Columbus, Ohio

eClinical Sciences, Durham, North Carolina, USA.

Correspondence to Janneke H.H.M. van de Wijgert, Department of Clinical Infection, Microbiology and Immunology, Institute of Infection and Global Health, University of Liverpool, Ronald Ross Building, West Derby Street, Liverpool L69 7BE, UK. Tel: +44 151 795 9613; fax: +44 151 795 5529; e-mail: j.vandewijgert@liv.ac.uk

Received 19 March, 2013

Revised 22 April, 2013

Accepted 24 April, 2013

Back to Top | Article Outline

Introduction

Hormonal contraception might increase HIV acquisition risk. A recent systematic review of 20 prospective studies concluded that combined oral contraception (COC) does not increase risk, but that the evidence for progestin-only injectables is conflicting [1]. In 2009, 145 million women worldwide were using hormonal contraception, most commonly COCs or progestin-only injectables, such as depot medroxyprogesterone acetate (DMPA) or norethisterone enanthate (NET-EN) [2]. Progestin-only injectables are the most frequently used form of contraception in sub-Saharan Africa, where the HIV burden is highest [2]. Several biological mechanisms have been postulated, but it is unclear which (if any) are clinically relevant. The proposed mechanisms include increased area of cervical ectopy, vaginal epithelial thinning, local immune activation, and changes in the composition of the vaginal microbiome [3,4].

Several prospective cohort studies have shown significant associations between bacterial vaginosis by Amsel [5] or Nugent criteria [6] and subsequent HIV acquisition in women, with adjusted hazard ratios (aHRs) ranging from 1.4 to 2.3 [7]. A meta-analysis of individual participant data of 13 African prospective cohort studies showed that intermediate vaginal microbiota (Nugent score 4–6) and bacterial vaginosis (Nugent score 7–10) were each associated with HIV acquisition in multivariable models (aHR 1.54 and 1.69, respectively) [8]. In addition, four studies have shown a positive association between vaginal candidiasis and HIV acquisition, with adjusted relative risks (aRRs) ranging from 1.8 to 3.3 [7]. Taken together, current evidence suggests that any deviation from a normal lactobacilli-dominated vaginal microbiome increases women's susceptibility to HIV. In the last decade, new molecular techniques have become available that can characterize the composition of the vaginal microbiome in much more detail (such as next-generation sequencing and DNA microarrays) [9], but to date, these have mostly been used in exploratory studies.

We conducted a systematic review of the published literature and reanalysed one of our own studies, to investigate whether hormonal contraception increases the risk of bacterial vaginosis and/or vaginal candidiasis. Both of these conditions could be on the causal pathway between hormonal contraception and HIV acquisition. We also reviewed exploratory studies describing the impact of hormonal contraception on the vaginal microbiome as characterized by molecular methods.

Back to Top | Article Outline

Materials and methods

Our aim was to assess whether women using hormonal contraception had a higher prevalence or incidence of bacterial vaginosis (assessed by Gram stain Nugent scoring or molecular methods) and/or vaginal candidiasis (all Candida species; assessed by culture or KOH wet mount), than women not using hormonal contraception. We could not reliably distinguish vaginal Candida colonization from symptomatic vaginal candidiasis and refer to these outcomes as (vaginal) candidiasis throughout this manuscript. We considered the exposure group inappropriate if hormonal methods and copper intrauterine devices (IUDs) were grouped together, but not when different types of hormonal methods were grouped together. For many studies, we could not be sure that ‘oral contraceptive (OC) users’ included progestin-only pill (POP) users or not. Throughout the manuscript, these study groups are indicated with OC instead of COC. However, in most populations, POPs are used by only a small percentage of pill users, and we therefore analysed the OC and COC studies together [10]. We considered the control group inappropriate if the entire group consisted of users of another contraceptive method (e.g. copper IUD users), but not if users of other methods or pregnant women made up a minority proportion of the larger control group (e.g. copper IUD users combined with women not using any contraception).

We conducted a systematic review according to the PRISMA 2009 guidelines [11]. Two researchers (M.C.V. and J.H.H.M.vdW.) designed the search terms (Fig. 1) and selection criteria (using predefined data extraction tables instead of a protocol), and independently selected and classified the articles. We searched the Medline (US National Library of Medicine, National Institutes of Health, Bethesda, Maryland, USA) and Embase (Elsevier B.V., Amsterdam, The Netherlands) databases for articles up to 1 July 2012, using the search terms in Fig. 1, and limiting our search to articles written in English. These searches yielded 1871 articles (883 in Medline and 988 in Embase), including 194 duplicates (Fig. 2). Fourteen additional articles were identified from reference lists. The title and abstract of all 1691 articles were read, and the following articles were eliminated: articles that included an intervention that might interfere with our research question, focused exclusively on the changes of the vaginal microbiome during pregnancy or the peri/postmenopausal period or did not address our research question (n = 1105) and articles published before 1990 (n = 472). A total of 114 articles were deemed of interest. Two of these could not be retrieved. After reading the full text of the remaining 112 articles, a further 86 were rejected because they did not address our research question appropriately (most commonly because bacterial vaginosis was assessed by Amsel criteria, Pap smear cytology or clinical signs and symptoms only; vaginal candidiasis was assessed by Pap smear cytology or clinical signs and symptoms only; there was no appropriate control group; or because the measure of effect was not and could not be calculated due to insufficient data presented in the article). A total of 36 articles were selected and categorized by clinical outcome: bacterial vaginosis only (n = 11), candidiasis only (n = 12), bacterial vaginosis and candidiasis as separate outcomes (n = 9) and the vaginal microbiome using molecular methods (n = 4).

Fig. 1

Fig. 1

Fig. 2

Fig. 2

We scored the quality of each study as follows: the exposure group consisted of only one hormonal method (1 point); the control group did not include any hormonal contraceptive users or consisted of the same women prior to exposure (1 point); the control group did not include any pregnant women (1 point); the study did not target symptomatic women (1 point); the study was prospective (1 point); and the measure of effect was adjusted in multivariable analysis (1 point). When studies had the same score, we ordered them in the tables from largest to smallest sample size.

When the number of cases and controls in each exposure group were presented in an article but no measure of effect was given, we calculated a crude odds ratio (OR), 95% confidence interval (CI) and two-sided P value by Fisher's exact test using STATA version 12 (StataCorp, College Station, Texas, USA).

We also reanalysed data from the Hormonal Contraception and HIV Acquisition (HC-HIV) Study. The HC-HIV study was implemented in Zimbabwe, Uganda and Thailand between 1999 and 2004 [12]. The HC-HIV study methodology, including bacterial vaginosis and candidiasis assessments, was described in a related article by van de Wijgert et al.[13]. The study was found to be well positioned to answer our research questions due to its study design (participants were selected based on their contraceptive method use), large study size (6109 women), long duration of follow-up (15–24 months per woman) and close attention paid to contraceptive method switches and pregnancies during follow-up. The HC-HIV study was approved by the ethics committees of all eight collaborating institutions in the United States, Uganda, Zimbabwe and Thailand [12]. All participants provided written informed consent.

In the current HC-HIV analysis, we used data from all three study countries and compared bacterial vaginosis (Nugent score 7–10) and intermediate microbiota (Nugent score 4–6) with normal microbiota (Nugent score 0–3) in separate multivariable binomial regression models with generalized estimating equations (GEEs). The same methodology was used to evaluate the vaginal candidiasis outcome. Contraceptive method switches during the study were taken into account and pregnant person-time was excluded. The following variables were considered for inclusion in the multivariable models (starting with the full model but removing variables that confounded the effect of hormonal contraception on the outcome of interest by less than 10%): study country, age, education, parity, living with a steady partner, current breastfeeding, any vaginal practices in previous interval, condom use in previous interval (consistent, inconsistent or none), age at first sex, sexual risk taking by the woman (composite variable), sexual risk taking by male partners (composite variable), self-reported lifetime history of sexually transmitted disease, any use of antimicrobials in previous interval and current HIV status (all women were HIV-negative at enrolment but 213 women acquired HIV infection during follow-up).

Back to Top | Article Outline

Results

Hormonal contraception and risk of bacterial vaginosis and intermediate microbiota by Nugent scoring

Eleven out of the 19 selected studies had a sample size larger than 200 (Table 1) [13–23]. Seven of these studies provided a measure of effect for COCs and/or DMPA separately or provided sufficient data for us to calculate the measure of effect [13–17,20,22]. Four studies combined all hormonal contraceptive methods into one exposure group [18,19,21,23]. Out of the six studies that evaluated COCs as a separate exposure group, five found a statistically significant reduction of bacterial vaginosis [13,14,17,20,22] and one found no effect [16]. All four studies that evaluated DMPA as a separate exposure group found a significant reduction of bacterial vaginosis [13–15,17]. One study also evaluated NET-EN as a separate exposure group and found a reduction in bacterial vaginosis in that group as well [15]. Two out of the four studies that grouped all hormonal methods into one exposure group found a reduction of bacterial vaginosis [18,19] and the other two studies found no association [21,23].

Table 1-a

Table 1-a

Table 1-b

Table 1-b

Another eight studies reporting on the association between hormonal contraception and bacterial vaginosis by Nugent scoring had a sample size below 200 participants [24–31]. Most of these showed no statistically significant results [24–27,29–31], but two showed a trend towards a reduction in bacterial vaginosis prevalence for COC use [25,27] and none showed a trend towards an increase in prevalence for either COC or DMPA use. A small prospective study [28] conducted in Mombasa, Kenya, including only 10 OC users and 43 DMPA users, found no statistically significant association between OC use and bacterial vaginosis but did find a significant reduction in bacterial vaginosis prevalence among DMPA users.

The three studies with a maximum quality score of 6 (Table 1) showed consistent results, reporting adjusted measures of effect for the association with bacterial vaginosis by Nugent scoring ranging from 0.80 to 0.90 for COC use and from 0.70 to 0.82 for DMPA use [13–15]. All associations were statistically significant. The HC-HIV study also evaluated the effects of COC and DMPA use on intermediate microbiota by Nugent scoring and found aRRs of 0.89 for each (COC: 95% CI 0.84–0.96, P = 0.002; DMPA: 95% CI 0.83–0.95; P < 0.001).

One large Italian study (n = 27 172) using stored electronic microbiology records deserves mention, even though this study evaluated cultures of Gram-negative and Gram-positive bacteria rather than Nugent scores for bacterial vaginosis diagnosis [32]. Women who were using COCs were less likely to have a positive culture for Gram-negative (aOR = 0.75; 95% CI 0.57–0.99) as well as Gram-positive bacteria (aOR = 0.86; 95% CI 0.68–1.09) than women who were not using any hormonal contraception, although the latter did not reach statistical significance.

Seven studies in Table 1, including all three studies with a maximum quality score of 6, recruited women on the basis of their contraceptive choices or HIV status, and not because they sought care for vaginal symptoms [13–16,18,20,21]. The other four studies recruited from gynaecology clinics [17,19,22,23], but only one study targeted women reporting vaginal symptoms [22]. None of the studies in Table 1 except for one [18] included postmenopausal women, but four studies did not exclude pregnant women [18,20–23]. All except for three studies in Table 1[13,16,17] compared a Nugent score of 7–10 (bacterial vaginosis) with a Nugent score of 0–6 (which includes both intermediate and normal microbiota). Our analysis of the HC-HIV study is the only one providing measures of effect for intermediate microbiota separately. Four articles explicitly mentioned that the microscopist who conducted the Nugent scoring was blinded to the exposure status of the participant [13,17,18,20] and three articles explicitly mentioned that external quality control of Nugent scoring was implemented [13,15,17]. All except for one study [23] presented adjusted measures of effect, but it was not always clear which adjustments were made and whether use of antibiotics or other antimicrobial drugs was taken into account.

Back to Top | Article Outline

Hormonal contraception and risk of vaginal candidiasis

Thirteen of the 21 selected studies had a sample size larger than 200 (Table 2): the presence of Candida species was assessed by culture in nine studies [32–40] and by KOH wet mount in four studies [13,14,21,41]. Only one study combined COC and injectables into one exposure group [21]; this study found no statistically significant effect (Table 2). All other studies evaluated COCs as the exposure group, and four studies also evaluated DMPA or injectables as a separate exposure group [13,14,33,35]. Of the 12 studies that evaluated COCs as a separate exposure group, seven found a statistically significant increase in vaginal candidiasis prevalence or incidence [14,34–36,38–40], two found a positive association that was close to significance [32,37], two found no association [33,41] and one found a statistically significant reduction of candidiasis by KOH wet mount [13]. Of the four studies that evaluated DMPA as a separate exposure group, one found a statistically significant increase in candidiasis [33], one found a statistically significant reduction [13] and two found no association [14,35].

Table 2-a

Table 2-a

Table 2-b

Table 2-b

Another eight studies reporting on the association between hormonal contraception and culture-positive candidiasis had a sample size below 200 participants [24,29–31,42–45]. Only one of these showed statistically significant results, reporting a significant increase in culture-positive candidiasis for COC use [24].

The three studies with a maximum quality score of 6 showed inconsistent results [13,14,33]. The one study using culture-positive candidiasis as the outcome found no association with COC use and an increased risk with DMPA use [33]. Of the two studies with Candida on KOH wet mount as the outcome, one found reduced risks in COC and DMPA users [13] and the other one an increased risk in COC users and no effect in DMPA users [14].

The majority of the studies in Table 2 were clinic-based, including five studies that only included symptomatic women [34,37–40]. All except one study with candidiasis by culture as the outcome [33] were cross-sectional or case–control studies, whereas three out of the four studies using Candida on KOH wet mount as the outcome were prospective cohort studies [13,14,41]. In contrast to the bacterial vaginosis studies presented in Table 1, the majority of the candidiasis studies in Table 2 did not present adjusted measures of effect. None of the articles except for one [41] described quality control procedures for the Candida assessments.

Our search yielded only four studies investigating the relationship between female sex hormones (but not hormonal contraceptives) and the vaginal microbiome characterized by molecular methods [46–49]. Two of these studies investigated the vaginal microbiome of women at various time points during an in-vitro fertilization (IVF) procedure [46,47] and two at various time points during the menstrual cycle [48,49]. The studies showed that high oestrogen levels favour a vaginal microbiome dominated by ‘healthy’ Lactobacillus species, but that data on progesterone levels are limited and not clear (Table 3). Three additional longitudinal menstrual cycle studies showed that the Lactobacillus crispatus load declines about 100-fold during menses, whereas the bacterial loads of Lactobacillus iners, Gardnerella vaginalis and Atopobium vaginae increase, thereby making women more vulnerable for bacterial vaginosis during and after menses [50–52].

Table 3

Table 3

Back to Top | Article Outline

Discussion

We conclude that COCs and DMPA (and most likely also NET-EN) reduce the prevalence and incidence of bacterial vaginosis in women, and that COCs (but not DMPA) may increase the prevalence of vaginal candidiasis. This is the first comprehensive review of these relationships, and our findings have not been widely recognized before. Therefore, the hypothesis that DMPA use may increase HIV acquisition risk by increasing bacterial vaginosis or vaginal candidiasis risk is not supported by the evidence. However, the potential role of candidiasis in HIV transmission among hormonal contraceptive users, and vaginal microbiome imbalances other than bacterial vaginosis or candidiasis, cannot yet be ruled out. Now that molecular methods have become more accessible, they should be incorporated into larger epidemiological studies to investigate these relationships more fully.

The evidence for a reduction of bacterial vaginosis risk in hormonal contraceptive users is much stronger than the evidence for a potential increased candidiasis risk in COC users. First, the results of the large bacterial vaginosis studies (with at least 200 participants) are consistent, and none of the smaller studies report a trend in the opposite direction. The three bacterial vaginosis studies with the highest quality score report bacterial vaginosis risk reductions of 10–20% for COCs and 18–30% for DMPA. The data on COCs and vaginal candidiasis are less consistent and in some cases even contradictory, and limited data are available for DMPA. Second, the results of the candidiasis studies were more difficult to interpret because it was not possible to differentiate between high and low levels of Candida species in the vagina, the quality of the candidiasis studies (in the context of our research questions) was low and Candida species were not assessed in the molecular studies. The mean quality score of the candidiasis studies with a sample size of at least 200 participants was 3.3, compared with 4.2 for the bacterial vaginosis studies. Candidiasis studies, compared with bacterial vaginosis studies, were less often prospective (31 vs. 55%), more often targeted symptomatic women (38 vs. 8%), less often presented adjusted measures of effect (38 vs. 91%) and less often described at least one aspect of laboratory quality control procedures (8 vs. 45%). It should be noted, however, that most prospective studies did not describe how contraceptive switches and repeat incident infections were handled, and that it was not always clear what adjustments were made in multivariable analyses.

The outcome of the bacterial vaginosis studies was more robust than the outcomes of the candidiasis studies, because the Nugent score is well standardized and takes quantities of three bacterial morphotypes into account (albeit crudely, by counting number of cells in microscopic fields). However, eight of the 11 bacterial vaginosis studies compared bacterial vaginosis (defined as a Nugent score of 7–10) with ‘no bacterial vaginosis’ (Nugent score 0–6), thereby including women with intermediate microbiota (Nugent score 4–6) in the negative outcome group. This most likely diluted the protective effect of hormonal contraception on bacterial vaginosis because molecular studies show that vaginal microbiome compositions associated with Nugent scores of 4–6 are more closely related to bacterial vaginosis than to a normal lactobacilli-dominated microbiome composition [53]. Our own analysis of the HC-HIV study data showed that COC and DMPA use was negatively associated with bacterial vaginosis as well as intermediate microbiota, and that the measures of effect were of similar magnitudes.

The main strength of the work described in this article is that we investigated the research questions in a comprehensive manner, using multiple assessments of the vaginal microbiome. We did not limit ourselves to bacterial vaginosis, but also included candidiasis and vaginal microbiome assessments using modern molecular methods. The most important limitation of the systematic review was that hardly any of the eligible studies were designed to investigate our research questions. We therefore had to work with different study designs, outcomes (see previous paragraph), exposure groups, control groups and analysis strategies. This is also the reason why we decided not to do meta-analyses. Although we found that the HC-HIV study was one of the best positioned studies to answer our research questions, that study had limitations too, most notably the fact that we did not assess candidiasis by culture but by KOH wet mount. In general, studies with large sample sizes and multiple follow-up visits tend to not use cultures for diagnosis due to the time-sensitive and labour-intensive nature of this procedure. Other limitations of our work include that we may have missed eligible studies, and that many of the eligible studies cannot firmly establish temporal sequences or rule out (residual) confounding.

Several hypotheses have been postulated to explain why oestrogen may inhibit bacterial vaginosis associated micro-organisms. The dominant hypothesis is that oestrogen increases the levels of available glycogen in epithelial cells, which facilitates lactobacilli growth and lactic acid production, thus lowering the vaginal pH [54]. Glycogen also provides a favourable environment for the growth of other micro-organisms, including Candida albicans[55]. However, these hypotheses do not explain why DMPA users and postmenopausal women, who have much lower levels of oestrogen, are also protected against bacterial vaginosis. It has been hypothesized that this may be due to a lack of menses [56]. As mentioned above, women are more vulnerable for bacterial vaginosis during and after menses due to lower L. crispatus loads and higher loads of bacterial vaginosis associated bacteria [50–52].

The effects of DMPA on HIV acquisition may be mediated by a high progesterone level, a low oestrogen level or the induction of a hypoestrogenic state [57]. Potential mechanisms include effects on physiological properties of the cervix (cervical ectopy) and vaginal wall (epithelial thinning or atrophy, reduced cornification of epithelial cells, increased availability of target cells for HIV) and immunomodulatory effects [3,4,58]. High-quality studies are needed to further investigate the role of these processes in HIV transmission.

In the last decade, the second-generation progestin levenorgestrel has become popular in family planning, and a recent report of the Initiative for Multipurpose Prevention Technologies endorsed the use of levenorgestrel in products that are being developed to prevent both HIV/sexually transmitted infections (STIs) and pregnancy [59]. The effect of levenorgestrel on the vaginal microbiome, cervix and vaginal wall and genital mucosal immune system therefore also needs to be studied as a matter of priority. In the meantime, women at risk of HIV using hormonal contraception, particularly progestin-only injectables, should be advised to also use condoms.

Back to Top | Article Outline

Acknowledgements

The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript. The content of this publication does not necessarily reflect their views or policies, nor does mention of trade names, commercial products or organizations imply endorsement by the US Government.

The authors thank the HC-HIV Study team for implementation of the HC-HIV study between 1999 and 2004, Pai-Lien Chen for providing input in the reanalysis of the HC-HIV data and Hanneke Borgdorff and Marc Bonten for critical review of the manuscript.

J.H.H.M.vdW. designed the study. A.N.T., J.H.H.M.vdW., C.S.M. analysed the HC-HIV data. M.C.V., J.H.H.M.vd.W. conducted the systematic review. J.H.H.M.vdW., M.C.V. wrote the first draft of the article. A.N.T., C.S.M. contributed to the writing of the article. J.H.H.M.vdW., M.C.V., A.N.T., C.S.M. read and met the ICMJE criteria for authorship. J.H.H.M.vdW., M.C.V., A.N.T., C.S.M. agreed with the manuscript's results and conclusions.

Ethics review was not required for the systematic review. The HC-HIV study was approved by the institutional review boards and/or ethics committees of all eight collaborating institutions in the United States, Uganda, Zimbabwe and Thailand. The approval letters are available upon request. All study participants provided written informed consent.

The HC-HIV study was funded with U.S. federal funds from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Department of Health and Human Services through a contract with Family Health International (contract N01-HD-0-3310).

Back to Top | Article Outline

Conflicts of interest

The authors have declared that no competing interests exist.

Back to Top | Article Outline

References

1. World Health Organization, Department of Reproductive Health and Research (2012). Hormonal contraception and HIV. Technical statement. http://http://www.who.int/reproductivehealth/publications/family_planning/rhr_12_8/en/index.html. [Accessed 5 February 2013].
2. United Nations, Department of Economic and Social Affairs, Population Division (2011). World contraceptive use 2011. http://http://www.un.org/esa/population/publications/contraceptive2011/contraceptive2011.htm. [Accessed 5 February 2013].
3. Blish CA, Baeten JR. Hormonal contraception and HIV-1 transmission. Am J Reprod Immunol 2011; 65:302–307.
4. Hel Z, Stringer E, Mestecky J. Sex steroid hormones, hormonal contraception, and immunobiology of human immunodeficiency virus-1 infection. Endocr Rev 2012; 31:79–97.
5. Amsel R, Totten PA, Spiegel CA, Chen KCS, Eschenbach D, Holmes KK. Nonspecific vaginitis: diagnostic criteria and microbial and epidemiologic associations. Am J Med 1983; 74:14–22.
6. 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.
7. Hayes R, Watson-Jones D, Celum C, van de Wijgert J, Wasserheit J. Treatment of sexually transmitted infections for HIV prevention: end of the road or new beginning?. AIDS 2010; 24 (Suppl 4):S15–S26.
8. Low N, Chersich MF, Schmidlin K, Egger M, Francis SC, van de Wijgert, et al. Intravaginal practices, bacterial vaginosis and HIV infection in women: individual person data meta-analysis. PloS Med 2011; 8:e1000416.
9. Fredricks DN, Fiedler TL, Marrazzo JM. Molecular identification of bacteria associated with bacterial vaginosis. N Engl J Med 2005; 353:1899–1911.
10. McCann MF, Potter LS. Progestin-only contraception: a comprehensive review. Contraception 1994; 50 (6 Suppl 1):S1–S195.
11. Moher D, Liberati A, Tetzlaff J, Altman DG. The PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009; 6:e1000097.
12. Morrison CS, Richardson BA, Mmiro F, Chipato T, Celentano DD, Luoto J, et al. Hormonal Contraception and the Risk of HIV Acquisition (HC-HIV) Study Group. Hormonal contraception and the risk of HIV acquisition. AIDS 2007; 21:85–95.
13. van de Wijgert JHHM, Morrison CS, Cornelisse PGA, Munjoma M, Moncada J, Awio P, et al. Bacterial vaginosis and vaginal yeast, but not vaginal cleansing, increase HIV-1 acquisition in African women. J Acquir Immune Defic Syndr 2008; 48:203–210.
14. Baeten JM, Nyange PM, Richardson BA, Lavreys L, Chonan B, Martin HL Jr, et al. Hormonal contraception and risk of sexually transmitted disease acquisition: results from a prospective study. Am J Obstet Gynecol 2001; 185:380–385.
15. Pettifor A, Delany S, Kleinschmidt I, Miller WC, Atashili J, Rees H. Use of injectable progestin contraception and risk of STI among South African women. Contraception 2009; 80:555–560.
16. Gupta K, Hillier SL, Hooton TM, Roberts PL, Stamm WE. Effects of contraceptive method on vaginal microbial flora: a prospective evaluation. J Infect Dis 2000; 181:595–601.
17. Riggs M, Klebanoff M, Nansel T, Zhang J, Schwebke J, Andrews W. Longitudinal association between hormonal contraceptives and bacterial vaginosis in women of reproductive age. Sex Transm Dis 2007; 34:954–959.
18. Watts H, Springer G, Minkoff H, Hillier SL, Jacobson L, Moxley M, et al. The occurrence of vaginal infections among HIV-infected and high-risk HIV-uninfected women. J Acquir Immune Defic Syndr 2006; 43:161–168.
19. Holzman C, Leventhal JM, Qiu H, Jones NM, Wang J. Factors linked to bacterial vaginosis in nonpregnant women. Am J Public Health 2001; 91:1664–1670.
20. Koumans EH, Sternberg M, Bruce C, McQuillan G, Kendrick J, Sutton M, et al. The prevalence of bacterial vaginosis in the United States, 2001-2004: associations with symptoms, sexual behaviors, and reproductive health. Sex Transm Dis 2007; 37:864–869.
21. Lan PT, Lundborg CS, Phuc HD, Sihavong A, Unemo M, Chuc NT, et al. Reproductive tract infections including sexually transmitted infections: a population-based study of women of reproductive age in a rural district of Vietnam. Sex Transm Infect 2008; 84:126–132.
22. Bradshaw CS, Morton AN, Garland SM, Morris MB, Moss LM, Fairley CK. Higher-risk behavioural practices associated with bacterial vaginosis compared with vaginal candidiasis. Obstet Gynecol 2005; 106:105–114.
23. Schwebke J, Desmond RA, Oh MK. Predictors of bacterial vaginosis in adolescent women who douche. Sex Transm Dis 2004; 31:433–436.
24. Egbe CA, Onwufor UC, Omoregie R, Enabulele OI. Female reproductive tract infections among vaginal contraceptive users in Benin City, Nigeria. Genomic Med Biomarkers Health Sci 2011; 3:49–52.
25. Schwebke JR, Desmond R. Risk factors for bacterial vaginosis in women at high risk for sexually transmitted diseases. Sex Transm Dis 2005; 32:654–658.
26. Plitt SS, Garfein RS, Gaydos CA, Strathdee SA, Sherman SG, Taha T. Prevalence and correlates of Chlamydia trachomatis, Neisseria gonorrhoeae, Trichomonas vaginalis infections, and bacterial vaginosis among a cohort of young injection drug users in Baltimore, Maryland. Sex Transm Dis 2005; 32:446–453.
27. Cohen CR, Duerr A, Pruithithada N, Rugpao S, Hillier S, Garcia P, et al. Bacterial vaginosis and HIV seroprevalence among female commercial sex workers in Chiang Mai, Thailand. AIDS 1995; 9:1093–1097.
28. McClelland RS, Richardson BA, Graham SM, Masese LN, Gitau R, Lavreys L, et al. A prospective study of risk factors for bacterial vaginosis in HIV-1 seronegative African women. Sex Transm Dis 2008; 35:617–623.
29. Ocak S, Cetin M, Hakverdi S, Dolapcioglu K, Gungoren A, Hakverdi AU. Effects of intrauterine device and oral contraceptives on vaginal flora and epithelium. Saudi Med J 2007; 28:727–731.
30. Eschenbach DA, Patton DL, Meier A, Thwin SS, Aura J, Stapleton A, et al. Effects of oral contraceptive use on vaginal flora and vaginal epithelium. Contraception 2000; 62:107–112.
31. Miller L, Patton DL, Meier A, Thwin SS, Hooton TM, Eschenbach DA. DMPA-induced hypoestrogenism and changes in vaginal flora and epithelium. Obstet Gynecol 2000; 96:431–439.
32. Tibaldi C, Cappello N, Latino MA, Masuelli G, Marini S, Benedetto C. Vaginal and endocervical microorganisms in symptomatic and asymptomatic nonpregnant females: risk factors and rates of occurrence. Clin Microbiol Infect 2009; 15:670–679.
33. Beigi RH, Meyn LA, Moore DM, Krohn MA, Hillier SL. Vaginal yeast colonization in nonpregnant women: a longitudinal study. Obstet Gynecol 2004; 104:926–930.
34. Cetin M, Ocak S, Gungoren A, Hakverdi AU. Distribution of Candida species in women with vulvovaginal symptoms and their association with different ages and contraceptive methods. Scand J Infect Dis 2007; 39:584–588.
35. Enweani IB, Gugnani HC, Okobia R, Ojo SB. Effects of contraceptives on the prevalence of vaginal colonization with Candida species in Edo State, Nigeria. Rev Iberoam Micol 2001; 18:171–173.
36. Grigoriou O, Baka S, Makraris E, Hassiakos D, Kapparos G, Kouskouni E. Prevalence of clinical vaginal candidiasis in a university hospital and possible risk factors. Eur J Obstet Gynecol Reprod Biol 2006; 126:121–125.
37. Geiger AM, Foxman B. Risk factors for vulvovaginal candidiasis: a case-control study among university students. Epidemiology 1996; 7:182–187.
38. Hart G. Factors associated with trichomoniasis, candidiasis and bacterial vaginosis. Int J STD AIDS 1993; 4:21–25.
39. Ahmad A, Khan AU. Prevalence of Candida species and potential risk factors for vulvovaginal candidiasis in Aligarh, India. Eur J Obstet Gynecol Reprod Biol 2009; 144:68–71.
40. Linhares LM, Witkin SS, Miranda SD, Fonseca AM, Pinotti JA, Ledger WJ. Differentiation between women with vulvovaginal symptoms who are positive or negative for Candida species by culture. Infect Dis Obstet Gynecol 2001; 9:221–225.
41. Bauters TG, Dhont MA, Temmerman MI, Nelis HJ. Prevalence of vulvovaginal candidiasis and susceptibility to fluconazole in women. Am J Obstet Gynecol 2002; 187:569–574.
42. Guzel AB, Ilkit M, Akar T, Burgut R, Demir SC. Evaluation of risk factors in patients with vulvovaginal candidiasis and the value of chromID Candida agar versus CHROMagar Candida for recovery and presumptive identification vaginal yeast species. Med Mycol 2011; 49:16–25.
43. Barousse MM, van der Pol BJ, Fortenberry D, Orr D, Fidel PL Jr. Vaginal yeast colonization, prevalence of vaginitis, and associated local immunity in adolescents. Sex Transm Infect 2004; 80:48–53.
44. Blum M, Gilerovitch M, Benaim J, Appelbaum T. The correlation between Chlamydia antigen, antibody, vaginal colonization and contraceptive method in young unmarried women. Adv Contracept 1990; 6:41–45.
45. Ginter G, Soyer HP, Rieger E. Vaginal yeast colonization and promiscuity. A study of 197 prostitutes. Mycoses 1992; 35:177–180.
46. Hyman RW, Herndon CN, Jiang H. The dynamics of the vaginal microbiome during infertility therapy with in vitro fertilization-embryo transfer. J Assist Reprod Genet 2012; 29:105–115.
47. Jakobsson T, Forsum U. Changes in the predominant human lactobacillus flora during in vitro fertilization. Ann Clin Microbiol Antimicrob 2008; 7:14.
48. Gustafsson RJ, Ahrne S, Jeppsson B, Benoni C, Olsson C, Stjernquist M, et al. The Lactobacillus flora in vagina and rectum of fertile and postmenopausal healthy Swedish women. BMC Womens Health 2011; 11:17.
49. Gajer P, Brotman RM, Gyoyun B, Sakamoto J, Schütte UME, Zhong X, et al. Temporal dynamics of the human vaginal microbiota. Sci Transl Med 2012; 4:132ra52.
50. Srinivisan S, Liu C, Mitchell CM, Fiedler TL, Thomas KK, Agnew KJ, et al. Temporal variability of human vaginal bacteria and relationship with bacterial vaginosis. PLoS One 2010; 5:e10197.
51. Lopes dos Santos Santiago G, Tency I, Verstraelen H, Verhelst R, Trog M, Temmerman M, et al. Longitudinal qPCR study of the dynamics of L. crispatus, L. iners, A. vaginae, (sialidase-positive) G. vaginalis, and P. Bivia in the vagina. PLoS One 2012; 7:e45281.
52. Jespers V, Menten J, Smet H, Poradosú S, Abdellati S, Verhelst R, et al. Quantification of bacterial species of the vaginal microbiome in different groups of women, using nucleic acid amplification tests. BMC Microbiol 2012; 12:83.
53. Ravel J, Gajer P, Abdo Z, Schneider GM, Koenig SSK, McCulle SL, et al. Vaginal microbiome of reproductive-age women. Proc Natl Acad Sci U S A 2011; 108 (Suppl 1):4680–4687.
54. Eschenbach DA, Thwin SS, Patton DL, Hooton TM, Stapleton AE, Agnew K, et al. Influence of the normal menstrual cycle on vaginal tissue, discharge, and microflora. Clin Infect Dis 2000; 30:901–907.
55. Sobel JD, Faro S, Force RW, Foxman B, Ledger WJ, Nyirjesi PR, et al. Vulvovaginal candidiasis; epidemiologic, diagnostic and therapeutic consideration. Am J Obstet Gynecol 1998; 178:303–311.
56. Hummelen R, Macklaim JM, Bisanz JE, Hammond JA, McMillan A, Vongsa R, et al. Vaginal microbiome and epithelial gene array in postmenopausal women with moderate to severe dryness. PLoS One 2011; 6:e26602.
57. Mishell DR, Dharma KM, Thorneycroft IH, Nakamura RM. Estrogenic activity in women receiving an injectable progestogen for contraception. Am J Obstet Gynecol 1972; 113:372–376.
58. Huijbregts RPH, Helton ES, Michel KG, Sabaj S, Richter HE, Goepfert PA, et al. Hormonal contraception and HIV-1 infection: medroxyprogesterone acetate suppresses innate and adaptive immune mechanisms. Endocrinology 2013; 154:1282–1295.
59. McGrory E, Romano J, Young Holt B, Manning J, Hemmerling A. multipurpose prevention technologies for reproductive health. Executive Summary of the Product Prioritization Stakeholder Meeting; 26 October 2012; Washington, DC. http://http://www.cami-health.org/documents/2012-SAWG-Report-ExecutiveReport.pdf. [Accessed 14 February 2013].
Keywords:

bacterial vaginosis; combined oral contraception; depot medroxyprogesterone acetate; HIV; systematic review; vaginal candidiasis; vaginal microbiome; vaginal microbiota

© 2013 Lippincott Williams & Wilkins, Inc.