Effects of combined oral contraceptives, depot medroxyprogesterone acetate and the levonorgestrel-releasing intrauterine system on the vaginal microbiome

Abstract

Objectives

Prior studies suggest that the composition of the vaginal microbiome may positively or negatively affect susceptibility to sexually transmitted infections (STIs) and bacterial vaginosis (BV). Some female hormonal contraceptive methods also appear to positively or negatively influence STI transmission and BV. Therefore, changes in the vaginal microbiome that are associated with different contraceptive methods may explain, in part, effects on STI transmission and BV.

Study design

We performed a retrospective study of 16S rRNA gene survey data of vaginal samples from a subset of participants from the Human Vaginal Microbiome Project at Virginia Commonwealth University. The subset included 682 women who reported using a single form of birth control that was condoms, combined oral contraceptives (COCs), depot medroxyprogesterone acetate (DMPA) or the levonorgestrel-releasing intrauterine system (LNG-IUS).

Results

Women using COCs [adjusted odds ratio (aOR) 0.29, 95% confidence interval (CI) 0.13–0.64] and DMPA (aOR 0.34, 95% CI 0.13–0.89), but not LNG-IUS (aOR 1.55, 95% CI 0.72–3.35), were less likely to be colonized by BV-associated bacteria relative to women who used condoms. Women using COCs (aOR 1.94, 95% CI 1.25–3.02) were more likely to be colonized by beneficial H₂O₂-producing Lactobacillus species compared with women using condoms, while women using DMPA (aOR 1.09, 95% CI 0.63–1.86) and LNG-IUS (aOR 0.74, 95% CI 0.48–1.15) were not.

Conclusions

Use of COCs is significantly associated with increased vaginal colonization by healthy lactobacilli and reduced BV-associated taxa.

Implications

COC use may positively influence gynecologic health through an increase in healthy lactobacilli and a decrease in BV-associated bacterial taxa.

Introduction

The vaginal microbiome appears to play an important role in women's health, and in vitro evidence suggests that certain Lactobacillus species may help to protect the vagina from pathogens. For example, Lactobacillus crispatus inhibits infection of HeLa cells by Chlamydia trachomatis and inhibits growth of Neisseria gonorrhoeae in a porcine model of vaginal mucosa [1], [2]. In addition, a recent report suggests that L. crispatus alters cervicovaginal mucus in such a way that trapping of HIV particles is enhanced [3]. These lactobacilli produce lactic acid, which is bactericidal, and some produce bacteriocins and/or H₂O₂. There is some evidence that lactic acid is the major microbicidal component produced by vaginal lactobacilli and that H₂O₂ does not play an important role in killing bacteria in the anoxic conditions of the vagina [4], [5]. However, whether H₂O₂ production plays an important role in controlling bacterial growth or whether it is just a marker, there is strong evidence that the presence of lactobacilli with the capacity to produce H₂O₂ is associated with vaginal “eubiosis” or a healthy microbiome [6], [7].

Bacterial vaginosis (BV) is a dysbiosis of the vaginal microbiome characterized by a decrease in lactobacilli and an increase in anaerobic neutralophiles. BV predisposes women to acquisition of HIV and other STIs from their sexual contacts [8], [9], [10], [11]. There is an association between susceptibility to STIs and the abundance of protective lactobacilli (reviewed in [12]) but also to damaging effects of the BV-associated taxa on the vaginal epithelium. Gardnerella vaginalis produces the cholesterol-dependent cytolysin vaginolysin, which can lyse vaginal epithelial cells, and this could compromise the barrier effect of the vaginal mucosa. There is also evidence that sialidase and mucin-degrading enzymes produced by the BV-associated bacteria disrupt the protective mucus layer [13]. In addition to higher rates of HIV acquisition by women with BV, the rate of transmission of HIV from women with BV to their sexual partners is also higher. BV-associated bacteria enhance HIV replication, and higher viral loads are found in cervicovaginal secretions from women with BV, Candida vaginitis and genital herpes simplex virus (HSV) [10], [14], [15].

Estrogen induces the accumulation of glycogen in the vaginal epithelium, and glycogen positively influences colonization by healthy lactobacilli [16], [17]. When estrogen levels are high in the follicular phase of the menstrual cycle and during pregnancy, the abundance of lactobacilli tends to increase. It would follow that estrogen-containing hormonal contraceptives might influence the vaginal microbiome. In fact, a number of studies have shown reduced rates of BV in women using combined oral contraceptives (COCs) [18], [19], [20].

Results from studies of the effect of progestin-only contraceptives have been less clear. A longitudinal study of intrauterine systems that release the progestin levonorgestrel (LNG-IUS) found no influence of the systems on vaginal microbiomes, and another study in baboons supported this finding [21], [22]. Results from studies of changes in the vaginal microbiome in women using depot medroxyprogesterone acetate (DMPA) injections as a contraceptive method have been inconsistent. A systematic review of 36 prior studies found that BV rates were lower in women using DMPA [23], but another study found no effect of hormonal contraception on the composition of the vaginal microbiome [24]. Yet another report described a decrease in the prevalence of H₂O₂-producing lactobacilli after 1 year of DMPA use [25]. Prior studies have relied on conventional microbiological techniques or phylogenetic microarrays to identify vaginal microbes, but to date, comprehensive 16S rRNA gene survey data have not been reported for women using hormonal contraceptives. Herein, we performed a retrospective comparison of 16S gene profiles from women using condoms, COCs (excluding progestin-only mini-pills), LNG-IUS and DMPA to determine whether hormonal contraceptives affect vaginal microbial community profiles or the abundance of vaginal lactobacilli.