The postmenopausal vaginal microbiome and genitourinary syndrome of menopause

Abstract

This review summarizes our current understanding of associations of the postmenopausal vaginal microbiome with genitourinary syndrome of menopause. We review the normal postmenopausal microbiota, examine the association of the microbiome with vulvovaginal symptoms, describe microbial communities associated with physical and laboratory findings, and report the impact of different treatments for GSM on microbiota and symptom improvement. Postmenopausal vaginal symptoms have an underlying pathophysiology that has not been fully elucidated. Estrogen treatment may not be sufficient to relieve symptoms of vaginal discomfort in all postmenopausal individuals. In addition, other interventions targeted at changing the microbiota or pH do not consistently improve symptom severity.

Introduction

The advent of molecular analysis of the microbiome, and the initial publications from the Human Microbiome Project in 2012¹ have expanded our understanding of the importance of commensal bacterial communities in the function of the human body. In the vagina, culture-based and gram-stain based analyses have long demonstrated consistent and contrasting patterns of microbial colonization in pre- vs. post-menopausal women. Next generation sequencing of the vaginal microbial community has expanded our understanding of its components, however many questions about its role in postmenopausal women remain.

Genitourinary syndrome of menopause (GSM) is the term used to describe numerous different signs and symptoms that can occur in postmenopausal women, which are associated in many cases with significant discomfort and bother and which are attributed to a decrease in systemic estrogen. Although all people have a reduction in circulating estrogen with menopause, only approximately 50% of postmenopausal women report GSM.² Patient-reported symptom severity does not reliably correlate with clinical exam, vaginal microbiota, or other objective findings such as vaginal pH and vaginal maturation index.^3-6 While it is well established that decreased serum estrogen contributes to GSM, the underlying mechanisms that lead to postmenopausal vaginal discomfort have not been fully elucidated.

Our current understanding of associations of the postmenopausal vaginal microbiome with GSM are examined herein. An examination of the urinary microbiome and urinary symptoms is outside the scope of this review. We will review the differences between pre- and postmenopausal microbiota, examine the association of the postmenopausal vaginal microbiome with bothersome vulvovaginal symptoms, and describe associations of microbial communities with physical findings (such as provider assessment of dryness and atrophy), and laboratory-based findings (pH, vaginal maturation index [VMI], and metabolome). Additionally, the impact of different treatments for GSM on microbiota, and the correlation with symptom improvement will be examined.

The vaginal microbiome before and after menopause

For decades, vaginal microbial communities were described either by culture or by using Nugent scoring of bacterial morphotypes on gram-stained slides of vaginal fluid, which gives three categories: Normal (mostly Lactobacillus morphotypes), Bacterial Vaginosis [BV] (No lactobacilli, and many Gardnerella or Mobiluncus morphotypes) or Intermediate (a mixture).⁷ Molecular sequencing studies of the vaginal microbiome among premenopausal individuals described five community types or “community state types (CSTs)” of bacteria in nearly 400 asymptomatic, premenopausal individuals.⁸ Four of the five CSTs (CST I, II, III, V) are Lactobacillus-dominant with low microbial diversity, and one (CST IV) is characterized by the lowest proportion of Lactobacillus species and the greatest diversity of bacterial taxa with higher proportions of anaerobic organisms.

Culture based studies, and evaluation by Nugent score demonstrated a lower detection of Lactobacillus in postmenopausal women not on systemic menopausal hormone therapy (MHT) compared to premenopausal women.^9-11 Sequencing based analyses have demonstrated the same patterns using CST classifications or simply Lactobacillus dominance by relative abundance.^3,4,12,13 However, between 20% and 50% of postmenopausal individuals, depending on the population studied and the prevalence of symptoms, will still have a Lactobacillus-dominant microbiome.^{3,4,6,9,10,13-15} MHT has a clear and well-established impact on the post-menopausal microbiota. Estrogen therapy (both systemic, with or without progestin, and local) increases the abundance of Lactobacillus species in the postmenopausal vaginal microbiome (Figure 1).^4,9,10,15

Figure 1:

The proportion of people with Lactobacillus dominance measured by molecular (solid bars) or gram-stain of vaginal fluid (cross-line pattern) in premenopausal women or those on MHT (darker bars) vs. postmenopausal people not on MHT (lighter bars).

In studies of premenopausal cohorts, Black women are less likely to have a CST I (L. crispatus dominant) community than White women. Additionally Black and Hispanic populations have a higher prevalence of bacterial vaginosis.^8,16 Race and ethnicity are social constructs that likely reflect unmeasured social stressors or exposures. In studies of postmenopausal individuals, vaginal microbial community composition has rarely been analyzed by race and ethnicity, and even in studies that do so, numbers of individuals studied are small.^3,17

GSM symptoms and the postmenopausal vaginal microbiota

We do not understand the true underlying mechanisms for why some people are troubled by GSM and others are not. The decrease in circulating estrogen happens for all women after menopause, but not all experience bothersome symptoms. It would be reasonable to hypothesize that the composition of the vaginal microbiome is related to symptoms, however few studies have tested that hypothesis.

The FDA outcome criteria suggested for treatment studies in GSM include pH, the vaginal maturation index (a scoring system for morphology of vaginal epithelial cells) and the patient’s most bothersome symptom (MBS). However, there are few studies that have assessed presence and severity of reported symptoms relative to microbial vaginal community composition.

Most studies in this area focus on vulvovaginal symptoms, usually vaginal dryness, rather than urinary symptoms. A small early study using molecular sequencing reported an association between a non-Lactobacillus-dominant vaginal community and examiner-rated vaginal dryness, but association with patient-reported dryness severity, though collected, was not analyzed.¹³ In a study including 375 postmenopausal women, self-reported vaginal dryness was associated with CST assignment and Shannon diversity: women with higher diversity and CST-IV were more likely to report dryness.¹⁴

A cross-sectional analysis of 88 individuals enrolled in one of the NIH funded Menopause Strategies: Finding Lasting Answers for Symptoms and Health (MsFLASH) clinical trials,¹⁸ that assessed therapies for hot flashes (75% postmenopausal and 25% perimenopausal) demonstrated little association between microbiota and the presence or severity of vulvovaginal symptoms at baseline among 47 individuals with symptoms and 41 without.³ In a secondary analysis, Lactobacillus dominance was more common in individuals whose vulvovaginal symptoms improved during the trial, regardless of intervention arm.¹⁹

One of the 5 main MsFLASH trials, the MsFLASH vaginal health trial, assessed interventions for moderate to severe bothersome vulvovaginal symptoms among 302 postmenopausal individuals with GSM.²⁰ All participants had at least one moderate to severe symptom and the majority had more than one symptom, with dryness and dyspareunia being the most common. In a subset analysis of 144 people, no consistent bacterial taxa were found to be associated with specific symptoms. Participants with no change or minimal changes in symptom severity were just as likely as those with maximum improvement to have a Lactobacillus-dominant microbiota after study intervention.⁶ Although there were not differences in the microbiota between people with the most vs. the least symptom improvement, and there were no differences in the magnitude of symptom improvement between treatment arms, there was a significant change in the microbial community composition only in the estrogen arm – demonstrating the disconnect between microbial community composition and symptom severity.

Physical findings of GSM (vulvovaginal atrophy) and association with the vaginal microbiota

The physical signs of GSM, also sometimes called vulvovaginal atrophy (VVA) include vaginal pallor, dryness, rugosity, blanching, friability, and petechiae. In a study of 30 pre-, 29 peri- and 28 postmenopausal individuals not using hormones, individuals without signs of VVA (n=62, a mixture of pre, peri and postmenopausal) were compared to those with VVA (n= 19, 16 of whom were postmenopausal). The vaginal microbiota of individuals with mild or moderate signs of VVA had 25-fold greater odds of being classified with the lowest abundance of Lactobacillus and highest diversity.¹² It is important to note that self-reported vaginal symptoms were not assessed in this study, and VVA without symptoms is physiologic among menopausal individuals with low estrogen levels.

Similarly, in a study of 67 postmenopausal Chinese women, those with symptoms and exam findings of VVA had significantly lower abundance of Lactobacillus species than asymptomatic women with no VVA on exam, however overall diversity was not different between the two groups.¹⁵ Hummelen et al (2011) examined microbiota in 32 postmenopausal individuals and found that those with the lowest scores for vaginal dryness (on exam) had the highest relative abundance of Lactobacillus.¹³

Vaginal maturation index (VMI) is a laboratory scoring of epithelial cell morphotypes, and is lower when the vaginal epithelium is thin and circulating estrogen is low. A higher VMI (reflective of more superficial cells and a thicker vaginal epithelium) has been associated with higher relative abundance of Lactobacillus spp.²¹ Vaginal pH reliably increases as serum estradiol levels decrease in menopause. Lactobacillus species produce lactic acid, which may play an antimicrobial role to protect against pathogen colonization.

As noted above, many studies have focused on clinical exam findings and their association with the vaginal microbiome. However, VVA is a normal physical finding among menopausal individuals, and more recent studies have shown that symptoms correlate poorly with exam findings. Some clinicians and researchers consider the physiologic vaginal changes during perimenopause and postmenopause, including VVA with loss of elasticity of vaginal tissue and reduced lubrication, as unhealthy or abnormal even though many menopausal individuals with VVA do not have GSM. Interventions aim to restore the menopausal vaginal environment to what is considered the “normal” premenopausal state. It is important that future studies focus on symptom improvement and not attempt to normalize the vaginal microbiota to a premenopausal state.

The impact of hormone therapy on the vaginal microbiome

The use of MHT profoundly affects the postmenopausal vaginal microbiome.^22-24 A study by Gliniewicz et al. (2019) showed that the vaginal microbiota of postmenopausal individuals using MHT was more likely to resemble that of premenopausal individuals than that of postmenopausal individuals without MHT.²³ These findings emphasize the role of estrogen in defining the composition of the postmenopausal vaginal microbiota.

A few RCTs that demonstrate a change in vaginal microbiota with oral or vaginal estrogen, oral ospemifene, oral tibolone (not a comprehensive list) are shown in Table 1. In a study of 60 postmenopausal women treated with oral conjugated equine estrogen 0.3 mg daily, after 4 weeks of treatment the relative abundance of Gardnerella, Prevotella and Fannyhessea (previously known as Atopobium) decreased and Lactobacillus increased.¹⁵ A comparison of people with VVA being treated with oral estrogen and the selective estrogen receptor modulator ospemifene showed that both had higher relative abundance of Lactobacillus than untreated participants.²¹ Lastly, although not an RCT, a comparison of people on tibolone therapy vs. no treatment demonstrated that people on tibolone had a higher relative abundance of Lactobacillus.²⁵

Table 1:

Change in vaginal microbiota with hormonal or non-hormonal interventions

Reference	N	Population	Intervention	Duration	Findings
Petricevic 2008	35	Austria	Oral probiotic (L. rhamnosus/L. reuteri)	14 days	63% improved to Nugent 0-3 (p < 0.05)
	37	Austria	Placebo	14 days	16% improved to Nugent 0-3
Jaisamrarn	44	Thailand	Vaginal estriol + L. acidophilus	12 days	Proportion with Grade 1 score (similar to Nugent 0-3) changed from 0% to 41% (p < 0.05)
	44	Thailand	Placebo	12 days	Proportion with Grade 1 score changed from 2.3% to 4.7%
Bisanz 2014	14	Canada	Crossover trial oral probiotic (L. rhamnosus/L. reuteri) vs. placebo	3 days	At 14 days, 20% in probiotic arm improved to Nugent score 0-3 vs. 0% in placebo
Shen 2016	29	China	Oral estradiol	4 weeks	Mean relative abundance of Lactobacillus increased from 0.11 ± 0.16 to 0.71 ± 0.38 (p < 0.0001)
Mitchell 2017	8	US	Oral estradiol	8 weeks	1/3 (33%) people who started with non-Lactobacillus dominant community gained Lactobacillus dominance
	13	US	Placebo	8 weeks	1/6 (17%) people who started with non-Lactobacillus dominant community gained Lactobacillus dominance
Ribeiro 2019	19	Brazil	Oral estradiol	16 weeks	Proportion with Grade 1 score (similar to Nugent 0-3) increased from 26.3% to 94.7% (p < 0.05)
	20	Brazil	Oral probiotic	16 weeks	Proportion with Grade 1 score (similar to Nugent 0-3) decreased from 44% to 29%
Srinivasan 2022	45	US	Vaginal estradiol tablet	12 weeks	Proportion with Lactobacillus dominance increased from 32% to 80% (p < 0.01)
	48	US	Vaginal moisturizer (pH = 3)	12 weeks	Proportion with Lactobacillus dominance increased from 31% to 34% (p > 0.05)
	51	US	Placebo gel (pH =3)	12 weeks	Proportion with Lactobacillus dominance increased from 24% to 26% (p > 0.05)
Alvisi 2023	20	Italy	Ospemifene	12 weeks	Mean relative abundance of Lactobacillus increased from 35% to 37% (p > 0.05)
	19	Italy	Systemic estradiol	12 weeks	Mean relative abundance of Lactobacillus increased from 23% to 24% (p > 0.05)

Topical vaginal estrogen is also associated with a change in microbiome composition. The MsFLASH trial demonstrated a significantly greater impact of vaginal estradiol on the vaginal microbiota, metabolome, vaginal maturation index and pH than low-pH vaginal moisturizer or low-pH placebo gel. The effect of vaginal estradiol was greatest in individuals with a high-diversity vaginal microbial community and low VMI at study entry.⁴

Vaginal pH, microbiota and GSM

Despite the mixed results of studies evaluating an association between vaginal pH and vulvovaginal symptoms, change in pH is one of three FDA-recommended primary outcomes for treatment trials in GSM. A significant decrease in vaginal pH has been seen in trials of vaginal estrogen,^26,27 vaginal DHEA,²⁸ vaginal laser,²⁹ hyaluronic acid, and vaginal gels containing lactic acid.^30-32 In pre-menopausal people, vaginal pH is primarily driven by the abundance of Lactobacillus.³³

Analyses from the MsFLASH vaginal health trial (comparing 10 μg vaginal estradiol tablet plus placebo gel vs placebo tablet plus vaginal moisturizer vs dual placebo) showed that a decrease in vaginal pH due to application of low pH vaginal gels is insufficient to change the microbiota in postmenopausal individuals. Additionally, the metabolites in vaginal fluid did not shift significantly with low-pH interventions, even among individuals with high-diversity microbial communities, suggesting that lowering pH alone may not alter underlying microbial or host metabolic processes.⁴ Although participants in all treatment groups had significant decreases in vaginal pH over 12 weeks, participants in the estradiol group had the largest decrease, associated with changes in the composition of the vaginal microbiota and metabolites. Additionally, vaginal estrogen contributed to an additional reduction in pH even among individuals with a low-diversity, Lactobacillus-dominant community, suggesting that estradiol facilitates an increase in metabolic activity of lactic acid–producers, such as Lactobacillus and Bifidobacteria spp.

Other studies have shown a concordance between improvement in symptoms and decrease in vaginal pH.^13,34,35 In the REJOICE trial, a randomized trial of intravaginal estrogen vs placebo, change in pH was strongly correlated with changes in the appearance of the vaginal mucosa and weakly correlated with severity of symptoms at the end of the trial, although the analysis was not adjusted for treatment arm.³⁵ Another study evaluating an oral estradiol formulation found that reduced vaginal pH was more highly associated with improved sexual function scores than with objective changes in the vaginal mucosa.³⁴ In the MsFLASH vaginal health trial, though individuals randomized to the estradiol arm had significantly lower pH than those in the other arms after 12 weeks of treatment, individuals with maximal changes in vulvovaginal symptom severity were not more likely to have a lower pH than individuals with no or minimal change in symptom severity. Thus, change in pH was correlated with the estrogen treatment arm but pH did not always decrease as symptoms improved.⁶

A significant decrease in pH over the course of a trial may not reflect the same underlying biological processes across different interventions, and thus, lowering pH should not be a primary goal. The metric of lowering vaginal pH per FDA guidance for primary efficacy analyses in clinical trials in GSM is a blunt measure.³⁶

Probiotics in postmenopausal individuals

Probiotics targeted for vaginal health are widely available as over-the-counter oral dietary supplements or vaginal capsules or suppositories. In vaginal applications, probiotics are applied directly to the site of action, whereas orally supplemented probiotics pass through the gastrointestinal tract; both application routes may be efficacious.^37,38 Some hypothesize that orally administered probiotics may provide additional beneficial effects to vaginal health by first benefiting the gut microbiota and thereby minimizing the ascension of urogenital pathogens from the rectum to vaginal tract.³⁹

There are relatively few studies focusing on perimenopausal or postmenopausal individuals and the use of probiotics for vaginal health (Table 1). A small study of people not using MHT, administered an oral product containing L. rhamnosus and L. reuteri vs. placebo for 14 days, showed a greater drop in Nugent score (i.e. increase in Lactobacillus morphotypes) in the active arm, though this was only measured during use of product.⁴⁰ A case-series of vaginal delivery of L. rhamnosus and L. reuteri for 3 days did not demonstrate any impact on Nugent scores in women not on MHT.⁴¹ A vaginal tablet containing 0.03 mg estriol and L. acidophilus, administered daily for 12 days was compared to placebo in 87 postmenopausal women with GSM and a VMI < 40%. VMI improved and symptoms improved over the course of treatment, but there was no control group with non-hormonal probiotic treatment, thus the effect cannot be attributed to the probiotic.⁴²

Few studies assess change in GSM symptoms after use of vaginal probiotics. In an open label randomized trial, 60 postmenopausal women were randomized to systemic MHT (1mg estradiol/0.5mg norethisterone acetate) vs. oral isoflavone vs. oral isoflavone and an oral probiotic containing Lactobacillus acidophilus, Lactobacillus casei, Lactococcus lactis, Bifidobacterium bifidum, and Bifidobacterium lactis. There was a significant reduction in severity of vaginal dryness in the MHT group, but not the other two groups.⁴³

Other microbial causes of symptoms in postmenopausal individuals

GSM can only be diagnosed once other causes for symptoms have been ruled out. Postmenopausal individuals with genitourinary symptoms should be evaluated for STIs. Clinicians may erroneously assume that older women are not sexually active or are in a mutually monogamous relationship. All individuals, regardless of menopausal status, should be queried regarding sexual history and risk factors for STIs so that appropriate testing and treatment can be provided. Vulvovaginal candidiasis should also be considered, and may have an increased prevalence in people receiving MHT.

BV, while highly prevalent among premenopausal individuals reporting abnormal vaginal discharge, can be particularly challenging to diagnose among postmenopausal individuals. A recent meta-analysis found that the prevalence of BV among postmenopausal women ranged from 2.0 to 57%, with a summary estimate of 16.9%.⁴⁴ The standard diagnostic criteria for BV are typically valid only for premenopausal individuals.¹⁰ Amsel’s criteria, commonly used in clinical practice, requires at least three of the four criteria: presence of non-inflammatory vaginal discharge (typically thin, homogeneous, and non-adherent), amine odor after addition of 10% potassium hydroxide to a sample of vaginal discharge, clue cells on saline microscopy, and vaginal pH > 4.5. In postmenopausal people, vaginal pH is normally above 4.5, in part because Lactobacillus relative and total abundance is often lower than in premenopausal people. Commonly used G. vaginalis-based diagnostic tests for BV may be less appropriate in postmenopausal people, though this has not been tested. More recently approved PCR-based tests often include detection of Lactobacillus in their algorithm for diagnosis of BV, and so these also may be less appropriate for postmenopausal people.

While significantly less common than BV and VVC, desquamative inflammatory vaginitis (DIV) is another possible etiology for genitourinary symptoms. Symptoms may be identical to those seen in BV, VVC, and GSM. Exam findings include erythema and petechiae, as well as elevated pH, which can be normal in the postmenopausal vagina. DIV is characterized by increased inflammatory cells and parabasal epithelial cells in saline microscopy of vaginal discharge. If the diagnosis is suspected, treatment can be initiated with vaginal clindamycin or corticosteroids.⁴⁵

Summary and Conclusions

A recent meta-analysis concluded that treatment with vaginal estrogen was associated with greater improvement in physical findings of GSM than treatment with nonhormonal therapies, but not necessarily greater improvements in vulvovaginal symptoms ²⁶. FDA guidance includes two measures of physical findings (pH and VMI) as recommended outcomes for treatment trials for GSM, and one symptom measure. Existing data suggest that estrogen treatment causes a significant change in the vaginal microbiome, pH and VMI, but that these changes are not necessarily causally linked to improvements in symptoms of GSM. This is most evident in an in-depth analysis performed by MsFLASH investigators, which demonstrated that estradiol was associated with significant, profound changes in microbiota, small molecule metabolites, VMI, and vaginal pH,⁴ but that these changes occurred in the same magnitude among women whose symptoms improved vs. those whose symptoms did not.¹⁹

In aggregate, most studies demonstrate an association with low serum estrogen concentrations, vaginal microbiota, and vaginal microenvironment; however, it appears that vaginal symptoms have an additional underlying pathophysiology that has yet to be fully elucidated. For clinicians, this means that vaginal estrogen treatment alone may not be sufficient to relieve symptoms of vaginal discomfort in all postmenopausal individuals. This suggests that interventions targeted at changing the postmenopausal microbiota or pH are unlikely to alter symptom severity for many individuals, particularly those with dyspareunia, the predominant symptom noted among those in the MsFLASH vaginal health trial. Conversely, nonhormonal therapies and lubricants targeting various symptoms may be sufficiently effective treatment for GSM, even though they do not alter the microbiome and do not act via a change in vaginal estrogen. Future studies should focus on understanding the underlying pathophysiology of symptoms, and the non-GSM benefits of the vaginal microbial composition for postmenopausal vaginal health.