The Vaginal Microbiome and Reproductive Health in Adolescents and Adults.

A literature search was conducted on March 22, 2024, using PubMed, Web of Science and Embase for articles published in peer-reviewed journals, in English, from inception ( Supplementary Figure 1 ). A PubMed search was repeated in September 2024 to identify more recent publications. A medical librarian assisted in developing and conducting the searches, which used a combination of controlled vocabulary and keywords to capture the main concepts of VM and adolescence. Variations of the following synonyms were also included in the search: vagina, microbiota, microbial communities, microbial ecology, microbial ecosystem, microbial environment, microbial interactions, microbial function, microecology, bacterial communities, flora, microflora, bacteriome, teen, preteen, youth, juvenile, pubescent, young women, minors, girl, schoolgirl, pediatric, paediatric, menarchal, perimenarchal, premenarchal, postmenarchal, and puberty. The first three pages of Google Scholar results were also searched as well as a selection of citing and cited works. Records were collected in EndNote and duplicates removed. Titles and abstracts were screened by 2 independent investigators (SMO and MC) to determine their relevance, and the final list was determined after discussion with NDS. Papers were included if they focused on healthy adolescent girls or adult women. Papers were excluded if they focused on pre-pubertal children, pregnancy, sex workers, disease states (e.g. vaginal infections), other microbiomes (e.g. skin, gut, oral), or animal models. Data were manually extracted and entered into an Excel file. Pre-defined extracted information included: article citation information (year of publication, author and title), study setting, subject demographics, study methodology (e.g., physician- or self-collected sampling, number of samples collected, reproductive hormone measurements, and sequencing method) and additional clinical information (e.g. sexual activity (active or inactive), hygiene practices, menstrual cycle regularity). Data quality and risk of bias was assessed by all investigators and a joint decision was made regarding study inclusion/exclusion.

The VM is defined as the complex microbial community that inhabits the vagina and exists in a symbiotic relationship with the host. The host provides oxygen, glucose, and other nutrients ( 20 ), and certain microbial profiles (e.g. Lactobacillus species predominant) are thought to promote vaginal health. In contrast to most healthy microbiomes of the body (e.g. the gut) that are characterized by great diversity, a healthy VM is generally thought to be one that is dominated by a single bacterial genus, Lactobacillus ( 21 ). Deviations from the normal Lactobacillus predominance have been associated with preterm birth, polycystic ovarian syndrome, dysmenorrhea, and increased susceptibility to Human papillomavirus (HPV), HIV, and cervical cancer ( 19 , 22 – 26 ), however, recent metagenomic studies suggest that Lactobacillus predominance may not be the only available “healthy” VM profile (see below). The VM is most often studied using vaginal swabs. Swabs can be collected in the clinic by a healthcare professional or self-collected. Importantly, self-collection of vaginal specimens has been shown to be highly acceptable to adolescent girls ( 27 , 28 ) and protocols have been developed to engage adolescents and young adults in VM sample self-collection ( 29 ). Microbial collection and stabilization kits are available that are specifically designed for vaginal samples (see, for example, OMNIgene•VAGINAL, DNAGenotek, Ottawa, Canada). DNA and RNA from microbiome samples can be extracted using commercially available kits (e.g. Qiagen blood and tissue kit) and sent for sequencing studies. Two of the most used sequencing methods in VM studies are: 1) 16S ribosomal RNA (rRNA) gene sequencing, and 2) whole-genome shotgun metagenomic analysis. 16S rRNA sequencing can only identify bacteria and archaea species because it targets and reads a region of the 16S rRNA gene that is only present within these microorganisms ( 30 ). Whole-genome shotgun metagenomic analysis can read all genomic DNA present in a sample and can therefore identify and profile bacteria, fungi, DNA viruses and other microorganisms simultaneously ( 31 ). Each of these methods has different strengths and weaknesses that must be considered depending on the research question being addressed ( 32 ). A Lactobacillus -dominated VM has traditionally been thought to indicate vaginal health ( 3 ). Lactobacilli, a gram-positive bacterium, ferments glycogen or glycogen breakdown products in the vagina, producing lactic acid. Lactic acid lowers vaginal pH which inhibits the growth of potential pathogens in the vagina ( 35 ); some species ( L. crispatus and L. gasseri ) also produce hydrogen peroxide ( 36 ) and bacteriocins (bactericidal proteins) that further protect against fungi and other harmful bacteria ( 21 , 35 ). Recent molecular studies, however, have demonstrated that the VM is more complex than previously thought. In a 16S RNA-based sequencing study of the VM in 144 asymptomatic Caucasian and Black women in North America, for example, Zhou et al found that 7% of Caucasian women and 33% of Black women did not demonstrate a Lactobacillus-dominant VM ( 37 ). A larger high-throughput sequencing study (n ≈ 400 adult women from North America representing four different ethnic groups) demonstrated that there are, in fact, five distinct VM profiles, called community state types (CSTs) ( 3 ). While most VMs (CSTs I, II, III, and V) were dominated by a single or multiple Lactobacillus species ( L. crispatus, L. gasseri, L. iners, and L. jensenii ), CST IV was a polymicrobial profile that included obligate anaerobic bacteria, such as, Gardnerella, Prevotella, Cryptobacterium, and Atopobium ( 3 ). CST-IV was more common in Black and Hispanic women ( 3 ). These ethnicity-based differences were recently confirmed by Fettweis et al. in an even larger VM study that included >1250 African American women and >400 women of European ancestry in North America ( 38 ) as well as in a smaller study (n=146) in young Black South African women demonstrating that the majority of women had cervicovaginal microbiota with low Lactobacillus abundance ( 39 ). The factors underlying these ethnicity-based differences remain to be determined but may be driven by cultural, lifestyle, and/or genetic factors ( 3 , 39 , 40 ). Importantly, these data have called into question the idea that Lactobacillus dominance equates to a “healthy” VM in all women. Studies across the menstrual cycle in adult women have consistently shown a Lactobacillus predominant VM ( 41 – 44 ). However, a recent metagenomic sequencing study by Hugerth et al. that included daily vaginal swab sampling in 49 Swedish women (average of 26 samples per participant) suggests that the VM is actually highly dynamic during the menstrual cycle; four so-called Vaginal Community Dynamics (VCDs) were identified, including constant eubiotic (defined as CST-I and CST-V; n=20/49=40.8%), constant dysbiotic (defined as CST-III and CST-IV, n=6/49=12.2%), menses-related dysbiotic (n=11, 22.4%), and unstable (n=12, 24.4%). Similar to previous studies, these data demonstrated that there is greater bacterial richness and diversity during the menstrual phase. This shift may occur because menstrual blood (which has a pH of 7.2–7.4) increases iron availability and vaginal pH; iron is an essential nutrient for many bacteria in the genital tract (e.g. Gardnerella vaginalis ) ( 45 ) and a neutral pH creates an ideal microenvironment for anaerobic commensal microbes ( 46 ). It is also possible that VM shifts during menses are related to use of menstrual hygiene products or to the decline in estrogen and progesterone levels. Additional work is needed in this area as only one study ( 43 ) (see also Table 1 ) has measured serum estrogen and progesterone levels in relation to VM samples in healthy adult women. Lifestyle factors have a significant impact on vaginal health ( Figure 1 ). Vegetarians exhibit higher microbial diversity than non-vegetarians ( 42 ). Several studies have suggested that higher levels of vitamins A, C, and E, and β-carotene may reduce the risk of bacterial vaginosis (BV) and that higher concentrations of zinc may lower the risk of Human Papilloma Virus (HPV) ( 47 ). Fettweis et al. identified a positive correlation between alcohol consumption and BV-associated bacteria whereas yogurt consumption was protective ( 38 ). High-intensity exercise ( 42 ), BMI ( 38 , 48 ), number of sexual partners and/or unprotected sexual intercourse ( 38 , 40 , 44 ), smoking status ( 38 , 49 ), and psychological stress ( 50 , 51 ) have also been shown to influence the VM composition. Hygiene practices can either support or disrupt the VM. Reusable products such as the menstrual cup stabilize the composition of the VM during menses and have also been shown to lower the risk of BV ( 41 , 52 , 53 ). Absorbent products such as sanitary pads and tampons have not been shown to alter the VM ( 54 , 55 ). Douching increases the risk for BV ( 56 ). Importantly, other vaginal hygiene products such as wipes, sprays, washes, and powders that are often marketed to adolescents and young women under the premise that they provide a “fresh, clean, and odorless” vagina may be detrimental to vaginal health ( 57 , 58 ). Studies in adult women using oral hormonal contraceptives have generally shown that high levels of lactobacillus are maintained ( 42 , 59 , 60 ), whereas one study found that progestin-only locally released contraceptives were associated with lower lactobacillus abundance ( 42 ). In a study of 164 reproductive-aged Chinese women, Ma et al. found that consistent condom use was associated with greater vaginal colonization with lactobacilli than was IUD use ( 10 ). Condoms are thought to be protective by preventing exposure to semen, which is alkaline (pH 7.0 to 8.0), and thereby maintaining an acidic vaginal pH that favors the lactobacilli population ( 10 ). The effects of IUDs and the vaginal ring on the VM have produced mixed results ( 61 – 63 ), with some studies showing no effects on the VM and others suggesting that copper IUDs can increase the risk of BV ( 64 ) and that hormonal IUDs may decrease lactobacillus abundance ( 42 , 65 ). For an in-depth discussion on the effects of contraception on the VM, we refer the reader to a recent mini review by Bakus et al. 2022 ( 66 ) (see also Figure 1 ). Antibiotics, antifungals, and vaginal antiseptics tend to disrupt vaginal composition by inhibiting the growth of lactobacilli and causing transitions between CSTs ( 8 , 67 ), and antibiotics are known to predispose to Candida vulvovaginitis ( 68 , 69 ). Intravaginal or oral probiotics have been shown to counteract vaginal dysbiosis and restore VM homeostasis ( 70 – 72 ). Few studies have investigated how the VM changes during puberty and early adolescence. In pre-pubertal girls, the vaginal microbiota is comprised primarily of anaerobic and aerobic rods and cocci with very few lactobacilli ( 73 ). It has been hypothesized that rising estrogen levels during mid-puberty facilitate the growth of lactobacilli, leading to a VM pattern that begins to resemble that of adults. In a 2-year longitudinal study of the VM in 49 adolescent girls in rural Uganda that used morphological-based (Nugent Gram-stain criteria) classification, there was a significant increase in Gram-positive rods (presumably Lactobacillus spp .) and a decrease in Gram-negative or variable rods (presumably Gardnerella vaginalis or Bacteroides species) over time even in pre-menarchal girls, whereas the VM composition was relatively stable during the peri- and postmenarchal period ( 74 ). Similarly, a longitudinal study of 31 peri-menarcheal girls (aged 10–12 yrs, Indiana, USA) that used 16S rRNA sequencing found that lactic acid-producing bacteria, primarily Lactobacillus spp. , dominated the VM in early to mid-puberty ( 12 ). Another 16S rRNA-based sequencing study of the VM in 90 girls (aged 13–18 years old, from the US and Canada, with regular menstrual cycles) identified four major clusters; three were dominated by Lactobacillus species and one was dominated by several lactic acid-producing, anaerobic bacteria such as Atopobium vaginae and Streptococcus ( 75 ). Taken together, these data suggest that the composition of the post-menarchal VM is quite similar to that of adult women, however, longitudinal studies are necessary in girls from early puberty through early adulthood to confirm these data. There have been very few studies to investigate the effect of internal and external factors on the VM in adolescent girls; most have focused on the effects of sexual activity or contraception. A microbial culture-based study of a limited number of bacterial species found no effect of oral contraceptives compared with barrier methods or no contraception on the VM of 171 healthy adolescent girls attending a University of California teen clinic ( 76 ). Sexually active subjects, however, had significantly more vaginal Gardnerella vaginalis, Lactobacillus, Mycoplasma, and Ureaplasma urealyticum compared with non-sexually active subjects in analyses controlling for age, age at sexual debut, and ethnicity ( 76 ). More recent studies among adolescent girls and young adult women in the USA ( 77 ), Australia ( 78 ), Belgium ( 79 ), and Tanzania ( 80 ) have similarly reported increased rates of colonization with G. vaginalis following sexual debut. Regarding contraception, a study of sexually active adolescent girls in the Washington DC metro area (n=59) that utilized 16S rRNA gene sequencing demonstrated that 3-months of progestin-based contraceptive use had no effect on the VM composition nor on vaginal secretions of immune biomarkers ( 81 ).

The VM is a complex and dynamic microbial community that plays an important role in vaginal health and disease. The majority of VM studies have focused on adult women. Thus, there is a critical need to conduct longitudinal studies using next generation sequencing methods to understand the composition of the normal VM during the pubertal transition as well as the potential influence of the adolescent VM on reproductive health in adulthood. Further studies are also required to understand the internal and external factors that influence the VM during adolescence, particularly the role of sex steroids, including estrogen, progesterone, and testosterone, and those factors contributing to racial differences. Adolescent self-sampling of the VM could be used clinically, for example, if a VM signature exists for regular, ovulatory cycles, PCOS (as already suggested by some studies ( 23 , 82 )), or endometriosis ( 83 ). The study by Hugerth et al. of vaginal community dynamics during the menstrual cycle in adult women is promising in that it suggests that clinicians would only need two to three mid-cycle samples and two samples during menses to classify the VM ( 44 ).

The vaginal microbiome (VM) is defined as the dynamic micro-ecosystem of bacteria, fungi, and viruses that inhabit the vagina. In recent years, the VM and its relationship to women’s health has gained increasing attention, and this has led to the formation of two major consortia to understand and promote vaginal health ( 1 , 2 ). While the VM is known to be dynamic across the lifespan and even across the menstrual cycle, most women exhibit communities dominated by Lactobacillus species. This species has been shown to maintain a healthy microbiome by lowering vaginal pH and by producing protective factors (such as hydrogen peroxides, lactic acid, and bacteriocins) against potential pathogens ( 3 – 7 ). The dynamic nature of the VM is related to a number of internal and external factors such as hormonal milieu (menstrual cycle phase, puberty, pregnancy, menopause, and contraception), ethnicity, age, diet, exercise, sexual activity, and antibiotics/antifungals ( Figure 1 ) ( 3 , 8 – 13 ). Fluctuations in such factors can either promote a healthy microbiome or lead to dysbiosis. Vaginal dysbiosis, an imbalance in the bacteria comprising the VM (generally an increase in diversity with loss of the typical Lactobacillus spp. predominance), has been tied to adverse gynecologic and obstetric outcomes, such as premature birth, sexually transmitted infections, pelvic inflammatory disease, cervical cancer, and dysmenorrhea ( 14 – 19 ). Prior research has largely focused on the VM of adult women rather than that of adolescent girls. Thus, little is known about the compositional nature of the adolescent VM and how it changes during the course of puberty. Research that has included adolescent populations has largely sought to understand the relationship between sexually transmitted infections (HIV, Chlamydia) or bacterial vaginosis [BV]) and the VM in high-risk populations as opposed to investigating the VM in healthy girls. Thus, it is unclear if the adult and adolescent VMs are similar or distinct, and if there is in fact a change in composition over time, what factors may mediate this change, what factors may interfere with this normal transition, and how dysbiosis during adolescence may impact future reproductive outcomes. This narrative review aims to provide an overview of the existing literature on the composition of the VM, particularly among adolescent girls and young adult women, to understand the internal and external factors that may alter its composition, and to propose how the VM might be utilized as an additional marker of reproductive health in clinical practice.

Supplementary Figure 1 Flow chart of the study. *Studies were excluded if they focused solely on children, sex workers, disease states, other microbiomes (e.g. skin, gut, oral), or animal models.