The main findings of the clinical trials evaluating intervention with fermented foods against BV and/or VVC are summarized in Table 1. A total of 6 clinical studies (17–22) were identified, two were published in the 90’s and the remaining were published between 2011 and 2017. Articles contained variations in clinical aspects (P/I/O), methodology and evaluation. Due to the high heterogeneity of the data, conducting a meta-analysis was deemed infeasible. The concept of a gut-vagina axis is relatively new and contributes to new insight in our understanding of the function and immunology of the female genital system (23). However, it is interesting to note that trials involving intervention with fermented foods in recent years (more than 5 y) have not been published. Of the 6 RCT trials (3 articles BV, 2 articles VVC and 1 both), all had positive outcomes for prevention or treatment of BV and/or VVC as observed in the primary and secondary outcomes listed in Table 1. All articles presented statements that supported effects were positive. However, the studies also had limitations to some extent. Some of the articles were performed on a limited number of participants (19, 20), applied intervention for a brief duration (17), published as pilot RCT’s (22), short communication (21) or brief report (19). The first three publications (19–21) did not report the specific starter cultures employed in the fermentation process, which likely included S. thermophilus and L. delbrueckii subsp. bulgaricus, as these are conventionally utilized in yogurt production. Additionally, the first two studies (19, 20) did not provide strain-level identification for L. acidophilus cultures. Shalev (19) investigating effects for both BV and VVC reported only beneficial effects of yogurt consumption against BV. On the other hand, Dols et al. (21) observed beneficial effects against BV for both the intervention (yogurt with L. rhamnosus GR-1) and control group (consuming standard yogurt). Hantoushzadeh et al. (17), found no significant differences between treatments when comparing the efficacies of medical treatment (clindamycin) versus yogurt consumption against BV, indicating ingestion of yogurt containing probiotic bacteria could be as effective as conventional medical intervention. To complement efficacy evaluations, several studies have employed established diagnostic tools to objectively assess changes in vaginal microbiota and clinical symptoms. The Nugent score is a Gram stain scoring system for vaginal swabs to diagnose BV and the Amsel criteria provide an alternative assessment for BV diagnosis based the presence of at least three of four findings: vaginal discharge, elevated vaginal pH, clue cells on microscopy, and a positive whiff test. Both Dols et al. (21) and the most recent study by Laue et al. (18) reported no significant change in the Nugent scores. However, Laue et al. (18) also confirmed that symptomatic relief was significant as indicated by the differences in Amsel criteria scores for intervention and control groups.

A total of 3 observational studies exploring the effect of fermented foods for BV/VVC outcomes were identified from the systematic search (Table 2). These studies were generally based on assessment of yogurt consumption patterns (among other factors) and vaginitis outcomes. All three papers reported positive effects related to yogurt consumption. To be more specific, Novikova and Mårdh (24) found that the VVC positive cohort had lower yogurt consumption pattern compared to their VVC negative and control cohorts. The use of certain antibiotics (for non-gynecologic intervention) can disrupt the balance of vaginal microbiota and lead to conditions such as yeast infections or bacterial vaginosis, namely post antibiotic vaginitis (PAV). Pirotta et al. (25) addressed the association between antibiotic use and vaginitis outcomes and reported that some women tend to self-medicate themselves by consuming yogurt and/or probiotic supplements containing lactobacilli. In this cross-sectional study approximately 40% of women resorted to this intervention for prevention and 43% for treatment. In the more recent study by Rosen et al. (26), it was implicated that higher consumption of low-fat dairy (a category of food that inherently may include fermented food such as yogurt, kefir etc.) could confer a healthier microbiome. The study also highlighted a significant scientific gap in understanding the mechanisms linking diet and vaginal microbiota composition.

The quality and risk of bias assessments (RoB 2.0 for RCTs and NOS for observational studies) are summarized in Figure 2. Among the six interventional studies assessed using the RoB 2.0 tool, only two were categorized as having a low overall risk of bias, while the remaining four exhibited high risk, most notably in domains such as deviations from intended interventions (n = 4), missing outcome data (n = 2), and measurement of outcomes (n = 2). Additionally, three studies showed concerns regarding selective reporting, randomization errors, or carryover effects. Moderate risks also arose from selective reporting (n = 4), errors in the randomization process (n = 3), and period or carryover effects in one crossover study. Furthermore, there was a high rate of participant drop-outs in two studies due to refusal to discontinue effective interventions (e.g., yoghurt) (20) or difficulties in adhering to study protocols (19). Observational studies, assessed via the NOS, scored lower overall: Rosen et al. (26) was rated as moderate quality (4/7 stars), while Novikova and Mårdh (24) and Pirotta et al. (25) were rated poor (3/9 and 2/7 stars, respectively), with consistent weaknesses in sample representativeness and lack of detailed comparator information.

These findings suggest that the evidence base is currently stronger for interventional studies, both in quantity (n = 6 vs. n = 3) and in methodological quality. For example, Laue et al. (18), one of the few low-risk RCTs, provided structured data with both Nugent and Amsel scoring systems, supporting a positive treatment effect. By contrast, the observational studies not only had limited statistical power but also relied heavily on self-reported data and lacked control for confounding factors. This undermines their capacity to establish causality and limits their robustness.

It is also important to acknowledge the heterogeneity in study designs and populations. Sample sizes varied widely (from <20 to 300), populations ranged from pregnant women to HIV-positive individuals, and treatment durations spanned from 1 week to several months. Such variability likely contributes to inconsistent findings and limits the generalizability of results. Furthermore, while all identified studies reported positive outcomes, no neutral or negative findings were captured. This may suggest that studies showing no/negative effect may remain underreported or unpublished. In any case, participant awareness of their assigned intervention was a major contributor to the high risk of bias identified in multiple clinical trials.

Taken together, the predominance of high-risk or poorly rated studies, small sample sizes, and absence of negative data highlight the need for cautious interpretation. While some RCTs, such as Laue et al. (18) and Dols et al. (21), present encouraging results with relatively low bias, the overall certainty of evidence remains moderate at best. Well-powered, rigorously designed trials with transparent reporting and inclusion of negative or neutral outcomes are urgently needed to better ascertain the true effect of fermented foods on vaginal health outcomes.

Yogurt is a fermented food produced by culturing certain types of dairy ingredients with a bacterial culture that includes Lactobacillus delbrueckii subsp. bulgaricus and Streptococcus thermophilus (27). These two species are not autochthonously present in the human gastrointestinal tract and are not inherently probiotic (28) although specific strains of these species with probiotic capacity have been identified (29). According to a joint report of the Food and Agriculture Organization and World Health Organization (30) probiotics are defined as ‘live microorganisms, which when administered in adequate amounts confer a health benefit on the host’. In order for a product to have probiotic quality it is expected that the probiotic strains retain viability throughout transit of the gastrointestinal tract. Therefore, although standard yogurt may not be probiotic, addition of probiotic cultures prior/post fermentation renders yogurt and other fermented milk products (i.e., milks fermented with starters other than yogurt cultures) carriers for probiotics. In all 9 (interventional and observational) human studies (17–22, 24–26), the food of interest was yogurt or a version thereof (yogurt drink) containing viable microbial load of lactobacilli. Intervention in all RCT studies involved the ingestion of yogurt fermented with standard yogurt culture (S. thermophilus and L. delbrueckii subsp. bulgaricus) and reformulated to include various lactic acid-producing bacteria. Fermented dairy drinks and yogurt are well known sources of lactic acid bacteria (LAB) and the presence of LAB strains are well documented. Indeed, dairy foods (specifically yogurt) are indicated as the carrier of choice for probiotic organisms (31–33) and it is thus reasonable that the clinical trials have explored effects of yogurt consumption. Although LAB are used extensively as culture organisms in dairy foods they are not exclusive to this category and many fermented foods such as fermented cereal drinks and vegetable/fruit juices can be produced through lactic fermentation (31). It is well known that many non-yogurt dairy foods and non-dairy foods may harbor LAB produced via autochthonous lactobacilli or with strains added at the start of fermentation. These may constitute part of the women’s diet. Interestingly, these have not been evaluated as confounders in human studies nor have they been investigated thus far for their efficacy in intervention against or prevention of BV/VVC. Nor have the standard cultures used for fermentation of yogurt in the RCT studies been accounted for or considered as a confounder.

The explanations on the mechanism of action centered on the identification of the active component followed by discussions related to how the active component results in the beneficial effects. Three domains were addressed; (1) the translocation mechanism of the probiotics, (2) the effects resulting from colonization of the gut, and (3) mechanisms related to temporal presence of the beneficial microbes in the vaginal econiche.

In all studies, the active component implicated in conferring the beneficial effect is the viable microbial load contained within fermented milks administered/ingested. Several mechanisms explaining how the microorganisms contained in the food matrices may impact the vaginal microenvironment have been proposed. The earliest identified mechanism was the translocation by route of anal contamination and the ascending of bacteria into the vagina (18, 34). This indicates to the rectal microbiota as a reservoir for colonization of the vaginal econiche, evidenced at the strain level for L. crispatus, L. gasseri, L. jensenii, and L. iners (35). Moreover, recent publications point to other mechanisms of translocation involving active transport of the bacterial cells. Miller refers to the hematogenous route of bacterial transfer from the gut (36) while other publications center on the cross-talk between the gut and women’s reproductive tract (37). Indeed, the translocation of lactobacilli from the gut of the nursing mother to the mammary glands and expression of the probiotic cells in mother’s milk (38) and the presence of probiotic DNA in the meconium (36) also suggest that there may be more complex, poorly understood translocation mechanisms involved. The involvement of immune system has received attention with possibilities of more directed cellular translocation (39, 40). It has also been proposed that IgA induced regulation for lactobacilli in the small intestine may promote colonization of these bacteria in the vagina (41). It should be stressed here that gut microbiota can also serve as an extravaginal reservoir of BV-associated bacteria (42), therefore, the properly balanced intestinal microbiota and healthy gut epithelium can help maintain a healthy vaginal environment.

The intestinal microbiota is modulated by ingested microorganisms and impacts the host immune system. As shown in the mouse model of Gardnerella vaginalis (GV)-induced BV, oral administration of L. rhamnosus HN001 and/or L. acidophilus GLa-14 more effectively activated innate and adaptive immunity compared to the intravaginal administration (43). Oral administration of lactobacilli more potently inhibited GV-induced myeloperoxidase activity, NF-κB activation, and TNF-α and IL-1β expression (involved in innate immunity), as well as inhibited GV-induced expression of RORγt, TNF-α, and IL-17 (involved in adaptive immunity). These results suggest that the anti-BV effect of orally administered probiotics may be due to its regulatory effects on immune responses through the gastrointestinal tract (43).

The newest proposed mechanism of action by which gut bacteria can beneficially influence the vaginal health involves extracellular vesicles (EV) released by bacteria. EV are small structures (below 300 nm) made of bilayer lipid membranes that cannot replicate themselves but carry a cargo of proteins, nucleic acids, and lipids. They play a key role in immune function, inflammatory reaction, and disease development by transporting active molecules to distant sites through the bloodstream (44). It has been suggested that EVs from commensal bacteria may have beneficial effects on the host by enhancing their mucosal tolerance and preventing disease progression, whereas EVs from pathogenic bacteria have proinflammatory effects on the host immune cells (44). While gut microbiota is restricted to the intestinal lumen, the secreted EVs can penetrate through the intestinal barriers, enter the systemic circulation, and affect both adjacent and distant organs (44). The potential of EVs in mediating lactobacilli beneficial effects was explored in in vitro studies in HeLa cervical cells model and showed that EVs from L. crispatus BC5 and L. gasseri BC12 (isolated from vagina of healthy women) significantly enhanced the cellular adhesion of other vaginal beneficial lactobacilli (45). The same EVs reduced the adhesion of pathogens: Escherichia coli, Staphylococcus aureus, S. agalactiae, and Enterococcus faecalis supporting the hypothesis that extracellular vesicles released by symbiotic lactobacilli may be implicated in sustaining a healthy vaginal homeostasis (45). Pili on the cell surface of Lacticaseibacillus rhamnosus GG (LGG) promotes adhesion to the mucosa and ensure close contact to host cells and emit EV carrying cargo of effector molecules. These molecules, including secreted proteins, surface-anchored proteins, polysaccharides, and lipoteichoic acids, which interact with host physiological processes have been identified and shown to stimulate epithelial cell survival and integrity, reduce oxidative stress, mitigate excessive mucosal inflammation, enhance IgA secretion, and provide long-term protection through epigenetic imprinting (46).

Temporal presence of lactobacilli in the vaginal epithelium can act protectively by competing with pathogens for nutrients and for adhesion sites at the surface of epithelial cells, producing of hydrogen peroxide, bacteriocins, and biosurfactants, along with organic acids (lactic acid, formic acid and other short chain fatty acids), which maintain the pH of the vagina too low for the growth of pathogens or by modulating local or regional immunological responses (47, 48). These postbiotic molecules could be considered effective against BV as well as VVC. On the other hand, Candida yeast morphogenesis and subsequent pathogenesis directed with quorum sensing activity may be disrupted with anti-film forming activity of probiotic enzymes (such as chitinases) and other postbiotics (49, 50). Intestinal colonization with bacteria can antagonize C. albicans by reshaping the metabolic environment, forcing metabolic adaptations that reduce fungal pathogenicity (51). Therefore, distinct mechanisms related to enrichment of lactobacilli in the vaginal microbiota can be beneficial for prevention and treatment of both BV and VVC.

S. thermophilus and L. delbrueckii spp. bulgaricus are generally recognized as safe (GRAS) microorganisms used in the production of yogurt and various dairy products (52). Some specific strains of these species have been studied for their probiotic efficacy (27) however some criteria need addressing for justification in using the term “probiotic” to describe such strains, e.g., the organism must be identified at the strain level and shown to express the relevant trait. Safety is a prerequisite for strains that have been identified as probiotic (53). Probiotic occurrence as normal commensals of the mammalian microbiota and their established safe use in diverse food and supplement products worldwide support their safety for oral consumption. Nevertheless, they are viable organisms, and therefore it is feasible that they could infect the host. Precaution is advised in the administration of probiotic organisms to some populations (i.e., immunocompromised patients) (54). Specific assessment of the probiotic strain provides a more in-depth understanding toward the safety of oral consumption. This has been demonstrated for individual strains such as L. crispatus CTV-05 (55) and L. rhamnosus HN001 (56). Furthermore, it is important to state that all species included in this review have received a qualified presumption of safety (QPS) status by the European Food Safety Authority (EFSA) (57). FDA regulations indicate the GRAS status of yogurt bacteria and specify granting of permission for the use of harmless lactic acid-producing bacteria, such as Lactobacillus acidophilus, as optional ingredients in specified standardized foods (58).