Preterm birth (PTB), occurring in approximately 10% of pregnancies worldwide, remains a major contributor to neonatal morbidity and mortality [1]. Despite advancements in medical and scientific research, the precise mechanisms underlying PTB are not yet fully understood. Recent research has increasingly focused on the vaginal microbiome, recognizing its potential influence on pregnancy outcomes [2]. The field of bacteriology has transitioned from traditional culture-based techniques to metagenomic analyses, providing deeper insights into microbial ecosystems [3]. Since the initiation of the human microbiome project, substantial progress has been made in characterizing microbial communities across various human body sites, including the vagina [4]. However, studies on the vaginal microbiome and its association with PTB have yielded inconsistent findings across different populations, with variations observed in research outcomes [5]. These discrepancies may arise from variations in study methodologies, gestational age at sample collection, host immune responses, genetic predisposition, and environmental factors [6]. Addressing these challenges through standardized research protocols with using standard materials, longitudinal studies, larger multi-ethnic cohort analyses, and multi-omics studies will be crucial for a more comprehensive understanding of the vaginal microbiome’s role in pregnancy [6]. Therefore, this review aims to synthesize current knowledge, identify gaps in existing research, and propose future directions to enhance our understanding of the vaginal microbiome’s impact on PTB risk.

The risk factors for PTB include ethnicity, socioeconomic status, maternal weight, smoking, periodontal disease, and underlying health conditions [7]. As the prevalence of advanced maternal age and comorbidities increases, so does the incidence of PTB [8,9]. sPTB accounts for approximately 70-75% of all PTB cases [1,10]. Notably, intra-amniotic infections are observed in one-third of sPTB cases, with microbial profiles similar to those found in the lower genital tract, suggesting an ascending infection mechanism [11]. Elevated levels of inflammatory cytokines in amniotic fluid, cervicovaginal fluid (CVF), and maternal blood further support this hypothesis [12-14]. Additionally, bacterial vaginosis (BV), characterized by reduced Lactobacillus dominance and increased anaerobic bacteria, has been identified as a significant risk factor for sPTB [15]. Advances in molecular diagnostic techniques, including 16S ribosomal RNA (rRNA) metagenomic sequencing, have enhanced our ability to detect diverse bacterial species in the vaginal microbiome, providing deeper insights into microbial contributions to PTB and potential biomarkers for early detection [3].

The vaginal microbiome undergoes dynamic changes throughout life, influenced by hormonal fluctuations, immune responses, sexual activity, and pregnancy [16]. During pregnancy, the vaginal microbiome typically stabilizes with reduced diversity, and Lactobacillus species including Lactobacillus crispatus (L. crispatus), Lactobacillus iners (L. iners), Lactobacillus jensenii (L. jensenii), and Lactobacillus gasseri (L. gasseri), become predominant [17]. These bacteria play a crucial role in maintaining a low vaginal pH through the secretion of lactic acid and hydrogen peroxide, which inhibit the growth of pathogenic microorganisms and help prevent infections [18,19].

Vaginal microbiota can be categorized into distinct community state types (CSTs) based on dominant bacterial species. CST I, II, III, and V are characterized by a dominance of different Lactobacillus species, whereas CST IV, associated with BV, consists of diverse anaerobes such as Gardnerella vaginalis (G. vaginalis), Prevotella (Prevotella spp.), and Atopobium vaginae (A. vaginae) [20]. An imbalance in this microbial composition-termed vaginal dysbiosis-has been linked to an increased risk of PTB [21-23]. A normal vaginal microbiome with the dominance of Lactobacillus species, maintains low microbial diversity and an acidic pH (approximately 3.5-4.5) by producing lactic acid. In contrast, depletion of Lactobacillus allows for an overgrowth of anaerobes such as G. vaginalis and Prevotella spp., which contribute to mucus degradation, biofilm formation, and inflammation, leading to elevated vaginal pH, production of pro-inflammatory metabolites, and increased susceptibility to infections and adverse pregnancy outcomes, including PTB [24].

The composition of the vaginal microbiome undergoes distinct changes across trimesters. In the first trimester, microbial diversity tends to be higher, with transitional microbial states observed. By the second trimester, the microbiome generally becomes more stable and Lactobacillus dominance increases, particularly L. crispatus and L. jensenii, which contribute to a protective environment through the production of lactic acid and hydrogen peroxide [25,26]. In the third trimester, the microbial community remains relatively stable in most cases, but dysbiotic shifts in late pregnancy have been linked to an increased risk of PTB [19,20,27-30]. Understanding trimester-specific microbial shifts and their influence on host immunity is crucial for identifying PTB biomarkers and developing preventive strategies. Late pregnancy dysbiosis, driven by microbial and immune interactions, has been linked to increased PTB risk.

Across different populations, studies have identified consistent patterns in the association between vaginal microbiota and PTB risk. Research on Asian cohorts, along with comparisons to African, American, and White populations, suggests that Lactobacillus-dominant microbiota generally confer protection against PTB, whereas anaerobe-dominant microbiota are linked to increased PTB risk. A shared finding across studies is that CST I, II, III, and V, characterized by dominance of L. crispatus, L. jensenii, and L. gasseri, are associated with term births, whereas CST IV, dominated by G. vaginalis, A. vaginae, and Prevotella spp., is correlated with a higher PTB risk (Fig. 1). These observations highlight the importance of maintaining microbial stability during pregnancy and the potential for targeted interventions that promote a Lactobacillus-dominant vaginal environment (Table 1) [31-34]. To compile the studies summarized in Table 1, we conducted a narrative selection of relevant cohort studies that investigated the relationship between the vaginal microbiome and PTB, with a particular focus on their potential use in predicting PTB risk. Studies were included based on the presence of key terms such as “vaginal microbiome”, “PTB”, “prediction”, and preference was given to large-scale cohort studies that enrolled diverse racial or ethnic populations.

Despite these commonalities, variations in study findings arise due to differences in study design, including sampling methods, sequencing techniques, and gestational age at sample collection. Some studies report L. iners as a transitional species that may either support a stable vaginal environment or contribute to dysbiosis, depending on host immune responses. Additionally, the composition of vaginal microbiota and its association with PTB differ slightly among ethnic groups, potentially due to genetic predisposition, dietary habits, and environmental exposures. These factors contribute to variations in PTB risk even within the same CST classification [32,35,36].

Furthermore, vaginal dysbiosis has been associated not only with PTB but also with conditions such as preterm prelabor rupture of membranes and intrauterine infections, suggesting that microbial imbalances have broader clinical implications. The inconsistencies observed across studies underscore the need for standardized methodologies and larger multi-ethnic cohort studies. Future research should integrate microbiome profiling with host immune and genetic factors to clarify the complex interactions contributing to PTB risk. A deeper understanding of these relationships will facilitate the development of more precise predictive models and targeted interventions for at-risk pregnancies.

Several methods are currently used to predict PTB, including microscopy, culture, and polymerase chain reaction-based tests [22,37,38]. The development of metagenomic sequencing has enabled comprehensive analysis of microbial communities, offering new avenues for PTB prediction [39]. Recent research has introduced non-invasive PTB prediction models that integrate CVF microbiome data with blood biomarkers and cervical length measurements [38,40]. Studies have explored various machine learning-based approaches, microbial biomarker panels combined with cytokine profiling, and metagenomic sequencing-based predictive models [41-43]. These models have demonstrated sensitivity ranging from 79% to 88% and specificity from 78% to 85%, highlighting their potential for improving PTB risk assessment [44]. Additionally, emerging computational techniques such as deep learning and neural networks are being investigated to enhance prediction accuracy by identifying complex microbial signatures [41,44-46]. However, despite these advancements, challenges remain in ensuring the reproducibility and generalizability of these models across diverse clinical settings, necessitating further validation through large-scale multi-ethnic cohort studies. Compared to conventional markers such as fetal fibronectin and phosphorylated insulin-like growth factor binding protein-1, which have reported sensitivities of approximately 56-64% and specificities of 75-85%, microbiome-based prediction models have demonstrated improved accuracy, with sensitivity ranging from 79% to 88% and specificity from 78% to 85% [47,48]. These findings suggest that integrating vaginal microbiome analysis with clinical factors such as cervical length and inflammatory biomarkers could enhance PTB risk assessment and early intervention strategies.

The association between BV and PTB has been well established in the literature [49]. In fact, due to this known relationship, it has been proposed that the Nugent scoring system should be incorporated into routine antenatal care [49]. The Nugent score is a standardized diagnostic method used to assess the vaginal microbiota based on Gram-stained vaginal smears. It evaluates the relative abundance of three bacterial morphotypes: large Gram-positive rods (Lactobacillus spp.), small Gram-variable rods (G. vaginalis and Bacteroides spp.), and curved Gram-variable rods (Mobiluncus spp.). Scores range from 0-10, with 0-3 indicating normal flora, 4-6 suggesting intermediate microbiota, and 7-10 consistent with BV. Despite the limitations of the scoring system, this has also shown a correlation with microbial diversity and dysbiosis in studies utilizing metagenomic approaches [50,51]. Given that PTB is a multifactorial syndrome, identifying vaginal dysbiosis as a contributing etiology may open new possibilities for precision-based prevention and management strategies aimed at reducing PTB risk.

Given the established link between vaginal microbiota composition and PTB, numerous therapeutic strategies have been investigated to restore and maintain a healthy microbial environment. Antibiotics have traditionally been used to treat BV, a condition associated with increased PTB risk due to its anaerobe-dominated microbiota. However, antibiotic treatments have shown inconsistent effects on reducing PTB incidence and carry concerns about promoting resistance and potentially disrupting beneficial microbial communities [52].

As a result, increasing attention has shifted toward probiotic and live biotherapeutic product interventions aimed at re-establishing a Lactobacillus-dominant microbiota [53-55]. Oral probiotics have produced variable outcomes in clinical trials, while vaginally administered L. crispatus has demonstrated more promising results, including reduction in BV recurrence and improved microbial stability [56,57].

Ongoing clinical trials are exploring the efficacy of different probiotic strains, routes of administration, dosages, and treatment durations in reducing PTB risk [58]. Although these alternative strategies are still under evaluation, they represent promising avenues for safe and targeted modulation of the vaginal microbiota during pregnancy [55,58]. Robust, placebo-controlled trials and mechanistic studies are needed to determine optimal therapeutic protocols and assess long-term maternal and neonatal outcomes [59]. However, based on the current evidence, in the short term, it is expected that early detection of vaginal dysbiosis-using either next-generation sequencing (NGS) or traditional methods such as the Nugent score-may allow for the identification of high-risk individuals and enable personalized treatment with probiotics or antibiotics to help prevent the occurrence of PTB.

Despite significant progress in microbiome research, several limitations continue to impede a complete understanding of its role in PTB. One major challenge is the difficulty in elucidating the precise mode of action by which specific microbiota influence pregnancy outcomes. The complex and dynamic interactions between microbial communities and the host immune system remain incompletely characterized [21]. Moreover, the potential interplay between the vaginal and gut microbiomes during pregnancy is an emerging area of interest, with early evidence suggesting that the gut microbiome may influence systemic immune responses that in turn affect vaginal microbial stability and pregnancy outcomes [60]. Diet, a major modulator of the gut microbiome, may thus exert indirect influence on the vaginal microbiota, offering a potential avenue for dietary interventions to prevent dysbiosis and adverse pregnancy outcomes [61]. Understanding this cross-talk may open new avenues for integrative maternal health strategies. Furthermore, NGS data, particularly from 16S rRNA amplicon sequencing or Shotgun metagenomic sequencing lack standardization in terms of sampling techniques, analytical pipelines, and bioinformatic processing, which hampers reproducibility and comparability across studies [38,62]. Inter-individual variability and geographic or ethnic influences further complicate the interpretation of results [63].

Future research should prioritize efforts to standardize NGS protocols and analytical frameworks, enabling cross-study harmonization and meta-analyses. Integrating multi-omics approaches, including proteomics, metabolomics, and transcriptomics, will be critical to elucidating host-microbe interactions and their role in pregnancy outcomes [64]. Investigating the immunological impact of vaginal microbiota through cytokine profiling and host gene expression studies will enhance our understanding of PTB pathogenesis [65]. In parallel, shotgun metagenomic sequencing [66], coupled with functional metagenomics and single-cell analysis, offers a high-resolution alternative to traditional sequencing, enabling detailed characterization of microbial communities and their functions [67]. As research advances, refining predictive models, developing targeted interventions, and validating findings through large-scale longitudinal studies will be essential to improving maternal and neonatal outcomes.

In light of emerging evidence, preventive strategies targeting the vaginal microbiota have gained increasing attention. While antibiotics have traditionally been used to treat abnormal microbiota, such as BV, their effectiveness in reducing PTB remains inconsistent and raises concerns about antibiotic resistance and disruption of beneficial microbes [49]. Recent clinical trials have explored probiotic therapy using Lactobacillus species, either orally or intravaginally administered, as a non-antibiotic approach to restore a healthy vaginal environment and reduce BV recurrence [53,58]. Moreover, studies investigating screening for dysbiosis during pregnancy, particularly among high-risk populations, suggest that early identification followed by targeted intervention could be a promising strategy for reducing the risk of PTB. Further large-scale, randomized controlled trials are needed to determine the optimal timing, route, and composition of probiotic or antimicrobial interventions for microbiome-based PTB prevention.

The vaginal microbiome plays a critical role in the risk and prevention of PTB. While Lactobacillus-dominant communities are generally protective, dysbiosis marked by anaerobic overgrowth increases susceptibility to PTB. Despite variations across studies, the consistent association between microbial composition and PTB underscores the need for continued investigation. Ongoing efforts to standardize methodologies and integrate multi-omics data will be essential in advancing microbial diagnostics and therapeutic strategies. The vaginal microbiome remains a promising frontier in improving maternal and neonatal health outcomes.