Vaginal Microbiome Study of 10,000 Women Reveals Lasting Microbial Scars of Infection

Two of the most common conditions affecting women’s health—bacterial vaginosis and urinary tract infection—have long been studied in isolation, in small cohorts and with limited genetic resolution. A new large-scale analysis published in Nature Communications now offers one of the most detailed population-level portraits to date of the vaginal microbiome and its relationship to these infections. By analyzing shotgun metagenomic profiles from 10,003 women across the United States, a team led by Sonia N. Whang and Nicole M. Gilbert of Washington University School of Medicine in St. Louis has shown that the microbial fingerprints of recent bacterial vaginosis and urinary tract infection remain detectable at the community level, even in a cross-sectional snapshot of the population.

The scale of the study is what sets it apart. Rather than enrolling patients through clinics, the researchers drew on self-reported clinical histories—whether participants had recently experienced bacterial vaginosis, a urinary tract infection, both, or neither—and paired those reports with high-resolution DNA sequencing of the vaginal microbiome. Shotgun metagenomics, which sequences all genetic material in a sample rather than targeting a single marker gene, allowed the team to identify bacteria at the species level and even distinguish among closely related strains. That resolution matters enormously in this context, because the organisms implicated in bacterial vaginosis are notoriously difficult to tell apart with older techniques.

The analysis revealed that women who reported a recent diagnosis of either condition displayed markedly different vaginal community structures compared with women who reported neither. Using the VALENCIA classification system, which assigns vaginal microbiomes to community state types based on their dominant organisms, the team found that the subtypes known as CST-IV—communities dominated by diverse anaerobic bacteria rather than protective Lactobacillus species—were significantly more prevalent among women with recent infections. Alpha diversity, a measure of how many different species coexist in a community, was also significantly elevated in these groups. In healthy vaginal environments, low diversity dominated by Lactobacillus crispatus or related species is generally considered protective, so elevated diversity and CST-IV prevalence are consistent with a disrupted microbial ecosystem.

One of the study’s most technically significant contributions lies in its species-level profiling of Gardnerella, a genus long associated with bacterial vaginosis. Historically, clinical microbiology treated Gardnerella vaginalis as a single entity, but genomic work over the past decade has shown that what was once called one species is in fact a collection of distinct species with different genetic repertoires and potentially different clinical behaviors. In this dataset, multiple Gardnerella species were enriched in women reporting recent bacterial vaginosis, confirming and refining that association. Notably, however, Gardnerella abundance did not differ significantly between women with and without recent urinary tract infections—a finding that tempers earlier mechanistic hypotheses suggesting a direct link between Gardnerella colonization and susceptibility to bladder infections.

Instead, the organisms that stood out in the urinary tract infection group were the classic uropathogens themselves. Escherichia coli, the leading cause of urinary tract infections worldwide, along with Enterococcus faecalis and Staphylococcus saprophyticus, were detectable at higher prevalence and higher relative abundance in the vaginal microbiomes of women who had recently experienced a urinary tract infection. This observation carries particular weight because the vagina is increasingly understood not merely as a bystander in urinary infections but as a potential reservoir from which uropathogens can seed the bladder. Detecting these organisms in vaginal samples from recently infected women fits that model of the vagina as a way station for pathogens that cause recurrent disease.

Perhaps the most striking finding concerns antibiotic exposure. Even among participants who reported recent antibiotic use, uropathogens remained detectable at elevated prevalence and abundance in the vagina. Antibiotics are the standard treatment for urinary tract infections, and one might expect a recent course of therapy to clear susceptible organisms from adjacent sites. The persistence of E. coli, E. faecalis, and S. saprophyticus in these samples is consistent with the possibility of residual colonization or rapid recurrence, suggesting that treatment directed at the bladder may leave the vaginal reservoir largely intact. If confirmed in longitudinal studies, this could help explain why so many women experience recurrent urinary tract infections within months of an initial episode.

The study also delivers a resource of considerable value to the research community. Large, species-level reference datasets for the human vaginal metagenome have been scarce, in part because vaginal samples are underrepresented in public databases compared with gut samples. By releasing profiles and linked clinical metadata from more than ten thousand participants, the team has created a high-resolution baseline against which future studies can compare their findings. The dataset spans women with and without recent urogenital infections, providing natural variation in community composition that can be used to benchmark bioinformatic tools, validate species classifiers, and generate hypotheses about microbial interactions in the urogenital tract.

The authors are careful about causality. Because the data are cross-sectional and rely on self-reported diagnoses, the study cannot determine whether the observed microbial signatures preceded the infections, resulted from them, or reflect shared underlying factors. A recent urinary tract infection could reshape the vaginal community through inflammation and antibiotic pressure, or a vagina already colonized by uropathogens could make infection more likely. Similarly, self-reported bacterial vaginosis may capture a spectrum of clinical presentations. What the study establishes firmly is that these signatures are detectable at population scale—a necessary foundation for the prospective studies that will be needed to untangle cause and effect. The authors also note the study’s funding sources, including the National Institute of Diabetes and Digestive and Kidney Diseases and Allora Health, the consumer microbiome testing company that employed some of the coauthors.

The practical implications point toward prevention strategies that treat the vagina as an active participant in urinary health. If uropathogens persist in the vaginal reservoir after treatment, then interventions aimed at restoring Lactobacillus-dominated communities, eliminating residual colonization, or screening for high-risk vaginal profiles could complement existing approaches to preventing recurrent infections. The refined Gardnerella findings sharpen the research agenda as well: rather than pursuing a single organism, investigators can now ask which specific species, and which genes they carry, matter for clinical outcomes. As metagenomic sequencing becomes more accessible, population-scale datasets like this one are likely to redefine how clinicians think about the boundary between vaginal and urinary health—one species at a time.

Subject of Research: Metagenomic characterization of the vaginal microbiome and uropathogens in women with recent bacterial vaginosis and urinary tract infections

Article Title: Large-scale metagenomic characterization of the vaginal microbiome and uropathogens in women with recent bacterial vaginosis and urinary tract infections

Article References: Whang, S. N., Wang, X., Thomas-White, K. J., Olmschenk, G., Garza, J. E., Navarro, P., & Gilbert, N. M. (2026). Large-scale metagenomic characterization of the vaginal microbiome and uropathogens in women with recent bacterial vaginosis and urinary tract infections. Nature Communications. https://doi.org/10.1038/s41467-026-78217-1

Image Credits: AI Generated

DOI: 10.1038/s41467-026-78217-1

Keywords: vaginal microbiome, bacterial vaginosis, urinary tract infection, metagenomics, Gardnerella, uropathogens, E. coli, community state types, microbiome diversity, women's health, antibiotic use, Nature Communications