A Closer Look at Bacterial Vaginosis: Is Gardnerella vaginalis Really Public Enemy Number One?

Table of Contents

• Key Points
• Understanding the Debate: One Germ or a Community of Germs?
• The Theory Under Fire: How Gardnerella Became BV's Prime Suspect
• How This Scientific Rebuttal Was Built
• What the Evidence Really Shows
• A Look Back: The Original Inoculation Experiments
• Alternative Explanations for Bacterial Vaginosis
• What This Means for Patients
• Limitations of This Analysis
• What Should Patients Take Away?
• Frequently Asked Questions
• Source Information

Key Points

• G. vaginalis is found in at least 25% of healthy women, so its presence alone does not mean infection.
• Studies show G. vaginalis occurs in sexually inexperienced girls, challenging the idea it's only sexually transmitted.
• Historical inoculation experiments showed pure G. vaginalis caused symptoms in only 8–24% of women.
• BV is likely polymicrobial; no single bacterial species appears in all cases.
• Testing positive for G. vaginalis without symptoms may not require treatment; focus on symptoms.

Understanding the Debate: One Germ or a Community of Germs?

Bacterial vaginosis (BV) is the most common vaginal infection in women of reproductive age, causing symptoms such as unusual discharge, a fishy odor, itching, and discomfort. It is also linked to serious health complications, including increased risk of sexually transmitted infections, pelvic inflammatory disease, and pregnancy complications.

For decades, scientists have debated exactly what causes BV. The leading suspect has been a bacterium called Gardnerella vaginalis (G. vaginalis), first identified in the 1950s as a possible culprit behind what was then called "nonspecific vaginitis."

Recently, a team led by Dr. Jane Schwebke and colleagues published a "conceptual model" in The Journal of Infectious Diseases arguing that G. vaginalis is the primary etiological agent — the main cause — of BV. Their model describes G. vaginalis as a sexually acquired pathogen that creates a bacterial biofilm (a thin, sticky layer of bacteria) on the vaginal wall, teams up with anaerobic bacteria (bacteria that thrive without oxygen), and outcompetes the lactobacilli that normally keep the vagina healthy.

But Roxana J. Hickey and Dr. Larry J. Forney, two researchers from the University of Idaho, are pushing back hard. In a sharply written letter to the editor, they argue that this "one germ, one disease" model doesn't fit the evidence. Their central message: G. vaginalis is not always a troublemaker. In fact, it appears to be a completely normal resident of the vaginal microbiome in a significant proportion of healthy women.

This debate may sound academic, but it has real consequences for patients. If G. vaginalis is truly the sole cause of BV, then treatments that specifically target this one bacterium would make sense. But if BV is actually caused by a combination of microbes working together, then current prevention and treatment strategies may need to be fundamentally rethought.

The Theory Under Fire: How Gardnerella Became BV's Prime Suspect

To understand the critique, it helps to first understand the model being challenged. Schwebke and colleagues made several major claims in their 2014 paper:

• G. vaginalis is not part of "normal" vaginal flora — because if it were, one would expect it to be present in all women and in pubertal girls as well.
• G. vaginalis is a sexually acquired pathogen — meaning it is transmitted through sexual activity.
• It causes BV by establishing a bacterial biofilm on the vaginal epithelium (the layer of cells lining the vagina).
• It has synergistic associations with vaginal anaerobes and competitive interactions with healthy vaginal lactobacilli.
• Koch's postulates are fulfilled — meaning the organism meets the strict scientific criteria required to prove it causes a specific disease.
• Screening and treatment should specifically target G. vaginalis in future prevention and treatment strategies.

Hickey and Forney take issue with virtually every one of these points. They point out that there is no biological reason to expect any single bacterial species to be present in 100% of healthy individuals. Just as no single type of bacteria dominates everyone's gut or skin, the vaginal microbiome varies substantially from woman to woman.

How This Scientific Rebuttal Was Built

This letter is what scientists call a "correspondence" — a formal written rebuttal published in a peer-reviewed journal. Rather than conducting new laboratory experiments, the authors carefully re-analyzed the existing body of evidence. Specifically, they:

1. Re-examined the studies cited by Schwebke and colleagues in support of their model.
2. Identified key studies that Schwebke and colleagues omitted, dismissed, or failed to take into account.
3. Re-analyzed historical inoculation experiments from the 1950s and 1960s that tested whether G. vaginalis actually causes BV when introduced into healthy women.
4. Reviewed modern cultivation-independent (DNA-based) studies that identify vaginal bacteria directly from samples, without needing to grow them in a laboratory dish.

Their goal was straightforward: determine whether the evidence truly supports naming G. vaginalis as the primary cause of BV — or whether the model was built on a selective reading of the data.

What the Evidence Really Shows

Finding 1: G. vaginalis is common in healthy women

The original model claims that G. vaginalis cannot be a member of "normal" vaginal flora because it isn't found in all women. Hickey and Forney call this claim illogical. No bacterial species is present in every single healthy person — not in the gut, not on the skin, and not in the vagina.

Numerous cultivation-independent studies — which use DNA sequencing techniques to identify bacteria in vaginal samples — have identified multiple different types of vaginal microbiota among healthy, symptom-free women. In these studies, G. vaginalis has been found in at least a quarter of healthy subjects (25% or more). These include landmark studies by Hyman and colleagues (2005), Fredricks and colleagues (2005), Zhou and colleagues (2007), and Ravel and colleagues (2011).

Schwebke and colleagues argue that these studies were "not rigorous" in their definition of "normal or optimal flora." But here's the crucial point: even Schwebke's own research contradicts her claims. In a study by Schwebke, Flynn, and Rivers (published in 2014), the researchers detected G. vaginalis in 38.5% of women who had a predominance of protective lactobacilli and a "Nugent score" of 0–3 — which are considered the hallmark conditions of a normal and healthy vagina. (The Nugent score is a standard laboratory method for diagnosing BV: 0–3 is considered normal, 4–6 intermediate, and 7–10 indicates BV.)

The authors of the original model emphasized that they were unable to detect G. vaginalis in the other 61.5% of healthy subjects. But Hickey and Forney argue this hardly proves that G. vaginalis is absent from healthy women — rather, it simply shows the bacterium isn't found in everyone, which is completely normal for any microbial species.

Finding 2: G. vaginalis is found in sexually inexperienced girls

The model posits that G. vaginalis is a sexually transmitted infection. But studies dating back to the 1980s consistently show that G. vaginalis is almost as common in sexually inexperienced adolescent females as in sexually active ones. Five separate studies support this finding:

• Bump and Buesching (1988) — found bacterial vaginosis in virginal and sexually active adolescent females, providing direct evidence against exclusive sexual transmission.
• Fethers and colleagues (2012) — examined BV candidate bacteria in sexually experienced and inexperienced women, published in PLoS One.
• Shafer and colleagues (1985) — studied the microbiology of the lower genital tract in postmenarchal (post-first-period) adolescent girls, published in The Journal of Pediatrics.
• Yamamoto and colleagues (2009) — examined bacterial populations in the vaginas of healthy adolescent women, published in the Journal of Pediatric and Adolescent Gynecology.
• Tabrizi and colleagues (2006) — specifically measured the prevalence of G. vaginalis and Atopobium vaginae in virginal women, published in Sexually Transmitted Diseases.

The authors note that Muzny and Schwebke acknowledged two of these studies in a recent review article (published in Current Infectious Disease Reports in 2013), but dismissed the findings by casting doubt on whether the young participants truthfully reported their virginal status. Hickey and Forney call this an unreasonable and implausible assumption.

Their more plausible interpretation: G. vaginalis (or at least some strains of it) may well be a member of the normal vaginal microbiota in a significant proportion of women. The fact that it can be transmitted through sexual activity does not preclude it from also being a harmless, normal resident of healthy vaginas in both girls and women.

A Look Back: The Original Inoculation Experiments

To understand the heart of this debate, we need to look at Koch's postulates — the classic scientific criteria, developed in the 19th century, for proving that a specific germ causes a specific disease. The four postulates require that:

1. The organism must be found in all individuals with the disease.
2. The organism must be isolated and grown in pure culture.
3. The organism, when introduced into a healthy, susceptible host, must elicit the disease.
4. The organism must be re-isolated from the newly infected host.

Schwebke and colleagues claimed that G. vaginalis fulfills all of these postulates for BV. Hickey and Forney argue that it fails at least the first, third, and fourth.

The first postulate fails for the reasons already discussed: G. vaginalis is commonly found in healthy individuals who show no signs of disease.

The third and fourth postulates fail based on historical evidence. Schwebke and colleagues referenced a 1969 study by Criswell and colleagues in which 29 healthy pregnant women were vaginally inoculated with pure cultures of G. vaginalis (then classified as Haemophilus vaginalis) and monitored for clinical and laboratory signs of "H. vaginalis vaginitis." The results, however, were far from convincing: only 7 of those 29 women (24%) developed clinical signs of BV. That means 76% of women did not develop symptoms after being directly inoculated with the suspected pathogen.

Even more striking are the earlier experiments by Gardner and Dukes (1955), the researchers who first linked this bacterium to "nonspecific vaginitis." Their results showed a dramatic difference between two types of inoculation:

• 11 of 15 women (73%) who were inoculated with vaginal secretions from affected individuals developed symptoms.
• But only 1 of 13 women (approximately 8%) who were inoculated with pure cultures of G. vaginalis developed symptoms.

This difference is striking. Vaginal secretions contain an entire community of bacteria, so they caused BV in a far higher percentage of women than the single organism alone. These early experiments strongly suggest that BV is not caused by G. vaginalis acting alone, but by a consortium of organisms working together. Hickey and Forney conclude that "so-called nonspecific vaginitis, or BV, is almost certainly polymicrobial in nature and, thus, not easily confirmed within the framework of Koch's postulates."

Modern research supports this view. A study by Srinivasan and colleagues (2012) using high-resolution phylogenetic (DNA-based) analyses found that multiple different bacterial species are associated with symptoms of BV, but no single species appears to be present in all cases. In other words, there is no "one germ" behind BV.

Alternative Explanations for Bacterial Vaginosis

If G. vaginalis isn't the sole cause of BV, what is going on? Hickey and Forney propose three alternative explanations, which they note are not mutually exclusive:

1. Strain-specific differences in virulence. Within the species G. vaginalis, there are genetic differences between strains. Some strains can elicit one or more symptoms of BV, while other strains are benign commensals — harmless residents found in healthy individuals. This would explain why G. vaginalis is found in healthy women and in women with BV.
2. Conditional expression of virulence genes. All strains of G. vaginalis may carry the genetic potential (virulence determinants) to cause disease, but these genes are only "switched on" under specific biological or environmental conditions in the vagina. Certain triggers — perhaps changes in pH, hormonal shifts, or the presence of other bacteria — could activate harmful behaviors.
3. Host susceptibility differences. All strains of G. vaginalis may express their virulence factors in the vaginal environment, but only some individuals are susceptible or respond negatively to these factors. Genetics, immune status, or other host factors could make certain women more vulnerable to developing BV.

The authors also acknowledge that strain-dependent differences in the virulence potential of G. vaginalis likely exist, and that potentially important synergistic relationships with other bacteria deserve further study. In other words, even if G. vaginalis plays a role in BV, it may not be the ringleader — it may be just one member of a gang of bacteria working together.

What This Means for Patients

This debate is not purely academic. It has potential consequences for how BV is diagnosed, treated, and prevented. Here's what the authors' analysis means for women:

• Treatment targets may be too narrow. If BV is truly a polymicrobial condition, then therapies designed to wipe out a single organism may fail to address the real underlying problem. Current treatments for BV (typically antibiotics like metronidazole or clindamycin) are fairly broad-spectrum, but if new targeted therapies are developed specifically against G. vaginalis, they might not work for all patients and could miss other bacteria involved in the infection.
• Recurrence rates make more sense. BV has a notoriously high recurrence rate. The authors' analysis suggests one reason: we may be treating the wrong target, or failing to restore a healthy, lactobacillus-dominant microbiome. Simply killing G. vaginalis without restoring the balance of the vaginal ecosystem may set the stage for recurrence.
• Testing positive for G. vaginalis doesn't mean you have an infection. The finding that G. vaginalis lives harmlessly in at least a quarter of healthy women — and even in a significant proportion of women with lactobacillus-predominant, "normal" flora — suggests that its mere presence is not a cause for alarm. A positive test for G. vaginalis alone, without symptoms or other signs of BV, may not require treatment.
• The "normal" vaginal microbiome is diverse. The authors emphasize that our understanding of "normal and healthy" with respect to the vaginal microbiome remains incomplete. Studies have documented different vaginal bacterial communities in healthy women of different racial and ethnic backgrounds (e.g., Zhou and colleagues 2007, Ravel and colleagues 2011), and this diversity needs to be respected in both research and clinical practice.
• Correlation does not equal causation. Hickey and Forney acknowledge that G. vaginalis is associated with BV and agree that this association warrants further investigation. But the association between the bacterium and the condition does not prove it causes the condition. Other factors — including the broader microbial community and host responses — may be more important than any single species.

Limitations of This Analysis

It's important to note what this letter is — and what it is not. This is a correspondence piece, not a new experimental study. The authors did not collect new data, enroll new patients, or perform new laboratory experiments. Instead, they re-analyzed existing published findings and constructed a logical argument. As with any scientific critique, there are limitations:

• The authors could only work with the data that other researchers chose to publish; publication bias is always possible in any field.
• Some of the historical studies cited (the 1969 Criswell study and the 1955 Gardner and Dukes studies) had small sample sizes — 29, 15, and 13 participants, respectively — and reflect the scientific standards of their era, including less rigorous diagnostic methods.
• This letter does not definitively prove any of the three alternative hypotheses it proposes. It simply demonstrates that the case for G. vaginalis as the primary cause of BV is far weaker than its proponents claim.
• The authors do not offer a fully developed alternative model that explains all cases of BV. They openly acknowledge that "there is still much work to be done in understanding the complex etiology of BV."

Still, the argument is compelling because it draws on multiple lines of evidence: modern DNA-based microbial ecology, historical inoculation studies, and the epidemiology of BV in sexually inexperienced populations.

What Should Patients Take Away?

So what does this mean for you, the patient? Here are practical takeaways from this scientific exchange:

• Don't panic about testing positive for G. vaginalis. If you have no symptoms, the presence of this bacterium alone does not mean you have an infection that needs treatment.
• Understand that BV is a complex condition. Science is still working out exactly how and why it develops. There may not be a single "cause" to identify and target.
• Focus on symptoms, not just lab tests. BV is diagnosed based on a combination of clinical criteria (vaginal discharge, odor, pH, and the presence of "clue cells" under the microscope) plus the Nugent score. If you have symptoms, seek care. If you don't, a positive test for G. vaginalis alone may not require treatment.
• Be aware of recurrence and talk to your doctor. BV often comes back after standard antibiotic treatment. Ask your healthcare provider about strategies for preventing recurrence, including whether probiotics or other approaches might help restore a healthy vaginal microbiome.
• Support healthy vaginal flora. Lactobacilli are protective, and maintaining a balanced vaginal microbiome is likely the best defense against BV. Avoid douching, which can disrupt the natural balance, and consider discussing vaginal health with your provider during routine visits.
• Stay tuned for more research. The vaginal microbiome is a rapidly evolving field. As DNA-sequencing technology improves and larger studies are conducted, we will likely learn much more about the complex microbial community that calls the vagina home — and how it relates to both health and disease.

Frequently Asked Questions

I tested positive for Gardnerella vaginalis but I have no symptoms. Do I need treatment?

Not necessarily. The article notes that G. vaginalis is found harmlessly in at least 25% of healthy women, and even in women with a normal, lactobacillus-predominant microbiome. A positive test alone, without symptoms or other signs of bacterial vaginosis, may not require treatment. Discuss your specific situation with your healthcare provider.

Is bacterial vaginosis a sexually transmitted infection?

The article challenges the idea that G. vaginalis is solely a sexually acquired pathogen. Studies dating back to the 1980s found G. vaginalis in sexually inexperienced adolescent girls almost as often as in sexually active ones. Sexual transmission may occur, but the bacterium can also be a normal resident in healthy women and girls.

Why does bacterial vaginosis keep coming back after antibiotic treatment?

The article suggests that the high recurrence rate of BV may be because we are treating the wrong target. BV appears to be caused by a community of bacteria working together, not just one species. Simply killing G. vaginalis without restoring a healthy, lactobacillus-dominant microbiome may set the stage for recurrence.

What does a Nugent score mean for diagnosing bacterial vaginosis?

The Nugent score is a standard laboratory method for diagnosing BV. A score of 0–3 is considered normal, 4–6 intermediate, and 7–10 indicates BV. The article discusses that even in women with a normal Nugent score of 0–3, G. vaginalis can be detected, showing it is not always a sign of infection.

Can Gardnerella vaginalis be found in women who have never had sex?

Yes. The article cites five separate studies from the 1980s onward showing G. vaginalis is almost as common in sexually inexperienced adolescent females as in sexually active ones. This includes a 2006 study that specifically measured its prevalence in virginal women, suggesting it can be part of normal vaginal flora.

If G. vaginalis isn't the main cause of BV, what is?

The article proposes that BV is almost certainly polymicrobial, meaning multiple bacteria working together cause it. Alternative explanations include strain-specific differences in G. vaginalis, conditional expression of virulence genes, or host susceptibility differences. No single species has been found in all cases of BV.

Should I worry about having a positive G. vaginalis test without symptoms?

The article emphasizes that the mere presence of G. vaginalis is not a cause for alarm. It lives harmlessly in at least a quarter of healthy women and even in women with lactobacillus-predominant normal flora. A positive test alone, without symptoms or other BV signs, may not require treatment.

I tested positive for Gardnerella vaginalis but have no symptoms. Should I get a second opinion before starting treatment?

A positive test for Gardnerella vaginalis alone does not necessarily mean you have bacterial vaginosis or need treatment. The bacterium is found in at least 25% of healthy women and even in those with normal lactobacillus-predominant flora. BV is likely caused by a community of bacteria, not a single species. If you have no symptoms, a second opinion can help clarify whether treatment is truly necessary, especially given the high recurrence rate of BV. Diagnostic Detectives Network provides independent expert second opinions.

Source Information

Original article title: TO THE EDITOR—Schwebke et al [1] present a conceptual model arguing the

Authors: Roxana J. Hickey and Larry J. Forney

Affiliations: Department of Biological Sciences and Institute for Bioinformatics and Evolutionary Studies, University of Idaho, Moscow, Idaho

Publication: The Journal of Infectious Diseases, 2014, Volume 210, Issue 10, Pages 1682–1683. Published by Oxford University Press on behalf of the Infectious Diseases Society of America.

DOI: 10.1093/infdis/jiu303

Published electronically: