Evvy maps six subtypes of bacterial vaginosis in 100,000-sample analysis
BV is the most common vaginal condition among reproductive-aged women in the US
Bacterial vaginosis has long been described in simple terms: a disruption of the vaginal microbiome, an overgrowth of certain bacteria, a common infection treated with antibiotics. But many researchers have argued for years that the reality is more complicated than that.
Now, new large-scale data from New York-based women’s health company Evvy adds weight to that view — and attempts to organise it into clearer categories.
After analysing more than 100,000 vaginal microbiome samples collected across the United States, Evvy says it has identified six distinct microbial “subtypes” that can sit behind a clinical diagnosis of BV. The company is now integrating those subtypes into its at-home testing platform, giving patients more detailed information about the bacterial patterns linked to their diagnosis. It also plans to peer review its initial research - an important step towards clinical validity.
“BV has historically been defined in very broad terms, despite advances in sequencing technology,” says Dr. Kate McLean, OBGYN and Chief Medical Advisor at Evvy.
“By applying high-resolution microbiome analysis across a large and diverse population, Evvy can identify reproducible microbial patterns that add important biological context to a BV diagnosis. This is how precision medicine begins to take shape in women’s health.”
A common condition for women
BV is the most common vaginal condition among reproductive-aged women in the US, affecting more than 30% each year. It is also one of the leading reasons people with vaginas visit a gynaecologist. Yet it is still broadly defined in clinical practice, and recurrence rates remain high.
The idea that BV is biologically heterogeneous is not new. Researchers have previously described different vaginal “community state types”, strain-level differences in key bacteria such as Gardnerella, and the role of biofilms in treatment resistance. But those insights have largely remained in academic circles.
Evvy’s contribution, it argues, lies in scale — and in translating microbiome science into something more patient-facing.
Six recurring patterns
Over four years, Evvy used shotgun metagenomic sequencing — a method that can detect bacteria at species level and beyond — to build what it describes as the world’s largest vaginal microbiome dataset.
When the company clustered microbiome profiles from people diagnosed with BV by clinicians, six recurring patterns emerged.
There is what Evvy calls “typical BV”, where bacteria commonly associated with the condition, including certain Gardnerella and Prevotella species, dominate and protective Lactobacillus bacteria are low.
In “transitional BV”, the microbiome appears to be in flux — either shifting toward a more typical BV profile or recovering from one. This subtype is dominated by Lactobacillus iners, a species often considered less stable than other lactobacilli.
“Lacto-dominant BV” describes cases where protective lactobacilli such as Lactobacillus crispatus remain dominant, but BV-associated bacteria are present at lower levels.
“Biofilm BV” includes bacterial species known to form biofilms — structured communities that can shield microbes from antibiotics and may help explain why BV frequently returns after treatment.
There is also “mixed BV”, where bacteria linked to BV coexist with microbes more commonly associated with aerobic vaginitis, and “atypical BV”, marked by less common gram-positive anaerobic bacteria that may not be well captured by traditional diagnostics.
Importantly, Evvy-affiliated clinicians will continue to diagnose BV using established criteria. The subtypes are presented as an additional layer of biological context, not a replacement for current standards.
Patients who use Evvy’s at-home vaginal microbiome test and are diagnosed with BV will now see their microbial subtype included in their results, alongside educational material and access to clinician-designed care pathways where eligible.
From research to real-world care?
For patients who experience repeated infections, the question is whether these distinctions will translate into different treatment strategies or improved outcomes.
Recurrence rates for BV are high, and antibiotic therapy does not work equally well for everyone. If some subtypes are more associated with biofilms, instability, or mixed microbial states, that could eventually influence how trials are designed or how clinicians approach care.
Priyanka Jain, Evvy’s co-founder and chief executive, said the goal is to move beyond a one-size-fits-all understanding of vaginal health.
“For too long, patients with BV have been treated as if a simple diagnosis tells the whole story,” she said.
“By investing in uncovering deeper biological context, we’re helping clinicians and patients better understand what’s actually happening beneath the surface — and laying the groundwork for more precise, evidence-driven care.”
The company says it has shared its methodology and initial findings and plans to publish larger peer-reviewed papers on BV heterogeneity and recurrence.
Currently these findings aren’t peer-reviewed so independent validation will be important. So too will evidence linking these subtypes to meaningful clinical differences — whether in symptoms, recurrence risk, or treatment response. Without that, subtyping risks becoming descriptive rather than transformative.
Still, the announcement reflects a wider shift in women’s health: long underfunded and often generalised conditions are being re-examined with higher-resolution tools. The science suggesting that BV is not a single biological state has been building for years. What may be changing now is the attempt to operationalise that complexity — and deliver it directly to patients.