Researchers tested an unexpected addition to vaginal treatment: oxygen

By Dr. Liji Thomas, MDReviewed by Lauren HardakerOct 28 2026

Can increasing oxygen levels in the vagina change the balance between protective bacteria and microbes associated with infection? Researchers tested the approach in women with bacterial vaginosis and vaginitis.

Study: Topical Oxygen Treatment Restores Vaginal Microbiota Balance in Women with Vaginitis and Bacterial Vaginosis: An Exploratory Study. Image Credit: Anusorn Nakdee / Shutterstock

A recent study published in the journal Microorganisms describes changes in the vaginal microbiome following topical oxygen treatment, alone or combined with antimicrobials. The study cannot establish whether oxygen itself caused these changes. The authors describe it as an exploratory study with findings that warrant further trials.

Background

Vaginal health depends on a Lactobacillus-dominated microbial community. These bacteria use glycogen-derived sugars to produce lactic acid, maintaining an acidic pH of roughly 3.5 to 4.5. They also produce bacteriocins and biosurfactants. Some species produce hydrogen peroxide, but its protective role within the vagina remains uncertain. Lactobacilli can prevent pathogens from adhering to the vaginal lining.

While L. crispatus is strongly associated with a resilient vaginal microbiome, L. iners is less associated with vaginal health, as it does not produce D-lactic acid and encodes a toxin upregulated under conditions of dysbiosis. Vaginal lactobacilli are aerotolerant, meaning they can survive exposure to oxygen.

Disruption of the vaginal lactobacillus-dominated commensal community is associated with bacterial vaginosis (BV) and mixed vulvovaginitis. BV is characterized by depleted lactobacilli, a pH above 4.5, clue cells on microscopy, and a polymicrobial overgrowth of anaerobic organisms such as Gardnerella vaginalis, Atopobium, Sneathia, Prevotella, and Megasphaera.

Conventional antimicrobial therapy can clear acute infection but may also deplete beneficial lactobacilli. Loss of these protective bacteria may contribute to the high recurrence rates seen in BV. The study investigates the effect of raising local oxygen tension.

Reported oxygen tension in a healthy vagina typically ranges from 15 to 30 mm Hg. Anaerobic overgrowth during BV may further reduce local oxygen availability.

Obligate anaerobes vary in their sensitivity to oxygen. Many possess antioxidant defenses, but oxygen exposure can still disrupt oxygen-sensitive metabolic processes and limit bacterial growth. Lactobacilli, by contrast, are aerotolerant and possess mechanisms that help them withstand oxidative stress.

The hypothesis is that oxygen can selectively suppress some dysbiosis-associated organisms while preserving or restoring protective species.

Study characteristics

This was an exploratory, randomized, open-label pilot study conducted at the Obstetrics Unit of the University Hospital of Modena, Italy. Twenty-two symptomatic women aged 18 to 45 with BV or non-specific vaginitis were allocated 1:1 by sealed envelope to one of two arms of eleven participants each.

One arm received topical oxygen alone, while the other also received targeted antibiotic or antimycotic treatment for 5 to 7 days.

The study delivered oxygen (concentration above 95%) intravaginally with the Caress Flow device at 3 L/min for 20 minutes. Sessions were scheduled twice weekly for two weeks, for a total of four sessions.

The device uses a sterile single-use cannula with lateral micro-apertures. The authors state that because it does not close off the vaginal canal from the atmosphere, this method is designed to minimize the risks of barotrauma and gas embolism. These risks were not independently assessed in this small study.

Women excluded from the study were those with vulvovaginal atrophy, amenorrhea of at least three months, and a history of antibiotic exposure within the previous 30 days.

Vaginal swabs were collected at baseline and 10 to 14 days after treatment. The researchers characterized microbiota composition using 16S rRNA gene sequencing, which identifies bacterial groups using genetic sequences. Pre- and post-treatment values were compared. Bacteria were also grouped as strict or facultative anaerobes.

Oxygen-alone arm

In the oxygen-only arm, pH remained unchanged at a median of 4.50. Swab results were negative for 4 of 11 women at baseline, rising to 9 of 11 after the oxygen treatment. This change approached statistical significance but did not reach it. These swab classifications alone do not establish that BV had resolved.

Total Lactobacillus relative abundance, the proportion of bacterial sequences assigned to this genus, was already high at baseline at a median of 92.6%, and remained stable at 91.5% afterward. Some increases were noted, including among sequences assigned to L. crispatus, L. iners and L. helveticus, though none individually was significant.

Facultative anaerobes changed little, with a small reduction in strict anaerobes from 6.9% to 5.6%. Small, non-significant reductions were reported for Bacteroides, Peptoniphilus and Streptococcus. The pattern is consistent with the researchers' oxygen-sensitivity hypothesis, but lactobacilli were already dominant at baseline.

Antimicrobial treatment plus oxygen

In the combined arm, the microbial profile shifted more substantially. This group began with considerably fewer lactobacilli and more strict anaerobes than the oxygen-only group, limiting comparisons between the regimens.

The median pH moved from 5.00 to 4.75, a change that was not statistically significant and left the median above 4.5. The number of negative swabs increased from 4 to 8 among the 11 women. At the end, no swabs were classified as positive, but white blood cells were still detected in the vaginal swabs of three women.

Total Lactobacillus abundance increased from a median of 49.1% to 85.2%, but the change was not statistically significant. L. iners increased significantly, from 18.9% to 47.9%.

Strict anaerobes also decreased significantly, from 24.8% to 9.0%. The analysis found significant reductions in Atopobium and Sneathia among individual genera, with a borderline reduction in Clostridium. Gardnerella did not change significantly. Atopobium and Sneathia were detected at appreciable levels in only some participants. The reported reductions reflected changes across individual samples rather than a decrease in the group median.

There were some opposing changes. Streptococcus and Staphylococcus both increased significantly, though Staphylococcus remained at very low relative abundance. The researchers considered the staphylococcal increase clinically negligible and suggested that both changes could reflect colonization of an ecological niche vacated by anaerobes. The short follow-up could not establish whether these increases were temporary. No participant developed symptoms of aerobic vaginitis.

Interpretation

The authors interpret these findings as suggesting that topical oxygen may help alter the vaginal microbiota, particularly as an adjunct to antimicrobial therapy. They highlight the reduction in Atopobium spp., a group containing organisms associated with biofilm formation and metronidazole resistance, as a potentially useful finding. The study did not measure biofilm disruption or demonstrate bacterial eradication.

The authors also frame the expansion of L. iners as a pioneer-species phenomenon. This term refers to its proposed role as a transitional colonizer that may help re-acidify the environment following the suppression of strict anaerobes. This could create conditions in which more protective species such as L. crispatus later become dominant. The study did not establish this sequence, and L. iners dominance can also occur in women with persistent or recurrent dysbiosis.

Safety

The study reports good short-term tolerability, with no adverse events reported among the 22 women. Participants described the procedure as comfortable and acceptable. The authors suggest that this approach may warrant further study as a non-invasive, antibiotic-sparing option, and call for larger, sham-controlled, multicenter trials with extended follow-up. The small sample cannot rule out uncommon adverse effects.

Limitations

Several constraints limit interpretation. The sample was very small, and there was no long-term follow-up. The study lacked an antimicrobial-only control arm or a sham-treatment group, so the additional effect of oxygen remains unknown. The two groups also differed considerably in microbiota composition at baseline, complicating comparisons between regimens. The analysis involved numerous comparisons of individual bacterial genera without a reported adjustment for multiple testing.

The sample had a heterogeneous dysbiotic profile, including BV and non-specific and potentially fungal vaginitis. 16S sequencing also offers limited species-level resolution and measures relative rather than absolute abundance. The study did not measure vaginal oxygen tension or redox potential, nor did it establish that bacterial counts fell, symptoms improved because of oxygen, or BV recurrence was prevented.

Conclusion

This pilot provides preliminary evidence that topical oxygen treatment of the vagina is feasible, well tolerated over a short period, and associated with changes that may favor a healthier vaginal microbial composition, especially a reduction in strict anaerobes when combined with antimicrobial therapy.

The small sample size, lack of appropriate control groups, baseline differences between treatment groups, and short follow-up mean the findings are hypothesis-generating only.

Future adequately powered controlled trials are necessary to investigate whether this approach adds benefit beyond standard antimicrobial treatment, reduces recurrence or improves clinical outcomes.