Back to Blog

Vaginal Microbiome and IVF: What Lactobacillus Dominance May Mean for Implantation

Dr. Senai Aksoy · · Updated · 8 min read
Medically reviewed on 7 November 2026 - Dr. Senai Aksoy
Stylised split illustration of a sheltered landscape with small bacteria-like figures beside a field of bright light

Key Takeaways

Studies link a Lactobacillus-dominant vaginal microbiome with better IVF outcomes, but a link is not proof. A 2025 trial found that treating abnormal vaginal bacteria before embryo transfer did not improve pregnancy or live-birth rates. Vaginal symptoms and recurrent bacterial vaginosis still deserve a standard clinical check. Commercial vaginal or endometrial microbiome testing is not recommended, even after recurrent implantation failure.

Key evidence: ESHRE Good Practice Recommendations on Recurrent Implantation Failure CDC STI Treatment Guidelines — Bacterial Vaginosis Haahr et al. 2025 — randomised trial of clindamycin and LACTIN-V in IVF

When the vaginal microbiome enters the IVF conversation

When IVF fails more than once, attention often goes first to embryo quality or endometrial thickness. Both matter, but neither explains every failure, and that is where questions about the vaginal microbiome come in.

Research has linked certain vaginal bacteria patterns with IVF outcomes in some patients. It has not shown that one pattern directly causes an embryo to implant or fail, so the microbiome is not a hidden key to every failed cycle.

What a “healthy” vaginal microbiome usually means

In many women of reproductive age, protective Lactobacillus bacteria dominate the vagina. Lactobacillus crispatus is the species most often linked with a stable, acidic environment. It produces lactic acid, which helps keep the vagina acidic.

Sometimes a more mixed group of bacteria that grow without oxygen takes over. It may include Gardnerella, Atopobium, or Prevotella. This pattern is often seen in vaginal dysbiosis (an imbalance in the usual bacteria) and bacterial vaginosis. A DNA test result alone, however, does not diagnose bacterial vaginosis.

Why it may matter for IVF

The embryo goes into the uterus, not the vagina, so any effect of the vaginal microbiome would be indirect.

Researchers have suggested that dysbiosis may be a sign of local inflammation and changed immune signals. Bacteria moving up into the uterus and changes in the local immune response are two possible explanations (Moreno & Simon, 2019). Both make biological sense, but neither has been shown to cause implantation failure.

Is there a separate uterine (endometrial) microbiome?

The uterine cavity is not a continuation of the vagina. Samples from the uterine lining hold far fewer bacteria. In one study with carefully collected samples, they rarely showed the Lactobacillus dominance seen in the vagina (Reschini et al., 2022). These samples are also easy to contaminate with bacteria from the vagina or cervix.

A low Lactobacillus percentage in a uterine sample is not, on its own, proof of a problem. In one study, pregnancy rates were similar whether or not the sample was labelled imbalanced. Some women conceived with no Lactobacillus found in the sample (Hashimoto & Kyono, 2019).

Evidence box: Reschini and Hashimoto study numbers

Reschini et al., 2022: researchers used embryo-transfer catheters to sample endometrial fluid from 53 women preparing for frozen embryo transfer, and took matched vaginal swabs from the same women. Lactobacillus dominance (more than 90% of sequencing reads) appeared in only 4 of 53 endometrial samples (8%). The dominant bacterial genera matched between vagina and endometrium in only 4 of 53 women (also 8%). The authors preferred double-lumen catheters to limit contamination.

Hashimoto & Kyono, 2019: this IVF cohort (a group followed over time) included 99 women under 40 having frozen blastocyst transfer. Pregnancy rates were similar between women labelled “dysbiotic” (imbalanced) and “eubiotic” (balanced) on testing of endometrial fluid.

We still do not know which bacteria truly live in the uterus, what levels should count as normal, or whether changing that environment changes the chance of live birth.

What the current data show

Observational studies (which compare groups without testing a treatment) link Lactobacillus dominance, especially L. crispatus, with higher implantation or pregnancy rates. A recent systematic review (a study that pools earlier studies) found the same pattern, including for live birth. Its authors also warned of low quality and risk of bias in the evidence (Samama et al., 2025).

A link is not proof. The most direct trial so far showed no benefit: treating abnormal vaginal bacteria before embryo transfer did not raise pregnancy or live-birth rates (Haahr et al., 2025).

Lactobacillus species may not all act the same way, and studies differ on which ones matter. A 2026 review judged the evidence too weak to support microbiome testing in patient care (Shabana et al., 2026). For now, the microbiome remains a research marker rather than a treatment target.

Evidence box: the main studies

Haahr et al., 2025: a randomised (patients assigned by chance), double-blind (neither patients nor clinicians knew who received what) trial of 338 IVF patients with abnormal vaginal bacteria on testing. Clinical-pregnancy rates (a pregnancy confirmed on ultrasound, which is not the same as live birth) were 42% with clindamycin plus a vaginal L. crispatus product, 46% with clindamycin plus placebo (an inactive look-alike treatment), and 45% with two placebos. Live-birth rates did not differ significantly. The vaginal product was investigational, and no oral probiotic was tested.

Shabana et al., 2026: a network meta-analysis (a review that compares several groups at once) of 12 observational studies and 2,289 women. Compared with L. crispatus profiles, the odds of pregnancy were lower with L. jensenii-dominated and low-Lactobacillus profiles, but the L. jensenii result rested on sparse data. Results for L. iners and L. gasseri were too imprecise to show a difference. The estimates were not adjusted for other factors, and only 57 live births were reported, across five studies.

Zhao et al., 2025: in one small IVF study, L. iners, itself a Lactobacillus species, was more abundant in women who did not conceive.

These findings come with limits:

What should be evaluated, and when?

Vaginal symptoms need a clinical check, but that does not automatically include microbiome testing. Unusual odour or discharge, irritation, or recurrent bacterial vaginosis call for a clinical history and the usual tests.

A commercial microbiome profile is a different test. Partly because so much about the uterine microbiome is unknown, the 2023 recommendations on recurrent implantation failure from the European Society of Human Reproduction and Embryology (ESHRE) state that uterine and vaginal microbiome profiling is not recommended.

Suspected chronic endometritis is a separate question. A vaginal microbiome result does not diagnose it. It is assessed separately with standard methods.

How to support a healthy vaginal microbiome

There is no standardised way to “raise” a Lactobacillus percentage. The practical steps are simpler:

  1. Avoid vaginal douching, scented wipes, antiseptic washes, and harsh soaps.
  2. Do not put food products such as yoghurt into the vagina.
  3. Have symptoms assessed rather than starting a probiotic on your own.
  4. Stop smoking if you smoke.
  5. Use antibiotics or antifungals only when there is evidence of infection or another specific indication.
  6. Do not use a probiotic as self-prescribed IVF preparation. If you are considering one, discuss the specific product and clinical reason with your clinician; an IVF benefit has not been established.

Studies have tested lactobacilli placed in the vagina and other probiotics for bacterial vaginosis (BV). Even so, guidance from the US Centers for Disease Control and Prevention (CDC) concludes that “overall, no studies support these products as an adjunctive or replacement therapy for women with BV.”

One small randomised study of women treated for BV found that adding oral probiotics to metronidazole did not improve cure rates. It was not an IVF trial, so it cannot show that probiotics improve implantation or live birth (Qi et al., 2023).

A randomised, placebo-controlled trial is now studying whether oral probiotics can shift a low-Lactobacillus vaginal profile in IVF patients. Only its protocol has been published, so there are no results yet (van Haren et al., 2025).

Evidence box: Qi et al. 2023 study details

The study randomised 67 women. Among women who did recover, more reached a Nugent score (a laboratory measure of vaginal bacteria) below 4 when probiotics were added. Women who were pregnant or planning pregnancy were excluded.

Dr. Aksoy’s Approach

Lactobacillus is protective in the vagina, but treating a low reading with probiotics is not a simple vitamin fix. Active bacterial vaginosis requires antibiotics, not probiotics.

The uterine cavity is a distinct environment with far fewer bacteria. Reaching a specific endometrial Lactobacillus percentage is not a standard IVF target. I treat diagnosed clinical conditions and the patient in front of me — not a commercial microbiome score.

Request a case review

If you have recurrent bacterial vaginosis, recurrent implantation failure, or a microbiome result you are unsure how to use, a structured review can help. It can separate symptoms that need standard testing from results that may not change treatment.

You can request a confidential case review before deciding on next steps.

Frequently Asked Questions for Dr. Aksoy

If bacterial vaginosis or a yeast infection is found shortly before embryo transfer, do you postpone the transfer?

When a vaginal infection is found just before embryo transfer, I decide according to the type of infection, how severe the symptoms are, and whether it is truly an active infection. If there is bacterial vaginosis with foul-smelling discharge and clear vaginal inflammation, I prefer to treat it first and, if needed, to delay the transfer briefly.

But when a patient has no complaints and a test simply shows a bacterial imbalance, that does not mean we must postpone the transfer. I think the same way about yeast infections. If there is itching, burning, and heavy discharge, I give the appropriate treatment; I generally do not treat Candida colonisation that causes no symptoms.

A delay of a few days is possible, especially with frozen embryo transfers, but there is no strong evidence that it improves the chance of pregnancy in every patient. What matters to me is not to overlook an active, clinically meaningful infection, and not to use antibiotics unnecessarily or delay the transfer just to correct a laboratory result.

If a patient is taking an oral or vaginal probiotic during IVF, do you ask her to stop?

I first ask my patients which product they are using and why. If it is an oral probiotic with known contents that the patient tolerates well, I do not insist on stopping it just because she has started IVF. But I also do not say it will improve embryo implantation or the chance of a live birth.

I am more careful with vaginal probiotics. Especially close to transfer day, I prefer not to use vaginal products that are not medically needed, because they can cause vaginal irritation and I want to avoid unnecessary applications around the transfer. Whether the patient is using vaginal progesterone also matters in this decision.

Today we do not have strong enough evidence that probiotics reliably improve IVF success. So I do not recommend them routinely, but I also do not automatically tell every patient who uses them that they are harmful. What matters is not to see probiotics as a proven infertility treatment, and not to use them in place of proper treatment when there is a real infection.

If an outside endometrial microbiome test shows low Lactobacillus, how do you approach it?

When an endometrial microbiome test done at another centre shows a low Lactobacillus percentage, I do not start antibiotics or probiotics on that result alone. These tests are not yet standardised enough to guide our clinical decisions reliably. Contamination of the uterine sample with vaginal bacteria, the analysis methods used, and results that can change over the menstrual cycle are all important problems. A low percentage also does not necessarily mean the embryo will fail to implant.

So I do not repeat the test or postpone the embryo transfer just to correct this report. The result does not diagnose chronic endometritis either. However, if there are repeated transfer failures, abnormal bleeding, or other findings that suggest chronic endometritis, I assess those separately with standard methods.

I tell my patients this in particular: a test being able to show the bacteria inside the uterus in detail does not mean we can use that information to raise the chance of pregnancy. At present, treating on a low Lactobacillus result alone has not been shown to increase the live-birth rate.

Evidence note: CDC guidance on bacterial vaginosis recommends treatment for women with symptoms, and its guidance on vulvovaginal candidiasis says that finding Candida without symptoms or signs is not a reason to treat. In a 2025 randomised, double-blind trial of 338 IVF patients with abnormal vaginal bacteria, clindamycin with or without a vaginal L. crispatus product did not raise clinical-pregnancy rates compared with placebo. ESHRE does not recommend microbiome profiling for recurrent implantation failure, NICE advises against endometrial microbiome analysis as an add-on for embryo transfer, and the UK fertility regulator, the HFEA, rates microbiome testing grey, meaning there is not enough good-quality evidence that it raises the chance of having a baby.

More questions about the vaginal microbiome

Should every IVF patient test the vaginal microbiome?

No. The European Society of Human Reproduction and Embryology (ESHRE) does not recommend vaginal or uterine microbiome profiling, even after recurrent implantation failure. Symptoms or recurrent bacterial vaginosis call for standard clinical assessment, not automatically for microbiome sequencing.

Is Lactobacillus crispatus always “good”?

L. crispatus is generally a favourable sign, because it is associated with a more stable, acidic vaginal environment. A single microbiome result still has to be read together with symptoms and fertility history.

Dysbiosis and chronic endometritis are separate clinical questions. Some inflammatory and bacterial findings have been studied in both, but a vaginal microbiome pattern neither proves nor diagnoses chronic endometritis. Suspected chronic endometritis is assessed separately with standard methods.

Should antibiotics be used before every embryo transfer?

No. Antibiotics should be used when there is evidence of infection or another specific indication. Routine, untargeted use is not a good strategy.

Sources

Next step

A question about your own case?

An article can set out the general picture, but not what applies to your own history. If you would like your situation looked at, you can send your questions and any previous reports to the medical team.

For privacy, please send only information needed for an initial reply. Ask the team which secure channel to use for medical reports or identity documents.

Request a medical review

Add as a Preferred Source on Google

You can add draksoyivf.com as one of your preferred health information sources on Google.

Add on Google
Dr. Senai Aksoy

Dr. Senai Aksoy studied and trained in France before returning to Turkey, where he was a member of the ICSI team at Sevgi Hospital, Ankara — the country's first ICSI centre (1994-95) — and a co-author on the first Turkish ICSI publications produced in collaboration with the Brussels Van Steirteghem group (Human Reproduction, 1996; PMID 8671323). He helped build the IVF programme at the American Hospital Istanbul and has been running his own fertility practice since 1998.

Verified profiles: PubMed ORCID LinkedIn

Instagram YouTube

The content has been created by Dr. Senai Aksoy and medically approved.

Before getting in touch

Please note that certain services, such as sperm donation, egg donation, surrogacy and gender selection (without medical necessity), are not permitted under Turkish law.

We would like to inform you that our clinic strictly adheres to these regulations and does not offer any of these services.

Continue

Privacy and measurement choices

Necessary storage keeps the site working. If you accept, Google Analytics and Microsoft Clarity may use optional cookies to help us understand site use and measure advertising. Personalized advertising remains off. If you decline, optional cookies stay off; Google Consent Mode may still send limited, cookieless measurement signals. You can change your choice later.

Privacy policy