Dysregulation of the vaginal microbiota in endometriosis and impact of endocrine disruptors: an exploratory case-control study

Background: Endometriosis is associated with chronic inflammation and immune dysregulation, yet its relationship with the vaginal microbiota remains poorly characterized. Emerging evidence suggests that endocrine-disrupting chemicals (EDCs) may further contribute to microbial imbalance. This study aimed to (1) characterize and compare vaginal microbiota composition between women with endometriosis and healthy controls, and (2) investigate associations between vaginal microbial profiles and self-reported EDC exposure.

This exploratory case-control study was conducted between January 2024 and June 2025 at Amiens University Hospital. This analysis includes 25 surgically confirmed endometriosis cases (stage III-IV), and 25 controls, all of whom completed the study questionnaire. Vaginal samples were analysed by culture, Nugent scoring, and quantitative PCR targeting 11 bacterial species with total 16S rRNA as endogenous reference. Group comparisons were performed using the Mann–Whitney U test and Fisher's exact test. EDC exposure was assessed by structured questionnaire in a survey cohort (n = 261; endometriosis n = 137, controls n = 124) which included the 50 women of the microbiota analysis.

Nugent scoring showed more intermediate-flora profiles in endometriosis (45.5% vs. 30.0%) and comparable bacterial vaginosis profiles (13.6% vs. 15.0%; p > 0.05). Total bacterial 16S biomass was comparable between groups (p = 0.17), validating the normalization. Quantitative PCR identified significant depletion of Atopobium vaginae (p = 0.001) and Lactobacillus crispatus (p = 0.018), alongside a significant increase of Lactobacillus gasseri (p = 0.008) in endometriosis, while total Lactobacillus genus abundance remained unchanged (p = 0.26), suggesting a qualitative rather than quantitative shift. In the survey cohort, insecticide diffuser use was more frequent among endometriosis patients (29% vs. 16%; p = 0.02).

This exploratory study identifies a qualitative restructuring of the vaginal Lactobacillus community in endometriosis, characterized by depletion of protective taxa and enrichment of a less protective species, distinct from classical bacterial vaginosis. By combining classical microbiology, molecular analysis and a self-reported exposure questionnaire in the same cohort, this study provides preliminary data on vaginal microbiota and potential EDC exposure in endometriosis, opening new perspectives for non-invasive diagnostic approaches and microbiota-targeted interventions. 1 Introduction Endometriosis is an estrogen-dependent chronic inflammatory disease affecting 10%–12% of women at reproductive age worldwide. Infertility is a major consequence, with approximately 50% of infertile women having endometriosis. Diagnostic delay of 3 to 10 years remains common due to the lack of non-invasive tools, and the condition is disproportionately represented among women undergoing in vitro fertilization. While this pathology is characterized by ectopic growth of endometrial-like tissue outside the uterine cavity (, ), this condition is associated with systemic inflammation, adhesions, fibrosis, and symptoms like pelvic pain and subfertility (–). Despite its prevalence, long-standing recognition (, ), and negative impact on quality of life (), its diagnosis remains challenging due to symptom heterogeneity. Recent guidelines advocate for non-invasive methods like imaging and biomarkers (, ), prompting investigations into alternative diagnostic approaches. The pathogenesis of endometriosis is multifactorial, with theories including retrograde menstruation, coelomic metaplasia, and stem cell involvement (, ). Emerging research highlights the role of the vaginal microbiota (), a low-diversity ecosystem dominated by Lactobacillus species (e.g., L. crispatus, L. iners, L. jensenii, L. gasseri), which maintain a protective acidic environment (pH 3.0–4.5) through lactic acid, H₂O₂, and bacteriocin production (, ). Dysbiosis, marked by reduced Lactobacillus dominance and increased pathogenic species, may contribute to chronic inflammation and oxidative stress which produce reactive oxygen species (ROS) that can potentially induce genome and epigenome disorders in endometriosis (–). Vaginal microbiota dysbiosis is also being investigated in gynecological oncology, notably in endometrial cancer, where higher vaginal alpha-diversity, enrichment of anaerobic species and depletion of Lactobacillus have been reported (, ). Yet, despite its central role in reproductive tract immunity and recent exploration in other gynecological diseases, the vaginal microbiota remains poorly characterized in endometriosis, representing a critical gap in current knowledge. In reproductive pathologies, it is well documented that endocrine-disrupting chemicals (EDCs), such as pesticides and industrial pollutant compounds, interfere with fertility (, 21), immune pathways and induce gut microbiota dysregulation (22, 23). Considering the similarities between gut and vaginal microbial ecosystems, it is hypothesized that EDCs could affect the vaginal microbiota. Chen et al. have reported that exposure to EDC reduces protective Lactobacillus strains and promote inflammation (24). EDCs can interfere with hormonal and immune pathways, potentially exacerbating endometriosis risk (25, 26). However, their specific impact on vaginal microbial imbalances in endometriosis remains largely unexplored. This study objectives are (i) characterizing the vaginal microbiota of women with surgically confirmed endometriosis compared with healthy controls using complementary microbiological approaches (culture, Nugent scoring, qPCR), (ii) determining whether endometriosis is associated with a distinct dysbiosis pattern, different from classical bacterial vaginosis and assessing lifestyle and to assess (iii) environmental exposures, including self-reported EDC use, to explore potential links between environmental factors and vaginal microbial restructuring 2 Methods 2.1 Study design and participants Women aged 18–45 years undergoing in vitro fertilization-embryo transfer were recruited between January 2024 and June 2025. Vaginal microbiota analyses were performed on 25 women with surgically and histologically confirmed endometriosis and 25 age-matched healthy controls. The planned recruitment target was 80 women with endometriosis and 80 controls. This article reports an exploratory analysis of the first 25 controls and 25 cases. Additional participants are being analysed separately. All participants included in the microbiota analysis had completed the questionnaire. Cases had surgically and histologically confirmed endometriosis staged III–IV by the revised ASRM classification (27). Controls were age-matched healthy volunteers with no history of endometriosis, chronic pelvic pain, or recent gynaecological infection. It should be noted that controls were not disease-free women but rather women seeking fertility treatment, who may present with other reproductive conditions unrelated to endometriosis. A flowchart summarizing screening, exclusions, and final enrolment is provided in Figure 1. Figure 1 2.2 Inclusion and exclusion criteria Eligible participants were women aged 18–45 years with surgically and histologically confirmed endometriosis (stages III-IV) presenting with symptoms such as dysmenorrhoea or pelvic pain. Women were excluded if they were pregnant or breastfeeding, had received systemic antibiotics, vaginal antimicrobials, probiotics or for the medical management of endometriosis (combined oral contraceptives, progestins, or GnRH analogues) within the preceding three months. This exclusion criterion did not apply to IVF-related gonadotrophin stimulation: vaginal sampling was performed before the initiation of ovarian stimulation, and no participant was under exogenous gonadotrophin or progesterone support at the time of sampling. 2.3 Ethics and consent The protocol was approved by the Institutional Ethics Committee of Amiens University Hospital (ID-RCB 2025-A02846-43). The questionnaire was reviewed and approved by the Institutional Ethics Committee of Université de Picardie Jules Verne (ID-RCB number: 2024-04-2) and was conducted in strict adherence to the principles outlined in the Declaration of Helsinki. Prior to enrolment, each participant was fully informed about the study's objectives and the role of the questionnaire. All participants provided written informed consent. Data handling complied with the Declaration of Helsinki and applicable data protection regulations. 2.4 Pre-sampling instructions and metadata Participants were instructed to avoid sexual intercourse, vaginal douching, and intravaginal products for 72 h prior to sampling. Sampling was scheduled outside active menstruation. Menstrual phase (follicular, ovulatory, luteal) and contraceptive status were recorded. Additional metadata collected included age, BMI, parity, smoking status, recent antibiotic use, and recent sexual activity. 2.5 Sample collection and handling Mid-vaginal swabs were collected using Sigma Transwab® flocked swabs (MWE, Corsham, Wilts, UK) in the following order: (1) swab for culture placed in Amies transport medium; (2) swab for molecular analysis placed in DNA stabilization buffer. Time from collection to laboratory processing, transport temperature, and storage conditions were recorded. Swabs for molecular analysis were stored at −20 °C until extraction (within 5 days). 2.6 Bacterial culture Swabs for culture were processed within 4 h of collection. Samples were inoculated onto blood agar and chocolate agar (incubated at 37 °C in 5% CO₂ for 24–48 h), MacConkey agar (aerobic, 24–48 h), and Sabouraud agar (30 °C, up to 72 h). Anaerobic cultures were incubated in an anaerobic chamber or jar for up to 72 h. Growth was reported as presence/absence and semi-quantitatively (scant, moderate, heavy); where quantitative plating was performed, colony counts were expressed as CFU/mL. Reference strains and internal quality controls were included on each run. 2.7 Phenotypic and biochemical identification Colonies were examined macroscopically and Gram-stained. Identification was performed using the Vitek 2 Compact system (bioMérieux, Marcy-l’Étoile, France) with GP/GN cards and ambiguous results were confirmed by API 20E (bioMérieux Marcy-l’Étoile, France) or MALDI-TOF (Bruker) when available. Species-level identifications were accepted at confidence scores ≥90%. 2.8 Microscopic evaluation using nugent scoring Gram-stained smears were scored according to Nugent criteria (0–10). All slides were evaluated by a single experienced microbiologist blinded to clinical status. Discrepant or ambiguous smears were reviewed by the same reader a second time. 2.9 DNA extraction and storage DNA was extracted using the Maxwell® RSC Buccal Swab DNA Purification Kit (Promega, Madison, WI, USA) with mechanical disruption using 200 µg of zircon beads in a FastPrep-24™ 5G (M.P. Biomedicals, LLC, Santa Ana, CA, USA), and the following modifications: bead-beating for 3 × 30 s at 6.0 m/s, proteinase K incubation at 56 °C for 20 min, and elution in 60 µL. Extraction blanks were included in each batch to monitor contamination. DNA concentration and purity were measured using a TECAN Infinite 200 Pro (Tecan Life Sciences, Switzerland). 2.10 Microbiological targets and rationale Bacterial targets were selected a priori based on established ecological and clinical relevance within the vaginal microbiota. Four Lactobacillus species (L. crispatus, L. iners, L. gasseri, L. jensenii) were included because they define the major Lactobacillus dominated community state types and differ in their protective functions (28). Gardnerella vaginalis and Atopobium (Fannyhessea) vaginae were selected as hallmark taxa of bacterial vaginosis and Nugent intermediate flora. Fusobacterium spp. was included due to recent reports of endometrial infiltration and increased vaginal prevalence in endometriosis (29). Finally, Enterococcus spp., Streptococcus spp., and Escherichia coli were selected as genera colonizing both the intestinal and genital tracts, to explore potential gut-vagina connections. Because the vaginal microbiota is dynamic and influenced by menstrual cycle phase, sexual activity, and hormonal status, sampling was scheduled outside menstruation, and standardized pre sampling instructions were provided. 2.11 qPCR and molecular assays qPCR assays targeted total bacteria and species-/genus-specific 16S regions for key vaginal strains (primer sequences and assay details are available in Table 1). Reactions were run in technical triplicate with no-template controls and extraction blanks on each plate. Standard curves from serial dilutions of genomic standards were used to determine assay efficiency, linear range, and limits of detection. Relative abundances were calculated using the ΔΔCt method with total bacterial 16S (All Eubacteria) as the reference and the mean of the control group as the calibrator. Results are reported as fold-change 2−ΔΔCt relative to controls. Table 1 | Bacteria phylum/species | Primer | Nucleotide Sequences (5′-3′) | Source | |---|---|---|---| | All Eubacteria | UniF | F: GTGSTGCAYGGYYGTCGTCA | (39) doi: 10.1017/S0007114508019880 | | UniR | R: ACGTCRTCCMCNCCTTCCTC | || | Lactobacillus spp. | 16SrRNA-F | F: AGCAGTAGGGAATCTTCCA | (40) doi: 10.1128/AEM.01895-09 | | 16SrRNA-R | R: CACCGCTACACATGGAG | || | L. crispatus | LcrisF | F: AGCGAGCGGAACTAACAGATTTAC | (33) doi: 10.1186/1471-2180-7-115 | | LcrisR | R: AGCTGATCATGCGATCTGCTT | || | L. iners | LInersF | F: GTCTGCCTTGAAGATCGG | | | LInersR | R: ACAGTTGATAGGCATCATC | || | L. jensenii | LABJENTR2F | F: CCTTAAGTCTGGGATACCATT | | | LABJENRT2R | R: ACGCCGCCTTTTAAACTTCTT | || | L. gasseri | LgassF | F: TGGAAACAGRTGCTAATACCG | | | LgassR | R: CAGTTACTACCTCTATCTTTCTTCACTAC | || | A. vaginae | ATOVAGRT3F | F: GGTGAAGCAGTGGAAACACT | | | ATOVAGRT3R | R: ATTCGCTTCTGCTCGCGCA | || | G. vaginalis | F-GV1 | F: TTACTGGTGTATCACTGTAAGG | | | R-GV3 | R: CCGTCACAGGCTGAACAGT | || | Enterococcus spp. | EnterocF | F: AGAGAGTAAGGTCCGATTGAAC | (41) doi: 10.1016/j.ajog.2018.02.012 | | EnterocR | R: GGTTGTTTCCCGTATTATGC | || | Streptococcus spp. | StrepF | F: GTACAGTTGCTTCAGGACGTATC | (42) doi: 10.12659/msm.902237 | | StrepR | R: ACGTTCGATTTCATCACGTTG | || | Fusobacterium spp. | 16SrRNA-F | F: CCCTTCAGTGCCGCAGT | (43) doi: 10.3892/etm.2019.8237 | | 16SrRNA-R | R: GTCGCAGGATGTCAAGAC | || | Escherichia coli | ybbW-F | F: TGATTGGCAAAATCTGGCCG | (44) doi: 10.1016/j.watres.2017.08.032 | | ybbW-R | R: GAAATCGCCCAAATCGCCAT | Primer sequences and amplicon details for qPCR assays used in this study. Columns Nucleotide Sequences: Forward primer (5′→3′); Reverse primer (5′→3′). 2.12 Questionnaire and EDC exposure assessment A structured questionnaire (Supplementary File 1) developed for this study, drawing on questions used in prior epidemiological surveys of environmental and occupational chemical exposure, captured demographics, reproductive history, household and occupational chemical exposure, dietary habits, and lifestyle factors. The questionnaire was presented to 302 participants; 261 were retained for analysis (endometriosis n = 137; controls n = 124), after excluding 41 incomplete submissions. All 50 women included in the vaginal microbiota analysis completed the questionnaire, and their responses were linked to their clinical records. The questionnaire data (n = 261) and the microbiota data (n = 50) are presented as two complementary analyses. Questionnaire data were anonymized prior to analysis. EDC-related variables collected through the questionnaire were self-reported and descriptive only, and no analytical verification or quantitative exposure assessment was performed. 2.13 Blinding and quality control Laboratory personnel performing culture, Nugent scoring, DNA extraction, and qPCR were blinded to clinical groups. Extraction blanks and no-template qPCR controls were processed in parallel. Any strain detected in negative controls were reported and considered when interpreting low-abundance results. 2.14 Statistical analysis Analyses were performed in R (v4.5.2; R Core Team, 2025). Continuous variables were assessed for normality using the Shapiro–Wilk test. Given the small sample sizes and non-normal distributions characteristic of qPCR relative quantification data, group comparisons were performed using the Mann–Whitney U test (two-sided; α = 0.05) throughout. Results are expressed as median relative quantification (RQ) with interquartile range [Q1–Q3]. Fold-change was calculated as the ratio of median RQ in the endometriosis group to median RQ in the control group. For each qPCR target, samples were excluded based on pre-defined quality criteria applied independently to the target and reference amplicons: (1) fewer than two valid replicates (Ct 40 cycles; or (3) coefficient of variation of valid replicates > 5%. A second pass of outlier exclusion applied a robust Z-score filter [Z = 0.6745 × (RQ−Median)/MAD; threshold |Z| > 3.5] calculated separately per group and per target, to identify samples with extreme RQ values attributable to partial DNA degradation or pipetting errors (30). These criteria were defined a priori and applied uniformly across all targets. Categorical variables were compared using Fisher's exact test (two-sided). For questionnaire data, χ2 tests were used where expected cell counts were ≥5; Fisher's exact test was applied otherwise. Effect sizes are reported as fold-change for continuous qPCR data and as odds ratios with 95% confidence intervals for categorical comparisons. Given the exploratory nature of the study and the small sample sizes, p-values were not adjusted for multiple comparisons; all reported p-values are uncorrected and should be interpreted in this context. Effect sizes, 95% confidence intervals, and exact p-values are reported. A two-sided p < 0.05 was considered statistically significant. 3 Results 3.1 Vaginal microbiota analysis 3.1.1 Participant characteristics After screening and exclusions (Figure 2), the study analysed vaginal samples from 25 women with surgically and histologically confirmed endometriosis and 25 age-matched controls (total n = 50). The mean age was 33 years in the endometriosis group and 31 years in the control group. All participants were in the follicular phase (>day 7) at recruitment. Due to per-assay losses, culture and Nugent scoring were available for 42 samples (endometriosis n = 22; controls n = 20). DNA was successfully extracted from all 50 vaginal swabs (endometriosis n = 25; controls n = 25). Two extracts (one per group) did not meet spectrophotometric purity and yield acceptance criteria after re-extraction and were excluded from quantitative analyses; therefore, quantitative qPCR results are based on 48 extracts (endometriosis n = 24; controls n = 24). Due to variable amplification efficiency and DNA availability, the number of samples included per target ranged from 3 to 21 per group after per-target quality filtering. Demographic and questionnaire data used for EDC analyses derive from the larger survey cohort (n = 261) described below. Figure 2 3.1.2 Bacterial culture Culture was attempted on vaginal swabs from all enrolled participants (endometriosis n = 25; controls n = 25). Viable bacteria were recovered from 42/50 swabs: 22/25 (88%) in the endometriosis group and 20/25 (80%) in controls. Lactobacillus spp. were isolated less frequently in the endometriosis group than in controls (22 vs. 20 positive cultures for any viable bacteria). Potentially pathogenic species including Mobiluncus spp., Enterococcus spp., and Escherichia coli were more commonly recovered from endometriosis samples. The difference in overall culture recovery between groups was not statistically significant (Fisher's exact test; p = 0.702). 3.1.3 Nugent scoring Gram-stained smears were available for 42 samples (endometriosis n = 22; controls n = 20) and were scored using the Nugent system (0–10). The distribution of Nugent categories by group is shown in Figure 2. In controls, 11/20 (55.0%) had normal flora (Nugent 0–3), 6/20 (30.0%) had intermediate flora (4–6), and 3/20 (15.0%) had scores consistent with bacterial vaginosis (7–10). In the endometriosis group, 9/22 (40.9%) had normal flora, 10/22 (45.5%) intermediate, and 3/22 (13.6%) bacterial vaginosis. Comparisons by Fisher's exact test did not reach statistical significance in this cohort. The prevalence of Nugent scores consistent with bacterial vaginosis was comparable between groups (3/22 vs. 3/20; Fisher's exact two-sided p = 1.00). The prevalence of normal, Lactobacillus-dominant profiles did not differ significantly (9/22 vs. 11/20; Fisher's exact two-sided p = 0.54) These data show a numerically higher proportion of intermediate-flora profiles in endometriosis (10/22 vs. 6/20; p = 0.35) and a lower proportion of Lactobacillus-dominant profiles (9/22 vs. 11/20; p = 0.54); however, the observed differences did not reach statistical significance in this cohort. 3.1.4 DNA yield and QC DNA was successfully extracted from all n = 50 vaginal swabs (endometriosis n = 25; controls n = 25). Spectrophotometric and fluorometric quality checks showed that two extracts (one endometriosis, one control) did not meet purity/yield acceptance criteria and were flagged for re-extraction. Re-extraction was attempted; when re-extraction failed to meet QC these two samples were excluded from quantitative qPCR analyses. Therefore, quantitative qPCR results reported below are based on 48 extracts (endometriosis n = 24; controls n = 24). 3.1.5 qPCR performance All qPCR assays met pre-defined plate quality thresholds (efficiency 90%–110%; R2 ≥ 0.99). No-template controls and extraction blanks were negative above the assay limit of detection in all accepted runs. Total bacterial 16S rRNA Ct values did not differ significantly between groups (median Ct: 16.7 in controls vs. 15.8 in endometriosis; Mann–Whitney U, p = 0.17; n = 17 controls/16 endometriosis), confirming comparable total bacterial biomass across groups and validating the use of relative abundance comparisons via the ΔΔCt method. 3.1.6 Relative abundance by ΔΔCt (fold-change relative to controls) Relative abundances were calculated using the 2−ΔΔCt method with total bacterial 16S rRNA as the endogenous reference and the mean ΔCt of included controls as the calibrator. Group comparisons were performed using the Mann–Whitney U test (two-sided; α = 0.05). Results are expressed as median RQ with interquartile range [Q1–Q3] and fold-change (FC = median RQ endometriosis/median RQ controls). Species-level analysis of vaginal Lactobacillus community composition revealed a specific restructuring pattern in endometriosis, characterized by depletion of two key protective species and relative enrichment of a less protective species (Figure 3). Figure 3 3.1.7 Lactobacillus community restructuring L. crispatus was significantly reduced in endometriosis (Mann–Whitney U, p = 0.018; median RQ: 0.133 vs. 0.001; FC = 0.009; n = 7/5; Figure 3A). L. crispatus is the species most strongly associated with vaginal health, producing both D- and L-lactic acid and maintaining a low protective pH. A. vaginae was markedly reduced in endometriosis (p = 0.001; median RQ: 0.057 vs. 0.013; FC = 0.023; n = 9/10; Figure 3B), representing the strongest quantitative signal in the dataset, an approximately 44-fold lower median abundance. L. gasseri was significantly more abundant in endometriosis (p = 0.008; median RQ: 0.105 vs. 1.145; FC = 10.95; n = 9/9; Figure 3C). While a member of the Lactobacillus genus, L. gasseri is considered less protective than L. crispatus, producing predominantly L-lactic acid with weaker antimicrobial activity. Total Lactobacillus genus abundance did not differ significantly between groups (p = 0.26; FC = 0.96; n = 17/16; Figure 3D), indicating that the specific changes above reflect qualitative community restructuring rather than a simple loss of Lactobacillus biomass. L. iners (p = 0.79; FC = 1.13; n = 9/9; Figure 3E) and L. jensenii (p = 0.48; FC = 0.66; n = 6/4; Figure 3F) did not differ significantly between groups. Due to the high per-target exclusion rate for G. vaginalis after quality filtering (n = 6/3), no formal statistical comparison was possible. 3.1.8 Gut-associated bacterial genera Given published evidence of endometrial Fusobacterium infiltration in endometriosis, we investigated whether a vaginal signature could be identified in our cohort. We additionally examined three genera known to colonize both the gut and vaginal tract (Staphylococcus, Enterococcus, E. coli) as an exploratory gut-vaginal axis analysis. Fusobacterium spp. showed a numerically lower median relative abundance in endometriosis (median RQ: 0.022 vs. 0.731 in controls; FC = 0.030; p = 0.21; n = 9/3; Figure 4A). Staphylococcus spp. showed no significant difference between groups (p = 0.14; FC = 0.209; n = 6/8; Figure 4B). Enterococcus spp. (p = 0.40; n = 4/3; Figure 4C) and E. coli (p = 0.86; n = 11/7; Figure 4D) were similarly non-significant. All results should be considered exploratory given the small per-target n, particularly for Fusobacterium (n = 3 in endometriosis) and Enterococcus (n = 3 in endometriosis). Figure 4 Streptococcus spp. were significantly reduced in endometriosis (p = 0.025; FC = 0.13; n = 10/9). Given the exploratory nature of this study and the limited n. The biological significance of vaginal Streptococcus depletion in endometriosis is unclear and warrants investigation in larger cohorts. 3.2 Questionnaire and EDC associations 3.2.1 Participant characteristics Our study included 261 participants (137 with endometriosis, 124 controls). In Table 2, significant age distribution differences were observed (p < 0.001), with the control group skewed younger (35% aged 18–24 vs. 7% in the endometriosis group). Geographic distribution also differed significantly (p < 0.001), with higher representation from Nord (44%) and Pas-de-Calais (27%) in the endometriosis group. Notably, insecticide diffuser use was more common in the endometriosis group (29% vs. 16%, p = 0.02). Dietary adaptations were more frequent among endometriosis patients (34% vs. 9% of controls), with specific avoidance of gluten, lactose, or other allergens. Dairy consumption patterns also diverged (p < 0.05): 13% of the endometriosis group avoided dairy entirely (vs. 2% of controls), while controls consumed dairy more frequently. Table 2 | Criteria | Endometriosis (n = 137) | Controls (n = 124) | p-value | |---|---|---|---| | Age Distribution | <0.001 | || | 18–24 | 10 (7%) | 44 (35%) | | | 25–30 | 35 (26%) | 14 (11%) | | | 31–35 | 28 (20%) | 9 (7%) | | | 36–40 | 33 (24%) | 15 (12%) | | | 41–45 | 16 (12%) | 10 (8%) | | | 46–50 | 9 (7%) | 11 (9%) | | | 51–60 | 6 (4%) | 15 (12%) | | | >60 | 0 | 6 (5%) | | | Department Distribution | <0.001 | || | Aisne | 9 (7%) | 3 (2%) | | | Oise | 18 (13%) | 9 (7%) | | | Somme | 13 (9%) | 46 (37%) | | | Nord | 60 (44%) | 9 (7%) | | | Pas-de-Calais | 37 (27%) | 57 (46%) | | | Insecticide Diffuser Use | <0.02 | || | Yes | 40 (29%) | 20 (16%) | | | Adapted Diet | <0.05 | || | Any Adaptation | 46 (34%) | 11 (9%) | | | Gluten-free | 21 (15%) | 3 (2%) | | | Lactose-free | 19 (14%) | 2 (2%) | | | Anti-inflammatory diet | 13 (9%) | 0 | | | Fodmap diet | 3 (2%) | 0 | | | Egg allergy | 5 (4%) | 1 (1%) | | | Other diets | 2 (1%) | 1 (1%) | | | Dairy Product Consumption | <0.05 | || | Several times a day | 11 (8%) | 28 (23%) | | | Daily | 60 (44%) | 66 (53%) | | | Occasionally | 48 (35%) | 27 (22%) | | | Never | 18 (13%) | 3 (2%) | | | Tobacco consumption | 0.3 | || | Yes | 26 (19%) | 17 (14%) | | | Cigarette per day | 0.6 | || | 10 | 10 (7%) | 8 (6%) | | | Electronic cigarette | 0.5 | || | Yes | 11 (8%) | 7 (6%) | | | Date of diagnosis | ||| | <6 months | 10 (7%) | N/A | | | <1 year | 15 (11%) | N/A | | | <2 years | 13 (9%) | N/A | | | <3 years | 22 (16%) | N/A | | | 5 years | 28 (20%) | N/A | | | >10 years | 27 (20%) | N/A | Demographic and behavioural characteristics of women with and without endometriosis in the French region of Hauts-de-France. 4 Discussion This exploratory study identified a specific pattern of vaginal microbiota alteration in women with endometriosis, characterized by a significant reduction of A. vaginae and L. crispatus and a significant increase of L. gasseri, consistent with a qualitative restructuring of the Lactobacillus community. A numerically higher prevalence of intermediate-flora profiles (Nugent 4–6) was additionally observed in the endometriosis group (45.5% vs. 30.0%; p = 0.35), with no increase in bacterial vaginosis-consistent profiles (13.6% vs. 15.0%; p = 1.00). An exploratory reduction of Streptococcus spp. (p = 0.025) was also detected, the biological significance of which remains unclear. These results are consistent with prior reports of microbial dysbiosis in endometriosis (31, 32) while also revealing patterns that diverge from classical bacterial vaginosis signatures. 4.1.1 Lactobacillus profiles and community stability Total Lactobacillus genus abundance did not differ significantly between groups (Mann–Whitney U, p = 0.26; FC = 0.96), indicating that overall Lactobacillus population was broadly comparable. However, species-level analysis revealed divergent patterns within this genus. L. crispatus was significantly reduced in the endometriosis group (p = 0.018; FC = 0.009). L. crispatus is the species most strongly associated with vaginal health: it produces D- and L-lactic acid, maintains a low pH, and is inversely associated with BV and pelvic inflammatory disease (, 28). Its specific depletion against a background of unchanged total Lactobacillus signal suggests a qualitative shift in community composition rather than a simple loss of Lactobacillus population. In contrast, L. gasseri was significantly more abundant in the endometriosis group (p = 0.008; FC = 10.95). While a member of the Lactobacillus genus, L. gasseri is considered less protective than L. crispatus: it produces predominantly L-lactic acid, has weaker antimicrobial activity, and is associated with more transitional community states (33). An inverse relationship between L. gasseri and L. crispatus has been previously reported, and our data are consistent with a competitive replacement of the more protective species. L. iners showed no significant difference (p = 0.79; FC = 1.13). L. jensenii did not differ significantly (p = 0.48; FC = 0.66), though a minor numerical trend toward lower abundance in endometriosis was observed. 4.1.2 A. vaginae depletion and divergence from classical BV A. vaginae was markedly and significantly reduced in the endometriosis group (p = 0.001; FC = 0.023), representing the strongest quantitative signal in the molecular dataset. This finding is, at first sight, paradoxical: A. vaginae is typically considered a BV-associated organism, yet our cohort showed a higher prevalence of intermediate-flora profiles in endometriosis (45.5% vs. 30.0%, ns). This apparent contradiction may be reconciled by the polymicrobial nature of BV: the Nugent score does not distinguish A. vaginae from other BV-associated anaerobes such as Mobiluncus spp., Prevotella spp., or Sneathia spp. In endometrial cancer, higher vaginal alpha-diversity and enrichment of anaerobes such as Prevotella, Anaerococcus and Mobiluncus have been reported (), although the comparison is limited by the marked differences in age and menopausal status between cohorts (mean age of 67 years in the endometrial cancer group). A such, intermediate Nugent profile with simultaneously reduced A. vaginae suggests that other anaerobic organisms may contribute to the dysbiotic signal, a pattern distinct from canonical BV and potentially specific to the endometriosis microenvironment. 4.1.3 Gut-associated and opportunistic genera A landmark study by Muraoka et al. (29) demonstrated that Fusobacterium infiltrates the endometrium in 64% of women with endometriosis but fewer than 10% of controls, and identified a pathogenic mechanism whereby Fusobacterium is facilitating the establishment of endometriotic lesions. The same study also found higher Fusobacterium prevalence in vaginal swabs of affected patients, providing direct rationale for our vaginal qPCR analysis. In our cohort, Fusobacterium spp. showed a numerically lower median abundance in endometriosis (FC = 0.030; p = 0.21), but this result is based on only a few valid endometriosis samples after quality filtering, critically insufficient to draw conclusions. The direction of the trend (higher in controls) is at odds with the Muraoka et al. findings, possibly reflecting true absence of vaginal enrichment, a type II error due to underpowering, or differential quality filtering between groups. A dedicated study with targeted Fusobacterium qPCR and larger n is required to clarify whether vaginal Fusobacterium enrichment accompanies the endometrial signature described by Muraoka et al. Regarding the other gut-associated bacteria, Streptococcus spp. were significantly less abundant in the endometriosis group (p = 0.025; FC = 0.13). Enterococcus spp. did not differ significantly after objective quality filtering (p = 0.40). Fusobacterium spp. similarly showed no significant between-group difference (p = 0.21). 4.1.4 Possible biological interpretations First of all, total bacterial biomass did not differ significantly between groups (p = 0.17), indicating that the four significant findings reported here reflect genuine shifts in vaginal microbial community composition rather than a non-specific reduction in overall bacterial load. This validates the use of the ΔΔCt normalization approach and strengthens the biological interpretation of the species-level differences observed. Taken together, three consistent findings, which are depletion of L. crispatus and A. vaginae, and the increase of L. gasseri, describe a specific qualitative restructuring of the vaginal microbial community in endometriosis. An additional exploratory signal of Streptococcus spp. reduction (p = 0.025) was observed but cannot be biologically interpreted at this stage and warrants investigation in larger cohorts. Rather than a simple loss of all protective bacteria, the data suggest a qualitative shift: replacement of the most protective lactobacillus species (L. crispatus) by a less protective congener (L. gasseri), alongside depletion of A. vaginae. Biologically, such changes could arise from altered local conditions in endometriosis inducing shifts in vaginal pH, mucosal secretions, oestrogen-mediated glycogen availability, immune mediators, or epithelial barrier function. The cross-sectional design prevents causal inference; microbiota alterations may be a cause, consequence, or epiphenomenon of disease. A further mechanistic route linking environmental exposure to the genital ecosystem is the estrobolome, the collection of gut bacterial genes involved in oestrogen deconjugation. Bacterial β-glucuronidase activity can reactivate hepatically inactivated oestrogens in the intestinal lumen, enabling their reabsorption and modulating systemic oestrogen availability (34, 35). Because endometriosis is an oestrogen-dependent disease and vaginal epithelial glycogen (which conditions Lactobacillus colonisation) is oestrogen-sensitive, perturbation of the estrobolome could theoretically influence both lesion maintenance and vaginal community composition (). EDCs can alter gut microbial communities and thus may indirectly affect oestrogen recycling via this axis, in addition to any direct receptor-mediated effects. 4.1.5 EDCs exposures and mechanistic hypotheses Self-reported indicators of EDC exposure, notably insecticide diffuser use, were more common among participants with endometriosis in the survey cohort (29% vs. 16%, p = 0.02). These variables should be interpreted cautiously, as no analytical confirmation of exposure or association testing was performed. Although our questionnaire data are descriptive and circumstantial, this observation aligns with emerging reports linking occupational or environmental chemical exposures to elevated endometriosis risk (36–38). If confirmed by objective biomonitoring, EDCs could plausibly influence the vaginal ecosystem through endocrine modulation, immune dysregulation, or indirect effects mediated by the gut microbiota and mucosal barrier integrity (23). Mechanistic pathways worth testing include EDC-driven changes in local estrogen signalling, altered antimicrobial peptide expression, shifts in mucosal glycogen availability, and immune cell recruitment that together could reduce L.crispatus colonization or stability (). These hypotheses remain speculative here because exposure data are self-reported and cross-sectional; prospective biomonitoring and experimental models are required to test causality. The geographic distribution of survey respondents differed significantly between groups (p < 0.001), with higher response rates from endometriosis patients in Nord (44%) and Pas-de-Calais (27%). This distribution reflects differential questionnaire response rates across departments rather than a designed spatial sampling strategy and should be interpreted with caution. Nevertheless, this data is contextually noteworthy in the context of Hauts-de-France, a region known for its intense agricultural activity, with heavy pesticide use. Several studies have suggested a potential link between chronic exposure to pesticides and the risk of endometriosis. In France, large quantities of EDCs are used in agriculture, particularly in the form of pesticides. In 2017, 68,000 tonnes were purchased, placing France first in Europe and third worldwide. EDCs are recognised for their genotoxicity and neurotoxicity, notably affecting brain development (). This study provides valuable insights into the relationship between endometriosis and exposure to endocrine disruptors. These findings highlight the need for further research to elucidate the underlying mechanisms and to develop effective prevention and treatment strategies. 4.1.6 Limitations and perspectives Although exploratory in nature, this study provides an important first step toward characterizing vaginal microbiota alterations in endometriosis. The sample size (n = 50) was intentionally designed to enable detailed molecular and classical microbiological analyses; however, larger, adequately powered cohorts will be essential to validate these findings and detect more subtle microbial shifts. Our study population was drawn from the Hauts-de-France region, an area with intensive agricultural activity and elevated pesticide use. While this may limit generalizability, it also offered a unique opportunity to investigate microbiota patterns in a population with potentially higher environmental exposure, thereby strengthening the rationale for examining endocrine-disrupting chemicals in relation to reproductive health. EDC exposure was assessed through structured self-reported questionnaires, which provided valuable insight into lifestyle and environmental factors but did not allow objective quantification of exposure. Future studies incorporating biomonitoring approaches (e.g., urinary or serum measurements) will be critical to refine exposure–microbiota associations. Likewise, our targeted qPCR strategy enabled precise quantification of key vaginal taxa, but broader community profiling using high-throughput sequencing will be necessary to capture the full complexity of microbial restructuring in endometriosis. Despite these considerations, our findings open several promising avenues for future research. Ongoing work in our laboratory includes high-throughput 16S rRNA gene sequencing of the endometrial microbiota to identify disease-specific microbial signatures, alongside an investigation of the gut-vaginal-endometrial microbiota axis and its role in endometriosis, as well as the evaluation of circulating cell-free DNA (cfDNA) as a potential non-invasive biomarker for endometriosis diagnosis and monitoring. 5 Conclusion This exploratory study identifies a specific pattern of vaginal microbiota restructuring in women with endometriosis: depletion of the keystone protective species L. crispatus and A. vaginae and increase of the less protective L. gasseri. An exploratory reduction of Streptococcus spp. was additionally observed and warrants confirmation in larger studies. Importantly, the reduction of A. vaginae in a context of increased intermediate-flora profiles argues against a classical BV pattern and suggests a distinct form of microbial dysbiosis specific to endometriosis. While hypothesis-generating, these results provide a molecular foundation for larger studies incorporating full metagenomics, objective EDC biomonitoring, and comprehensive clinical phenotyping. Abbreviations are listed in alphabetical order. All abbreviations are defined at first use in the manuscript text. 7 Additional requirements 7.1.1 Ethical approval and consent to participate For additional requirements for specific article types and further information please refer to ‘Article types’ on every Frontiers journal page. This study was carried out at the Amiens University Hospital, in Reproductive Medicine Department. ‘PESTIENDOMICRO’ protocol was approved by Institutional Ethical Committee on 28/02/2026 (ID-RCB number: 2025-A02846-43). The questionnaire was reviewed and approved by the Institutional Ethics Committee of Amiens University Hospital (ID-RCB number: 2024-04-2) and was conducted in strict adherence to the principles outlined in the Declaration of Helsinki. Prior to enrolment, each participant was fully informed about the study's objectives and the role of the questionnaire. Participation in this study was entirely voluntary, with individuals free to withdraw at any point without facing any consequences. Prior to enrolment, participants received a comprehensive explanation of the study's objectives, procedures, potential risks, and anticipated benefits. To ensure privacy, all data collected were anonymized, and no information was retained that could directly or indirectly identify any participant. The study protocol was designed to prevent any linkage between personal identities and the anonymized data, meaning no correspondence table or identifying code was used. As a result, the data are not classified as personal under applicable data protection standards All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee (ID-RCB number: 2025-A02846-43) and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Statements Data availability statement The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation. Ethics statement The studies involving humans were approved by Institutional Ethics Committee of Amiens University Hospital (ID-RCB 2025-A02846-43). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study. Author contributions ST-Z: Writing – original draft, Resources, Writing – review & editing, Software, Investigation, Formal analysis, Visualization, Funding acquisition, Data curation, Validation, Project administration, Methodology, Conceptualization, Supervision. MK: Writing – review & editing, Investigation, Writing – original draft, Software, Validation, Data curation, Visualization, Resources, Methodology, Formal analysis. AV: Investigation, Writing – review & editing, Writing – original draft. MS: Data curation, Investigation, Formal analysis, Writing – review & editing, Writing – original draft. ZH: Writing – original draft, Data curation, Software, Conceptualization, Validation, Writing – review & editing, Formal analysis. CN: Supervision, Project administration, Writing – review & editing, Writing – original draft, Resources. AC: Validation, Writing – review & editing, Writing – original draft. RC: Writing – original draft, Supervision, Validation, Resources, Funding acquisition, Project administration, Writing – review & editing. VB: Supervision, Writing – review & editing, Funding acquisition, Project administration, Validation, Writing – original draft, Resources, Visualization, Methodology. FL: Software, Resources, Writing – review & editing, Writing – original draft, Funding acquisition, Methodology, Project administration. MB: Methodology, Project administration, Formal analysis, Validation, Visualization, Data curation, Supervision, Writing – review & editing, Conceptualization, Writing – original draft, Funding acquisition, Software, Investigation, Resources. HK-C: Methodology, Supervision, Formal analysis, Data curation, Resources, Writing – original draft, Writing – review & editing, Investigation, Visualization, Conceptualization, Validation, Project administration, Software, Funding acquisition. Funding The author(s) declared that financial support was received for this work and/or its publication. Sophian Tricotteaux-Zarqaoui received a doctoral fellowship from the French Region Haut-de-France (50%), the Mutualité Sociale Agricole of Picardie (25%) and ATL Laboratories (25%). Acknowledgments The authors are deeply grateful to all the women who generously provided vaginal samples and completed our survey, making this research possible. Special thanks are extended to the research administration team at Picardy Jules Verne University for their guidance and assistance in securing ethical approval for the study. Conflict of interest The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Generative AI statement The author(s) declared that generative AI was not used in the creation of this manuscript. Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us. Publisher’s note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. Supplementary material The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/frph.2026.1978124/full#supplementary-material Abbreviations ASRM, American Society for Reproductive Medicine; BV, Bacterial vaginosis; BMI, Body mass index; cfDNA, Circulating cell-free DNA; CFU, Colony-forming unit; DNA, Deoxyribonucleic acid; ΔΔCt, Delta-delta cycle threshold (relative quantification method); EDC, Endocrine-disrupting chemical; EDCs, Endocrine-disrupting chemicals; ESHRE, European Society of Human Reproduction and Embryology; FC, Fold-change; GnRH, Gonadotrophin-releasing hormone analogue; IQR, Interquartile range; MAD, Median absolute deviation; MALDI-TOF, Matrix-assisted laser desorption/ionization time-of-flight (mass spectrometry); PCR, Polymerase chain reaction; qPCR, Quantitative polymerase chain reaction; QC, Quality control; RNA, Ribonucleic acid; ROS, Reactive oxygen species; RQ, Relative quantification; 16S rRNA, 16S ribosomal ribonucleic acid.

1. BurneyROGiudiceLC. Pathogenesis and pathophysiology of endometriosis. Fertil Steril. (2012) 98(3):511–9. 10.1016/j.fertnstert.2012.06.029PubMed PMID: 22819144. 2. ZondervanKTBeckerCMMissmerSA. Endometriosis. N Engl J Med. (2020) 382(13):1244–56. 10.1056/NEJMra1810764 3. RižnerTL. Estrogen metabolism and action in endometriosis. Mol Cell Endocrinol. (2009) 307(1–2):8–18. 10.1016/j.mce.2009.03.022 4. BulunSMonsavaisDPavoneMDysonMXueQAttarEet al. Role of estrogen receptor-β in endometriosis. Semin Reprod Med. (2012) 30:39–45. 10.1055/s-0031-1299596 5. SymonsLKMillerJEKayVRMarksRMLiblikKKotiMet al. The immunopathophysiology of endometriosis. Trends Mol Med. (2018) 24(9):748–62. 10.1016/j.molmed.2018.07.004 6. SampsonJA. Metastatic or embolic endometriosis, due to the menstrual dissemination of endometrial tissue into the venous circulation. Am J Pathol. (1927) 3(2):93–110.43. PubMed PMID: 19969738; PubMed Central PMCID: PMC1931779. Available online at: https://pmc.ncbi.nlm.nih.gov/articles/PMC1931779/ 7. NezhatCNezhatFNezhatC. Endometriosis: ancient disease, ancient treatments. Fertil Steril. (2012) 98(6):S1–62. 10.1016/j.fertnstert.2012.08.001 8. NnoahamKEHummelshojLWebsterPd’HoogheTDe Cicco NardoneFDe Cicco NardoneCet al. Impact of endometriosis on quality of life and work productivity: a multicenter study across ten countries. Fertil Steril. (2011) 96(2):366–373.e8. 10.1016/j.fertnstert.2011.05.090 9. GuptaDHullMLFraserIMillerLBossuytPMMJohnsonNet al. Endometrial biomarkers for the non-invasive diagnosis of endometriosis. Cochrane Database Syst Rev. (2016) 2016(4):CD012165. 10.1002/14651858.CD012165PubMed PMID: 27094925; PubMed Central PMCID: PMC6953323. 10. BeckerCMBokorAHeikinheimoOHorneAJansenFKieselLet al. ESHRE guideline: endometriosis†. Hum Reprod Open. (2022) 2022(2):hoac009. 10.1093/hropen/hoac009 11. SourialSTempestNHapangamaDK. Theories on the pathogenesis of endometriosis. Int J Reprod Med. (2014) 2014(1):179515–9. 10.1155/2014/179515 12. UzunerCMakJEl-AssaadFCondousG. The bidirectional relationship between endometriosis and microbiome. Front Endocrinol. (2023) 14:14. 10.3389/fendo.2023.1110824 13. MaBForneyLJRavelJ. The vaginal microbiome: rethinking health and diseases. Annu Rev Microbiol. (2012) 66:371–89. 10.1146/annurev-micro-092611-150157PubMed PMID: 22746335; PubMed Central PMCID: PMC3780402. 14. AmabebeEAnumbaDOC. The vaginal microenvironment: the physiologic role of lactobacilli. Front Med. (2018) 5. 10.3389/fmed.2018.00181 15. FredricksDNFiedlerTLMarrazzoJM. Molecular identification of bacteria associated with bacterial vaginosis. N Engl J Med. (2005) 353(18):1899–911. 10.1056/NEJMoa043802 16. SehringJBeltsosAJeelaniR. Human implantation: the complex interplay between endometrial receptivity, inflammation, and the microbiome. Placenta. (2022) 117:179–86. 10.1016/j.placenta.2021.12.015 17. BenkhalifaMMenoudPAPiquemalDHazoutJYMahjoubSZarquaouiMet al. Quantification of free circulating DNA and differential methylation profiling of selected genes as novel non-invasive biomarkers for endometriosis diagnosis. Biomolecules. (2025) 15(1):69. 10.3390/biom15010069 18. AquinoCINicosiaALigoriAVolpicelliAISuricoD. Microbiota status and endometrial cancer: a narrative review about possible correlations in affected versus healthy patients. Sci. (2024) 6(4):75. 10.3390/sci6040075 19. AquinoCILa VecchiaMPasolliESalaGLigoriABoldoriniRet al. Decoding the microbial landscape of endometrial cancer: a case-control study. BMC Microbiol. (2026) 26(1):492. 10.1186/s12866-026-05017-4 20. LahimerMAbou DiwanMMontjeanDCabryRBachVAjinaMet al. Endocrine disrupting chemicals and male fertility: from physiological to molecular effects. Front Public Health. (2023) 11. 10.3389/fpubh.2023.1232646 21. Tricotteaux-ZarqaouiSLahimerMAbou DiwanMCoronaACandelaPCabryRet al. Endocrine disruptor chemicals exposure and female fertility declining: from pathophysiology to epigenetic risks. Front Public Health. (2024) 12:12. 10.3389/fpubh.2024.1466967 22. PuCLiuYMaJHouLChengYZhangBet al. Bisphenol S exposed changes in intestinal microflora and metabolomics of freshwater crayfish, procambarus clarkii. Aquat Toxicol. (2024) 272:106957. 10.1016/j.aquatox.2024.106957 23. Abou DiwanMHuetAPoiriezJJoly CondetteCDelanaudSSevinEet al. Effects of chlorpyrifos on gut dysbiosis and barriers integrity in women with a focus on pregnancy and prebiotic intervention: insights from advanced in vitro human models. Environ Pollut. (2025) 367:125533. 10.1016/j.envpol.2024.125533 24. ChenPLiTYangXChenLGuoYChenPet al. Association between monoaromatic hydrocarbons exposure and vaginal microbiota disorder in recurrent implantation failure. J Hazard Mater Adv. (2023) 10:100320. 10.1016/j.hazadv.2023.100320 25. KuoCHYangSNKuoPLHungCH. Immunomodulatory effects of environmental endocrine disrupting chemicals. Kaohsiung J Med Sci. (2012) 28(7, Supplement):S37–42. 10.1016/j.kjms.2012.05.008 26. LiuZLuYZhongKWangCXuX. The associations between endocrine disrupting chemicals and markers of inflammation and immune responses: a systematic review and meta-analysis. Ecotoxicol Environ Saf. (2022) 234:113382. 10.1016/j.ecoenv.2022.113382 27. American Society for Reproductive Medicine. Revised American society for reproductive medicine classification of endometriosis: 1996. Fertil Steril. (1997) 67(5):817–21. 10.1016/S0015-0282(97)81391-X 28. RavelJGajerPAbdoZSchneiderGMKoenigSSKMcCulleSLet al. Vaginal microbiome of reproductive-age women. Proc Natl Acad Sci U S A. (2011) 108(supplement_1):4680–7. 10.1073/pnas.1002611107 29. MuraokaASuzukiMHamaguchiTWatanabeSIijimaKMurofushiYet al. Fusobacterium infection facilitates the development of endometriosis through the phenotypic transition of endometrial fibroblasts. Sci Transl Med. (2023) 15(700):eadd1531. 10.1126/scitranslmed.add1531 30. IglewiczBHoaglinDC. How to Detect and Handle Outliers. Milwaukee, Wis: ASQC Quality Press (1993). p. 87. (The ASQC basic references in quality control: statistical techniques; no. 16). 31. AtaBYildizSTurkgeldiEBrocalVPDinleyiciECMoyaAet al. The endobiota study: comparison of vaginal, cervical and gut microbiota between women with stage 3/4 endometriosis and healthy controls. Sci Rep. (2019) 9(1):2204. 10.1038/s41598-019-39700-6 32. ColonettiTSaggiorattoMCGrandeAJColonettiLJuniorJCDCerettaLBet al. Gut and vaginal microbiota in the endometriosis: systematic review and meta-analysis. BioMed Res Int. (2023) 2023(1):2675966. 10.1155/2023/2675966 33. De BackerEVerhelstRVerstraelenHAlqumberMABurtonJPTaggJRet al. Quantitative determination by real-time PCR of four vaginal lactobacillus species, gardnerella vaginalis and atopobium vaginae indicates an inverse relationship between L. Gasseri and L. Iners. BMC Microbiol. (2007) 7(1):115. 10.1186/1471-2180-7-115 34. FuhrmanBJFeigelsonHSFloresRGailMHXuXRavelJet al. Associations of the fecal microbiome with urinary estrogens and estrogen metabolites in postmenopausal women. J Clin Endocrinol Metab. (2014) 99:4632–40. 10.1210/jc.2014-2222 35. ErvinSMLiHLimLRobertsLRLiangXManiSet al. Gut microbial β-glucuronidases reactivate estrogens as components of the estrobolome that reactivate estrogens. J Biol Chem. (2019) 294(49):18586–99. 10.1074/jbc.RA119.010950PubMed PMID: 31636122. 36. SmarrMMKannanKLouisGMB. Endocrine disrupting chemicals and endometriosis. Fertil Steril. (2016) 106(4):959–66. 10.1016/j.fertnstert.2016.06.034PubMed PMID: 27424048. 37. RumphJTStephensVRArchibongAEOsteenKGBruner-TranKL. Environmental endocrine disruptors and endometriosis. Adv Anat Embryol Cell Biol. (2020) 232:57–78. 10.1007/978-3-030-51856-1_4PubMed PMID: 33278007; PubMed Central PMCID: PMC7978485. 38. CaporossiLCapannaSViganòPAlteriAPapaleoB. From environmental to possible occupational exposure to risk factors: what role do they play in the etiology of endometriosis?Int J Environ Res Public Health. (2021) 18(2):532. 10.3390/ijerph18020532 39. Ramirez-FariasCSlezakKFullerZDuncanAHoltropGLouisP. Effect of inulin on the human gut microbiota: stimulation of bifidobacterium adolescentis and faecalibacterium prausnitzii. Br J Nutr. (2008) 101(4):541–50. 10.1017/S0007114508019880 40. MartiRDabertPZiebalCPourcherAM. Evaluation of lactobacillus sobrius/L. Amylovorus as a new microbial marker of pig manure. Appl Environ Microbiol. (2010) 76(5):1456–61. 10.1128/AEM.01895-09 41. MorenoICicinelliEGarcia-GrauIGonzalez-MonfortMBauDVilellaFet al. The diagnosis of chronic endometritis in infertile asymptomatic women: a comparative study of histology, microbial cultures, hysteroscopy, and molecular microbiology. Am J Obstet Gynecol. (2018) 218(6):602.e1–602.e16. 10.1016/j.ajog.2018.02.012PubMed PMID: 29477653. 42. LinXChenXTuYWangSChenH. Effect of probiotic lactobacilli on the growth of streptococcus mutans and multispecies biofilms isolated from children with active caries. Med Sci Monit. (2017) 23:4175–81. 10.12659/MSM.902237PubMed PMID: 28851857. 43. WangHChenTChenYLuoTTanBChenHet al. Evaluation of the inhibitory effects of vaginal microorganisms on sperm motility in vitro corrigendum in/10.3892/etm.2025.12996. Exp Ther Med. (2019) 19(1):535–44. 10.3892/etm.2019.8237 44. WalkerDIMcQuillanJTaiwoMParksRStentonCAMorganHet al. A highly specific Escherichia coli qPCR and its comparison with existing methods for environmental waters. Water Res. (2017) 126:101–10. 10.1016/j.watres.2017.08.032 Summary

endocrine disruptor chemicals, endometriosis, reproductive health, vaginal dysbiosis, vaginal microbiota Citation Tricotteaux-Zarqaoui S, Kabbour M, Vandecandelaere A, Soudain M, Halbouty Z, Nadifi C, Cervino A, Cabry R, Bach V, Lahlou FA, Benkhalifa M and Khorsi-Cauet H (2026) Dysregulation of the vaginal microbiota in endometriosis and impact of endocrine disruptors: an exploratory case-control study. Front. Reprod. Health 8:1978124. doi: 10.3389/frph.2026.1978124 Received 25 August 2026 Revised 01 October 2026 Accepted 05 October 2026 Published 23 October 2026 Volume 8 - 2026 Edited by Maria Grazia Porpora, Sapienza University of Rome, Italy Reviewed by Alejandra García-Gasca, National Council of Science and Technology (CONACYT), Mexico Carmen Imma Aquino, University of Eastern Piedmont, Italy Updates Copyright © 2026 Tricotteaux-Zarqaoui, Kabbour, Vandecandelaere, Soudain, Halbouty, Nadifi, Cervino, Cabry, Bach, Lahlou, Benkhalifa and Khorsi-Cauet. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. *Correspondence: Hafida Khorsi-Cauet [email protected] † These authors share first authorship ‡ These authors share last authorship Disclaimer All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.