{"paper_id":"31cec61b-c303-4059-bf55-6f0d9617534e","body_text":"ARTICLE IN PRESS\nARTICLE IN PRESS\nhttps://doi.org/10.1038/s44294-026-00144-9\nReceived: 18 July 2025\nAccepted: 10 May 2026\nCite this article as: Rahou, I.,\nGehrmann, T., Ahannach, S. et al.\nCitizen science reveals comorbidities\nin endometriosis with no shift in\nvaginal microbiome composition. npj\nWomens Health (2026). https://\ndoi.org/10.1038/s44294-026-00144-9\nInas Rahou, Thies Gehrmann, Sarah Ahannach, Camille Nina Allonsius, Ilke De Boeck,\nTim Van Rillaer, Francesca Donders, Veronique Verhoeven, Gilbert Donders, Stijn\nWittouck & Sarah Lebeer\nWe are providing an unedited version of this manuscript to give early access to its\nfindings.  Before final  publication, the manuscript will undergo further editing. Please\nnote there may be errors present which affect  the content, and all legal disclaimers\napply.\nIf this paper is publishing under a Transparent Peer Review model then Peer\nReview reports will publish with the final  article.\n© The Author(s) 2026. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International\nLicense, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit\nto the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified  the licensed material. You do not\nhave permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this\narticle are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the\narticle's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain\npermission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.\nnpj Women's Health\nArticle in Press\nCitizen science reveals comorbidities in\nendometriosis with no shift in vaginal microbiome\ncomposition\n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nCitizen science reveals comorbidities in endometriosis with no shift in vaginal \nmicrobiome composition \nInas Rahou1, Thies Gehrmann1, Sarah Ahannach1,2, Camille Nina Allonsius1, Ilke De Boeck1, Tim Van \nRillaer1, Francesca Donders3,4, Veronique Verhoeven2,5, Gilbert GG Donders4,6, Stijn Wittouck1, Sarah \nLebeer1,2 *  \n1Laboratory of Microbiology & Biotechnology, Department of Bioscience Engineering, Antwerp University, \nGroenenborgerlaan 171, 2020 Antwerp, Belgium \n2U-MaMi Centre of Excellence, University of Antwerp, Antwerp, Belgium \n3Department of Gynaecology, Brugmann University Hospital, A. Van Gehuchtenplein 4, 1020 Brussel, Belgium \n4Femicare VZW, Clinical Research for Women, Gasthuismolenstraat 33, B-3300 Tienen, Belgium.  \n5Department of Family Medicine and Population Health, University Hospital of Antwerp, Drie Eikenstraat 655, 2650 \nEdegem, Belgium \n6Department of Obstetrics and Gynecology, Faculty of Medicine, University Hospital Antwerp, Wilrijkstraat 10, B-\n2650 Edegem, Belgium. \n \n*Corresponding author: sarah.lebeer@uantwerpen.be  \n \nKeywords \nEndometriosis, citizen science, vaginal microbiome, symptom profiles, comorbidities \n \n \n \n \n \n \n \n \n \n \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nAbstract \nEndometriosis is a chronic inflammatory condition affecting 2 –10% of reproductive -aged women, most \ncommonly presenting with pelvic pain and subfertility. While its impact on reproductive health is \nincreasingly recognized, the vaginal microbiome’s role in the pathogenesis of endometriosis is still unclear \nand a comprehensive map of potential co-occurring conditions remains underexplored. Leveraging data \nfrom the Isala citizen -science platform in Flanders (Belgium), we analy sed vaginal microbiome profiles \nobtained through 16S rRNA sequencing and health data from 95 women with self-reported endometriosis \nand 2,279 without. While no differences were observed in vaginal microbiome composition or diversity, \nwe identified significant associations between endometriosis and polycystic ovarian syndrome (OR = 2.92, \n95% CI 1.71-4.78, p < 0.001), migraine (OR = 3.75, 95% CI 1.38-8.60, p = 0.025), irritable bowel syndrome \n(OR=2.57, 95% CI 1.43 -4.36, p = 0.008) and dyspareunia ( OR = 1.67, 95% CI 1.14 -2.40, p = 0.033). These \nfindings suggest that the vaginal microbiome composition plays at most a limited role in endometriosis \nand highlights how citizen science can effectively complement clinical research by capturing \nunderrecognized comorbidities.  \n \n  \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nIntroduction \nEndometriosis is a chronic inflammatory condition that most commonly affects reproductive -aged \nwomen, with an estimated prevalence of 2-10%1–5. It is defined by the presence of endometrial-like tissue \noutside the uterus, with lesions varying in size, depth and location, affecting both reproductive and non -\nreproductive organs . Pelvic pain 6, subfertility7, gastrointestinal issues 8,9 and mental health problems 10 \nhave been frequently reported in different endometriosis cohorts across the world (with sizes between \n174 and 188,461), imposing a significant burden on both patient and society with substantial healthcare \ncosts11–13. Yet, a comprehensive map of the f ull range of symptoms and co -occurring conditions remains \nlacking.          \nThe pathogenesis of endometriosis is multifactorial and remains incompletely understood. Beyond the \nwidely accepted theory of retrograde menstruation 14, current understanding embraces several \ncomplementary hypotheses including coelomic metaplasia 15, vascular or lymphatic dissemination of \nendometrial cells 16, hormonal imbalances involving oestrogen and progesterone 17, immune \ndysregulation18 and (epi)genetic predisposition 19 – with chronic low -grade systemic inflammation as a \ncentral driver of disease progression 20–23. This heterogeneous pathophysiology is mirrored in the \nchallenges of diagnosis and treatment.  \nCurrent guidelines recommend transvaginal ultrasound and/or MRI as first-line imaging modalities due to \ntheir high sensitivity and specificity for deep and ovarian endometriosis, while laparoscopy followed by \nhistologic confirmation, an invasive and expensive procedure, i s only recommended for individuals with \nnegative imaging or unsuccessful empirical treatment 24. However, imaging techniques require expertise \nand frequently miss superficial peritoneal endometriosis, which accounts for approximately 80% of \nendometriosis subtypes25, highlighting the need  for improved non -invasive diagnostic tests. Current \ntherapeutic strategies include hormone -suppressive therapy, surgical removal of ectopic endometrial \nlesions and in some cases neuromodulation 24. However, these approaches are not curative and can lead \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nto various side effects; in the case of surgery, high risk of complications related to laparoscopic surgery \nand a risk of persisting chronic postsurgical pain 26,27; in the case of hormonal treatment, different side \neffects can be noted, most prevalent are depressive symptoms, vasomotor symptoms, weight gain, \nheadaches and bleeding irregularities 28–30. Together, this underscores the urgent need for non -invasive \ndiagnostics and novel treatment strategies.  \nRecently, the microbiome has emerged as a potential source f or diagnostic biomarkers. Bacterial DNA - \nand metabolite-based signatures have been shown to be implicated in the development and progression \nof several inflammatory conditions such as irritable bowel disease 31,32 and polycystic ovarian syndrome \n(PCOS)33. Emerging research from small-scale studies (n=35 in Jiminez et al., 2024; n=19 in Ata et al. 2019) \nindicate that endometriosis is associated with compositional differences in the microbiome across \nanatomical sites, including the  lower and upper reproductive tracts , as well as the gastrointe stinal \nsystem34 using 16S rRNA amplicon sequencing. At a functional level , reduced concentrations of gut \nmicrobiota-derived n-butyrate in feces were found in a mouse injection model of endometriosis compared \nto mice without endometriosis using targeted metabolomics 35. Others have reported altered levels of 4 -\nhydroxyindole in stool samples of individuals with endometriosis (n = 18) compared to healthy controls (n \n= 33) through untargeted metabolomics36.  \nA growing area of interest is the vaginal microbiome, which plays an essential role in reproductive health \nand is involved in several reproductive inflammatory conditions37,38. In most women of reproductive age, \nthe vaginal microbiome is predominantly composed of Lactobacillus species, such as Lactobacillus \ncrispatus and Lactobacillus iners, but also a substantial proportion of Gardnerella39. Lactobacilli appear \nbeneficial for maintaining a healthy vaginal environment based on various epidemiological 40,41 and lab-\nbased studies42–46, where their absence has been linked to an increased risk for various reproductive issues \nsuch as preterm birth 47–49, bacterial vaginosis 50, endometritis37  and aerobic vaginitis 38, mostly using 16S \nrRNA amplicon sequencing. Reduced levels of Lactobacillus and shifts towards a non -optimal vaginal \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nmicrobial community have been associated with inflammation51–53, potentially contributing to the chronic \ninflammatory environment characterist ic of endometriosis.   However, it remains unclear whether and \nhow the vaginal microbiome is associated with endometriosis.  \nRecently we have adopted a citizen -science approach to investigate the vaginal microbiome and general \nvaginal health of women in Fla nders (Belgium)39. This catalogue has enabled us in our present study to \nanalyse associations between the vaginal microbiome a nd endometriosis based on 16S rRNA amplicon \nsequencing of samples from 95 women with endometriosis and 2,279 women without endometriosis. \nFurthermore, to characterize the multifactorial nature of endometriosis and identify conditions that may \nreflect shared underlying mechanisms, we assessed a broad range of infections, clinical conditions, and \nvaginal symptoms through questionnaires in individuals with endometriosis (n = 144) and compared these \nwith women without endometriosis (n = 3,337).  \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nResults \nSubfertility affects up to 47% of individuals with endometriosis in a population-based cohort \nTo assess whether and how endometriosis symptom status influenced the vaginal microbiome \ncomposition and general health , participants with endometriosis were stra tified into two groups based \non their symptom status: individuals experiencing current and past endometriosis-related symptoms (n = \n66, n = 78). ‘Endometriosis-related symptoms’ in our study refer to the participant’s self-perceived disease \nburden, as the survey did not request specific symptom types and was not designed to capture \nendometriosis-related symptomatology.  Current endometriosis-related symptoms  were defined as \nburden experienced at the time of study participation, and past symptoms as burden th at had been \nexperienced previously but was no longer present when the questionnaire was completed. This subset of \nindividuals with endometriosis is not representative of the broader endometriosis population as the Isala \nproject targeted a predominantly healthy population and likely includes milder disease presentations. The \ncontrol group (n = 3,337) consisted of participants who reported not having endometriosis. To minimize \nthe presence of potential undiagnosed endometriosis cases in the control group, we excluded individuals \nwho reported dysmenorrh oea and requiring non -steroidal anti -inflammatory drugs (NSAIDs) during \nmenstruation. Menopausal individuals were also excluded from this study. Information on the location of \nendometriotic lesions and treatment history was not available because health data were only collected \nvia general surveys in  the Isala platform39. A detailed overview of the participants’ demographics and \nadditional variables are summarized in Table 1.  \nWithin the endometriosis group, individuals with past symptoms were significantly older with a median \nage of 36 (IQR 31-41) than those with current symptoms (31 [27-35], p < 0.001). Additionally, compared \nto the control group (2 8 [24-33]), individuals with endometriosis were significantly older regardless of \nsymptom status (p < 0.001 for past symptoms group, and p = 0.006 for current symptoms group). After \nadjusting for age, demographic, reproductive, and lifestyle characteristics did not significantly differ across \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nthe groups, except for subfertility and self-reported health. Subfertility was significantly more prevalent \nin individuals with endometriosis compared to the control group (p < 0.001).  Self-reported health was \nassessed through a five -point Likert -scale (1=very poor, 2=poor, 3=fair, 4=good, 5=very good)  with a \npoorer perceived health in individuals with current and past endometriosis-related symptoms compared \nto the control group ( p < 0.001; p = 0.003). Overall, after adjustment for age, endometriosis  symptom \nstatus was primarily associated with subfertility and poorer self-reported health, while other demographic \nand lifestyle characteristics were comparable across groups. \nEndometriosis shows no association with Lactobacillus dominance, genus-level diversity or abundance of \nindividual taxa in the vagina \nTo assess whether the vaginal microbiome composition is associated with endometriosis and its symptom \nstatus, vaginal microbiome profiles were obtained for 95 of the 144 participants with self -reported \nendometriosis, including 43 profiles from individuals with current endometriosis -related symptoms and \n52 from individuals with past symptoms. In comparison, 2,279 vaginal microbiome profiles were analysed \nfrom the control group (Figure 2a). Overall, 79.1% of participants with current endometriosis -related \nsymptoms had a Lactobacillus-dominated microbiota, compared to 71.2% in partici pants with past \nendometriosis-related symptoms and 80.1% in the control group, with no significant differences ( p = \n0.371). We applied different embedding methods including t-distributed stochastic neighbour embedding \n(t-SNE, Figure 2b), uniform manifold approximation and projection (UMAP, Supplementary Figure 1a) and \nprincipal coordinates analysis (PCoA, Supplementary Figure 1b), but no clear grouping of bacterial profiles \nwas observed based on current or past symptom status compared to the control group.  In addition, we \nexamined whether symptom status in participants with endometriosis  was associated with the \n(sub)genera present in at least 10% of all participants. Each test (Maaslin, Limma, DESeq, ANCOM-BC89, \nand a linear regression on the centered  log-ratio (CLR) transformed abundance data) was adjusted for \ntechnical confounders, age, recent sexual intercourse , number of pregnancies,  phase of the menstrual \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \ncycle and use of hormonal contraceptives, and associations were considered significant only if consensus \nwas reported by at least three differential abundance tools as described in Methods and Materials.  No \nsignificant associations were observed between endometriosis symptom status and relative abundance \nof specific microbial taxa. Similarly, no significant associations were observed between endometriosis and \ngenus-level diversity, including both alpha (Shannon) diversity and beta diversity, irrespective of symptom \nstatus. These findings confirm that the vaginal microbiome composition and its key f eatures play only a \nlimited role – if any – in the pathogenesis of endometriosis. \nPolycystic ovary syndrome shows strong co -occurrence with endometriosis, alongside other comorbid \nconditions \nComplementing the vaginal microbiome analyses and drawing on the uniquely rich survey data collected \nwithin the Isala programme, we next assessed comorbidities to better understand the broader health \nprofile associated with endometriosis.  To characterize the spectrum of comorbidities associated with \nendometriosis, we analysed extensive health questionnaires from individuals with endometriosis  \nreporting past or current endometriosis -related symptoms and compared them with controls with no \nreported endometriosis. The prevalence of several vaginal symptoms, urogenital infections and a range \nof conditions related to reproductive, metabolic and gastrointestinal health was compared between these \ngroups, with statistical significance determined after correction for multiple testing using the Benjamini–\nHochberg procedure.  A summary of the results is shown in Figure 2 with the corresponding counts \nprovided in Supplementary Table 1. \nDyspareunia was significantly more common among individuals with endometriosis, with an odds ratio of \n1.67 (95% C I 1.1 4–2.40; p = 0.033), as was irritable bowel syndrome (OR 2.57, 95% CI 1. 34–4.36,   \np = 0.008) and migraine (OR 3.75, 95% CI 1.38-8.60, p = 0.025). Interestingly, a significant co-occurrence \nof polycystic ovarian syndrome  in individuals with endometriosis was observed (OR 2. 92, 95% CI 1. 71–\n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n4.78, p < 0.001). By contrast, no significant associations were observed between endometriosis and \ncardiovascular, dermatological, or respiratory conditions, nor with urogenital infections or self-reported \nvaginal symptoms. None of the comorbid conditions associated with endometriosis showed a significant \nassociation with vaginal microbiome composition or diversity (Supplementary Figure 3). Together, these \nfindings underscore that , beyond the well -recognised menstrual -related pain, individuals with \nendometriosis experience a higher prevalence of several inflammatory comorbidities. This complexity \nhighlights the multifaceted nature of the disease and supports the need for integrated , multimodal \napproaches to clinical management. \nDiscussion \nDespite growing recognition of endometriosis as a complex and multifactorial condition affecting \nreproductive health, the potential role of the vaginal microbiome in the pathogenesis of endometrios is \nremains largely unexplored. In this study, we leveraged data from the citizen science-driven Isala platform \nto investigate whether endometriosis and its symptom status was  associated with differences in vaginal \nmicrobiome composition. Investigating the broad range of conditions in this population is a critical step \ntoward understanding potential shared pathologies with comorbidities and the underlying mechanisms \nof endometriosis. \nCompared with population ‑based studies reporting endometriosis prevalence of approximately 8 -\n10%54,55, the Isala cohort showed a lower overall prevalence of 3.5% . This may be explained by the setup \nof the Isala project, which was  designed to capture a general, predominantly healthy populatio n, as \ndefined by the World Health Organization (i.e., a state of complete physical, mental, and social well-being, \nnot merely t he absence of disease or infirmity).  Interestingly, despite the  call for healthy women, a \nconsiderable number of participants with self-reported endometriosis enrolled in the study. Among these \nindividuals, 1.6% experienced a significant physical burden of disease compared with controls (p < 0.001), \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nindicating that disease burden is present even within this nominally healthy cohort.  This highlights the \ncomplexity of self‑perceived health and may reflect broader societal factors such as the normalization of \npain and stigma surrounding reproductive health, which can lead to prolonged diagnostic delays.  \nIn our study, we analysed microbiome profiles from 95 individuals with self-reported endometriosis. Since \nlactobacilli are documented to hav e an anti -inflammatory and antimicrobial activity in the vagina 46,56,57 \nand because endometriosis is clearly linked to inflammation20–23, we hypothesized that lactobacilli would \nbe decreased in our Isala endometriosis cohort. Of interest, we observed no significant differences in \nLactobacillus dominance between the endometriosis group and controls  (p = 0.371 ). Importantly, the \npresence of a Lactobacillus-dominated microbiome do es not necessarily imply the absence of an \nunderlying inflammatory condition.  Elevated oestrogen levels, characteristic of the hormonal milieu in \nendometriosis58–60, promote glycogen accumulation in vaginal epithelial cells 61. The breakdown of \nglycogen releases simple sugars that serve as a substrate for Lactobacillus species44, potentially supporting \ntheir dominance even in the presence of underlying inflammation or disease. T his could explain why, \ndespite the presence of endometriosis, the vaginal microbiome may still appear Lactobacillus-dominated \nin our study. Similarly, we found no significant associations between endometriosis – regardless of \nsymptom status – and alpha or beta diversity, nor were there significant differences in the relative \nabundance of specific bacterial taxa when compared to controls. Similar findings have also been observed \nby a prior study with 21 individuals with endometriosis using 16S rRNA amplicon sequencing62. However, \nseveral other studies using the same technique have reported associations between specific bacterial taxa \nand endometriosis. For instance, a recent study comparing the vaginal microbiota of 35 individuals with \nendometriosis and chronic pelvic pain to 23 individuals with chronic pelvic pain alone observed an \nincreased relative abunda nce of Streptococcus anginosus in the endometriosis group 63. Additionally, a \nlarger study of 78 individuals with endometriosis found a higher abundance of Fusobacterium nucleatum \nin vaginal samples 64. Other studies, with sample sizes ranging from 10 to 37 individuals, have described \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nelevated levels of Anaerococcus65,66, Escherichia34,67, Streptococcus34,63,68, Blautia69, as well as conflicting \nfindings on the presence of Fannyhessea34 (formerly classified as Atopobium) in vaginal samples of \nindividuals with endometriosis. While these studies provide valuable insights, many are limited by \nmethodological constraints, including a lack of adjustment for confounding factors such as age and \ncorrection for multiple testin g. Reliance on laparoscopy or imaging for diagnosis, while ensuring clinical \naccuracy, restricts sample sizes in these studies, limiting the statistical power. Furthermore, while a few \nstudies34,66,70 have stratified vaginal microbiota profiles based on the revised American Society for \nReproductive Medicine (rASRM) classification, it is well established that endometriosis disease stage does \nnot correlate with symptom status71,72. Given the substantial symptom burden experienced by individuals \nwith endometriosis, stratifying cohorts based on symptom status – rather than disease stage – may \nfacilitate the identification of clinically relevant microbiome -based biomarkers. Conversely, our present \nstudy includes a relatively large sample size, accounts for key confounders of the vaginal microbiome, and \nstratifies participants with endometriosis based on symptom status, offering a more rigorous and \ncomprehensive assessment of microbial taxa potentially associated with endometriosis.  \nThrough extensive health questionnaires, we observed significant differences in the prevalence of certain \ncomorbid conditions between individuals with and without self -reported endometriosis.  For instance, \nirritable bowel syndrome was reported more than twice as often among individuals with endometriosis \n(11.1%) compared to controls ( 4.8%, OR 2.57, 95% CI  1.43 - 4.36, p = 0.008), which is consistent with \nranges reported in previous studies and might reflect shared pathogenic mechanisms, such as mast cell \nactivation73,74. In addition, migraine was found to be significantly associated with endometriosis (OR=3.75 \n95% CI 1.38-8.60, p = 0.025), exceeding the effect sizes reported in previous nationwide studies (OR 1.70, \n95% CI 1.59 -1.82 in Yang et al . 2012; OR 1.50, CI 95%  1.29-1.74 in Gete  et al . 2023)75,76. Dyspareunia \naffected 34.0% of individuals with endometriosis compared to 25.1% among controls  (OR 1.67, 95% CI \n1.14-2.40, p = 0.033), which aligns with previous findings from Singh et al. (2020) (38.3% vs. 17.7%) 77. \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nInterestingly, our data showed a  nearly three -fold higher odds ratio  of PCOS among participants with \nendometriosis (16.9%) compared to controls ( 4.7%, 95% CI 1.71-4.78, p < 0.001), with the latter falling  \nwithin the globally reported prevalence range of 3 –11%, depending on the diagnostic criteria used 78. \nNotably, the association we observed was approximately twice as high as that reported in both a \npopulation-based cohort (n = 127) and an operative cohort (n = 473)79. Recent findings reported a slightly \nhigher PCOS co‑occurrence of 24.5% among individuals with endometriosis, further support the relevance \nof thi s association 80. Shared risk genes and overlapping inflammatory pathways, characterized by \ndysregulated cytokines and chemokines, between endometriosis and PCOS may partly explain the \nobserved comorbidity, as has been reported before 81–83. At the same time, the observed association \nshould be interpreted with caution, as  differences in healthcare utilization and diagnostic opportunities \nmay also contribute. PCOS is commonly diagnosed using vaginal ultrasound, and individuals undergoing \nimaging as part of endometriosis evaluation may therefore have increased opportunities for PCOS \ndetection.   \nFurthermore, we found that among women trying to conceive, 47.2% with current endometriosis-related \nsymptoms and 36.1% with past endometriosis -related symptoms reported requiring fertility treatment \ncompared to 7.7% controls  (p < 0.001). While these rates fall within ranges described in literature, it is \nimportant to highlight that our findings are derived from a population-based cohort, unlike most previous \nstudies with subfertility cohorts 84,85, which may overestimate the true co -occurrence of subfertility in \nindividuals with endometriosis. In addition, endometriosis is often diagnosed during fertility evaluations, \nwhich may lead to increased detection among individuals presenting with subfertility and thus partially \ninflate the observed association. \nThis study has several limitations.  First, the control group may include asymptomatic individuals with \nundiagnosed endometriosis, which could lead to an underestimation of the true prevalence of \nendometriosis in our cohort and distort the group comparisons. Vice versa, the endometriosis group may \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \ncontain participants without a clinically confirmed diagnosis, some of whom may be false positives. \nSecond, individuals with severe forms of endometriosis, both in terms of clinical extent and impact on \nquality of life, may be underrepresented in ou r cohort considering the setup of the Isala project which \nwas intentionally designed to characterize the vaginal microbiome in a large and predominantly healthy \npopulation. Third, the available data lacked sufficient resolution to further characterize the endometriosis \nsubgroups, and information on treatment history or symptom management was not available.  Fourth, \nthis study did not account for the potential influence of prior or current treatments on the reported \nsymptom burden and healthcare utilization , which may have affected our findings. Finally, although we \ndid not observe a significant association between the vaginal microbiome composition and endometriosis, \nfunctional differences at the metabolic or transcriptomic level may still be present and relevant. \nDespite inherent limitations, self -reported data remains a valuable tool in endometriosis research, \nparticularly given the challenges and delays in obtaining formal diagnoses. Many individuals experience \nsymptoms for years before receiving clinical confirmation, and self-reporting can offer a more immediate \nand inclusive snapshot of lived experiences across diverse populations. Moreover, capturing self-reported \nsymptom severity and comorbid conditions can help identify underrecognized patterns and gen erate \nhypotheses for further study in clinical cohorts. Taken together, our findings do not support an association \nbetween the vaginal microbiome composition and endometriosis. However, we identified a pronounced \nco-occurrence of polycystic ovarian syndrom e with endometriosis, in addition to other comorbid \nconditions. These results suggest potential shared mechanisms worth further investigating and \ndemonstrate the value of citizen science in advancing endometriosis research. We suggest that follow-up \nresearch should move beyond microbial composition focusing on functional aspects such as host -\nmicrobiome interactions, immune profiling, and metabolomic signatures, as well as explore anatomical \nsites beyond the vagina, which may offer more promising pathways to ward identifying biomarkers or \ntherapeutic targets. \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nMethods \nStudy cohort and data collection \nThis study was conducted in accordance with the ethical principles of the Declaration of Helsinki. The \nstudy protocol received ethical approval from the Ethical Committee of the Antwerp University \nHospital/University of Antwerp (reference number B300201942076) and is registered on clinicaltrials.gov \nunder the unique identifier NCT04319536. Informed consent was obtained from all participants. \nThis cross -sectional s tudy draws its study cohort from the Isala citizen science project \n(https://isala.be/en/). For this study, 144 participants with self-reported endometriosis were included and \nstratified into two groups based on their symptom status: individuals experiencing current endometriosis-\nrelated symptoms (n = 66) and individuals experiencing symptoms in the past but without active \nsymptoms during the study (n = 78). This was complemented by a group of 3,337 participants withou t \nreported endometriosis, excluding individuals who reported dysmenorrhoea and requiring NSAIDs during \nmenstruation. Menopausal individuals were excluded from this study. All participants filled in a health \nquestionnaire with General Data Protection Regula tion-compliant questions on the Qualtrics platform. \nVaginal swabs were collected and processed as described before and stored in the in -house biobank \ndecentralized hub to comply with the most recent GDPR -regulations in Belgium on biobanking human \nsamples ( KB 2018/30209) 39. A total of 95 microbiome profiles were obtained from the endometriosis \ncohort, including 43 profiles from in dividuals with current endometriosis -related symptoms and 52 \nprofiles from individuals with past symptoms. Additionally, 2,279 microbiome profiles were collected from \nparticipants without self-reported endometriosis.   \n16S rRNA amplicon sequencing, reference database and quality control \nVaginal swabs were collected and processed as previously described by Lebeer et al. (2023)39. Briefly, DNA \nwas extracted using the DNeasy PowerSoil Pro Kit (Qiagen), followed by amplification of the V4 region of \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nthe 16S rRNA gene using standard barcoded primers adapted for dual -index sequencing. PCR products \nwere purified, quantified, and po oled in equimolar concentrations to generate sequencing libraries. \nLibraries were sequenced using dual -index paired -end sequencing on an Illumina MiSeq platform, \nincluding appropriate negative controls for DNA extraction and PCR.   \nFor the construction of the custom 16S reference database and processing of amplicon sequencing data, \na refined taxonomic framework was developed to improve resolution within the Lactobacillus genus by \ndefining subgenera based on phylogenetic relationships. A custom 16S rRNA refe rence database was \ngenerated using sequences from the Genome Taxonomy Database (GTDB) and adapted for use with \nDADA286. Sequence quality control and processing were conducted using the DADA2 pipeline, including \nfiltering of low -quality reads, merging of paired -end reads, and removal of chimeras. Taxonomic \nassignment was performed using a custom 16S rRNA reference database , followed by reclassification \nsteps to align with updated Lactobacillaceae taxonomy and the defined Lactobacillus subgenera. \nAdditional quality control steps included removal of non -bacterial and low -quality amplicon sequence \nvariants, as well as filterin g of samples based on read count and normalized sequencing depth. For full \nmethodological details, we refer to Lebeer et al. (2023)39. \nEmbeddings \nTo explore the vaginal microbiome composition of participants with and without self -reported \nendometriosis three different embedding methods were applied using Bray -Curtis dissimilarity based on \nthe relative abundances of (sub)genera within samples: t-distributed stochastic neighbour embedding (t-\nSNE), uniform manifold approximation and projection (UMAP), and principal coordinates analysis (PCOA). \nPlots were generated with ggplot2 R package.   \nStatistical analyses \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nThrough survey data, we  compared demographic variables between participants with and without self -\nreported endometriosis , further distinguishing between individuals reporting current versus past \nendometriosis-related symptoms. Normality of continuous variables including age, BMI , and number of \npregnancies was assessed using the Shapiro-Wilk test. None of these variables were normally distributed \nand were therefore presented as median and interquartile range (IQR). For the ordinary variable ‘self -\nrated health’ the mean and standard deviation were calculated based on a five-point ordinary scale (very \ngood = 5, good = 4, fair = 3, poor = 2, very poor = 1). Depending on the data type of the variable s, an \nappropriate age-adjusted model ( linear, logistic, Poisson, multinomial, or ordina l regression ) was \nimplemented. Overall significance was calculated through likelihood ratio -tests for categorical variables \nfollowed by correction for multiple testing through the Benjamini-Hochberg procedure.   \nAlpha and beta diversity measures were compa red as previously described by Lebeer et al. (2023)39. The \ndifferential abundance of taxa was performed using the in -house R package multidiffabundance, version \n0.0.1 (publicly available at https://github.com/thiesgehrmann/multidiffabundance), using Maaslin2 87, \nLimma88, DESeq2 89, ANCOM -BC90, and a linear regression on the centered log -ratio (CLR) transformed \nabundance data, reporting the consensus of the five tools. Each test was adjusted for technical \nconfounders, age, recent sexual intercourse (last 24 hours), number of pregnancies, phase of cycle, use of \nhormonal contraceptives, and  library read concentratio n. Correction for multiple testing was applied \nwithin each tool using the Benjamini –Hochberg false discovery rate procedure. Data was processed in R \nversion 4.2.2 using the in-house developed package Tidytacos91 and the tidyverse set of packages. Results \nwere visualised in Python 3, selecting only a subset of taxa based on their known importance in the vaginal \nmicrobiome.  \nTo investigate associations between endometriosis and a range of symptoms, infections, and clinical \nconditions, we conducted multiple regression analyses, with all outcomes defined as lifetime (‘ever’) self-\nreported occurrence of the respective infection, condition, or vaginal symptom. The selection of potential \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nconfounders was determined using a directed acyclic graph (DAG; Supplementary Figure 2), which allows \nthe identification of the minimum adjustment required, while avoiding inappropriate ‘overadjustment’ \nfor mediator variables 92. The following covariates were included in our models: age, BMI and education \nlevel. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated and presented in a forest plot. \nP-values were corrected for multiple testing through the Benjamini–Hochberg procedure.  \nData availability  \nSequencing data are available at the European Nucleotide Archive (ENA) under bioproject PREB50407. \nSample me tadata are available with access control through the European Genome –Phenome Archive \n(EGA) under dataset ID EGAD00001009890. Access is granted as described, upon agreement to the \nharmonised Data Access Agreement developed by EU-STANDS4PM (European Union standards for in silico \nmodels for personalised medicine; https://doi.org/10.6084/m9.figshare.23904300). \nAcknowledgments  \nThe authors would like to thank all Isala participants  and following colleagues and students who were \ninstrumental for the Isala sampling campaign and processing : I. Tuyaerts, N. Van Vliet, L. Van Ham, M. \nLegein, D. Vandenheuvel, E. Cauwenberghs, L. Delanghe, A. Groenwals, S. El Messaoudi, J. Hiers, L. Van \nDyck, C. Dricot,  L. Leysen and L. Martin Diaz.  Strategic support was provided by L. Talboom and L. \nHaesevoets (Studio Maria, communication), C. Varszegi (Little Big Things, website, \nhttps://littlebigthings.be), R. Broms and S. Vergauwen (Sensoa vzw, sexual lifestyle quest ions), E. Den \nHond and C. Franken (Provinciaal Instituut voor Hygiëne Antwerpen, population survey), J. Raes (KU \nLeuven, Flemish gut flora project), K. Scott (food-related questions), K. Wuyts and R. Samson (urbanization \nand contact with urban green), the Antwerp Biobank (University Hospital Antwerp) and Centre of Medical \nGenetics (University Hospital Antwerp, sequencing support). The authors acknowledge the European \nResearch Council (starting grant Lacto -Be 852600 of S.L., with S.A., T.G., S.W., W.V.B., T. E. and J.D. \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nappointed on the project), the Special Research Fund of the Universiteit Antwerpen (UA BOF; DOCPRO \n37054 grant of S.A. and temporary mandate grant 48145 of S.W.), the Inter -University Special Research \nFund of Flanders (iBOF; POSSIBL project), BOF funding and proof -of-concept VALERIE (Horizon, grant ID \n101213306), the industrial research fund UAntwerpen (IOF service platform microbiome sequencing) and \nthe Research Foundation —Flanders (FWO; aspirant fundamental research grant 11A0620N and \npostdoctoral fellowship 12AZ624N  of S.W., postdoctoral fellowship 12S4222N of I.D.B , senior post -\ndoctoral research grant 1277222N of I.S., aspirant strategic basic research grant 1SD0622N of L.V.D. and \nResearch projects G049022N, G031222N and S006426 of S.L.). The funders had no role in study design, \ndata collection, data analysis, data interpretation, or writing of the manuscript. \nAuthor contributions \nS.A., S.W., G.D., V.V. and S.L. designed the study and worked on the conceptualization of the research \nproject. S.A. and S.L. worked on the survey set-up. I.R. and T.G. cleaned the answers. S.A. and S.L. carried \nout the experimental and logistical work. I.D.B. was responsible for the biobanking of all collected \nsamples. I.R. analy sed the health questionnaire data and performed the statistical analyses. T.G., T.V.R. \nand S.W. processed the sequencing data and performed the biostatistical analyses. T.V.R. adjusted the \nmicrobiome data with the new taxonomy. I.R. and T.G. worked on the visualizations. I.R., T.G., S.A., I.D.B., \nC.N.A., T.V.R., S.W., F.D. and S.L. contributed to the interpretation of the results. I.R. and S.L. wrote the \noriginal manuscript. All authors contributed to reviewing and editing of the final manuscript.  \nCompeting interests \nS.L. declares to be a voluntary academic board member of the International Scientific Association on \nProbiotics and Prebiotics (ISAPP, www.isappscience.org), cofounder of YUN and scientific advisor for Freya \nBiosciences. The team of S.L. declare s research funding from YUN, Bioorg, Puratos, DSM I -Health and \nLesaffre/Gnosis. None of these organizations or companies were involved in the design or data analysis \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nof this study, which was fully funded by the university, governmental, and European funding. S.A. declares \nto be a voluntary member of the student and fellows association of ISAPP. The other authors declare no \ncompeting interests. \nDeclaration of Generative AI and AI-assisted technologies in the writing process \nDuring the preparation of this wor k, the author(s) used ChatGPT by OpenAI exclusively to assist with \nparaphrasing and to gather inspiration for scientific rephrasing. After using this tool/service, the author(s) \nreviewed and edited the content as needed and take(s) full responsibility for the content of the \npublication. \nReferences  \n1. Zondervan, K. T., Becker, C. M. & Missmer, S. A. Endometriosis. N Engl J Med 382, 1244–1256 \n(2020). \n2. Von Theobald, P., Cottenet, J., Iacobelli, S. & Quantin, C. Epidemiology of Endometriosis in France: A \nLarge, Nation-Wide Study Based on Hospital Discharge Data. BioMed Research International 2016, \n1–6 (2016). \n3. Reid, R. et al. The prevalence of self-reported diagnosed endometriosis in the Australian population: \nresults from a nationally-representative survey. BMC Res Notes 12, 88 (2019). \n4. Christ, J. P. et al. Incidence, prevalence, and trends in endometriosis diagnosis: a United States \npopulation-based study from 2006 to 2015. American Journal of Obstetrics and Gynecology 225, \n500.e1-500.e9 (2021). \n5. Illum, L. R. H. et al. Temporal and regional differences in the incidence of hospital‑diagnosed \nendometriosis: a Danish population‑based study. Acta Obstet Gynecol Scand 101, 737–746 (2022). \n6. Becker, K. et al. Real world data on symptomology and diagnostic approaches of 27,840 women \nliving with endometriosis. Sci Rep 11, 20404 (2021). \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n7. Prescott, J. et al. A prospective cohort study of endometriosis and subsequent risk of infertility. \nHum. Reprod. 31, 1475–1482 (2016). \n8. Ek, M. et al. Gastrointestinal symptoms among endometriosis patients—A case-cohort study. BMC \nWomen’s Health 15, 59 (2015). \n9. Yang, F. et al. Evidence of shared genetic factors in the etiology of gastrointestinal disorders and \nendometriosis and clinical implications for disease management. Cell Reports Medicine 4, 101250 \n(2023). \n10. Gambadauro, P., Carli, V. & Hadlaczky, G. Depressive symptoms among women with endometriosis: \na systematic review and meta-analysis. American Journal of Obstetrics and Gynecology 220, 230–\n241 (2019). \n11. Nnoaham, K. E. et al. Impact of endometriosis on quality of life and work productivity: A multicenter \nstudy across ten countries. Fertility and Sterility 96, 366-373.e8 (2011). \n12. Simoens, S. et al. The burden of endometriosis: Costs and quality of life of women with \nendometriosis and treated in referral centres. Human Reproduction 27, 1292–1299 (2012). \n13. Gete, D. G. et al. Hospitalisations and length of stays in women with endometriosis: a data linkage \nprospective cohort study. EClinicalMedicine 80, 103030 (2025). \n14. Sampson, J. A. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue \ninto the peritoneal cavity. American Journal of Obstetrics and Gynecology 14, 422–469 (1927). \n15. Konrad, L., Dietze, R., Kudipudi, P. K., Horné, F. & Meinhold-Heerlein, I. Endometriosis in MRKH \ncases as a proof for the coelomic metaplasia hypothesis? Reproduction 158, R41–R47 (2019). \n16. Sampson, J. A. Metastatic or Embolic Endometriosis, due to the Menstrual Dissemination of \nEndometrial Tissue into the Venous Circulation. Am J Pathol 3, 93-110.43 (1927). \n17. Attia, G. R. et al. Progesterone Receptor Isoform A But Not B Is Expressed in Endometriosis1. The \nJournal of Clinical Endocrinology & Metabolism 85, 2897–2902 (2000). \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n18. Lin, Y.-J., Lai, M.-D., Lei, H.-Y. & Wing, L.-Y. C. Neutrophils and Macrophages Promote Angiogenesis \nin the Early Stage of Endometriosis in a Mouse Model. Endocrinology 147, 1278–1286 (2006). \n19. Rahmioglu, N. et al. The genetic basis of endometriosis and comorbidity with other pain and \ninflammatory conditions. Nat Genet 55, 423–436 (2023). \n20. Ahn, S. H. et al. IL-17A Contributes to the Pathogenesis of Endometriosis by Triggering \nProinflammatory Cytokines and Angiogenic Growth Factors. The Journal of Immunology 195, 2591–\n2600 (2015). \n21. Chang, K.-K. et al. IL-27 triggers IL-10 production in Th17 cells via a c-Maf/RORγt/Blimp-1 signal to \npromote the progression of endometriosis. Cell Death Dis 8, e2666–e2666 (2017). \n22. Wilson, T. R. et al. Neutrophils initiate proinflammatory immune responses in early endometriosis \nlesion development. JCI Insight 10, e186133 (2025). \n23. Vallvé-Juanico, J., Santamaria, X., Vo, K. C., Houshdaran, S. & Giudice, L. C. Macrophages display \nproinflammatory phenotypes in the eutopic endometrium of women with endometriosis with \nrelevance to an infectious etiology of the disease. Fertility and Sterility 112, 1118–1128 (2019). \n24. Becker, C. M. et al. ESHRE guideline: endometriosis. Human Reproduction Open 2022, hoac009 \n(2022). \n25. Whitaker, L. H. R. et al. Laparoscopic treatment of isolated superficial peritoneal endometriosis for \nmanaging chronic pelvic pain in women: study protocol for a randomised controlled feasibility trial \n(ESPriT1). Pilot Feasibility Stud 7, (2021). \n26. Leborne, P. et al. Clinical outcomes following surgical management of deep infiltrating \nendometriosis. Sci Rep 12, (2022). \n27. Martinez, V. et al. Chronic postsurgical pain: A European survey. European Journal of \nAnaesthesiology 41, 351–362 (2024). \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n28. De Corte, P. et al. Endometriosis and risk of depression among oral contraceptive users: a pooled \nanalysis of cohort studies from 13 countries. Human Reproduction 40, 479–486 (2025). \n29. Lockhat, F. B., Emembolu, J. O. & Konje, J. C. The efficacy, side-effects and continuation rates in \nwomen with symptomatic endometriosis undergoing treatment with an intra-uterine administered \nprogestogen (levonorgestrel): a 3 year follow-up. Human Reproduction 20, 789–793 (2005). \n30. Donnez, J. et al. Treatment of endometriosis-associated pain with linzagolix, an oral gonadotropin-\nreleasing hormone–antagonist: a randomized clinical trial. Fertility and Sterility 114, 44–55 (2020). \n31. Phan, J. et al. Alterations in Gut Microbiome Composition and Function in Irritable Bowel Syndrome \nand Increased Probiotic Abundance with Daily Supplementation. mSystems 6, e01215-21 (2021). \n32. Jacobs, J. P. et al. Multi-omics profiles of the intestinal microbiome in irritable bowel syndrome and \nits bowel habit subtypes. Microbiome 11, 5 (2023). \n33. Yang, Y. et al. Gut microbiota in women with polycystic ovary syndrome: an individual based \nanalysis of publicly available data. eClinicalMedicine 77, 102884 (2024). \n34. Ata, B. et al. The Endobiota Study: Comparison of Vaginal, Cervical and Gut Microbiota Between \nWomen with Stage 3/4 Endometriosis and Healthy Controls. Scientific Reports 9, 1–9 (2019). \n35. Chadchan, S. B. et al. Gut microbiota and microbiota-derived metabolites promotes endometriosis. \nCell Death Discovery 9, (2023). \n36. Talwar, C. et al. Identification of distinct stool metabolites in women with endometriosis for non-\ninvasive diagnosis and potential for microbiota-based therapies. Med 6, 100517 (2025). \n37. Haggerty, C. L., Hillier, S. L., Bass, D. C., Ness, R. B., & the PID Evaluation and Clinical Health (PEACH) \nStudy Investigators. Bacterial Vaginosis and Anaerobic Bacteria Are Associated with Endometritis. \nCLIN INFECT DIS 39, 990–995 (2004). \n38. Oerlemans, E. F. M. et al. The dwindling microbiota of aerobic vaginitis, an inflammatory state \nenriched in pathobionts with limited TLR stimulation. Diagnostics 10, 1–15 (2020). \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n39. Lebeer, S. et al. A citizen-science-enabled catalogue of the vaginal microbiome and associated \nfactors. Nat Microbiol 8, 2183–2195 (2023). \n40. Huo, Y., Jiang, Q. & Zhao, W. Meta-analysis of metagenomics reveals the signatures of vaginal \nmicrobiome in preterm birth. Medicine in Microecology 14, 100065 (2022). \n41. Huang, C. et al. Meta-analysis reveals the vaginal microbiome is a better predictor of earlier than \nlater preterm birth. BMC Biol 21, 199 (2023). \n42. Spurbeck, R. R. & Arvidson, C. G. Lactobacillus gasseri Surface-Associated Proteins Inhibit Neisseria \ngonorrhoeae Adherence to Epithelial Cells. Infect Immun 80, 3743–3743 (2012). \n43. Nunn, K. L. et al. Enhanced Trapping of HIV-1 by Human Cervicovaginal Mucus Is Associated with \nLactobacillus crispatus-Dominant Microbiota. mBio 6, e01084-15 (2015). \n44. Jenkins, D. J. et al. Bacterial amylases enable glycogen degradation by the vaginal microbiome. Nat \nMicrobiol https://doi.org/10.1038/s41564-023-01447-2 (2023) doi:10.1038/s41564-023-01447-2. \n45. Izadifar, Z. et al. Mucus production, host-microbiome interactions, hormone sensitivity, and innate \nimmune responses modeled in human cervix chips. Nat Commun 15, 4578 (2024). \n46. Decout, A. et al. Lactobacillus crispatus S-layer proteins modulate innate immune response and \ninflammation in the lower female reproductive tract. Nat Commun 15, 10879 (2024). \n47. Gudnadottir, U. et al. The vaginal microbiome and the risk of preterm birth: a systematic review and \nnetwork meta-analysis. Sci Rep 12, 7926 (2022). \n48. Chopra, C., Kumar, V., Kumar, M. & Bhushan, I. Role of vaginal microbiota in idiopathic infertility: a \nprospective study. Microbes and Infection 26, 105308 (2024). \n49. Fettweis, J. M. et al. The vaginal microbiome and preterm birth. Nature Medicine 25, 1012–1021 \n(2019). \n50. Hardy, L. et al. A fruitful alliance: the synergy between Atopobium vaginae and Gardnerella vaginalis \nin bacterial vaginosis-associated biofilm. Sex Transm Infect 92, 487–491 (2016). \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n51. Berard, A. R. et al. Vaginal epithelial dysfunction is mediated by the microbiome, metabolome, and \nmTOR signaling. Cell Reports 42, 112474 (2023). \n52. Łaniewski, P. & Herbst-Kralovetz, M. M. Bacterial vaginosis and health-associated bacteria modulate \nthe immunometabolic landscape in 3D model of human cervix. npj Biofilms Microbiomes 7, (2021). \n53. Delgado-Diaz, D. J. et al. Distinct Immune Responses Elicited From Cervicovaginal Epithelial Cells by \nLactic Acid and Short Chain Fatty Acids Associated With Optimal and Non-optimal Vaginal \nMicrobiota. Front. Cell. Infect. Microbiol. 9, (2020). \n54. Vos, T. et al. Global, regional, and national incidence, prevalence, and years lived with disability for \n301 acute and chronic diseases and injuries in 188 countries, 1990–2013: a systematic analysis for \nthe Global Burden of Disease Study 2013. The Lancet 386, 743–800 (2015). \n55. Wang, M.-H., Chen, J.-H., Qi, X.-Y., Li, Z.-X. & Huang, Y. Global prevalence of adenomyosis and \nendometriosis: a systematic review and meta-analysis. Reprod Biol Endocrinol 23, 148 (2025). \n56. Argentini, C. et al. Evaluation of Modulatory Activities of Lactobacillus crispatus Strains in the \nContext of the Vaginal Microbiota. Microbiol Spectr 10, (2022). \n57. Hearps, A. C. et al. Vaginal lactic acid elicits an anti-inflammatory response from human \ncervicovaginal epithelial cells and inhibits production of pro-inflammatory mediators associated \nwith HIV acquisition. Mucosal Immunology 10, 1480–1490 (2017). \n58. Han, S. J., Lee, J. E., Cho, Y. J., Park, M. J. & O’Malley, B. W. Genomic Function of Estrogen Receptor \nβ in Endometriosis. Endocrinology 160, 2495–2516 (2019). \n59. Noble, L. S. et al. Prostaglandin E2 Stimulates Aromatase Expression in Endometriosis-Derived \nStromal Cells1. The Journal of Clinical Endocrinology & Metabolism 82, 600–606 (1997). \n60. Han, S. J. et al. Estrogen Receptor β Modulates Apoptosis Complexes and the Inflammasome to \nDrive the Pathogenesis of Endometriosis. Cell 163, 960–974 (2015). \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n61. Navarro, S. et al. Glycogen availability and pH variation in a medium simulating vaginal fluid \ninfluence the growth of vaginal Lactobacillus species and Gardnerella vaginalis. BMC Microbiol 23, \n186 (2023). \n62. Hernandes, C. et al. Microbiome Profile of Deep Endometriosis Patients: Comparison of Vaginal \nFluid, Endometrium and Lesion. Diagnostics 10, 163 (2020). \n63. Jimenez, N. et al. Vaginal and rectal microbiome contribute to genital inflammation in chronic pelvic \npain. BMC Med 22, 283 (2024). \n64. Muraoka, A., Yokoi, A. & Kajiyama, H. Emerging bacterial factors for understanding pathogenesis of \nendometriosis. iScience 27, 108739 (2024). \n65. MacSharry, J. et al. Endometriosis specific vaginal microbiota links to urine and serum N-glycome. \nSci Rep 14, 25372 (2024). \n66. Perrotta, A. R. et al. The Vaginal Microbiome as a Tool to Predict rASRM Stage of Disease in \nEndometriosis: a Pilot Study. Reproductive Sciences 27, 1064–1073 (2020). \n67. Sessa, R. et al. Characterization of the vaginal microbiota in Italian women with endometriosis: \npreliminary study. Arch Gynecol Obstet 310, 2141–2151 (2024). \n68. Akiyama, K. et al. Molecular detection of microbial colonization in cervical mucus of women with \nand  without endometriosis. American journal of reproductive immunology (New York, N.Y. : 1989) \n82, e13147 (2019). \n69. Hicks, C. et al. Oral, Vaginal, and Stool Microbial Signatures in Patients With Endometriosis as \nPotential Diagnostic Non‑Invasive Biomarkers: A Prospective Cohort Study. BJOG 1471–0528.17979 \n(2024) doi:10.1111/1471-0528.17979. \n70. Wei, W., Zhang, X., Tang, H., Zeng, L. & Wu, R. Microbiota composition and distribution along the \nfemale reproductive tract of women with endometriosis. Annals of Clinical Microbiology and \nAntimicrobials 19, 1–9 (2020). \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n71. Schliep, K. C. et al. Pain typology and incident endometriosis. Hum Reprod 30, 2427–2438 (2015). \n72. Johnson, N. P. et al. World Endometriosis Society consensus on the classification of endometriosis. \nHum. Reprod. 32, 315–324 (2017). \n73. Guilarte, M. et al. Peripheral Corticotropin-Releasing Factor Triggers Jejunal Mast Cell Activation and \nAbdominal Pain in Patients With Diarrhea-Predominant Irritable Bowel Syndrome. Am J \nGastroenterol 115, 2047–2059 (2020). \n74. Mariuzzi, L. et al. Functional expression of aryl hydrocarbon receptor on mast cells populating \nhuman endometriotic tissues. Laboratory Investigation 96, 959–971 (2016). \n75. Yang, M.-H. et al. Women with Endometriosis Are More Likely to Suffer from Migraines: A \nPopulation-Based Study. PLoS ONE 7, e33941 (2012). \n76. Gete, D. G., Doust, J., Mortlock, S., Montgomery, G. & Mishra, G. D. Associations between \nendometriosis and common symptoms: findings from the Australian Longitudinal Study on \nWomen’s Health. American Journal of Obstetrics and Gynecology 229, 536.e1-536.e20 (2023). \n77. Singh, S. et al. Prevalence, Symptomatic Burden, and Diagnosis of Endometriosis in Canada: Cross-\nSectional Survey of 30 000 Women. Journal of Obstetrics and Gynaecology Canada 42, 829–838 \n(2020). \n78. Neven, A. C. H. et al. Prevalence and accurate diagnosis of polycystic ovary syndrome in adolescents \nacross world regions: a systematic review and meta-analysis. European Journal of Endocrinology \n191, S15–S27 (2024). \n79. Schliep, K. C. et al. Examining the co-occurrence of endometriosis and polycystic ovarian syndrome. \nAJOG Global Reports 3, 100259 (2023). \n80. Kaveh, M. et al. Epidemiological study and clinical consequences of endometriosis and PCOS co-\noccurrence in a clinical population of women: a cross-sectional study. BMC Women’s Health 26, 182 \n(2026). \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n81. Tan, H., Long, P. & Xiao, H. Dissecting the shared genetic architecture between endometriosis and \npolycystic ovary syndrome. Front. Endocrinol. 15, 1359236 (2024). \n82. Eriksson, G. et al. Single-cell profiling of the human endometrium in polycystic ovary syndrome. Nat \nMed https://doi.org/10.1038/s41591-025-03592-z (2025) doi:10.1038/s41591-025-03592-z. \n83. Omidvar-Mehrabadi, A., Ebrahimi, F., Shahbazi, M. & Mohammadnia-Afrouzi, M. Cytokine and \nchemokine profiles in women with endometriosis, polycystic ovary syndrome, and unexplained \ninfertility. Cytokine 178, 156588 (2024). \n84. Saraswat, L. et al. O-250 Infertility and endometriosis: a 30-yearlong national population-based \nstudy of prevalence, association and pregnancy outcomes. Human Reproduction 40, deaf097.250 \n(2025). \n85. Leibson, C. L. et al. Incidence and characterization of diagnosed endometriosis in a geographically \ndefined population. Fertility and Sterility 82, 314–321 (2004). \n86. Callahan, B. J. et al. DADA2: High-resolution sample inference from Illumina amplicon data. Nat \nMethods 13, 581–583 (2016). \n87. Mallick, H. et al. Multivariable association discovery in population-scale meta-omics studies. PLoS \nComput Biol 17, e1009442 (2021). \n88. Law, C. W., Chen, Y., Shi, W. & Smyth, G. K. voom: precision weights unlock linear model analysis \ntools for RNA-seq read counts. Genome Biol 15, R29 (2014). \n89. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq \ndata with DESeq2. Genome Biol 15, 550 (2014). \n90. Lin, H. & Peddada, S. D. Analysis of compositions of microbiomes with bias correction. Nat Commun \n11, 3514 (2020). \n91. Wittouck, S., Rillaer, T. V., Smets, W. & Lebeer, S. Tidytacos: An R package for analyses on taxonomic \ncomposition of microbial communities. JOSS 10, 6313 (2025). \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n92. Textor, J., Van Der Zander, B., Gilthorpe, M. S., Liśkiewicz, M. & Ellison, G. T. H. Robust causal \ninference using directed acyclic graphs: the R package ‘dagitty’. Int. J. Epidemiol. dyw341 (2017) \ndoi:10.1093/ije/dyw341. \n \n \n \n \n \n \n \nTables and Figures \nTable 1. Demographic characteristics of participants with and without self -reported endometriosis. The former \ngroup is further categorised into participants experiencing current and past endometriosis -related symptoms. \nPercentages were calculated based on the tota l number of participants within each group. Significantly different \nvariables between the three pairwise comparisons: control vs. past symptoms (1), control vs. current symptoms (2) \nand current symptoms vs. past symptoms (3) are represented in the last column with an asterisk. Significant results \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \nare indicated with an asterisk. IQR: interquartile range, BMI: body mass index, HPV: humanpapillomavirus. Self-rated \nhealth is a five-point ordinary scale (very good = 5, good = 4, fair = 3, poor = 2, very poor = 1). \n     \n     \n \nCharacteristic Control \n(N = 3337) \nEndometriosis, \npast complaints \n(N = 78) \nEndometriosis,  \ncurrent complaints \n(N = 66) \nP-value  \n \nAge (years) Median [IQR] 28 [24-33] 36 [31-41] 31 [27-35] < 0.001 (1;3) 0.006 (2) \nBMI (kg/m2) Median [IQR] 23.1 [21.0-26.2] 23.0 [21.3-26.4] 23.6 [21.4-27.5] 0.311 \nBirth delivery mode    0.766 \n   Vaginal  2964 (88.8%) 72 (92.3%) 57 (86.4%)  \n   Caesarean section 332 (10.0%) 4 (5.1%) 7 (10.6%)  \n   Missing  41 (1.2%) 2 (2.6%) 2 (3.0%)  \nContraception    0.142 \n   Combined oral contraceptives 805 (24.1%) 12 (15.4%) 13 (19.7%)  \n   Progesterone only pill 31 (0.9%) 4 (5.1%) 3 (4.5%)  \n   Hormonal intrauterine device 294 (8.8%) 10 (12.8%) 9 (13.6%)  \n   Other or none 2207 (66.2%) 52 (66.7%) 41 (62.2%)  \nMarital status during last 3 months1    0.063 \n   No partner 757 (22.7%) 15 (19.2%) 8 (12.1%)  \n   One partner 2510 (75.2%) 55 (70.5%) 55 (83.3%)  \n   Multiple partners 70 (2.1%) 8 (10.3%) 3 (4.6%)  \nSubfertility  85/1108 (7.7%) 17/47 (36.2%) 17/36 (47.2%) < 0.001 (1;2) \nNumber of pregnancies Median [IQR] 0 [0-1] 2 [0-2] 0 [0-2] 0.285 \nBiological child(ren) 1080 (32.4%) 47 (60.7%) 25 (37.9%) 0.813 \nAllergies and intolerances     \n   Gluten allergy 65 (2.0%) 3 (3.9%) 4 (6.1%) 0.236 \n   Lactose intolerant 253 (7.6%) 11 (14.1%) 9 (13.6%) 0.142 \nAntibiotics during last 3 months 690 (0.7%) 20 (25.6%) 17 (25.8%) 0.467 \n(1) (3) \n(2) \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n 1 Without implying cohabitation. \n \n \n \n \n \n \n \n \n \n \n \n \n \n \nCurrent smoker 298 (8.9%) 10 (12.8%) 2 (3.0%) 0.213 \nCurrent drug user 319 (9.6%) 3 (3.9%) 4 (6.1%) 0.467 \nHPV vaccinated 1416 (42.4%) 14 (18.0%) 17 (25.8%) 0.236 \nSelf-rated health 4.12 ± 0.64 3.85 ± 0.72 3.79 ± 0.65 0.003 (1) < 0.001 (2) \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n \nFigure 1. Overview of cohort size, microbiome composition, and bacterial diversity in vaginal samples of \nindividuals with self -reported endometriosis (Endo) and controls.  (a) Overview of participants, collected \nmetadata and vaginal microbiome profiles. (b) t-SNE analysis. (c) Bar plot showing the vaginal microbiome profiles \nof participants with self -reported endometriosis (current versus past endometriosis -related symptoms), and \nwithout self-reported endometriosis (control group) on genus level. d) Associations on the level of beta diversity \nbetween the samples; Associations on the level of alpha diversity of the samples ; Associations on the level of \nabundance of specific taxa analysed by five different differential abundance testing methods (Limma, Maaslin2, \nDESeq2, ANCOM-BC, and a linear regression on the centered log-ratio transformed abundance data). \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n \n                   \n        \n                      \n        \n                          \n       \n                            \n        \n                                  \n      \n                                    \n      \n        \n      \n      \n         \n                        \n        \n                          \n       \n                            \n       \n         \n                        \n         \n                        \n                        \n  \n   \n   \n \n  \n  \n      \n      \n      \n  \n       \n                          \n                             \n  \n  \n         \n            \n               \n           \n                           \n                         \n                       \n                          \n          \n             \n        \n          \n                                                         \n    \n    \n    \n    \n    \n    \n    \n    \n    \n    \n      \n                  \n          \n            \n               \n           \n          \n                             \n                           \n                         \n                            \n          \n        \n             \n                                                         \n                 \n    \n    \n    \n    \n    \n                 \n       \n    \n    \n    \n    \n    \n    \n    \n    \n    \n    \n      \n                  \n          \n            \n               \n           \n          \n                             \n                           \n                         \n                            \n          \n        \n             \n       \n      \n    \n    \n    \n    \n    \n                    \n    \n                 \n             \n          \n          \n           \n            \n               \n               \n                   \n         \n           \n              \n              \n             \n          \n             \n        \n                 \n                   \n                    \n                                                  \n                          \n       \n       \n       \n       \n       \n       \n\nARTICLE IN PRESS\nARTICLE IN PRESS\n \n   \n \n \nFigure 2.  Associations between endometriosis and a range of infections, clinical conditions, and vaginal \nsymptoms. Forest plot shows adjusted odds ratios (ORs) with 95% confidence intervals derived from multivariable \nlogistic regression models. All outcomes refer to ever having had the infection, condition, or vaginal symptom, as \nself-reported by participants. Analyses were adjusted for age, body mass index, and educational level. P -values \nwere corrected for multiple testing using the Benjamini–Hochberg false discovery rate procedure. *P-value < 0.05, \n**P-value < 0.01, ***P-value < 0.001","source_license":"CC0","license_restricted":false}