Citizen science reveals comorbidities in endometriosis with no shift in vaginal microbiome composition

Endometriosis, citizen science, vaginal microbiome, symptom profiles, comorbidities ARTICLE IN PRESS ARTICLE IN PRESS

Endometriosis is a chronic inflammatory condition affecting 2 –10% of reproductive -aged women, most commonly presenting with pelvic pain and subfertility. While its impact on reproductive health is increasingly recognized, the vaginal microbiome’s role in the pathogenesis of endometriosis is still unclear and a comprehensive map of potential co-occurring conditions remains underexplored. Leveraging data from the Isala citizen -science platform in Flanders (Belgium), we analy sed vaginal microbiome profiles obtained through 16S rRNA sequencing and health data from 95 women with self-reported endometriosis and 2,279 without. While no differences were observed in vaginal microbiome composition or diversity, we identified significant associations between endometriosis and polycystic ovarian syndrome (OR = 2.92, 95% CI 1.71-4.78, p < 0.001), migraine (OR = 3.75, 95% CI 1.38-8.60, p = 0.025), irritable bowel syndrome (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 findings suggest that the vaginal microbiome composition plays at most a limited role in endometriosis and highlights how citizen science can effectively complement clinical research by capturing underrecognized comorbidities. ARTICLE IN PRESS ARTICLE IN PRESS

Endometriosis is a chronic inflammatory condition that most commonly affects reproductive -aged women, with an estimated prevalence of 2-10%1–5. It is defined by the presence of endometrial-like tissue outside the uterus, with lesions varying in size, depth and location, affecting both reproductive and non - reproductive organs . Pelvic pain 6, subfertility7, gastrointestinal issues 8,9 and mental health problems 10 have been frequently reported in different endometriosis cohorts across the world (with sizes between 174 and 188,461), imposing a significant burden on both patient and society with substantial healthcare costs11–13. Yet, a comprehensive map of the f ull range of symptoms and co -occurring conditions remains lacking. The pathogenesis of endometriosis is multifactorial and remains incompletely understood. Beyond the widely accepted theory of retrograde menstruation 14, current understanding embraces several complementary hypotheses including coelomic metaplasia 15, vascular or lymphatic dissemination of endometrial cells 16, hormonal imbalances involving oestrogen and progesterone 17, immune dysregulation18 and (epi)genetic predisposition 19 – with chronic low -grade systemic inflammation as a central driver of disease progression 20–23. This heterogeneous pathophysiology is mirrored in the challenges of diagnosis and treatment. Current guidelines recommend transvaginal ultrasound and/or MRI as first-line imaging modalities due to their high sensitivity and specificity for deep and ovarian endometriosis, while laparoscopy followed by histologic confirmation, an invasive and expensive procedure, i s only recommended for individuals with negative imaging or unsuccessful empirical treatment 24. However, imaging techniques require expertise and frequently miss superficial peritoneal endometriosis, which accounts for approximately 80% of endometriosis subtypes25, highlighting the need for improved non -invasive diagnostic tests. Current therapeutic strategies include hormone -suppressive therapy, surgical removal of ectopic endometrial lesions and in some cases neuromodulation 24. However, these approaches are not curative and can lead ARTICLE IN PRESS ARTICLE IN PRESS to various side effects; in the case of surgery, high risk of complications related to laparoscopic surgery and a risk of persisting chronic postsurgical pain 26,27; in the case of hormonal treatment, different side effects can be noted, most prevalent are depressive symptoms, vasomotor symptoms, weight gain, headaches and bleeding irregularities 28–30. Together, this underscores the urgent need for non -invasive diagnostics and novel treatment strategies. Recently, the microbiome has emerged as a potential source f or diagnostic biomarkers. Bacterial DNA - and metabolite-based signatures have been shown to be implicated in the development and progression of several inflammatory conditions such as irritable bowel disease 31,32 and polycystic ovarian syndrome (PCOS)33. Emerging research from small-scale studies (n=35 in Jiminez et al., 2024; n=19 in Ata et al. 2019) indicate that endometriosis is associated with compositional differences in the microbiome across anatomical sites, including the lower and upper reproductive tracts , as well as the gastrointe stinal system34 using 16S rRNA amplicon sequencing. At a functional level , reduced concentrations of gut microbiota-derived n-butyrate in feces were found in a mouse injection model of endometriosis compared to mice without endometriosis using targeted metabolomics 35. Others have reported altered levels of 4 - hydroxyindole in stool samples of individuals with endometriosis (n = 18) compared to healthy controls (n = 33) through untargeted metabolomics36. A growing area of interest is the vaginal microbiome, which plays an essential role in reproductive health and is involved in several reproductive inflammatory conditions37,38. In most women of reproductive age, the vaginal microbiome is predominantly composed of Lactobacillus species, such as Lactobacillus crispatus and Lactobacillus iners, but also a substantial proportion of Gardnerella39. Lactobacilli appear beneficial for maintaining a healthy vaginal environment based on various epidemiological 40,41 and lab- based studies42–46, where their absence has been linked to an increased risk for various reproductive issues such as preterm birth 47–49, bacterial vaginosis 50, endometritis37 and aerobic vaginitis 38, mostly using 16S rRNA amplicon sequencing. Reduced levels of Lactobacillus and shifts towards a non -optimal vaginal ARTICLE IN PRESS ARTICLE IN PRESS microbial community have been associated with inflammation51–53, potentially contributing to the chronic inflammatory environment characterist ic of endometriosis. However, it remains unclear whether and how the vaginal microbiome is associated with endometriosis. Recently we have adopted a citizen -science approach to investigate the vaginal microbiome and general vaginal health of women in Fla nders (Belgium)39. This catalogue has enabled us in our present study to analyse associations between the vaginal microbiome a nd endometriosis based on 16S rRNA amplicon sequencing of samples from 95 women with endometriosis and 2,279 women without endometriosis. Furthermore, to characterize the multifactorial nature of endometriosis and identify conditions that may reflect shared underlying mechanisms, we assessed a broad range of infections, clinical conditions, and vaginal symptoms through questionnaires in individuals with endometriosis (n = 144) and compared these with women without endometriosis (n = 3,337). ARTICLE IN PRESS ARTICLE IN PRESS

Subfertility affects up to 47% of individuals with endometriosis in a population-based cohort To assess whether and how endometriosis symptom status influenced the vaginal microbiome composition and general health , participants with endometriosis were stra tified into two groups based on their symptom status: individuals experiencing current and past endometriosis-related symptoms (n = 66, n = 78). ‘Endometriosis-related symptoms’ in our study refer to the participant’s self-perceived disease burden, as the survey did not request specific symptom types and was not designed to capture endometriosis-related symptomatology. Current endometriosis-related symptoms were defined as burden experienced at the time of study participation, and past symptoms as burden th at had been experienced previously but was no longer present when the questionnaire was completed. This subset of individuals with endometriosis is not representative of the broader endometriosis population as the Isala project targeted a predominantly healthy population and likely includes milder disease presentations. The control group (n = 3,337) consisted of participants who reported not having endometriosis. To minimize the presence of potential undiagnosed endometriosis cases in the control group, we excluded individuals who reported dysmenorrh oea and requiring non -steroidal anti -inflammatory drugs (NSAIDs) during menstruation. Menopausal individuals were also excluded from this study. Information on the location of endometriotic lesions and treatment history was not available because health data were only collected via general surveys in the Isala platform39. A detailed overview of the participants’ demographics and additional variables are summarized in Table 1. Within the endometriosis group, individuals with past symptoms were significantly older with a median age of 36 (IQR 31-41) than those with current symptoms (31 [27-35], p < 0.001). Additionally, compared to the control group (2 8 [24-33]), individuals with endometriosis were significantly older regardless of symptom status (p < 0.001 for past symptoms group, and p = 0.006 for current symptoms group). After adjusting for age, demographic, reproductive, and lifestyle characteristics did not significantly differ across ARTICLE IN PRESS ARTICLE IN PRESS the groups, except for subfertility and self-reported health. Subfertility was significantly more prevalent in individuals with endometriosis compared to the control group (p < 0.001). Self-reported health was assessed through a five -point Likert -scale (1=very poor, 2=poor, 3=fair, 4=good, 5=very good) with a poorer perceived health in individuals with current and past endometriosis-related symptoms compared to the control group ( p < 0.001; p = 0.003). Overall, after adjustment for age, endometriosis symptom status was primarily associated with subfertility and poorer self-reported health, while other demographic and lifestyle characteristics were comparable across groups. Endometriosis shows no association with Lactobacillus dominance, genus-level diversity or abundance of individual taxa in the vagina To assess whether the vaginal microbiome composition is associated with endometriosis and its symptom status, vaginal microbiome profiles were obtained for 95 of the 144 participants with self -reported endometriosis, including 43 profiles from individuals with current endometriosis -related symptoms and 52 from individuals with past symptoms. In comparison, 2,279 vaginal microbiome profiles were analysed from the control group (Figure 2a). Overall, 79.1% of participants with current endometriosis -related symptoms had a Lactobacillus-dominated microbiota, compared to 71.2% in partici pants with past endometriosis-related symptoms and 80.1% in the control group, with no significant differences ( p = 0.371). We applied different embedding methods including t-distributed stochastic neighbour embedding (t-SNE, Figure 2b), uniform manifold approximation and projection (UMAP, Supplementary Figure 1a) and principal coordinates analysis (PCoA, Supplementary Figure 1b), but no clear grouping of bacterial profiles was observed based on current or past symptom status compared to the control group. In addition, we examined whether symptom status in participants with endometriosis was associated with the (sub)genera present in at least 10% of all participants. Each test (Maaslin, Limma, DESeq, ANCOM-BC89, and a linear regression on the centered log-ratio (CLR) transformed abundance data) was adjusted for technical confounders, age, recent sexual intercourse , number of pregnancies, phase of the menstrual ARTICLE IN PRESS ARTICLE IN PRESS cycle and use of hormonal contraceptives, and associations were considered significant only if consensus was reported by at least three differential abundance tools as described in Methods and Materials. No significant associations were observed between endometriosis symptom status and relative abundance of specific microbial taxa. Similarly, no significant associations were observed between endometriosis and genus-level diversity, including both alpha (Shannon) diversity and beta diversity, irrespective of symptom status. These findings confirm that the vaginal microbiome composition and its key f eatures play only a limited role – if any – in the pathogenesis of endometriosis. Polycystic ovary syndrome shows strong co -occurrence with endometriosis, alongside other comorbid conditions Complementing the vaginal microbiome analyses and drawing on the uniquely rich survey data collected within the Isala programme, we next assessed comorbidities to better understand the broader health profile associated with endometriosis. To characterize the spectrum of comorbidities associated with endometriosis, we analysed extensive health questionnaires from individuals with endometriosis reporting past or current endometriosis -related symptoms and compared them with controls with no reported endometriosis. The prevalence of several vaginal symptoms, urogenital infections and a range of conditions related to reproductive, metabolic and gastrointestinal health was compared between these groups, with statistical significance determined after correction for multiple testing using the Benjamini– Hochberg procedure. A summary of the results is shown in Figure 2 with the corresponding counts provided in Supplementary Table 1. Dyspareunia was significantly more common among individuals with endometriosis, with an odds ratio of 1.67 (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, p = 0.008) and migraine (OR 3.75, 95% CI 1.38-8.60, p = 0.025). Interestingly, a significant co-occurrence of polycystic ovarian syndrome in individuals with endometriosis was observed (OR 2. 92, 95% CI 1. 71– ARTICLE IN PRESS ARTICLE IN PRESS 4.78, p < 0.001). By contrast, no significant associations were observed between endometriosis and cardiovascular, dermatological, or respiratory conditions, nor with urogenital infections or self-reported vaginal symptoms. None of the comorbid conditions associated with endometriosis showed a significant association with vaginal microbiome composition or diversity (Supplementary Figure 3). Together, these findings underscore that , beyond the well -recognised menstrual -related pain, individuals with endometriosis experience a higher prevalence of several inflammatory comorbidities. This complexity highlights the multifaceted nature of the disease and supports the need for integrated , multimodal approaches to clinical management.

Despite growing recognition of endometriosis as a complex and multifactorial condition affecting reproductive health, the potential role of the vaginal microbiome in the pathogenesis of endometrios is remains largely unexplored. In this study, we leveraged data from the citizen science-driven Isala platform to investigate whether endometriosis and its symptom status was associated with differences in vaginal microbiome composition. Investigating the broad range of conditions in this population is a critical step toward understanding potential shared pathologies with comorbidities and the underlying mechanisms of endometriosis. Compared with population ‑based studies reporting endometriosis prevalence of approximately 8 - 10%54,55, the Isala cohort showed a lower overall prevalence of 3.5% . This may be explained by the setup of the Isala project, which was designed to capture a general, predominantly healthy populatio n, as defined by the World Health Organization (i.e., a state of complete physical, mental, and social well-being, not merely t he absence of disease or infirmity). Interestingly, despite the call for healthy women, a considerable number of participants with self-reported endometriosis enrolled in the study. Among these individuals, 1.6% experienced a significant physical burden of disease compared with controls (p < 0.001), ARTICLE IN PRESS ARTICLE IN PRESS indicating that disease burden is present even within this nominally healthy cohort. This highlights the complexity of self‑perceived health and may reflect broader societal factors such as the normalization of pain and stigma surrounding reproductive health, which can lead to prolonged diagnostic delays. In our study, we analysed microbiome profiles from 95 individuals with self-reported endometriosis. Since lactobacilli are documented to hav e an anti -inflammatory and antimicrobial activity in the vagina 46,56,57 and because endometriosis is clearly linked to inflammation20–23, we hypothesized that lactobacilli would be decreased in our Isala endometriosis cohort. Of interest, we observed no significant differences in Lactobacillus dominance between the endometriosis group and controls (p = 0.371 ). Importantly, the presence of a Lactobacillus-dominated microbiome do es not necessarily imply the absence of an underlying inflammatory condition. Elevated oestrogen levels, characteristic of the hormonal milieu in endometriosis58–60, promote glycogen accumulation in vaginal epithelial cells 61. The breakdown of glycogen releases simple sugars that serve as a substrate for Lactobacillus species44, potentially supporting their dominance even in the presence of underlying inflammation or disease. T his could explain why, despite the presence of endometriosis, the vaginal microbiome may still appear Lactobacillus-dominated in our study. Similarly, we found no significant associations between endometriosis – regardless of symptom status – and alpha or beta diversity, nor were there significant differences in the relative abundance of specific bacterial taxa when compared to controls. Similar findings have also been observed by a prior study with 21 individuals with endometriosis using 16S rRNA amplicon sequencing62. However, several other studies using the same technique have reported associations between specific bacterial taxa and endometriosis. For instance, a recent study comparing the vaginal microbiota of 35 individuals with endometriosis and chronic pelvic pain to 23 individuals with chronic pelvic pain alone observed an increased relative abunda nce of Streptococcus anginosus in the endometriosis group 63. Additionally, a larger study of 78 individuals with endometriosis found a higher abundance of Fusobacterium nucleatum in vaginal samples 64. Other studies, with sample sizes ranging from 10 to 37 individuals, have described ARTICLE IN PRESS ARTICLE IN PRESS elevated levels of Anaerococcus65,66, Escherichia34,67, Streptococcus34,63,68, Blautia69, as well as conflicting findings on the presence of Fannyhessea34 (formerly classified as Atopobium) in vaginal samples of individuals with endometriosis. While these studies provide valuable insights, many are limited by methodological constraints, including a lack of adjustment for confounding factors such as age and correction for multiple testin g. Reliance on laparoscopy or imaging for diagnosis, while ensuring clinical accuracy, restricts sample sizes in these studies, limiting the statistical power. Furthermore, while a few studies34,66,70 have stratified vaginal microbiota profiles based on the revised American Society for Reproductive Medicine (rASRM) classification, it is well established that endometriosis disease stage does not correlate with symptom status71,72. Given the substantial symptom burden experienced by individuals with endometriosis, stratifying cohorts based on symptom status – rather than disease stage – may facilitate the identification of clinically relevant microbiome -based biomarkers. Conversely, our present study includes a relatively large sample size, accounts for key confounders of the vaginal microbiome, and stratifies participants with endometriosis based on symptom status, offering a more rigorous and comprehensive assessment of microbial taxa potentially associated with endometriosis. Through extensive health questionnaires, we observed significant differences in the prevalence of certain comorbid conditions between individuals with and without self -reported endometriosis. For instance, irritable bowel syndrome was reported more than twice as often among individuals with endometriosis (11.1%) compared to controls ( 4.8%, OR 2.57, 95% CI 1.43 - 4.36, p = 0.008), which is consistent with ranges reported in previous studies and might reflect shared pathogenic mechanisms, such as mast cell activation73,74. In addition, migraine was found to be significantly associated with endometriosis (OR=3.75 95% CI 1.38-8.60, p = 0.025), exceeding the effect sizes reported in previous nationwide studies (OR 1.70, 95% 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 affected 34.0% of individuals with endometriosis compared to 25.1% among controls (OR 1.67, 95% CI 1.14-2.40, p = 0.033), which aligns with previous findings from Singh et al. (2020) (38.3% vs. 17.7%) 77. ARTICLE IN PRESS ARTICLE IN PRESS Interestingly, our data showed a nearly three -fold higher odds ratio of PCOS among participants with endometriosis (16.9%) compared to controls ( 4.7%, 95% CI 1.71-4.78, p < 0.001), with the latter falling within the globally reported prevalence range of 3 –11%, depending on the diagnostic criteria used 78. Notably, the association we observed was approximately twice as high as that reported in both a population-based cohort (n = 127) and an operative cohort (n = 473)79. Recent findings reported a slightly higher PCOS co‑occurrence of 24.5% among individuals with endometriosis, further support the relevance of thi s association 80. Shared risk genes and overlapping inflammatory pathways, characterized by dysregulated cytokines and chemokines, between endometriosis and PCOS may partly explain the observed comorbidity, as has been reported before 81–83. At the same time, the observed association should be interpreted with caution, as differences in healthcare utilization and diagnostic opportunities may also contribute. PCOS is commonly diagnosed using vaginal ultrasound, and individuals undergoing imaging as part of endometriosis evaluation may therefore have increased opportunities for PCOS detection. Furthermore, we found that among women trying to conceive, 47.2% with current endometriosis-related symptoms and 36.1% with past endometriosis -related symptoms reported requiring fertility treatment compared to 7.7% controls (p < 0.001). While these rates fall within ranges described in literature, it is important to highlight that our findings are derived from a population-based cohort, unlike most previous studies with subfertility cohorts 84,85, which may overestimate the true co -occurrence of subfertility in individuals with endometriosis. In addition, endometriosis is often diagnosed during fertility evaluations, which may lead to increased detection among individuals presenting with subfertility and thus partially inflate the observed association. This study has several limitations. First, the control group may include asymptomatic individuals with undiagnosed endometriosis, which could lead to an underestimation of the true prevalence of endometriosis in our cohort and distort the group comparisons. Vice versa, the endometriosis group may ARTICLE IN PRESS ARTICLE IN PRESS contain participants without a clinically confirmed diagnosis, some of whom may be false positives. Second, individuals with severe forms of endometriosis, both in terms of clinical extent and impact on quality of life, may be underrepresented in ou r cohort considering the setup of the Isala project which was intentionally designed to characterize the vaginal microbiome in a large and predominantly healthy population. Third, the available data lacked sufficient resolution to further characterize the endometriosis subgroups, and information on treatment history or symptom management was not available. Fourth, this study did not account for the potential influence of prior or current treatments on the reported symptom burden and healthcare utilization , which may have affected our findings. Finally, although we did not observe a significant association between the vaginal microbiome composition and endometriosis, functional differences at the metabolic or transcriptomic level may still be present and relevant. Despite inherent limitations, self -reported data remains a valuable tool in endometriosis research, particularly given the challenges and delays in obtaining formal diagnoses. Many individuals experience symptoms for years before receiving clinical confirmation, and self-reporting can offer a more immediate and inclusive snapshot of lived experiences across diverse populations. Moreover, capturing self-reported symptom severity and comorbid conditions can help identify underrecognized patterns and gen erate hypotheses for further study in clinical cohorts. Taken together, our findings do not support an association between the vaginal microbiome composition and endometriosis. However, we identified a pronounced co-occurrence of polycystic ovarian syndrom e with endometriosis, in addition to other comorbid conditions. These results suggest potential shared mechanisms worth further investigating and demonstrate the value of citizen science in advancing endometriosis research. We suggest that follow-up research should move beyond microbial composition focusing on functional aspects such as host - microbiome interactions, immune profiling, and metabolomic signatures, as well as explore anatomical sites beyond the vagina, which may offer more promising pathways to ward identifying biomarkers or therapeutic targets. ARTICLE IN PRESS ARTICLE IN PRESS

Study cohort and data collection This study was conducted in accordance with the ethical principles of the Declaration of Helsinki. The study protocol received ethical approval from the Ethical Committee of the Antwerp University Hospital/University of Antwerp (reference number B300201942076) and is registered on clinicaltrials.gov under the unique identifier NCT04319536. Informed consent was obtained from all participants. This cross -sectional s tudy draws its study cohort from the Isala citizen science project (https://isala.be/en/). For this study, 144 participants with self-reported endometriosis were included and stratified into two groups based on their symptom status: individuals experiencing current endometriosis- related symptoms (n = 66) and individuals experiencing symptoms in the past but without active symptoms during the study (n = 78). This was complemented by a group of 3,337 participants withou t reported endometriosis, excluding individuals who reported dysmenorrhoea and requiring NSAIDs during menstruation. Menopausal individuals were excluded from this study. All participants filled in a health questionnaire with General Data Protection Regula tion-compliant questions on the Qualtrics platform. Vaginal swabs were collected and processed as described before and stored in the in -house biobank decentralized hub to comply with the most recent GDPR -regulations in Belgium on biobanking human samples ( KB 2018/30209) 39. A total of 95 microbiome profiles were obtained from the endometriosis cohort, including 43 profiles from in dividuals with current endometriosis -related symptoms and 52 profiles from individuals with past symptoms. Additionally, 2,279 microbiome profiles were collected from participants without self-reported endometriosis. 16S rRNA amplicon sequencing, reference database and quality control Vaginal swabs were collected and processed as previously described by Lebeer et al. (2023)39. Briefly, DNA was extracted using the DNeasy PowerSoil Pro Kit (Qiagen), followed by amplification of the V4 region of ARTICLE IN PRESS ARTICLE IN PRESS the 16S rRNA gene using standard barcoded primers adapted for dual -index sequencing. PCR products were purified, quantified, and po oled in equimolar concentrations to generate sequencing libraries. Libraries were sequenced using dual -index paired -end sequencing on an Illumina MiSeq platform, including appropriate negative controls for DNA extraction and PCR. For the construction of the custom 16S reference database and processing of amplicon sequencing data, a refined taxonomic framework was developed to improve resolution within the Lactobacillus genus by defining subgenera based on phylogenetic relationships. A custom 16S rRNA refe rence database was generated using sequences from the Genome Taxonomy Database (GTDB) and adapted for use with DADA286. Sequence quality control and processing were conducted using the DADA2 pipeline, including filtering of low -quality reads, merging of paired -end reads, and removal of chimeras. Taxonomic assignment was performed using a custom 16S rRNA reference database , followed by reclassification steps to align with updated Lactobacillaceae taxonomy and the defined Lactobacillus subgenera. Additional quality control steps included removal of non -bacterial and low -quality amplicon sequence variants, as well as filterin g of samples based on read count and normalized sequencing depth. For full methodological details, we refer to Lebeer et al. (2023)39. Embeddings To explore the vaginal microbiome composition of participants with and without self -reported endometriosis three different embedding methods were applied using Bray -Curtis dissimilarity based on the relative abundances of (sub)genera within samples: t-distributed stochastic neighbour embedding (t- SNE), uniform manifold approximation and projection (UMAP), and principal coordinates analysis (PCOA). Plots were generated with ggplot2 R package. Statistical analyses ARTICLE IN PRESS ARTICLE IN PRESS Through survey data, we compared demographic variables between participants with and without self - reported endometriosis , further distinguishing between individuals reporting current versus past endometriosis-related symptoms. Normality of continuous variables including age, BMI , and number of pregnancies was assessed using the Shapiro-Wilk test. None of these variables were normally distributed and were therefore presented as median and interquartile range (IQR). For the ordinary variable ‘self - rated health’ the mean and standard deviation were calculated based on a five-point ordinary scale (very good = 5, good = 4, fair = 3, poor = 2, very poor = 1). Depending on the data type of the variable s, an appropriate age-adjusted model ( linear, logistic, Poisson, multinomial, or ordina l regression ) was implemented. Overall significance was calculated through likelihood ratio -tests for categorical variables followed by correction for multiple testing through the Benjamini-Hochberg procedure. Alpha and beta diversity measures were compa red as previously described by Lebeer et al. (2023)39. The differential abundance of taxa was performed using the in -house R package multidiffabundance, version 0.0.1 (publicly available at https://github.com/thiesgehrmann/multidiffabundance), using Maaslin2 87, Limma88, DESeq2 89, ANCOM -BC90, and a linear regression on the centered log -ratio (CLR) transformed abundance data, reporting the consensus of the five tools. Each test was adjusted for technical confounders, age, recent sexual intercourse (last 24 hours), number of pregnancies, phase of cycle, use of hormonal contraceptives, and library read concentratio n. Correction for multiple testing was applied within each tool using the Benjamini –Hochberg false discovery rate procedure. Data was processed in R version 4.2.2 using the in-house developed package Tidytacos91 and the tidyverse set of packages. Results were visualised in Python 3, selecting only a subset of taxa based on their known importance in the vaginal microbiome. To investigate associations between endometriosis and a range of symptoms, infections, and clinical conditions, we conducted multiple regression analyses, with all outcomes defined as lifetime (‘ever’) self- reported occurrence of the respective infection, condition, or vaginal symptom. The selection of potential ARTICLE IN PRESS ARTICLE IN PRESS confounders was determined using a directed acyclic graph (DAG; Supplementary Figure 2), which allows the identification of the minimum adjustment required, while avoiding inappropriate ‘overadjustment’ for mediator variables 92. The following covariates were included in our models: age, BMI and education level. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated and presented in a forest plot. P-values were corrected for multiple testing through the Benjamini–Hochberg procedure. Data availability Sequencing data are available at the European Nucleotide Archive (ENA) under bioproject PREB50407. Sample me tadata are available with access control through the European Genome –Phenome Archive (EGA) under dataset ID EGAD00001009890. Access is granted as described, upon agreement to the harmonised Data Access Agreement developed by EU-STANDS4PM (European Union standards for in silico models for personalised medicine; https://doi.org/10.6084/m9.figshare.23904300). Acknowledgments The authors would like to thank all Isala participants and following colleagues and students who were instrumental for the Isala sampling campaign and processing : I. Tuyaerts, N. Van Vliet, L. Van Ham, M. Legein, D. Vandenheuvel, E. Cauwenberghs, L. Delanghe, A. Groenwals, S. El Messaoudi, J. Hiers, L. Van Dyck, C. Dricot, L. Leysen and L. Martin Diaz. Strategic support was provided by L. Talboom and L. Haesevoets (Studio Maria, communication), C. Varszegi (Little Big Things, website, https://littlebigthings.be), R. Broms and S. Vergauwen (Sensoa vzw, sexual lifestyle quest ions), E. Den Hond and C. Franken (Provinciaal Instituut voor Hygiëne Antwerpen, population survey), J. Raes (KU Leuven, Flemish gut flora project), K. Scott (food-related questions), K. Wuyts and R. Samson (urbanization and contact with urban green), the Antwerp Biobank (University Hospital Antwerp) and Centre of Medical Genetics (University Hospital Antwerp, sequencing support). The authors acknowledge the European Research Council (starting grant Lacto -Be 852600 of S.L., with S.A., T.G., S.W., W.V.B., T. E. and J.D. ARTICLE IN PRESS ARTICLE IN PRESS appointed on the project), the Special Research Fund of the Universiteit Antwerpen (UA BOF; DOCPRO 37054 grant of S.A. and temporary mandate grant 48145 of S.W.), the Inter -University Special Research Fund of Flanders (iBOF; POSSIBL project), BOF funding and proof -of-concept VALERIE (Horizon, grant ID 101213306), the industrial research fund UAntwerpen (IOF service platform microbiome sequencing) and the Research Foundation —Flanders (FWO; aspirant fundamental research grant 11A0620N and postdoctoral fellowship 12AZ624N of S.W., postdoctoral fellowship 12S4222N of I.D.B , senior post - doctoral research grant 1277222N of I.S., aspirant strategic basic research grant 1SD0622N of L.V.D. and Research projects G049022N, G031222N and S006426 of S.L.). The funders had no role in study design, data collection, data analysis, data interpretation, or writing of the manuscript. Author contributions S.A., S.W., G.D., V.V. and S.L. designed the study and worked on the conceptualization of the research project. 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 out the experimental and logistical work. I.D.B. was responsible for the biobanking of all collected samples. I.R. analy sed the health questionnaire data and performed the statistical analyses. T.G., T.V.R. and S.W. processed the sequencing data and performed the biostatistical analyses. T.V.R. adjusted the microbiome data with the new taxonomy. I.R. and T.G. worked on the visualizations. I.R., T.G., S.A., I.D.B., C.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 original manuscript. All authors contributed to reviewing and editing of the final manuscript. Competing interests S.L. declares to be a voluntary academic board member of the International Scientific Association on Probiotics and Prebiotics (ISAPP, www.isappscience.org), cofounder of YUN and scientific advisor for Freya Biosciences. The team of S.L. declare s research funding from YUN, Bioorg, Puratos, DSM I -Health and Lesaffre/Gnosis. None of these organizations or companies were involved in the design or data analysis ARTICLE IN PRESS ARTICLE IN PRESS of this study, which was fully funded by the university, governmental, and European funding. S.A. declares to be a voluntary member of the student and fellows association of ISAPP. The other authors declare no competing interests. Declaration of Generative AI and AI-assisted technologies in the writing process During the preparation of this wor k, the author(s) used ChatGPT by OpenAI exclusively to assist with paraphrasing and to gather inspiration for scientific rephrasing. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication.

1. Zondervan, K. T., Becker, C. M. & Missmer, S. A. Endometriosis. N Engl J Med 382, 1244–1256 (2020). 2. Von Theobald, P., Cottenet, J., Iacobelli, S. & Quantin, C. Epidemiology of Endometriosis in France: A Large, Nation-Wide Study Based on Hospital Discharge Data. BioMed Research International 2016, 1–6 (2016). 3. Reid, R. et al. The prevalence of self-reported diagnosed endometriosis in the Australian population:

from a nationally-representative survey. BMC Res Notes 12, 88 (2019). 4. Christ, J. P. et al. Incidence, prevalence, and trends in endometriosis diagnosis: a United States population-based study from 2006 to 2015. American Journal of Obstetrics and Gynecology 225, 500.e1-500.e9 (2021). 5. Illum, L. R. H. et al. Temporal and regional differences in the incidence of hospital‑diagnosed endometriosis: a Danish population‑based study. Acta Obstet Gynecol Scand 101, 737–746 (2022). 6. Becker, K. et al. Real world data on symptomology and diagnostic approaches of 27,840 women living with endometriosis. Sci Rep 11, 20404 (2021). ARTICLE IN PRESS ARTICLE IN PRESS 7. Prescott, J. et al. A prospective cohort study of endometriosis and subsequent risk of infertility. Hum. Reprod. 31, 1475–1482 (2016). 8. Ek, M. et al. Gastrointestinal symptoms among endometriosis patients—A case-cohort study. BMC Women’s Health 15, 59 (2015). 9. Yang, F. et al. Evidence of shared genetic factors in the etiology of gastrointestinal disorders and endometriosis and clinical implications for disease management. Cell Reports Medicine 4, 101250 (2023). 10. Gambadauro, P., Carli, V. & Hadlaczky, G. Depressive symptoms among women with endometriosis: a systematic review and meta-analysis. American Journal of Obstetrics and Gynecology 220, 230– 241 (2019). 11. Nnoaham, K. E. et al. Impact of endometriosis on quality of life and work productivity: A multicenter study across ten countries. Fertility and Sterility 96, 366-373.e8 (2011). 12. Simoens, S. et al. The burden of endometriosis: Costs and quality of life of women with endometriosis and treated in referral centres. Human Reproduction 27, 1292–1299 (2012). 13. Gete, D. G. et al. Hospitalisations and length of stays in women with endometriosis: a data linkage prospective cohort study. EClinicalMedicine 80, 103030 (2025). 14. Sampson, J. A. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. American Journal of Obstetrics and Gynecology 14, 422–469 (1927). 15. Konrad, L., Dietze, R., Kudipudi, P. K., Horné, F. & Meinhold-Heerlein, I. Endometriosis in MRKH cases as a proof for the coelomic metaplasia hypothesis? Reproduction 158, R41–R47 (2019). 16. Sampson, J. A. Metastatic or Embolic Endometriosis, due to the Menstrual Dissemination of Endometrial Tissue into the Venous Circulation. Am J Pathol 3, 93-110.43 (1927). 17. Attia, G. R. et al. Progesterone Receptor Isoform A But Not B Is Expressed in Endometriosis1. The Journal of Clinical Endocrinology & Metabolism 85, 2897–2902 (2000). ARTICLE IN PRESS ARTICLE IN PRESS 18. Lin, Y.-J., Lai, M.-D., Lei, H.-Y. & Wing, L.-Y. C. Neutrophils and Macrophages Promote Angiogenesis in the Early Stage of Endometriosis in a Mouse Model. Endocrinology 147, 1278–1286 (2006). 19. Rahmioglu, N. et al. The genetic basis of endometriosis and comorbidity with other pain and inflammatory conditions. Nat Genet 55, 423–436 (2023). 20. Ahn, S. H. et al. IL-17A Contributes to the Pathogenesis of Endometriosis by Triggering Proinflammatory Cytokines and Angiogenic Growth Factors. The Journal of Immunology 195, 2591– 2600 (2015). 21. Chang, K.-K. et al. IL-27 triggers IL-10 production in Th17 cells via a c-Maf/RORγt/Blimp-1 signal to promote the progression of endometriosis. Cell Death Dis 8, e2666–e2666 (2017). 22. Wilson, T. R. et al. Neutrophils initiate proinflammatory immune responses in early endometriosis lesion development. JCI Insight 10, e186133 (2025). 23. Vallvé-Juanico, J., Santamaria, X., Vo, K. C., Houshdaran, S. & Giudice, L. C. Macrophages display proinflammatory phenotypes in the eutopic endometrium of women with endometriosis with relevance to an infectious etiology of the disease. Fertility and Sterility 112, 1118–1128 (2019). 24. Becker, C. M. et al. ESHRE guideline: endometriosis. Human Reproduction Open 2022, hoac009 (2022). 25. Whitaker, L. H. R. et al. Laparoscopic treatment of isolated superficial peritoneal endometriosis for managing chronic pelvic pain in women: study protocol for a randomised controlled feasibility trial (ESPriT1). Pilot Feasibility Stud 7, (2021). 26. Leborne, P. et al. Clinical outcomes following surgical management of deep infiltrating endometriosis. Sci Rep 12, (2022). 27. Martinez, V. et al. Chronic postsurgical pain: A European survey. European Journal of Anaesthesiology 41, 351–362 (2024). ARTICLE IN PRESS ARTICLE IN PRESS 28. De Corte, P. et al. Endometriosis and risk of depression among oral contraceptive users: a pooled analysis of cohort studies from 13 countries. Human Reproduction 40, 479–486 (2025). 29. Lockhat, F. B., Emembolu, J. O. & Konje, J. C. The efficacy, side-effects and continuation rates in women with symptomatic endometriosis undergoing treatment with an intra-uterine administered progestogen (levonorgestrel): a 3 year follow-up. Human Reproduction 20, 789–793 (2005). 30. Donnez, J. et al. Treatment of endometriosis-associated pain with linzagolix, an oral gonadotropin- releasing hormone–antagonist: a randomized clinical trial. Fertility and Sterility 114, 44–55 (2020). 31. Phan, J. et al. Alterations in Gut Microbiome Composition and Function in Irritable Bowel Syndrome and Increased Probiotic Abundance with Daily Supplementation. mSystems 6, e01215-21 (2021). 32. Jacobs, J. P. et al. Multi-omics profiles of the intestinal microbiome in irritable bowel syndrome and its bowel habit subtypes. Microbiome 11, 5 (2023). 33. Yang, Y. et al. Gut microbiota in women with polycystic ovary syndrome: an individual based analysis of publicly available data. eClinicalMedicine 77, 102884 (2024). 34. Ata, B. et al. The Endobiota Study: Comparison of Vaginal, Cervical and Gut Microbiota Between Women with Stage 3/4 Endometriosis and Healthy Controls. Scientific Reports 9, 1–9 (2019). 35. Chadchan, S. B. et al. Gut microbiota and microbiota-derived metabolites promotes endometriosis. Cell Death Discovery 9, (2023). 36. Talwar, C. et al. Identification of distinct stool metabolites in women with endometriosis for non- invasive diagnosis and potential for microbiota-based therapies. Med 6, 100517 (2025). 37. Haggerty, C. L., Hillier, S. L., Bass, D. C., Ness, R. B., & the PID Evaluation and Clinical Health (PEACH) Study Investigators. Bacterial Vaginosis and Anaerobic Bacteria Are Associated with Endometritis. CLIN INFECT DIS 39, 990–995 (2004). 38. Oerlemans, E. F. M. et al. The dwindling microbiota of aerobic vaginitis, an inflammatory state enriched in pathobionts with limited TLR stimulation. Diagnostics 10, 1–15 (2020). ARTICLE IN PRESS ARTICLE IN PRESS 39. Lebeer, S. et al. A citizen-science-enabled catalogue of the vaginal microbiome and associated factors. Nat Microbiol 8, 2183–2195 (2023). 40. Huo, Y., Jiang, Q. & Zhao, W. Meta-analysis of metagenomics reveals the signatures of vaginal microbiome in preterm birth. Medicine in Microecology 14, 100065 (2022). 41. Huang, C. et al. Meta-analysis reveals the vaginal microbiome is a better predictor of earlier than later preterm birth. BMC Biol 21, 199 (2023). 42. Spurbeck, R. R. & Arvidson, C. G. Lactobacillus gasseri Surface-Associated Proteins Inhibit Neisseria gonorrhoeae Adherence to Epithelial Cells. Infect Immun 80, 3743–3743 (2012). 43. Nunn, K. L. et al. Enhanced Trapping of HIV-1 by Human Cervicovaginal Mucus Is Associated with Lactobacillus crispatus-Dominant Microbiota. mBio 6, e01084-15 (2015). 44. Jenkins, D. J. et al. Bacterial amylases enable glycogen degradation by the vaginal microbiome. Nat Microbiol https://doi.org/10.1038/s41564-023-01447-2 (2023) doi:10.1038/s41564-023-01447-2. 45. Izadifar, Z. et al. Mucus production, host-microbiome interactions, hormone sensitivity, and innate immune responses modeled in human cervix chips. Nat Commun 15, 4578 (2024). 46. Decout, A. et al. Lactobacillus crispatus S-layer proteins modulate innate immune response and inflammation in the lower female reproductive tract. Nat Commun 15, 10879 (2024). 47. Gudnadottir, U. et al. The vaginal microbiome and the risk of preterm birth: a systematic review and network meta-analysis. Sci Rep 12, 7926 (2022). 48. Chopra, C., Kumar, V., Kumar, M. & Bhushan, I. Role of vaginal microbiota in idiopathic infertility: a prospective study. Microbes and Infection 26, 105308 (2024). 49. Fettweis, J. M. et al. The vaginal microbiome and preterm birth. Nature Medicine 25, 1012–1021 (2019). 50. Hardy, L. et al. A fruitful alliance: the synergy between Atopobium vaginae and Gardnerella vaginalis in bacterial vaginosis-associated biofilm. Sex Transm Infect 92, 487–491 (2016). ARTICLE IN PRESS ARTICLE IN PRESS 51. Berard, A. R. et al. Vaginal epithelial dysfunction is mediated by the microbiome, metabolome, and mTOR signaling. Cell Reports 42, 112474 (2023). 52. Łaniewski, P. & Herbst-Kralovetz, M. M. Bacterial vaginosis and health-associated bacteria modulate the immunometabolic landscape in 3D model of human cervix. npj Biofilms Microbiomes 7, (2021). 53. Delgado-Diaz, D. J. et al. Distinct Immune Responses Elicited From Cervicovaginal Epithelial Cells by Lactic Acid and Short Chain Fatty Acids Associated With Optimal and Non-optimal Vaginal Microbiota. Front. Cell. Infect. Microbiol. 9, (2020). 54. Vos, T. et al. Global, regional, and national incidence, prevalence, and years lived with disability for 301 acute and chronic diseases and injuries in 188 countries, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013. The Lancet 386, 743–800 (2015). 55. Wang, M.-H., Chen, J.-H., Qi, X.-Y., Li, Z.-X. & Huang, Y. Global prevalence of adenomyosis and endometriosis: a systematic review and meta-analysis. Reprod Biol Endocrinol 23, 148 (2025). 56. Argentini, C. et al. Evaluation of Modulatory Activities of Lactobacillus crispatus Strains in the Context of the Vaginal Microbiota. Microbiol Spectr 10, (2022). 57. Hearps, A. C. et al. Vaginal lactic acid elicits an anti-inflammatory response from human cervicovaginal epithelial cells and inhibits production of pro-inflammatory mediators associated with HIV acquisition. Mucosal Immunology 10, 1480–1490 (2017). 58. Han, S. J., Lee, J. E., Cho, Y. J., Park, M. J. & O’Malley, B. W. Genomic Function of Estrogen Receptor β in Endometriosis. Endocrinology 160, 2495–2516 (2019). 59. Noble, L. S. et al. Prostaglandin E2 Stimulates Aromatase Expression in Endometriosis-Derived Stromal Cells1. The Journal of Clinical Endocrinology & Metabolism 82, 600–606 (1997). 60. Han, S. J. et al. Estrogen Receptor β Modulates Apoptosis Complexes and the Inflammasome to Drive the Pathogenesis of Endometriosis. Cell 163, 960–974 (2015). ARTICLE IN PRESS ARTICLE IN PRESS 61. Navarro, S. et al. Glycogen availability and pH variation in a medium simulating vaginal fluid influence the growth of vaginal Lactobacillus species and Gardnerella vaginalis. BMC Microbiol 23, 186 (2023). 62. Hernandes, C. et al. Microbiome Profile of Deep Endometriosis Patients: Comparison of Vaginal Fluid, Endometrium and Lesion. Diagnostics 10, 163 (2020). 63. Jimenez, N. et al. Vaginal and rectal microbiome contribute to genital inflammation in chronic pelvic pain. BMC Med 22, 283 (2024). 64. Muraoka, A., Yokoi, A. & Kajiyama, H. Emerging bacterial factors for understanding pathogenesis of endometriosis. iScience 27, 108739 (2024). 65. MacSharry, J. et al. Endometriosis specific vaginal microbiota links to urine and serum N-glycome. Sci Rep 14, 25372 (2024). 66. Perrotta, A. R. et al. The Vaginal Microbiome as a Tool to Predict rASRM Stage of Disease in Endometriosis: a Pilot Study. Reproductive Sciences 27, 1064–1073 (2020). 67. Sessa, R. et al. Characterization of the vaginal microbiota in Italian women with endometriosis: preliminary study. Arch Gynecol Obstet 310, 2141–2151 (2024). 68. Akiyama, K. et al. Molecular detection of microbial colonization in cervical mucus of women with and without endometriosis. American journal of reproductive immunology (New York, N.Y. : 1989) 82, e13147 (2019). 69. Hicks, C. et al. Oral, Vaginal, and Stool Microbial Signatures in Patients With Endometriosis as Potential Diagnostic Non‑Invasive Biomarkers: A Prospective Cohort Study. BJOG 1471–0528.17979 (2024) doi:10.1111/1471-0528.17979. 70. Wei, W., Zhang, X., Tang, H., Zeng, L. & Wu, R. Microbiota composition and distribution along the female reproductive tract of women with endometriosis. Annals of Clinical Microbiology and Antimicrobials 19, 1–9 (2020). ARTICLE IN PRESS ARTICLE IN PRESS 71. Schliep, K. C. et al. Pain typology and incident endometriosis. Hum Reprod 30, 2427–2438 (2015). 72. Johnson, N. P. et al. World Endometriosis Society consensus on the classification of endometriosis. Hum. Reprod. 32, 315–324 (2017). 73. Guilarte, M. et al. Peripheral Corticotropin-Releasing Factor Triggers Jejunal Mast Cell Activation and Abdominal Pain in Patients With Diarrhea-Predominant Irritable Bowel Syndrome. Am J Gastroenterol 115, 2047–2059 (2020). 74. Mariuzzi, L. et al. Functional expression of aryl hydrocarbon receptor on mast cells populating human endometriotic tissues. Laboratory Investigation 96, 959–971 (2016). 75. Yang, M.-H. et al. Women with Endometriosis Are More Likely to Suffer from Migraines: A Population-Based Study. PLoS ONE 7, e33941 (2012). 76. Gete, D. G., Doust, J., Mortlock, S., Montgomery, G. & Mishra, G. D. Associations between endometriosis and common symptoms: findings from the Australian Longitudinal Study on Women’s Health. American Journal of Obstetrics and Gynecology 229, 536.e1-536.e20 (2023). 77. Singh, S. et al. Prevalence, Symptomatic Burden, and Diagnosis of Endometriosis in Canada: Cross- Sectional Survey of 30 000 Women. Journal of Obstetrics and Gynaecology Canada 42, 829–838 (2020). 78. Neven, A. C. H. et al. Prevalence and accurate diagnosis of polycystic ovary syndrome in adolescents across world regions: a systematic review and meta-analysis. European Journal of Endocrinology 191, S15–S27 (2024). 79. Schliep, K. C. et al. Examining the co-occurrence of endometriosis and polycystic ovarian syndrome. AJOG Global Reports 3, 100259 (2023). 80. Kaveh, M. et al. Epidemiological study and clinical consequences of endometriosis and PCOS co- occurrence in a clinical population of women: a cross-sectional study. BMC Women’s Health 26, 182 (2026). ARTICLE IN PRESS ARTICLE IN PRESS 81. Tan, H., Long, P. & Xiao, H. Dissecting the shared genetic architecture between endometriosis and polycystic ovary syndrome. Front. Endocrinol. 15, 1359236 (2024). 82. Eriksson, G. et al. Single-cell profiling of the human endometrium in polycystic ovary syndrome. Nat Med https://doi.org/10.1038/s41591-025-03592-z (2025) doi:10.1038/s41591-025-03592-z. 83. Omidvar-Mehrabadi, A., Ebrahimi, F., Shahbazi, M. & Mohammadnia-Afrouzi, M. Cytokine and chemokine profiles in women with endometriosis, polycystic ovary syndrome, and unexplained infertility. Cytokine 178, 156588 (2024). 84. Saraswat, L. et al. O-250 Infertility and endometriosis: a 30-yearlong national population-based study of prevalence, association and pregnancy outcomes. Human Reproduction 40, deaf097.250 (2025). 85. Leibson, C. L. et al. Incidence and characterization of diagnosed endometriosis in a geographically defined population. Fertility and Sterility 82, 314–321 (2004). 86. Callahan, B. J. et al. DADA2: High-resolution sample inference from Illumina amplicon data. Nat

13, 581–583 (2016). 87. Mallick, H. et al. Multivariable association discovery in population-scale meta-omics studies. PLoS Comput Biol 17, e1009442 (2021). 88. Law, C. W., Chen, Y., Shi, W. & Smyth, G. K. voom: precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol 15, R29 (2014). 89. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15, 550 (2014). 90. Lin, H. & Peddada, S. D. Analysis of compositions of microbiomes with bias correction. Nat Commun 11, 3514 (2020). 91. Wittouck, S., Rillaer, T. V., Smets, W. & Lebeer, S. Tidytacos: An R package for analyses on taxonomic composition of microbial communities. JOSS 10, 6313 (2025). ARTICLE IN PRESS ARTICLE IN PRESS 92. Textor, J., Van Der Zander, B., Gilthorpe, M. S., Liśkiewicz, M. & Ellison, G. T. H. Robust causal inference using directed acyclic graphs: the R package ‘dagitty’. Int. J. Epidemiol. dyw341 (2017) doi:10.1093/ije/dyw341. Tables and Figures Table 1. Demographic characteristics of participants with and without self -reported endometriosis. The former group is further categorised into participants experiencing current and past endometriosis -related symptoms. Percentages were calculated based on the tota l number of participants within each group. Significantly different variables between the three pairwise comparisons: control vs. past symptoms (1), control vs. current symptoms (2) and current symptoms vs. past symptoms (3) are represented in the last column with an asterisk. Significant results ARTICLE IN PRESS ARTICLE IN PRESS are indicated with an asterisk. IQR: interquartile range, BMI: body mass index, HPV: humanpapillomavirus. Self-rated health is a five-point ordinary scale (very good = 5, good = 4, fair = 3, poor = 2, very poor = 1). Characteristic Control (N = 3337) Endometriosis, past complaints (N = 78) Endometriosis, current complaints (N = 66) P-value Age (years) Median [IQR] 28 [24-33] 36 [31-41] 31 [27-35] < 0.001 (1;3) 0.006 (2) BMI (kg/m2) Median [IQR] 23.1 [21.0-26.2] 23.0 [21.3-26.4] 23.6 [21.4-27.5] 0.311 Birth delivery mode 0.766 Vaginal 2964 (88.8%) 72 (92.3%) 57 (86.4%) Caesarean section 332 (10.0%) 4 (5.1%) 7 (10.6%) Missing 41 (1.2%) 2 (2.6%) 2 (3.0%) Contraception 0.142 Combined oral contraceptives 805 (24.1%) 12 (15.4%) 13 (19.7%) Progesterone only pill 31 (0.9%) 4 (5.1%) 3 (4.5%) Hormonal intrauterine device 294 (8.8%) 10 (12.8%) 9 (13.6%) Other or none 2207 (66.2%) 52 (66.7%) 41 (62.2%) Marital status during last 3 months1 0.063 No partner 757 (22.7%) 15 (19.2%) 8 (12.1%) One partner 2510 (75.2%) 55 (70.5%) 55 (83.3%) Multiple partners 70 (2.1%) 8 (10.3%) 3 (4.6%) Subfertility 85/1108 (7.7%) 17/47 (36.2%) 17/36 (47.2%) < 0.001 (1;2) Number of pregnancies Median [IQR] 0 [0-1] 2 [0-2] 0 [0-2] 0.285 Biological child(ren) 1080 (32.4%) 47 (60.7%) 25 (37.9%) 0.813 Allergies and intolerances Gluten allergy 65 (2.0%) 3 (3.9%) 4 (6.1%) 0.236 Lactose intolerant 253 (7.6%) 11 (14.1%) 9 (13.6%) 0.142 Antibiotics during last 3 months 690 (0.7%) 20 (25.6%) 17 (25.8%) 0.467 (1) (3) (2) ARTICLE IN PRESS ARTICLE IN PRESS 1 Without implying cohabitation. Current smoker 298 (8.9%) 10 (12.8%) 2 (3.0%) 0.213 Current drug user 319 (9.6%) 3 (3.9%) 4 (6.1%) 0.467 HPV vaccinated 1416 (42.4%) 14 (18.0%) 17 (25.8%) 0.236 Self-rated health 4.12 ± 0.64 3.85 ± 0.72 3.79 ± 0.65 0.003 (1) < 0.001 (2) ARTICLE IN PRESS ARTICLE IN PRESS Figure 1. Overview of cohort size, microbiome composition, and bacterial diversity in vaginal samples of individuals with self -reported endometriosis (Endo) and controls. (a) Overview of participants, collected metadata and vaginal microbiome profiles. (b) t-SNE analysis. (c) Bar plot showing the vaginal microbiome profiles of participants with self -reported endometriosis (current versus past endometriosis -related symptoms), and without self-reported endometriosis (control group) on genus level. d) Associations on the level of beta diversity between the samples; Associations on the level of alpha diversity of the samples ; Associations on the level of abundance of specific taxa analysed by five different differential abundance testing methods (Limma, Maaslin2, DESeq2, ANCOM-BC, and a linear regression on the centered log-ratio transformed abundance data). ARTICLE IN PRESS ARTICLE IN PRESS ARTICLE IN PRESS ARTICLE IN PRESS Figure 2. Associations between endometriosis and a range of infections, clinical conditions, and vaginal symptoms. Forest plot shows adjusted odds ratios (ORs) with 95% confidence intervals derived from multivariable logistic regression models. All outcomes refer to ever having had the infection, condition, or vaginal symptom, as self-reported by participants. Analyses were adjusted for age, body mass index, and educational level. P -values were corrected for multiple testing using the Benjamini–Hochberg false discovery rate procedure. *P-value < 0.05, **P-value < 0.01, ***P-value < 0.001