{"paper_id":"6a27eae2-0bcc-4caa-bc6f-8aeaa46362f2","body_text":"RESEARCH Open Access\n© The Author(s) 2024. Open Access  This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, \nsharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and \nthe source, provide a link to the Creative Commons licence, and indicate if changes were made. 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 \nin the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will \nneed to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The \nCreative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available \nin this article, unless otherwise stated in a credit line to the data.\nDang et al. BMC Women's Health          (2024) 24:123 \nhttps://doi.org/10.1186/s12905-024-02945-z\nBMC Women's Health\n†Chunxiao Dang and Zhenting Chen contributed equally to this \nwork.\n*Correspondence:\nYan Liu\n200962000811@sdu.edu.cn\nJinxing Liu\nLjx276@sina.com\nFull list of author information is available at the end of the article\nAbstract\nBackground An increasing body of observational studies have indicated an association between gut microbiota and \nendometriosis. However, the causal relationship between them is not yet clear. In this study, we employed Mendelian \nrandomization method to investigate the causal relationship between 211 gut microbiota taxa and endometriosis.\nMethods Independent genetic loci significantly associated with the relative abundance of 211 gut microbiota \ntaxa, based on predefined thresholds, were extracted as instrumental variables. The primary analytical approach \nemployed was the IVW method. Effect estimates were assessed primarily using the odds ratio and 95% confidence \nintervals. Supplementary analyses were conducted using MR-Egger regression, the weighted median method, the \nsimple mode and the weighted mode method to complement the IVW results. In addition, we conducted tests for \nheterogeneity, horizontal pleiotropy, sensitivity analysis, and MR Steiger to assess the robustness of the results and the \nstrength of the causal relationships.\nResults Based on the IVW method, we found that the family Prevotellaceae, genus Anaerotruncus, genus Olsenella, \ngenus Oscillospira, and order Bacillales were identified as risk factors for endometriosis, while class Melainabacteria \nand genus Eubacterium ruminantium group were protective factors. Additionally, no causal relationship was observed \nbetween endometriosis and gut microbiota. Heterogeneity tests, pleiotropy tests, and leave-one-out sensitivity \nanalyses did not detect any significant heterogeneity or pleiotropic effects.\nConclusions Our MR study has provided evidence supporting a potential causal relationship between gut \nmicrobiota and endometriosis, and it suggests the absence of bidirectional causal effects. These findings could \npotentially offer new insights for the development of novel strategies for the prevention and treatment of \nendometriosis.\nAssessing the relationship between gut \nmicrobiota and endometriosis: a bidirectional \ntwo-sample mendelian randomization \nanalysis\nChunxiao Dang1†, Zhenting Chen2†, Yuyan Chai3, Pengfei Liu1, Xiao Yu4, Yan Liu5* and Jinxing Liu1*\n\nPage 2 of 10\nDang et al. BMC Women's Health          (2024) 24:123 \nIntroduction\nEndometriosis (EMs) is a chronic, estrogen-dependent \ninflammatory condition characterized by the presence \nof endometrial tissue outside the uterus [ 1]. Approxi -\nmately 6–10% of women of reproductive age are affected \nby EMs, and about 50% of infertile women have EMs \n[2, 3]. Due to the secretive and diverse nature of EMs \nsymptoms, and the lack of reliable non-invasive meth -\nods for detecting endometriosis, it often goes unno -\nticed. In recent years, the gut microbiota has emerged \nas a research hotspot, with scholars [ 4– 6] discovering its \nassociations with various diseases such as gastrointesti -\nnal disorders, cardiovascular diseases, respiratory dis -\neases, and more. Research on the relationship between \ngut microbiota and endometriosis has spanned over two \ndecades, starting as early as the 1990s and continuing to \nthe present day. Many scholars have observed significant \ndifferences in the types, distribution, and abundance of \ngut microbiota between patients with EMs and healthy \nwomen [ 7, 8]. Additionally, up to 90% of EMs patients \nexperience gastrointestinal issues such as nausea, vom -\niting, diarrhea, and bloating [ 9], suggesting a potential \nimbalance in the gut microbiota. In fact, in a large-scale \nstudy, EMs patients were found to have a 50% increased \nrisk of developing inflammatory bowel disease (IBD) \ncompared to the general population [ 10]. Furthermore, \necological imbalances in the gut, vagina, or uterus in EMs \npatients may impact estrogen metabolism, immune sys -\ntem balance, and exacerbate the condition [11, 12]. How-\never, in observational studies, the relationship between \ngut microbiota and endometriosis can be influenced by \nconfounding factors (such as age and surgical history) \nand reverse causality, making it uncertain whether these \nassociations are causal in nature.\nRandomized controlled trials (RCTs) are considered \nthe gold standard in epidemiology for inferring causal \nrelationships. However, due to ethical constraints, \nimplementing RCTs can be challenging [ 13]. Mendelian \nrandomization (MR) utilizes single nucleotide polymor -\nphism (SNP) loci as instrumental variables to infer causal \nassociations between exposures and outcomes. It does \nso by adhering to the genetic principle of “random allo -\ncation of parental alleles to offspring, ” achieving similar \nrandomization effects without being influenced by exter -\nnal environmental factors, thus compensating for the \nlimitations of observational studies [14].\nCurrently, there are no MR reports regarding a causal \nrelationship between gut microbiota and endometriosis. \nAlthough previous observational studies have suggested \nan association between gut microbiota and the incidence \nand progression of endometriosis, the causal relationship \nis not yet clear. This study is the first application of a two-\nsample Mendelian randomization approach to explore \nthe causal association between gut microbiota and endo -\nmetriosis. It aims to provide new insights into the treat -\nment and prevention of endometriosis.\nMaterials and methods\nResearch design\nIn a scenario where the genome wide association study \n(GWAS) summary data for the exposure variable and \nthe GWAS summary data for the outcome variable are \nmutually independent, this study employed the TwoSam-\npleMR package in R programming language to conduct a \ntwo-sample bidirectional Mendelian randomization anal-\nysis. The objective was to investigate the causal associa -\ntion between gut microbiota and endometriosis, with the \nspecific design as shown in Fig.  1. MR analysis adheres \nto three crucial assumptions [ 15]: First, the instrumental \nvariables are strongly correlated with the exposure vari -\nable. Second, the instrumental variables are independent \nof observed or unobserved confounding factors. Third, \nthe instrumental variables affect the outcome solely \nthrough the exposure.\nData source\nThe GWAS summary data for endometriosis were \nobtained from the Finngen database, which includes data \nfrom 77,257 European participants and covers 16,377,306 \nSNPs ( https://gwas.mrcieu.ac.uk/datasets/finn-b-N14_\nENDOMETRIOSIS/). The statistical data on gut micro -\nbiota were derived from the research conducted by the \nMiBioGen Consortium ( http://www.mibiogen.org/), \nwhich incorporated 18,340 individuals from 24 cohorts, \nmainly from Europe [ 16]. Microbial composition was \nanalyzed using three distinct variable regions of the tar -\ngeted 16  S rRNA gene, namely V4 (10,413 samples, 13 \ncohorts), V3-V4 (4,211 samples, 6 cohorts), and V1-V2 \n(3,716 samples, 5 cohorts). Supplementary File 1 shows \na description of the participants in each cohort in a data -\nset of gut microbiota. Both gut microbiota and endome -\ntriosis were selected as exposure and outcome variables, \nrespectively, for the MR analysis. As our study is based on \npublicly available databases, ethical committee approval \nwas not required.\nInstrumental variable selection\n(1) IVs Selection: To obtain strongly related expo -\nsure data, SNPs with a significance level of P < 5 × 10− 8 \nwere selected as conditions. Given that gut microbiota \nSNPs rarely have P < 5 × 10− 8, gut microbiota SNPs were \nselected with a threshold of P < 1 × 10− 5. (2) Independence \nKeywords Mendelian randomization study, Gut microbiota, Endometriosis, Causal effects\n\nPage 3 of 10\nDang et al. BMC Women's Health          (2024) 24:123 \nCriterion: The PLINK aggregation method was used to \ncalculate linkage disequilibrium (LD) between each risk \nfactor’s SNPs. SNPs with an LD coefficient r2 > 0.001 and \na physical distance of less than 10,000 kb were removed \nto ensure that the SNPs were mutually independent \nand to eliminate the influence of genetic pleiotropy on \nthe results [ 17, 18]. (3) Statistical Strength Criteria: The \nstrength of the instrumental variables was calculated \nusing the F-statistic, with the formula: F = β2 / SE2 (where \nβ is the allele effect size and SE is the standard error). \nInstrumental variables with F < 10 were removed to \nensure that the instrumental variables were unrelated to \nunmeasured confounding factors [ 19]. Finally, the “har -\nmonise_data” function from the TwoSampleMR package \nwas used to align the direction of alleles between expo -\nsure and outcome, remove palindromic and incompatible \nSNPs [ 20], and exclude SNPs with confounding factors \nthrough the PhenoScanner database ( http://www.phe-\nnoscanner.medschl.cam.ac.uk/).\nMendelian randomization analysis\nIn this study, the inverse variance weighted (IVW) \nmethod [ 21] was employed as the primary analyti -\ncal approach for establishing causal relationships. This \nmethod, assuming the validity of all instrumental vari -\nables, calculates weighted estimates by taking the recip -\nrocal of their variances as weights. It provides the most \naccurate results when there is no heterogeneity or \nFig. 1 Flowchart of instrumental variable screening for MR method analysis\n \n\nPage 4 of 10\nDang et al. BMC Women's Health          (2024) 24:123 \nhorizontal pleiotropy present. Additionally, MR-Egger \nregression, the weighted median (WME) method, the \nsimple mode (SM) and the weighted mode (WM) method \nwere used as supplementary analyses to complement the \nIVW results. MR-Egger regression method performs \nweighted linear regression of the exposure and outcome \neffect estimates, providing a causal effect assessment \neven when all SNPs are invalid instruments. The WME \nmethod leverages the intermediate effects of all available \ngenetic variations, estimating them by weighting each \nSNP by the inverse variance of its correlation with the \noutcome. SM and WM are mode-based methods. The \nmode-based estimation model clusters SNPs with similar \ncausal effects and returns causal effect estimates for the \nmajority of clustered SNPs. Specifically, WM weights the \ninfluence of each SNP on the cluster by the inverse vari -\nance of its outcome effect. These methods complement \nthe IVW results and provide additional insights into the \ncausal relationships between exposure and outcome vari-\nables. Finally, we conducted reverse MR analysis for EMs \nand gut microbiota. The methods and settings used in \nthese reverse MR analysis were consistent with those of \nforward MR.\nSensitivity analysis\nHeterogeneity testing [ 22] assesses the presence of dif -\nferences among various IVs. It utilizes the P-value from \nCochran’s Q test to evaluate heterogeneity, with P > 0.05 \nindicating the absence of heterogeneity. If heterogeneity \nis detected, the MR pleiotropy residual sum and outlier \n(MR-PRESSO) test is employed to assess potential out -\nliers [ 23], eliminate them, and then reanalyze the data. \nMultiplicity testing [24] verifies the reliability of MR anal-\nysis results. MR-Egger intercept is used to detect hori -\nzontal pleiotropy, with P > 0.05 indicating the absence \nof horizontal pleiotropy and, thus, the reliability of the \nMR analysis results. Sensitivity testing [ 25] is conducted \nusing a “leave-one-out” approach, sequentially removing \neach SNP . If the MR results derived from the remain -\ning SNPs do not exhibit significant differences from the \noverall result, it demonstrates the robustness of the MR \nresults. Furthermore, the MR Steiger directional test was \nemployed to further assess the correlation between the \nexposure and the outcome.\nResults\nCausal effect of gut microbiota on EMs\nIn this study, 211 gut microbiota relative abundances \nwere selected as the exposure variable from gut microbi -\nota GWAS data involving 18,340 participants. These 211 \ntaxa include 9 phylums, 16 classes, 20 orders, 35 families, \nand 131 genuses. As both heterogeneity and pleiotropy \ntests yielded negative results, the IVW analysis results \nwere considered the primary reference indicator. The MR \nanalysis results indicate that seven different gut micro -\nbiota at various taxonomic levels (1 class, 1 order, 1 fam -\nily, and 4 genuses) may be associated with endometriosis, \nas shown in Fig.  2. The main MR analysis results for the \nassociation between all gut microbiota and the risk of \nEMs, as well as the results of heterogeneity and pleiot -\nropy tests, can be found in Supplementary File 2.\nWe identified associations between endometriosis and \nfive microbial taxonomic groups with positive correla -\ntions: family Prevotellaceae (OR = 1.19, 95%CI 1.02 ∼ 1.40, \nP = 0.026), genus Anaerotruncus  (OR = 1.25, 95%CI \n1.03 ∼ 1.53, P = 0.025), genus Olsenella  (OR = 1.11, 95%CI \n1.01 ∼ 1.22, P = 0.036), genus Oscillospira  (OR = 1.21, \n95%CI 1.01 ∼ 1.46, P = 0.035), order Bacillales (OR = 1.11, \n95%CI 1.00 ∼ 1.22, P = 0.042). Simultaneously, two micro-\nbial taxonomic groups showed negative associations with \nendometriosis: class Melainabacteria  (OR = 0.86, 95%CI \n0.75 ∼ 0.99, P = 0.036), genus Eubacterium ruminantium \ngroup (OR = 0.88, 95%CI 0.79 ∼ 0.98, P = 0.015) (Figs.  2, \n3 and 4). For detailed results of all SNPs related to these \nseven gut microbiota (including specific chromosomes, F \nvalues, and R2), please refer to Supplementary File 3.\nAs indicated in Supplementary File 3, we noted that \nthe contribution of total variation (R 2 values) for the 7 \ngut microbiota ranged from 0.13 to 0.21%, with F values \nspanning from 18.27 to 29.81. This range effectively rules \nout the possibility of weak genetic instrumental variables. \nHeterogeneity testing was conducted with a distribu -\ntion = 10,000 setting. The Cochran’s Q test for both IVW \nand MR-Egger regressions indicated the absence of het -\nerogeneity among the SNPs of each microbial taxonomic \ngroup. Multiple-effect tests revealed that the MR-Egger \nregression intercepts were all less than 0.05, and their \nP-values were greater than 0.05, suggesting the absence \nof horizontal pleiotropy. Furthermore, all MR Steiger \ndirectional tests consistently indicated that the direction \nfrom gut microbiota to endometriosis was robust for all \noutcomes (Table  1). Sensitivity analysis was performed \nusing a “leave-one-out” test, and a forest plot was gener -\nated. The results indicated that removing any single SNP \ndid not significantly influence the remaining SNP results, \nall remained on the same side of the null line. This sug -\ngests that the MR results in this study are robust. Refer to \nFig. 5 for visualization of the sensitivity analysis results.\nReverse-direction MR analyses\nFinally, a reverse mendelian randomization analysis was \nconducted, with endometriosis as the exposure factor \nand gut microbiota as the outcome variables. The results \nof each SNP of endometriosis and 7 gut microbiota are \nshown in Supplementary File 4. Heterogeneity and mul -\ntiple-effect tests yielded negative results. The IVW analy -\nsis revealed that there is no causal relationship between \nendometriosis and the seven different gut microbiota \n\nPage 5 of 10\nDang et al. BMC Women's Health          (2024) 24:123 \nat various taxonomic levels. The MR Steiger directional \ntests for the 7 gut microbiota with respect to endometri -\nosis yielded TRUE results. Detailed results can be found \nin Table 2.\nDiscussion\nMain findings and interpretation\nIn this study, we assessed for the first time the potential \nrelationship between gut microbiota and endometriosis \nby a bidirectional MR method, and identified the pres -\nence of specific microbial groups at the level of phy -\nlum, order, family, and genus that are closely related to \nEMs, family Prevotellaceae , genus Anaerotruncus , genus \nOlsenella, genus Oscillospira  and order Bacillales  had a \nrisk effect on endometriosis, and class Melainabacteria , \ngenus Eubacterium ruminantium group  was a protective \nfactor against endometriosis. Sensitivity analyses showed \nno horizontal pleiotropy, indicating that our MR analy -\nses were not affected by confounding factors, and “leave-\none-out” analyses confirmed the robustness of the study.\nDuring menstruation, when endometrial tissue retro -\ngrades into the peritoneal cavity and implants into sur -\nrounding tissues, such as the intestines or peritoneum, it \nleads to the formation of endometriotic lesions [ 26]. In \napproximately 10% of women, the immune system fails \nto clear these ectopic endometrial cells, leading to the \nactivation of macrophages, secretion of pro-inflamma -\ntory cytokines and growth factors, and the spread of the \nlesions [ 27, 28]. The gut microbiota is a crucial compo -\nnent of the human immune system, with immunomod -\nulatory functions mediated through interactions with \nstromal cells and epithelial cells. Research has shown that \nmicrobial metabolites act as messengers between the gut \nmicrobiota and immune functions [ 29– 31]. In studies \ninvolving mice with endometriosis, alterations in micro -\nbial metabolites were observed. The consumption of gut \nmicrobiota suppressed inflammation related to endome -\ntriosis [32] and influenced immune cell populations, sug -\ngesting that gut microbiota can influence endometriosis \nthrough immune pathways.\nFig. 2 Forrest plot for summary causal effects of gut microbiota on EMs risk based on IVW method for the primary analysis\n \n\nPage 6 of 10\nDang et al. BMC Women's Health          (2024) 24:123 \nThe abnormal endocrine microenvironment within \nEMs lesions is considered a key characteristic of endo -\nmetriosis. Estrogen [ 33] has a direct cell anti-apoptotic \nand proliferative effect on EMs lesions and promotes the \nformation of a pro-inflammatory microenvironment, \ncontributing to the chronic progression of the disease. \nEstrogen is a major regulatory factor for gut microbiota, \nand the gut microbiome’s genetic repertoire involved in \nFig. 4 Scatter plots of two taxa of gut microbiota negatively associated with EMs. (A) class Melainabacteria (B) genus Eubacterium ruminantium group\n \nFig. 3 Scatter plots of five taxa of gut microbiota positively associated with EMs. ( A) family Prevotellaceae (B) genus Anaerotruncus (C) genus Olsenella \n(D) genus Oscillospira (E)order Bacillales\n \n\nPage 7 of 10\nDang et al. BMC Women's Health          (2024) 24:123 \nestrogen metabolism is often referred to as the “estrobo -\nlome” [ 34]. It participates in estrogen regulation by \nsecreting beta-glucuronidase [ 35], forming the estro -\ngen-gut microbiota axis. Research has shown significant \ndifferences in the expression of 17β-estradiol, 16-keto-\n17β-estradiol, 2-hydroxyestrone, and 2-hydroxyestradiol \nin individuals with EMs. Additionally, there is a clear \npositive correlation between the gut microbiota of EMs \npatients and urinary estrogen levels [36]. Family Prevotel-\nlaceae belongs to the Bacteroidetes phylum, and a meta-\nanalysis [ 37] found that the abundance of Bacteroidetes \nis positively correlated with estrogen levels. When the \nFirmicutes/Bacteroidetes ratio in the gut decreases, there \nis an increase in the secretion of beta-glucuronidase in \nthe intestine, leading to elevated estrogen levels. High \nTable 1 Heterogeneity and pleiotropy evaluations for genetically causal associations of gut microbiota with EMs risk\nGut microbiota nSNP Cochran’s Q Pval MR-Egger MR Steiger\nIVW MR-Egger egger_intercept Pval Direction Pval\nclass Melainabacteria 10 10.645 0.329 −0.026 0.288 TRUE 1.17E−61\nfamily Prevotellaceae 16 15.496 0.346 0.002 0.933 TRUE 6.98E−56\ngenus Anaerotruncus 13 13.755 0.405 0.028 0.166 TRUE 6.22E−42\ngenus Eubacterium ruminantium group 18 12.733 0.692 < 0.001 0.983 TRUE 5.80E−61\ngenus Olsenella 10 7.374 0.524 0.011 0.629 TRUE 1.16E−33\ngenus Oscillospira 8 3.269 0.824 0.023 0.555 TRUE 1.22E−27\norder Bacillales 9 2.759 0.935 0.020 0.561 TRUE 3.45E−31\nTable 2 Results of reverse MR analysis of EMs on gut microbiota\nGut microbiota OR 95%CI Pval Cochran’s Q Pval Egger_Pval MR Steiger\nDirection Pval\nclass Melainabacteria 1.012866671 0.927–1.106 0.776483371 0.850 0.305 TRUE 3.15E-14\nfamily Prevotellaceae 1.038144984 0.982–1.098 0.18802718 0.452 0.596 TRUE 3.31E-11\ngenus Anaerotruncus 0.968896866 0.912–1.030 0.307702166 0.186 0.035 TRUE 3.29E-11\ngenus Eubacterium ruminantium group 1.041735583 0.962–1.128 0.312730065 0.398 0.620 TRUE 1.38E-11\ngenus Olsenella 1.101249839 0.987–1.229 0.084877138 0.564 0.766 TRUE 8.30E-12\ngenus Oscillospira 1.037769604 0.970–1.110 0.279033778 0.474 0.644 TRUE 3.48E-12\norder Bacillales 0.998659244 0.999−0.886 0.982427034 0.585 0.586 TRUE 2.52E-12\nFig. 5 Results of a leave-one-out analysis of the association of gut microbiota with EMs MR. (A) class Melainabacteria (B) family Prevotellaceae (C) genus \nAnaerotruncus (D) genus Eubacterium ruminantium group (E) genus Olsenella (F) genus Oscillospira (G) order Bacillales\n \n\nPage 8 of 10\nDang et al. BMC Women's Health          (2024) 24:123 \nestrogen levels are directly associated with the develop -\nment of EMs, and our study provides similar findings.\nMultiple studies have indicated [ 7, 33] that individu -\nals with endometriosis experience dysbiosis in their gut \nmicrobiota. The gut microbiota, when fermenting carbo -\nhydrates, produces short-chain fatty acids (SCFAs) that \ncan activate G protein-coupled receptors. This activation \nhas beneficial effects by reducing food intake, improv -\ning insulin sensitivity, inhibiting fat accumulation, and \nreducing systemic inflammation [ 38]. However, in cases \nof gut microbiota dysbiosis, there is a reduction in SCFA \nproduction. Simultaneously, certain neuroactive metab -\nolites, such as glutamate and butyric acid, increase in \nlevel. These metabolites can stimulate brain neurons and, \nthrough the hypothalamus-pituitary-ovary axis, increase \novarian estrogen secretion, exacerbating the condition of \npatients [39, 40].\nIt is noteworthy that PERROTTA et al. [ 41] estab -\nlished an EM classification model based on random for -\nest, revealing that the vaginal microbiota could predict \nthe severity of endometriomas (EMs), with Anaerococ-\ncus identified as the most crucial factor, while the gut \nmicrobiota lacked corresponding accuracy. Furthermore, \nCHEN et al. [42] built a model based on the female repro-\nductive tract microflora, which can distinguish whether \ninfertility is caused by EMs. Considering the potential \ninfluences on the gut microbiota from factors such as \ndiet, antimicrobial drugs, and psychological stress, rely -\ning on it as a tool for early diagnosis and screening of \nEMs is unreliable. Similarly, the reproductive tract micro-\nbiota can be affected by different physiological stages and \ndiseases like vaginal infections. Therefore, exploration of \nnon-invasive diagnostic methods for EMs is still needed, \nand using saliva for diagnosis may be more helpful [ 43]. \nHowever, what can be confirmed is the causal associa -\ntion between gut microbiota and endometriosis, with a \ndynamic interplay between the two, which holds poten -\ntial implications for future bacteria-based therapies.\nLimitation\nHowever, our study has several limitations: (1) Human \nbehavior is complex, and while understanding the genetic \nrisk of a disease can help prevent its occurrence to some \nextent, environmental factors also play a role in the \ndevelopment of the disease [ 44], and MR can only par -\ntially eliminate the interference of confounding factors \nsuch as the environment [ 45]. (2) The current study may \nnot comprehensively explore the entire spectrum of the \ngut microbiota, from phylum to genus level, potentially \nmissing other microbial taxa that could have a causal \nrelationship with endometriosis, especially those associ -\nated with increased risk. (3) The outcome data used in \nthe study is derived from European populations, and cau-\ntion should be exercised when extrapolating the results \nto other populations with different lifestyles, cultural \nbackgrounds, and genetic backgrounds, as specific traits \nmay vary across different racial and ethnic groups driven \nby their distinct living environments and genetic back -\ngrounds. Efforts should be made to include populations \nof all ethnicities globally in genetic studies of this nature. \n(4) Although we have demonstrated a causal relationship \nbetween gut microbiota and endometriosis, the under -\nlying mechanism is still unclear and requires further \nresearch.\nConclusions\nThe study collected data from GWAS databases and used \na two-sample bidirectional MR approach to confirm the \npotential causal relationship between gut microbiota and \nendometriosis, providing new insights into the patho -\ngenesis and treatment of endometriosis. Future research \nshould aim to further elucidate the underlying mecha -\nnisms by which these microbial communities influence \nendometriosis, explore potential treatment strategies tar-\ngeting gut microbiota.\nAbbreviations\nMR  Mendelian randomization\nEMs  Endometriosis\nRCT  Randomized controlled trials\nSNPs  Single nucleotide polymorphisms\nIVs  Instrumental variables\nGWAS  Genome wide association study\nLD  Linkage disequilibrium\nIVW  Inverse variance weighted\nWM  Weighted median\nSupplementary Information\nThe online version contains supplementary material available at https://doi.\norg/10.1186/s12905-024-02945-z.\nSupplementary Material 1\nSupplementary Material 2\nSupplementary Material 3\nSupplementary Material 4\nAcknowledgements\nThis work benefited from the publicly available statistics of GWAS. We thank \nthe contributors to the original GWAS database.\nAuthor contributions\nPengfei Liu and Jinxing Liu conceived the study. Chunxiao Dang and Yuyan \nChai provided the design of the study. Zhenting Chen and Pengfei Liu \ncollected the data. Xiao Yu and Yan Liu conducted the main analyses of the \nstudy. Chunxiao Dang and Zhenting Chen wrote the body of the manuscript. \nYan Liu and Jinxing Liu revised the manuscript. All authors reviewed the the \nmanuscript.\nFunding\nThis work was supported by the Natural Science Foundation of China \n(No.82104917), the Natural Science Foundation of Shandong Province (No.\nZR2021MH079).\n\nPage 9 of 10\nDang et al. BMC Women's Health          (2024) 24:123 \nData availability\nAll data generated or analysed during this study are included in this published \narticle and its supplementary information files.\nDeclarations\nEthics approval and consent to participate\nNot applicable.\nConsent for publication\nNot applicable.\nCompeting interests\nThe authors declare no competing interests.\nAuthor details\n1First Clinical Medical College, Shandong University of Traditional Chinese \nMedicine, Jinan 250355, Shandong, China\n2Department of eugenic genetics, Dongying People’s Hospital (Dongying \nHospital of Shandong Provincial Hospital Group), Dongying 257091, \nShandong, China\n3Department of obstetrics, The People’s Hospital of Dongying Distric, \nDongying 257091, Shandong, China\n4Department of gynaecology, Affiliated Hospital of Shandong University \nof Traditional Chinese Medicine, Jinan 250000, Shandong, China\n5National Key Laboratory for Innovation and Transformation of Luobing \nTheory, The Key Laboratory of Cardiovascular Remodeling and Function \nResearch, Department of Cardiology, Chinese Ministry of Education, \nChinese National Health Commission and Chinese Academy of Medical \nSciences, Qilu Hospital of Shandong University, Jinan 250000, Shandong, \nChina\nReceived: 8 October 2023 / Accepted: 1 February 2024\nReferences\n1. Filby CE, Rombauts L, Montgomery GW, Giudice LC, Gargett CE. Cellular \norigins of endometriosis: towards Novel Diagnostics and therapeutics. Semin \nReprod Med. 2020;38:201–15.\n2. Engemise S, Gordon C, Konje JC, Endometriosis. BMJ. 2010;340:c2168.\n3. La Rosa VL, Barra F, Chiofalo B, Platania A, Di Guardo F, Conway F, et al. An \noverview on the relationship between endometriosis and infertility: the \nimpact on sexuality and psychological well-being. J Psychosom Obstet \nGynaecol. 2020;41:93–7.\n4. Barandouzi ZA, Starkweather AR, Henderson WA, Gyamfi A, Cong XS. Altered \ncomposition of Gut Microbiota in Depression: a systematic review. Front \nPsychiatry. 2020;11:541.\n5. Marques FZ, Jama HA, Tsyganov K, Gill PA, Rhys-Jones D, Muralitharan RR, et \nal. Guidelines for transparency on gut Microbiome studies in essential and \nexperimental hypertension. Hypertension. 2019;74:1279–93.\n6. Woodall CA, McGeoch LJ, Hay AD, Hammond A. Respiratory tract infec-\ntions and gut microbiome modifications: a systematic review. PLoS ONE. \n2022;17:e0262057.\n7. Bauşic AIG, Creţoiu SM, Bauşic V, Matasariu DR, Stănculescu RV, Brătilă E. The \nrole of gut dysbiosis in endometriosis’ diagnosis and treatment approaches - \ncase report. Rom J Morphol Embryol. 2023;64:263–9.\n8. Iavarone I, Greco PF, La Verde M, Morlando M, Torella M, de Franciscis P , et al. \nCorrelations between Gut Microbial Composition, Pathophysiological and \nSurgical aspects in Endometriosis: a review of the literature. Med (Kaunas). \n2023;59:347.\n9. Benagiano G, Brosens I, Lippi D. Endometriosis: ancient or modern disease? \nIndian J Med Res. 2015;141:236–8.\n10. Chiaffarino F, Cipriani S, Ricci E, Roncella E, Mauri PA, Parazzini F, et al. \nEndometriosis and inflammatory bowel disease: a systematic review of the \nliterature. Eur J Obstet Gynecol Reprod Biol. 2020;252:246–51.\n11. Forbes JD, Chen CY, Knox NC, Marrie RA, El-Gabalawy H, de Kievit T, et al. A \ncomparative study of the gut microbiota in immune-mediated inflammatory \ndiseases-does a common dysbiosis exist? Microbiome. 2018;6:221.\n12. West CE, Renz H, Jenmalm MC, Kozyrskyj AL, Allen KJ, Vuillermin P , et al. The \ngut microbiota and inflammatory noncommunicable diseases: associa-\ntions and potentials for gut microbiota therapies. J Allergy Clin Immunol. \n2015;135:3–13.\n13. Hariton E, Locascio JJ. Randomised controlled trials - the gold standard for \neffectiveness research: study design: randomised controlled trials. BJOG. \n2018;125:1716.\n14. Carnegie R, Zheng J, Sallis HM, Jones HJ, Wade KH, Evans J, et al. Mendelian \nrandomisation for nutritional psychiatry. Lancet Psychiatry. 2020;7:208–16.\n15. Lawlor DA, Harbord RM, Sterne JA, Timpson N, Davey Smith G. Mendelian \nrandomization: using genes as instruments for making causal inferences in \nepidemiology. Stat Med. 2008;27:1133–63.\n16. Kurilshikov A, Medina-Gomez C, Bacigalupe R, Radjabzadeh D, Wang J, Demir-\nkan A, et al. Large-scale association analyses identify host factors influencing \nhuman gut microbiome composition. Nat Genet. 2021;53:156–65.\n17. Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, et al. The \nMR-Base platform supports systematic causal inference across the human \nphenome. Elife. 2018;7:e34408.\n18. Slatkin M. Linkage disequilibrium–understanding the evolutionary past and \nmapping the medical future. Nat Rev Genet. 2008;9:477–85.\n19. Zhuang Z, Yu C, Guo Y, Bian Z, Yang L, Millwood IY, et al. Metabolic Signatures \nof Genetically Elevated Vitamin D among Chinese: observational and mende-\nlian randomization study. J Clin Endocrinol Metab. 2021;106:e3249–60.\n20. Hartwig FP , Davies NM, Hemani G, Davey Smith G. Two-sample mendelian \nrandomization: avoiding the downsides of a powerful, widely applicable but \npotentially fallible technique. Int J Epidemiol. 2016;45:1717–26.\n21. Wu F, Huang Y, Hu J, Shao Z. Mendelian randomization study of inflammatory \nbowel disease and bone mineral density. BMC Med. 2020;18:312.\n22. Bowden J, Spiller W, Del Greco MF, Sheehan N, Thompson J, Minelli C, et al. \nImproving the visualization, interpretation and analysis of two-sample sum-\nmary data mendelian randomization via the Radial plot and radial regression. \nInt J Epidemiol. 2018;47:2100.\n23. Verbanck M, Chen CY, Neale B, Do R. Publisher correction: detection of \nwidespread horizontal pleiotropy in causal relationships inferred from \nmendelian randomization between complex traits and diseases. Nat Genet. \n2018;50:1196.\n24. Carter AR, Sanderson E, Hammerton G, Richmond RC, Davey Smith G, Heron \nJ, et al. Mendelian randomisation for mediation analysis: current methods \nand challenges for implementation. Eur J Epidemiol. 2021;36:465–78.\n25. Gronau QF, Wagenmakers EJ. Limitations of bayesian leave-one-out Cross-\nvalidation for Model Selection. Comput Brain Behav. 2019;2:1–11.\n26. Giudice LC. Clinical practice. Endometriosis. N Engl J Med. 2010;362:2389–98.\n27. Ahn SH, Monsanto SP , Miller C, Singh SS, Thomas R, Tayade C. Patho-\nphysiology and Immune Dysfunction in Endometriosis. Biomed Res Int. \n2015;2015:795976.\n28. Han SJ, O’Malley BW. The dynamics of nuclear receptors and nuclear receptor \ncoregulators in the pathogenesis of endometriosis. Hum Reprod Update. \n2014;20:467–84.\n29. Arpaia N, Rudensky AY. Microbial metabolites control gut inflammatory \nresponses. Proc Natl Acad Sci U S A. 2014;111:2058–9.\n30. Marsland BJ. Regulating inflammation with microbial metabolites. Nat Med. \n2016;22:581–3.\n31. Chang PV, Hao L, Offermanns S, Medzhitov R. The microbial metabolite \nbutyrate regulates intestinal macrophage function via histone deacetylase \ninhibition. Proc Natl Acad Sci U S A. 2014;111:2247–52.\n32. Chadchan SB, Cheng M, Parnell LA, Yin Y, Schriefer A, Mysorekar IU, et al. \nAntibiotic therapy with metronidazole reduces endometriosis disease \nprogression in mice: a potential role for gut microbiota. Hum Reprod. \n2019;34:1106–16.\n33. Shan J, Ni Z, Cheng W, Zhou L, Zhai D, Sun S, et al. Gut microbiota imbalance \nand its correlations with hormone and inflammatory factors in patients with \nstage 3/4 endometriosis. Arch Gynecol Obstet. 2021;304:1363–73.\n34. Qi X, Yun C, Pang Y, Qiao J. The impact of the gut microbiota on the reproduc-\ntive and metabolic endocrine system. Gut Microbes. 2021;13:1–21.\n35. Ervin SM, Li H, Lim L, Roberts LR, Liang X, Mani S, et al. Gut microbial \nβ-glucuronidases reactivate estrogens as components of the estrobolome \nthat reactivate estrogens. J Biol Chem. 2019;294:18586–99.\n36. Le N, Cregger M, Brown V, Loret de Mola J, Bremer P , Nguyen L, et al. Associa-\ntion of microbial dynamics with urinary estrogens and estrogen metabolites \nin patients with endometriosis. PLoS ONE. 2021;16:e0261362.\n\nPage 10 of 10\nDang et al. BMC Women's Health          (2024) 24:123 \n37. d’Afflitto M, Upadhyaya A, Green A, Peiris M. Association between sex \nhormone levels and gut microbiota composition and Diversity-A systematic \nreview. J Clin Gastroenterol. 2022;56:384–92.\n38. Cong J, Zhou P , Zhang R. Intestinal microbiota-derived short chain fatty acids \nin host health and disease. Nutrients. 2022;14:1977.\n39. Mikhael S, Punjala-Patel A, Gavrilova-Jordan L. Hypothalamic-pituitary-ovarian \nAxis disorders Impacting Female Fertility. Biomedicines. 2019;7:5.\n40. Baj A, Moro E, Bistoletti M, Orlandi V, Crema F, Giaroni C. Glutamatergic signal-\ning along the Microbiota-Gut-Brain Axis. Int J Mol Sci. 2019;20:1482.\n41. Perrotta AR, Borrelli GM, Martins CO, Kallas EG, Sanabani SS, Griffith LG, et \nal. The vaginal microbiome as a Tool to predict rASRM Stage of Disease in \nEndometriosis: a pilot study. Reprod Sci. 2020;27:1064–73.\n42. Chen C, Song X, Wei W, Zhong H, Dai J, Lan Z, et al. The microbiota con-\ntinuum along the female reproductive tract and its relation to uterine-related \ndiseases. Nat Commun. 2017;8:875.\n43. Bendifallah S, Dabi Y, Suisse S, Delbos L, Spiers A, Poilblanc M, et al. Validation \nof a salivary miRNA signature of endometriosis-interim data. NEJM Evid. \n2023;2:EVIDoa2200282.\n44. Meisel SF, Beeken RJ, van Jaarsveld CH, Wardle J. Genetic susceptibility testing \nand readiness to control weight: results from a randomized controlled trial. \nObes (Silver Spring). 2015;23:305–12.\n45. Agustí A, Melén E, DeMeo DL, Breyer-Kohansal R, Faner R. Pathogenesis of \nchronic obstructive pulmonary disease: understanding the contributions \nof gene-environment interactions across the lifespan. Lancet Respir Med. \n2022;10:512–24.\nPublisher’s Note\nSpringer Nature remains neutral with regard to jurisdictional claims in \npublished maps and institutional affiliations.","source_license":"CC0","license_restricted":false}