Genetic liability to endometriosis and pregnancy outcomes: a two-sample Mendelian randomization study with maternal–fetal effect decomposition

Background Endometriosis is associated with adverse pregnancy outcomes in standard observational studies, including placental complications, preterm birth, and caesarean delivery. However, causal inference from these studies is complicated by residual confounding, differential clinical management, and the presence of intermediate factors such as subfertility and the use of assisted reproductive technologies, which may lie on the causal pathway between endometriosis and adverse outcomes.We applied Mendelian randomization (MR) to estimate the causal effects of genetic liability to endometriosis on a broad range of maternal and perinatal outcomes.

We conducted a two-sample MR study using summary-level GWAS data. Forty-one independent genetic instruments for endometriosis were derived from the largest available GWAS meta-analysis (60,674 cases; 701,926 controls; mean F-statistic = 279). SNP–outcome associations were obtained for 30 outcomes from the MR-PREG collaboration, FinnGen Release 12, and a postpartum haemorrhage GWAS meta-analysis, spanning placental disorders, pregnancy timing, labour and delivery, hypertensive disorders, fetal growth, and neonatal outcomes. Primary analyses used the inverse-variance weighted method, complemented by MR-Egger, weighted median, weighted mode, and MR-PRESSO. Trio-based models disentangled maternal from fetal genetic contributions. Multiple testing was addressed using false discovery rate correction. Findings Across 30 outcomes, only placenta praevia reached FDR-corrected significance, with a robust and consistent causal signal across four of five sensitivity methods (IVW OR 1·62, 95% CI 1·33–1·97; q<0·001). Within the placental disorders domain, estimates for premature placental separation and the broader placental disorders phenotype were directionally concordant but imprecise. For premature rupture of membranes, estimates were concordant across three methods, though the association was sensitive to cohort exclusion and did not survive multiple testing correction and should be interpreted cautiously. By contrast, hypertensive disorders, gestational diabetes, postpartum haemorrhage, stillbirth, and most neonatal outcomes showed estimates consistently close to the null across all methods. Trio-based analyses suggested predominantly maternal genetic pathways for most outcomes; fetal genetic contributions were not significant after correction for multiple testing, with exploratory signals observed for birthweight-related outcomes requiring independent replication. Interpretation A robust causal signal for placenta praevia alongside directionally consistent estimates across the placental disorders domain, suggests that mechanisms related to abnormal implantation and placentation may constitute a major mechanism for how endometriosis liability influences pregnancy. These results suggest that previously reported associations with broader obstetric outcomes may partly reflect confounding or clinical management patterns, and support targeted surveillance for abnormal placentation rather than a generalised elevation of obstetric risk. Funding TSC, MCB, EA, and DAL are members of the MRC Integrative Epidemiology Unit at the University of Bristol (MC_UU_00032/5). MCM is supported by the Research Council of Norway through its Centres of Excellence funding scheme (project No 262700); and the research project “Endometriosis and adenomyosis throughout the life-course” (project No 351058). Evidence before this study We searched PubMed and MEDLINE from inception to March 2025 using combinations of the terms “endometriosis” and key pregnancy outcomes (including placenta praevia, preterm birth, postpartum haemorrhage, caesarean delivery, fetal growth, and pregnancy complications), without language restrictions. We included observational studies, systematic reviews, meta-analyses, and Mendelian randomisation studies. Observational evidence consistently suggests that women with endometriosis have increased risks of placenta praevia, preterm birth, caesarean delivery, and impaired fetal growth, with reported effect sizes for placenta praevia typically ranging from two- to four-fold. However, findings for other outcomes, including postpartum haemorrhage and hypertensive disorders, are inconsistent. Taken together, these findings suggest that endometriosis may primarily affect early implantation and placentation processes. Rather than supporting a general intensification of obstetric surveillance, our results highlight the need to better characterise the underlying biological mechanisms of abnormal placental implantation, with the aim of identifying targets for prevention. Mendelian randomisation studies have attempted to address these limitations, but results remain inconsistent. Some studies report little evidence for causal effects on pregnancy outcomes, while others suggest associations restricted to severe disease or specific outcomes. These discrepancies likely reflect differences in statistical power—particularly for rare outcomes—and outcome definitions. Importantly, no previous study has jointly examined a broad spectrum of pregnancy outcomes while disentangling maternal, fetal, and paternal genetic effects. Added value of this study Using a two-sample Mendelian randomisation design across 30 pregnancy and perinatal outcomes, we provide robust genetic evidence that liability to endometriosis causally increases the risk of placenta praevia. This association was consistent across multiple sensitivity analyses and supported by the absence of pleiotropy and heterogeneity. Across the broader placental disorders domain, estimates were directionally concordant, suggesting that the causal signal extends beyond a single phenotype and reflects a shared underlying mechanism related to implantation and placentation. In contrast, we found little evidence supporting a causal role of endometriosis liability in hypertensive disorders, gestational diabetes, postpartum haemorrhage, fetal growth restriction, or most neonatal outcomes. Associations observed for preterm birth and caesarean delivery were not robust to sensitivity analyses and appeared driven by pleiotropy or cohort-specific effects. By incorporating trio-based analyses, we further show that associations are predominantly driven by maternal genetic pathways, supporting a uterine mechanism rather than fetal genetic effects. Implications of all the available evidence Taken together, the available evidence supports a more targeted approach to pregnancy care in women with endometriosis. The robust and specific association with placenta praevia—together with directionally consistent findings across the placental disorders domain—supports targeted surveillance of placental location rather than a generalised increase in obstetric monitoring. The absence of consistent causal effects for most other outcomes suggests that previously reported associations in observational studies may largely reflect residual confounding, differences in the population of women who achieve pregnancy (for example due to subfertility or the use of assisted reproductive technologies), and variations in clinical management (such as increased surveillance or intervention rates), rather than a direct biological effect of endometriosis itself. From a mechanistic perspective, these findings point to impaired implantation and early placentation as the principal pathway linking endometriosis to adverse pregnancy outcomes. Future research should focus on the biology of uterine receptivity and placentation, including the role of disease severity and co-existing adenomyosis, to identify opportunities for early intervention. Competing Interest Statement MCB and DAL received funding from Novartis for unrelated research. JV, TSC, MCM, LA, ZK, DB, and NP declare no competing interests. Funding Statement This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. TSC, MCB, LA, DAL are members of the MRC Integrative Epidemiology Unit at the University of Bristol funded by the UK Medical Research Council (Grant ref: MC_UU_00032/5). MCM is supported by the Research Council of Norway through its Centres of Excellence funding scheme (project No 262700); and the research project "Endometriosis and adenomyosis throughout the life-course" (project No 351058). All cohort-specific funding information is detailed in the online supplemental file. Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: Does the study describe the use of any human data, samples, or any research involving human subjects? Answer 'YES' if you have a decision from your Ethics Committee or Institutional Review Board (IRB) or if you used (or will use) human datasets Important! Failure to disclose the use of human data, samples or involvement of human subjects (see below) will significantly delay the processing of your submission The use of human data includes but is not limited to: - Studies of publicly available human data - Surveys and interview-based studies - Any analyses of patient records (including electronic health records) - Any research involving living or deceased human subjects or samples (even discarded or anonymized) - Clinical research design protocols (even those not yet initiated) Yes No Did (or will) this work use ONLY simulated data or ONLY data that were openly available to the public before the initiation of the study? Answer 'NO' if you have a decision from your Ethics Committee or Institutional Review Board (IRB) or if you used datasets or databases that were obtained after a request, or required application, screening or registration for access. Yes Example text: The study used (or will use) ONLY openly available human data that were originally located at: The study used summary-level human genetic data from publicly available sources, including the GWAS Catalog (https://www.ebi.ac.uk/gwas/ ) and FinnGen (https://www.finngen.fi/en/access_results ), as well as data obtained through established research collaborations (MR-PREG). Detailed access information for all datasets is provided in the manuscript. [Accented Characters] Important! Processing of your submission will be significantly delayed if you do not provide complete details about the source of the data (with links, if applicable). Links should also be included in the Data Availability Links section below. I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes Data Availability Analytical R code is available on GitHub at: https://github.com/jonasvibert/endoMR-PREG Genetic association data for adverse pregnancy outcomes generated by the MR-PREG collaboration can only be used for research that is covered by data agreements with current contributing studies. The ALSPAC access policy that describes the proposal process in detail, including any costs associated with conducting research at ALSPAC, which may be updated from time to time, is available at: https://www.bristol.ac.uk/medialibrary/sites/alspac/documents/researchers/dataaccess/ALSPAC_Access_Policy.pdf Data are available upon request from BiB, and information is available at: https://borninbradford.nhs.uk/research/how-to-access-data/ Data from MoBa are available upon application to Helsedata, administered by the Norwegian Institute of Public Health (see: https://helsedata.no for details). Researchers can apply for access to UK Biobank data via the Access Management System (AMS): https://www.ukbiobank.ac.uk/enable-your-research/apply-for-access Summary statistics from the endometriosis GWAS meta-analysis (Rahmioglu et al., 2023) excluding 23andMe are available from the EBI GWAS Catalog (Study Accession GCST90205183). GWAS summary statistics from 23andMe, Inc. were made available under a data use agreement that protects participant privacy; please contact dataset-request{at}23andme.com or visit research.23andMe.com/collaborate for more information and to apply to access the data. Meta-analysis summary statistics for PPH (Westergaard et al., 2024), including all subtypes, early bleeding, and antepartum haemorrhage, are deposited at: https://www.decode.com/summarydata FinnGen Release 12 summary statistics are publicly available; full details are provided at: https://finngen.gitbook.io/documentation/