Is polycystic ovary syndrome associated with uterine malformations? A systematic review using Bradford Hill’s causality framework

STUDY QUESTION: Is there a causal relationship between polycystic ovary syndrome (PCOS) and the occurrence of congenital uterine anomalies (CUAs)? SUMMARY ANSWER: Bradford Hill criteria did not support the causal relationship between PCOS and CUAs. WHAT IS KNOWN ALREADY: PCOS and CUAs are both linked to infertility and complications during pregnancy, but it is unclear whether PCOS increases the risk of developing CUAs. STUDY DESIGN SIZE DURATION: A systematic review with qualitative analysis using Bradford Hill criteria was performed. PARTICIPANTS/MATERIALS SETTING METHODS: Studies evaluating CUAs in women with PCOS or mechanisms linking PCOS-related factors to Müllerian development were selected. Comprehensive searches were conducted in MEDLINE, Web of Science, Cochrane Library, and Scopus for studies published in English from 1 January 2000 up to 30 August 2025 using terms related to PCOS, CUAs, and Müllerian anomalies. The nine Bradford Hill criteria were applied using modern epidemiological tools and concepts. For each Bradford Hill criterion, we prioritized studies according to the highest available level of evidence as defined by the Oxford Centre for Evidence-Based Medicine (CEBM). Study quality and risk of bias were systematically assessed using validated tools appropriate to each study design. MAIN RESULTS AND THE ROLE OF CHANCE: Twenty-one studies were included (5 systematic reviews with meta-analysis, 4 case-control, 11 cohort, 1 cross-sectional). Evidence supported the Bradford Hill criteria of biological gradient, coherence, plausibility, and analogy, whereas consistency, specificity, temporality, and experimental evidence were not met, and the strength of association only partially fulfilling. Overall, the findings provide weak qualitative support for a potential causal relationship between PCOS and CUAs. LIMITATIONS REASONS FOR CAUTION: Findings are limited by heterogeneity in diagnostic criteria and imaging methods, and potential selection bias. In addition, the strength of association is characterized by large coefficient interval and modest significance. WIDER IMPLICATIONS OF THE FINDINGS: Routine uterine imaging for identifying/screening CUAs is not justified by the PCOS diagnosis alone until standardized prospective studies demonstrate a causal inference between diseases. STUDY FUNDING/COMPETING INTERESTS: Departmental funds (Departments of Medical-Surgical Sciences and Translational Medicine, Sapienza University of Rome, and of Experimental and Clinical Medicine of 'Magna Graecia', University of Catanzaro; Italy) were used to support the authors throughout the study period and manuscript preparation. No specific external funding was received. The authors declare no competing interests. REGISTRATION NUMBER: PROSPERO CRD420251049920.