Mapping the causal landscape of polycystic ovary syndrome

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Mendelian randomization (MR) studies have increasingly investigated potential predicted antecedents and consequences of polycystic ovary syndrome (PCOS), yet published evidence remains scattered across diverse exposures, outcomes, and analytical frameworks. This study aimed to systematically organize published PCOS-related MR evidence and develop an interactive platform to support evidence retrieval, comparison, and interpretation. We searched PubMed, Web of Science, Embase, and the Cochrane Library from inception to June 1, 2026. Original MR Studies were eligible if PCOS was either the exposure or the outcome, with extractable effect estimates. Each causal pathway was extracted separately for direction, definitions, effect estimates, primary MR method, Genome-Wide Association Study sources, ancestry, sample size, PCOS phenotype, sensitivity analyses, and sample overlap. Methodological appraisal followed Strengthening the Reporting of Observational Studies in Epidemiology-Mendelian Randomization. Evidence was summarized by direction, variable category, quality, and consistency. An interactive MR-PCOS Atlas was developed using R Shiny. Of 728 records identified, 306 original MR studies were included, yielding 1081 extractable causal pathways (763 with PCOS as outcome, 318 as exposure). Evidence concentrated in physiological traits and diseases/clinical phenotypes (312 and 299 pathways, respectively). Methodological quality was high in 13 studies, moderate in 279, and low in 14. Evidence most frequently addressed metabolic, endocrine, reproductive, and comorbidity pathways, while lifestyle, gut microbiota, immune phenotypes, and interventions remained limited. The MR-PCOS Atlas enables interactive browsing and comparison of published findings. This systematic synthesis provides a structured map of genetically supported PCOS associations and highlights areas of consistency, uncertainty, and methodological heterogeneity. Findings should be interpreted considering PCOS definitions, Genome-Wide Association Study sources, ancestry, and MR assumptions. The MR-PCOS Atlas serves as an evidence resource for retrieval and hypothesis generation, not for direct clinical recommendations.