Childhood trauma, interpersonal violence, and their association with endometriosis: systematic review and meta-analysis

Observational studies link trauma-related exposures to endometriosis, but findings vary across exposures and diagnostic criteria. We synthesized associations and examined differences by diagnostic method, exposure definition, trauma subtype, and timing. We searched PubMed, Embase, Web of Science, and Cochrane Library from inception to April 27, 2026 (updated October 2, 2026) for observational studies reporting relative effect estimates. One estimate per study, selected by a predefined hierarchy, was log-transformed and pooled using DerSimonian-Laird random effects. Hartung-Knapp-Sidik-Jonkman confidence intervals and a 95% prediction interval were calculated. Prespecified subgroups examined diagnosis and exposure definition; sensitivity analyses examined effect measure, design, outcome ascertainment, quality, and alternative estimate selection. Exploratory trauma-subtype and timing analyses were descriptive. Quality was assessed using the Newcastle-Ottawa Scale and domain-based bias appraisal. Ten studies involving over 1.5 million women were included; six contributed to meta-analysis. Trauma-related exposures were associated with higher relative occurrence of endometriosis (pooled estimate 1.32, 95% CI 1.18–1.48; Hartung–Knapp 95% CI 1.12–1.56; I²=58.1%), with a 95% prediction interval of 0.96–1.82. Sensitivity estimates ranged from 1.14 to 1.42, all above 1. In an exploratory, underpowered comparison by design without a formal between-group test, the pooled estimate was numerically lower in four cross-sectional, case–control or surgical-cohort analyses (1.24, 95% CI 1.03–1.49) than in two prospective cohorts (1.42, 95% CI 1.19–1.69). Prespecified subgroup estimates were numerically higher for registry-, record-, or self-report-based diagnoses than surgical confirmation, and composite trauma measures than sexual trauma alone; differences were nonsignificant, with three studies per subgroup. Exploratory subtype and timing analyses used two to five estimates each and are hypothesis-generating. Observational evidence suggests trauma-related exposures are associated with modestly higher relative occurrence of an endometriosis diagnosis. The small evidence base, heterogeneous exposure and outcome definitions, reliance on retrospective self-report, confounding, and diagnostic bias preclude etiological interpretation. Prospective studies with standardized trauma assessment and consistent diagnostic criteria are needed before drawing conclusions about causal pathways or routine trauma assessment in endometriosis care. Not applicable. PROSPERO CRD420261379461.