Thyroid autoimmunity, thyroid function, and endometriosis: reproductive immune-endocrine crosstalk and causal uncertainty.

Background: Endometriosis is an estrogen-dependent inflammatory disorder associated with pelvic pain, infertility, repeated treatment, and impaired quality of life. Thyroid dysfunction and thyroid autoimmunity are also common in reproductive-aged women and may affect menstrual cyclicity, ovarian function, implantation, pregnancy maintenance, and assisted reproduction. Observational studies suggest an association between thyroid-related phenotypes and endometriosis, but its directionality, reproductive relevance, and biological basis remain uncertain. Main body: This narrative review synthesizes epidemiological, mechanistic, reproductive, environmental, genetic, and causal-inference evidence on the thyroid-endometriosis interface. It distinguishes biochemical thyroid function traits, thyroid dysfunction, thyroid autoimmunity, and diagnosis-based thyroid disease as separate but related constructs. Current evidence suggests that thyroid autoimmunity may be more relevant to endometriosis than isolated thyroid hormone abnormalities. Cohort, register-based, case-control, and infertility-population studies have reported associations with subsequent thyroid morbidity, thyroid autoantibody positivity, and assisted reproduction outcomes, although findings remain heterogeneous across populations, study designs, and phenotype definitions. Mechanistic evidence is best interpreted at three levels: established endometriosis biology, indirect thyroid-related reproductive plausibility, and shared upstream susceptibility involving endocrine-disrupting chemicals and genetic or epigenetic vulnerability. Direct thyroid-to-lesion evidence remains limited. Most available evidence is observational. A recent bidirectional Mendelian randomization study reported several phenotype-specific associations, but the estimates require independent replication and do not establish a uniform thyroid-to-endometriosis causal relationship. Conclusion: The thyroid-endometriosis association is better viewed as a phenotype-specific reproductive immune-endocrine interface than as a simple thyroid hormone-driven causal pathway. Current evidence does not justify universal thyroid screening solely because endometriosis is present. Thyroid evaluation should instead follow established endocrine, reproductive, pregnancy-related, or autoimmune indications, including infertility, recurrent pregnancy loss, menstrual disturbance, planned assisted reproduction, symptoms of thyroid dysfunction, previous abnormal thyroid tests, or coexisting autoimmune disease. Future studies should combine standardized phenotyping, prospective follow-up, mechanistic biomarkers, and causal-inference methods to determine whether thyroid-related pathways are clinically useful markers, modifiable contributors, or parallel manifestations of systemic dysregulation in endometriosis.