Endocrine Abstracts

Endometriosis is a common gynaecological condition caused by the abnormal growth of endometrial-like tissue outside the uterus. There are currently no biomarkers validated for non-invasive diagnosis and the exact aetiology is unknown; however, the association between endometriosis and thyroid dysfunction has recently been reported. We aimed to investigate the value of thyroid hormone testing in predicting the odds of an endometriosis diagnosis. Retrospective data was taken from normo-ovulatory women who undertook a Hertility Health at-home blood test to measure free thyroxine (FT4) and thyroid-stimulating hormone (TSH) among other reproductive hormones. Via a virtual questionnaire, cases (n = 17) had a self-reported diagnosis of endometriosis whilst controls (n = 208) confirmed that they had not been diagnosed. Exclusion criteria for both cohorts included diagnoses of polycystic ovary syndrome, primary ovarian insufficiency, fibroids, pelvic inflammatory disease and fallopian tube blockage. Capillary blood sampling was performed on day 3 of the menstrual cycle; serum FT4 and TSH concentrations were measured using enzyme immunoassays. Multivariate logistic regression and ROC curves were used to assess whether FT4 or TSH concentrations were associated with an endometriosis diagnosis. Thyroid hormone measurements were log-transformed prior to analysis. P values, adjusted odds ratios [OR] and areas under ROC curves [AUC] have been reported. A multivariate logistic regression model which included TSH concentration, age, exercise frequency, recreational drug use, smoking status and BMI was found to be statistically significant (P = 0.0016, OR [1.80], AUC [0.94]). Similarly, statistical significance was found when FT4 was incorporated in a logistic regression model with the same demographic covariates (P = 0.0089, OR [0.01], AUC [0.89]). These data suggest that TSH or FT4 concentrations in combination with the demographic variables described above may have some predictive value in assessing the likelihood of an endometriosis diagnosis in normo-ovulatory women. However, more data is required to evaluate the accuracy and clinical impact of this model.