Subclinical thyroid dysfunction and clinical severity in perimenopausal abnormal uterine bleeding: a retrospective cohort study

Aims: To examine the association between thyroid dysfunction and bleeding severity, anemia burden, and management outcomes in perimenopausal women with abnormal uterine bleeding (AUB).Methods: In this retrospective cohort study, 742 perimenopausal women aged 40-55 years with AUB were evaluated between January 2014 and December 2024. Thyroid status was categorized as euthyroid, subclinical hypothyroidism, overt hypothyroidism, subclinical hyperthyroidism, or overt hyperthyroidism based on serum thyroid-stimulating hormone and free thyroxine levels. Severe AUB, anemia, and management outcomes (iron therapy, blood transfusion, and procedural interventions) were analyzed using multivariable logistic regression adjusted for demographic and metabolic covariates. Receiver operating characteristic analyses were used to explore the standalone discriminatory performance of thyroid-stimulating hormone for severe AUB and moderate-to-severe anemia.Results: Thyroid dysfunction was identified in 29.4% of participants. Severe AUB and moderate-to-severe anemia were more frequent among women with subclinical and overt hypothyroidism. Thyroid dysfunction was independently associated with severe AUB (aOR 2.21; 95% CI 1.54-3.18), anemia (aOR 1.89; 95% CI 1.33-2.68), and escalation of care (aOR 1.97; 95% CI 1.28-3.02), as well as increased likelihood of iron therapy and blood transfusion. A thyroid-stimulating hormone threshold near 4.0 mIU/L showed moderate discrimination for severe AUB and anemia and should be interpreted as an exploratory risk-enrichment value rather than a definitive diagnostic or treatment cut-off.Conclusion: Thyroid dysfunction, particularly subclinical hypothyroidism, was independently associated with increased bleeding severity, anemia burden, and greater healthcare utilization in perimenopausal AUB. However, given the retrospective observational design, these findings do not establish causality.