Diabetes and uterine fibroid diagnosis in midlife: Study of Women's Health Across the Nation (SWAN).

Abstract

BACKGROUND Fibroids are non-cancerous uterine tumors potentially associated with cardiovascular risk factors. We examined prospectively associations of glucose, insulin, sex hormone binding globulin (SHBG), and diabetes with incidence of fibroid diagnoses in midlife. METHODS Participants in the Study of Women's Health Across the Nation (SWAN) cohort (n=2570) reported fibroid diagnoses at enrollment (1996-1997) and 13 follow-up visits (1996-2013). At all visits, we measured glucose, insulin, and SHBG in fasting blood samples and calculated homeostatic model assessment for insulin resistance (HOMA-IR). Diabetes was defined using glucose levels, self-reported diabetes, or diabetes medication use. We used discrete-time survival models to estimate hazard ratios (HR) and 95% confidence intervals (CI) for associations of time-varying biomarkers and diabetes with incident fibroid diagnoses, adjusted for demographics and healthcare utilization. We also evaluated effect modification by menopausal status. RESULTS At baseline, 2.7% of participants (n=70) were using diabetes medication. Time-varying glucose, insulin, HOMA-IR, and SHBG were not associated with fibroid diagnosis. However, diabetes was associated with a 28% lower incidence of fibroid diagnosis (adjusted HR: 0.72, 95% CI: 0.44, 1.17), driven by participants using metformin (adjusted HR: 0.49, 95% CI: 0.21, 1.12), though precision was limited. After stratification by menopausal status, higher HOMA-IR and insulin were associated with greater incidence of fibroid diagnosis during premenopause but not perimenopause, while the inverse association between diabetes and fibroids was strongest during perimenopause. CONCLUSION The effect of diabetes and biomarkers on fibroids may vary by menopausal status. Fibroid risk may increase with insulin resistance and decrease with diabetes treatment.