Ovarian volume is more closely related to the different manifestations of polycystic ovary syndrome than follicle number per ovary

Abstract

Objective: Polycystic ovary (PCO), a diagnostic component of polycystic ovary syndrome (PCOS), requires either an ovarian volume (OV) criterion or a follicle number per ovary (FNPO) criterion. This study investigated the association of OV and FNPO criteria with various manifestations of PCOS. Methods: This cross-sectional study was conducted at a university hospital among 100 patients newly diagnosed with PCOS (according to the revised Rotterdam criteria). Fasting blood samples were collected to measure glucose, total testosterone (TT), luteinizing hormone (LH), follicle-stimulating hormone (FSH), lipid, insulin, and hemoglobin A1c levels. An oral glucose tolerance test was performed. Transabdominal or transvaginal ultrasound of the ovaries was done, depending on patients’ marital status. All investigations were conducted in the follicular phase of the menstrual cycle. OV >10 mL and/or FNPO ≥12 indicated PCO. A homeostasis model assessment of insulin resistance (IR) value ≥2.6 indicated IR, and metabolic syndrome (MS) was defined according to the international harmonization criteria. Results: Seventy-six participants fulfilled the OV criterion, 70 fulfilled the FNPO criterion, and 89 overall had PCO. Both maximum OV and mean OV had a significant correlation with TT levels (<italic>r</italic>=0.239, <italic>p</italic>=0.017 and <italic>r</italic>=0.280, <italic>p</italic>=0.005, respectively) and the LH/FSH ratio (<italic>r</italic>=0.212, <italic>p</italic>=0.034 and <italic>r</italic>=0.200, <italic>p</italic>=0.047, respectively). Mean OV also had a significant correlation with fasting insulin levels (<italic>r</italic>=0.210, <italic>p</italic>=0.036). Multivariate binary logistic regression analysis showed that IR (odds ratio [OR], 9.429; 95% confidence interval [CI], 1.701 to 52.271; <italic>p</italic>=0.010) and MS (OR, 7.952; 95% CI, 1.821 to 34.731; <italic>p</italic>=0.006) had significant predictive associations with OV alone, even after adjustment for age and body mass index. Conclusion: OV may be more closely related to the androgenic and metabolic characteristics of PCOS than FNPO.