LAUSR

This study aimed to evaluate the efficacy of SHBG in predicting insulin resistance (IR) in newly diagnosed, untreated patients with polycystic ovary syndrome (PCOS). Hundred newly diagnosed, untreated patients with PCOS and 61 subjects without PCOS (41 healthy volunteers with normal BMI and 20 subjects with overweight/obese) were included in the study. Receiver-operating characteristic (ROC) analysis was used to assess the effectiveness of SHBG in predicting IR in overweight/obese and non-overweight PCOS patients and the optimal cut-off values of SHBG. The results showed negative correlations between log-SHBG and log-I0 (r = − 0.372, P < 0.001) and log-SHBG and log-Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) (r = − 0.393, P < 0.001) after adjusting for blood pressure, serum lipid, age, and body mass index (BMI) in all of the PCOS patients. In patients with IR (defined as HOMA-IR ≥2.29), the area under the ROC curves (AUCs) of the SHBG for ROC analysis in the non-overweight group, overweight/obese group, and all PCOS patients were 0.774 (P = 0.0001), 0.922 (P = 0.0001), and 0.885 (P = 0.0001), respectively. The optimal cut-off value of SHBG was 37 nmol/L with a sensitivity of 97.62% and specificity of 80.85% in the overweight group. In patients with IR (HOMA-IR ≥2.5), the AUCs of SHBG for ROC analysis in the non-overweight group, overweight/obese group, and all PCOS patients were 0.741 (P = 0.0003), 0.928 (P = 0.0001), and 0.880 (P = 0.0001), respectively. The optimal cut-off value of SHBG was 30.2 nmol/L with a sensitivity of 97.44% and specificity of 82.69% in the overweight/obese group. In conclusion, this study observed a negative correlation between SHBG and HOMA-IR in PCOS patients after adjustment of confounding factors. SHBG was an independent influential factor of HOMA-IR and can be used as a positive predictive marker for IR in PCOS patients, especially in those who are overweight/obese.