Body mass index is a good predictor of metabolic abnormalities in polycystic ovary syndrome.

AIM: To assess which parameters among hyperandrogenism (total testosterone-tT-or free androgen index-FAI), sex hormone binding globulin (SHBG) or body mass index (BMI) could better predict a worse metabolic profile in women with polycystic ovary syndrome (PCOS).

METHODS: Five hundred and eighty-six women with PCOS and clinical or biochemical hyperandrogenism were included. Receiver Operating Characteristics (ROC) curves with tT, FAI, SHBG and BMI were performed for metabolic parameters and a cut-off with sensitivity and specificity was obtained for each parameter. The women were then divided into three groups and compared according to their BMI.

RESULTS: Based on the ROC curves, tT proved not to be a good predictor of metabolic alterations. FAI and SHBG resulted to be good predictors of some markers of metabolic damage. The area under the curves (AUC) of SHBG were greater than those of FAI. SHBG levels affects the values of homeostasis model assessment of insulin resistance (HOMA-IR), fasting insulin, high density lipoproteins (HDL), low density lipoproteins (LDL), and total cholesterol also when corrected for BMI. However, the highest AUCs of the ROC curves were observed when BMI was used, which was significantly related to all the metabolic parameters analyzed. Dividing women according to their BMI, BMI between 25.00 and 30.00 kg/m2 had a worse metabolic profile but still in a normal range, while BMI ≥ 30 kg/m2 women had a significant metabolic derangement.

DISCUSSION: BMI is a good predictor factor of metabolic changes in PCOS women at any age, and obesity is associated to the appearance of metabolic complications. Overweight and obese PCOS women should be addressed to perform a complete metabolic assessment.