Development and validation of a nomogram for the prediction of metabolic syndrome in polycystic ovary syndrome in a Chinese population

To construct a nomogram for predicting metabolic syndrome (MetS) in women with polycystic ovary syndrome. In this retrospective study, we analyzed clinical and biochemical data from 859 Chinese women diagnosed with PCOS. Univariable logistic regression and forward stepwise logistic regression were employed to identify independent predictors of MetS. A predictive nomogram was developed that integrates age, acne status, body mass index (BMI), fasting insulin levels (FINS), and the ApoB/ApoA ratio. The model’s discriminative performance, calibration accuracy, and clinical utility were assessed using the area under the receiver operating characteristic curve (AUC), calibration curves accompanied by Brier scores, Hosmer – Lemeshow tests, decision curve analysis (DCA), and clinical impact curves (CIC). Internal validation was conducted through bootstrap resampling over 1,000 iterations. The nomogram exhibited strong discriminative capability with an AUC of 0.874 (95% CI: 0.850–0.899), surpassing BMI alone which had an AUC of 0.824 (<i>p</i> &lt; 0.0001). Both the calibration curve and Hosmer – Lemeshow test indicated satisfactory model fit. DCA and CIC analyses suggested that the nomogram could provide net clinical benefits for risk stratification among PCOS patients. The proposed nomogram accurately predicts MetS risk in PCOS patients, supporting early identification and individualized management.