摘要目的:评价代谢综合征（metabolic syndrome，MetS）及其组分与甲状腺结节之间的相关性，并构建其诊断预测模型。方法:回顾性收集2020年12月至2022年1月就诊于新疆医科大学第一附属医院的539例18周岁以上MetS患者的人口统计学指标、血液生化检查指标以及甲状腺B超检查结果，数据纳入单变量分析，并用LASSO算法进行变量优化的选择分析。利用多变量Logistic回归分析构建预测模型和绘制列线图。甲状腺结节的危险因素以森林图表示。通过自举迭代法（1000次）对模型进行内部验证，用C统计量确定预测模型的判别能力，应用决策曲线分析来评估预测模型的校正能力。结果:MetS患者中甲状腺结节的患病率为21.33%（115/539）。预测模型列线图包括六个变量，即性别、年龄、甲状腺疾病史、糖化血红蛋白（HbA1c）、高血压诊断和甲状腺体积。C统计量为0.75，表明列线图预测模型具有良好的鉴别能力。校准图显示模型具有良好的校正能力（Hosmer-Lemeshow检验， P=0.228）。决策曲线分析图显示，该列线图的应用与临床实践中的净收益相关，有助于临床决策。 结论:女性、年龄、甲状腺病史、HbA1c%、高血压诊断和甲状腺体积过大是MetS患者甲状腺结节的危险因素。所建立的列线图可以有效预测MetS患者发生甲状腺结节的风险，因此可以指导减少MetS患者罹患甲状腺癌的预防工作。

abstractsObjective:To evaluate the correlation between metabolic syndrome(MetS) and hyroid nodules. Additionally, the establishment of nomogram model is needed to predict the risk factors of thyroid nodules in patients with Mets.Methods:A total of 539 MetS-patients(age >18 years old) who admitted to the First Affiliated Hospital of Xinjiang Medical University from December 2020 to January 2022 were enrolled in this retrospective study. Their demographics, blood biochemical and Doppler ultrasonography examination results were collected into univariate analysis. The LASSO algorithm was used to optimize the selection of variables. Subsequently, multivariable Logistic regression analyses(containing logit scale transformation) were applied to construct a predictive model and a nomogram. The risk factors of thyroid nodules were presented as a forest plot. Internal validation of the model was performed by bootstrapping(1000 times). C-statistic was used to determine the discriminatory ability of the predictive model. The decision curve analysis(DCA) was applied to evaluate the calibration ability of the prediction model.Results:The prevalence of thyroid nodules was 21.33%(115 cases) among patients with MetS. Six variables, embracing gender, age, history of thyroid disease, HbA1c%, hypertension, and thyroid volume, were included in the nomogram. The c-statistic was 0.75, indicating the upstanding discriminative ability of the nomogram. The calibration plot showed the model had good calibration ability(Hosmer-Lemeshow test, P=0.228). The DCA diagram showed that the application of this nomogram is associated with net benefit gains in clinical practice. Conclusions:Female, age, history of thyroid disease, HbA1c%, hypertension and large thyroid volume are the risk factors of thyroid nodule in MetS patients. The established nomogram model can effectively predict the risk of occurrence of thyroid nodule in MetS patients; hence, it can guide prevention efforts to reduce thyroid cancer in MetS patients.