基于Irisin、Kisspeptin、SHBG构建妊娠期糖尿病早产的风险模型及其验证  

Construction and validation of a risk model for preterm delivery in gestational diabetes mellitus based on Irisin,Kisspeptin,and SHBG

作　　者：唐余[1] 邓夏 陈玉兰 李念 TANG Yu;DENG Xia;CHEN Yulan;LI Nian(Department of Obstetrics and Gynecology,Mianyang Central Hospital,Mianyang 621000,China)

机构地区：[1]绵阳市中心医院妇产科,四川绵阳621000

出　　处：《东南大学学报（医学版）》2023年第6期813-819,共7页Journal of Southeast University(Medical Science Edition)

摘　　要：目的:基于鸢尾素(Irisin)、人吻素-1编码的肽类激素Kisspeptin、性激素结合球蛋白(SHBG)构建妊娠期糖尿病(GDM)早产的风险模型,并验证模型的预测价值。方法:选取2017年2月至2022年1月237例GDM患者,根据是否发生早产分为发生组(n=78)、未发生组(n=159),比较两组一般资料、Irisin、Kisspeptin、SHBG,采用Logistic回归分析早产的相关影响因素,使用R语言绘制预测早产的列线图模型,并对模型进行内部与外部验证。结果:发生组空腹血糖、餐后2 h血糖高于未发生组(P<0.05);发生组孕24周、孕28周、分娩前Irisin、SHBG低于未发生组,Kisspeptin高于未发生组(P<0.05);空腹血糖、餐后2 h血糖、Irisin、Kisspeptin、SHBG均与早产相关(P<0.05);基于Logistic回归分析绘制预测早产的列线图模型,该模型预测早产的C-index为0.972(95%CI:0.718~0.994),区分度为0.801,校准度为0.925,提示分类结果与实际结果具有较高的符合度。结论:Irisin、SHBG降低及Kisspeptin升高均与GDM患者早产风险增加相关,基于各指标构建的预测早产列线图模型能为临床早期预测妊娠结局提供可靠参考,从而指导临床及时管理、干预,促进结局的改善。Objective:To construct a risk model for preterm delivery in gestational diabetes mellitus(GDM)based on Irisin,the human kissin-1 encoded peptide hormone Kisspeptin,and sex hormone-binding globulin(SHBG),and to validate the predictive value of the model.Methods:237 patients with GDM from February 2017 to January 2022 were selected and divided into incident group(n=78)and non-incident group(n=159)according to whether preterm delivery occurred,comparing general information,Irisin,Kisspeptin and SHBG between the two groups,analyzing the influencing factors related to preterm delivery using Logistic regression,drawing a nomogram model for predicting preterm delivery using R language,and validating the model internally and externally.Results:Fasting glucose and 2 h postprandial glucose were higher in the incident group than those in the non-incident group(P<0.05);Irisin and SHBG at 24 weeks of gestation,28 weeks of gestation and before delivery were lower in the incident group than those in the non-incident group,and Kisspeptin was higher than that in the non-incident group(P<0.05);fasting glucose,2 h postprandial glucose,Irisin,Kisspeptin and SHBG were all associated with preterm delivery(P<0.05);based on Logistic regression analysis to plot the nomogram model for predicting preterm labor,the C-index of the model for predicting preterm delivery was 0.972(95%CI:0.718-0.994),with a discrimination of 0.801 and a calibration of 0.925,suggesting that the predicted classification results have a high degree of compliance with the actual results.Conclusion:The decreased Irisin,SHBG and increased Kisspeptin are all associated with increased risk of preterm delivery in patients with GDM,and the nomogram model of predicting preterm delivery constructed based on each index can provide a reliable reference for clinical early prediction of pregnancy outcome,thus guiding clinical timely management,intervention and promoting improvement of outcome.

关 键 词：妊娠期糖尿病 早产 鸢尾素 人吻素-1编码的肽类激素 性激素结合球蛋白 列线图模型 

分 类 号：R587.1[医药卫生—内分泌] R714.256[医药卫生—内科学] R195.1[医药卫生—临床医学]