个体化预测甲状腺功能减退孕妇发生妊娠期糖尿病的风险预测Nomogram模型的建立与验证

Establishment and validation of an individualized risk prediction nomogram model for gestational diabetes mellitus in pregnant women with hypothyroidism

目的建立个体化预测甲状腺功能减退孕妇发生妊娠期糖尿病的风险预测Nomogram模型并验证。方法选取甲状腺功能减退孕产妇160例作为研究对象,其中合并妊娠期糖尿病患者(观察组)85例,血糖正常患者(对照组)75例。比较2组患者的年龄、孕次、产次以及孕前体质量指数、甲状腺过氧化物酶抗体(TPOAb)、胰岛素抵抗指数(HOMA-IR)、胰岛β细胞功能指数(HOMA-β)、游离四碘甲腺原氨酸(FT4)、游离三碘甲腺原氨酸(FT3)和促甲状腺激素(TSH)。采用多因素分析法分析甲状腺功能减退孕妇发生妊娠期糖尿病的危险因素。采用Nomogram模型预测甲状腺功能减退孕妇发生妊娠期糖尿病的风险。采用受试者工作特征(ROC)曲线评估Nomogram模型对状腺功能减退孕妇发生妊娠期糖尿病的预测价值。结果观察组患者的年龄、孕次、产次以及孕前体质量指数、TPOAb、HOMA-IR、HOMA-β高于对照组,孕前FT4、FT3、TSH低于对照组,差异有统计学意义(P<0.05)。多因素分析显示,较高的年龄、孕次、产次以及孕前体质量指数、TPOAb、HOMA-IR、HOMA-β和较低的孕前FT4、FT3、TSH均是甲状腺功能减退孕妇发生妊娠期糖尿病的危险因素。Nomogram模型预测显示,较高的年龄(28~31岁)、孕次(2次以上)、产次(1次以上)和孕前体质量指数(24.00 kg/m^(2)以上)、TPOAb(>421.33 IU/mL)、HOMA-IR(>2.21)、HOMA-β(>115.66)以及较低的孕前FT4(<0.33 ng/dL)、FT3(<1.65 pg/mL)、TSH(<0.26 mIU/mL)均是甲状腺功能减退孕妇发生妊娠期糖尿病的危险因素。ROC曲线分析显示,Nomogram模型预测甲状腺功能减退孕妇发生妊娠期糖尿病的曲线下面积为0.785,预测效能较好。结论风险预测Nomogram模型在个体化预测甲状腺功能减退孕妇发生妊娠期糖尿病中的价值较高。

Objective To establish and validate an individualized risk prediction nomogram model for the occurrence of gestational diabetes mellitus in pregnant women with hypothyroidism.Methods A total of 160 pregnant women with hypothyroidism were selected as the study subjects,including 85 patients with gestational diabetes mellitus(observation group)and 75 patients with normal blood glucose levels(control group).The age,gravidity,parity as well as pre-pregnancy body mass index,thyroid peroxidase antibody(TPOAb),homeostatic model assessment of insulin resistance(HOMA-IR),homeostatic model assessment of beta-cell function(HOMA-β),free thyroxine(FT4),free triiodothyronine(FT3)and thyroid stimulating hormone(TSH)were compared between the two groups.Multivariable analysis was used to analyze the risk factors for gestational diabetes mellitusin pregnant women with hypothyroidism.The nomogram model was used to predict the risk of gestational diabetes mellitus in pregnant women with hypothyroidism.The receiver operating characteristic(ROC)curve was used to evaluate the predictive value of the nomogram model for gestational diabetes mellitus in pregnant women with hypothyroidism.Results The age,gravidity,parity as well as pre-pregnancy body mass index,TPOAb,HOMA-IR and HOMA-βwere higher in the observation group than in the control group,while pre-pregnancy FT4,FT3 and TSH were lower(P<0.05).Multivariable analysis showed that older age,gravidity,parity as well as pre-pregnancy body mass index,TPOAb,HOMA-IR and HOMA-β,and lower pre-pregnancy FT4,FT3 and TSH were risk factors for gestational diabetes mellitusin pregnant women with hypothyroidism.The nomogram model prediction showed that older age(28 to 31 years),gravidity(more than 2 times),parity(more than 1 time)as well as pre-pregnancy body mass index(above 24.00 kg/m^(2)),TPOAb(>421.33 IU/mL),HOMA-IR(>2.21)and HOMA-β(>115.66),and lower pre-pregnancy FT4(<0.33 ng/dL),FT3(<1.65 pg/mL)and TSH(<0.26 mIU/mL)were risk factors for gestational diabetes mellitus in pregnant women with hypothyroidism.ROC curve analysis showed that the area under the curve of the nomogram model for predicting gestational diabetes mellitus in pregnant women with hypothyroidism was 0.785,indicating good predictive performance.Conclusion The risk prediction nomogram model has high value in individualized prediction of gestational diabetes mellitus in pregnant women with hypothyroidism.