GDM患者分娩前血糖变异性参数对巨大儿的预测价值

Predictive value of glucose variability parameters before delivery for macrosomia in patients with gestational diabetes mellitus

目的探讨妊娠糖尿病(GDM)患者分娩前血糖变异性参数与子代出生体重的相关性。方法选取2019年3月—2021年2月连云港市东方医院GDM患者100例,以各项指标均正常的孕妇261例作为对照。根据GDM患者和正常孕妇分娩的新生儿是否为巨大儿(体重≥4000 g)分为GDM巨大儿组、GDM健康儿组、巨大儿对照组、健康儿对照组。收集所有孕妇一般资料和实验室检测结果[血脂、血糖、糖化血红蛋白(HbA_(1c))等],并检测血糖变异性参数[平均血糖波动幅度(MAGE)、日间血糖平均绝对差(MODD)、餐后血糖波动幅度(PPGE)]。采用Logistic回归分析评估GDM患者子代巨大儿的危险因素。构建列线图预测模型,并进行评价。采用Pearson相关分析评估新生儿体重与产妇血糖变异性参数之间的相关性。通过曲线拟合和阈值效应分析确定MAGE、MODD和PPGE的阈值效应。结果GDM巨大儿组新生儿出生30 min的血糖水平显著低于巨大儿对照组(P<0.05)。GDM巨大儿组孕妇空腹血糖(FBG)、餐后2 h血糖(2 h PG)、三酰甘油(TG)、低密度脂蛋白胆固醇(LDL-C)、HbA_(1c)、MAGE、PPGE、MODD、载脂蛋白B(apo B)显著高于其他3组孕妇(P<0.05),高密度脂蛋白胆固醇(HDL-C)、载脂蛋白A(apo A)显著低于其他3组孕妇(P<0.05)。多因素Logistic回归分析结果显示,MAGE、PPGE、MODD和HbA_(1c)均是GDM患者子代巨大儿的危险因素[比值比(OR)值分别为1.116、1.169、1.072、1.061,95%可信区间(CI)分别为1.021~1.246、1.051~1.301、1.012~1.134、1.004~1.121,P<0.05]。Pearson相关分析结果显示,新生儿体重与MAGE、PPGE、MODD和HbA_(1c)均呈正相关(r值分别为0.77、0.68、0.72、0.75,P<0.05)。曲线拟合和阈值效应分析结果显示,当MAGE、PPGE、MODD分别高于2.75、2.06、3.27 mmol·L^(-1)时,GDM患者子代巨大儿的发生率随3项指标的升高而呈上升趋势。列线图预测模型的受试者工作特征(ROC)曲线下面积为0.887,一致性指数(C-index)为0.887,校正曲线与理想曲线拟合度良好。结论血糖变异性参数可作为GDM患者分娩巨大儿的预测指标,具有较高的临床应用价值。

Objective To investigate the correlation between blood glucose variability parameters before delivery and birth weight of newborns in gestational diabetes mellitus(GDM)patients.Methods A total of 100 GDM patients from Lianyungang Dongfang Hospital from March 2019 to February 2021 were enrolled,and 261 pregnant women with normal indicators were enrolled as controls.According to whether the newborns were macrosomia(weight≥4000 g),they were classified into GDM macrosomia group,GDM healthy newborn group,macrosomia control group and healthy control group.The general data and laboratory determination results[blood lipid,blood glucose,glycated hemoglobin A1c(HbA_(1c))and so on]were collected,and blood glucose variability parameters[mean amplitude of glycemic excursions(MAGE),mean of daily differences(MODD),postprandial glucose excursion(PPGE)]were determined.Logistic regression analysis was used to evaluate the risk factors of macrosomia in GDM patients.The prediction model of nomogram was constructed and evaluated.Pearson correlation analysis was used to evaluate the correlation between neonatal weight and maternal blood glucose variability.The threshold effects of MAGE,MODD and PPGE were determined by curve fitting and threshold effect analysis.Results The blood glucose level at 30 min after birth in GDM macrosomia group was lower than that in macrosomia control group(P<0.05).Fasting blood glucose(FBG),2 h postprandial blood glucose(2 h PG),triglyceride(TG),low-density lipoprotein cholesterol(LDL-C),HbA_(1c),MAGE,PPGE,MODD and apolipoprotein B(apo B)in GDM macrosomia group were higher than those in the other 3 groups(P<0.05).High-density lipoprotein cholesterol(HDL-C)and apolipoprotein A(apo A)were lower than those in the other 3 groups(P<0.05).MAGE,PPGE,MODD and HbA_(1c)were all risk factors for macrosomia of GDM patients[odds ratios(OR)were 1.116,1.169,1.072 and 1.061,95%confidence intervals(CI)were 1.021-1.246,1.051-1.301,1.012-1.134,1.004-1.121,respectively,P<0.05].Neonatal weight was positively correlated with MAGE,PPGE,MODD and HbA_(1c)(r=0.77,0.68,0.72 and 0.75,P<0.05).The results of curve fitting and threshold effect analysis show that when MAGE,PPGE and MODD were higher than 2.75,2.06 and 3.27 mmol·L^(-1),respectively,the incidence of macrosomia in GDM patients showed an increasing trend with the increase of the 3 indicators.The area under receiver operating characteristic(ROC)curve(AUC)and consistency index(C-index)of the nomogram prediction model were 0.887 and 0.887,respectively,and the calibration curve fitted the ideal curve well.Conclusions Blood glucose variability parameters can be used as predictors of macrosomia in GDM patients,which have high clinical application value.