Abstract: Background　In China，with the implementation of the universal two-child policy，the prevalence of pregnancy in multiparae has risen dramatically，followed by an increase in the incidence of adverse pregnancy outcomes.Multiparae are at higher risk for developing gestational diabetes mellitus （GDM） than primiparas.Objective　To investigate the risk factors of GDM in multiparae，to establish a risk prediction model of GDM for this population.Methods　This prospective study was conducted in a cohort of women（including multiparae and primiparas） who delivered a singleton baby in Beijing Obstetrics and Gynecology Hospital，Capital Medical University，between January and September 2017.Basic personal information，history of pregnancy and childbirth，history of GDM，family history of diabetes，physical parameters，fasting plasma glucose （FPG），and fasting plasma lipids during the first trimester were collected.75-g OGTT was conducted at 24-28 weeks of gestation，and the results were stratified.GDM was diagnosed following the IADPSG 2010 criteria.The above-mentioned parameters and incidence of GDM were compared between multiparae and primiparas.Influencing factors of GDM in multiparae were identified with univariate and multivariate analyses，and were used to establish a risk prediction model of GDM.Results　A total of 5 110 cases were included，and 34.8% （1 779/5 110）of them were multiparae.Compared with primiparas，multiparae had an older age，more pregnancies，higher prepregnancy BMI，TG，and FPG levels，higher percentages of history of adverse pregnancy and childbirth outcomes，history of GDM，and junior college or below education level，and higher GDM incidence （P<0.05）．Among the multiparae，those with GDM showed an older age，more pregnancies，higher prepregnancy BMI，TG， LDL-C and FPG levels，lower HDL-C levels, and higher percentages of history of GDM，and of family history of diabetes than those without （P<0.05）．Logistic regression analysis showed that age，history of GDM，prepregnancy BMI，and FPG in early pregnancy were the influencing factors of GDM in multiparae （P<0.05）．The established Logistic regression model was as follows：PGDM=1/{1+exp〔-（-8.305 3+0.075 2×age+0.530 0×history of GDM+0.125 2×prepregnancy BMI+0.398 4×PFG）〕}.ROC analysis indicated that the AUC of the predictive model was 0.704，the corresponding sensitivity，specificity and Youden's index were 65.84%，68.90% and 0.347 respectively.Conclusion　Multiparae，especially those with glycemic abnormalities during the previous pregnancy，have a high risk of developing GDM，and a population should be intervened in priority.GDM risk prediction model established with variables of age，prepregnancy BMI and FPG in early pregnancy，can provide a theoretical basis for early screening，monitoring，and intervention of GDM in high-risk pregnant women.

Key words: Diabetes, gestational；Multiparity；Age factors；Obesity

摘要： 背景　随着二孩政策的放开，经产妇妊娠比例大幅上升，不良妊娠结局发生率随之增长，经产妇发生妊娠糖尿病（GDM）的风险高于初产妇。目的　探讨经产妇GDM发生风险并建立相关预测模型。方法　通过队列研究的方式，前瞻性收集2017年1—9月于首都医科大学附属北京妇产医院分娩的单胎孕妇基本信息、孕产史、既往史、家族史、身体测量指标及孕早期血糖、血脂水平，结局指标为妊娠糖尿病，诊断采用2010年国际糖尿病与妊娠研究组标准，对孕24~28周75 g口服葡萄糖耐量结果进行分组。分析比较经产妇与初产妇各项指标及GDM发生率，采用单因素、多因素分析探讨经产妇发生GDM的影响因素，并建立风险预测模型。结果　研究共包括5 110例孕妇，经产妇占比34.8%（1 779/5 110）。经产妇年龄、孕次、孕前体质指数（BMI）、三酰甘油（TG）、空腹血糖水平及有不良孕产史、GDM史、受教育程度为专科及以下者占比、GDM发生率高于初产妇，差异有统计学意义（P<0.05）。经产妇中，发生CDM者年龄、孕次、孕前BMI、TG、低密度脂蛋白胆固醇（LDL-C）、空腹血糖水平及有GDM史、糖尿病家族史者占比高于未发生GDM者，高密度脂蛋白胆固醇（HDL-C）水平低于低于未发生GDM者，差异有统计学意义（P<0.05）。Logistic回归分析结果显示：年龄、GDM史、孕前BMI及孕早期空腹血糖是经产妇发生GDM的影响因素（P<0.05）。建立Logistic回归模型为：PGDM=1/{1+exp〔-（-8.305 3+0.075 2×年龄+0.530 0×GDM史+0.125 2×孕前BMI+0.398 4×空腹血糖）〕}。利用上述指标进行受试者工作特征（ROC）曲线建模分析，其曲线下面积（AUC）为0.704，预测的灵敏度为65.84%，特异度为68.90%，约登指数为0.347。结论　经产妇，尤其是前次妊娠血糖异常患者是GDM的高危人群，也是预防干预的重点人群。GDM史结合年龄、孕前BMI及孕早期空腹血糖建立的风险预测模型，可为高危患者早期筛查、监测及干预提供理论依据。

关键词: 糖尿病, 妊娠；经产；年龄因素；肥胖症