妊娠早期空腹血浆葡萄糖水平与妊娠期糖尿病诊断的相关性

Relationship between fasting plasma glucose in early pregnancy and diagnosis of gestational diabetes mellitus

目的探讨妊娠早期空腹血浆葡萄糖（fasting plasma glucose，FPG）水平与妊娠期糖尿病（gestafional diabetes mellitus，GDM）诊断的关系，分析同际糖尿病与妊娠研究组（International Association of Diabetesand Pregnancy Study Groups，IADPSG）新GDM诊断标准中妊娠早期FPG作为GDM诊断标准的适用性。方法收集2011年4月1日至12月31日在北京大学第一医院行产前检查的非孕前糖尿病孕妇2761例临床资料，比较GDM与非GDM2组孕妇妊娠早期FPG水平；同时依据妊娠早期FPG水平分为FPG〈5．1mmol／L组（2431例）和FPG≥5．1mmol／L组（330例），比较GDM发生率。采用t或X^2检验比较各组妊娠结局，FPG对GDM发生风险的预测行Logistic回归分析及受试者工作特性曲线分析。结果（1）2761例孕妇中，诊断GDM515例，发生率18．7％。GDM组妊娠早期FPG水平显著高于非GDM组[（4．84±0．46）mmol／L与（4．57±0．35）mmol／L,t=11．924，P=0．000]，妊娠早期FPG每升高lmmol／L，发生GDM风险增加7．984倍（OR=8．984，95％CI：6．605～12．220）。（2）FPG〈5．1mmol／L组与≥5．1mmol／L组在妊娠中晚期被诊断GDM的比例分别是15．2％（370／2431）和43．9％（145／330），差异有统计学意义（X^2=123．976，P=0．000）。FPG≥6．1mmol／L者共5例，均于妊娠中期诊断GDM。（3）妊娠早期FPG与GDM诊断的受试者工作特性曲线分析：最大曲线下面积0．718，95％CI：0．690～0．747；以4．795mmol／L和4．785mmol／L为界值时，诊断GDM的敏感性和特异性分别是0．600、0．612和0．735、0．726。（4）2761例孕妇中已分娩1208例，其中GDM227例，非GDM981例，2组剖宫产率分别为54．2％（123／227）和39．2％（385／981），差异有统计学意义（X^2=16．884，P=0．000），巨大儿、新生儿高胆红素血症、低出生体重儿、早产、胎儿生长受限、子痫前期的发生率差异均无统计学意义（P均〉．05）；FPG（5．1mmol／L组和≥5．1mmol／L组中GDM分别为173例和54例，早产发生率分别是5．8％（10／173）和14．8％（8／54），≥5．1mmol／L组显著升高（X^2=4．601，P〈0．05），剖宫产、胰岛素应用、巨大儿、子痫前期的发生率差异均无统计学意义（P均〉0．05）。结论妊娠早期以FPG≥5．1mmol／L作为GDM诊断标准会出现过度诊断，不建议推广，但有评估发生GDM风险的价值，值得临床重视。

Objective To investigate the relationship between fasting plasma glucose （FPG） in early pregnancy and diagnosis of gestational diabetes mellitus （GDM） and to confirm the rationality of the new standard for GDM diagnosis in early pregnancy set by the International Association of Diabetes and Pregnancy Study Groups （1ADPSG）. Methods Clinical materials of 2761 pregnant women without diabetes mellitus, who accepted prenatal cares in Peking University First Hospital from April 1, 2011 to December 31, 2011, were collected and analyzed. The difference between FPG levels of GDM and non GDM women was compared. According to the early pregnancy FPG level, the subjects were divided into group A （FPG〈5.1 mmol/L, n 2431） and B （FPG≥5. 1 mmol/L, n=330）. The incidence of GDM and pregnant outcomes of the two groups were compared with t or Chi square test. Relationship between FPG and GDM was analyzed by Logistic regression and receiver operating characteristic curve. Results （1） Among the 2761 subjects, 515 were diagnosed as GDM （18.7%） and the early pregnancy FPG level in GDM group was significantly higher than that in nonGDM group [（4.84±0.46） mmol/L vs （4.57±0.35） retool/L, t±11.924, P=0.0001. In early pregnancy, the risk of GDM increased by 7. 984-fold （OR=8. 984, 95±CI： 6. 605-12. 220） with every 1 mmol/L increase of the FPG level. （2） The diagnostic rate of GDM during mid-and last trimester in group A （15.2%, 370/2431） was lower than that of group B （43.9%, 145/330）, Z2=123.976,P=0.000. （3） Receiver operating characteristic curve analysis of FPG in early pregnancy and diagnosis of GDM： The largest area under the curve was 0. 718 （95%CI 0. 690-0. 747）. The sensitivity and specificity were 0. 600 and 0. 612, or 0. 735 and 0. 726 respectively, when 4. 795 mmol/L or 4. 785 mmol/L were set as the cut-off value. （4） Among the 1208 cases delivered, GDM was diagnosed in 227 cases. The cesarean section rate （54.2%, 123/227） of GDM women was higher than that （39. 2%, 385/981） of non GDM women （X^2= 16. 884,P=0. 000）. There were no differences in the incidences of macrosomia, neonatal hyperbilirubinemia, low birthweight infant, premature delivery, fetal growth restriction and preeclampsia between GDM and non-GDM group （all P〉0.05）. The incidence of premature birth in GDM women with FPG〈5. 1 mmol/L was lower （5.8%, 10/173） than that （14.8%, 8/54） of women with FPG≥5.1 mmol/L （X^2=4. 601,P〈0.05）. The incidence of cesarean section, insulin administration, macrosomia and preeclampsia increased from low FPG group to high FPG group, however there was no statistical significances. Conclusions Diagnosing GDM with FPG≥5.1 mmol/L in early pregnancy is not recommended as over diagnosis might happen. But this cut off value might indicate that the patient are at risk of GDM, and this population should not be ignored.