血红蛋白糖化指数对糖尿病诊断影响的研究  

Impact of hemoglobin glycation index on the diagnosis of diabetes mellitus

作　　者：林璐[1] 王安平[1] 窦京涛[1] 陈予龙 刘洋[1] 马芳铃[1] 郑华[1] 东黎光[2] 王淑玉[3] 母义明[1] Lin Lu;Wang Anping;Dou Jingtao;Chen Yulong;Liu Yang;Ma Fangling;Zheng Hua;Dong Liguang;Wang Shuyu;Mu Yiming(Department of Endocrinology,First Medical Centre,Chinese People's Liberation Army General Hospital,Beijing 100853,China;Physical Examination Center,Peking University Shougang Hospital,Beijing 100144,China;Beijing Hypertension League Institute,Beijing 100039,China)

机构地区：[1]解放军总医院第一医学中心内分泌科,北京100853 [2]北京大学首钢医院体检中心,北京100144 [3]北京高血压联盟研究所,北京100039

出　　处：《中华糖尿病杂志》2022年第11期1251-1257,共7页CHINESE JOURNAL OF DIABETES MELLITUS

摘　　要：目的探讨血红蛋白糖化指数(HGI)对糖尿病诊断的影响。方法本研究利用中国2型糖尿病患者恶性肿瘤发生风险的流行病学(REACTION)研究北京分中心的基线数据,选取2012年1至8月来自北京市石景山区苹果园社区资料完整的9705名受试者作为研究对象。对受试者进行问卷调查及体格检查,检测空腹血糖、糖化血红蛋白(HbA_(1c))、糖负荷后2 h血糖以及血脂。利用空腹血糖和HbA_(1c)绘制散点图并建立线性回归方程,将每名受试者的FPG代入方程即得到个人的预测HbA_(1c)值,计算HGI,HGI=实测HbA_(1c)-预测HbA_(1c)。最终纳入8480名≥40岁无糖尿病史受试者。依据HGI水平三分位切点将最终纳入分析的8480名受试者分为低HGI、中HGI、高HGI组3组,分别依据HbA_(1c)和口服葡萄糖耐量试验(OGTT)结果诊断糖尿病,采用χ^(2)检验比较3组间的糖尿病检出率,Cohen's kappa系数评价2种标准的诊断一致性。使用有序多分类logisitc回归分析法分析HGI的影响因素。结果低HGI组(HGI<-0.1973)2841例,中HGI组(-0.1973≤HGI<0.1359)2979例,高HGI组(HGI≥0.1359)2660例。单独以HbA_(1c)≥6.5%作为诊断标准,共有905例(10.7%)被诊断为糖尿病,其中高HGI组受试者占比高达74.8%(677/905),远高于中HGI组[15.5%(140/905)]和低HGI组[9.7%(88/905)],3组间差异有统计学意义(P<0.001)。HbA_(1c)和OGTT的诊断一致性较差(κ=0.488),HGI是影响两者诊断一致性的主要因素。以HGI三分位分组作为因变量,有序多分类logistic回归分析结果显示,高龄、女性性别、体重指数、低密度脂蛋白胆固醇、糖负荷后2 h血糖和服用调脂药物是高HGI的影响因素(均P<0.05)。结论单独以HbA_(1c)为糖尿病诊断标准可能会导致高HGI表型个体过度诊断,而对低HGI表型则诊断不足。Objective To investigate the impact of hemoglobin glycation index(HGI)on the diagnosis of diabetes mellitus.Methods Using the baseline data from the Risk Evaluation of Cancers in Chinese Diabetic Individuals:a longitudinal(REACTION)study,9705 participants with complete data,who were selected from Pingguoyuan community of the Shijingshan district in Beijing between January and August 2012,received questionnaires,physical examination and laboratory tests.Laboratory tests consisted of fasting plasma glucose(FPG),glycated hemoglobin A_(1c)(HbA_(1c)),2-hour plasma glucose(2hPG),and lipid profile.A linear regression equation was fitted based on the scatter plots of HbA_(1c) against FPG.The HGI was then calculated as the difference between the measured HbA_(1c) value and that predicted by plugging FPG in the above equation(HGI=measured HbA_(1c)-predicted HbA_(1c)).Ultimately,8480 participants of aged≥40 years and no history of diabetes were included in the final analysis.The participants were categorized into three groups according to tertiles(low,moderate and high)of HGI.We compared the prevalence of diabetes among the three HGI groups with chi-square test,using HbA_(1c) and oral glucose tolerance test(OGTT)criteria respectively,and examined the diagnostic concordance between the two measures by Cohen's kappa test.Factors associated with the high HGI phenotype were identified by order multi-classification logistic regression model.Results Participants were categorized into three groups according to their HGI values:low HGI(HGI<-0.1973,2841 cases)group,moderate HGI(-0.1973≤HGI<0.1359,2979 cases)group,and high HGI(HGI≥0.1359,2660 cases)group.Among patients with diabetes detected by HbA_(1c) alone,the high HGI group accounted for 74.8%(677/905),which was much higher than the moderate(15.5%,140 cases)and the low(9.7%,88 cases)HGI groups.The agreement between HbA_(1c) and OGTT for the diagnosis of diabetes was poor(κ=0.488),which was mostly influenced by HGI.The high HGI phenotype was correlated with age,female gende

关 键 词：糖尿病 诊断 糖化血红蛋白 血红蛋白糖化指数 

分 类 号：R587.1[医药卫生—内分泌]