非传统糖代谢指标（GA、FA、1，5-AG）的临床应用研究

发布时间：2018-01-03 13:26

本文关键词：非传统糖代谢指标（GA、FA、1，，5-AG）的临床应用研究　出处：《安徽医科大学》2017年硕士论文　论文类型：学位论文

【摘要】：第一部分酶法检测糖化白蛋白主要方法学性能评估及白蛋白检测对糖化白蛋白检测的影响目的评估液态酶法检测血清糖化白蛋白(Glycated albumin,GA)的主要方法学性能;探讨溴甲酚绿法(BCG法)、改良溴甲酚紫法(m BCP法)检测血清白蛋白(Alb)对GA测定的影响;比较BCG法、m BCP法与免疫比浊法(ITA法)检测血清Alb结果的差异。方法评估液态酶法检测血清GA的准确度、精密度、线性范围、回收试验、参考区间;随机抽取306份不同白蛋白(Alb)浓度新鲜血清标本,分别用溴甲酚绿法(BCG)、改良溴甲酚紫法(m BCP)、免疫比浊法(ITA)检测Alb浓度,液态酶法检测GA,并计算GA值(100*GA/Alb)。结果液态酶法检测血清GA的准确度良好,以批内均值为测得值计算GA的平均偏倚为2.75%;GA各浓度的重复性及实验室内精密度的CV值均小于试剂盒声明的精密度(3%);所选40例正常体检对象的GA水平也均在厂商说明书提供的参考范围(11%~16%)内。线性范围评价中,回归方程为Y=1.0153X+0.1608,R2=0.9991,GA浓度在0.45g/L~31.2g/L范围内呈线性,满足线性要求;回收试验中,对照样本与加标样本间平均偏差为1.57g/L,加标平均回收率为102.46%。正常Alb浓度(Alb≥40 g/L)时,BCG法与m BCP法、BCG法与ITA法、m BCP法与ITA法检测结果比较,临床差异不明显且GABCG与GAm BCP结果间差异无统计学意义(p=0.537);低Alb浓度(Alb40 g/L)时,BCG法与m BCP、ITA法间差异有统计学意义(p0.01),GABCG明显低于GAm BCP(p0.01)。结论该法分析性能良好,满足临床需求。正常Alb浓度时,BCG法、m BCP法与ITA法三者结果差异无明显临床意义,各法用于GA计算无明显差异;但低Alb浓度时,m BCP法与ITA法二者一致性优于BCG法与ITA法,若使用BCG法则GA结果可能被低估。第二部分线性回归、非线性拟合评估糖代谢指标间关系及糖代谢指标对糖尿病诊断性能比较目的比较线性拟合和非线性拟合评估糖代谢指标间关系,并比较糖代谢指标诊断糖尿病性能。方法回顾性分析549例住院患者的糖化血红蛋白(Hb A1c)、空腹血糖(fasting glucose,FPG)、GA以及糖化血清蛋白(fructosamine,FA)间关系。非线性拟合模型中,采用确定系数确定最适拟合方法。线性拟合及非线性拟合间差异采用费歇尔Z转化。应用受试者特征工作曲线(ROC曲线)分析糖代谢指标诊断糖尿病的曲线下面积(the weighted area under the curve,AUC)。结果GA,FA与Hb A1c,FPG相关性高。线性回归及非线性拟合中,GA与Hb A1c相关性均要强于GA与FA(均p0.001)。总体上,与线性回归相比,最佳非线性拟合GA与Hb A1c相关性由0.828提高至0.871(p0.05),GA与FPG相关性由0.618提高至0.792(p0.01)。糖尿病组中,线性回归与非线性拟合评估糖代谢指标间关系性能相当。而非糖尿组中,非线性拟合较线性回归更好地评估糖代谢指标间关系。以患者临床诊断为糖尿病诊断依据时,Hb A1c的AUC最大为0.831(0.796~0.867),其次为GA 0.811(0.774~0.849)、FA 0.758(0.717~0.798),FPG最小为0.695(0.651~0.740)。结论与线性回归相比,非线性拟合可能更适用于评估GA,FA与Hb A1c,FPG关系。GA、FA可作为Hb A1c和FPG的良好补充,甚至为糖尿病潜在诊断指标。第三部分适合本地区人群的非传统糖代谢指标参考区间研究目的建立适合本地区人群GA、FA及1,5-脱水葡萄糖醇(1,5-AG)的参考区间。方法收集458例健康体检者,为参考人群。其中男226例,女232例,年龄在20~79岁(中位年龄43岁)。根据性别(男、女)、年龄(20~39岁、40~59岁及60~79岁)分组,结合Lathi算法以及适当的临床考虑评估参考区间分组必要性。结果GA性别间差异有统计学意义(p0.001),但参考区间上限的绝对差异仅0.31%,不足以纳入临床考虑。GA水平在三个年龄分组中差异有统计学意义(p0.001),Lathi算法建议就20~59岁及60~79岁分组,故建议GA参考区间分别为10.38~13.89%,10.23~14.79%。1,5-AG水平男性要显著高于女性(p0.001),平均水平绝对差异为51μmol/L(8.5μg/m L),故按照性别分组,男性为107~367μmol/L,女性为79~306μmol/L。FA性别间参考下限绝对差异仅为7μmol/L,年龄间差异无统计学意义(P0.05),故FA参考区间为220~298μmol/L.结论成功建立适合本地区人群的非传统糖代谢指标参考区间。
[Abstract]:The first part mainly enzymatic detection of glycated albumin detection method performance evaluation and effect on albumin glycated albumin detection to evaluate serum glycated albumin (Glycated enzymatic method albumin, GA) the main method performance study; bromocresol green method (BCG method), modified bromocresol purple method (m BCP method) to detect serum albumin (Alb) effect on the determination of the GA; BCG m method, BCP method and immunoturbidimetry (ITA) differences in detection of serum Alb results. The evaluation method of the detection of serum GA liquid enzymatic method for accuracy, precision, linear range, recovery test and reference interval; 306 random samples of different albumin (Alb) concentrations of fresh serum samples respectively by bromocresol green method (BCG), modified bromocresol purple method (m BCP), immune turbidimetric method (ITA) to detect the concentration of Alb, detection of GA liquid enzymatic method, and calculate the GA value (100 *GA/Alb). The detection of serum GA liquid enzymatic method accurately Well, the batch mean measured value calculated the average bias of 2.75% GA; repeatability of different concentration of GA and laboratory precision CV values were less than the precision of the kit statement (3%); the reference range of the selected 40 cases of normal healthy object GA levels also are provided in the instructions of the manufacturer (11%~16%). The linear range of the evaluation, the regression equation was Y=1.0153X+0.1608, R2=0.9991, GA concentration was linear in the range of 0.45g/L~31.2g/L, to meet the requirements of linear; recovery test, the average deviation control samples and spiked samples was 1.57g/L, the average recovery rate of 102.46%. in normal Alb concentration (Alb = 40, g/L) BCG method and m BCP method, BCG method and ITA method, comparison of the test results of M BCP method and ITA method, no significant difference in clinical difference is not obvious and the GABCG and GAm results between BCP (p=0.537); low concentration of Alb (Alb40 g/L), BCG and m BCP, there was statistically significant difference between the ITA method Yi (P0.01), GABCG was significantly lower than that of GAm BCP (P0.01). Conclusion the method has good performance, meet the clinical needs. The normal concentration of Alb, BCG method, the results of the three m BCP method and ITA method were not obvious clinical significance, the method for calculation of GA had no significant difference; but the low concentration of Alb, m BCP method and ITA method two is more consistent than the BCG method and ITA method, if using the BCG rule GA results may be underestimated. In the second part, linear regression, nonlinear fitting evaluation index the relationship between glucose metabolism and glucose metabolism in diabetes mellitus to compare the diagnostic performance of linear fitting and nonlinear fitting evaluation indexes of glucose metabolism and the relationship between. Comparison of glucose metabolism index in the diagnosis of diabetes performance. Methods: a retrospective analysis of 549 cases of hospitalized patients with HbA1c (Hb A1c), fasting blood glucose (fasting, glucose, FPG, GA) and glycosylated serum protein (fructosamine, FA). The relationship between nonlinear fitting model in mining With the coefficient of determination of the optimum fitting method. The difference between the linear fitting and nonlinear fitting using Fischer Z transformation. By using the receiver operating characteristic curve (ROC curve) analysis of glucose metabolism index in the diagnosis of diabetic area under the curve (the weighted area under the curve, AUC). The results of GA, FA and Hb A1c, FPG correlation high. Linear regression and nonlinear fitting, the correlation between GA and Hb A1c were stronger than GA and FA (p0.001). On the whole, compared with the linear regression, nonlinear fitting and Hb A1c GA the best correlation is increased from 0.828 to 0.871 (P0.05), GA and FPG correlation increased from 0.618 to 0.792 (P0.01). The diabetes group, linear regression and nonlinear fitting evaluation indexes of glucose metabolism. The relationship between the performance and non diabetes group compared with the linear regression, nonlinear fitting better assessment indexes of glucose metabolism. The relationship between the clinical diagnosis of patients with diabetes diagnosis, Hb A1c A The maximum UC was 0.831 (0.796~0.867), followed by GA 0.811 (0.774~0.849), FA 0.758 (0.717~0.798) FPG, the minimum was 0.695 (0.651~0.740). Conclusion compared with linear regression, nonlinear fitting may be more suitable for the evaluation of GA, FA and Hb A1c, FPG.GA, FA can be used as a good supplement of Hb A1c and FPG even as a potential diagnostic index, diabetes. Non traditional Glycmetabolism reference interval objective third part for the local population to establish local populations of GA, FA and 1,5- Ag (1,5-AG) of the reference interval. Methods 458 healthy subjects, as the reference population. There were 226 male and 232 female patients and at the age of 20~79 years (median age 43 years). According to gender (male, female), age (20~39 years, 40~59 years and 60~79 years) group, combined with the Lathi algorithm and the appropriate clinical assessment is necessary to consider the reference interval of packet. There was a significant difference in GA between sexes (P results 0.001), but the absolute difference of reference range limit is only 0.31%, not enough to consider in clinical.GA level difference was statistically significant in three age groups (p0.001), the proposed Lathi algorithm of 20~59 and 60~79 years old group, it is suggested that the GA reference intervals were 10.38~13.89%, 10.23~14.79%.1,5-AG levels of male is significantly higher than that of female (p0.001) the average absolute difference, 51 mol/L (8.5 g/m L), it is grouped according to the gender, the male is 107~367 mol/L, mol/L.FA 79~306 for female gender difference is only the absolute reference limit of 7 mol/L, there was no significant difference between the age (P0.05), the FA reference range for 220~298 mol/L. successfully. The establishment of non traditional Glycmetabolism reference interval region suitable for the population.

【学位授予单位】：安徽医科大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R587.1

【参考文献】