Hu Yun, Shen Yun, Yan Rengna, Li Fengfei, Ding Bo, Wang Huiying, Su Xiaofei, Ma Jianhua
Department of Endocrinology, Nanjing First Hospital, Nanjing Medical University, Nanjing, Jiangsu, China.
Diabetes Ther. 2020 Sep;11(9):2019-2027. doi: 10.1007/s13300-020-00879-x. Epub 2020 Jul 21.
The aim of this study was to investigate the relationship between actual measured glycated hemoglobin (HbA1c) and estimated glycated hemoglobin (EA1c) in the flash glucose monitoring (FGM) system in Chinese patients with type 2 diabetes.
This study was conducted in Nanjing First Hospital. Each patient used FGM twice in a 3-month period (during the first 14 days immediately after baseline and during a second 14-day period from days 76 to 90 after baseline). HbA1c measurements were made using a high-performance liquid chromatography assay before the start of the first FGM period (baseline) and at the end of the second FGM period.
A total of 74 patients (35 men; mean age ± standard deviation [SD] 67.6 ± 5.2 years) were enrolled in the study. The mean (± SD) duration of diabetes was 11.9 ± 7.8 months. The first and second HbA1c measurements were both higher than the EA1c (both p < 0.001). Mean glucose (MG) gradually decreased over time and was the lowest on day 14. Linear regression showed that only HbA1c at baseline affected the gap between HbA1c and EA1c (β = 0.319, p = 0.01) when the educational level, age, gender, duration of diabetes, body mass index, HbA1c at baseline, and number of scans daily were included as independent variables. The best model for calculating EA1c was EA1c% = MG mmol/L × 0.669 - 0.213 × 8th MG + 3.351 when MG > 9.7 mmol/L, and EA1c % = (MG mmol/L + 2.590)/1.590 when MG ≤ 9.7 mmol/L. The correlation coefficient for EA1c and HbA1c in this model (model 7) is higher than that reported the original model in the FGM system 1 (0.955 vs. 0.822, respectively; p < 0.001).
The EA1c used by FreeStyle Libre™ is lower than the actual measured HbA1c. Improvement in the glucose levels during FGM in these patients may contribute to the lowering of EA1c.
The study is registered with ClinicalTrials.gov, number NCT03785301.
本研究旨在调查中国2型糖尿病患者在瞬感葡萄糖监测（FGM）系统中实际测量的糖化血红蛋白（HbA1c）与估算糖化血红蛋白（EA1c）之间的关系。
本研究在南京第一医院开展。每位患者在3个月内使用FGM两次（在基线期后的前14天以及基线期后第76至90天的第二个14天期间）。在第一个FGM期开始前（基线）和第二个FGM期结束时，采用高效液相色谱法测量HbA1c。
共有74例患者（35例男性；平均年龄±标准差[SD]为67.6±5.2岁）纳入本研究。糖尿病平均病程为11.9±7.8个月。首次和第二次HbA1c测量值均高于EA1c（均p<0.001）。平均血糖（MG）随时间逐渐下降，在第14天时最低。线性回归分析显示，当将教育程度、年龄、性别、糖尿病病程、体重指数、基线HbA1c以及每日扫描次数作为自变量纳入时，仅基线HbA1c会影响HbA1c与EA1c之间的差值（β=0.319，p=0.01）。计算EA1c的最佳模型为：当MG>9.7 mmol/L时，EA1c% = MG mmol/L×0.669 - 0.213×第8次MG + 3.351；当MG≤9.7 mmol/L时，EA1c% = (MG mmol/L + 2.590)/1.590。该模型（模型7）中EA1c与HbA1c的相关系数高于FGM系统1中原始模型所报告的相关系数（分别为0.955和0.822；p<0.001）。
FreeStyle Libre™所使用的EA1c低于实际测量的HbA1c。这些患者在FGM期间血糖水平的改善可能有助于降低EA1c。
本研究已在ClinicalTrials.gov注册，注册号为NCT03785301。
