Department of Endocrinology and Diabetes, Xiamen Diabetes Institute, Xiamen Clinical Medical Center for Endocrine and Metabolic Diseases, Xiamen Diabetes Prevention and Treatment Center, Fujian Key Laboratory of Diabetes Translational Medicine, The First Affiliated Hospital of Xiamen University, School of Medicine, Xiamen University, Xiamen, China.
Fujian Provincial Key Laboratory of Ophthalmology and Visual Science, Eye Institute of Xiamen University, School of Medicine, Xiamen University, Xiamen, China.
Front Endocrinol (Lausanne). 2022 Apr 14;13:858912. doi: 10.3389/fendo.2022.858912. eCollection 2022.
Nocturnal asymptomatic hypoglycemia (NAH) is a serious complication of diabetes, but it is difficult to be detected clinically. This study was conducted to determine the largest amplitude of glycemic excursion (LAGE) to predict the episodes of NAH in outpatients with type 2 diabetes.
Data were obtained from 313 outpatients with type 2 diabetes. All subjects received continuous glucose monitoring (CGM) for consecutive 72 hours. The episodes of NAH and glycemic variability indices (glucose standard deviation [SD], mean amplitude of plasma glucose excursion [MAGE], mean blood glucose [MBG]) were accessed CGM. LAGE was calculated from self-monitoring blood glucose (SMBG).
A total of 76 people (24.3%) had NAH. Compared to patients without NAH, patients with NAH showed higher levels of glucose SD (2.4 ± 0.9 mmol/L vs 1.7 ± 0.9 mmol/L, p <0.001), MAGE (5.2 ± 2.1 mmol/L vs 3.7 ± 2.0, p<0.001) and LAGE (4.6 ± 2.3 mmol/L vs 3.8 ± 1.9 mmol/L, p=0.007), and lower level of MBG (7.5 ± 1.5 mmol/L vs 8.4 ± 2.2 mmol/L, p=0.002). LAGE was significantly associated with the incidence of NAH and time below rang (TBR) in model 1 [NAH: 1.189 (1.027-1.378), p=0.021; TBR: 0.008 (0.002-0.014), p=0.013] with adjustment for age, BMI, sex, work, hyperlipidemia, complication and medication, and in model 2 [NAH: 1.177 (1.013-1.367), p=0.033; TBR: 0.008 (0.002-0.014), p=0.012] after adjusting for diabetes duration based on model 1, as well as in model 3 [NAH: 1.244 (1.057-1.464), p=0.009; TBR: 0.009 (0.002-0.016), p=0.007] with further adjustment for HbA1c based on model 2. In addition, no significant interactions were found between LAGE and sex, age, HbA1c, duration of diabetes, BMI and insulin therapy on the risk of NAH. The receiver operator characteristic (ROC) curve shows the ideal cutoff value of LAGE for the prediction of NAH was 3.48 mmol/L with 66.7% sensitivity, 50% specificity and 0.587 (95% CI: 0.509-0.665) of area under the ROC curve.
High glycemic variability is strongly associated with the risk of NAH. The LAGE based on SMBG could be an independent predictor of NAH for outpatients with type 2 diabetes, and LAGE greater than 3.48 mmol/L could act as a warning alarm for high risk of NAH in daily life.
夜间无症状性低血糖(NAH)是糖尿病的一种严重并发症,但临床上很难检测到。本研究旨在确定最大血糖波动幅度(LAGE)以预测 2 型糖尿病门诊患者的 NAH 发作。
从 313 例 2 型糖尿病门诊患者中获取数据。所有患者均接受连续 72 小时的连续血糖监测(CGM)。通过 CGM 评估 NAH 和血糖变异性指数(血糖标准差[SD]、平均血糖波动幅度[MAGE]、平均血糖[MBG])。LAGE 由自我监测血糖(SMBG)计算得出。
共有 76 人(24.3%)发生 NAH。与无 NAH 患者相比,有 NAH 患者的血糖 SD(2.4±0.9mmol/L 比 1.7±0.9mmol/L,p<0.001)、MAGE(5.2±2.1mmol/L 比 3.7±2.0mmol/L,p<0.001)和 LAGE(4.6±2.3mmol/L 比 3.8±1.9mmol/L,p=0.007)水平更高,MBG(7.5±1.5mmol/L 比 8.4±2.2mmol/L,p=0.002)水平更低。LAGE 在模型 1 中与 NAH 发生率和血糖低于目标范围时间(TBR)显著相关[NAH:1.189(1.027-1.378),p=0.021;TBR:0.008(0.002-0.014),p=0.013],并在模型 2 中进行调整[NAH:1.177(1.013-1.367),p=0.033;TBR:0.008(0.002-0.014),p=0.012],并在基于模型 1 调整糖尿病病程后,在模型 3 中进行进一步调整[NAH:1.244(1.057-1.464),p=0.009;TBR:0.009(0.002-0.016),p=0.007],并基于模型 2 进一步调整 HbA1c。此外,在 NAH 风险方面,LAGE 与性别、年龄、HbA1c、糖尿病病程、BMI 和胰岛素治疗之间没有显著的交互作用。受试者工作特征(ROC)曲线显示,LAGE 预测 NAH 的理想截断值为 3.48mmol/L,其灵敏度为 66.7%,特异性为 50%,ROC 曲线下面积为 0.587(95%CI:0.509-0.665)。
高血糖变异性与 NAH 风险密切相关。基于 SMBG 的 LAGE 可能是 2 型糖尿病门诊患者 NAH 的独立预测因子,LAGE 大于 3.48mmol/L 可能是日常生活中 NAH 高危的预警信号。