Harvengt Antoine, Bastin Marie, Toussaint Cédric, Beckers Maude, Helleputte Thibault, Lysy Philippe
Pôle EDIN, Institut de Recherche Expérimentale et Clinique, UCLouvain, 1200 Brussels, Belgium.
DNAlytics, 1348 Ottignies-Louvain-la-Neuve, Belgium.
Nutrients. 2025 Aug 12;17(16):2610. doi: 10.3390/nu17162610.
Severe hypoglycemia (SH) is a critical complication in children and adolescents with type 1 diabetes (T1D), associated with cognitive impairment, coma, and significant psychosocial burden. Despite advances in glucose monitoring, predicting SH remains challenging, as most models focus on milder hypoglycemic events. To develop a machine learning model for early prediction of SH using continuous glucose monitoring (CGM) data in children and adolescent T1D patients. This retrospective study analyzed CGM data from 67 patients (37 SH episodes, 1430 non-SH segments). Glycemic curves were segmented into 5-day windows, and 21 features were extracted, including glycemic mean, variability, time below range (TBR < 60 mg/dL), and PCA components of glucose trends. A support vector machine (SVM) model was trained using repeated cross-validation to predict SH 15 min before onset. Model performance was evaluated using sensitivity, specificity, balanced classification rate (BCR), and area under the ROC curve (AUC). The model achieved robust performance, with a median AUC of 90% (IQR: 87-93%) and median BCR of 84% (IQR: 80-89%). Sensitivity and specificity exceeded 80%, demonstrating reliable detection of impending SH. However, the positive predictive value (PPV) was low (12%), with false alarms frequently triggered during descending glucose trends or near-hypoglycemic values (end glucose <54 mg/dL). SH episodes were stratified into two subgroups: group 1 (<45 mg/dL, = 26) and group 2 (>52 mg/dL, = 15). Notably, false alarms occurred at a median interval of 25 days, minimizing alarm fatigue. These findings confirm the feasibility of SH prediction in clinical practice, prioritizing high-risk events over milder hypoglycemia. By alerting patients and medical teams early on, this tool could facilitate individualized treatment adjustments, reduce the risk of serious hypoglycemic events, and thus contribute to more personalized management of pediatric diabetes, while improving patients' quality of life.
严重低血糖（SH）是1型糖尿病（T1D）儿童和青少年的一种关键并发症，与认知障碍、昏迷及显著的社会心理负担相关。尽管血糖监测取得了进展，但预测SH仍然具有挑战性，因为大多数模型关注的是较轻的低血糖事件。本研究旨在利用儿童和青少年T1D患者的连续血糖监测（CGM）数据开发一种用于早期预测SH的机器学习模型。这项回顾性研究分析了67例患者的CGM数据（37次SH发作，1430个非SH时段）。血糖曲线被分割为5天的窗口，并提取了21个特征，包括血糖均值、变异性、低于范围的时间（TBR<60mg/dL）以及葡萄糖趋势的主成分分析成分。使用重复交叉验证训练支持向量机（SVM）模型，以在SH发作前15分钟进行预测。使用灵敏度、特异性、平衡分类率（BCR）和ROC曲线下面积（AUC）评估模型性能。该模型表现出稳健的性能，AUC中位数为90%（IQR：87 - 93%），BCR中位数为84%（IQR：80 - 89%）。灵敏度和特异性超过80%，表明对即将发生的SH具有可靠的检测能力。然而，阳性预测值（PPV）较低（12%），在血糖下降趋势或接近低血糖值（末梢血糖<54mg/dL）时经常触发误报。SH发作被分为两个亚组：第1组（<45mg/dL，n = 26）和第2组（>52mg/dL，n = 15）。值得注意的是，误报的中位间隔为25天，最大限度地减少了警报疲劳。这些发现证实了在临床实践中预测SH的可行性，将高风险事件置于比轻度低血糖更优先的位置。通过尽早提醒患者和医疗团队，该工具可以促进个体化治疗调整，降低严重低血糖事件的风险，从而有助于更个性化地管理儿童糖尿病，同时提高患者的生活质量。
