摘要：

目的: 探讨以甘油三酯-葡萄糖（TyG）指数替代洛桑急性脑卒中登记分析（ASTRAL）评分中血糖指标的ASTRAL-TyG模型（即改良洛桑评分）对缺血性卒中患者预后的预测价值。方法: 基于郑州大学第一附属医院缺血性卒中数据库前瞻性纳入2019年1月至2021年12月共3393例缺血性卒中患者，入院24 h内测定空腹血糖和甘油三酯以计算TyG指数并进行ASTRAL评分，将TyG指数替代ASTRAL评分中血糖指标构建ASTRAL-TyG模型；发病后1年采用改良Rankin量表（mRS）评估功能预后。采用单因素和多因素Logistic回归分析筛查预后不良的影响因素，绘制受试者工作特征（ROC）曲线并计算曲线下面积评估ASTRAL-TyG模型的预测效能。结果: 根据mRS评分分为预后良好（≤ 2分）组（2961例）和预后不良（> 2分）组（432例）。Logistic回归分析显示，入院24 h内ASTRAL评分增加（OR =1.156，95%CI：1.133 ~ 1.179；P = 0.000）和TyG指数增加（OR = 1.731，95%CI：1.489 ~ 2.013；P = 0.000）是缺血性卒中患者预后不良的危险因素（模型1，TyG指数以连续变量纳入）；入院24 h内ASTRAL评分增加（OR = 1.156，95%CI：1.133 ~ 1.179；P = 0.000）、TyG指数8.69 ~ 9.08（OR = 1.445，95%CI：1.025 ~ 2.039；P =0.036）、TyG指数≥ 9.55（OR = 2.103，95%CI：1.532 ~ 2.888；P = 0.000）是缺血性卒中患者预后不良的危险因素（模型2，TyG指数以分类变量纳入）。ROC曲线显示，ASTRAL评分和ASTRAL-TyG模型预测功能预后不良的曲线下面积分别为0.754（95%CI：0.728 ~ 0.779，P = 0.000）和0.768（95%CI：0.743~0.794，P =0.000），灵敏度分别为65.05% 和65.97%、特异度为74.23% 和78.01%；ASTRAL-TyG模型的预测效能高于ASTRAL评分（Z = 2.084，P = 0.037）。结论: 以TyG指数替代ASTRAL评分中血糖指标拟合出的ASTRAL-TyG模型可以提高对缺血性卒中患者预后不良的预测价值。

关键词: 缺血性卒中, 甘油三酯类, 葡萄糖, 预后, 危险因素, Logistic模型

Abstract:

Objective: To investigate the predictive value of a modified Acute Stroke Registry and Analysis of Lausanne (ASTRAL) score, in which the triglyceride glucose (TyG) index replaces the glucose parameter of ASTRAL score, for poor prognosis in ischemic stroke patients. Methods: This study was based on the ischemic stroke database of The First Affiliated Hospital of Zhengzhou University. A total of 3393 ischemic stroke patients from January 2019 to December 2021 were enrolled. Baseline information was collected, and the TyG index and ASTRAL score were calculated. Replace the glucose parameter of ASTRAL score with the TyG index to construct ASTRAL-TyG model. Modified Rankin Scale (mRS) score was used to evaluate the functional outcome. Univariate and multivariate Logistic regression analyses were performed to screen for influencing factors of poor prognosis. Receiver operating characteristic (ROC) curve was plotted, and the area under the curve (AUC), net reclassification improvement (NRI), and integrated discrimination improvement (IDI) were calculated to evaluate the predictive value of the ASTRAL - TyG model. Results: Based on the mRS score, patients were stratified into a good prognosis group (mRS ≤ 2, n = 2961) and a poor prognosis group (mRS > 2, n = 432). Logistic regression analysis showed that both increased ASTRAL score witnin 24 h of admission (OR = 1.156, 95%CI: 1.133-1.179; P = 0.000) and increased TyG index (OR = 1.731, 95%CI: 1.489-2.013; P = 0.000) were risk factors for poor prognosis (Model 1, TyG index was included as a continuous variable). Increased ASTRAL score within 24 h of admission (OR = 1.156, 95%CI: 1.133-1.179; P = 0.000), TyG index of 8.69-9.08 (OR = 1.445, 95%CI: 1.025-2.039; P = 0.036), and TyG index ≥ 9.55 (OR = 2.103, 95%CI: 1.532-2.888; P = 0.000) were risk factors for poor prognosis (Model 2, TyG index was included as a categorical variable). The ROC curve demonstrated that the AUC for predicting one year poor prognosis was 0.754 (95%CI: 0.728-0.779, P = 0.000) for the ASTRAL score and 0.768 (95%CI: 0.743-0.794, P = 0.000) for the ASTRAL-TyG model. The sensitivity was 65.05% and 65.97%, while the specificity was 74.23% and 78.01%, respectively. The predictive value of the ASTRAL - TyG model was significantly higher than that of the ASTRAL score (Z = 2.084, P = 0.037). Conclusions: The ASTRAL score modified by the TyG index (ASTRAL - TyG model) improved the predictive value for poor prognosis in ischemic stroke patients.

Key words: Ischemic stroke, Triglycerides, Glucose, Prognosis, Risk factors, Logistic models