目的:探讨高敏C反应蛋白/白蛋白比值(hypersensitive C-reactive protein to albumin ratio, HAR)与卒中后认知损害(post-stroke cognitive impairment, PSCI)的相关性及其预测价值。方法:回顾性纳入2016年1月至2020年1月期间广州医科大学附属脑科医院神经内科收治的缺血性卒中患者。在入院次日清晨抽取空腹静脉血检测HAR。在卒中发病后3个月时采用蒙特利尔认知功能量表(Montreal Cognitive Assessment, MoCA)评估认知功能，MoCA总分<22分定义为PSCI。比较PSCI组与非PSCI组基线临床资料，采用多变量n logistic回归分析确定PSCI的独立危险因素。通过受试者工作特征(receiver operator characteristic, ROC)曲线分析和列线图评估HAR对PSCI的预测价值。n 结果:共纳入208例患者，年龄(68.79±9.96)岁，男性116例(55.8%)。HAR中位数为0.182×10n -3，88例(42.3%)患者存在PSCI。PSCI组年龄(n P＝0.001)、高半胱氨酸水平(n P＝0.007)、HAR(n P＝0.001)以及高血压(n P＝0.001)和糖尿病(n P＝0.018)的构成比均显著高于非PSCI组。多变量n logistic回归分析显示，在校正混杂因素后，以HAR第1四分位数组为参考，HAR第4四分位数组是缺血性卒中发病后3个月时PSCI的独立危险因素(优势比8.805，95%可信区间3.215～24.117；n P＝0.001)。ROC曲线分析显示，HAR预测PSCI的曲线下面积为0.734(95%可信区间0.668～0.793；n P<0.001)，最佳截断值为0.284×10n -3，对应的敏感性和特异性分别为60.23%和84.17%。列线图分析同样提示HAR是PSCI的独立预测因素(一致性指数0.802，95%可信区间0.749～0.853；n P<0.001)。n 结论:基线HAR较高可用于预测缺血性卒中患者的PSCI风险。“,”Objective:To investigate correlation between high sensitivity C-reactive protein/albumin ratio (HAR) and post-stroke cognitive impairment (PSCI) in patients with ischemic stroke and its predictive value.Methods:Patients with ischemic stroke admitted to the Department of Neurology, the Affiliated Brain Hospital of Guangzhou Medical University from January 2016 to January 2020 were enrolled retrospectively. Fasting venous blood was drawn for HAR detection in the morning the day after admission. The Montreal Cognitive Assessment (MoCA) was used to assess cognitive function at 3 months after the onset of stroke. A total MoCA score of <22 was defined as PSCI. The baseline clinical data of the PSCI group and the non-PSCI group were compared, and multivariate n logistic regression analysis was used to identify the independent risk factors for PSCI. The predictive value of HAR for PSCI was evaluated through Receiver Operator Characteristic (ROC) curve analysis and Nomogram.n Results:A total of 208 patients were enrolled in the study, their age was 68.79±9.96 years, and 116 were male (55.8%). The median HAR was 0.182×10n -3. Eighty-eight patients (42.3%) had PSCI. The age (n P=0.001), homocysteine level (n P=0.007), HAR (n P=0.001) and the proportion of patients with hypertension (n P=0.001) and diabetes (n P=0.018) in the PSCI group were significantly higher than those of the non-PSCI group. Multivariate n logistic regression analysis showed that after adjusting for confounding factors, the first quartile of HAR was used as a reference, and the fourth quartile of HAR was an independent risk factor for PSCI at 3 months after the onset of ischemic stroke (odds ratio 8.805, 95% confidence interval 3.215-24.117; n P=0.001). ROC curve analysis showed that the area under the curve of HAR for predicting PSCI was 0.734 (95% confidence interval 0.668-0.793; n P<0.001), the best cut-off value was 0.284×10n -3, and the corresponding sensitivity and specificity were 60.23% and 84.17%, respectively. The Nomogram analysis also suggested that HAR was an independent predictor of PSCI (concordance index 0.802, 95% confidence interval 0.749-0.853; n P<0.001).n Conclusion:The higher baseline HAR can be used to predict the risk of PSCI in patients with ischemic stroke.