目的 探讨房颤合并糖尿病患者轻度认知功能障碍（mild cognitive impairment，MCI）的影响因素，及预测模型的建立，为房颤合并糖尿病患者MCI的治疗提供指导。 方法 分析2023年1月至2024年1月在郑州大学第五附属医院心血管内科二病区就诊并确诊为房颤合并糖尿病的患者199例，采用单因素分析和多元logistic回归分析房颤合并糖尿病患者影响MCI的相关因素，根据多元logistic回归分析结果建立房颤合并糖尿病患者影响MCI的预测模型。 结果 单因素分析结果显示，年龄（P = 0.002 3）、同型半胱氨酸（P < 0.000 1）、空腹血糖（P = 0.022 5）、糖化血红蛋白（P = 0.006 6）、血尿酸（P = 0.032 2）为MCI发生的影响因素。多因素logistic回归分析：年龄（OR = 1.08，P = 0.000 4）、同型半胱氨酸（OR = 1.37，P < 0.000 1）、空腹血糖（OR = 1.22，P = 0.023 5）、糖化血红蛋白（OR = 1.61，P = 0.004 2）、血尿酸（OR = 1.29，P = 0.009 1）为MCI发生的独立影响因素。在约登指数（Youden index，YI）最大时，为最佳阈值。本研究在最佳阈值处，灵敏度为0.74，特异度为0.80，曲线下面积（AUC）=0.809，说明该模型能较好预测MCI发生。 结论 年龄、空腹血糖、血同型半胱氨酸、血尿酸、血糖化血红蛋白是房颤合并糖尿病患者MCI发生的独立危险因素。基于多因素logistic 回归构建临床预测模型对房颤合并糖尿病患者MCI发生具有一定的预测价值。

Objective To explore the influencing factors of mild cognitive impairment （MCI） in patients with atrial fibrillation and diabetes mellitus， and to establish the prediction model， so as to provide guidance for the treatment of MCI in patients with atrial fibrillation and diabetes mellitus. Methods 199 patients with atrial fibrillation and diabetes diagnosed in the second ward of Cardiovascular Department of the Fifth Affiliated Hospital of Zhengzhou University from January 2023 to January 2024 were analyzed. The related factors of MCI in patients with atrial fibrillation and diabetes mellitus were analyzed by univariate analysis and multivariate logistic regression. According to the results of multivariate logistic regression analysis， the prediction model of MCI in patients with atrial fibrillation and diabetes mellitus was established. Results Univariate analysis showed that age （P =0.002 3）， homocysteine （P < 0.000 1）， fasting blood glucose （P = 0.022 5）， glycated hemoglobin （P = 0.006 6）， and blood uric acid （P = 0.032 2） were the influencing factors of MCI. Multivariate logistic regression analysis： age （OR = 1.08， P = 0.000 4）， homocysteine （OR = 1.37， P < 0.000 1）， fasting blood glucose （OR = 1.22， P =0.023 5）， glycated hemoglobin （OR = 1.61， P = 0.004 2）， and blood uric acid （OR = 1.29， P = 0.009 1） were the independent influencing factors of MCI. The optimal threshold is when the Youden index （YI = sensitivity + specificity） is maximum. At the optimal threshold， the sensitivity was 0.74， the specificity was 0.80， and the area under the curve （AUC） was 0.809， indicating that the model can effectively predict the occurrence of MCI. Conclusion Age， fasting blood glucose， blood homocysteine， blood uric acid and glycosylated hemoglobin are independent risk factors for MCI in patients with atrial fibrillation and diabetes. The clinical prediction model based on multivariate logistic regression has a certain predictive value for the occurrence of MCI in patients with atrial fibrillation and diabetes mellitus.