2型糖尿病并发认知功能障碍的危险因素分析及Nomogram模型构建

Establishment of nomogram model to predict the risks for cognitive dysfunction in type 2 diabetes mellitus patients

目的:探讨2型糖尿病(T2DM)并发认知功能障碍的危险因素并构建Nomogram预测模型。方法:回顾性分析海安市人民医院2021年1月~2022年12月治疗的193例T2DM患者的临床资料,根据有无并发认知功能障碍将其分为并发组和未并发组。采用多因素Logistic回归分析T2DM患者并发认知功能障碍的影响因素,采用R 3.5.3软件构建风险预测Nomogram模型,并评估模型区分度和预测效能。结果:T2DM患者认知功能障碍发生率为38.86%(75/193);年龄≥65岁、病程≥10年、糖尿病视网膜病变、糖化血红蛋白(HbA1c)≥8%、超敏-C反应蛋白(hs-CRP)≥10 mg/L均是T2DM患者并发认知功能障碍的危险因素(P<0.05);基于5项危险因素构建风险预测Nomogram模型,Bootstrap自抽样验证的一致性指数为0.834,校准曲线图显示校准曲线与标准曲线贴合较好,受试者工作特征曲线(ROC)显示,Nomogram模型预测的灵敏度、特异度、曲线下面积分别为82.67%、86.44%、0.859。结论:年龄≥65岁、病程≥10年、糖尿病视网膜病变、HbA1c≥8%、hs-CRP≥10 mg/L均是T2DM患者并发认知功能障碍的独立危险因素,基于以上5项危险因素构建的风险预测Nomogram模型具有良好一致性和预测效能。

Objective:To investigate the risk factors of cognitive dysfunction in type 2 diabetes mellitus(T2DM)and develop a nomogram model.Methods:The clinical data were retrospectively analyzed in 193 T2DM patients treated in our hospital between January 2021 and December 2022.The patients were divided into complicated group and non-complicated group based on cognitive dysfunction.Multivariate logistic regression analysis was used to analyze the influencing factors of cognitive dysfunction in T2DM patients.A risk prediction nomogram model was then developed using R 3.5.3 software,and its discrimination and predictive accuracy was assessed.Results:The incidence rate of cognitive dysfunction in T2DM patients was 38.86%(75/193).Age≥65 years,disease duration≥10 years,diabetic retinopathy,glycosylated hemoglobin(HbA1c)≥8%,and high sensitivity-C reactive protein(hs-CRP)≥10 mg/L were independent risk factors for cognitive dysfunction in T2DM patients(P<0.05),and a risk prediction nomogram model was developed based on the above five risk factors.The consistency index obtained from Bootstrap self-sampling validation was 0.834.The calibration curve showed that the calibration curve fitted well with the standard curve.The receiver operating characteristic curve showed that the sensitivity,specificity,and area under the curve predicted by the nomogram model were 82.67%,86.44%,and 0.859,respectively.Conclusion:Age≥65 years,disease duration≥10 years,diabetic retinopathy,HbA1c≥8%,and hs-CRP≥10 mg/L are independent risk factors for cognitive dysfunction in T2DM patients.The risk prediction nomogram model based on these factors has good consistency and prediction efficiency.