基于美国国家健康和营养调查数据库构建糖尿病并发冠心病风险列线图预测模型  

Construction of a nomogram prediction model for the risk of coronary heart disease in patients with diabetes based on the National Health and Nutrition Examination Survey database

作　　者：李超 邹花一阳 郑宏喜 王亮[2] Li Chao;Zou Huayiyang;Zheng Hongxi;Wang Liang(Department of Cardiology,Yizheng People's Hospital,Jiangsu211401,China)

机构地区：[1]江苏省仪征市人民医院心内科,211401 [2]南京医科大学第一附属医院心血管内科 [3]天长市人民医院心内科

出　　处：《临床内科杂志》2025年第3期209-212,共4页Journal of Clinical Internal Medicine

基　　金：国家自然科学基金资助项目(82200474)。

摘　　要：目的 基于美国国家健康和营养调查(NHANES)数据库构建糖尿病并发冠心病风险列线图预测模型,为临床诊断和干预提供依据。方法 基于美国NHANES数据库中2013~2018年的数据,最终纳入1 136例糖尿病患者,根据是否合并冠心病将其分为冠心病组(326例)和非冠心病组(810例)。收集所有患者的一般临床资料并分组进行比较。采用单因素和多因素logistic回归分析评估糖尿病并发冠心病的危险因素,并构建列线图模型。采用受试工作者特征(ROC)曲线评估模型的预测效能。结果 冠心病组患者年龄、年收入、受教育年限、BMI、收缩压(SBP)、舒张压(DBP)、总胆固醇(TC)、甘油三酯(TG)、低密度脂蛋白胆固醇(LDL-C)、空腹血糖(FPG)、糖化血红蛋白(HbA1c)、尿素氮(BUN)、肌酐(Cr)、尿酸(UA)、C反应蛋白(CRP)水平均高于非冠心病组,男性、吸烟史、饮酒史患者比例及高密度脂蛋白胆固醇(HDL-C)水平均低于非冠心病组,两组患者种族构成比例比较差异有统计学意义(P<0.05)。多因素logistic回归分析结果显示,年龄、BMI、吸烟史、SBP、TC及FPG均是糖尿病患者并发冠心病的危险因素(P<0.05)。列线图预测糖尿病患者并发冠心病的ROC曲线下面积(AUC)为0.90,敏感度为92%、特异度为78%,提示列线图模型具有较好的预测效能。结论 基于美国NHANES数据库构建的糖尿病并发冠心病风险列线图预测模型有较高的预测效能,可为临床诊断和干预提供参考。Objective To construct a nomogram prediction model for the risk of diabetes patients complicated with coronary heart disease(CHD)based on the National Health and Nutrition Examination Survey(NHANES)database,and to provide a basis for clinical diagnosis and intervention.Methods Based on the data from 2013 to 2018 from NHANES database,1136 diabetic patients were finally included,and divided into CHD group(326 cases)and non-CHD group(810 cases)according to whether they were complicated CHD or not.General clinical data of all patients were collected and grouped for comparison.Univariate and multifactorial logistic regression analyses were used to assess the risk factors for diabetes mellitus complicated with CHD,and a nomogram prediction model was constructed.Recevier operating characteristics(ROC)curve as used to assess the predictive efficacy of the model.Results Age,annual income,years of education,BMI,systolic blood pressure(SBP),diastolic blood pressure(DBP),total cholesterol(TC),triglyceride(TG),low-density lipoprotein cholesterol(LDL-C),fasting glucose(FPG),glycosylated hemoglobin(HbAlc),urea nitrogen(BUN),creatinine(Cr),uric acid(UA),C-reactive protein(CRP),and uric acid(UA)in CHD group were higher than those in non-CHD group,male,the proportion of smoking and drinking histroy,and high-density lipoprotein cholesterol(HDL-C)levels were lower than those in the non-CHD group,comparison of the proportions of the racial composition of the patients in the two groups showed a statistically significant difference(P<0.05).The results of multifactorial logistic regression analysis showed that age,BMI,smoking,SBP,TC,FPG were risk factors for the complication of CHD in diabetic patients(P<0.05).The area under the ROC curve(AUC)of the nomogram for predicting concurrent CHD in patients with diabetes was 0.90,with a sensitivity of 92%and specificity of 78%,suggesting that the nomogram prediction model had a good predictive efficacy.ConclusionThe nomogram prediction model for the risk of diabetes mellitus complicating CHD co

关 键 词：国家健康和营养检查调查数据库 糖尿病 冠心病 危险因素 列线图 预测模型 

分 类 号：R587.1[医药卫生—内分泌]