健康体检人群糖尿病发病风险的评估方法  

Risk prediction of type 2 diabetes mellitus in a health check-up population

作　　者：刘海霞[1] 胡冬梅[2] 冷松[3] 刘颖[3] 赵辉[3] 孙广慧[3] 

机构地区：[1]大连医科大学附属第二医院内分泌科,116027 [2]大连医科大学卫生统计教研室 [3]大连医科大学附属第二医院体检中心

出　　处：《中华健康管理学杂志》2012年第3期178-181,共4页Chinese Journal of Health Management

基　　金：基金项目：大连市科学技术基金计划项目（2010E15SF182）（2011J21DW018）

摘　　要：目的建立适合健康体检个体糖尿病发病风险的评估方法。方法结合国内外的研究结果分析近年糖尿病发病率资料及发病风险因素，得出我国成年人糖尿病发病的主要风险因素，将目标人群随机分为两组队列，队列1用于危险因素评分方法的建立，采用非条件logistic回归分析筛选糖尿病的危险因素，根据危险因素的回归系数确定不同变量的积分值，以累计积分值的大小判断个体患病的危险性，并以队列2的资料验证该方法进行人群筛检的价值。结果研究结果表明男性与女性分别当积分值为65．0与65．5时，约登指数最大，故将65．0与65．5作为判别不同性别发生糖尿病风险的阈值，其筛检的灵敏度分别为90．6％与83．3％，特异度分别为89．4％与97．7％，阳性预测值分别为58．8％与80．0％，阴性预测值为98．3％与98．2％。曲线下面积分别为0．955与0．899，危险因素评分法用于筛查糖尿病，差异有统计学意义（P=0．000）。结论该方法简易经济，可操作性强，有望成为健康体检机构筛查人群糖尿病发病的高危人群的实用、有效的方法。Objective To develop a simple scoring system to identify individuals with higher risk of type 2 diabetes mellitus （T2DM）. Methods Major risk factors of T2DM were fixed after literature review. Logistic regression was used to select risk factors of T2DM. Target populations were randomly assigned to 2 groups, and group A was used for the development of risk factor scoring method and group B for the confirmation of this new method. Results The threshold for T2DM risk was 65.0 for men and 65.5 for women （sensitivity 90. 6% and 83.3% ,respectively; specificity 89.4% and 97.7% ,respectively; positive predictive value 58.8% and 80. 0%, respectively; negative predictive value 98.3% and 98.2%, respectively ; area under the curve 0. 955 and 0. 899, respectively）. There was significant difference of T2DM detection between screening and risk scoring system （ P = 0. 000 ）. Conclusion Our T2DM risk scoring system may provide an effective tool to identify individuals with a higher risk of developing T2DM.

关 键 词：糖尿病 2型 危险因素 风险评分 

分 类 号：R587.1[医药卫生—内分泌]