本文選題：型糖尿病 + 風險預測模型　； 參考：《山東大學學報(醫(yī)學版)》2017年06期

【摘要】：目的構(gòu)建健康管理人群2型糖尿病3年發(fā)病風險預測模型。方法依托山東多中心健康管理縱向觀察大數(shù)據(jù)庫,選擇20~75歲的基線未患2型糖尿病者構(gòu)建隊列。采用Cox比例風險回歸構(gòu)建2型糖尿病預測模型,以受試者工作特征曲線下面積(AUC)評價模型的預測效能,以十折交叉驗證法檢驗模型的穩(wěn)定性。結(jié)果隨訪期間共新發(fā)糖尿病1 624例,男性和女性的發(fā)病密度分別為15.00‰、10.83‰。男性預測模型最終納入的變量包括年齡、體質(zhì)量指數(shù)、空腹血糖、甘油三酯、谷丙轉(zhuǎn)氨酶、白細胞計數(shù)。納入女性預測模型的變量包括年齡、空腹血糖、甘油三酯、高密度脂蛋白、谷丙轉(zhuǎn)氨酶。男性和女性預測模型的AUC分別為0.795(95%CI:0.764~0.827)和0.707(95%CI:0.654~0.759)。結(jié)論分性別建立的2型糖尿病發(fā)病風險預測模型在健康管理人群中均具有較好預測能力。
[Abstract]:Objective to establish a 3-year risk prediction model for type 2 diabetes mellitus in health management population. Methods based on the large longitudinal observation database of Shandong multi-center health management, a cohort of 20 ~ 75 years old patients without type 2 diabetes mellitus was selected to construct a cohort. The predictive model of type 2 diabetes mellitus was constructed by Cox proportional risk regression. The predictive effectiveness of the model was evaluated by the area under the operating characteristic curve, and the stability of the model was tested by a ten fold cross-validation method. Results during the follow-up period, 1 624 cases of new diabetes mellitus were found, the incidence density of male and female were 15.00 鈥，

本文編號：1828417