2型糖尿病性肾动脉狭窄危险因素的预测研究

The study of prediction model of renal artery stenosis in type 2 diabetic patients

目的:研究建立2型糖尿病(DM)患者合并肾动脉粥样硬化性狭窄(ARAS)的预测模型。方法:经肾动脉造影证实的2型糖尿病ARAS患者129例和无ARAS者122例。对两组年龄、DM病程等指标进行统计学分析。结果:最终进入模型的ARAS独立预测因子是年龄(≥65岁)(OR=42.66,P<0.01),加速性高血压(OR=24.78,P<0.01),DM病程≥5年(OR=4.76,P<0.05),颈动脉内膜中层厚度(IMT)≥1 mm(OR=3.41,P<0.01)及Ccr≥90 mL/min(OR=0.91,P<0.01)。预测ARAS的最佳概率值是0.54(敏感度86.0%,特异度93.4%)。Logistic回归方程ROC曲线下面积为0.868(95%CI=0.816～0.92,P<0.01),Cox and Snell决定系数R2=0.586。结论:年龄≥65岁、AH、DM病程≥5年、颈动脉IMT≥1 mm、Ccr≥90 mL/min是DM患者并发ARAS的独立预测因子,该预测模型可以对2型糖尿病患者进行肾动脉狭窄的风险预测,不仅有助于提高彩超及有创检查的诊断阳性率,而且可以减少不必要的漏诊以提高治疗的成本效益。

Objective：To establish prediction model of type 2 diabetes mellitus （DM） in patients with atherosclerotic renal ste- nosis （ARAS）. Methods： The renal artery angiography in type 2 diabetes patients with ARAS 129 cases and 122 cases without ARAS. The age, DM duration and other indicators for statistical analysis in two groups. Results： Independent predictors of the final prediction model inculde the age （≥65 years ） （ OR = 42. 66, P 〈 0. 01 ）, accelerated hypertension （ OR = 24. 78, P 〈 0. 01 ）, DM du- ration ≥5 years （OR=4.76 ,P〈0.05）, carotid IMT≥1 mm （OR=3.41,P〈O. O1） and Ccr ≥90 mlMmin （OR=O. 91,P〈 0.01）. The regression model is P/（1 -P） =EXP（ -2.68 ＋ 3.75Age（l〉65岁） ＋ 3.21AH-2.40Ccr ＋1.56DM（≥5年） ＋ 1.23IMT）. The best probability forecast of ARAS is 0. 54 （ sensitivity 86.0%, specificity 93.4% ）. Logistic regression equation for the area under the ROC curve was 0. 868 （ 95 % C1 0. 816-0. 92, P 〈 0. 01 ） , Cox and Snell coefficient of determination R2 = 0. 586. Conclusion ： The years/〉65, AH, DM duration ≥5 years, carotid IMT ≥ 1 mm, Ccr ≥90 mL/min are independent predictor of ARAS for DM patients. This prediction model can be carried out in patients with type 2 diabetes risk prediction of ARAS. It not only helps to improve the positive rate of color Doppler ultrasound and other invasive examinations and can reduce unnecessary missed diagnosis and improve the cost-effectiveness of treatment.