慢性肾炎患者发生骨质疏松的危险因素及诊断模型的构建

Risk factors and diagnostic model establishment of suffering from osteoporosis in patients with chronic nephritis

目的分析慢性肾炎患者发生骨质疏松的危险因素,并构建诊断模型。方法回顾性分析138例慢性肾炎患者的临床资料,根据骨密度值将患者分为骨质疏松组(n=50)与非骨质疏松组(n=88)。记录两组患者的性别、年龄、体质指数、慢性肾炎病程、糖尿病史、吸烟史、激素服用时间,并检测两组患者的骨代谢和生化指标,即血磷、血钙、血清尿酸、血肌酐、骨碱性磷酸酶(BALP)、碱性磷酸酶(ALP)、全段甲状旁腺激素(iPTH)、25-羟基维生素D_(3)[25(OH)D_(3)]、估算肾小球滤过率(eGFR)。采用Logistic回归分析慢性肾炎患者发生骨质疏松的危险因素并构建诊断模型;采用受试者工作特征(ROC)曲线评价模型的诊断效能并确定诊断阈值。结果单因素分析结果显示,相比于非骨质疏松组,骨质疏松组激素服用时间≥3年的患者比例更高,BALP、iPTH水平更高,25(OH)D_(3)水平、eGFR更低(均P<0.05)。多因素Logistic回归分析结果显示,激素服用时间≥3年、BALP和iPTH水平升高是慢性肾炎患者发生骨质疏松的危险因素,25(OH)D_(3)水平、eGFR升高是慢性肾炎患者发生骨质疏松的保护因素(均P<0.05)。Logistic回归诊断模型为Logit P=-7.955+1.583×激素服用时间+0.850×BALP+0.240×iPTH-0.384×25(OH)D_(3)-0.071×eGFR,该模型的ROC曲线下面积为0.935,灵敏度为94.00%,特异度为92.05%,正确诊断率为92.75%。结论激素服用时间≥3年、BALP和iPTH水平升高的慢性肾炎患者发生骨质疏松的风险增加,25(OH)D_(3)水平、eGFR升高者发生骨质疏松的风险降低;根据上述影响因素构建的Logistic诊断模型能有效评估慢性肾炎患者发生骨质疏松的风险。

Objective To analyze the risk factors for suffering from osteoporosis in patients with chronic nephritis,and to establish a diagnostic model.Methods The clinical data of 138 patients with chronic nephritis were retrospectively analyzed,and the patients were assigned to osteoporosis group(n=50)or non-osteoporosis group(n=88)according to bone mineral density.The gender,age,body mass index,disease course of chronic nephritis,history of diabetes mellitus,smoking history,hormone administration duration were recorded in both groups;in addition,bone metabolism and biochemical indexes,namely,serum phosphorus,serum calcium,serum uric acid,serum creatinine,bone alkaline phosphatase(BALP),alkaline phosphatase(ALP),intact parathyroid hormone(iPTH),25-hydroxyvitamin D_(3)(25(OH)D_(3)),and estimated glomerular filtration rate(eGFR)were measured in the two groups.The Logistic regression was performed to analyze the risk factors for suffering from osteoporosis in patients with chronic nephritis,and to establish a diagnostic model.The receiver operating characteristic(ROC)curve was employed to evaluated the diagnostic efficiency of the model,and determine the diagnostic threshold.Results The results of univariate analysis revealed that the osteoporosis group exhibited a higher proportion of patients with hormone administration duration≥3 years,higher levels of BALP and iPTH,and lower 25(OH)D_(3) level and eGFR as compared with the non-osteoporosis group(all P<0.05).The results of multivariate Logistic regression analysis revealed that hormone administration duration≥3 years,the elevations of BALP and iPTH levels were the risk factors for suffering from osteoporosis in patients with chronic nephritis,and the elevations of 25(OH)D_(3) level and eGFR were the protective factors for suffering from osteoporosis in patients with chronic nephritis(all P<0.05).The Logistic regression diagnostic model was Logit P=-7.955+1.583×hormone administration duration+0.850×BALP+0.240×iPTH-0.384×25(OH)D_(3)-0.071×eGFR,and the area under the curve of this model was 0.935,the sensitivity was 94.00%,the specificity was 92.05%,and the diagnostic accuracy was 92.75%.Conclusion The chronic nephritis patients with hormone administration duration≥3 years,and with the elevations of BALP and iPTH levels are at a increased risk of suffering from osteoporosis,and the patients with the elevations of 25(OH)D_(3) level and eGFR are at a decreased risk of suffering from osteoporosis.The established Logistic diagnostic model based on aforesaid influencing factors can effectively evaluate the risk of suffering from osteoporosis in patients with chronic nephritis.