维持性血液透析患者肌少-骨质疏松症列线图诊断模型的构建与验证

Construction and validation of a nomogramdiagnostic model for osteosarcopenia in maintenance hemodialysis patients

目的探讨维持性血液透析患者肌少-骨质疏松症的相关危险因素,构建列线图诊断模型并验证效果。方法采用便利抽样法,选取2020年7月—2021年4月在上海市6所医院行规律血液透析的697例患者作为建模集,选取2020年11月在上海市同济医院行规律血液透析的132例患者作为验证集,收集患者的一般资料、实验室指标、人体参数、躯体功能、营养状况、体力活动、认知功能和抑郁情绪。采用Logistic回归分析探讨维持性血液透析患者肌少-骨质疏松症的危险因素并构建列线图模型,通过受试者工作特征曲线下面积、校准曲线、决策曲线检验评价模型的效果。结果共纳入建模集维持性血液透析患者697例,其中发生肌少-骨质疏松症的患者171例,肌少-骨质疏松症的发生率为24.53%(171/697)。二项Logistic回归分析结果显示,年龄、体重指数、体力活动强度、查尔森共病指数是维持性血液透析患者发生肌少-骨质疏松症的影响因素(P<0.05)。建模集、十折交叉验证和验证集的受试者工作特征曲线下面积分别为0.835、0.827、0.851;建模集和验证集的校准曲线拟合良好;决策曲线显示,列线图的临床效用较好。结论维持性血液透析患者容易发生肌少-骨质疏松症。高龄、低体重指数、高查尔森共病指数、低强度体力活动是维持性血液透析患者发生肌少-骨质疏松症的危险因素,基于上述因素构建的列线图诊断模型可以帮助医护人员早期识别高风险人群、制定防治措施。

Objective To explore the risk factors of osteosarcopenia in maintenance hemodialysis patients,construct a diagnostic nomogram model and verify the effect.Methods Usingthe convenient sampling method,a total of 697 patients who underwent regular hemodialysis in six hospitals in Shanghai from July 2020 to April 2021 were selected as the modeling set,and 132 patients who underwent regular hemodialysis in Tongji Hospital in Shanghai in November 2020 were selected as the validation set.General information,laboratory indicators,human parameters,physical functions,nutritional status,physical activity,cognitive function,and depression were collected.Logistic regression was used to analyze the risk factors of osteosarcopenia in maintenance hemodialysis patients and to construct a nomogram model.The effect of the model was evaluated by the area under the receiver operating characteristic curve,calibration curve,and decision curve.Results A total of 697 maintenance hemodialysis patients were included in the modeling set,including 171 patients with osteosarcopenia,with an incidence rate of 24.53%(171/697).The results of the binomial logistic regression analysis showed that age,body mass index,physical activity intensity,and Charlson Comorbidity Index(CCI)were the influencing factors for the occurrence of osteosarcopenia in maintenance hemodialysis patients(P<0.05).The area under the receiver operating characteristic curve in the modeling set,ten-fold cross-validation,and validation set were 0.835,0.827,and 0.851,respectively.The calibration curves of the modeling and validation sets fitted well.The decision curve showed that the clinical utility of the nomogram was good.Conclusions Maintenance hemodialysis patients are prone to osteosarcopenia.Old age,low body mass index,high Charlson Comorbidity Index,and low-intensity physical activity are risk factors for osteosarcopenia in maintenance hemodialysis patients.A nomogramdiagnostic model based on the above-mentioned influencing factors can help medical staff identify high-risk populations early and develop prevention and treatment measures.