目的 构建持续不卧床腹膜透析(continuous ambulatory peritoneal dialysis，CAPD)患者蛋白质能量消耗(protein-energy wasting，PEW)的预测模型并进行验证。方法 选取2025年5月～2025年7月天津中医药大学第一附属医院的382例CAPD患者作为研究对象，收集相关资料。按照7∶3的比例随机分为训练集(267例)和验证集(115例)。将训练集分为PEW组与非PEW组，进行单因素分析和LASSO回归筛选预测变量。再通过Logistic回归构建CAPD患者PEW的列线图模型，采用受试者工作特征曲线下面积评估区分度，通过Hosmer-Lemeshow检验评价校准度。 结果 单因素分析显示训练集患者年龄(Z=-3.565，P＜0.001)、性别(χ²=4.157，P=0.041)、体质量指数(Z=-6.446，P＜0.001)、工作状态(χ²=5.989，P=0.014)、握力(Z=-4.578，P＜0.001)、社会支持(χ²=8.246，P=0.016)、抑郁(χ²=38.272，P＜0.001)、血红蛋白(Z=-2.790，P=0.005)、白蛋白(Z=-5.223，P＜0.001)、三酰甘油(Z=-2.222，P=0.026)、肌酐(Z=-2.983，P=0.003)、血钾(Z=-1.981，P=0.048)为CAPD患者PEW相关的潜在影响因素。LASSO回归进一步精确筛选出体质量指数(-0.125)、握力(-0.006)、社会支持水平(-0.033)、抑郁水平(0.665)和白蛋白(-0.088)共5个预测CAPD患者PEW的关键变量。训练集和验证集的受试者工作特征曲线下面积分别为0.852(95% CI：0.805～0.899)和0.875(95% CI：0.813～0.937)，Hosmer-Lemeshow检验显示拟合良好(训练集χ2=10.224，P=0.250，验证集χ2=6.961，P=0.541)。 结论 本研究构建的列线图模型具有良好的预测效能，有助于临床早期识别PEW高危患者，为个体化营养干预提供决策依据。

Objective To construct and validate a predictive model for protein-energy wasting (PEW) in patients undergoing continuous ambulatory peritoneal dialysis (CAPD).  Methods  A total of 382 CAPD patients treated at the First Teaching Hospital of Tianjin University of Traditional Chinese Medicine from May 2025 to July 2025 were enrolled as study subjects, and relevant data were collected. The patients were randomly divided into a training set (267 cases) and a validation set (115 cases) at a 7:3 ratio. The training set was further divided into a PEW group and a non-PEW group. Univariate analysis and least absolute shrinkage and selection operator (LASSO) regression were performed to screen for predictive variables. A nomogram model for predicting PEW in CAPD patients was then constructed using Logistic regression. The discriminative ability of the model was assessed using the area under the receiver operating characteristic curve (AUC), and calibration was evaluated using the Hosmer-Lemeshow test.  Results   Univariate analysis revealed that age (Z=    -3.565, P<0.001), sex (χ²=4.157, P=0.041), body mass index (BMI) (Z=-6.446, P<0.001), employment status (χ²=5.989, P=0.014), handgrip strength (Z=-4.578, P<0.001), social support (χ²=8.246, P=0.016), depression (χ²=38.272, P<0.001), hemoglobin (Z=-2.790, P=0.005), albumin  (Z=-5.223, P<0.001), triglycerides  (Z=-2.222, P=0.026), creatinine (Z=-2.983, P=0.003), and serum potassium (Z=-1.981, P=0.048) were potential factors associated with PEW in CAPD patients. LASSO regression further refined the selection to five key predictive variables for PEW: BMI (coefficient: -0.125), handgrip strength (-0.006), social support level(-0.033), depression level (0.665), and albumin (-0.088). The AUC values for the training set and validation set were 0.852 (95% CI: 0.805～0.899) and 0.875 (95% CI: 0.813～0.937), respectively. The Hosmer-Lemeshow test indicated good calibration (training set: χ²=10.224, P=0.250; validation set: χ²=6.961, P=0.541).  Conclusion  The nomogram model constructed in this study demonstrates good predictive performance, which may facilitate early clinical identification of high risk PEW patients and provide a basis for decision making in individualized nutritional intervention.