目的 采用全球领导人营养不良倡议(GLIM)标准调查住院放疗患者的营养状况,并构建放疗后营养恶化的预测模 型。 方法 选择 2022 年 1 月至 2022 年 12 月在中国人民解放军空军特色医学中心放射治疗科放疗的 97 例住院肿瘤患者,于放 疗前行 GLIM 营养不良评估,后依据放疗后体重是否丢失超过 5%,分为营养稳定组(n = 57)和营养恶化组(n = 40)。 利用单因 素和多因素 Logistic 回归分析放疗后患者营养恶化的危险因素,并基于危险因素构建列线图预测模型,随后应用受试者操作 特征(ROC)曲线下面积(AUC)、拟合优度检验及决策曲线(DCA)对模型的区分度、精准度及应用价值进行验证。 结果 纳入 观察的 97 例放疗患者中 38 例存在营养风险或中/ 重度营养不良(39. 1%),放疗后 40 例患者出现营养恶化(41. 2%)。 分析营 养稳定与营养恶化组患者的基线情况,年龄、性别、学历、肿瘤部位、基础疾病、是否存在远处转移及入院 BMI 均无差异。 Logistic 回归分析示放疗后营养恶化的独立危险因素是:GLIM 中/ 重度营养不良、放疗次数及前白蛋白(P<0. 05)。 将上述 3 项 危险因素引入 R 软件建立列线图模型,AUC 为 0. 723(95%CI = 0. 622 ~ 0. 825),效准曲线和理想曲线均较好。 DCA 提示模型 具备良好的应用价值。 结论 GLIM 标准是预测肿瘤患者放疗后营养状态恶化的有效工具。 基于 GLIM 评估、放疗次数及前白 蛋白构建的列线图可有效预测放疗后患者的营养恶化风险,有望为医务人员采取积极的营养干预提供参考和指导。

Objective To investigate the nutritional status of cancer patients undergoing radiotherapy using the Global Leadership Initiative on Malnutrition GLIM criteria and establish a predictive model for post-radiotherapy nutritional deterioration. Method A total of 97 hospitalized cancer patients who received radiotherapy at the Fourth Military Medical University of the Chinese People′s Liberation Army from January 2022 to December 2022 were selected. The GLIM criteria for malnutrition were applied before radiotherapy and patients were categorized into stable nutrition n = 57 or deteriorated nutrition groups n = 40 based on whether they experienced more than a 5% body weight loss after radiotherapy. Univariate and multivariate Logistic regression analysis was conducted to identify factors influencing nutritional deterioration during hospitalization and a nomogram prediction model was developed based on these factors. Discrimination AUC Hosmer-Lemeshow test and decision curve analysis DCA were used to assess the model′s accuracy and clinical utility. Result Among the 97 patients 38 39. 1% had nutritional risk or malnutrition while 40 41. 2% experienced nutritional deterioration following radiotherapy. There were no significant differences in age gender education level tumor location underlying diseases distant metastasis or BMI at admission between the two groups. Univariate and multivariate logistic regression analyses revealed that GLIM malnutrition status radiotherapy frequency and prealbumin levels independently predicted nutritional deterioration after radiotherapy P < 0. 05 . The nomogram model demonstrated an AUC value of 0. 723 95% CI = 0. 622-0. 825 with good calibration curves observed for both validation data sets as well as ideal curves in terms of discrimination ability. Conclusion The GLIM criteria is an effective tool for predicting the deterioration of nutritional status in tumor patients after radiotherapy. The nomogram model constructed based on the GLIM evaluation radiotherapy frequency and prealbumin levels can effectively predict the risk of nutritional deterioration in patients after radiotherapy. This provides valuable guidance and reference for medical staff in the radiotherapy department to implement proactive preventive nursing interventions.