| 李刚,刘红梅,贾钦尧,等.基于 LASSO回归的慢性阻塞性肺疾病老年住院病人合并肌少症风险预测模型的建立[J].安徽医药,2026,30(7):1430-1435. |
| 基于 LASSO回归的慢性阻塞性肺疾病老年住院病人合并肌少症风险预测模型的建立 |
| A LASSO regression-based nomogram model for predicting the risk of comorbid sarcopenia in elderly hospitalised COPD patients |
| |
| DOI：10.3969/j.issn.1009-6469.2026.07.031 |
| 中文关键词: 肺疾病，慢性阻塞性 肌少症 LASSO回归 多因素分析 列线图模型 |
| 英文关键词: Pulmonary disease, chronic obstructive Sarcopenia LASSO regression Multifactorial analysis Nomogram model |
| 基金项目:四川省基层卫生事业发展研究中心科研项目( SWFZ21-C-76) |
|
| 摘要点击次数: 289 |
| 全文下载次数: 105 |
| 中文摘要: |
| 目的建立基于 LASSO回归的慢性阻塞性肺疾病( chronic obstructive pulmonary disease，COPD)老年住院病人合并肌少症风险预测模型，并进行验证。方法回顾性选择 2022年 1月至 2023年 12月川北医学院附属医院武胜医院接诊的 COPD老年住院病人 584例病人进行研究。入组病人以 7∶3比例分为模型组 408例，验证组 176例。根据有无肌少症将模型病人分为两组，比较两组病人各指标，以 LASSO回归筛选潜在影响因素后行多因素 logisitc回归获得 COPD老年住院病人合并肌少症的独立性影响因素，以此建立列线图模型并进行验证。结果研究模型组共 94例( 23.04%)病人确诊为肌少症。 LASSO回归基础上行多因素 logistic回归分析结果显示，性别、年龄、身体质量指数、吸烟史、慢性阻塞性肺疾病全球倡议分级、改良版英国医学研究委员会呼吸问卷得分及血钙为 COPD老年住院病人合并肌少症的独立性影响因素( P<0.05)。用列线图形式展示多因素 logistic回归分析结果。模型组 AUC 95%CI为 0.84(0.80，0.89)；验证组为 0.81(0.76，0.86)。校准曲线结果显示:模型曲线与理想模型基本拟合成对角线。 H-L拟合优度检验显示本研究模型拟合度较好。决策曲线显示当预测概率阈值 0.05~0.75时使用本研究模型预测 COPD病人合并肌少症的净获益最高。结论该研究建立的列线图模型用于预测 COPD老年住院病人合并肌少症风险具有较高准确度与区分度。 |
| 英文摘要: |
| Objective To develop and validate a predictive model for the risk of comorbid sarcopenia in elderly hospitalised chronicobstructive pulmonary disease (COPD) patients based on LASSO regression.Methods Retrospective selection a total of 584 elderlyhospitalized patients with COPD, who were admitted to Wusheng Hospital, Affiliated Hospital of North Sichuan Medical College fromJanuary 2022 to December 2023. The enrolled patients were divided into a model group (n=408) and a validation group (n=176) in a 7:3 ratio. Patients in the model were divided into 2 groups according to the presence or absence of sarcopenia, and various indicatorswere compared between the 2 groups. LASSO regression was used to screen the potential influencing factors and then multifactorial Lo-gisitc regression was performed to obtain the independent influencing factors of sarcopenia in elderly hospitalized patients with COPD.A nomogram model was then established based on these factors and subsequently validated.Results A total of 94 (23.04%) patients inthe study model group were diagnosed with sarcopenia. The results of the multivariate logistic regression analysis based on LASSO re-gression showed that gender, age, body mass index (BMI), smoking history, Global Initiative for Chronic Obstructive Lung Disease(GOLD) grade, modified British medical research council (mMRC) score, and serum calcium were independent influencing factors forsarcopenia in elderly hospitalized COPD patients (P<0.05). The results of multifactor logistic regression analysis were presented in the form of Nomogram. The AUC 95%CI 0.84 (0.80, 0.89) for the model group and 0.81 (0.76, 0.86) for the validation group. The results of calibration curves showed that the model curves were basically fitted to the diagonal line of the ideal model. The Hosmer-Lemeshow goodness-of-fit test showed that the model in this study was well fitted. Decision curves showed the highest net benefit of using the pres-ent study model to predict comorbid sarcopenia in patients with COPD when the predictive probability threshold ranged from 0.05 to0.75.Conclusion The nomogram model developed in this study has high accuracy and discrimination for predicting the risk of comor-bid sarcopenia in elderly hospitalised patients with COPD. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
