骨关节炎患者跌倒危险因素分析和列线图模型的建立与验证

Analysis of Fall Risk Factors and Development and Validation of a Nomogram Model in Patients with Osteoarthritis

摘要:

目的：骨关节炎（OA）患者由于关节功能受限、疼痛及肌力下降等因素，跌倒风险显著高于普通人群，可能导致骨折、软组织损伤等严重并发症，增加医疗负担。识别跌倒的危险因素并开发可靠的预测工具对于改善患者管理并预防不良后果至关重要。 方法：本研究基于中国健康与退休纵向研究（CHARLS 2015）数据，纳入年龄、性别、体重、视听功能、疼痛、睡眠、抑郁等多维变量。采用LASSO回归筛选变量，通过多因素logistic回归分析确定独立预测因子，构建列线图模型，并通过ROC曲线、校准曲线和决策曲线分析（DCA）评估模型性能。 结果：共纳入1,980例骨关节炎患者，其中477例（24.1%）报告跌倒。多因素分析显示，听力障碍、工具性日常生活活动（IADL）功能障碍、抑郁、肝脏疾病和肾脏疾病是跌倒的独立预测因素。列线图模型在训练集和测试集中的AUC分别为0.675（95%CI: 0.642–0.707）和0.633（95%CI: 0.579–0.687），校准曲线显示良好的一致性，DCA表明模型具有临床实用性。 结论：本研究建立的列线图模型能够有效预测骨关节炎患者的跌倒风险，具有较强的临床适用性和预测性能，可为临床提供个体化风险评估工具，辅助制定干预策略。

Abstract:

Objective:?Patients with osteoarthritis (OA) face a significantly higher risk of falls compared to the general population due to factors such as limited joint function, pain, and reduced muscle strength. Falls may lead to serious complications such as fractures and soft tissue injuries, increasing the healthcare burden. Identifying risk factors for falls and developing reliable prediction tools are crucial for improving patient management and preventing adverse outcomes. Methods:This study utilized data from the China Health and Retirement Longitudinal Study (CHARLS 2015), incorporating multidimensional variables including age, sex, body weight, visual and auditory function, pain, sleep, and depression. Variables were screened using LASSO regression, and independent predictors were identified through multivariate logistic regression analysis to construct a nomogram model. The model's performance was evaluated using ROC curves, calibration curves, and decision curve analysis (DCA). Results:A total of 1,980 OA patients were included, among whom 477 (24.1%) reported falls. Multivariate analysis revealed that hearing impairment, instrumental activities of daily living (IADL) dysfunction, depression, liver disease, and kidney disease were independent predictors of falls. The nomogram model achieved AUC values of 0.675 (95% CI: 0.642–0.707) in the training set and 0.633 (95% CI: 0.579–0.687) in the test set. Calibration curves demonstrated good consistency, and DCA indicated the model’s clinical utility. Conclusion:?The established nomogram model effectively predicts the risk of falls in OA patients, demonstrating strong clinical applicability and predictive performance. It can serve as a personalized risk assessment tool to support the development of intervention strategies in clinical practice.