【摘要】 目的 探讨老年人内在能力(intrinsic capacity, IC)降低的危险因素，构建并验证老年人IC降低风险预测模型。方法 于2020年1月—2022年5月在新疆维吾尔自治区社区、以流行病学调查收集到的1 072名老年人(年龄≥60岁)为研究对象，随机选择其中750名(70%)为建模集，322名(30%)为验证集。利用logistic回归分析探讨老年人IC降低相关因素，运用R4.2.0软件建立风险预测模型，并进行内部验证，采用C-index指数、ROC曲线下面积、校准曲线评价模型的预测效果。结果 多因素logistic回归分析结果显示：高握力(OR=0.958)、6 m步行时间(OR=0.492)及教育程度(OR=0.460)是其保护因素。在此基础上，成功构建老年人IC降低风险的列线图预测模型。建模集ROC曲线下面积为0.742,最佳临界值为2.099,灵敏度为0.519%,特异度为0.858%。验证集ROC曲线下面积为0.734,最佳临界值为0.100,灵敏度为0.678%,特异度为0.700%。校正曲线分析显示模型的区分度和校准度良好，C-index指数为0.742。结论 本研究成功建立并验证具有良好预测精准度的老年人IC降低风险预测模型，可用于临床医护工作者早期筛查IC下降的高风险老年人群。更多还原

【Abstract】 Objective To explore the risk factors of intrinsic capacity(IC) decline in older adults, and to establish and verify the risk prediction model for the reduction of IC in older adults.Methods From January 2020 to May 2022, a total of 1,072 elderly people(aged≥60 years) collected through epidemiological surveys in communities in Xinjiang Uygur Autonomous Region were selected as the research subjects, 750(70%) of which were randomly selected as the modeling case and 322(30%) as the verification case. The related factors of IC decline in the elderly were explored by logistic regression analysis, and R 4.2.0 software was used to establish a nomogram model for risk prediction. Internal validation of the model was performed using the enhanced Bootstrap calculation, and the established validation group was used for external validation. C-index, the area under the ROC curve, and calibration curve were used to evaluate the predictive effect of the model. Results The results of multivariate logistic regression analysis showed that high handgrip strength(OR=0.958), 6-meter walking time(OR=0.492) and education level(OR=0.460) were protective factors. On this basis, a nomogram of prediction model for the redution of IC in old adults was successfully constructed.The area under the ROC curve for the modeling set is 0.742, with the optimal threshold at 2.099, sensitivity at 0.519%, and specificity at 0.858%. The area under the ROC curve for the validation set is 0.734, with the optimal threshold at 0.100, sensitivity at 0.678%, and specificity at 0.700%. Calibration curve analysis shows that the model has good discrimination and calibration, with a C-index of 0.742.Conclusion The risk reduction prediction model of IC with good prediction accuracy in older adults issuccessfully established and verified in this study, which can be used for clinical care workers to early screen the high-risk groups of the intrinsic capacity decline in older adults.更多还原