Wang Hui, Wu Sensen, Pan Dikang, Ning Yachan, Wang Cong, Guo Jianming, Gu Yongquan
Department of Vascular Surgery, Xuanwu Hospital, Capital Medical University, Beijing, China.
Department of Intensive Care Medicine, Xuanwu Hospital, Capital Medical University, Beijing, China.
Front Public Health. 2025 Jan 13;12:1447366. doi: 10.3389/fpubh.2024.1447366. eCollection 2024.
Changes in cognitive function are commonly associated with aging in patients with cardiovascular diseases. The objective of this research was to construct and validate a nomogram-based predictive model for the identification of cognitive impairment in older people suffering from cardiovascular diseases.
This retrospective study included 498 participants with cardiovascular diseases aged >60 selected from the NHANES 2011-2014. The study employed the Minor Absolute Shrinkage and Selection Operator (LASSO) regression model, in conjunction with multivariate logistic regression analysis, to identify relevant variables and develop a predictive model. We used statistical techniques as in the Minor Absolute Shrinkage (MAS) and the Selection Operator (LASSO) regression model, in conjunction with multivariate logistic regression analysis, to identify variables that were significantly predictive of the outcome. After which, based on the selected relevant variables, we developed a machine learning model that was predictive of cognitive impairment such as Alzheimer's diseases in the older people. The effectiveness of the resultant nomogram was evaluated by assessing its discriminative capability, calibration, and conducting decision curve analysis (DCA). The constructed predictive nomogram included age, race, educational attainment, poverty income ratio, and presence of sleep disorder as variables. The model demonstrated robust discriminative capability, achieving an area under the receiver-operating characteristic curve of 0.756, and exhibited precise calibration. Consistent performance was confirmed through 10-fold cross-validation, and DCA deemed the nomogram clinically valuable.
We constructed a NHANES cardiovascular-based nomogram predictive model of cognitive impairment. The model exhibited robust discriminative ability and validity, offering a scientific framework for community healthcare providers to assess and detect the risk of cognitive decline in these patients prematurely.
认知功能的变化通常与心血管疾病患者的衰老有关。本研究的目的是构建并验证一种基于列线图的预测模型，用于识别患有心血管疾病的老年人的认知障碍。
这项回顾性研究纳入了从2011 - 2014年美国国家健康与营养检查调查（NHANES）中选取的498名年龄大于60岁的心血管疾病患者。该研究采用最小绝对收缩与选择算子（LASSO）回归模型，并结合多因素逻辑回归分析，以确定相关变量并建立预测模型。我们使用了如最小绝对收缩（MAS）和选择算子（LASSO）回归模型中的统计技术，并结合多因素逻辑回归分析，来识别对结果有显著预测作用的变量。之后，基于所选的相关变量，我们开发了一种机器学习模型，用于预测老年人中的认知障碍，如阿尔茨海默病。通过评估其判别能力、校准情况并进行决策曲线分析（DCA）来评价所得列线图的有效性。构建的预测列线图包括年龄、种族、教育程度、贫困收入比和睡眠障碍的存在情况作为变量。该模型显示出强大的判别能力，受试者工作特征曲线下面积达到0.756，并且表现出精确的校准。通过10倍交叉验证证实了一致的性能，并且DCA认为该列线图具有临床价值。
我们构建了一个基于NHANES心血管疾病的认知障碍列线图预测模型。该模型表现出强大的判别能力和有效性，为社区医疗服务提供者评估和检测这些患者过早出现认知衰退的风险提供了一个科学框架。
