Huang Zhi-Xin, Chen Lijuan, Chen Ping, Dai Yingyi, Lu Haike, Liang Yicheng, Ding Qingguo, Liang Piaonan
Department of Neurology, The Affiliated Guangdong Second Provincial General Hospital of Jinan University, Guangzhou, Guangdong, China.
Guangzhou University of Chinese Medicine, Guangzhou, Guangdong, China.
Front Cardiovasc Med. 2024 Jul 25;11:1392752. doi: 10.3389/fcvm.2024.1392752. eCollection 2024.
This study aimed to investigate the prevalence of carotid atherosclerosis (CAS), especially among seniors, and develop a precise risk assessment tool to facilitate screening and early intervention for high-risk individuals.
A comprehensive approach was employed, integrating traditional epidemiological methods with advanced machine learning techniques, including support vector machines, XGBoost, decision trees, random forests, and logistic regression.
Among 1,515 participants, CAS prevalence reached 57.4%, concentrated within older individuals. Positive correlations were identified with age, systolic blood pressure, a history of hypertension, male gender, and total cholesterol. High-density lipoprotein (HDL) emerged as a protective factor against CAS, with total cholesterol and HDL levels proving significant predictors.
This research illuminates the risk factors linked to CAS and introduces a validated risk scoring tool, highlighted by the logistic classifier's consistent performance during training and testing. This tool shows potential for pinpointing high-risk individuals in community health programs, streamlining screening and intervention by clinical physicians. By stressing the significance of managing cholesterol levels, especially HDL, our findings provide actionable insights for CAS prevention. Nonetheless, rigorous validation is paramount to guarantee its practicality and efficacy in real-world scenarios.
本研究旨在调查颈动脉粥样硬化（CAS）的患病率，尤其是在老年人中的患病率，并开发一种精确的风险评估工具，以促进对高危个体的筛查和早期干预。
采用了一种综合方法，将传统流行病学方法与先进的机器学习技术相结合，包括支持向量机、XGBoost、决策树、随机森林和逻辑回归。
在1515名参与者中，CAS患病率达到57.4%，集中在老年人中。发现与年龄、收缩压、高血压病史、男性性别和总胆固醇呈正相关。高密度脂蛋白（HDL）成为预防CAS的保护因素，总胆固醇和HDL水平被证明是重要的预测指标。
本研究阐明了与CAS相关的危险因素，并引入了一种经过验证的风险评分工具，逻辑分类器在训练和测试期间的一致表现突出了该工具。该工具在社区健康项目中识别高危个体方面显示出潜力，简化了临床医生的筛查和干预工作。通过强调管理胆固醇水平，尤其是HDL的重要性，我们的研究结果为预防CAS提供了可操作的见解。尽管如此，严格的验证对于确保其在实际场景中的实用性和有效性至关重要。
