变系数模型简介及其在医学研究领域中的应用  

Introduction and application of varying-coefficient models in medical research

作　　者：魏可成 余勇夫 秦国友 WEI Ke-cheng;YU Yong-fu;QIN Guo-you(Department of Biostatistics,School of Public Health,Fudan University,Shanghai 200032,China)

机构地区：[1]复旦大学公共卫生学院生物统计学教研室,上海200032

出　　处：《复旦学报（医学版）》2024年第3期439-442,共4页Fudan University Journal of Medical Sciences

基　　金：国家自然科学基金(82173612);上海市市级科技重大专项(ZD2021CY001)。

摘　　要：简要介绍变系数模型,并以变系数Cox模型为例,通过实际数据分析展示其在医学和公共卫生领域的应用,为相关研究提供方法学参考。实例基于某疾病预防控制中心部分慢病管理数据,拟合变系数Cox模型,探索高血压人群中体重指数(body mass index,BMI)与死亡风险之间的时变关联。结果显示,体重过低(BMI<18.5 kg/m^(2))与较高的死亡风险相关,但其关联程度随着随访时间的延长而逐渐减弱;超重(23 kg/m^(2)≤BMI<25 kg/m^(2))与较低的死亡风险相关,其关联程度随着随访时间的延长而逐渐减弱。变系数模型能够捕捉到暴露因素对于结局的影响如何随着其他变量的改变而变化,有助于更为全面地理解变量间的复杂关系,在医学和公共卫生研究中有很好的应用和推广价值。This paper briefly introduced the varying coefficient model and used the varying coefficient Cox model as an example to demonstrate its application in the fields of medicine and public health through real data analysis,thereby provided methodological references for related research.The example was based on chronic disease management data from a Center for Disease Control and Prevention,fitting a varying coefficient Cox model to explore the time-varying association between body mass index(BMI)and mortality risk among a hypertensive population.The results showed that being underweight(BMI<18.5 kg/m^(2))was associated with a higher risk of mortality,but this association weakened over time;being overweight(23 kg/m^(2)≤BMI<25 kg/m^(2))was associated with a lower risk of mortality,and this association also weakened over time.The varying coefficient model captured how the impact of exposure factors on outcomes changed with other variables,helping to comprehensively understand the complex relationships between variables,and had significant application and promotion value in medical and public health research.

关 键 词：COX模型 高血压 体重指数(BMI) 全因死亡 时变效应 

分 类 号：R311[医药卫生—基础医学]