基于纵向研究数据的非线性混合效应模型在中医证候演变规律研究中的应用  

Application of Nonlinear Mixed Effect Model Based on Longitudinal Study Data in the Study on TCM Syndrome Evolution Law

作　　者：伊明洋 关建新 闫丽祯 倪园园 王至婉[2,3,4] YI Mingyang;GUAN Jianxin;YAN Lizhen;NI Yuanyuan;WANG Zhiwan(Henan University of Chinese Medicine,Zhengzhou 450046,China;Co-construction Collaborative Innovation Center for Chinese Medicine and Respiratory Diseases by Henan&Education Ministry of China,Zhengzhou 450046,China;Henan Province Key Laboratory of TCM for Prevention and Treatment of Respiratory Diseases,Zhengzhou 450046,China;The First Affiliated Hospital of Henan University of Chinese Medicine,Zhengzhou 450000,China)

机构地区：[1]河南中医药大学,河南郑州450046 [2]河南中医药大学呼吸疾病中医药防治省部共建协同创新中心,河南郑州450046 [3]河南省中医药防治呼吸病重点实验室,河南郑州450046 [4]河南中医药大学第一附属医院,河南郑州450000

出　　处：《中国中医药信息杂志》2023年第10期17-21,共5页Chinese Journal of Information on Traditional Chinese Medicine

基　　金：国家重点研发计划(2017YFC1700103);国家自然科学基金面上项目(81973791)。

摘　　要：近年来,中医证候演变规律研究领域蓬勃发展,纵向研究作为前瞻性研究的一种,具有完整性、科学性等特点,可动态捕捉证候变化全过程。纵向研究所采集数据具有时间序列性、复杂多态性、相关性和非线性,如何处理和分析纵向研究中采集的大量临床数据是该领域难点。非线性混合效应模型既可处理非线性重复测量数据,又允许随机效应与固定效应进入模型非线性部分,可较好拟合证候随时间变化趋势,且能将影响证候演变的诸多因素作为协变量加入模型,观察对证候变化的影响。本文将纵向研究数据特点与非线性混合效应模型的优势相结合,通过实例分析,探讨该方法应用于证候演变规律研究的可行性,以期为后续研究提供方法学思路。In recent years,the research field of TCM syndrome evolution law has been booming.Longitudinal study,as a kind of prospective study,has the characteristics of completeness and scientific nature,and can dynamically capture the whole process of syndrome change.The data collected in longitudinal studies are characterized by time series,complex polymorphism,correlation and nonlinearity.How to deal with and analyze the large amount of clinical data collected in longitudinal studies is a difficulty in this field.The nonlinear mixed effects model can not only deal with nonlinear repeated measurement data,but also allow random effects and fixed effects to enter the nonlinear part of the model,which can better fit the changing trend of syndromes over time.Moreover,many factors that affect the evolution of syndromes can be added into the model as covariates to observe their influence on the changes of syndromes.This article combined the characteristics of longitudinal study data with the advantages of nonlinear mixed effects model,and explored the feasibility of applying this method to the study of syndrome evolution law through an example analysis,in order to provide methodological ideas for subsequent research.

关 键 词：非线性混合效应模型 证候演变规律 纵向研究 应用 

分 类 号：R2-03[医药卫生—中医学] R241