重复测量资料非线性混合效应模型应用与实现

Application of Nonlinear Mixed Effects Model for Repeated Measurement Data and Implement

目的探讨重复测量资料非线性分析技术、SAS软件NLMIXED过程及在群体药动学的应用。方法结合重复测量数据特点,采用最大似然原理进行参数估计,建立非线性混合效应参数模型。结果该模型不仅考虑了个体内和个体间变异,而且也考虑了参数间的非线性,允许固定效应和随机效应进入模型的非线性部分;可方便地分析随机缺失等非均衡数据;有助于引入其他解释变量时最佳模型的选择,更客观地解释其对代谢过程的影响。结论当重复测量资料不满足线性条件时,使用非线性混合效应模型能更客观地反映原数据特征,挖掘资料蕴藏的信息,弥补线性理论分析非线性重复测量资料之不足。

Objective To study the nonlinear analysis methods for repeated measurement data, implement parameter estimate in PPK by NLMIXED procedure in SAS soft. Methods We fit parameter model applying the theory of nonlinear mixed effects model, consider the characteris- tic of the kinds of data, carry out the parameter estimations using maximum likelihood estimators. Results The nonlinear mixed effects model for repeated measurement data not only recognize and estimate variability both between and within individuals, but also consideration the nonlinear relation ship between the explanatory variable and the reponse variable; allowing both fixed and random effects enter nonlinearly the models; permits us to handle applications involving both missing at random and unbalanced data; easy to introduce other explanatory variable to the model, analysis the effeetion of metabolize. Conclusion When the data is nonlinear repeated measurement data, nonlinear mixed effect model can fit the reed data better, dig all of the information, make up the linear theory analying the nonlinear repeated measurement data.