论文部分内容阅读
目的介绍统计学相加模型以及相乘模型在分析生物学交互作用中的应用。方法本文首先对交互作用的概念和统计学检验原理进行概述,然后通过一项病例-对照研究实例,利用多种统计方法,应用SAS软件的相应模块,分析相乘模型和相加模型的不同结果,评价它们在解释生物学交互作用时的差异;其中对于相加模型可信区间及统计学检验,还提出了几种简单易行的评价方法。结果两因素对疾病风险的交互作用在不同模型表现不同。本文成功建立利用非线性混合效应模型进行相加模型可信区间估计的方法,并在病例对照研究实例中发现了具有统计学意义(P<0.05)的交互作用。结论利用非线性混合效应模型进行相加尺度交互作用的分析方法切实可行。在分析生物学交互作用过程中,应慎重选择适宜模型。
Objective To introduce the statistical addition model and multiplication model in the analysis of biological interactions. Methodology This paper first outlines the concept of interaction and the principle of statistical testing, and then uses a case-control study case, using various statistical methods, the corresponding modules of SAS software to analyze the different results of the multiplicative model and the additive model , To evaluate their differences in the explanation of biological interactions. For the confidence interval of additive model and statistical test, several simple and easy evaluation methods are also proposed. Results The interaction of two factors on disease risk differed in different models. In this paper, a method to estimate the confidence interval of the additive model using the nonlinear mixed-effects model has been successfully established and a statistically significant (P <0.05) interaction was found in case-control studies. Conclusion It is practicable to use non-linear mixed-effects model to analyze the additive scale interactions. In the analysis of biological interactions, the appropriate model should be carefully chosen.