广义可加混合模型在婴幼儿生长发育研究中的应用

Application of Generalized Additive Mixed Model in Researches of Infant Growth

目的使用广义可加混合模型,分析婴幼儿体重增长的影响因素,探讨广义可加混合模型在婴幼儿生长发育研究中的应用。方法采用出生队列随访资料,通过R语言软件mgcv包拟合广义可加混合模型。结果本研究共纳入儿童819名,儿童体重的影响因素包括:月龄(F=516.17,P<0.001),出生体重(F=13.19,P<0.001),辅食添加频率得分(F=18.10,P<0.001),母亲产后体重(F=30.55,P<0.001),配方奶摄入量(F=11.28,P<0.001),月龄×配方奶摄入量(F=6.47,P<0.001),月龄×辅食添加频率得分(F=72.70,P<0.0001),月龄×睡眠总时长(F=2.63,P<0.05)。结论广义可加混合模型拟合效果较好,可探测变量间的复杂关系,适用于婴幼儿生长发育研究。

Objective To investigate determinants of infant weight gain and explore the application in the researches of infant growth using the generalized additive mixed model.Methods We analyzed the follow-up data of birth cohort and fit generalize additive mixed model through the mgcv package in the R software.Results Factors associated with infant weight gain were as follows:month age(F=516.17,P<0.001),birth weight(F=13.19,P<0.001),complementary feeding frequency(F=18.10,P<0.001),maternal postpartum weight(F=30.55,P<0.001),formula consumption(F=11.28,P<0.001),(month age)×(formula consumption)(F=6.47,P<0.001),(month age)×(complementary feeding frequency)(F=72.70,P<0.0001),(month age)×(sleep duration)(F=2.63,P<0.05).Conclusion The generalized additive mixed model was available for the researches of infant growth because it could detect the complex relationship between variables.