论文部分内容阅读
以中国８省人口、经济与营养健康调查的资料为例，利用多水平模型探讨了在平衡个人因素后省、县、家庭三个水平上的因素对动物蛋白摄入比例的影响。发现家庭因素如平均收入、饮食习惯、对饮食营养的认识程度等是影响动物蛋白摄入的较重要因素；此外，还发现不同省、不同县动物蛋白的摄入存在差异，揭示出摄入动物蛋白比例的增长还可能与各地区饮食文化和种植特点有关。从而可以为改善我国营养结构提供一些较为客观的依据。对于不同模型的比较认为多水平模型可以有效地用于当前许多大型卫生调查中具有层次结构的数据的分析，克服普通回归模型的不足
Taking the 8 provinces in China as an example, this paper uses the multi-level model to investigate the impact of factors on animal protein intake after balancing individual factors with provincial, county and family levels. Found that the family factors such as average income, eating habits, awareness of diet nutrition are the more important factors affecting animal protein intake; In addition, also found that different provinces, different county animal protein intake differences, revealing the intake of animals The increase in protein proportion may also be related to the dietary culture and planting characteristics in different regions. Which can provide some more objective basis for improving our country’s nutrition structure. Comparison of different models suggests that a multi-level model can be effectively used for the analysis of hierarchically structured data in many current large-scale health surveys, overcoming the deficiencies of the normal regression model