Yenerall Jackie, You Wen, Hill Jennie
Department of Agricultural and Resource Economics, University of Tennessee, 2621 Morgan Circle, Knoxville, TN, 27996, USA.
Department of Public Health Sciences, University of Virginia, 200 Jeanette Lancaster Way, Charlottesville, VA, 22903, USA.
BMC Public Health. 2020 Nov 19;20(1):1747. doi: 10.1186/s12889-020-09882-0.
Modifying a household's food environment by targeting a single retailer type, like supermarkets, has a limited impact on dietary outcomes. This may be because the food environment has a limited impact on shopping behaviors, or because households are not as reliant on supermarkets as we assume. However, our understanding of how households shop for food, especially when considering the use of both food at home (FAH) retailers, such as supermarkets, and away from home retailers (FAFH), such as restaurants, is limited. Thus, understanding how households shop for food is a necessary first step when developing programs to modify food purchasing behavior.
K-means cluster analysis was used to identify weekly food shopping trip patterns based on the percentage of trips to FAH and FAFH retailers in the 2013 Food Acquisition and Purchase Survey (FoodAPS) dataset (n = 4665 households). Multinomial logistic regression was used to examine the relationship between shopping trip patterns, household and food environment characteristics.
Three patterns emerged: primarily supermarket, primarily supercenter, or mix (i.e. no dominant retailer type, but high FAFH use). Households with incomes below 185% of the federal poverty line were evenly divided between patterns that rely primarily on FAH retailers, and the mix pattern. While nearly 70% of households with incomes above 185% of the federal poverty line are in the mix cluster. Supermarket and superstore availability significantly influenced the likelihood of belonging to those clusters respectively, while having a child, higher income, and attitudes towards healthy meal preparation time or taste significantly influenced the likelihood of belonging to the mix cluster.
Although lower-income households are more likely to rely primarily on FAH retailers, household's, regardless of income, that primarily utilize FAH retailers show a strong preference for either superstores or supermarkets suggesting a need for interventions to reach both retailer types. However, altering the food environment alone may not be sufficient to discourage use of FAFH retailers as households relying on FAFH retailers are significantly influenced by meal preparation time and healthy food taste.
通过针对单一零售商类型(如超市)来改变家庭的食物环境,对饮食结果的影响有限。这可能是因为食物环境对购物行为的影响有限,或者是因为家庭对超市的依赖程度并不像我们假设的那样高。然而,我们对家庭如何购买食物的理解有限,尤其是在考虑同时使用家庭内食物(FAH)零售商(如超市)和家庭外食物(FAFH)零售商(如餐馆)的情况下。因此,了解家庭如何购买食物是制定改变食物购买行为计划的必要第一步。
使用K均值聚类分析,基于2013年食物获取与购买调查(FoodAPS)数据集(n = 4665户家庭)中前往FAH和FAFH零售商的出行百分比,确定每周的食物购物出行模式。使用多项逻辑回归来检验购物出行模式、家庭和食物环境特征之间的关系。
出现了三种模式:主要是超市模式、主要是超级中心模式或混合模式(即没有主导的零售商类型,但FAFH使用率高)。收入低于联邦贫困线185%的家庭在主要依赖FAH零售商的模式和混合模式之间平均分配。而收入高于联邦贫困线185%的家庭中,近70%属于混合模式集群。超市和大型超市的可获得性分别显著影响了属于这些集群的可能性,而有孩子、较高收入以及对健康膳食准备时间或口味的态度则显著影响了属于混合模式集群的可能性。
虽然低收入家庭更有可能主要依赖FAH零售商,但无论收入如何,主要使用FAH零售商的家庭对超级中心或超市都有强烈偏好,这表明需要针对这两种零售商类型进行干预。然而,仅改变食物环境可能不足以抑制FAFH零售商的使用,因为依赖FAFH零售商的家庭受膳食准备时间和健康食物口味的影响很大。