Association of the food environment with obesity: a systematic review of geographical and statistical methods

Abstract

BackgroundObesity is one of the main risk factors for many non-communicable diseases, and it is influenced by a complex, multifaceted system of determinants. Many studies have found an association between the food environment and obesity; however, the role of different types of food outlets and their contribution to obesity is unclear. MethodsWe performed a systematic review to explore the geographical and statistical methods used to evaluate the link between food environment and obesity, to assess the association between the availability of food outlets with obesity, and to determine gaps and limitations in the literature. We searched Medline via OvidSP, Scielo, Scopus, and Google Scholar databases in December, 2014, and February, 2016, searching for papers published since 1946 in Spanish or English. Search terms included spatial analysis, obesity, and food environment. Findings595 papers were retrieved and after inclusion and exclusion criteria, 65 studies were kept. Density (29 studies, 45%) and distance (21, 32%) were the most used geographical measures. Multilevel analysis was used by 20 studies (31%); three (5%) considered space, spatial, autocorrelation, or clustering phenomena whereas 34 (52%) did not. Longitudinal design was used by nine studies (14%). Positive associations (p<0·05) between the food environment and obesity were found by 39 studies (56%). Fast-food outlets (22 studies, 34%) and convenience stores (7, 11%) were the most associated with obesity. Limitations of the reviewed studies included non-representative populations (24 studies, 37%), self-reported data on height and weight (50, 77%), lack of verification of the food environment data (41, 63%), not considering physical activity or ability to walk (29, 45%), and assumptions about food outlet use. InterpretationAlthough many studies found a positive association of the food environment with obesity, inconsistencies remain. Several methodological issues can affect the quality of evidence of the association of the food environment with obesity. The selection of geographical methods and statistical methods must comply with the characteristics of the geographical location and the type of data being used. Public policy aimed at regulating food outlets could contribute to increase the availability and accessibility of healthy food and therefore decrease the burden of obesity. FundingNational Council for Science and Technology, Mexico.