Methods of measuring associations between the Retail Food Environment and weight status: Importance of classifications and metrics. (August 2019)

• Record Type: Journal Article
• Authors: Wilkins, Emma; Morris, Michelle; Radley, Duncan; Griffiths, Claire
• Is Part Of: SSM - population health. Volume 8(2019)
• Publication Date: 2019-08
• Subjects: Community nutrition environment -- Foodscape -- GIS -- Food access -- Obesogenic environment -- Methodology -- United Kingdom; Social medicine -- Periodicals; Medical anthropology -- Periodicals; Public health -- Periodicals; Psychology -- Periodicals; Medicine -- Periodicals; 362.105
• DOI: 10.1016/j.ssmph.2019.100404
• Languages: English
• ISSNs: 2352-8273

Abstract

Abstract: Despite considerable research, evidence supporting associations between the 'Retail Food Environment' (RFE) and obesity remains mixed. Differences in the methods used to measure the RFE may explain this heterogeneity. Using data on a large (n = 10, 111) sample of adults from the Yorkshire Health Study (UK), we modelled cross-sectional associations between the RFE and weight status using (i) multiple definitions of 'Fast Food', 'Convenience' and 'Supermarkets' and (ii) multiple RFE metrics, identified in a prior systematic review to be common in the literature. Both the choice of outlet definition and the choice of RFE metric substantively impacted observed associations with weight status. Findings differed in relation to statistical significance, effect sizes, and directions of association. This study provides novel evidence that the diversity of RFE measurement methods is contributing to heterogeneous study findings and conflicting policy messages. Greater attention is needed when selecting and communicating RFE measures in research. Highlights: This study compared different Retail Food Environment (RFE) measures. Food outlet definitions substantively impacted RFE-obesity associations. Associations also differed substantively when employing different RFE metrics. Existing literature uses diverse RFE measures, contributing to mixed study findings. Overlooking measurement differences impairs evidence synthesis and translation.