Spatial and spatio-temporal analyses of neighborhood retail food environments: evidence for food planning and interventions
The food system has been increasingly recognized as an indispensable component in professional planning in Canada. As its retailing part, the Retail Food Environment (RFE) has recently gained considerable attention, since it plays an important role in shaping residents’ eating behaviors and diet-related health outcomes, especially obesity. Identifying the strengths and weaknesses of the RFE in a neighborhood is essential for successful food planning and interventions. Yet current neighborhood RFE assessment mainly uses secondary food outlet datasets to evaluate absolute food access, largely overlooks the dynamic nature of the RFE and the variations of in-store features between food outlets, and predominantly applies descriptive RFE measures. Comprised of three articles that focus on a common theme, neighborhood RFE assessment, this dissertation uses novel spatial and spatio-temporal statistical modeling approaches to explore neighborhood RFE in the Regional Municipality of Waterloo with food outlet datasets that include the information of both the community and consumer nutrition environments. Firstly, this research explores spatio-temporal variations of relative healthy food access (RHFA) with a multiple-year RFE dataset. The results suggest that food swamps are more prevalent than food deserts in the study region and that food swamps have become more prevalent during the study period. Spatio-temporal food swamps, neighborhoods where RHFA is decreasing faster than the regional trend, are highlighted for interventions. Secondly, this research investigates the association between marginalization and neighborhood RFE at various geographical scales. ‘Healthy’ and ‘less healthy’ food outlets are differentiated based on in-store features from a primary food outlet dataset. RFE ‘healthfulness’ is a relative measure of healthy food access, which is modeled via probability distributions. The results indicate that neighborhoods with higher residential instability, material deprivation, and population density are more likely to have access to healthy food outlets within a walkable distance from a binary ‘have’ or ‘not have’ access perspective. At the walkable distance scale however, materially deprived neighborhoods are found to have less healthy RFE (i.e., lower RHFA). Finally, this research applies a spatial factor analysis model to assess neighborhood restaurant environment (NRE) for the city of Kitchener with multiple restaurant assessment indicators. Neighborhoods with least healthy NRE (simultaneously suffer from lower relative availability of healthy eating options, higher prices of healthy eating, and lower/higher healthy eating facilitator/barrier) are identified. Facilitator/barrier is found to be most relevant with NRE healthfulness. This research significantly advances our understanding of neighborhood RFE. Conceptually, it extends the definition of food swamps by incorporating a temporal dimension and provides empirical evidence that the deprivation amplification hypothesis in the RFE context holds only at specific geographical scales when neighborhood RFE is assessed with specific strategies. It also challenges the uncertainties associated with descriptive RFE measures that purport to represent the underlying concept – the ‘healthfulness’ of neighborhood RFE. Methodologically, this research facilitates the application of spatial and spatio-temporal statistical approaches in RFE studies. Findings from this research could assist planners and policy makers in developing food intervention programs to improve neighborhood RFE and promote population-wide healthy eating in the Region of Waterloo.
Cite this version of the work
Hui Luan (2016). Spatial and spatio-temporal analyses of neighborhood retail food environments: evidence for food planning and interventions. UWSpace. http://hdl.handle.net/10012/11079