Food-Led Neighborhood Regeneration and the Fruit and Vegetable Intake of Local Residents: The Mediating Role of Food Access

Food-led regeneration has emerged as a strategy for revitalizing declining urban neighborhoods. Such spatial regeneration is associated with variations in local food access, which further correlates with residents’ diets. This study employs structural equation modeling and explainable machine learning to analyze how food-led neighborhood regeneration in Chengdu, China, is associated with local residents’ fruit and vegetable (F&V) intake via its correlations with access to healthy, unhealthy, and online-to-offline food delivery service (OFDS) outlets. The findings reveal no direct correlation between food-led regeneration and F&V consumption; instead, food access mediates the association. Specifically, food-led regeneration has negative indirect associations with F&V intake through higher unhealthy and OFDS food density and lower healthy food diversity, and positive indirect associations via higher healthy food density, improved OFDS food diversity, and lower unhealthy food proximity. Furthermore, these indirect associations are level-dependent, shaped by nonlinear associations between regeneration level and food access, and between food access and F&V intake. A prominent negative association exists at moderate levels, mediated by higher unhealthy and OFDS food density and lower healthy food diversity. The study uncovers level-dependent, multi-faceted associations of the regeneration-food access-F&V intake nexus. It proposes level-adaptive strategies to facilitate health-oriented neighborhood regeneration.