Mapping food access: how neighborhood deprivation shapes healthy food availability in the United Kingdom
摘要
Limited access to affordable, nutritious food is a key structural determinant of health inequality. In the United Kingdom, food desert risk is often assumed to mirror socioeconomic deprivation, yet their alignment across small-area geographies remains unclear.
ObjectivesTo characterize the spatial distribution of food desert risk across England, Wales, and Scotland; quantify its association and concordance with area-level deprivation; and identify deprivation domains most strongly linked to food desert risk.
MethodsWe integrated the E-Food Desert Index (EFDI) with national Indices of Multiple Deprivation (IMD) across 41,729 Lower Layer Super Output Areas and Data Zones using 2011 Census boundaries. Standardized scores and quartiles were analyzed using spatial clustering, correlation, multinomial regression, ordinary least squares, adaptive LASSO with interaction terms, and mixed-effects and spatial sensitivity analyses.
ResultsFood desert risk and deprivation were moderately correlated (r = 0.40, p < 0.001) but showed substantial spatial discordance. Nearly half (48.2%) of the most deprived areas were also in the highest food desert quartile, while about one-third showed discordant classifications. Deprivation showed a strong graded association with food desert risk (RRR per IMD quartile = 3.07; 95% CI: 2.98–3.16). Employment deprivation was the strongest predictor across nations, followed by education, housing/access to services, and living environment. Antagonistic employment–income interactions were observed. Rural areas with low deprivation but high food desert risk were especially evident in Scotland.
ConclusionsFood desert risk and deprivation represent overlapping but distinct dimensions of structural disadvantage. Reliance on deprivation indices alone may miss communities with substantial food access barriers, particularly in rural settings. Integrating food environment metrics alongside deprivation measures may improve place-based policy targeting.