Background Exposure to a poor-quality food environment is thought to influence diet-related behavioural and health outcomes, but evidence is mixed. Better evidence may be generated by exploring associations with more proximal outcomes, such as food purchasing. This study explores associations between food environment exposures household take-home and out-of-home (OOH) food and drink purchasing.

Methods Item-level food and drink purchase data from the 2019 Kantar Fast Moving Consumer Goods panel for London and Northern England, UK, were used to assess associations between food environment exposures and household-level take-home (n=2118) and individual-level OOH (n=447) food and drink purchasing. Residential neighbourhood food environment exposures were created using address-level data from take-home (chain and non-chain supermarkets) and OOH (restaurants and takeaways) outlets retrieved from Ordnance Survey Points of Interest. Exposures were derived using a 1 km network buffer around the population-weighted centroid of households’ residential postcode districts. Density, proximity and relative composition measures were created for both take-home and OOH outlets. Negative binomial regression models explored associations between food environment exposure measures and frequency of take-home food and drink purchasing, total take-home calories, calories from fruits and vegetables, high fat, salt and sugar products, and ultra-processed foods (UPF), volume of take-home alcoholic beverages, and frequency of OOH food and drink purchasing. Models were adjusted for individual and household socio-economic characteristics, area deprivation and population density.

Results Fully adjusted models found no consistent patterns of association between food environment exposures and take-home and OOH food and drink purchasing. There was some evidence for an inverse relationship between the distance to OOH outlets and purchased calories from ultra-processed foods (UPF), with a 500 m increase in distance to the nearest OOH outlet associated with a 1.1% reduction in take-home UPF calories (Incidence Rate=0.989; 95% confidence interval 0.982–0.997). A lower density of independent supermarkets was linked to higher volume of purchased alcohol in Northern England only (IR=0.952, 95%CI 0.927–0.978). Greater distance to OOH outlets was associated with higher volumes of purchased alcohol in both regions separately, but more strongly in London (IR=1.298, 95%CI 1.089–1.549 vs. IR=1.139, 95%CI 1.039–1.248).

Conclusion This study, using highly granular objective consumer purchase data, finds limited evidence for the role of the food environment on food and drink purchasing, despite some evidence for the role of exposure of OOH outlets in UPF purchasing. Region-specific effects on alcohol purchases indicate the importance of spatial context for research and policy.