Physical and online food outlet availability and its influence on out-of-home dietary behaviours in Great Britain: A repeated cross-sectional study

Abstract

Background: As online food delivery service (OFDS) platforms gain popularity, understanding their impact on diet alongside physical food outlets is important for addressing suboptimal dietary quality. This study examined the independent and combined associations between physical and online food outlet availability and out-of-home dietary behaviours in 2019 and 2022. We also explored whether associations between physical outlet availability and dietary behaviours are modified by online food outlet availability.

Methods: In this repeated cross-sectional analysis, we used British data from the adult International Food Policy Study (IFPS) in 2019 (n = 2912) and 2022 (n = 3544). Postcodes were used to assess neighbourhood food outlet availability using Ordnance Survey data and to determine OFDS availability on three platforms through web scraping. Associations were examined between neighbourhood outlet and OFDS availability with self-reported frequency of physical food outlet use, online food outlet use, and consuming meals prepared out-of-home.

Results: In 2019 and 2022, both neighbourhood and OFDS availability were positively associated with all outcome measures. In 2019, after mutual adjustment, both availability measures remained associated with online food outlet use and consuming meals prepared out-of-home. However, in 2022, only OFDS availability was associated with these outcomes. For example, a one standard deviation increase in OFDS availability was associated with a 9% (95%CI 3%-14%) increase in frequency of consuming meals prepared out-of-home after adjusting for neighbourhood outlet availability. OFDS availability also modified associations between neighbourhood outlets and both online food outlet use and out-of-home meal consumption. As OFDS availability increased, the link between neighbourhood outlets and out-of-home meal consumption weakened.

Conclusion: Neighbourhood outlet availability may influence out-of-home dietary behaviours, but its impact appears to weaken when OFDS availability is considered. Public health strategies should address the growing influence of OFDS platforms to improve dietary quality.

Keywords: Digital foodscape; Fast food; Fast-food; Takeaways.

Fig. 1

Incidence rate ratios and 95%CI of the association between neighbourhood OOH and OFDS availability, and OOH-related dietary behaviours in the Great Britain analytic IFPS weighted sample, 2022 (n = 3544). Neighbourhood OOH and OFDS availability measures were standardised by subtracting the mean to ensure comparability. Model 1 was adjusted for age, sex at birth, income adequacy, educational level, ethnicity, children in household, deprivation index, and supermarket availability (only for models with neighbourhood OOH availability due to their direct competition for customers, whereas we hypothesized that online OOH outlets and supermarkets cater to different consumer needs). Model 2 was additionally adjusted for the alternative food outlet availability measure. Abbreviations: OOH = out-of-home and OFDS = Online Food Delivery Service.

Fig. 2

Incidence rate ratios and 95%CI of the association between neighbourhood OOH availability, and online food outlet use and frequency of consuming meals prepared OOH stratified by quartiles of OFDS availability in the Great Britain analytic IFPS weighted sample, 2022. There were n = 886 participants in each stratified sample and the neighbourhood OOH availability was standardised separately by each subgroup of OFDS availability quartile by subtracting the mean to ensure comparability. Models were adjusted for age, sex at birth, income adequacy, educational level, ethnicity, children in household, deprivation index and supermarket availability. Abbreviations: Q = quartile, OOH = out-of-home and OFDS = Online Food Delivery Service.

Fig. 3

Incidence rate ratios and 95%CI of the association between neighbourhood OOH and OFDS availability, and OOH-related dietary behaviours in the England analytic IFPS weighted samples, 2019 (n = 2912) and 2022 (n = 3047). Neighbourhood OOH and OFDS availability measures were standardised by subtracting the mean to ensure comparability. Due to data availability, analyses were only conducted among those living in England. Model 1 was adjusted for age, sex at birth, income adequacy, educational level, ethnicity, deprivation index, and supermarket availability (only for models with neighbourhood OOH outlets due to their direct competition for customers, whereas online OOH outlets and supermarkets cater to different consumer needs). Model 2 was additionally adjusted for the alternative availability measure. Abbreviations: OOH = out-of-home.

Fig. 4

Incidence rate ratios and 95%CI of the association between neighbourhood OOH availability and dietary behaviours stratified by quartiles of OFDS availability in the England analytic IFPS weighted sample, 2019 and 2022. In 2019, there were n = 743, n = 726, n = 717 and n = 726 participants in Q1 through Q4, respectively. In 2022, this was n = 774, n = 755, n = 756 and n = 762. Due to data availability, analyses were only conducted among those living in England. Neighbourhood OOH availability was standardised separately for each subgroup of OFDS availability quartile by subtracting the mean to ensure comparability. Models were adjusted for age, sex at birth, income adequacy, educational level, ethnicity, deprivation index and supermarket availability. Abbreviations: Q = quartile and OOH = out-of-home.