Skip to main content

Table 1 Food environment exposure metrics compared in this study, and precedent for their use in the literature

From: Creating ‘obesogenic realities’; do our methodological choices make a difference when measuring the food environment?

Variable

Description

Precedent for usea

1000 population

Density of food outlets per 1000 population/per km2, per LSOA

O’Dwyer & Coveney [25] | Maddock [26] | Mehta & Chang [27] Ball et al.[6] | Moore & Diez Roux [28] | Chou et al.[29] Mobley et al.[30] | Simmons et al.[31] Sturm & Datar [32] Black et al.[33] | Powell et al.[34] | Burgoine et al.[10] Cummins et al.[35] | Reidpath et al.[36] | Macdonald et al.[37] Macdonald et al.[16]

km2

 

Block et al.[38] | Maddock [26] | Moore and Diez Roux [28]

LSOA centroid

  

Density per buffer at:

  

400 m Euclidean radius

Counts of food outlets within 400/800/1000 m Euclidean radius buffers from population-weighted centroids of LSOA

Austin et al.c[39] | Currie et al.c[40]

800 m Euclidean radius

 

Austin et al.c[39] | Jeffery et al.c[41] | Laraia et al.c[42] Timperio et al.c[43] | Currie et al.c[40] Spence et al.c[44]

1000 m Euclidean radius

 

Apparicio et al.[11] | Smoyer-Tomic et al.[45] | Seliske et al.c[46]

400 m Street Network

Counts of food outlets within 400/800/1000 m street network buffers from population-weighted centroids of LSOA

Smith et al.c[47]

800 m Street Network

 

Smoyer-Tomic et al.[50] | Harrison et al.[48]

1000 m Street Network

 

Larsen & Gillilandc[49] | Seliske et al.[46]

Proximity using:

  

Euclidean distance

Euclidean or street network distance (m) from LSOA population-weighted centroids to nearest food outlet

Apparicio et al.[11] | Winkler et al.[12] | Bodor et al.[13]

Street network distance

 

Zenk et al.b[14] | Pearce et al.[15] | Macdonald et al.[16] | Smith et al.[17] Pearce et al.[18] | Sharkey et al.[19] | Pearce et al.d[20] Sharkey & Horel [21] | Block et al.c[24] | Burdette & Whitaker [22] | Frank et al.c[23]

  1. aMethodologically similar, not necessarily using LSOAs; using imperial or metric measurements; geographic- or population-weighted centroids.
  2. bManhattan block distance used as an alternative to street network distance.
  3. cBuffer sizes employed around known home or school address.
  4. dUsed travel time rather than travel distance.