Are secondary data sources on the neighbourhood food environment accurate? Case-study in Glasgow, UK.

OBJECTIVES: To assess the validity of a publicly available list of food stores through field observations of their existence, in order to contribute to research on neighbourhood food environments and health. METHODS: All multiple-owned supermarkets, and a 1 in 8 sample of other food outlets, listed in 1997 and 2007 in the public register of food premises held by Glasgow City Council, Scotland, were visited to establish whether they were trading as foodstores. Postcode sectors in which foodstores were located were classified into least, middling and most deprived neighbourhoods. RESULTS: In total, 325 listed foodstores were visited in 1997 and 508 in 2007. Of these 87% and 88%, respectively, were trading as foodstores. There was a very slight gradient in validity by deprivation, with validity higher in least deprived neighbourhoods, though this was not statistically significant. CONCLUSIONS: There was reasonable, but not perfect, agreement between the list of food premises and field observations, with nearly 1 in 9 of sampled foodstores not present on the ground. Since the use of inaccurate secondary data sources may affect estimates of relationships between the neighbourhood food environment and health, further work is required to establish the validity of such data in different contexts.