Abstract
We recently showed that across the mainland USA there is no association between the density of fast food and full service restaurants and the prevalence of obesity. In a recent editorial it was suggested there are 4 problems with our analysis. The suggested problems were the area of analysis may not reflect adequately the exposure to different outlets, using the absolute numbers of restaurants rather than their ratio, using a global model which assumes the same relationship across all sites and finally the potential for residual confounding. In this short note we address all four of these issues and provide some new analysis of the impact of the ratio of restaurant types on obesity prevalence showing there is only a very weak association (r2 = 0.006). We conclude that none of the supposed weaknesses in our original analysis are valid.
Abbreviations
- FFR
- fast food restaurant
- FSR
- full service restaurant