Higher densities of fast-food and full-service restaurants are not associated with obesity prevalence

Background: The obesity epidemic in the United States has been mirrored by an increase in calories consumed outside of the home and by expansions in the numbers of, and portion sizes at, both fast-food restaurants (FFRs) and full-service restaurants (FSRs), leading some to blame the epidemic on the restaurant industry. If this were indeed true, one would predict that greater per capita densities of FFRs and FSRs would lead to greater obesity prevalence. Objective: We evaluated the population-level association between both FSRs and FFRs and the prevalence of obesity and calculated the proportion of calories consumed in these establishments. Design: In this ecological cross-sectional study, we used county-level data (aggregate-level data) for obesity prevalence across the mainland United States in 2012 and matched these data to county-level per capita densities of FFRs and FSRs in the same year. Multiple linear regression was used to determine the relation between the prevalence of obesity and the densities of FFRs and FSRs after adjustment for confounding factors. Results: Contrary to expectations, obesity prevalence was highly significantly negatively related to the densities of both FFRs and FSRs (combined-effect R-2 = 0.195). This was principally because greater numbers of both FFRs and FSRs were located in areas in which individuals were on average wealthier and more educated. When we normalized for these factors (and additional socioeconomic variables), the associations between restaurant densities and obesity effectively disappeared (pooled R-2 = 0.008). Our calculations showed that the percentage of total calories consumed in FFRs and FSRs is a mean of only 15.9% of the total intake (maximum: 22.6%). Conclusions: Variations in the densities of FFRs and FSRs are not linked to the prevalence of obesity in the United States, and food consumed in these establishments is responsible for <20% of total energy intake. This finding has implications for policy decisions regarding how we aim to tackle the obesity epidemic.