Open Access Highly Accessed Short paper

Calorie labeling and consumer estimation of calories purchased

Glen B Taksler1* and Brian Elbel23

Author Affiliations

1 Medicine Institute, Cleveland Clinic, 9500 Euclid Avenue, G1-40F, Cleveland 44195, OH, USA

2 Departments of Population Health and Medicine, New York University School of Medicine, New York, NY, USA

3 New York University Wagner School of Public Service, New York, NY, USA

For all author emails, please log on.

International Journal of Behavioral Nutrition and Physical Activity 2014, 11:91  doi:10.1186/s12966-014-0091-2

Published: 12 July 2014



Studies rarely find fewer calories purchased following calorie labeling implementation. However, few studies consider whether estimates of the number of calories purchased improved following calorie labeling legislation.


Researchers surveyed customers and collected purchase receipts at fast food restaurants in the United States cities of Philadelphia (which implemented calorie labeling policies) and Baltimore (a matched comparison city) in December 2009 (pre-implementation) and June 2010 (post-implementation). A difference-in-difference design was used to examine the difference between estimated and actual calories purchased, and the odds of underestimating calories.

Participants in both cities, both pre- and post-calorie labeling, tended to underestimate calories purchased, by an average 216–409 calories. Adjusted difference-in-differences in estimated-actual calories were significant for individuals who ordered small meals and those with some college education (accuracy in Philadelphia improved by 78 and 231 calories, respectively, relative to Baltimore, p = 0.03-0.04). However, categorical accuracy was similar; the adjusted odds ratio [AOR] for underestimation by >100 calories was 0.90 (p = 0.48) in difference-in-difference models. Accuracy was most improved for subjects with a BA or higher education (AOR = 0.25, p < 0.001) and for individuals ordering small meals (AOR = 0.54, p = 0.001). Accuracy worsened for females (AOR = 1.38, p < 0.001) and for individuals ordering large meals (AOR = 1.27, p = 0.028).


We concluded that the odds of underestimating calories varied by subgroup, suggesting that at some level, consumers may incorporate labeling information.

Diet; Health policy; Energy intake; Caloric restriction; Obesity