TABLE 1—

Characteristics of Studies that Assessed the Impact of Menu Calorie Labeling on Energy Purchased in Restaurant Settings in a Systematic Review and Meta-Analysis of Studies Through October 2013

Study (Year)	Design	Intervention	Setting	Population	Duration	Assessment	Outcome	Results
Bollinger et al. (2011)50	Natural experiment with control	Data collected before and after introduction of menu calorie labeling in New York City with no calorie labeling in comparison cities	Chain coffee shop	All Starbucks chain locations in New York City, Boston, and Philadelphia (∼100 million transactions)	Complete transaction records for 3 mo prelabeling and 11 mo postlabeling	Transaction records	kcal purchased per transaction	Difference-in-difference: log(kcal)b = −0.060 (SE = 0.001)
								−14.4 kcal (SE = 0.247)
Downs et al. (2013)26	Natural experiment without control	Data collected before and after introduction of menu calorie labeling in New York City; subjects randomized to receive per meal anchor, daily anchor, or no calorie anchor	Fast-food restaurants	1094 adults	2 mo prelabeling and 2 mo postlabeling	Intercept receipts	kcal purchased per meal	Model with adjustment for neighborhood, gender, age, race/ethnicity, calorie recommendation:
								Labeling b = 17.74 (SE = 28.20)
								Daily anchor b = 61.44 (SE = 34.22)
								Combined effect = 79.18 (SE = 62.42)
Dumanovsky et al. (2011)51	Natural experiment without control	Data collected before and after introduction of menu calorie labeling in New York City	Fast-food restaurants	15 798 adults	1 y prelabeling and 9 mo postlabeling	Intercept receipts	kcal purchased per meal	Model with adjustment for chain, gender, neighborhood poverty: b = 15.3 kcal (95% CI = 34.5, −3.8)
Elbel et al. (2011)28	Natural experiment with control	Data collected before and after introduction of menu calorie labeling in New York City with no calorie labeling in comparison city	Fast-food restaurants	349 children and adolescents aged 1–17 y	2 wk prelabeling, 1 mo postlabeling	Intercept receipts	kcal purchased per meal	Calorie labeling
								Prelabeling mean: 643 kcal (SD = 334)
								Postlabeling mean: 652 kcal (SD = 330)
								Control
								Prelabeling mean: 611 kcal (SD = 366)
								Postlabeling mean: 673 kcal (SD = 265)
Elbel et al. (2009)52	Natural experiment with control	Data collected before and after introduction of menu calorie labeling in New York City with no calorie labeling in comparison city	Fast-food restaurants	1125 adults	2 wk prelabeling, 1 mo postlabeling	Intercept receipts	kcal purchased per meal	Difference-in-difference:
								b = 19 kcal (SE = 58)
Ellison et al. (2013)24	Randomized controlled trial	Participants randomized to order from college restaurant menu with calorie labeling, labeling plus traffic light, or no labeling	Sit-down college restaurant	138 adults	Single exposure	In-person order	kcal purchased per meal	Calorie labeling mean: 817 kcal (SD = 328)
								Control mean: 765 kcal (SD = 368)
Finkelstein et al. (2011)29	Natural experiment with control	Data collected before and after introduction of menu calorie labeling in King County, WA, with no calorie labeling in comparison counties	Fast-food restaurants	Monthly transaction data (∼10 000 per store per month) from 7 intervention and 7 control restaurants	Complete transaction records for 12 mo prelabeling and 13 mo postlabeling	Transaction records	kcal purchased per transaction	Difference-in-difference adjusting for season:
								b = 18.5 kcal (SE = 15.11)
Krieger et al. (2013)30	Natural experiment without control	Data collected before and after introduction of menu calorie labeling in King County, WA	Fast-food restaurants and coffee chain	7325 participants aged ≥ 14 y	Data collected 1–3 mo prelabeling, 4–6 mo and 16–18 mo postlabeling	Intercept receipts	kcal purchased per meal	Post period 2 change controlling for chain type, gender, race/ethnicity, age, and location:
								Food chains:
								b = −35.5 kcal (95% CI = −75.5, 4.4)
								Coffee chains:
								b = −26.3 kcal (95% CI = −40.0, −12.7)
Tandon et al. (2011)53	Natural experiment with control	Data collected before and after introduction of menu calorie labeling in King County, WA, with no calorie labeling in comparison county	Fast-food restaurants	133 parent–child pairs with children aged 6–11 y	One meal prelabeling and 1 meal 3–6 mo postlabeling	Mailed receipt with telephone survey	kcal purchased per meal	Difference-in-difference; Model:
								Children: b = 34 kcal (95% CI = −90, 159)
								Adults: b = 5 kcal (95% CI = −119, 129)

Notes. CI = confidence interval.