Calories Purchased by Hospital Employees After Implementation of a Cafeteria Traffic Light-Labeling and Choice Architecture Program

JAMA Netw Open. 2019 Jul 3;2(7):e196789. doi: 10.1001/jamanetworkopen.2019.6789.


Importance: Nearly one-third of the 150 million US adults who are employed are obese. A workplace program that reduces employees' caloric intake could help prevent obesity.

Objective: To determine if implementation of a hospital workplace cafeteria healthy eating program was associated with reduction in calories purchased by employees over 2 years.

Design, setting, and participants: A longitudinal cohort study of 5695 employees who purchased food with their employee identification card before and after implementation of a workplace cafeteria healthy eating program was conducted from December 1, 2009, to February 29, 2012, in a hospital in Boston, Massachusetts. Statistical analysis was conducted from April 6, 2018, to May 14, 2019.

Exposure: After a 3-month baseline period, "traffic light" food labels (in which green indicates healthy, yellow indicates less healthy, and red indicates least healthy) and choice architecture (product placement) changes were implemented permanently.

Main outcome and measures: Main outcomes were changes in calories of employees' purchases from baseline (December 1, 2009, to February 28, 2010) to the same quarter 1 year (December 1, 2010, to February 28, 2011) and 2 years (December 1, 2011, to February 29, 2012) later. Purchases were obtained from sales data linked to employee identification cards. The hypothesized association of estimated changes in daily calories consumed with employees' weight was calculated using a dynamic model of weight change. Analyses were adjusted for age, sex, race/ethnicity, and job type.

Results: Among the 5695 employees in the study, 4057 were women and 1638 were men, and the mean (SD) age was 40 (12) years. Baseline purchases had a mean of 565 kcal per transaction (95% CI, 558-572 kcal) and decreased 19 kcal per transaction (95% CI, -23 to -15 kcal) at 1 year and 35 kcal per transaction (95% CI, -39 to -31 kcal) at 2 years (-6.2%; P < .001) relative to baseline. The largest reduction was from red-labeled items, decreasing 42 kcal per transaction at 2 years (95% CI, -45 to -39 kcal) from a mean of 183 kcal per transaction (95% CI, 177-188 kcal) at baseline (-23.0%; P < .001). Calories from green-labeled items increased 6 kcal per transaction (95% CI, 3-9 kcal) from a mean of 152 kcal per transaction (95% CI, 149-155 kcal) (4.0%; P < .001), and calories from yellow-labeled items did not change. Among frequent purchasers (≥36 transactions per quarter; n = 453), total kilocalories per quarter at baseline was 41 784, which, averaged over the entire follow-up, decreased a mean of 4275 kcal/quarter (95% CI, -5325 to -3224 kcal) over 2 years. Assuming no compensatory changes in diet or activity, this equated to a reduction of 47 kcal per day and estimated a 2.0-kg weight loss over 3 years.

Conclusions and relevance: A workplace cafeteria traffic light-labeling and choice architecture program was associated with a sustained decrease in calories purchased, particularly from unhealthy foods. Point-of-purchase programs could help improve dietary intake and prevent obesity in employees.

Publication types

  • Research Support, N.I.H., Extramural

MeSH terms

  • Adult
  • Choice Behavior
  • Diet, Healthy / methods*
  • Energy Intake*
  • Female
  • Food Labeling / methods*
  • Health Promotion / methods*
  • Humans
  • Male
  • Massachusetts
  • Obesity* / prevention & control
  • Obesity* / psychology
  • Occupational Health
  • Occupational Health Services / methods*