Poster Presentation Australian and New Zealand Obesity Society Annual Scientific Conference 2023

Designing school e-canteens to promote healthier lunch purchases: effects of traffic-light labelling and choice architecture.     (#213)

Helen G Dixon 1 2 , Maree L Scully 1 , Rebecca Wyse 3 , Claire Hardi 4 , Kristy Mills 4 , Tessa Delaney 3 , Melanie Wakefield 1 2
  1. Centre for Behavioural Research in Cancer, Cancer Council Victoria, Melbourne, VIC, Australia
  2. Melbourne School of Psychological Sciences., The University of Melbourne, Parkville, Victoria, Australia
  3. School of Medicine & Public Health, University of Newcastle, Newcastle, New South Wales, Australia
  4. Prevention Division, Cancer Council Victoria, Melbourne, VIC, Australia

BACKGROUND:  Schools increasingly use online canteens (e-canteens), offering a potentially low cost, high-reach channel for delivering healthy eating interventions to parents and students.

AIMS: This study tested whether re-designing the e-canteen interface to include traffic light nutrition labelling, or additionally incorporating choice architecture, promotes healthier school lunch purchases. 

METHODS: Using a cluster-randomised controlled trial, ten Victorian primary schools were randomised to one of three e-canteen conditions:  Control (standard); Traffic light labelling (green, amber, or red traffic light emojis displayed alongside menu items to show their level of healthiness); Traffic light labelling + Choice architecture (visibility and ease of access to healthier items maximised to ‘nudge’ consumers towards healthier choices).  N=19 weeks of e-canteen purchasing data for the  baseline (Term 2, 2022) and intervention periods (Term 3, 2022) was collected. Separate linear mixed models were conducted to test for differences between conditions in the nutritional content (energy, saturated fat, sugar and sodium) of e-canteen orders at follow-up, controlling for baseline values.

RESULTS: Orders for students in the Traffic light labelling condition were higher in energy, saturated fat and sodium (no change in sugar), had a higher proportion of red light items (+5%, p<0.05) relative to controls and no significant improvement in the proportion of green light items ordered.  Orders for students in the Traffic light labelling + Choice architecture had a higher proportion of green light items (+7%, p<0.05) relative to controls; however, there was no significant difference in energy, saturated fat, sugar, sodium, nor in the proportion of red light items ordered. 

CONCLUSIONS:  Traffic light labelling + Choice architecture appeared to positively reinforce healthier options on e-canteen menus, prompting students to order a higher proportion of healthy, green light items. However, contrary to our hypothesis, Traffic light labelling in isolation, appeared to prompt less healthy purchases. Further investigation is recommended.