Change in the Healthiness of Foods Sold in an Australian Supermarket Chain Following Implementation of a Shelf Tag Intervention Based on the Health Star Rating System

Abstract

Introduction: Most people in Australia buy most of their food in supermarkets. Marketing techniques promoting healthy foods in supermarkets can be important to encourage healthy eating at a population level. Shelf tags that highlight the healthiness of products have been identified as one such promising initiative. The aim of this study was to assess changes in the healthiness of foods sold in an Australian supermarket chain following implementation of a shelf tag intervention based on the Australian Health Star Rating (HSR) system. Methods: A controlled, non-randomised trial was undertaken in seven supermarkets (intervention: n = 3; control: n = 4) of a single chain in Victoria, Australia, over 12 weeks (4 weeks baseline, 8 weeks intervention period) between August and November 2015. The intervention involved provision of a shelf tag indicating the HSR of all packaged products that scored 4.5 or 5 stars (‘high-HSR products’) using the Australian HSR system. Posters indicating the healthiness of fresh fruits and vegetables (not eligible for an HSR rating, as they are not packaged) were also installed. Weekly per store sales data were provided by the retailer. In an intention-to-treat analysis (with intervention status of individual products based on their eligibility to be tagged), the proportion (%) of all ‘high-HSR’ packaged food sold and the volume of key nutrients (saturated fat, total fat, sodium, total sugar, protein, carbohydrates and energy) per 100 g sold were assessed. Difference-in-difference analyses were conducted to determine the difference between intervention and control stores in terms of mean outcomes between baseline and intervention periods. Customer exit surveys (n = 304) were conducted to evaluate awareness and use of the shelf tags and posters. Results: The proportion of ‘high-HSR products’ sold increased in the intervention period compared to the baseline period in each of the three intervention stores (average increase of 0.49%, 95% CI: −0.02, 0.99), compared to a decrease of −0.15% (−0.46, 0.15) in control stores (p = 0.034). The overall increase in intervention compared to control stores (difference-in-difference) of 0.64% represents an 8.2% increase in the sales of ‘high-HSR products’. Sales of total sugar, total fat, saturated fat, carbohydrates, sodium, protein and total energy in packaged food all decreased significantly more in intervention stores compared to control stores. Sales of fresh fruits and vegetables decreased in intervention stores compared to control stores. Customer surveys found that 34.4% noticed the shelf tags. Of those who noticed the tags, 58% believed the shelf tags influenced their purchases. Conclusions: With this study, we found that the use of shelf tags that highlight the healthiest packaged foods in a supermarket setting showed promise as a mechanism to improve the healthiness of purchases. Opportunities to scale up the intervention warrant exploration, with further research needed to assess the potential impact of the intervention on overall population diets over the longer term.