Plain language summary
Nutrition labels provide point-of-purchase information on the nutritional content of pre-packaged foods. This study is a post-hoc exploratory analysis of recorded label information viewing behaviour and associated packaged food purchases of study participants over the four-week intervention period. For this study combined data from all three intervention groups carried out in a previous study were analysed. Participants who scanned at least one product label and/or purchased at least one packaged food or non-alcoholic beverage over the four-week study intervention period were included. Results show that label information was viewed for approximately one fifth of all purchased products. Shoppers were most likely to view labelling information for convenience foods, cereals, snack foods, breads, and oils. Authors conclude that nutrition labels may influence healthier food purchases by those consumers who choose to use them.
BACKGROUND There are few objective data on how nutrition labels are used in real-world shopping situations, or how they affect dietary choices and patterns. DESIGN The Starlight study was a four-week randomised, controlled trial of the effects of three different types of nutrition labels on consumer food purchases: Traffic Light Labels, Health Star Rating labels, or Nutrition Information Panels (control). Smartphone technology allowed participants to scan barcodes of packaged foods and receive randomly allocated labels on their phone screen, and to record their food purchases. The study app therefore provided objectively recorded data on label viewing behaviour and food purchases over a four-week period. A post-hoc analysis of trial data was undertaken to assess frequency of label use, label use by food group, and association between label use and the healthiness of packaged food products purchased. RESULTS Over the four-week intervention, study participants (n = 1255) viewed nutrition labels for and/or purchased 66,915 barcoded packaged products. Labels were viewed for 23% of all purchased products, with decreasing frequency over time. Shoppers were most likely to view labels for convenience foods, cereals, snack foods, bread and bakery products, and oils. They were least likely to view labels for sugar and honey products, eggs, fish, fruit and vegetables, and meat. Products for which participants viewed the label and subsequently purchased the product during the same shopping episode were significantly healthier than products where labels were viewed but the product was not subsequently purchased: mean difference in nutrient profile score -0.90 (95% CI -1.54 to -0.26). CONCLUSIONS In a secondary analysis of a nutrition labelling intervention trial, there was a significant association between label use and the healthiness of products purchased. Nutrition label use may therefore lead to healthier food purchases.