Food labelling data and reformulation tools

Food labelling is regulated to protect and support consumers’ informed food choices, but food labels also provide useful data for user communities. Cross-sectional monitoring of labelling provides important insights into supply trends and, whilst progress has been made with harmonisation of data collection, only a few countries have extensive monitoring programmes, often due to cost and time implications. The traditional approach is to take photographs of pre-packed foods in key food stores and, using data arising from this image, compile food labelling and composition databases. The approach generates hundreds of thousands of images and involves data extraction and manipulation. Data management and analyses in such studies is complex. Furthermore, such studies need to be repeated every few years. With the growing use of smart phones and associated infrastructure (data roaming, Wi-Fi), crowdsourcing became a realistic alternative for support in compilation of labelling databases. Advantages include reduced costs, speed, flexibility, scalability, and diversity, and participation of citizens, while disadvantages are related with accuracy and duplication.