Project 456911

Background: Unhealthy food marketing has a negative impact on what children eat and on their health. Health Canada is going to ban digital marketing of unhealthy food that appeals to children <13 years. However, it is not clear which types of marketing should be banned and it will be hard to know if companies are obeying the new law. To address these problems, we are developing an artificial intelligence (AI) system that can identify digital marketing of unhealthy food. Right now, the system can only identify a few marketing strategies that companies use to target children. Our system needs to be expanded so it can identify all of the ways that marketers try to appeal to children. Goals: This research will expand our AI system by training it to 'think like a child' and know which digital marketing of unhealthy food appeals to children. We will then use the new AI system to study how this marketing affects what children eat. Methods: Children (6-12 years; n=2,600) will tell us which marketing appeals to them. This information will be used to train our AI system to 'think like a child.' We will also check that the system is accurate and not biased against some groups of children, such as ethnic minority groups. Next, we will use the AI system to do a randomized controlled trial in Calgary, Toronto and Montreal. In this trial, 678 children will be randomized to see digital marketing of unhealthy food that does (n=226) or does not (n=226) appeal to children, or to non-food marketing (n=226). We will measure how this marketing affects children's intake of snacks, their diet quality, and their awareness of, and attitudes towards marketed foods. Expected results: This research will develop the first AI system that can identify digital marketing of unhealthy food that appeals to children from their own perspective. It will also help us to understand how policies can better protect children from the negative impacts of unhealthy food marketing.