(PO08-56-22) What Factors Shape Whether Digital Food Marketing Appeals to Children?

Objectives: Children are exposed to large amounts of unhealthy food marketing on digital media. This marketing contains elements that marketing experts consider child-appealing, such as cartoons, bold colors, and childish font styles. Although these elements capture children's attention, there is still ambiguity regarding which additional factors also play a role. We aimed to examine the effects of child characteristics (sociodemographic, behavioural, and dietary intake factors) and marketing instance features on whether digital food marketing instances appealed to children.

Methods: Thirty-nine children from Calgary, Alberta, Canada indicated whether 1660 digital food marketing instances (~130 per child) appealed to them ('Is this ad for kids like you'; yes vs. no). Each instance was evaluated by three children. Information on children's sociodemographic characteristics, screen-related behaviours, and dietary intake was also collected. Agreement between children was measured using the Fleiss' kappa statistic. Text, labels, objects, and logos extracted from the food marketing instances were combined with the variables collected from the children to fit logistic regression, random forest, conditional inference tree, and gradient boosting models to assess which variables were the most important determinants of child appeal.

Results: Agreement between children was low, with an average Fleiss' kappa of -0.01 (95% CI: -0.09, 0.08). The models indicated that the text contained in the food marketing instances was the most important determinant of child appeal. Besides text, the three most important predictors of child appeal were the household's highest level of education, children’s vegetable consumption and BMI. The conditional inference tree indicated that children with low consumption of vegetables, high consumption of unhealthy snacks and more time spent using screen devices tended to consider more food marketing instances as child appealing.

Conclusions: There is substantial variability among children in which marketing instances appeal to them; even children with similar characteristics disagreed. However, children with poorer dietary intakes reported that more food marketing instances appealed to them.

Funding Sources: Canadian Institutes of Health Research