Marketing of unhealthy foods and beverages to youth: an investigation of domestic and international approaches to child-directed marketing restriction

Abstract

Background Diet quality during childhood is a critical factor for healthy growth and development. Children are vulnerable to food marketing, which impacts nutrition behaviours by influencing product awareness, attitudes, and purchasing intent. Restrictions on marketing of unhealthy foods to children are a potential population-level policy implemented in several countries. In Canada, marketing practices are guided by voluntary industry self-regulation; Québec is the only province which has mandatory restrictions on child-directed advertising (the Consumer Protection Act). In 2023, Health Canada proposed new regulations to restrict advertising of foods that contribute to excess intakes of sodium, sugars, and saturated fat to children under 13 years of age across some media and settings. Scientific literature on the effectiveness of mandatory marketing restrictions is mixed. Few studies have examined Québec’s existing provincial advertising law and there is limited evidence on nuanced components of marketing restrictions, including ‘brand-only’ marketing (i.e., marketing that relies solely on attributes recognizable as a representation of the brand, such as logos or mascots, without any identifiable food/beverage products). Additionally, though several countries require messages promoting healthy diet and exercise to be displayed on advertisements for unhealthy foods as an alternative policy measure, their potential to mitigate the promotional influence of advertisements is currently not well understood.

Research Objectives The overall objective of this thesis was to examine key characteristics of policies restricting food marketing to children that may strengthen or weaken policy impact. This was done using three sub-studies. Study 1 sought to compare self-reported exposure to child-directed unhealthy food marketing among youth in Québec (which has mandatory restrictions on child-directed advertising) to youth in the other Canadian provinces where no mandatory marketing restrictions exist. Study 2 sought to examine the impact of ‘brand-only’ marketing, which would be permitted under proposed federal regulations in Canada, on product appeal and selection among youth. Study 3 sought to examine the efficacy of public health messages displayed on sugar-sweetened beverage (SSB) advertisements on advertised product appeal and perceived healthiness among youth.

Methods The three studies were conducted as part of the International Food Policy Study (IFPS) Youth Surveys, which are annual cross-sectional surveys conducted online in Australia, Canada, Chile, Mexico, the United Kingdom, and the United States with youth aged 10-17 years. Study 1 used an observational design, consisting of cross-sectional analysis of data from 15,199 youth in Canada (n=3,757 in Québec) only. Survey waves 2019-2022 assessed self-reported exposure to unhealthy food marketing, including marketing locations, marketed products, marketing techniques, brands marketed, perceptions of marketing as targeting children under 13, and desire to consume advertised products. Logistic and linear regression models investigated differences by geographic jurisdiction, adjusting for age, sex-at-birth, race or ethnicity, perceived income adequacy, screentime, and survey year. Studies 2 and 3 both used experimental designs and were embedded within the 2023 IFPS Youth Surveys for all six countries. Study 2 consisted of a between-group experimental design conducted with 5,744 youth. Participants were randomised to view an advertisement according to one of four conditions: control (video game), ‘brand-only’ (McDonald’s logo), ‘brand and unhealthy food’ (McDonald’s logo and hamburger), and ‘brand and healthy food’ (McDonald’s logo and salad). Logistic and linear regression models tested the effect of ‘brand-only’ marketing on product appeal (measured through desire to consume the advertised product) and self-reported menu item selection in a hypothetical product selection task, as compared with other conditions. Study 3 consisted of a between-group 2-by-2 experimental design conducted with 11,521 youth. Participants were randomised to view one of two SSB advertisements (soda or a fruit drink) either with or without a health message. Logistic and linear regression models examined the efficacy of health message presence on advertised product appeal (measured through desire to consume the advertised beverage) and perceived healthiness, including 2-way interactions between beverage type and message presence.

Results Study 1: Compared to youth in other provinces, Québec youth reported seeing advertisements for unhealthy foods in fewer locations (β: -0.46), and for fewer beverage (β: -0.18) and restaurant brands (β: -0.15; p<0.001 for all); this was true across marketing locations regardless of whether they were covered under restrictions. Québec youth also reported seeing advertisements for sugary drinks (β: -0.40), fast-food (β: -0.30), sugary cereals (β: -0.38), snacks (β: -0.27), and desserts/treats (β: -0.28; p<0.001 for all) less frequently than those in other provinces. Québec youth had lower odds of reporting seeing advertisements featuring child-appealing marketing techniques including cartoons/characters (adjusted odds ratio (AOR): 0.57), celebrities (AOR: 0.57), unhealthy food/drink company branding (AOR: 0.75), and restaurant toys (AOR: 0.53; p<0.001 for all). Québec youth had lower odds of identifying beverage brand advertisements as ‘usually aimed at kids 12 and under’ (AOR: 0.72, p<0.001) and reported less desire to consume branded beverages than youth in other provinces (β: -0.10, p=0.002). Overall, youth’s self-reported exposure to unhealthy food and beverage advertisements was substantially lower in Québec, the only Canadian province with mandatory restrictions on child-directed advertising.

Study 2: No differences by condition were observed for product appeal (F=2.0, p=0.111). In terms of menu item selection, youth who were randomly assigned to view the ‘brand and healthy food’ advertisement were more likely to select the Garden Salad (5.7%; product featured in ‘brand and heathy food’ condition and only menu item meeting marketing nutrient profile model thresholds) than those in the ‘brand and unhealthy food’ (3.8%, odds ratio (OR):1.53) and ‘brand-only (3.5%, OR:1.49) conditions (p<0.05). Youth in the ‘brand and healthy food’ condition also selected items with less saturated fat and calories than those in the ‘brand and unhealthy food’ (saturated fat β: -0.66; calories β: -25.50) and ‘brand-only’ (saturated fat β: -0.62; calories β: -26.06) conditions (p<0.01 for all). Moderate-to-weak evidence suggests that youth in the ‘brand and unhealthy food’ condition (17.5%) were more likely to select the Quarter Pounder Deluxe (product featured in ‘brand and unhealthy food’ condition) than those in the ‘brand and healthy food’ (14.2%, OR: 1.29, p=0.013) and ‘brand-only’ (15.0%, OR: 1.20, p=0.068) conditions. Overall, brand-only advertisements likely have similar impacts on youth’s selection of unhealthy products as advertisements featuring unhealthy foods, whereas advertisements featuring healthy products may increase selection of healthier items.

Study 3: No differences in advertised beverage appeal were observed by health message presence (F=1.13, p=0.287). Youth who saw an advertisement with a health message rated the advertised beverage as healthier (mean:1.60) than those who saw an advertisement without a health message (mean:1.54, β:0.06, p=0.001). Health message presence slightly increased perceived healthiness for fruit punch (β: 0.09, p=0.002) but not Sprite (p=0.171); however, there was no evidence of an interaction between SSB beverage type and message presence for either perceived healthiness (F=1.68, p=0.195) or appeal (F=1.13, p=0.288). Overall, health messages may be ineffective in discouraging unhealthy food consumption and may even erroneously reinforce perceptions of SSBs as healthy.

Contribution/Significance The findings provide evidence in support of mandatory marketing policies having a protective effect, while highlighting the need to include brand-only marketing in marketing regulations. The results also suggest that health messages may not be an effective regulatory alternative to marketing restriction. Priorities for future research include: 1) observational research with Anglophone and Francophone youth across Canada, including robust sample sizes of non-dominant language speakers in each province, to discern language-related differences in child-directed marketing exposure; 2) experimental studies using ‘real world’ selection tasks to assess impact of brand-only and healthy food advertisements on menu item selection across a variety of brands and food processing levels; and 3) experimental studies investigating whether framing and design of health messages according to best practice guidelines may increase message efficacy. Overall, this thesis provides timely evidence to inform efficacy of policies seeking to restrict marketing of unhealthy foods to children in Canada and abroad.