Ready-to-eat cereals from 4 supermarkets (ICA, Coop, Hemköp, and Lidl), representing 90% of the market share in Sweden, were analyzed [16]. The largest supermarkets in the Stockholm area were targeted to ensure the presence of the entire range of products. Prior to data collection, which took place in February and March 2019, consent forms were sent to the store managers. One store denied us permission without giving more details on why they chose to do so, and thus this retailer was not included. On site, the cereal department and all shelves containing breakfast cereals, including gluten-free sections, were identified. Oatmeal for porridge was regarded as a grain and not used as a breakfast cereal per se. Muesli, however, was included. There is a thin line between cereal bars and candy, and cereal bars are categorized differently from breakfast cereal on the FoodSwitch platform. Thus, they were not deemed breakfast cereal and were excluded.

The DataCollector app of the FoodSwitch platform (Android, version 2.7) from the George Institute for Global Health was used to collect information from packages. First, the bar codes were scanned using a smartphone camera. Then, numerous pictures of the front of the pack, the Nutrition Information Panel (NIP), and the ingredients were taken to collect all relevant information.

All products were manually entered in the system by 1 researcher (AM). The NIP, the gluten status and the Keyhole symbol, if present, were recorded. All child-targeted information on the package was identified. The criteria were as follows: the presence of cartoons, games, toys, children’s movie references, or text addressed to children on the package. A second researcher (V-ML) reviewed all the information entered and confirmed the presence of marketing targeted at children.

Finally, products were categorized and an HSR was generated. HSR scores of 0.5 to 5 stars in 10 half-star increments were assigned to all scanned products, where a higher number of stars represents healthier products. In the HSR system, each packaged food item is categorized into 1 of 6 categories depending on food type. All breakfast cereals were assigned to category 2. The HSR score was calculated via baseline and modifying points using the following formula: HSR score = baseline points – modifying points. Baseline points depended on energy content, saturated fat, sugar, and salt, and modifying points were based on protein, fiber, fruit, and vegetable content. The final assignment of the HSR score depended on which category the product was assigned to [17].

Categorical variables were summarized as the number of products and corresponding percentages, and continuous variables were summarized as mean (SD) and median (IQR). A Kolmogorov-Smirnov test was used to determine whether the data were normally distributed or not. A Mann-Whitney U test and an independent-samples t test were used to compare the nutritional values and HSR of child-targeted cereals to the nontargeted ones for nonnormal and normal distribution, respectively. In addition, the HSR scores were divided into two groups, healthier (HSR≥3.5) and less healthy (HSR<3.5) [18]. After dichotomization of HSR scores, the Fisher exact test was performed to assess correlation to the presence of the Keyhole symbol. The significance level was set to P<.05. The statistical analysis was performed using SPSS Statistics 25 (IBM Corp).

This was a study of packaged food supplies in supermarkets in Sweden and did not involve study participants or animal testing. Therefore, no ethical permission was sought.

This study compared the healthiness of ready-to-eat breakfast cereals targeted at children to cereal products not specifically targeted at children across supermarkets in Sweden using the FoodSwitch platform. Despite lower levels of saturated fat, the child-targeted cereals were overall less healthy, according to the nutritional content. Child-targeted products had a greater amount of sugar, with a mean nearly 2 times higher than the cereals not branded toward children. Furthermore, they were lower in fiber and protein. Our results are in line with 3 similar studies conducted in Canada, Australia, New Zealand, Guatemala, and the United States [19-23]. However, this is the first study focusing on the healthiness of packaged cereal products in Europe. Interestingly, the number of products targeted at children is relatively small. Compared to the markets analyzed in other countries, we obtained the smallest ratio between child-targeted cereals and non–child-targeted ones [19-23]. A few cereals rated high in healthiness according to the HSR system although these had a high sugar content. This aligns with previous findings showing that the HSR rating algorithm does not sufficiently penalize products with a high added sugar content [24].

In a context where childhood obesity is at the heart of public health concerns, our findings are highly relevant. The direct link between added sugar intake and obesity is well established [25,26]. Recently, the American Heart Association reported associations between added sugars and cardiovascular disease risk factors among children at levels well below the actual average intake [27]. Therefore, sugar levels in foods intended for children should be as low as possible or at least be similar to products not branded toward children. Dietary fibers are known to improve satiety, and fiber consumption is associated with lower body weight [28]. Even though it has been shown that consumers of breakfast cereals are more likely to reach the recommended daily amount of fiber, there should be no difference in fiber between food items branded toward children and cereals not targeted at children [29]. The presence of more total fat and saturated fat in non–child-targeted products was probably due to cereal type. When the products were categorized, we noticed that granola and muesli accounted for a majority. These types of cereals often contain nuts and seeds, which are high in fat. Salt content was similarly low in all products. In fact, many sodium-reduction initiatives have been introduced as part of the European Union salt reduction framework [30]. One study showed a significant drop in salt levels in breakfast cereals in the United Kingdom [31]. We can assume that the same happened in Sweden since the brands sold in both countries are very similar; however, this would be possible to investigate in the future by using this study as the reference values for 2019.

Marketing toward children often depicts cartoon characters to influence their food preferences [32]. In addition, children exposed to child-targeted cereals via TV commercials tend to have higher intake of these advertised cereals [33]. Problematically, products advertised to children are often of poor nutritional quality, as demonstrated in our study [34]. Unfortunately, packaging also affects parents, and parents being role models influence the eating habits of their children [35]. In fact, children depend on their parents’ food choices. Finally, it has been shown that NIPs and front-of-pack labels can sometimes be difficult to interpret [36]. Therefore, there is a need to provide easy-to-understand solutions to better guide consumers and implement healthy dietary habits from the earliest age.

A substantial part of the foods sold and eaten in high-income countries is preprepared packaged foods, and processed foods are largely responsible for an excessive intake of saturated fat, energy, added sugars, and sodium in the diet [37]. Thus, the food industry plays an increasingly important role in public health. The aim of the FoodSwitch platform is to change consumers’ behaviors and legislators’ actions. Front-of-pack labels are well established in influencing consumers toward healthier food choices. The Keyhole symbol used in Sweden was implemented 30 years ago. This front-of-pack logo is well known by Swedes as well as Scandinavians [38]. Although it is hard to assess its effect, it seems to positively affect the consumers’ behavior [39]. Unfortunately, not all manufacturers use the logo even though it is free of charge. The criteria required to make a product eligible for the Keyhole symbol are numerous and vary according to product type. Thus, producers might not be aware when a product meets the standard. Moreover, displaying the symbol induces extra costs for new packaging. Manufacturers often prioritize the sales of their foods. Therefore, another reason for the lack of compliance to the Keyhole symbol might be that producers are not convinced of any added value by the logo on their products. Throughout our investigation, we realized that many products considered as “healthy” by the HSR system (>3.5) did not display the Keyhole symbol. It is very likely that some of them were suitable to be branded with the Keyhole symbol.

Because of this, introducing the FoodSwitch app in Sweden could be of interest to Swedish consumers. First, it would allow a virtual front-of-pack labeling of all products (Figure 2). In fact, with information from the database, the app could generate the HSR score and the Traffic Lights label for any packaged product and suggest a healthier alternative. However, these two front-of-pack labeling systems used in the FoodSwitch app to direct consumers toward a healthier food choice are not used in Sweden. Fortunately, both the HSR and Traffic Lights label are self-explanatory and should not impede compliance of Swedish users to the mobile app’s recommendations. In addition, it would be possible to add a filter on the app that would generate the Keyhole symbol. Second, the implementation of FoodSwitch in Sweden could pressure manufacturers to reformulate their products. If consumers choose healthier packaged foods, producers would have to adapt to evolving client demand. Finally, the FoodSwitch database would be a very useful tool for conducting research on many aspects of the packaged food market. Through crowdsourcing, the information from packaged foods are always updated, which allows a tracking of the nutritional composition of the food supply over time.

We consider the conclusions drawn from our primary investigation to be reliable for the Swedish market because we visited the largest grocery retailers, representing 90% of market shares [16]. Another strength is the reliability of our data. During data collection, the data were double-checked by the collector prior to data entry. Once entered, the data were checked by a second researcher.

Limitations include possible transcribing errors and, in some cases, minor discrepancies that can occur between the NIP information shown on the package and the real value [40]. In terms of the method of assessing the healthiness of products, in Europe, displaying fiber content on the package is not mandatory [41]; however, it is often included on the NIP and only one product in our sample did not display fiber content. Finally, another limitation is that our results are limited to the Swedish market.

A further research objective could be to expand this assessment of nutrition to all child-targeted products. Additionally, data collection of breakfast cereals can be redone to provide a longitudinal perspective.

In conclusion, we showed that the nutritional quality of ready-to-eat cereals targeted at children was overall not significantly unhealthier than ready-to-eat breakfast cereals not targeted at children. However, cereals targeted at children were high in sugar and low in fiber. Thus, we conclude that there is a need to improve the dietary quality of child-targeted breakfast cereals in Sweden.