According to the WHO, food environment is one of the key factors to promote a healthy diet(1). Recent studies showed the influence of the food environment on food choices and food consumption which revealed the importance of monitoring food outlets and food quality in stores(Reference Ni Mhurchu, Vandevijvere and Waterlander2,Reference Minaker, Shuh and Olstad3) . Moreover, processed foods represent a third of the total volume of food purchased in supermarkets and grocery stores in the province of Quebec (Canada)(Reference Plamondon, Durette and Paquette4). Processed foods are often high in fat, sugar and Na, and studies show that chronic overconsumption can lead to an increased risk of non-communicable diseases(Reference Srour, Fezeu and Kesse-Guyot5).
To better understand the food environment to which populations are exposed, many countries are monitoring food’s nutritional value and its evolution over time. The International Network for Food and Obesity/non-communicable diseases Research, Monitoring and Action Support (INFORMAS)(Reference Neal, Sacks and Swinburn6) and the Food Monitoring Group(Reference Dunford, Webster and Metzler7) already monitor food composition in different countries in order to support governmental regulation and voluntary commitments by industry in creating healthier food environments. In parallel, the Observatory of Food Quality in France (OQALI) evaluates the nutritional composition of different food categories. For instance, they observed an improvement in the nutritional composition of pizzas and ready-to-eat (RTE) meals (i.e. lower fat and saturated fat content) between 2009–2012 and 2015–2016(8). The observed differences were mainly explained by product reformulation by the industry rather than changes in consumers’ choices towards healthier options. OQALI as well as other researchers in Europe and in the USA found that there is a high variability in nutritional quality among processed foods, even among products within a same food category(Reference Neal, Sacks and Swinburn6,Reference Dunford, Webster and Metzler7,Reference Baldridge, Huffman and Taylor9,10) . Although an increase in use of sales data has been observed over the past 5 years(Reference Bandy, Adhikari and Jebb11), access remains very difficult for researchers and thus, a limited number of studies has combined such data with the nutritional composition found on food labels in order to estimate what consumers actually buy and eat. This combination is highly relevant since it could contribute to better target products for reformulation. Indeed, it may help to focus policy-makers efforts on the types of products that sell the most, since a small improvement in nutrient content of great sellers can have a large and significant impact on public health(Reference Korošec and Pravst12). Furthermore, it may allow to monitor and assess the impact of the introduction of new regulations (i.e. nutrition facts table, front-of-pack (FOP) labelling) on food purchases. Except for Na(Reference Arcand, Jefferson and Schermel13) and sugar(Reference Bernstein, Christoforou and Weippert14), no study reporting overall nutritional value of food products in Canada has been yet published, which supports the relevance of monitoring the food environment and the nutritional quality of the food supply in provinces such as Quebec.
In 2016, the Quebec government created the very first government health prevention policy(15), a policy of major importance aiming to improve population health and quality of life. Several actions have been established within a framework to improve the nutritional quality of the food supply and further encourage food companies to improve the nutritional quality of their products. Meanwhile, the implementation of an Observatory aiming to monitor the nutritional quality of the food supply was deemed a priority for a network of researchers and knowledge users (representatives from governmental, non-governmental, parapublic and private organisations). Hosted by the Institute of Nutrition and Functional Foods at Université Laval, the Food Quality Observatory (henceforth Observatory) was thus officially launched in 2016 and is currently supported by the Quebec’s Ministry of Health and Social Services as well as the Quebec’s Ministry of Agriculture, Fisheries and Food. By generating reliable and useful information on the nutritional quality of food products available, the Observatory aims to contribute to the creation of healthier food environments, which will in turn facilitate healthier food choices and likely improve the overall health of the population.
The food category of RTE breakfast cereals has been selected to test the methodology and the feasibility of the studies to be undertaken by the Observatory. Indeed, breakfast is an important meal of the day(Reference Barr, Vatanparast and Smith16–Reference Gibson and Gunn20), and RTE breakfast cereals are part of the daily diet of a large proportion of the population(Reference Leech, Boushey and McNaughton21 ,Reference Albertson, Anderson and Crockett22) . Several studies which have analysed the impact of RTE breakfast cereals on dietary intakes and human health show variations in their nutritional value. Despite many positive impacts on diet quality (since they may provide whole grains, nuts, fruits, fibre, etc.) and cardiometabolic health(Reference Williams19,Reference Gibson and Gunn20,Reference Albertson, Thompson and Franko23–Reference Holmes, Kaffa and Campbell27) , some RTE breakfast cereals are highly processed and high in some nutrients of public health concern (e.g. added sugar, Na and preservatives)(Reference Maschkowski, Hartmann and Hoffmann28,Reference Cordain, Eaton and Sebastian29) .
It is also known that information on processed food products such as nutrition facts table, claims and other nutritional information can be difficult to understand for consumers(Reference Campos, Doxey and Hammond30). Strategies have been implemented to facilitate consumers’ food choices and to improve the nutritional quality of the food supply. Among these strategies, UK adopted in 2006 a voluntary FOP traffic light system coded for fat, saturated fat, sugar and Na(Reference Sacks, Rayner and Swinburn31). In 2013, the UK government published guidelines for uniform FOP colour-coded labelling(32). Since then, most of the UK supermarkets and many food manufacturers provide that label(33). Australia and New Zealand adopted in 2014 a voluntary Health Star Rating FOP system(34). Two years after the adoption of the FOP system, energy density and Na content were found to be lower whereas fibre content was higher in labelled products compared with their composition prior to the adoption of the FOP system(Reference Mhurchu, Eyles and Choi35). In 2016, Chile adopted a FOP warning symbol for food products exceeding specified amounts of energy, saturated fat, sugar and Na(36). After implementation, Chile observed that food companies reformulated products to adapt to the new regulation(Reference Quintiliano Scarpelli, Pinheiro Fernandes and Rodriguez Osiac37). All food categories combined, total sugar content showed the highest reduction after the FOP implementation, suggesting that sugar content of RTE breakfast cereals would thus be of major interest. The number of products with FOP ‘high in sugar’ before (with a simulation) and after implementation in Chile was significantly reduced in cereal products (e.g. cookies, crackers, cakes, breads) which included breakfast cereals. Indeed, median of total sugar almost had dropped by 50 % between 2013 and 2019, with 51 % of cereal products having a warning symbol in 2013 in comparison to 47 % in 2019. In 2017, a voluntary FOP label using letters from A to E was adopted to characterise the nutritional quality of food products in France (Nutri-Score)(Reference Ducrot, Méjean and Julia38). Since then, Belgium, Switzerland and Spain also adopted the Nutri-Score. In Canada, a standardised FOP warning symbol on food exceeding 15 % of the daily value (DV) for saturated fats, total sugar and Na is currently under consideration by the government(Reference Taillie, Hall and Popkin39). While it remains difficult at this point to confirm if nutrient-specific warning labels are more successful than summary labels(Reference Temple40), such a warning symbol could potentially act as a nudge for the food industry to reformulate their products while easily and rapidly informing consumers about less healthy food options. Moreover, this warning symbol would target specifically nutrients of interest which are known to be consumed in excess by consumers. In the case of RTE breakfast cereals, a FOP warning symbol – particularly for total sugar which is a nutrient of interest in this food category – could potentially affect a large number of products, including great sellers(Reference Goglia, Spiteri and Menard41).
Therefore, the aims of the present study were to generate a methodology to (1) test the use of sales data combined with nutrient values to characterise the nutritional composition of RTE breakfast cereals offered and purchased in the province of Quebec (Canada) and (2) verify the extent to which a FOP label based on the percentage of DV for total sugar, as a strategy to improve the food supply, would be distributed in this food category.
Methods
Data collection
In order to reach the objectives described above, a database containing the nutritional value of each RTE breakfast cereal was created by Protégez-Vous – a Quebec-based non-profit organisation specialising in consumer information and product testing – and was used by the Observatory following a data-sharing agreement. Nutritional and labelling information were obtained by Protégez-Vous by purchasing every RTE breakfast cereal in supermarkets, grocery stores and specialty grocery stores from the Greater Montreal area (Quebec, Canada) in September 2016. Cereals that were considered in this study were only cold breakfast cereals available in individual packaging (no multiple packages with several varieties of cereals) and those with nutritional information available on packaging. All information present on the product packaging (e.g. brand, nutrition facts table, list of ingredients, nutrition and health claims, serving size, etc.) were coded in the database using double coders. The reference portion of 55 g was chosen because it represented the reference amount for cereals (i.e. amount typically consumed in one occasion) at the time of the study. Nutritional value variables listed for the purposes of this study were as follows: energy (kJ), total fat (g), saturated fat (g), total sugar (g), fibre (g), protein (g) and Na (mg). The price per reference portion (55 g) and per unit (e.g. one box) were also documented by calculating the average of the prices observed in the various stores visited.
This nutritional value database was merged with a sales database (provided by Nielsen company(42)) of RTE breakfast cereals sold in the province of Quebec for 52 weeks between May 2016 and May 2017 using unique product codes (UPC). For each product, the database included the following data: sales in Canadian dollars (CAD$), sales in kilograms (kg) and sales in unit. Sales information comes from the optical reading of the products purchased in the main food chains of Quebec markets.
Classifications
RTE breakfast cereals were grouped by two different coders into different classifications to facilitate comparisons (e.g. muesli, sweetened, granola, etc.). The classifications were adapted from OQALI in France(Reference Goglia, Spiteri and Menard41). Each classification includes products with common characteristics in terms of their type of ingredients and/or technology used during processing. Definitions of these classifications are presented in supplemental materials.
Statistical analyses
To provide a general description of the nutritional value and the price per serving of RTE breakfast cereals found on the shelves available in Quebec (food supply), means and standard deviations illustrating the distribution of each of these variables were first calculated (n 331). The descriptive analyses were then repeated by weighting by sales volume in kg (food purchases; n 306)). Weighting the averages for sales better represents what Quebecers buy – and eventually consume – by giving a higher weight to the most popular cereals and a lower weight to the cereals which are less purchased. Since the analyses weighted for sales were produced from the combined database, the number of products analysed was lower than the one for the unweighted analysed (in which only the nutritional value data are available). Kruskal–Wallis tests and ANOVA were used to compare means and weighted for sales means nutrient content and prices between different cereal types. For all statistical tests, the significance threshold was corrected using the Bonferroni correction method to compensate for multiple comparisons. Statistical tests were conducted using SAS software version 9.4.
Results
A total of 331 different RTE breakfast cereals were identified in the Quebec food supply. Nutritional value and all packaging information were referenced for these cereals. Using the UPC, this dataset was merged with the sales database which contains more than 700 RTE breakfast cereals sold over 1 year. A total of 306 products with sales information were successfully cross-referenced with the 331 cereals identified in the food supply representing 92 % of RTE breakfast cereals identified in the Quebec food supply. The sales volume of products for which nutritional and purchasing information were both available amounted to CAD$ 230 million which represents 90 % of all sales of RTE breakfast cereals in Quebec. Products present in the sales database but missing in the nutritional database were mostly multiple packages with several varieties, discontinued products or different sizes of the products already included in the study.
Table 1 shows the variety of RTE breakfast cereals according to their type and purchase percentage. The most represented types of cereals were granola (36·6 %), sweetened (19·6 %), plain (15·4 %) and muesli cereals (10·3 %), whereas the most purchased were the sweetened (40·9 %) and the granola type (19·8 %).
Table 1 Availability and purchases of ready-to-eat (RTE) breakfast cereals according to their type
Table 2 shows the nutritional value and price per portion of all types of RTE breakfast cereals for both offered (as found on the shelves; n 331) and purchased RTE breakfast cereals (weighted by sales volume; n 306). A large variability was observed in saturated fat, total sugar and Na content of the different types of cereals. The variability was even higher in purchased than offered RTE breakfast cereals, emphasising once again on the importance of monitoring both food offered and purchased at the same time. Mean saturated fat content weighted for sales varied between 0·1 g (plain) and 2·7 g (granola) per 55 g of RTE breakfast cereals, while mean total sugar content varied between 6 g (plain) and 18 g (chocolate) and mean Na between 117 mg (granola) and 328 mg (plain). More particularly, when compared with others, granola cereals purchased had a higher energy, fat, saturated fat, protein content, a lower Na content and a similar total sugar content. Results remained similar when the unweighted nutritional composition of cereals was considered. Selling price of granola cereals was higher than other RTE breakfast cereals. When compared with others, sweetened cereals purchased had a higher total sugar content, and lower energy, fat and protein content. In addition to the previous results, when the unweighted nutritional composition was considered, Na content was higher and saturated fat and fibre contents were lower. Selling price of sweetened cereal was similar to others.
Table 2 Nutritional value and price (per 55 g serving) of ready-to-eat (RTE) breakfast cereals according to their type†,‡
* Significantly different from other cereals (P < 0·00078). This threshold equals to the Bonferroni correction for supply and purchases separately.
† Supply represents the average nutritional value of the cereals found on the shelves (n 331).
‡ Purchases represents the average nutritional value of cereals weighted by sales volume (n 306).
Figure 1 shows the large variation of total sugar content between the different types of cereal as well as between products within the same category. One RTE breakfast cereal out of five exceeded 15 % of the DV for total sugar (i.e. 15 g) and would get a FOP warning symbol according to Health Canada policy under review. More specifically, chocolate (65 %) and sweetened cereals (49 %) were those exceeding 15 % DV most often. Moreover, this figure illustrates the sales for each product and shows that many important sellers were sweetened cereals and most of them exceeded 15 % of the DV for total sugar. Actually, 65 % of sweetened cereals’ sales are above the 15 % of DV for total sugar.
Fig. 1 Distribution of sugar content of different types of ready-to-eat (RTE) breakfast cereals and their sales volume. The bigger the circles, the higher the sales (kg). Signs + represent RTE breakfast cereals for which sales data were not available
Discussion
The overview of RTE breakfast cereals offered and purchased in the province of Quebec generated in this study confirmed that the methodology of combining nutritional and sales data in Quebec and relating this data to the percentages of DV is feasible and relevant. First, 331 different RTE breakfast cereals have been identified and sales data were available for 306 of them which allowed coverage of a large part of the total cereals market (90 %) in the analyses. This overview also represents a wide variety of RTE breakfast cereals, similar to what has been observed in other countries(10,Reference Nieto, Rincon-Gallardo Patiño and Tolentino-Mayo43–Reference Vermote, Bonnewyn and Matthys45) . Moreover, the present study showed that granola and sweetened cereals are the most frequently found in the market with, respectively, 36·6 % and 19·6 % of the RTE breakfast cereals offered, as well as 19·8 % and 40·9 % of the RTE breakfast cereals purchased. These findings differ from what is observed elsewhere since, in comparison, chocolate cereals and light cereals were the most frequent RTE breakfast cereals in France(Reference Goglia, Spiteri and Menard41), while those two types of cereal each represented only 5% of the supply in Quebec. We also found that the nutritional value differs greatly between types of RTE breakfast cereals offered in the province of Quebec, as it has also been observed in other countries(Reference Goglia, Spiteri and Menard41,Reference Chepulis, Hill and Mearns46) . Total sugar content of Quebec RTE breakfast cereals greatly varies between types of cereals and even within same type. For example, the large range observed for total sugar content among sweetened cereals (i.e. from 1·8 g to 30·6 g/55 g serving) clearly demonstrates interesting opportunities for improvement. While large variations had also been observed elsewhere in the world, the mean total sugar content is higher in Quebec than in Australia(Reference Louie, Dunford and Walker47), UK(Reference Pombo-Rodrigues, Hashem and He48) and Belgium(Reference Vermote, Bonnewyn and Matthys45), which again underline the need for reformulation in this food category.
Additionally, a great proportion of chocolate (65 %) and sweetened (49 %) cereals are above 15 % of DV for total sugar (i.e. 15 g/55 g serving). This means that most of these cereals would carry, on the front of the package, the FOP warning symbol currently under consideration by Health Canada. Since chocolate cereals represent only 3·5 % of total RTE breakfast cereals purchases, reformulating total sugar content of these products – even if desirable – would have little impact in terms of public health. However, 65 % of sweetened cereals’ sales (sweetened cereals represent 40·9 % of total RTE breakfast cereals purchases) are above the 15 % of DV for total sugar. Thus, small reduction of total sugar content in these products may have a major public health impact. These cereals should therefore be closely monitored in the future to ensure that improvement efforts through reformulation are made by the industry. Reformulated products without nutritional warnings were perceived as more healthful and had higher purchase intention scores than their regular counterparts with warning, while nutrient claims did not have a relevant effect on consumers’ perception(Reference Schnettler, Ares and Sepúlveda49). Such a FOP symbol could thus be a win-win for consumers and companies.
Monitoring the evolution of the nutritional composition of RTE breakfast cereals is of major importance. No improvement has been seen between 2006 and 2010 in nutritional composition of RTE breakfast cereals in Australia(Reference Louie, Dunford and Walker47) nor in New Zealand between 2013 and 2017, suggesting that industry self-regulation of the nutritional composition of this food category needs reconsideration(Reference Chepulis, Hill and Mearns46). However, in the UK, a significant Na reduction was observed in breakfast cereals between 2004 and 2015 confirming the success of the UK voluntary Na reduction programme(Reference Pombo-Rodrigues, Hashem and He50). Similarly, OQALI had observed a significant Na reduction in chocolate and sweetened RTE breakfast cereals (30 mg and 60 mg/100 g, respectively) between 2008 and 2011(51). However, those changes were not significant after weighting for sales, suggesting that the biggest sellers did not change the nutritional composition of their products.
Since few studies have combined nutritional data with sales data to monitor the actual food supply, the present study is the first in Canada to assess the nutritional value of RTE breakfast cereals that are both offered and purchased. However, the present study has some limitations. The nutritional database is an overview at a given time that may not represent the whole portrait of the food supply during the whole year. Different products may not have been identified, such as the products that entered the market after data collection or those sold at another moment during the year but that were discontinued before the data collection. Additionally, not all RTE breakfast cereals were successfully matched to sales data (n 25). In fact, the sales database available through the Nielsen company does not include some RTE breakfast cereals, such as certain private labels of specific grocery stores. Moreover, even if food sales data can give an overview of food intakes(Reference Ransley, Donnelly and Khara52,Reference Martin, Howell and Duan53) , it is not possible to ensure that RTE breakfast cereals purchased are actually eaten by the consumers who bought them.
In terms of perspectives, the Observatory will use the methodology described in this paper to address other food categories that can have a significant impact on population health with the aim of monitoring the evolution of the nutritional value of the food supply in years to come. Currently, sliced breads, luncheon meats, RTE soups, granola bars, frozen meals, pasta sauces, yogurts and dairy desserts, sausages, cookies and crackers have also been analysed which sums up to more than 4000 food items (www.offrealimentaire.ca). These databases will give the possibility to characterise Quebec and Canadian food categories from different angles: target consumers, presence of claims, artificial sweeteners or food additives, etc.
In conclusion, the methodology used in this study provides an overview of the RTE breakfast cereals offered and purchased in the province of Quebec. This also leads to the identification of general findings regarding the nutritional value as well as to the information available on food packaging. Consequently, with these results, it becomes possible to identify areas of improvement regarding the nutritional composition of processed foods, which is of great relevance for policy-makers and public health nutrition advocates for healthier food choices. These findings form the basis for monitoring nutritional value of RTE breakfast cereals in the future. The same methodology will also be used for other food categories and will then allow the monitoring of a significant portion of the food supply in Canada.
The full report is freely available (in French only) on www.offrealimentaire.ca.
Acknowledgements
Acknowledgements: The authors are grateful to Raphaëlle Jacob for their involvement in data collection and classification, as well as to Laurélie Trudel, JoAnne Labrecque and the Observatory’s Scientific Committee and Board of Directors for their support. Financial support: The Food Quality Observatory received financial support from the ministère de la Santé et des Services sociaux du Québec (MSSS), the ministère de l’Agriculture, des Pêcheries et de l’Alimentation du Québec (MAPAQ), Québec en forme, ministère de l’Économie et de l’Innovation, Canada Foundation for Innovation and the Institut sur la nutrition et les aliments fonctionnels at Université Laval. Conflict of interest: All authors have no conflict of interest. Authorship: Conceptualisation: V.P., S.L., C.P. and M.P.; Carry out the study: S.P.; Statistical analysis: P.G.; Writing: J.P. and S.P.; Review and editing: all authors. Ethics of human subject participation: Not applicable.
Supplementary material
For supplementary material accompanying this paper, visit https://doi.org/10.1017/S1368980021001361
