Development of the Food Label Information Program: A Comprehensive Canadian Branded Food Composition Database

Evaluating the application of front-of-package labelling regulations to menu labelling in the Canadian restaurant sector using menu food label information and price (Menu-FLIP) 2020 data

OBJECTIVE

To evaluate the application of front-of-package (FOP) labelling regulations to menu labelling in the Canadian restaurant sector by assessing the proportion of menu items that would be required to display the 'high-in' FOP symbol if the policy were extended to the restaurant sector.

DESIGN

Nutrition information of 18 760 menu items was collected from 141 chain restaurants in Canada. Menu items were evaluated using the mandatory FOP labelling regulations promulgated in Canada Gazette II by Health Canada in July of 2022.

SETTING

Chain restaurants with ≥20 establishments in Canada.

PARTICIPANTS

Canadian chain restaurant menu items including beverages, desserts, entrées, sides and starters.

RESULTS

Overall, 77 % of menu items in the Canadian restaurant sector would display a 'high-in' FOP symbol. Among these menu items, 43 % would display 'high-in' one nutrient, 54 % would display 'high-in' two and 3 % would display 'high-in' all three nutrients-of-concern. By nutrient, 52 % were 'high-in' sodium, and 24 and 47 % were 'high-in' total sugars and saturated fat, respectively.

CONCLUSIONS

Given the poor nutritional quality of restaurant foods, the current regulations, if applied to restaurant foods, would result in most menu items displaying a FOP symbol. Therefore, expanding the Canadian FOP labelling regulations to the restaurant sector can be key to ensuring a healthy food environment for Canadians. Furthermore, menu labelling along with other multi-faceted approaches such as reformulation targets are necessary to improve the dietary intake of Canadians from restaurant foods.

Comparing Canada's 2018 proposed and 2022 final front-of-pack labelling regulations using generic food composition data and a nationally representative dietary intake survey

OBJECTIVE

The objective of the study was to compare the potential dietary impact of proposed and final front-of-pack labelling (FOPL) regulations (published in Canada Gazette I (CG1) and Canada Gazette II (CG2), respectively) by examining the difference in the prevalence of foods that would require a 'High in' front-of-pack nutrition symbol and nutrient intakes from those foods consumed by Canadian adults.

DESIGN

Foods in a generic food composition database (n 3676) were categorised according to the details of FOPL regulations in CGI and CGII, and the differences in the proportion of foods were compared. Using nationally representative dietary survey data, potential intakes of nutrients from foods that would display a 'High in' nutrition symbol according to CGI and CGII were compared.

SETTING

Canada.

PARTICIPANTS

Canadian adults (≥ 19 years; n 13 495).

RESULTS

Compared with CGI, less foods would display a 'High in' nutrition symbol (Δ = -6 %) according to CGII (saturated fat = -4 %, sugars = -1 %, sodium = -3 %). Similarly, potential intakes of nutrients-of-concern from foods that would display a 'High in' nutrition symbol were reduced according to CGII compared with CGI (saturated fat = -21 %, sugars = -2 %, sodium = -6 %). Potential intakes from foods that would display a 'High in' nutrition symbol were also reduced for energy and nutrients-to-encourage, including protein, fibre, calcium and vitamin D.

CONCLUSIONS

Changes to FOPL regulations may have blunted their potential to limit intakes of nutrients-of-concern; however, they likely averted potential unintended consequences on intakes of nutrients-to-encourage for Canadians (e.g. calcium and vitamin D). To ensure policy objectives are met, FOPL regulations must be monitored regularly and evaluated over time.

Examining the role of industry lobbying on Canadian front-of-pack labelling regulations

Health Canada recently issued a Marketing Authorization to expand the eligibility of the dairy-related exemption for Canadian front-of-pack labelling (FOPL) regulations. The 2024 Marketing Authorization exempts dairy-related products that are a 'source of calcium,' rather than only 'high in' calcium as previously regulated, from displaying a 'High in' front-of-pack nutrition symbol, regardless of their saturated fat and sodium levels. The Marketing Authorization, heavily influenced by the food industry, lacks strong scientific evidence to support its adoption. Although there is a high prevalence of inadequate calcium intakes among Canadians, the Marketing Authorization will exempt more dairy-related products that are significant contributors of saturated fat and sodium for Canadians. While providing very little calcium, many dairy-related products, particularly cheese products, are 'high in' saturated fat and/or sodium. Expanding the exemption criteria will allow dairy-related products with little health benefits to be reflected as 'healthy' (i.e., not display a 'High in' nutrition symbol), blunting the potential impact that FOPL regulations could have on improving the diets of Canadians. We strongly urge Health Canada to reconsider the expansion of the exemption and encourage others to conduct policy-relevant research and participate in the policy decision-making process to promote evidence-informed public health policies for the health of Canadians.

Monitoring sodium content in processed and ultraprocessed foods in Argentina 2022: compliance with National Legislation and Regional Targets

OBJECTIVE

To assess the current Na levels in a variety of processed food groups and categories available in the Argentinean market to monitor compliance with the National Law and to compare the current Na content levels with the updated Pan American Health Organisation (PAHO) regional targets.

DESIGN

Observational cross-sectional study.

SETTING AND PARTICIPANTS

Argentina. Data were collected during March 2022 in the city of Buenos Aires in two of the main supermarket chains. We carried out a systematic survey of pre-packaged food products available in the food supply assessing Na content as reported in nutrition information panels.

RESULTS

We surveyed 3997 food products, and the Na content of 760 and 2511 of them was compared with the maximum levels according to the Argentinean law and the regional targets, respectively. All food categories presented high variability of Na content. More than 90 % of the products included in the National Sodium Reduction Law were found to be compliant. Food groups with high median Na, such as meat and fish condiments, leavening flour and appetisers are not included in the National Law. In turn, comparisons with PAHO regional targets indicated that more than 50 % of the products were found to exceed the regional targets for Na.

CONCLUSIONS

This evidence suggests that it is imperative to update the National Sodium Reduction Law based on regional public health standards, adding new food groups and setting more stringent legal targets.

Advancing sustainability in the food and nutrition system: a review of artificial intelligence applications

Promoting sustainability in food and nutrition systems is essential to address the various challenges and trade-offs within the current food system. This imperative is guided by key principles and actionable steps, including enhancing productivity and efficiency, reducing waste, adopting sustainable agricultural practices, improving economic growth and livelihoods, and enhancing resilience at various levels. However, in order to change the current food consumption patterns of the world and move toward sustainable diets, as well as increase productivity in the food production chain, it is necessary to employ the findings and achievements of other sciences. These include the use of artificial intelligence-based technologies. Presented here is a narrative review of possible applications of artificial intelligence in the food production chain that could increase productivity and sustainability. In this study, the most significant roles that artificial intelligence can play in enhancing the productivity and sustainability of the food and nutrition system have been examined in terms of production, processing, distribution, and food consumption. The research revealed that artificial intelligence, a branch of computer science that uses intelligent machines to perform tasks that require human intelligence, can significantly contribute to sustainable food security. Patterns of production, transportation, supply chain, marketing, and food-related applications can all benefit from artificial intelligence. As this review of successful experiences indicates, artificial intelligence, machine learning, and big data are a boon to the goal of sustainable food security as they enable us to achieve our goals more efficiently.

Accelerating the Classification of NOVA Food Processing Levels Using a Fine-Tuned Language Model: A Multi-Country Study

The consumption and availability of ultra-processed foods (UPFs), which are associated with an increased risk of noncommunicable diseases, have increased in most countries. While many countries have or are planning to incorporate UPF recommendations in their national dietary guidelines, the classification of food processing levels relies on expertise-based manual categorization, which is labor-intensive and time-consuming. Our study utilized transformer-based language models to automate the classification of food processing levels according to the NOVA classification system in the Canada, Argentina, and US national food databases. We showed that fine-tuned language models using the ingredient list text found on food labels as inputs achieved a high overall accuracy (F1 score of 0.979) in predicting the food processing levels of Canadian food products, outperforming traditional machine learning models using structured nutrient data and bag-of-words. Most of the food categories reached a prediction accuracy of 0.98 using a fined-tuned language model, especially for predicting processed foods and ultra-processed foods. Our automation strategy was also effective and generalizable for classifying food products in the Argentina and US databases, providing a cost-effective approach for policymakers to monitor and regulate the UPFs in the global food supply.

Food Composition Databases (FCDBs): A Bibliometric Analysis

Food composition databases (FCDBs) are important tools that provide information on the nutritional content of foods. Previously, it was largely unclear what nutritional contents and which FCDBs were involved in highly cited papers. The bibliometric study aimed to identify the most productive authors, institutions, and journals. The chemicals/chemical compounds with high averaged citations and FCDBs used by highly cited papers were identified. In July 2023, the online database Web of Science Core Collection (WoSCC) was queried to identify papers related to FCDBs. A total of 803 papers were identified and analyzed. The first paper indexed in WoSCC was published in 1992 by Pennington, which described the usefulness of FCDB for researchers to identify core foods for their own studies. In that paper, the FCDB described was the USDA 1987-88 NFCS (the United States Department of Agriculture 1987-88 Nationwide Food Consumption Survey). The most productive author was Dr. Paul M. Finglas, the Head of the Food Databanks National Capability at the Quadram Institute (Norwich, UK) and the Managing Director of EuroFIR. His most cited paper among this dataset was about the development of an online Irish food composition database together with EuroFIR. The most productive institutions were the USDA and the World Health Organization (WHO) instead of universities. Flavonoid was the most recurring chemical class among the highly cited ones. The anti-oxidative properties and protective effects against heart disease and cancer of flavonoids might be some of the reasons for their popularity in research. Among the highly cited papers, the most heavily used FCDBs were the USDA database for the flavonoid content of selected foods, Fineli, the USDA National Nutrient Database for Standard Reference (USNDB), EuroFIR eBASIS-Bioactive Substances in Food Information Systems, and Phenol-Explorer. High-quality national and international FCDBs should be promoted and made more accessible to the research and public communities to promote better nutrition and public health on a global scale.

Nutritional quality and price of plant-based dairy and meat analogs in the Canadian food supply system

There has been an increased consumer interest and public health emphasis on plant-based protein foods, resulting in a rise in the availability of highly processed plant-based analogs. The objectives of this study were to assess the nutritional quality and the price of plant-based dairy and meat analogs compared to their respective animal-derived products and to examine the association between processing levels and the nutritional quality among these products. Using a branded food composition database, products in cheese, yogurt, milk, and meat categories were examined (n = 3231). Products were categorized as plant-based analogs versus animal-derived products using the ingredient list. Products were examined for their nutrient content, overall nutritional quality using the Food Standards Australia New Zealand nutrient profiling model, price, and processing levels using the NOVA classification. All plant-based analogs had lower protein and higher total carbohydrate, sugar, and fiber content compared to their respective animal-derived products. Compared to their respective animal-derived products, plant-based milk and meat analogs had lower energy, total fat, and saturated fat content; plant-based yogurt and meat analogs had lower sodium content; and all plant-based dairy analogs had lower calcium content. Plant-based cheese and yogurt analogs were more expensive than animal-based products; however, there was no significant difference among milk and meat products. There was no association between processing levels and overall nutritional quality among dairy and meat products. Plant-based analogs may be part of a healthy and affordable diet to reduce the intakes of nutrients of concern; however, additional compositional guidelines and/or labeling may be needed to highlight the differences in the levels of nutrients to encourage.

Nutritional information on the labels of processed and ultra-processed foods and beverages marketed in a supermarket chain in Lima in 2022

OBJECTIVES.

Motivation for the study. Peruvian Law No. 30021 establishes the use of warning octagons for foods with high content of critical nutrients (sugar, sodium, saturated and trans fats); however, the declaration of nutritional information is not mandatory. Main findings. Of a total of 4404 processed and ultra-processed foods marketed in supermarkets in Lima, only 71.4% declared some type of nutritional information. In addition, only 46.0% declared information on the content of critical nutrients regulated by Law No. 30021. Implications. There is a need for a mandatory and standardized declaration of nutritional information on packaged foods marketed in Peru, in order to allow the population to make healthy decisions when choosing their food and to monitor the correct use of warning octagons. . To estimate the number of processed and ultra-processed beverages and foods that provide nutritional information on their packaging, and to describe the characteristics of this information, as well as to determine the presence of nutritional information on products with octagons.

MATERIALS AND METHODS.

Photographs were taken of the labels of 4404 processed and ultra-processed beverages and foods marketed in supermarkets in Metropolitan Lima. The information on the label was collected and registered in the mobile and web version of the Food Label Information Program (FLIP). We analyzed variables related to the nutritional information, the way in which such information is declared and the information in beverages and foods with octagons.

RESULTS.

Only 71.4% of the products had some type of nutritional information. Of these, 13.8% provided the nutritional information as a text and not in a table, and only 56.3% declared it per 100 grams or milliliters. Of the total number of foods with the octagon "Contains trans fats", only 19.2% declared their content.

CONCLUSIONS.

More than a quarter of the beverages and packaged foods in the Peruvian market did not provide nutritional information of any kind, and of those that did, only one did so in different formats and units. In addition, we found that a proportion of beverages and foods for each type of octagon did not declare information of the nutrient that is mentioned in the octagon.

Natural language processing and machine learning approaches for food categorization and nutrition quality prediction compared with traditional methods

BACKGROUND

Food categorization and nutrient profiling are labor intensive, time consuming, and costly tasks, given the number of products and labels in large food composition databases and the dynamic food supply.

OBJECTIVES

This study used a pretrained language model and supervised machine learning to automate food category classification and nutrition quality score prediction based on manually coded and validated data, and compared prediction results with models using bag-of-words and structured nutrition facts as inputs for predictions.

METHODS

Food product information from University of Toronto Food Label Information and Price Database 2017 (n = 17,448) and University of Toronto Food Label Information and Price Database 2020 (n = 74,445) databases were used. Health Canada's Table of Reference Amounts (TRA) (24 categories and 172 subcategories) was used for food categorization and the Food Standards of Australia and New Zealand (FSANZ) nutrient profiling system was used for nutrition quality score evaluation. TRA categories and FSANZ scores were manually coded and validated by trained nutrition researchers. A modified pretrained sentence-Bidirectional Encoder Representations from Transformers model was used to encode unstructured text from food labels into lower-dimensional vector representations, followed by supervised machine learning algorithms (i.e., elastic net, k-Nearest Neighbors, and XGBoost) for multiclass classification and regression tasks.

RESULTS

Pretrained language model representations utilized by the XGBoost multiclass classification algorithm reached overall accuracy scores of 0.98 and 0.96 in predicting food TRA major and subcategories, outperforming bag-of-words methods. For FSANZ score prediction, our proposed method reached a similar prediction accuracy (R²: 0.87 and MSE: 14.4) compared with bag-of-words methods (R²: 0.72-0.84; MSE: 30.3-17.6), whereas structured nutrition facts machine learning model performed the best (R²: 0.98; MSE: 2.5). The pretrained language model had a higher generalizable ability on the external test datasets than bag-of-words methods.

CONCLUSIONS

Our automation achieved high accuracy in classifying food categories and predicting nutrition quality scores using text information found on food labels. This approach is effective and generalizable in a dynamic food environment, where large amounts of food label data can be obtained from websites.

The Expansion of the Hellenic Food Thesaurus; Allergens Labelling and Allergens-Free Claims on Greek Branded Food Products

Branded food composition databases (BFCDs) are valuable information tools that meet multiple user needs. Recently, recognising allergies and intolerances as an emerging concern for various stakeholders, BFCDs evolve to embed information on allergens. This study aims to expand the Greek BFCD, HelTH, to include allergen information for its 4002 products. A new file was added to the structure of HelTH, and data were curated to record label information. In 68.4% of products, at least one allergen was present in the ingredient list and in 38.9% at least one allergen in a precautionary statement. Milk (38.8%), gluten (32.7%), and soybeans (17.4%) were most commonly declared in the ingredient list; nuts (18.3%), eggs (13.1%), and milk (12.2%) were most commonly declared in precautionary statements. Allergen-free claims were present in 5.3% of the products and referred mostly on gluten and milk. In general, no statistically significant differences were identified between the nutritional composition of allergen-free claimed products and their equivalents. This study delivers an expanded BFCD that provides organised and detailed allergen information; new insights on the presence of food allergens in branded foods and issues of concern regarding allergen declaration that need to be addressed in order to improve label information.