Sodium intake in Argentina has been estimated to be at least double the dose of 2000 mg/day recommended by WHO, mostly coming from processed foods. Argentina is one of the few countries in the world that have regulated sodium content in certain food products. This study presents an assessment of sodium content in a selection of food groups and categories as reported in the nutrient information panels. We surveyed 3674 food products, and the sodium content of 864 and 1375 of them was compared to the maximum levels according to the Argentinean law and the regional targets, respectively. All food categories presented high variability of sodium content. Over 90% of the products included in the national sodium reduction law were found to be compliant. Food groups with high median sodium, such as condiments, sauces and spreads, and fish and fish products, are not included in the national law. In turn, comparisons with the lower regional targets indicated that almost 50% of the products analyzed had sodium contents above the recommended values. This evidence suggests that enhancing sodium reduction in processed foods may be a necessity for public health objectives and it is also technically feasible in Argentina.
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.
26 September 2022 in Revista Panamericana de Salud Pública
Objectives. To identify the front-of-package scheme—Multiple Traffic Light (MTL), Nutri-Score (NS), and black octagon Warning System (WS)—most effective in reducing purchase intention and perceived product healthfulness of drinking yogurts, cookies and cheese spreads, and to assess the joint influence of nutrient claims on the effects. Methods. Randomized-controlled experiment. A within-subjects factorial design was used to evaluate the influence of three independent factors on perceived product healthfulness and purchase intention. A total of 704 adults in Argentina were shown three mock-up products and asked to indicate which product they would buy choosing between pairs of products from different categories and to rate product healthfulness on a 7-point Likert scale. Results. The WS was the most effective in reducing purchase intention in drinking yogurts (OR:0.16, 95%CI: 0.09;0.28), cookies (OR:0.10, 95%CI: 0.05;0.18) and cheese spreads (OR:0.10, 95%CI: 0.05;0.18), and the perception of healthfulness (WS Mean score 3,63 vs No label 4,24, p<0.001), regardless of the participants’ gender, age, and level of education. NS was inefficacious in reducing product healthfulness perception, while MTL had significantly increased how healthful the product was perceived (p<0.001). Nutrient claims increased purchase intention and perception of healthfulness, thus reducing the effectiveness of front-of-package labels (p<0.001). Conclusions. In line with growing evidence, our findings support that WS perform better than NS and MTL in reducing purchase intention and healthfulness perception of products with excessive amounts of critical nutrients associated with the greatest burden of diseases. Front-of-package WS are expected to facilitate the population in Argentina to make healthier decisions.
In 2015, the Pan American Health Organization (PAHO) published sodium targets for packaged foods, which included two distinct levels: one “regional” and one “lower” target. Changes to the sodium content of the food supply in Latin American Countries (LAC) has not been evaluated. A repeated cross-sectional study used food label data from 2015 (n = 3859) and 2018 (n = 5312) to determine changes in the proportion of packaged foods meeting the PAHO sodium targets and the distribution in the sodium content of foods in four LAC (Argentina, Costa Rica, Paraguay, Peru). Foods were classified into the 18 food categories in the PAHO targets. The proportion of foods meeting the regional targets increased from 82.9% to 89.3% between 2015 and 2018 (p < 0.001). Overall, 44.4% of categories had significant decreases in mean sodium content. Categories with a higher proportion of foods meeting the regional and lower targets in 2018 compared to 2015 (p < 0.05) were breaded meat and poultry, wet and dry soups, snacks, cakes, bread products, flavored cookies and crackers, and dry pasta and noodles. While positive progress has been made in reducing the sodium content of foods in LAC, sodium intakes in the region remain high. More stringent targets are required to support sodium reduction in LAC.
As children are particularly vulnerable to marketing, this study analyzes marketing techniques and health/nutrition claims in food packaging and evaluates the nutritional quality in three food categories: sweet biscuits, breakfast cereals and dairy-based desserts. This descriptive study analyzed marketing techniques and claims included in food packaging (n = 301) in one of the largest retailers in Argentina. Trained researchers coded data following an adapted version of the INFORMAS protocol to account for local food packaging regulations. Nutritional quality was assessed using both Pan American Health Organization Nutrient Profile (PAHO NPM) and the WHO Regional Office for Europe Nutrient Profile (WHO Euro NPM) models. Under the PAHO model, 87% (n 262) of the product sample presented excess content of at least one nutrient (“less healthy” products), and 91% (n = 273) should not be marketed to children according to the WHO Euro model. Almost 40% of less healthy food products displayed nutrition claims on their package. Characters or celebrity endorsements, which are particularly attractive to children, featured in 32% of less healthy products, being more frequent in less healthy food products than in healthier ones. Results indicate that packaging for food products with low nutritional value often includes powerful marketing elements in Argentina, which renders young children very vulnerable to obesogenic influence. Moreover, the real nutritional value of the products analyzed were often at odds with the health claims shown on its package. Food labeling policies must be improved in Argentina to guarantee people’s health protection against deceptive advertising.