Artigos

Conteúdo (article):

Adesão aos acordos voluntários de redução de sódio no Brasil
Adherence to voluntary sodium reduction agreements in Brazil

Introduction
Non-communicable diseases (NCDs) are the leading cause of death worldwide, being responsible for more than 40 million deaths in 2016, which is equivalent to 71% of all deaths1. The NCDs that contribute most to the morbidity and mortality burden have several common modifiable lifestyle risk factors, including harmful use of alcohol, smoking, physical inactivity and unhealthy diet1.
One of the diet-related factors is excessive sodium consumption, which is associated with high blood pressure and the development of cardiovascular disease (CVD)2, 3. Worldwide mean sodium intake is 10 g/day, which is double the amount recommended by the World Health Organization (WHO). Government programs to reduce population intake of sodium are a cost effective means of preventing CVD and premature deaths4. The WHO has prioritized sodium reduction to prevent and control NCDs, setting a target of a 30% relative reduction in mean population intake by 20255. The Pan American Health Organization (PAHO) has also recognized that sodium consumption is a priority, launching a series of recommendations to reduce intake to less than 2 g or 5 g salt per person per day by 20206.
The situation in Brazil in relation to sodium intake, prevalence of NCDs and NCD deaths is similar to global trends. In 2016, it was estimated that NCDs accounted for 74% of total deaths, 28% of which were caused by CVD1, while sodium intake was double the WHO recommended limit1, 7, 8. Although the main sources of sodium in the country are table salt and salt-based condiments, data from Brazil’s household budget surveys conducted between 2003 and 2009 have shown that the contribution of processed and ultra-processed foods to sodium availability has increased significantly8. A study using consumption data from a household budget survey of 34,000 Brazilians aged over 10 years in 2009 showed that processed and ultra-processed food products accounted for more than half of dietary sodium9.
In view of the above, since 2010, the Ministry of Health has been discussing strategies to reduce sodium intake, defining priority actions such as promoting the consumption of minimally processed staple foods, food education, guidance on nutrition labeling, and the reformulation of processed foods10. In 2011, the Ministry of Health and Brazilian Food Industry Association (ABIA) signed the first voluntary agreement aimed at reducing the sodium content of packaged foods sold in the country. Five agreements were signed between 2011 and 2017, setting sodium reduction targets for 35 food categories11-15. It is worth mentioning that these agreements are part of the strategy to reduce population salt consumption to less than 5 g/day by 2020, in line with PAHO and WHO targets11-15.
With the aim of providing evidence on sodium levels in packaged foods available on the Brazilian market and helping shape national policies to reduce population intake of this nutrient, this study assessed compliance with voluntary sodium reduction targets for packaged food products sold in the country and compared the amount of sodium in these foods to the limits recommended by the PAHO for the prevention of NCDs.
Methods
The data were collected between April and July 2017 by photographing the labels of foods sold by Brazil’s biggest supermarket chains in São Paulo and Salvador16. São Paulo was chosen because it is the country’s largest city. As one of the selected supermarket chains only has stores in the country’s Northeast region, the data from the products sold by this chain were collected in Salvador, the largest city in the region.
All stores owned by the respective supermarket chains in the selected locations were georeferenced and the socioeconomic status of the surrounding areas was determined based on the average family head income in census tracts within a 1,000-meter radius, using data from the latest census (2010)17. The stores were then divided into income terciles and the largest outlets in terms of area in the first and third terciles were included in the sample. Two outlets from each chain were visited, except for one chain, which stipulated that data could only be collected in its distribution centers, where all the products were available. Data collection followed the approach proposed by the International Network for Food and Obesity/Non-Communicable Diseases Research, Monitoring and Action Support (INFORMAS), described in detail by Kanter et al.18. Formal permission for data collection was obtained from the supermarket chains.
Photographs were taken of all sides of the packaging. The data were entered into the REDCap platform using a form developed by the University of North Carolina at Chapel Hill and Institute of Nutrition and Food Technology in Chile adapted for use in the present study. Duplicate products, different package sizes, packages that contained multiple individual items, and products without information were excluded, resulting in 11,434 products.
Only products from the 35 food categories included in the voluntary agreements signed before data collection were analyzed (Box 1). The agreement signed in 2017, setting new targets for instant noodles, sliced bread and mini bread rolls for 2018 and 2020, was therefore not considered15. We did not find any items in the categories ‘bread roll’, ‘roulade’ and ‘liquid/gelatinous stock’. The final sample comprised 1,553 items from 32 food categories.
The following descriptive statistics were used to analyze the data: means and standard deviations; 25th, 50th and 75th percentiles; and minimum and maximum sodium content in each food category. The Shapiro-Wilk test was used to determine whether the sodium content data were normally distributed, adopting a significance level of p ≤ 0.05. For most of the categories (20 of 32, or 62.5%), the data did not have a normal distribution. As the reduction targets were set based on mean sodium content, both mean and median sodium content are presented.
To assess adherence to the voluntary agreements, we calculated the prevalence of products with sodium content below the latest targets set for each respective category. As we collected all products, regardless of manufacturer, we analyzed whether brands with at least five products with sodium content above the targets belonged to companies that signed the agreements.
We also verified compliance with the PAHO Nutrient Profile Model, which establishes the following criterion for identifying products excessive in sodium: ≥ 1 mg of sodium per 1 kcal 19. This tool was developed to help PAHO member countries identify inappropriate nutrient profiles in non-alcoholic foods and beverages covered by regulations to prevent and control diet-related NCDs19.
We chose this criterion to assess the rigor of the voluntary agreements, as it is consistent with discussions on policies in the region and because, as far as we know, the PAHO and WHO have yet to develop specific thresholds for the reformulation of products containing sodium.
Level of agreement for the classification of the foods based on the thresholds proposed by the voluntary agreements and PAHO was determined using the Cohen’s kappa coefficient (k). The following classifications were used to assess the extent of agreement between the methods: poor, k<0.00; slight, 0.00 ≤ k ≤ 0.20; fair, 0.21 ≤ k ≤ 0.40; moderate, 0.41 ≤ k ≤ 0.60; substantial, 0.61 ≤ k ≤ 0.80; and almost perfect, k > 0.8020.
The analyses were performed using Stata version 16.0.
Results
Table 1 shows sodium content by food category. The categories with the highest sodium content were condiments [‘rice seasoning’ (mean 20,075.7; median 19,730.0 mg/100g), ‘stock cubes/powder’ (mean 20,029.7; median 20,505.3 mg/100g), ‘seasoning paste’ (mean 19,596.2; median 26,840.0 mg/100g), ‘other types of seasoning’ (mean 13,807.7; median 14,780.0 mg/100g)] and meat products [‘refrigerated mortadella’ (mean 1,443.5 ; median 1,320.0 mg/100g ), ‘mortadella kept at room temperature’ (mean 1,417.9; median 1,350.0 mg/100g), ‘hams’ (mean 1,329,9; median 1,160.0 mg/100g), ‘cooked sausage kept at room temperature’ (mean 1,193.6; median 1,187.0 mg/100g) and ‘refrigerated cooked sausage’ (mean 1,136.6; median 1,320.0 mg/100g)]. Mean sodium content was above the voluntary targets in 25.0% of the 32 categories (‘sponge cake mix’, ‘moist cake mix’, ‘salty corn snacks’, ‘savory cookies’, ‘cheese spread’, ‘refrigerated mortadella’, ‘mortadella kept at room temperature’ and ‘hams’). Median sodium content was above the targets in four categories (‘sponge cake mix’, ‘moist cake mix’, ‘uncooked sausage’ and ‘refrigerated mortadella’).
Table 2 shows the proportion of foods that met the limits proposed by the voluntary sodium reduction agreements and PAHO Nutrient Profile Model. Considering the overall sample, 77.7% met the voluntary sodium reduction targets. The categories with the largest proportion of products that met the voluntary targets were ‘mini bread rolls’ (100.0%) and ‘rice seasoning’ (100.0%), followed by ‘breakfast cereal’ (97.5%), ‘breaded foods’ (95.3%) and ‘mayonnaise’ (95.2%). The category with the smallest proportion of products that met the limits was ‘refrigerated mortadella’ (23.8%), followed by ‘sponge cake mix’ (29.5%), ‘moist cake mix’ (37.5%) and ‘uncooked sausage’ (45.0%). The products with sodium content above the limits belonged to more than 100 brands. Seventeen of these brands had at least five products that failed to comply with the limits. Together, these brands accounted for 52.7% of the items that that failed to comply (183 of 347). Seven of these brands belonged to companies linked to the food associations that signed the voluntary agreements. Of the remaining 10, four were supermarket brands.
Only 35.9% of the foods met the limits proposed by the PAHO. The category with the largest proportion of items within the limit was ‘moist cake mix’ (100.0%), followed by ‘sandwich cookies’ (99.0%), ‘sweet cookies’ (94.6%), ‘breakfast cereal’ (92.4%) and ‘sandwich cakes’ (91.3%). In almost a third of the categories (31.3%), none of the products were within the limits proposed by the PAHO Nutrient Profile Model, and in 16 categories (50.0%) the proportion of foods with appropriate sodium content was 50% or less.
Considering the overall sample, agreement between the two criteria was slight (kappa = 0.199). Only five categories showed substantial agreement or above (k>0.60):‘chips/potato sticks’, ‘sponge cake mix’, ‘cake without filling’, ‘sandwich cakes’ and ‘sweet cookies’. The large majority of the categories (23 of 32, or 71.9%) showed slight or poor agreement.

Discussion
Our findings show that around a quarter of the products belonging to the categories included in the voluntary agreements failed to meet the sodium reduction targets set to be achieved by 2017. An even larger proportion (64.1%) had a sodium content above the limit proposed by the PAHO Nutrient Profile Model. Agreement between the two criteria for the classification of products with high sodium content was slight, despite the fact that both criteria are aimed at preventing NCDs.
Recent studies have pointed to a progressive reduction in the sodium content of foods sold in Brazil and, overall, a large proportion of products comply with the voluntary targets. In a study assessing 20 categories included in the first voluntary agreements, Nilson et al. found a significant reduction in the mean sodium content of foods in 13 categories between 2011 and 2017, ranging from 8 to 34%21. Another study including the same food categories reported that more than 85% of the products analyzed in each category met the sodium targets between 2011 and 201322.
The difference between our results and those of these studies may be explained by methodological differences. For example, we included all the categories and products covered by the agreements and targets set to be achieved by 2017. Some of the categories in our study with a lower proportion of products that meet the targets, such as sausages, mortadellas and hams, were recently added to the agreements and were not assessed by Nilson et al., who analyzed food categories in agreements signed in 201122. However, comparisons of some of the same categories (‘cake without filling’, ‘sponge cake mix’, ‘moist cake mix’, ‘salty corn snacks’ and ‘savory cookies’, for example) show a lower level of compliance than that reported by Nilson et al. This difference may be at least partially explained by the fact that, unlike Nilson et al., we included all of the relevant products found in the supermarkets, regardless of manufacturer, and not just those produced by ABIA member companies 22. Our findings show that most of the brands that contributed most to the items that failed to meet the targets did not belong to companies that are members of the food associations that signed the agreements. Despite the expected spillover effects of these agreements on non-signatory companies, our findings show that these initiatives had less effect on this group than on ABIA member companies. Thus, the changes to the sodium content of foods on the market may not be as pronounced as previously shown by Nilson et al. Brazil, for example, was the country with the lowest proportion of products meeting the lower regional target among 14 Latin American and Caribbean Countries23.
The comparison of the voluntary agreements and PAHO Nutrient Profile Model indicated that the two methods were discordant when classifying whether the sodium content of a product is compatible with NCD prevention. The sodium reduction targets in the voluntary agreements were set based on mean values in each category after excluding outliers, leaving out a large proportion of the products that already met the targets. The Brazilian targets were set mainly focusing on products with high levels of sodium and are not capable of ensuring that products from different categories have appropriate nutrient profiles24. Salt reduction targets in the United Kingdom, for example, cover a larger number of food categories (around 80) and are generally more rigorous, particularly for meat products such as sausages and hamburgers25. Some of the included categories in the United Kingdom, such as pizzas and sandwiches, also contain high levels of sodium in Brazil26, yet are not covered by the agreements. Deficiencies in monitoring is a factor that limits the effectiveness of voluntary agreements in the prevention and control of NCDs. A study that evaluated technical reports published by Brazil’s National Health Surveillance Agency (ANVISA), which, together with the Ministry of Health, is responsible for monitoring reduction targets, showed that food categories lacked standardization and the number of samples and regions included in the process were insufficient27.
Various countries have adopted measures to reduce population intake of sodium28; however, estimates suggest that consumption exceeds recommended levels for the prevention of NCDs in all countries and regions29. Some countries, such as the United Kingdom and Finland, have been successful in implementing voluntary programs with the food industry and have managed to reduce salt levels, population intake, blood pressure levels, and associated diseases over recent decades30-32. Despite progress in Finland, which reduced population salt intake from 12 to around 9 g/day between 1979 and 2002, Laatikainen et al. estimate that, if the decrease in intake were to follow the same trend, it would take another 35 years for men and 26 years for women to achieve an intake of 5 g31. More recently, Argentina and South Africa introduced legislation limiting salt levels in a range of food categories. The advantages of mandatory reformulation include the possibility of imposing fiscal penalties, the fact that regulations apply to all manufacturers, and the stability of measures with changes of government28. However, critics of reformulations aimed at reducing the intake of critical nutrients such as sodium suggest that the strategy has important limitations. These include the fact that it is not possible to ensure that reformulated ultra-processed products have an appropriate nutrient profile and one of the policy rationales behind such measures is that reformulation is a means of reducing the consumption of these nutrients without changing dietary patterns, which can legitimate the consumption of these foods33.
Other strategies can be employed to reduce population intake of sodium besides reformulation. These include the use of taxes and adequate nutrient labelling, including front-of-package warnings indicating that a product is high in one or more nutrients of concern. In 2014, Mexico created a tax on sugary drinks and non-essential high energy density foods such as salty snacks, candies and breakfast cereals. A study showed that one year after the introduction of the measure, purchases of the taxed products had fallen34. In Hungry, a tax on unhealthy foods such as salty snacks was also effective in reducing the consumption of products in the food categories and promoting more healthy food choices, both due to price and increased awareness about the products35. The use of front-of-package warnings is also mentioned by the report “Salt-smart Americas: a guide for country-level action” as a useful complement to the reformulation of foods with the aim of reducing population intake of sodium6. In Latin America, Chile, Uruguay, Peru and Mexico have adopted front-of-package warning labels. Studies in Brazil demonstrate that the presence of warnings indicating that a product is high in critical nutrients can help consumers make more healthy food choices36, 37.
The present study stands out because of the size and broad scope of the sample of packaged foods. However, it does have some limitations. The number of products found in some food categories was small and the possibility of losses of some foods cannot be ruled out. However, previous studies also found a small number of foods in certain categories21, 22, indicating limited product variety. Another limitation is the fact that we did not include other types of food outlets. However, supermarkets are the most commonly used outlet for food shopping in Brazil and provide around 60% of the calories purchased for household consumption38. Furthermore, we used the information displayed by the manufacturers on the food labels, without performing laboratorial analyses to verify the stated content. Finally, we did not consider the market share of the products and were therefore unable to identify the sodium content of the most commonly consumed products in Brazil.
Considering that sodium intake in the country is more than twice the recommended threshold and the growing consumption of ultra-processed foods, more rigorous regulations need to be put in place in order to achieve Brazil’s population sodium intake reduction target. The deadline for the voluntary sodium reduction targets was 2020 and trends indicate that Brazil is far from achieving the desired population intake. We recommend that measures be expanded to cover more product categories, focusing especially on ultra-processed foods like pizzas and sandwiches, and the development of stricter targets as in other countries, ideally closer to the level proposed by the PAHO for the prevention of NCDs. Finally, it is important to highlight the limitations of voluntary agreements in comparison to mandatory regulations, meaning that priority should be given to the latter. Other measures such as high sodium content warning labels and education to raise public awareness of the use and consumption of salt should be implemented alongside these strategies in order to reduce population intake of sodium.
References

1. World Health Organization (WHO). Noncommunicable diseases country profiles 2018. Geneva: WHO; 2018.
2. Strazzullo P, D\'Elia L, Kandala NB, Cappuccio FP. Salt intake, stroke, and cardiovascular disease: meta-analysis of prospective studies. BMJ. 2009;339:b4567.
3. He FJ, Tan M, Ma Y, MacGregor GA. Salt Reduction to Prevent Hypertension and Cardiovascular Disease: JACC State-of-the-Art Review. J Am Coll Cardiol. 2020;75(6):632-47.
4. Webb M, Fahimi S, Singh GM, Khatibzadeh S, Micha R, Powles J, et al. Cost effectiveness of a government supported policy strategy to decrease sodium intake: global analysis across 183 nations. BMJ. 2017;356:i6699.
5. World Health Organization (WHO). Global action plan for the prevention and control of noncommunicable diseases 2013-2020. Geneva: WHO; 2013.
6. Pan American Health Organization (PAHO). Salt-Smart Americas: A Guide for Country-Level Action. Washington, DC: PAHO; 2013.
7. Mill JG, Malta DC, Machado Í, Pate A, Pereira CA, Jaime PC, et al. Estimation of salt intake in the Brazilian population: results from the 2013 National Health Survey. Rev Bras Epidemiol. 2019;22Suppl 02(Suppl 02):E190009.SUPL.2.
8. Sarno F, Claro RM, Levy RB, Bandoni DH, Monteiro CA. [Estimated sodium intake for the Brazilian population, 2008-2009]. Rev Saude Publica. 2013;47(3):571-8.
9. de Moura Souza A, Bezerra IN, Pereira RA, Peterson KE, Sichieri R. Dietary Sources of Sodium Intake in Brazil in 2008-2009. Journal of the Academy of Nutrition and Dietetics. 2013;113(10):1359-65.
10. Nilson EAF, Jaime PC, Resende DdO. Iniciativas desenvolvidas no Brasil para a redução do teor de sódio em alimentos processados. Rev Panam Salud Publica. 2012;32:6.
11. Brasil. Termo de Compromisso 004/2011 de 7 de abril de 2011. Estabelece as metas nacionais para a redução do teor de sódio em alimentos processados no Brasil. Brasília: Diário Oficial da União; 2011. p. 81.
12. Brasil. Termo de Compromisso 035/2011 de 13 de dezembro de 2011. Estabelece as metas nacionais para a redução do teor de sódio em alimentos processados no Brasil. Brasília: Diário Oficial da União; 2011.
13. Brasil. Termo de Compromisso de 05 de novembro de 2013. Estabelece as metas nacionais para a redução do teor de sódio em alimentos processados no Brasil. Brasília: Diário Oficial da União; 2013.
14. Brasil. Termo de Compromisso de 28 de agosto de 2012. Estabelece as metas nacionais para a redução do teor de sódio em alimentos processados no Brasil. Brasília: Diário Oficial da União; 2012. p. 124.
15. Brasil. Termo de Compromisso 005/2017 de 13 de junho de 2017. Estabelece as metas nacionais para a redução do teor de sódio em alimentos processados no Brasil. Brasília: Diário Oficial da União; 2017.
16. Euromonitor International. Grocery retailers in Brazil. 2016.
17. Instituto Brasileiro de Geografia e Estatística (IBGE). Censo demográfico 2010. Rio de Janeiro: IBGE; 2010.
18. Kanter R, Reyes M, Corvalán C. Photographic methods for measuring packaged food and beverage products in supermarkets2017. cdn.117.001016 p.
19. Pan American Health Organization (PAHO). Pan American Health Organization Nutrient Profile Model. Washington, DC: PAHO; 2016.
20. Watson PF, Petrie A. Method agreement analysis: A review of correct methodology. Theriogenology. 2010;73(9):1167-79.
21. Nilson EAF, Spaniol AM, Gonçalves VSS, Moura I, Silva SA, L\'Abbé M, et al. Sodium Reduction in Processed Foods in Brazil: Analysis of Food Categories and Voluntary Targets from 2011 to 2017. Nutrients. 2017;9(7).
22. Nilson EAF, Spaniol AM, Goncalves VSS, Oliveira ML, Campbell N, L\'Abbe M, et al. The impact of voluntary targets on the sodium content of processed foods in Brazil, 2011-2013. J Clin Hypertens (Greenwich). 2017;19(10):939-45.
23. Arcand J, Blanco-Metzler A, Benavides Aguilar K, L\'Abbe MR, Legetic B. Sodium Levels in Packaged Foods Sold in 14 Latin American and Caribbean Countries: A Food Label Analysis. Nutrients. 2019;11(2).
24. Martins APB. Redução de sódio em alimentos: uma análise dos acordos voluntários no Brasil. São Paulo: Instituto Brasileiro de Defesa do Consumidor; 2014.
25. Public Health England (PHE). Salt reduction targets for 2017. London: PHE; 2017.
26. de Moura Souza A, Bezerra IN, Pereira RA, Peterson KE, Sichieri R. Dietary sources of sodium intake in Brazil in 2008-2009. J Acad Nutr Diet. 2013;113(10):1359-65.
27. Martins APB, Andrade GC, Bandoni DH. Avaliação do monitoramento do teor de sódio em alimentos: uma análise comparativa com as metas de redução voluntárias no Brasil. Vigil sanit debate. 2015;3:9.
28. Webster J, Trieu K, Dunford E, Hawkes C. Target salt 2025: a global overview of national programs to encourage the food industry to reduce salt in foods. Nutrients. 2014;6(8):3274-87.
29. Powles J, Fahimi S, Micha R, Khatibzadeh S, Shi P, Ezzati M, et al. Global, regional and national sodium intakes in 1990 and 2010: a systematic analysis of 24 h urinary sodium excretion and dietary surveys worldwide. BMJ Open. 2013;3(12):e003733.
30. He FJ, MacGregor GA. A comprehensive review on salt and health and current experience of worldwide salt reduction programmes. J Hum Hypertens. 2009;23(6):363-84.
31. Laatikainen T, Pietinen P, Valsta L, Sundvall J, Reinivuo H, Tuomilehto J. Sodium in the Finnish diet: 20-year trends in urinary sodium excretion among the adult population. Eur J Clin Nutr. 2006;60(8):965-70.
32. Laverty AA, Kypridemos C, Seferidi P, Vamos EP, Pearson-Stuttard J, Collins B, et al. Quantifying the impact of the Public Health Responsibility Deal on salt intake, cardiovascular disease and gastric cancer burdens: interrupted time series and microsimulation study. J Epidemiol Community Health. 2019;73(9):881-7.
33. Scrinis G, Monteiro CA. Ultra-processed foods and the limits of product reformulation. Public Health Nutr. 2018;21(1):247-52.
34. Batis C, Rivera JA, Popkin BM, Taillie LS. First-Year Evaluation of Mexico\'s Tax on Nonessential Energy-Dense Foods: An Observational Study. PLoS Med. 2016;13(7):e1002057.
35. World Health Organization (WHO). Assessment of the impact of a public health product tax. Budapest: WHO; 2015.
36. de Morais Sato P, Mais LA, Khandpur N, Ulian MD, Bortoletto Martins AP, Garcia MT, et al. - Consumers’ opinions on warning labels on food packages: A qualitative study in Brazil. - Plos One. (- 6):- e0218813.
37. Khandpur N, de Morais Sato P, Mais LA, Bortoletto Martins AP, Spinillo CG, Garcia MT, et al. Are Front-of-Package Warning Labels More Effective at Communicating Nutrition Information than Traffic-Light Labels? A Randomized Controlled Experiment in a Brazilian Sample. Nutrients. 2018;10(6):688.
38. Machado PP, Claro RM, Canella DS, Sarti FM, Levy RB. Price and convenience: The influence of supermarkets on consumption of ultra-processed foods and beverages in Brazil. Appetite. 2017;116:381-8.