Contents lists available at ScienceDirect

Food Chemistry

journal homepage: www.elsevier.com/locate/foodchem

Compliance of declared vs. analysed values with EU tolerance limits formandatory nutrients in prepacked foods

Tânia Gonçalves Albuquerquea,b, M. Antónia Nunesb, M. Beatriz P.P. Oliveirab,Helena S. Costaa,b,⁎

a Department of Food and Nutrition, National Institute of Health Dr. Ricardo Jorge, I.P., Lisbon, Portugalb REQUIMTE/LAQV, Dept. of Chemical Sciences, Faculty of Pharmacy, University of Porto, Porto, Portugal

A R T I C L E I N F O

Keywords:Processed foodsPrepackedNutrition labelAccuracyTolerances

A B S T R A C T

This study assessed compliance between declared and analysed values on prepacked foods, considering thetolerance limits for salt, fat and saturated fatty acids. Foods were distributed by food categories (e.g. snacks,ready-to-eat meals, potato and potato-products, bakery and pastry products) and a total of 209 products wereanalysed. Only half of the samples with a declared value ≥1.25 g/100 g were within tolerance limits for saltcontent. The lowest number of samples outside tolerance limits was observed for fat content; for saturated fattyacids, 27% of the samples were outside of tolerance limits. Only amongst “cereal products” were 100% ofproducts compliant for fat and salt declarations. It is of utmost importance that manufacturers update con-tinuously values declared for prepacked foods, because this information is crucial for consumers and food policy-makers as well as being a legal requirement.

1. Introduction

Some consumers are more aware of the relationship between dietand health, and their dietary choices, as well as a healthy lifestyle, andthese are increasingly important to them (Miller & Cassady, 2015). Infact, this is also a concern for policy-makers, because scientific evidenceindicates that healthy diets and lifestyles are important not only forindividuals, but impacts public healthcare expenses and productivity(Albert, 2010).

The European Strategy for Prevention and Control of Non-commu-nicable Diseases 2012–2016 recommends: (i) promoting healthy con-sumption through fiscal policies and marketing controls; (ii) elim-inating trans fatty acids from processed foods and replacing them withpolyunsaturated fats; and (iii) reducing diet salt intake to less than 5 g(2000mg sodium) per person per day (World Health Organization,2012). Thus, food authorities worldwide have established regulationsfor nutritional labelling, as a public health tool. In 1985, and subse-quently amended, the Codex Alimentarius Commission adopted Guide-lines on Nutritional Labelling, providing consumers with reliable in-formation about the nutritional profile of products on the market. Later,other countries developed mandatory nutritional labelling standardsand initiated approaches for specified foods and products bearing nu-trition and health claims, including the European Union (European

Parliament and Council of the European Union, 2011; CodexAlimentarius, 2013; Talati et al., 2017).

Codex Alimentarius aims to contribute to global harmonisation of thefood law and, ultimately, overcome the complexity of internationaltrade in food (Frohlich, 2017). Codex Alimentarius Commission has es-tablished guidelines and/or regulations on nutrition labelling to helpprotect consumers. More specifically, the purpose of label informationis to help consumers understand the nutritional profile and quality ofproducts and, hence, make better dietary decisions, based on the in-formation provided and public health advice (Cheftel, 2005; Kleef &Dagevos, 2015).

Nutritional labelling is a societal compromise, where food manu-factures are challenged to develop and/or reformulate products toachieve a favourable nutritional profile e.g. reduced salt or sugar con-tents (Hawley et al., 2013). In exchange, provided the industry ensuresnutritional labelling is accurate, the competitiveness of products in-creases (Kok & Radzi, 2017).

In Europe, increased awareness amongst some consumers and im-proved transparency by the food industry have, together, resulted in acomprehensive regulatory framework (Kasapila & Shaarani, 2016).Outside EU, different label requirements have been adopted in theUnited States and Canada, Mexico, Australia and New Zealand, China,and MERCOSUR Member States (Argentina, Brazil, Chile, Colombia,

https://doi.org/10.1016/j.foodchem.2019.125330Received 3 April 2018; Received in revised form 1 August 2019; Accepted 4 August 2019

⁎ Corresponding author at: Department of Food and Nutrition, National Institute of Health Dr. Ricardo Jorge, I.P., Av. Padre Cruz, 1649-016 Lisbon, Portugal.E-mail address: helena.costa@insa.min-saude.pt (H.S. Costa).

Food Chemistry 302 (2020) 125330

Available online 05 August 20190308-8146/ © 2019 Elsevier Ltd. All rights reserved.

T

Ecuador, Paraguay, and Uruguay); others follow Codex Standards(Kasapila & Shaarani, 2016). For instance, in Brazil and China, Codex-based nutrition labelling became mandatory in 2001 and January 2013,respectively (Block, Arisseto-Bragotto, & Feltes, 2017; Huang et al.,2016).

European legislation also sought to harmonise EU Member State(MS) laws to allow free circulation of products and maintain conditionswithin the internal market (Edinger, 2016). Regulation (EC) No. 178/2002 of the European Parliament and of the Council of 28 January 2002as General Food Law (Edinger, 2016) established a higher standard inprotection of human life and health within EU countries across foodsystems: production, processing and distribution (European Parliamentand of the Council of the European Union, 2002). However, the Eur-opean Commission (EC) identified irregularities in the presentation oflabel information amongst EU MS. Thus, on 25 October 2011, ECRegulation (EU) No. 1169/2011 was published on the provision of foodinformation to consumers, setting out common definitions, principlesand procedures (European Parliament and Council of the EuropeanUnion, 2011). Above all, the Regulation states that food informationshould not mislead consumers about the nature, characteristics, andfood effects (European Parliament and Council of the European Union,2011). This Regulation is applicable to all EU MS and has entered intoforce on 13 December 2016. Fig. 1 shows the mandatory information onlabels of prepacked foods and optional information concerning nutri-tion declarations.

Nutritional labelling should provide information about the compo-sition of products, allowing consumers to make an informed, healthychoice (European Parliament and Council of the European Union, 2011;Himmelsbach, Allen, & Francas, 2014). In accordance with the Reg-ulation (EU) No. 1169/2011 Article 31 Item 4, “declared values shall beaverage values based on: the manufacturer’s analysis of the food or a cal-culation from the known or actual average values of the ingredients used or acalculation from generally established and accepted data”.

The most accurate method to determine nutritional composition ischemical analysis, preferably undertaken by an accredited laboratorythat participates in proficiency testing schemes (Costa, Vasilopoulou,Trichopoulou, & Finglas, 2010; Machackova et al., 2018). However,

large numbers of new products, and the cost of time-consuming ana-lyses, as well as the demographics of the European food manufacturingbase, means most producers choose to calculate nutritional composition(Machackova et al., 2018), although few have the knowledge to un-dertake such a nuanced approach.

Foods are naturally subject to compositional changes due to avariety of factors from production to consumption. Thus, it is not pos-sible to guarantee the precise amounts of nutrients in any batch ofproducts. However, differences between declared values should notdeviate substantially from actual nutrient content in order to avoidmisleading the consumer (European Commission, 2012). Therefore, in2012, the EC released a guidance to EU MS authorities and food busi-ness operators on acceptable tolerances for nutrition labelling purposes(European Commission, 2012).

The aim of this work was to evaluate compliance between declaredand analysed values in prepacked foods, considering permitted toler-ance limits for salt, fat and saturated fatty acids (SFA).

2. Materials and methods

2.1. Overall design, sampling and data collection

Between 2015 and 2016, 209 prepacked food products available inPortugal were acquired from different stores (supermarkets, retailstores and take-away restaurants) (Albuquerque, Santos, Silva, Oliveira,& Costa, 2018). Whenever possible, foods from different brands andretailers were selected. On receiving the products at the laboratory, anExcel® spreadsheet was completed including the following information:(i) food identification and description; (ii) place, date and time of col-lection; (iii) list of ingredients; and (iv) nutrition declaration. Moreinformation is available in the Supplementary material.

The samples included prepacked foods from the major food cate-gories, namely snacks (n=13; corn chips, sweet and salty popcorn);fast-food products (n=15; burgers, pizzas); nuts and oilseeds (n=29);cereal products (n=12; granola, muesli, cereal bars, breakfast cereals);ready-to-eat meals (n=33); sauces (n=4); potato and potato-products(n= 20; potato chips, French fries); bakery products (n= 24;

Mandatory food

information

Name of the food

List of ingredients

Allergens

Quantity or categories of ingredients

Net quantity of the food

Data of minimum

durability or the 'use by'

date

Any special storage

conditions and/or

conditions of use

Name and address of

the business operator

Country of origin or place of

provenance

Instructions for use

Volume of alcohol

(beverages containing more than

1.2%)

Nutrition declaration

Nutrition declaration information

Optional

Mono-unsaturates; polyunsaturates; polyols; starch; f ibre; any of the

vitamins or minerals

Mandatory

Energy value; and the amounts of fat,

saturates, carbohydrate, sugars, protein and salt

Fig. 1. List of mandatory food information for prepacked foods and list of mandatory and optional nutrients for the nutrition declaration according to Regulation(EU) No. 1169/2011.

T.G. Albuquerque, et al. Food Chemistry 302 (2020) 125330

2

croissants, sweet and salty breads, cakes); and pastry products (n=59;cookies, biscuits and wafers). All the items were stored in accordancewith the label instructions.

A composite sample of approximately 1 kg was prepared using atleast three units. The selected products were homogenised in a blender(GM200, RETSCH, Germany) at 5000 rpm for 1 to 3min, depending onthe food matrix (Albuquerque et al., 2018). Then, samples were storedbased in their perishability (e.g. high perishable foods were frozen at−20 °C) until analysis.

2.2. Reagents and standards

All chemicals and reagents were of analytical grade. Hydrochloricacid, nitric acid, n-heptane, methanol, potassium hydroxide, potassiumferrocyanide, potassium thiocyanate, zinc acetate and silver nitratewere purchased from Merck (Darmstadt, Germany). Petroleum ether40-60° was purchased from Panreac (Barcelona, Spain). Fatty acidsmethyl esters standards (FAME Mix C4:0–C24:0, and Linoleic acidmethyl ester mix, Supelco®, Pennsylvania, CA, USA) were acquired atSigma-Aldrich (Madrid, Spain). Ultrapure water was prepared using aMilli-Q filter system (Millipore, Bedford, MA, USA).

2.3. Analytical methods

The salt content was determined using Charpentier-Volhard’s titra-tion (Albuquerque, Sanches-Silva, Santos, & Costa, 2012). This methoddetermines salt content based on the concentration of chloride ions ti-trated with silver nitrate solution. Sodium chloride was estimated fromthe amount of chloride ion, as determined by the end-point of titration.

Total fat was obtained using acid hydrolysis followed by extractionusing a Soxhlet apparatus (Soxtec™ 2050, Auto Fat Extraction System,FOSS Analytical, Hilleroed, Denmark) with petroleum ether 40–60°, asthe extraction solvent. The residue was dried for 1 h 30min at101 ± 2 °C, until constant weight, according to the acid hydrolysismethod (AOAC 948.15, 2000; Albuquerque, Santos, Sanches-Silva,Oliveira, Bento, & Costa, 2016).

To determine fatty acids composition, a cold transesterificationmethod with a methanol solution of potassium hydroxide (2M) fol-lowed by gas chromatography analysis was performed according to theISO 5509:2000 and Commission Regulation (EC) N° 796/2002, withslight modifications (Commission Regulation, 2002). This method isbased on the conversion of fatty acids naturally present in foods intofatty acids methyl esters (Albuquerque, Oliveira, Sanches-Silva, Bento,& Costa, 2016). For the conversion of fatty acid methyl esters into theirfatty acids, AOAC 996.06 was used (AOAC, 2000).

A gas chromatograph (Hewlett Packard 6890 series GC-Systems,Waldbronn, Germany) coupled to flame ionization detector was used. Astabilised polyphase (90% biscyanopropyl/10% cyanopropylphenyl si-loxane; 100m×0.25mm i.d., 0.2 µm film thickness) Supelco® 2560column (Supelco®, Sigma-Aldrich) was employed. Helium was used ascarrier gas at 1.0 mL/min flow rate. The split ratio was 1:50. The oventemperature was programmed as follows: 60 °C for 1min, then to 168 °Cat 17 °C/min, held for 28min, then to 235 °C at 4 °C/min and, finally,held for 15min. The peaks were identified by comparison of retentiontimes with their standards.

For quality assurance purposes, the following reference materialswere used: NIST® SRM® 1548a Typical Diet and NIST® SRM® 1546aMeat Homogenate. Each sample was analysed in triplicate and chro-matographic analyses performed in duplicate. Salt, fat and fatty acidscontents are expressed as g/100 g of edible portion on a fresh weightbasis.

2.4. Comparison of labelled and determined analytical values

Labelled or declared values for salt, fat and SFA (in g/100 g) wereobtained based on the manufacturers’ label declarations or from the

manufacturers’ websites. For salt, if the values were reported as sodium,these were multiplied by 2.5 to obtain the corresponding salt content(European Parliament and Council of the European Union, 2011).Considering the EC guidance document with respect to setting of tol-erances for nutrient values declared on product labels, the first step incomparison was to apply rounding guidelines to the declared values(European Commission, 2012). For example, if the value for fat was12.6 g, rounding rules state no decimals and, therefore, the declaredvalue should be 13 g. The second step was to calculate upper and lowerbounds of these rounded values. For example, continuing with theprevious value for fat (13 g), the lower and upper bounds are 12.5 g and13.4 g, respectively. The third step was to calculate lower and uppertolerances, considering established tolerances for foods (EuropeanCommission, 2012), including measurement uncertainty, since manu-facturers did not provide an uncertainty range for any of the selectedproducts. For a total fat declared value of between 10 g and 40 g per100 g of food, the tolerance is± 20%. Therefore, the analytical valuemust be 13 ± 20% (10 g and 16 g) per 100 g of food. The fourth stepwas to compare declared and analytical values in order to evaluatewhich were within or beyond tolerance limits. Finally, declared valueswere analysed with respect to any deviation. For example, if the labelvalue was greater than the analytical value, it was considered an over-estimation.

In the present study, samples were also grouped into nine categories(snacks; fast-food; nuts and oilseeds; ready-to-eat meals; cereal pro-ducts; sauces; potato and potato-products; bakery products; and pastryproducts), taking into account the main ingredients and features of theselected foodstuffs. A detailed analysis of compliance with EU tolerancelimits was also carried out.

3. Results and discussion

Most of the processed foods are recognised as sources of salt and fat,specifically SFA and trans fatty acids. Therefore, in recent years, greatattention has been devoted to foods of this type and different countrieshave performed a variety of actions to produce healthier foods, i.e. withlower salt and trans fatty acids contents (Mouratidou, Livaniou,Saborido, Wollgast, & Caldeira, 2014; World Health Organization,2013).

In our study, 209 food products were analysed for salt, fat and fattyacids (saturated, monounsaturated, polyunsaturated and trans fattyacids). Not all of these samples were used to evaluate compliance withEU tolerance limits since some values were missing. For instance, forsalt, 78.0% (n=163) of foods were checked while for total fat and SFA85.6% (n= 179) and 79.9% (n=167) of selected products, respec-tively, were considered. For some of these products the information wasmissing because nutrition declaration was voluntary before 13December 2016 and the products entered in the food chain before thisdate (European Parliament and Council of the European Union, 2011).

In this work, compliance between declared values and analyticalvalues was performed for salt, fat and SFA because these are mandatoryparameters of European nutritional labelling (European Parliament andCouncil of the European Union, 2011). Regarding trans fatty acids, itwas only at the end of 2018 that the Commission published a draftCommission Regulation, amending Annex III to Regulation (EC) No1925/2006 of the European Parliament and of the Council, consideringthe content of trans fat, other than trans fat occurring naturally in an-imal fat, stating foods intended for the final consumer shall not exceed2 g per 100 g of fat.

In 2011, the EC White Paper highlighted nutrients with negativeimpact on public health, such as saturated fat, sugars or sodium, andintroduced a mandatory standardised presentation (EuropeanParliament and Council of the European Union, 2011). However, Reg-ulation (EU) No. 1169/2011 adopted salt in preference to sodium, be-cause salt is more readily understood by consumers and does not re-quire any further manipulation (European Parliament and Council of

T.G. Albuquerque, et al. Food Chemistry 302 (2020) 125330

3

the European Union, 2011). Nutritional policies aimed at decreasingsalt intake have been growing. Indeed, Members of the World HealthOrganization (WHO), like Portugal, have agreed a global reduction insalt intake of a 30% by 2025 as a priority to reduce non-communicablediseases. To that end, mandatory targets, such as salt in bread, havebeen introduced in several countries and regions as well as programmesin partnership with the food industry to reduce the salt content of otherprocessed foods (Webster, Trieu, Dunford, & Hawkes, 2014).

The most accurate method for measuring salt concentrations ischemical analysis. Values in food products are useful for different sec-tors involved in reduction of salt intake, such as health professionals,governments, and food industry. They also help the industry identifyproducts for reformulation (Dunford et al., 2012) as well as stimulatingengagement in such task. Currently, processed foods are the primarysources of dietary salt. In fact, 75–80% of salt intake comes from pro-cessed products rather than salt added during cooking (10–15%) whilstnaturally occurring salt contributes only 5–10% of total dietary saltintake (Dötsch-Klerk, Goossens, Meijer, & Van Het Hof, 2015).

WHO recommends not more than 5 g of salt for adults (corre-sponding to 2 g of sodium) per day (Dötsch-Klerk et al., 2015) whilstRegulation (EU) No. 1169/2011 established 6 g of salt as daily re-ference intake for an adult (European Parliament and Council of theEuropean Union, 2011). Nevertheless, in many countries, the averagedaily salt consumption is estimated to vary between 9 and 12 g (Dötsch-Klerk et al., 2015). Amongst the samples collected for this study, thesalt content varied between<0.01 and 4.58 g/100 g (meanvalue=0.970 g/100 g). Most foods analysed (59%) had salt contentsless than 1 g/100 g. In the food categories, mean salt content increasedin the order: cereals (0.400 ± 0.2 g/100 g) > cookies, biscuits andwafers (0.718 ± 0.5 g/100 g) > bakery products (0.758 ± 0.5 g/100 g) > nuts and oilseeds (0.875 ± 0.8 g/100 g) > ready-to-eatmeals (1.03 ± 0.2 g/100 g) > potato and potato products(1.11 ± 0.8 g/100 g) > fast-food (1.31 ± 0.3 g/100 g) > sauces(1.97 ± 0.4 g/100 g) > snacks (2.09 ± 1.2 g/100 g).

Over-consumption of foods with high fat and SFA contents is asso-ciated with poor health outcomes (Kraak, Swinburn, Lawrence, &Harrison, 2014). Total fat contents ranged from 1.89 to 67.3 g/100 g(mean value= 21.0 g/100 g) with the highest values determined fornuts and oilseeds (42.2 ± 17.1 g/100 g), sauces (29.9 ± 6.4 g/100 g)and snacks (22.5 ± 8.6 g/100 g). However, predictably, nut and oil-seeds did not contain the most SFA (5.06 ± 2.3 g/100 g), which is whyit is important to evaluate the type of fat, i.e. in this case, nuts andoilseeds contain mainly unsaturated fatty acids that are linked withhealth benefits.

In 2012, the EU Framework for National Initiatives on SelectedNutrients introduced saturated fat as a target for reduction initiatives. Itwas proposed that intakes of saturated fat should be reduced by at least5% (in 4 years, 2016) and a further 5% by 2020 (Fuster, 2016). For anaverage adult, the reference SFA intake is 20 g and 70 g of total fat(European Parliament and Council of the European Union, 2011). For ahealthy diet, intakes of SFA should be as low as possible. In our study,around 62% of foods had a SFA content ≥4 g/100 g. Cookies, snacksand bakery products were the categories with the highest values, spe-cifically 8.52 ± 4.1, 7.83 ± 5.2 and 7.06 ± 3.4 g/100 g, respec-tively.

3.1. Compliance with EU tolerance limits

Tolerances refer to acceptable differences between nutrient valuesdeclared on the label and those established in the course of officialcontrols, as stated in Regulation (EU) No. 1169/2011 (EuropeanCommission, 2012). Tolerance limits make allowances for the naturalvariations in ingredients and final products, but the actual amountsmay not vary significantly from the value declared on the label(European Commission, 2012). According to the Guidance Documentthere are several factors to take in account when measured values are

outside the tolerance for the declared value, such as natural variation ofthe nutrient (e.g., by the influence of seasons), nutrient lability, andrates of degradation of nutrients in a specific food matrix (e.g., vitaminC) (European Commission, 2012). Other reasons to be considered areanalytical variability and sampling procedures (European Commission,2012).

In the EU, calculation methods are a legally acceptable process fornutritional labelling (European Parliament and Council of the EuropeanUnion, 2011). However, manufacturers can also have products analysedby laboratories, external or in-house. However, in 2017, Kok and Radzi(2017) presented reasons that can cause non-compliance between de-termined and labelled nutrient values, not least (1) methodologies ap-plied in the determination of the nutrients content, despite followingstandard AOAC methods and (2) samples preparation and homogeneity,which are of particular importance, since these can lead to false results,especially in canned foods, where the solids may be or not be separatedfrom the liquid in the nutritional analysis.

Overall, in our study, declared values for fat content were leastlikely to be non-compliant; 88% of products were in compliance for fatcompared to 74% for salt and 73% for SFA.

For salt, the established tolerance limits are± 0.375 g, if the foodhas a content< 1.25 g/100 g, and± 20% if the food has a salt content≥1.25 g/100 g (European Commission, 2012). With respect to the de-clared values for salt, 26% (42 out of 163) of the products were not incompliance with EU tolerance limits, i.e. on analysis, foods containedsalt concentrations that deviated significantly from declared values.More specifically, those with a declared content of ≥1.25 g of salt/100 g had the most products (52%) outside tolerance limits (Fig. 2).Considering the nutrients under analysis, all food categories had thehighest percentage of products within the tolerance limits for salt, ex-cept fast-food and nuts and oilseeds (Fig. 3).

Three tolerance limits have been set for fat content, based on de-clared values (European Commission, 2012). For instance, tolerancelimits for foods containing 10 g fat per 100 g are ± 2 g; for thoseproducts with fat contents 10–40 g/100 g or> 40 g/100 g, tolerancelimits are± 20% or± 8 g, respectively (European Commission, 2012).With respect to fat contents, samples selected for this study were allo-cated to one of three groups, according to their declared fat con-tent:< 10 g/100 g (n=39, 21.8%); 10–40 g/100 g (n=118; 65.9%);and> 40 g/100 g (n=22, 12.7%). Our results showed that foods withfat contents ranging from 10 to 40 g/100 g were most compliant (92%,108 out of 118). For the other groups,< 10 g and> 40 g/100 g, 18% offoods were outside tolerance limits (Fig. 2). The highest percentage offoods with fat values outside the tolerance limits were observed in fast-food (45%) followed by ready-to-eat meals (24%) and snacks (15%)while, for SFA, it was snacks (54%), fast-food (50%) and ready-to-eatmeals (38%) (Fig. 3).

The values declared for fat in all cereal products and sauces were inagreement within tolerance limits. This category of food has been underpublic scrutiny over recent years, due to their high content in addedsugars, saturated fat, and salt (Williams, 2014). Therefore, food busi-ness operators have been reducing these nutrients. For example, inbread, cheese, chips, soups, and sauces, salt reduction has been between10 and 40% (Kloss, Meyer, Graeve, & Vetter, 2015). The tolerance limitfor SFA is ± 0.8 g, if the declared value is less than 4 g/100 g, and±20%, if the declared value is greater than or equal to 4 g/100 g. In ourstudy, 110 of the foods (65.9%) had declared SFA contents ≥4 g/100 g.However, the percentage of declared values within and beyond toler-ance limits was very similar for both groups. For instance, 26% of foodswith declared SFA values< 4 g/100 g were outside the tolerance limitwhile, for the foods with a declared value ≥4 g/100 g, 27% were out-side the tolerance limit (Fig. 2).

3.2. Over- and underestimation within the declared values

Accuracy of nutritional labels is essential for consumers who want

T.G. Albuquerque, et al. Food Chemistry 302 (2020) 125330

4

to control energy and nutrients intakes (Jumpertz et al., 2013). Forinstance, individuals that are in a programme of weight loss and/ormaintenance, energy and fat intakes can be useful information(Jumpertz et al., 2013). Likewise, for salt, the accurate label nutritiondeclaration is extremely important for individuals that control the saltintake due, for example, to chronic diseases that require low sodiumdiets. Furthermore, in accordance with the Guidance Document, “fornutrients where consumers are generally interested in reducing their intakes(such as fats, sugars and salt/sodium), the declared values should not beestablished at the lower tolerance range” (European Commission, 2012).Therefore, ideally, declared values for salt, fat and SFA should be aboveanalytically determined and/or estimated and/or calculated values.

In our study, evaluation of over- or under-estimated declared valuesby food categories, showed that, for fast-food and potato and potato-products, all declared values were over-estimated, i.e. declared valuesfor salt, fat and SFA were higher than analytically determined values, asper recommendations. However, sauces had the highest percentage ofunder-estimated declared values for fat (100%) and salt (75%) while,for SFA, values for pastry products were under-estimated (44%).

For salt content, 63% (103 out of 163) of declared values were over-estimated in relation to analytically determined values, meaning they

are in line with recommendations (Fig. 4). Furthermore, in six out ofnine food categories, declared values for salt content were over-esti-mated (Fig. 5). However, the percentages of under-estimated valueswere 59% and 70% for foods with a declared salt contents< 1.25 g and≥1.25 g/100 g, respectively.

In our study, 58% (103 out of 179) of declared fat values wereunder-estimated. It is noticeable that, for fat values, all groups (fatcontent< 10 g; 10–40 g and>40 g) had the highest percentage offoods with declared values lower than amounts determined analyti-cally, especially those foods with declared values< 10 g/100 g (74% ofunder-estimated values). Only fast-food, potato and potato-products,and bakery products categories declared over-estimated values,meaning that analytically determined values of other food categorieswere higher than the declared values (Fig. 5), which is not in line withthe recommendations.

For the foods with SFA declared values less than 4 g/100 g, 61%were over-estimated while, for SFA content ≥4 g/100 g, this increasedto 72% (Fig. 4). All the categories in our study had over-estimated SFAdeclared values, as per recommendations.

The issues identified in our study are not unique to Europe. An as-sessment of the accuracy of nutritional labelling in 30 products present

Fig. 2. Prepacked processed foods for which the declared values for salt, fat and saturated fatty acids (SFA) are within or outside the EU tolerance limits.

Fig. 3. Prepacked processed foods by categories for which the declared values are within or outside the EU tolerance limits for salt, fat and saturated fatty acids(SFA).

T.G. Albuquerque, et al. Food Chemistry 302 (2020) 125330

5

in the Australia market was carried out in 2006 (Fabiansson, 2006).Significant differences between the analysed and the declared valueswere found, averaging between −13% and +61% (tolerance limitsare± 20%). A later study, by the Food Standards Australia NewZealand, examined 165 food products obtained from Australian retailoutlets and showed that 85% of products presented sodium contentsless than or within 20% of the label claim (Cunningham & Sobolewski,2011). In comparison, verification of label compliance for sodiumcontent of processed foods selected from Brazilian market showed highrates of non-compliance between chemical analysis and the labelledvalue. Indeed, 13 of the 17 products analysed had sodium contents thatexceeded the nutrition declaration by more than 20% (Ribeiro, Ribeiro,Vasconcelos, Andrade, & Stamford, 2013). Fitzpatrick et al. (2014)analysed the accuracy of Canadian food labels for calories, trans fat,saturated fat, and sugar and showed that actual sodium content of foodwas higher (≈18%) when compared to the declared values meaning anunderestimation trend of the declared values for that nutrient(Fitzpatrick et al., 2014).

4. Conclusions

WHO recognised product reformulation as a viable approach forfood industries to participate, with health organisations, in the reduc-tion of salt, sugar, and fat intakes. It is a challenge to reduce, for ex-ample product salt content, without compromising sensory character-istics and/or shelf-life and consumer acceptance and technologicalalternatives are limitations that have to be considered. Regardless, inthe EU, the legislative framework requires that information provided toconsumers is not misleading, which includes nutrient values declaredon packaging. A snapshot of actual fat, saturated fat, and salt contentsas well as compliance with EU tolerance limits between declared andanalysed values are presented in this work. Considering the results,additional efforts are still needed to ensure better compliance of de-clared values, especially for fast-food and snacks. Also, monitoring ofother food categories and through time is advisable. In the meantime,authors also recognize that the results reported are noteworthy forconsumers, public health organisations, and the food industry.

Fig. 4. Prepacked processed foods that have under-estimated or over-estimated declared values for salt, fat and saturated fatty acids (SFA) in comparison withanalytically determined values.

Fig. 5. Prepacked processed foods by categories that have under-estimated or over-estimated declared values for salt, fat and saturated fatty acids (SFA) in com-parison with analytically determined values.

T.G. Albuquerque, et al. Food Chemistry 302 (2020) 125330

6

Declaration of Competing Interest

The authors declare that they have no known competing financialinterests or personal relationships that could have appeared to influ-ence the work reported in this paper.

Acknowledgements

This work has been funded by National Institute of Health Dr.Ricardo Jorge, I.P., under the project PTranSALT (Reference number2012DAN828). This work was also supported by the project UID/QUI/50006/2013 – POCI/01/0145/FEDER/007265 with financial supportfrom FCT/MEC through national funds and co-financed by FEDER,under the Partnership Agreement PT2020. Tânia GonçalvesAlbuquerque acknowledges the PhD fellowship (SFRH/BD/99718/2014) funded by the FCT, FSE and MEC. Maria Antónia Nunes ac-knowledges the PhD fellowship (SFRH/BD/130131/2017) funded byFCT.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.foodchem.2019.125330.

References

Albert, J. (2010). Innovations in food labelling. The Food and Agriculture Organization of theUnited Nations and Woodhead Publishing Limited(1st ed.). WoodheadPublishinghttps://doi.org/10.1016/B978-1-84569-754-9.50027-9.

Albuquerque, T. G., Oliveira, M. B. P. P., Sanches-Silva, A., Bento, A. C., & Costa, H. S.(2016). The impact of cooking methods on the nutritional quality and safety ofchicken breaded nuggets. Food & Function, 7(6), 2736–2746. https://doi.org/10.1039/C6FO00353B.

Albuquerque, T. G., Sanches-Silva, A., Santos, L., & Costa, H. S. (2012). An update onpotato crisps contents of moisture, fat, salt and fatty acids (including trans -fattyacids) with special emphasis on new oils/fats used for frying. International Journal ofFood Sciences and Nutrition, 63, 713–717. https://doi.org/10.3109/09637486.2011.644768.

Albuquerque, T. G., Santos, F., Sanches-Silva, A., Oliveira, M. B., Bento, A. C., & Costa, H.S. (2016). Nutritional and phytochemical composition of Annona cherimolaMill. fruitsand by-products: Potential health benefits. Food Chemistry, 193, 187–195. https://doi.org/10.1016/j.foodchem.2014.06.044.

Albuquerque, T. G., Santos, J., Silva, M. A., Oliveira, M. B. P. P., & Costa, H. S. (2018). Anupdate on processed foods: Relationship between salt, saturated and trans fatty acidscontents. Food Chemistry, 267, 75–82. https://doi.org/10.1016/j.foodchem.2018.01.029.

AOAC (2000). Association of official analytical chemists. In Official methods of analysischemists (17th ed.). Washington, DC.

Block, J. M., Arisseto-Bragotto, A. P., & Feltes, M. M. C. (2017). Current policies in Brazilfor ensuring nutritional quality. Food Quality and Safety, 1(4), 275–288. https://doi.org/10.1093/fqsafe/fyx026.

Cheftel, J. C. (2005). Food and nutrition labelling in the European Union. Food Chemistry,93(3), 531–550. https://doi.org/10.1016/j.foodchem.2004.11.041.

Codex Alimentarius. (2013). Guidelines on Nutrition Labelling. Cac/Gl 2-1985, 53(9),1689–1699. DOI:10.1017/CBO9781107415324.004.

Commission Regulation (EC) N° 796/2002 of 6 May 2002 amending Regulation (EEC) No2568/91 on the characteristics of olive oil and olive-pomace oil and on the relevantmethods of analysis and the additional notes in the Annex to Council Regulation(EEC) No 2658/87 on the tariff and statistical nomenclature and on the CommonCustoms Tariff. Official Journal of the European Communities, L128, 8-28.

Costa, H. S., Vasilopoulou, E., Trichopoulou, A., & Finglas, P. (2010). New nutritionaldata on traditional foods for European food composition databases. European Journalof Clinical Nutrition, 64, S73–S81. https://doi.org/10.1038/ejcn.2010.215.

Cunningham, J., & Sobolewski, R. (2011). Food composition databases for nutrition la-belling: Experience from Australia. Journal of Food Composition and Analysis, 24,682–685. https://doi.org/10.1016/j.jfca.2010.07.002.

Dötsch-Klerk, M., Goossens, W., Meijer, G. W., & Van Het Hof, K. H. (2015). Reducing saltin food; Setting product-specific criteria aiming at a salt intake of 5 g per day.European Journal of Clinical Nutrition, 69(7), 799–804. https://doi.org/10.1038/ejcn.2015.5.

Dunford, E., Webster, J., Metzler, A. B., Czernichow, S., Mhurchu, C. N., Wolmarans, P., ...Neal, B. (2012). International collaborative project to compare and monitor the nu-tritional composition of processed foods. European Journal of Preventive Cardiology,19(6), 1326–1332. https://doi.org/10.1177/1741826711425777.

Edinger, W. H. (2016). Promoting educated consumer choices. Has EU food information

legislation finally matured? Journal of Consumer Policy, 39(1), 9–22. https://doi.org/10.1007/s10603-015-9307-3.

European Commission. (2012). Guidance document for competent authorities for thecontrol of compliance with EU legislation on: Regulation (EU) No. 1169/2011 of theEuropean Parliament and of the Council of 25 October 2011 on the provision of foodinformation to consumers.

European Parliament and Council of the European Union. (2011). Regulation (EU) No.1169/2011 of the European Parliament and of the Council of 25 October 2011 on theprovision of food information to consumers.

European Parliament and of the Council of the European Union. (2002). Regulation (EC)No. 178/2002 of 28 January 2002 laying down the general principles and require-ments of food law, establishing the European Food Safety Authority and laying downprocedures in matters of food safety. Official Journal of the European Communities,L31, 1–24. https://doi.org/2004R0726 - v.7 of 05.06.2013.

Fabiansson, S. U. (2006). Precision in nutritional information declarations on food labelsin Australia. Asia Pacific Journal of Clinical Nutrition, 15(4), 451–458.

Fitzpatrick, L., Arcand, J. A., L’Abbe, M., Deng, M., Duhaney, T., & Campbell, N. (2014).Accuracy of Canadian food labels for sodium content of food. Nutrients, 6(8),3326–3335. https://doi.org/10.3390/nu6083326.

Frohlich, X. (2017). The informational turn in food politics: The US FDA’s nutrition labelas information infrastructure. Social Studies of Science, 47(2), 145–171. https://doi.org/10.1177/0306312716671223.

Fuster, V. (2016). Unlikely allies: Partnering with the food industry to fight the obesityepidemic. Journal of the American College of Cardiology, 68(8), 871–873. https://doi.org/10.1016/j.jacc.2016.07.712.

Hawley, K. L., Roberto, C. A., Bragg, M. A., Liu, P. J., Schwartz, M. B., & Brownell, K. D.(2013). The science on front-of-package food labels. Public Health Nutrition, 16(3),430–439. https://doi.org/10.1017/S1368980012000754.

Himmelsbach, E., Allen, A., & Francas, M. (2014). Study on the impact of food in-formation on consumers’ decision making.

Huang, L., Neal, B., Dunflord, E., Wu, J. H. Y., Crino, M., & Trevena, H. (2016).Completeness of nutrient declarations and the average nutritional composition of pre-packaged foods in Beijing, China. Preventive Medicine Reports, 4, 397–403. https://doi.org/10.1016/j.pmedr.2016.08.002.

Jumpertz, R., Venti, C. A., Le, D. S., Michaels, J., Parrington, S., Krakoff, J., & Votruba, S.(2013). Food label accuracy of common snack foods. Obesity, 21(1), 164–169.https://doi.org/10.1002/oby.20185.

Kasapila, W., & Shaarani, S. M. (2016). Legislation - Impact and trends in nutrition la-beling: A global overview. Critical Reviews in Food Science and Nutrition, 56(1), 56–64.https://doi.org/10.1080/10408398.2012.710277.

Kleef, E. Van, & Dagevos, H. (2015). The growing role of front-of-pack nutrition profilelabeling: A consumer perspective on key issues and controversies. Critical Reviews inFood Science and Nutrition, 55(3), 291–303. https://doi.org/10.1080/10408398.2011.653018.

Kloss, L., Meyer, J. D., Graeve, L., & Vetter, W. (2015). Sodium intake and its reduction byfood reformulation in the European Union – A review. NFS Journal, 1, 9–19. https://doi.org/10.1016/j.nfs.2015.03.001.

Kok, S. C., & Radzi, C. W. J. M. (2017). Accuracy of nutrition labels of pre-packaged foodsin Malaysia. British Food Journal, 119(2), 230–241. https://doi.org/10.1108/JHOM-09-2016-0165.

Kraak, V. I., Swinburn, B., Lawrence, M., & Harrison, P. (2014). An accountability fra-mework to promote healthy food environments. Public Health Nutrition, 17(11),2467–2483. https://doi.org/10.1017/S1368980014000093.

Machackova, M., Giertlova, A., Porubska, J., Roe, M., Ramos, C., & Finglas, P. (2018).EuroFIR Guideline on calculation of nutrient content of foods for food business op-erators. Food Chemistry, 238, 35–41. https://doi.org/10.1016/j.foodchem.2017.03.103.

Miller, L. M. S., & Cassady, D. L. (2015). The effects of nutrition knowledge on food labeluse. A review of the literature. Appetite, 92, 207–216. https://doi.org/10.1016/j.appet.2015.05.029.

Mouratidou, T., Livaniou, A., Saborido, C. M., Wollgast, J., & Caldeira, S. (2014). Transfatty acids in Europe: Where do we stand? A synthesis of the evidence: 2003–2013.European Commission Joint Research Centre Science and Policy Report. https://doi.org/10.2788/1070.

Ribeiro, V. F., Ribeiro, M.de A., Vasconcelos, M. A. da S., Andrade, S. A. C., & Stamford, T.L. M. (2013). Processed foods aimed at children and adolescents: Sodium content,adequacy according to the dietary reference intakes and label compliance. Revista deNutrição, 26(4), 397–406. https://doi.org/10.1590/S1415-52732013000400002.

Talati, Z., Pettigrew, S., Neal, B., Dixon, H., Hughes, C., Kelly, B., & Miller, C. (2017).Consumers’ responses to health claims in the context of other on-pack nutrition in-formation: A systematic review. Nutrition Reviews, 75(4), 260–273. https://doi.org/10.1093/nutrit/nuw070.

Webster, J., Trieu, K., Dunford, E., & Hawkes, C. (2014). Target salt 2025: A globaloverview of national programs to encourage the food industry to reduce salt in foods.Nutrients, 6(8), 3274–3287. https://doi.org/10.3390/nu6083274.

Williams, P. G. (2014). The benefits of breakfast bereal bonsumption: A systematic reviewof the evidence base 1–4. American Society for Nutrition, 5, 636–673. https://doi.org/10.3945/an.114.006247.animal.

World Health Organization (2012). Action Plan for implementation of European Strategy forthe Prevention and Control of Noncommunicable Diseases 2012–2016. WHO RegionalOffice for Europe33.

World Health Organization (2013). Mapping salt reduction initiatives in the WHO EuropeanRegion. Copenhagen: WHO Publications.

T.G. Albuquerque, et al. Food Chemistry 302 (2020) 125330

7