Does the Nutri-Score 2023 algorithm provide consistent classification across European food supplies? A cross-sectional study using branded food composition data from 14 countries Authors: Chantal Julia, Théo Vasseur, Hélène Alexiou, Joline W J Beulens, Anette E Buyken, Torsten Bohn, Pauline Ducrot, Ann Katrin Engelbert, Marie-Noëlle Falquet, Esther Infanger, Benedikt Merz, Elisabeth H M Temme, Stefanie Vandevijvere Published: 2026 ABSTRACT AbstractIntroductionThe Nutri-Score is a summary graded and colour-coded front of pack nutritional label providing an overall assessment of the nutritional value of prepacked products. While its 2023 update has been shown to be consistent with food-based dietary guidelines in the countries in which it has been adopted, its application has not been investigated in a wider geographical area. The objective of this study was to describe the classification of products according to the Nutri-Score in a large number of European Union (EU) countries.MethodsBranded food composition from 14 European countries was obtained from Euromonitor. Food products were classified according to the Nutri-Score algorithm, based on their content in energy (kJ/100 g), saturated fats (g/100 g), sugars (g/100 g), salt (g/100 g), proteins (g/100 g), dietary fibres (g/100 g) and fruit, vegetables and legumes (%/100 g). The distribution of the 2023 and 2015 Nutri-Score classifications was described across countries and food categories.ResultsOverall, classification according to the 2023 Nutri-Score was consistent across 408 399 products from 14 countries and consistent with food-based dietary guidelines: >90% of frozen fruit and vegetables or >95% pulses were classified in the more favourable classes of the Nutri-Score (A/B), while >90% of discretionary foods such as chocolate confectionery and sweet biscuits were classified in the less favourable classes (D/E). Cross-country differences in classification reflected underlying variability in nutrient composition of products (eg, salt content in refined breads ranged from 1.05 to 1.23 g/100 g across countries).ConclusionThe Nutri-Score classification appeared consistent across a large geographical area, confirming its potential as an EU-wide front-of-package label. While some residual challenges remain for a limited number of specific product groups (eg, whole-grain pasta or rice), our findings indicate that the 2023 update represents a substantial improvement over the 2015 algorithm.Keywords: Public Health, Nutritive Value, Primary Prevention WHAT IS ALREADY KNOWN ON THIS TOPIC Nutri-Score is a front-of-package nutrition label providing a summary indicator of the nutritional value of prepacked foods. It has been shown to be consistent with food-based dietary guidelines in the countries in which it has been adopted. The 2023 update of the Nutri-Score algorithm is associated with an improved classification of foods according to food-based dietary guidelines. Nutri-Score is a front-of-package nutrition label providing a summary indicator of the nutritional value of prepacked foods. It has been shown to be consistent with food-based dietary guidelines in the countries in which it has been adopted. The 2023 update of the Nutri-Score algorithm is associated with an improved classification of foods according to food-based dietary guidelines. WHAT THIS STUDY ADDS This study extended the analysis of the classification of foods according to the Nutri-Score outside of the countries having adopted it as their official front-of-package nutritional label. Classification in 14 European countries showed a global consistency with food-based dietary guidelines and the ability to highlight variability in the nutritional values of prepacked food products. This study extended the analysis of the classification of foods according to the Nutri-Score outside of the countries having adopted it as their official front-of-package nutritional label. Classification in 14 European countries showed a global consistency with food-based dietary guidelines and the ability to highlight variability in the nutritional values of prepacked food products. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY This study highlights the ability of the Nutri-Score to be applied consistently across European countries, confirming its status as a main contender for a harmonised system in the European Union. This study highlights the ability of the Nutri-Score to be applied consistently across European countries, confirming its status as a main contender for a harmonised system in the European Union. Introduction Low diet quality is recognised as one of the leading causes of mortality and morbidity in the European region.1 While the overall burden of diseases associated with dietary risks has tended to decrease in volume since the 90s, its proportion has remained stable over the same period of time.2 As a modifiable behavioural risk factor, improvement of dietary choices has been a key target for public health policies over the last decades.3,4 Most countries in the European Union (EU) have implemented public health policies to improve population diets. However, the observed stability in the share of burden attributable to dietary risks questions the ability of these policies to lead to sustained change and highlights the necessity of engaging in additional and more efficient strategies.5 Indeed, despite the wide development and dissemination of food-based dietary guidelines (FBDGs) to the population for many years, adherence to these remains relatively low.6 This may partly be due to a focus of nutrition programmes to target individual behaviour rather than implementing public health policies targeting the creation of healthy food environments.5,7,8 Additionally, food environments have changed over time, with some studies showing a shift towards a higher accessibility of less healthy food options.9 Among the ‘best-buy’ policies to improve public health nutrition, interpretative front-of-pack nutrition labels (FoPLs) stand out as one of the cost-effective strategies,10,11 as they allow for a better understanding of the potential contribution of prepacked products to a healthy diet. Indeed, with the increase in availability, accessibility and affordability of processed products in Western countries, consumers may struggle to identify products with a favourable nutritional profile and integrate them within their diet. While FBDGs provide guidance on healthy dietary patterns, they usually focus on messages regarding overall food groups (eg,‘consume at least five fruit and vegetables a day’) and general frequencies of consumption over a day/week period of time. For prepacked foods, consumers require additional support for identifying nutritionally favourable food choices within a food group (eg, fresh, frozen, canned in syrup, included within breakfast cereals).12 FoPL aim at reducing the information gap by providing guidance as to the selection of individual foods within food groups. Additionally, they may incentivise manufacturers to innovate or reformulate foods towards products with a healthier nutrient profile. Both worldwide and in the EU region, FoPL policies have progressively been adopted at the national level13 and advice for harmonised FoPLs has been issued by the European Commission.14 In the European region, the Nutri-Score—a summary graded FoPL system—is one of the most widespread, as it has been adopted as the official voluntary scheme by six countries, that is: Belgium, France, Germany, Luxembourg, the Netherlands, Switzerland and its use was announced in Spain. Moreover, as many companies with multicountry reach adopted the system, Nutri-Score is present in markets even in countries where the government has not adopted it. As such, the Nutri-Score was identified by the EU Commission in its 2020 inception impact assessment as one of the potential options for a harmonised and mandatory FoPL for the EU.15 In 2023, the algorithm underpinning the Nutri-Score was updated from the one initially developed in 2015 to increase the consistency between the classification of products in the FoPL and the national FBDG of the countries having adopted the scheme.16 The update of the algorithm was supported by detailed analyses of the classification of products in national databases from Belgium, France, Germany and the Netherlands.16 Research showed that the updated version of the Nutri-Score algorithm had strengthened alignments with FBDG compared with the 2015 version.17,18 Considering the potential for the Nutri-Score to be adopted in other EU countries, and its current dissemination through multinational brands that have already adopted it, it appeared necessary to explore the consistency of the Nutri-Score classification across various markets in the EU. No study to date has examined the performance of the updated 2023 Nutri-Score algorithm across a wide range of EU markets using a harmonised branded food dataset. Additionally, while FBDG are largely similar in their key messages across countries,19 food markets may differ in the nutritional composition of available products, owing to cultural differences, potential taste preferences and company practices.20 The objective of the present study is therefore to describe the classification of products according to nutritional quality from a large number of EU countries with a diversified food market by applying the 2023 Nutri-Score algorithm, and to evaluate the consistency of the classification across countries as part of a validation exercise of the updated algorithm. Additionally, a secondary objective was to describe the differences in classifications observed between applying the 2015 and the 2023 Nutri-Score algorithms. Methods Data collection Branded food composition data was obtained from Euromonitor through a licensing agreement with Euromonitor International. Data from 2022 and 2023 from 14 EU countries were included (Austria, Belgium, Czech Republic, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Netherlands, Poland, Spain, Sweden). The data provided by Euromonitor was collected using the web scraping method from a varied selection of retailer websites in the various countries concerned. The data was validated by the Euromonitor team, using a proprietary algorithm to flag data containing inconsistent information, which was then checked manually before being accessed by the authors. Data extraction from Euromonitor included product name and characteristics (country, brand, supplier, retailer), nutrient composition (energy, fats, saturated fats, carbohydrates, sugars, fibres, proteins and salt) and list of ingredients (in text form). The data did not include fresh produce. Once data was accessed by the authors, additional quality checks were conducted (see below). Statistical analysis For comparison purposes foods were categorised to correspond to the harmonised classification of foods proposed by the Scientific Committee of the Nutri-Score in previous works (see online supplemental table 1).21,22 Some food categories were excluded entirely either because they fell outside the scope of the Nutri-Score (eg, baby food, milk formula) or because their declared nutritional composition for the purpose of the Nutri-Score included both products ‘as sold’ and ‘as prepared’ (eg, evaporated milk, dessert mixes), precluding a meaningful interpretative analysis of the food group. Additional food subcategories were created in order to explore the classification of Nutri-Score according to the specific nutritional composition for breads (three subcategories according to dietary fibre content: refined bread for products with <3 g dietary fibres/100 g, bread source of fibres for products with 3–<6 g fibres/100 g; bread rich in fibres for products with ≥6 g fibres/100 g) and breakfast cereals (three subcategories according to sugar content: Low sugar for products ≤5 g sugar/100 g, medium sugar for products 5–<13 g sugar/100 g, high sugar for products with ≥13 g sugar/100 g). Additionally, specific vegetable oils were identified and analysed separately (palm, olive, canola, etc), as well as various types of milk (skimmed for products with ≤0.5 g fat/100 g, semiskimmed for products 1.5–1.8 g fat/100 g, whole fat for products with >3.5 g fat/100 g—as per EU regulation23). Individual products were excluded in the case of: mismatch between categorisation and nutritional composition (eg, beverages classified as ‘non-sugar’, but with a substantial reported nutritional value for sugar); missing values for fat, saturated fats, carbohydrates, sugars, proteins or salt; and outlier values of nutritional composition (based on nutrient distribution in the database). Missing data for dietary fibre were imputed to the mean of the subcategory of product and verified through comparison with Ter Borg et al.24 The percentage of the ‘fruits, vegetables and legumes’ component was attributed following the method by Ter Borg et al,24 based on average content in subcategories of products. The presence of non-nutritive sweeteners was evaluated using the list of ingredients and subcategory of products. The Nutri-Score was calculated using both the 2015 and 2023 algorithms.16,21,22 Cross-tabulations of the Nutri-Score 2023 and 2015 classes according to categories and subcategories of foods were computed for each country and for the whole database. For individual country data, food categories with less than 10 products were not presented. Finally, distribution of nutritional values (energy, proteins, total sugar, saturated fats, dietary fibre, salt) were detailed for each food category and subcategory (mean, P5, P25, P50, P75, P95). Results Overall distribution of Nutri-Score classification Out of 575 015 food and non-alcoholic beverage products extracted by Euromonitor, exclusion of food categories not in the perimeter of the Nutri-Score or products with missing or outlier data for nutritional composition led to a final dataset of 408 399 products (figure 1). Exclusion rates differed across countries, ranging from 45.4% for Denmark to 16% in Greece (see online supplemental table 5). Flow chart of included data. National datasets included from a minimum of 6949 products for Greece to a maximum of 69 651 products for France (table 1). Table 1Number of food products included in the analysis of Euromonitor data (N=408 399)CountryNPercentageAustria88182.2Belgium20 3135.0Czech Republic27 7496.8Denmark74511.8Finland80492.0France69 65117.1Germany47 45411.6Greece69491.7Ireland11 7432.9Italy36 2568.9Netherlands33 4568.2Poland38 1559.3Spain56 63913.9Sweden35 7168.7Total408 399 Number of food products included in the analysis of Euromonitor data (N=408 399) The most represented food categories (across all countries) in the database were chocolate and chocolate confectionery (n=34 819), followed by sweet biscuits (n=25 917), savoury snacks (n=21 175) and processed red meat (n=18 409), highlighting the wide variety of foods available in these sectors of the market. Overall, the classification of products according to the Nutri-Score 2023 was consistent with dietary recommendations (tables 2–4 and detailed online supplemental tables 2–4 in Supplemental material). Foods corresponding to food categories recommended in FBDG (eg, fruits, vegetables, pulses) were largely classified in the most favourable classes of the Nutri-Score: >90% of frozen fruit and vegetable products were classified as A, as well as >95% of dry pulses and canned legumes. For these categories, no products reached the least favourable class of the Nutri-Score, and products classified in C or D classes were largely products with high contents in salt or sugar (eg, fruits in syrup of canned legumes with high levels of salt). Conversely, discretionary foods for which consumption should be limited in FBDG were mostly classified in the least favourable classes of the Nutri-Score: a large majority of sweet snacking products were classified as D or E (eg, >90% of products for chocolate confectionery, sweet biscuits, cakes, sweets and pies, colas). The limited share that reached the A or B categories were largely sugar-free confectionery and gums, or high-protein, low-sugar snack bars (sports/protein bars). Additionally, the Nutri-Score allowed for a good discrimination between products within the same food group category. Indeed, most categories of foods were distributed across two or more categories of the Nutri-Score: bread, breakfast cereals, ready meals or processed meat or poultry were classified across A to E classes of the Nutri-Score. Discrimination based on nutritional composition was particularly apparent in food categories for which subgroups were created based on nutrient composition: while low-sugar breakfast cereals were mainly classified as A (78.9%), high-sugar breakfast cereals were mainly classified as C/D classes; while bread rich in dietary fibre was mostly classified in A/B classes (65.0%), refined bread was mostly classified as C/D (93.8%); low-fat cheese was mostly classified as C (49.2%) while regular fat unprocessed cheese was mostly classified as D (79.6%); skimmed and semiskimmed milk were classified as B (>99%), whereas whole fat milk was classified as C (93.1%). Table 2Distribution of food products from main food categories according to the 2023 and 2015 NS for products in the main algorithm for general foods (selected food groups Euromonitor, Total number of products from detailed table, N=343 477)N2023 NS algorithmABCDE Cereal products and legumes Bread 8974 15.5817.6648.5216.991.25 Bread fibre source40034.4513.2960.0320.681.55 Bread rich in fibre341035.1929.7729.884.840.32 Refined bread15611.282.4359.7134.082.50Dried pasta14 26878.2215.133.951.750.95 Breakfast cereals 14 177 35.148.5327.5724.484.27 Breakfast cereals low sugar400378.946.678.295.450.65 Breakfast cereals medium sugar282642.4314.2328.8414.120.39 Breakfast cereals high sugar73488.487.3637.5938.847.73Rice445124.7866.057.391.260.52Canned legumes111397.840.541.350.270Dry pulses11391.1504.424.420 Fruit and vegetable products Frozen fruit25394.861.581.981.580Shelf stable fruit148187.173.854.054.660.27Shelf stable vegetables437477.484.879.306.361.99 Meat and fish products Processed red meat18 4091.081.1914.4129.2154.11Meat substitutes268933.4317.7427.8517.43.57Processed poultry641714.7713.0120.8836.4314.9Processed seafood15 71624.3216.8121.2426.6910.94 Ready meals Pizza46560.751.8043.952.381.16 Prepared salads 1915 11.599.9753.6322.772.04 Prepared salad meals17262.4910.6059.3925.262.26 Prepared salad greens18994.714.231.0600Readymade meals16 7066.5514.6854.0018.596.18Soups, sauce and dressingsSauces, dressings and condiments17 1868.133.1419.7332.9536.06Cooking sauces67411.884.8247.6021.224.49Soup63126.7049.2640.911.201.93 Snacking products Sugar confectionery17 2011.4914.554.5719.1160.28Chocolate confectionery34 8190.050.050.787.7891.34Ice cream78571.030.5715.3543.0140.04Sweet biscuits25 9171.171.517.1725.2064.95Honey196900.101.5727.3770.95Fruit snacks57462.782.0719.3462.8113.00Savoury snacks21 1755.444.2616.6752.8020.84Snack bars70534.100.9514.1855.6825.10Cakes, sweet pies and tarts49070.770.987.4829.2861.48Sweet spreads87791.561.6218.3662.0716.39Pastries23031.091.179.6433.8754.23 Dairy and plant-based products Dairy and plant-based desserts60457.595.7761.4418.596.60Plant-based yoghurt111118.7235.7339.245.670.63Yoghurt13 35819.6423.2152.643.501.01Fromage frais and quark325725.1510.8127.1423.2713.63Processed cheese35740.781.2013.4063.3521.26 Unprocessed cheese 13 986 0.840.5913.9973.3211.27 Unprocessed cheese low fat19685.744.1749.2435.215.64 Unprocessed cheese low salt84431.390.9822.3174.490.83 Unprocessed cheese regular fat12 0180.030.018.2179.5612.19 Unprocessed cheese regular salt5543001.3071.5327.17Condensed milk29404.0817.013.0675.85Bold entries correspond to higher-level food categories.NS, Nutri-Score. Distribution of food products from main food categories according to the 2023 and 2015 NS for products in the main algorithm for general foods (selected food groups Euromonitor, Total number of products from detailed table, N=343 477) Bold entries correspond to higher-level food categories. NS, Nutri-Score. Table 4Distribution of food products from main food categories according to the 2023 and 2015 NS for products in the algorithm for beverages (Euromonitor, N=39 625)N2023 NS algorithmABCDE Milk, fermented milk-based and plant-based beverages Cow milk 2784 72.9925.830.540.65 Cow milk skimmed51399.220.7800 Cow milk half skimmed115999.8300.170 Cow milk whole5942.6993.101.522.69Flavoured milks137020.5128.0324.7426.72Plant-based drinks438023.0633.8421.2121.89Goat milk13551.8545.931.480.74Drinkable yoghurt133423.0110.9422.7943.25 Water-based beverages Water621000000Other juices53580.9317.201.310.56Energy drinks17226.568.369.2975.78Flavoured water13409.7863.2814.6312.31Soda with fruits36670.8215.4926.2957.4Fruit nectars23344.5022.6226.7446.14Smoothies53934.3214.2839.7441.67Fruit juice410114.5661.7718.704.97Sodas other than colas47940.9616.1232.7750.15Reduced sugar cola carbonates with non-nutritional sweeteners78080.778.9710.26Colas13190.152.2713.9583.62Sports drinks5564.1436.1552.527.19Ready-to-drink tea17713.2243.8245.237.74Liquid concentrates9233.930.7729.7935.54Powder concentrates541.8533.3324.0740.74Ready-to-drink coffee93310.4031.7342.7715.11Asian specialty drinks400015.0085.00Bold entries correspond to higher-level food categories.NS, Nutri-Score. Distribution of food products from main food categories according to the 2023 and 2015 NS for products in the algorithm for beverages (Euromonitor, N=39 625) Bold entries correspond to higher-level food categories. NS, Nutri-Score. Differences between 2015 and 2023 Nutri-Score When comparing the classification between the 2023 Nutri-Score and the 2015 Nutri-Score, the results showed a better discrimination between products in closer alignment with FBDGs for the updated Nutri-Score (detailed online supplemental tables 2–4 in Supplemental material). For specific food categories, the 2023 classification shifted towards more favourable classes: vegetable oils—olive and canola oils reaching the B class of the Nutri-Score 2023; nuts and seeds shifted towards the A category. However, for most categories the Nutri-Score 2023 led to stricter classifications of products high in nutrients of concern, with reallocation of the modal frequency for some food categories: high-sugar breakfast cereals’ modal class was shifted from C to D, pizza from C to D, ready-to-eat meals from B to C, pastries from D to E, yoghurt from B to C, reduced sugar colas with non-nutritional sweeteners from B to C. Cross-country differences in food supplies The classification of products in the Nutri-Score 2023 across countries was largely consistent, in particular for food and non-alcoholic beverage products for which nutritional composition was not expected to vary: olive oil (consistently classified as B—table 3 and detailed online supplemental table 3 in Supplemental material), skimmed and semiskimmed milk (consistently classified as B—table 4 and detailed online supplemental table 4 in Supplemental material), dried pasta (consistently classified largely in A—table 2 and detailed online supplemental table 2 in Supplemental material), dried pulses (consistently classified as A—table 2 and detailed online supplemental table 2 in Supplemental material), rice (consistently classified in B—table 2 and detailed online supplemental table 2 in Supplemental material) or chocolate confectionery (consistently classified in E—table 2 and detailed online supplemental table 2 in Supplemental material). Table 3Distribution of food products from main food categories according to the 2023 and 2015 NS for products in the algorithm for fats, oils, nuts and seeds (Euromonitor, N=25 302)N2023 NS algorithmABCDE Fats and oils Butter152000.330.268.4990.92 Cooking fats31603.1616.7721.2058.86 Cream and alternatives195102.824.4182.1610.61 Margarine17510.112.0636.9528.3832.5 Olive oil33020.2196.882.360.480.06 Other oil2462022.9547.120.2829.65 Palm oil2000090.0010.00 Sunflower oil415021.6977.350.480.48 Rapeseed oil205093.666.3400 Corn oil590098.311.690 Nuts and seeds Nuts and seeds10 03039.7114.2533.6610.41.98 Nut/seed spreads326642.514.8526.5512.123.98NS, Nutri-Score. Distribution of food products from main food categories according to the 2023 and 2015 NS for products in the algorithm for fats, oils, nuts and seeds (Euromonitor, N=25 302) NS, Nutri-Score. For other foods, some differences appeared across countries, highlighting the variability in the nutritional composition of the food offer included in the database across countries. This was particularly apparent for bread products, for which compositions varied substantially across countries (see additional material). The proportion of breads rich in dietary fibres classified in the A category ranged from 53.2% in Denmark or 51.4% in Ireland and 35.5% in the Netherlands to 16.5% in Poland or 25% in Austria. The proportion of refined breads classified in the D category ranged from 65.6% in Greece or 40.6% in France to 20% in Finland or 15% in Ireland. These differences stemmed principally from the composition with respect to salt, energy and sugars, for which mean compositions and variability were higher in the countries in which classifications were less favourable (eg, average salt content in refined bread was 1.23±0.45 g/100 g in Greece and 1.05±0.29 g/100 g in Ireland; sugar content in refined bread was 7.15±5.43 g/100 g in France and 2.82±1.92 g/100 g in Finland—see online supplemental table 6). Similarly, 40.4% of ready-to-eat meals were classified as A/B classes of the Nutri-Score in Ireland while 38.7% were classified as D/E in the Netherlands. Average salt content in this category was 2.20±4.06 g/100 g in the Netherlands compared with 0.74±0.67 g/100 g in Ireland (see online supplemental table 3). Discussion The results of the classification of products according to the 2023 and 2015 Nutri-Score algorithm using the Euromonitor database is consistent with other reports and in particular with the previous publication by the Scientific Committee of the Nutri-Score based on a selection of nationally available branded food composition databases.16 Classification of foods from the various food categories such as breakfast cereals, snacks and sweets or beverages are largely similar across countries, underpinning the validity of the Nutri-Score outside of the four countries in which the Nutri-Score 2023 algorithm was tested (Belgium, France, Germany, the Netherlands). However, Nutri-Scores of some food categories such as bread products and ready-to-eat meals showed differences across countries, suggesting different compositions for the foods available in these categories over different European countries. These results are in line with previous studies in multiple EU countries using generic food composition data,25 branded food data26 or in analyses of individual country-level food composition data from branded foods on the market.17,18,24,27–32 Of note, except for the initial analysis from the scientific committee of the Nutri-Score regarding the 2023 update,16 previous multicountry analyses used the 2015 Nutri-Score algorithm.25,26 All of these previous studies reported similar Nutri-Score classifications, with favourable classes for fruit and vegetable products, legumes and whole-grain foods, and less favourable classes for discretionary foods, sugar-sweetened beverages and processed meat. Similarly, all reported wide variability in the nutrient profile of foods across categories, with several classes of the Nutri-Score represented within each food category, thereby underlying the usefulness of the Nutri-Score for consumers as a guide when choosing foods within a given food group. Studies comparing the 2023 updated classification in the Nutri-Score and the 2015 algorithm also showed similar results, with an overall more aligned classification of products with the FBDGs, in particular for products high in fat, sugars and salt and a more favourable classification for specific categories of products such as vegetable oils or nuts.17,18 Overall, multicountry comparisons and single country studies using the 2015 classification of the Nutri-Score concluded that the classification was mainly in accordance with FBDG,24,25,27,29,33 with some food groups for which consistency could be improved. In particular, the Nutri-Score classification was found to be mostly in agreement with the Mediterranean food pyramid,27,33 as well as Dutch,24 German29 or international25 FBDG. This ability to align overall with a wide range of FBDG underpins the consistency of these guidelines, which may have specific areas of focus depending on the country, but are based on the same core scientific corpus.19 Additionally, more recent single country studies focusing on the comparison between the 2015 and 2023 Nutri-Score concluded that the 2023 algorithm allowed for a higher consistency with FBDG, with the 2023 version addressing some limitations identified using the 2015 version.17,18,28,31 Previous limitations that have been improved with the 2023 version of the Nutri-Score include in particular the classification of artificially-sweetened beverages (shifted towards less favourable classes), of favourable vegetable oils (shifted towards more favourable classes) and a better discrimination between whole-grain and refined grain breads. Beyond these changes, further tightening of thresholds for nutrients of concern could be considered to increase discrimination in specific food categories, for example, stricter sugar thresholds could shift breakfast cereals towards less favourable Nutri-Score categories, but might also reduce discrimination across inherently less healthy product groups (eg, sweet biscuits mostly containing less fibre and/or more saturated fat than breakfast cereals). As such, the current algorithm allows an adequate balance between penalising high-sugar content and preserving discrimination across and within product categories. Similarly, the consistently favourable classification of dried pasta, including refined varieties, reflects a known limitation of the algorithm, as already highlighted by the Scientific Committee of the Nutri-Score, which examined additional modifications for this specific category but judged that the added complexity and unintended consequences on other food groups would outweigh the relatively limited gains in discrimination.1 This study confirms these previous works and extends these results to a wider range of European countries, showing a substantial improvement in the alignment of the 2023 Nutri-Score with FBDGs. In this sense, our work primarily contributes to the validation of the 2023 Nutri-Score as a single algorithm applied across different national food supplies, rather than to a country-by-country comparison of its alignment with specific FBDGs. Nonetheless, some residual challenges persist for a limited number of specific product groups. In line with recent national work from Norway, which identified remaining challenges such as limited differentiation between refined and whole-grain products, fat quality and fish, these findings suggest that country-specific or region-specific evaluations could usefully inform future refinements of the Nutri-Score 2023 to further improve its alignment with diverse FBDGs. Additionally, the Nutri-Score 2023 classification was able to reveal variations in the nutritional composition of the foods generally available across European countries. This variability is in line with results by Vin et al who identified differences in the nutritional composition of beverages and breakfast cereals between three European countries (Romania, Austria and France), with both differences in terms of the food offering between the countries but also differing nutrient contents for similar products.34 Indeed, beyond variability in the food offer that may be related to differences due to taste preferences or cultural differences and practices, concern has recently been raised in the EU over the potential ‘dual food quality’ of food products across countries, in particular between Western and Eastern countries in the EU.35–37 These concerns, initially raised by consumers, led to a series of systematic analyses of products in several EU countries by the EU Joint Research Centre.38,39 These analyses showed that though a limited proportion of products displaying identical packs had significantly different nutrient compositions (9% in the 2021 analysis), a substantial proportion of products with similar package designs had varying nutrient composition (22% in the 2021 analysis).38 Considering the initial purpose of the Nutri-Score to visualise differences in nutrient composition of foods within the respective food group, its implementation across the EU would potentially allow consumers to directly identify products falling under the ‘dual food quality’ terminology and therefore empower consumers to report them under the ‘Unfair Commercial Practices Directive’.40 Additionally, the Nutri-Score could be a tool for both reformulation in the industry and monitoring by public authorities in the EU region.41,42 The main strength of this study is its wide scope in terms of countries and food categories covered, including branded products as sold in the EU market. This study is among the first to evaluate the 2023 Nutri-Score algorithm using a dataset of this scale. However, limitations to the use of the Euromonitor database need to be acknowledged. First, although the Euromonitor database is intended to capture the food market, the representativeness of the set for each country compared with the overall food supply cannot be guaranteed. The datasets may under-represent or over-represent some food categories or retailers. This variability in the data sources and quality is apparent in the number of products available and then with complete data for the analysis depending on the country. Second, despite extensive data cleaning, the possibility that some misclassification of products occurred cannot be excluded. Misclassifications can be easily identified for products for which very low variability in nutrient composition is expected such as olive oil but this is more difficult for categories with high variability such as ready-to-eat meals. Also, some differences in classifications were observed between national databases explored beforehand by the Scientific Committee and the results obtained in Euromonitor. However, for the most part, distributions were consistent. A notable exception can be found for bread rich in dietary fibres, of which only 35.5% reached the A category in the Netherlands with the 2023 algorithm using the Euromonitor database, while in the national Dutch database underlying the 2023 analysis of this group, 89% of whole-grain bread were classified in the A category.16 Such differences may be related to the methodology for collecting and correcting data (selection of retailers and product categories). Indeed, the present analysis identified breads rich in dietary fibres content by the nutrient content, whereas the Dutch national database used ingredient lists to identify legally defined whole-grain products. Third, while nutritional data was mostly complete, data on ingredient lists was not so exhaustive. As such, it was not possible to extract detailed data from ingredient lists to compute the ‘Fruits, vegetables, legumes’, necessitating the use of reasonable assumptions of content for large food groups. Finally, the Euromonitor database, though very large, did not cover all food categories. In particular, the database did not cover fresh produce, in particular fresh fruit and vegetables, fish or meat. However, the database covered extensively prepacked products such as ready-to-eat meals or snacking products, for which the Nutri-Score is primarily intended. Overall, the present analysis confirms the validity of the Nutri-Score across multiple countries in the EU region, with both a good consistency with FBDG and a satisfactory discrimination of products according to their nutritional value. While some residual challenges remain for a limited number of specific product groups, our findings indicate that the 2023 update represents a substantial improvement over the 2015 algorithm. Taken together, these results support the expansion of the Nutri-Score outside of the countries currently having adopted it, and its main contender status for a harmonised and mandatory scheme for the EU. Supplementary material Acknowledgements We would like to particularly thank Manon Egnell (French Directorate of Health) for acting as Secretariat for the group, efficiently organising the workflow of the group (meeting agenda and minutes, logistical help in setting in-person and remote meetings). Footnotes Funding: The ScC members did not receive funding for the work conducted in the Scientific Committee of the Nutri-Score. The members of the Steering Committee of the Nutri-Score funded the data acquisition and analysis, with equal participation of all countries involved in the governance (Grant/Award Number Not Applicable). The funders had no role in the design, data collection, data analysis and reporting of this study. Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Ethics approval: Ethical approval was not required for this study, as the dataset comprised exclusively food composition data and did not involve human participants or personal data. Data availability free text: The authors cannot publicly share data as it is proprietary of Euromonitor International. Access can be made by purchasing data directly from Euromonitor at https://www.euromonitor.com/. While every attempt has been made to ensure accuracy and reliability, Euromonitor International cannot be held responsible for omissions or errors of historic figures or analyses. Patient and public involvement: Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research. Author note: Supplementary material is available at https://lookerstudio.google.com/s/lU17MqT96hY. Data availability statement Data may be obtained from a third party and are not publicly available. All data relevant to the study are included in the article or uploaded as supplementary information. References Associated Data Supplementary Materials Data Availability Statement Data may be obtained from a third party and are not publicly available. All data relevant to the study are included in the article or uploaded as supplementary information.