*3.2. Food Classification According to Their Sodium Content*

As shown in Figure 1A, only 12.7% of foods complied with the conditions for the nutrient claim sodium-free and most of them belonged to G9 (non-alcoholic drinks), G10 (one type of ingredient) and G16 (sweets) (Table S4). Around one third of foods could be categorized as very low in sodium (32.4%), while 48.2% of foods were low in sodium (Figure 1A).

**Figure 1.** Foods in conformity with the nutrition claims regulated by the European Regulation No 1924/2006 [39] and the Codex Alimentarius [40] (**A**) or exceeding the NPMs thresholds (**B**) for sodium, by group.

Five groups had more than 90% of their items classified as low in sodium (G5, G9, G10, G12 and G16) (Figure 1A). Not surprisingly, none of the foods in G8 (meat) qualified for any nutrient claim on sodium. Only one could be classified as low in sodium in G14 (snacks), while less than 5% in G4 (cheese) and G7 (fish/seafood). G9 (non-alcoholic drinks) was the only group with a considerable proportion of sodium-free foods (58.9%), while for G10 (one type of ingredient) and G16 (sweets) values were below 40% (Figure 1A).

When the Nutrient Profile Models (NPMs) were applied, opposite results were obtained (Figure 1B). Near half of all foods were considered high in sodium according to the PAHO-NPM (47.2%), while around one third with the Chile-NPM (31.9%). More than 10% of all foods classified as high in sodium by both NPMs belonged to G7 (fish/seafood), G8 (meat) and G15 (snacks) each (Table S4). Only 1 and 3 foods out of 280 in the meat group (G8) was not considered high in sodium according to the PAHO-NPM and Chile-NPM respectively (Figure 1B). G4 (cheese), G7 (fish) and G14 (sauces) had also a very large proportion of foods high in sodium (Figure 1B). On the contrary, no food exceeded the maximum values in G12 (pasta), while only a few in G16 (sweets) (3 and 6 foods according to the PAHO-NPM and Chile-NPM respectively) (Figure 1B, Table S4).

It is interesting to note that the results obtained by applying both NPMs strongly differed in some groups (Figure 1B, Table S4). The κ statistics and the disagreement probability were used to compare both NPMs on foods (Table 4). The agreement was considered "substantial" for the total food database. However, important discrepancies were obtained for some food groups. On one hand, both NPMs were in accord for the G12 and had a "near perfect" agreement for G3, G6, G8 and G15. On the other hand, it was only "slight" for G2, G5, G7, G9, G13 and G14.


**Table 4.** Agreement between the PAHO-NPM [39] and the Chile-NPM [40].

<sup>1</sup> Values: 0 to 1. <sup>2</sup> Agreement was assessed using the κ statistic as follows: 0.01–0.20 'slight'; 0.21–0.40 'fair'; 0.41–0.60 'moderate'; 0.61–0.80 'substantial'; 0.81–0.99 'near perfect'. \* Agreement was assessed using the disagreement probability: >0.1 'substantial'; <0.1 'near perfect'; 0 'perfect'.
