Essay: Non-communicable diseases

EXECUTIVE SUMMARY
‘ Non-communicable diseases are the leading cause of death on a global and national level. The main types of non-communicable diseases in the Netherlands, are responsible for a substantial proportion of the disease burden and mostly attributable to over-nutrition and poor diet quality.
‘ The majority of the food eaten is being processed or pre-prepared and high in energy, saturated fat, sugar and sodium.
‘ To improve the tremendous burden of disease, the government and food industry are under increasing pressure to enhance the quality of the food supply.
‘ Recently, joint ambitions of the Dutch government and the private sector have been formulated in an ‘agreement on improvements in product composition’, in which mandatory reformulation targets have been defined to be reached before the end of 2020. This agreement is in line with the existing initiative of the Choices programme, which has been stimulating product innovation in the Netherlands with product specific nutrient criteria.
‘ The National Institute for Public Health and the Environment (RIVM) is responsible for monitoring food composition and relies on national food composition tables. However, with the increasing amount of (innovated) processed foods, resources are limited to update these databases (at low costs).
‘ An international collaborative project can offer an opportunity to collect data on a detailed product level: The Global Food Monitoring Group has established a new (online) global branded food composition database(1)(1), with a standardized and sustainable methodology to collect accurate data at low cost. It aims to collate data on nutrient information (or lack thereof) from labels of packaged foods. This can be used to drive national (and international) category-wide changes in the composition of processed food products, which even if they are small, can deliver significant advances in public health.
‘ The overall goal of the current project was to collate and analyse data on saturated fat and sodium levels of meat products in the Netherlands, using the application and barcode scanning methodology from the Global Food Monitoring Group to asses changes over time and highlight the absence of certain data.
‘ Data was compared to data available from reformulation monitoring reports (2011 & 2014) and evaluated against the reformulation targets in the Netherlands that have been agreed for sodium and saturated fat levels of meat products and that should be met by January 1, 2015.
‘ Data were available for 863 meat products, 86% (n=745) displaying an Nutritional Information Panel (NIP).
‘ No progress was observed with sodium levels and little progress with saturated fat reductions in the Netherlands over the past 4 years in processed meat products. In the light of the reformulation covenant of 2014, more focus is needed if the Netherlands is to meet its national commitment to reduce sodium and saturated fat in meat products.
‘ All this information can be used to create more insight, motivate government and food industry and support efforts of reformulation.
‘ Conclusively, it is a versatile database. The database is up-to-date on nutrient information, but also on other label information (claims, logo’s, portion sizes, etc.) and independent from the industry for data collection. Data collection is easy and efficient and is internationally comparable (also over time). The database can stimulate manufactures to put a (correct) nutrient declaration on their product packaging, because the product label is the interface of a the industry towards the consumer. Finally, the database can be linked to a consumer application for smartphones, which stimulates consumers in making the healthier decision.
ABSTRACT
Aim: To collate and analyse data on nutrient information for meat products in the Netherlands, using a standardized methodology established by the Global Food Monitoring Group. The primary objective was to compare changes in levels of saturated fat and sodium from 2012-2015. The secondary objective was to evaluate reformulation targets for sodium and saturated fat levels that should have been met by January 1, 2015.
Methodology: Data was collected by photographing the Nutritional Information Panels (NIPs) from product labels of processed foods in-store using The George Institute’s Data Collector smartphone application. These photos were uploaded into a central database and data were entered and checked manually. Data were then cleaned and analysed. Mean levels of sodium and saturated fat were calculated and compared (ANOVA) with data available from reformulation monitoring reports and to the reformulation targets.
Results: Data were collected for 911 meat products. After data checking, information was available for 863 meat products, 86% (n=745) displaying an Nutritional Information Panel (NIP). Sodium levels in 2015 were not significantly different in any of the subcategories of meat products compared to previous years. Mean levels of saturated fat were significantly different within the subcategory combined heated meat products (10 g/100g (SD=3) in 2014 versus 8 g/100g (SD=3) in 2015; P<0.001). When comparing the newest 2015 data with the reformulation targets, 14%-93% of product sodium levels complied with these targets. Saturated fat percentages of complying products varied between 25% and 88%. Conclusion: We observed no progress with sodium and little progress with saturated fat reductions in the Netherlands over the past 4 years in processed meat products. In the light of the reformulation covenant of 2014, more focus is needed if the Netherlands is to meet its national commitment to reduce sodium and saturated fat in meat products. This method of evaluation should also be used for other product categories to monitor progress and to ultimately decrease the burden of nutrition-associated disease in the population in the Netherlands. ' 1. INTRODUCTION Non-communicable diseases (NCD's), also known as chronic diseases, are globally the leading cause of premature death (63% of all reported deaths in the world compared to 89% in the Netherlands)(2, 3). Next to the main types of NCD's in the Netherlands, respectively cancer and cardiovascular disease, obesity and (type 2) diabetes 'are responsible for a substantial proportion of the disease burden in the Netherlands. These are mostly attributable to poor lifestyle with one of the major causes: over-nutrition and poor diet quality (3-5). This has led to a more focussed attention on the quality of processed foods (6).The majority of the food eaten in the Netherlands (and other developed countries) is being processed or pre-prepared by the food industry (7). Nowadays, the majority of the population is exposed to foods high in energy, saturated fat, sugar and sodium (4). The industry is an important stakeholder in the food eaten by the population and has an important role to play in public health (8) (Appendix 1). To improve the tremendous burden of disease, the government and food industry are under increasing pressure to enhance the quality of the food supply (9). Since 2006, the Choices programme has been stimulating product innovation in the Netherlands (10): products from participating food companies carry a health logo if their products comply with nutrient criteria for sodium, trans fat, saturated fat, caloric content and added sugars (11-13). At the same time the Dutch government (Ministry of Health, Welfare and Sports, VWS) has been urging the food industry to improve their products by lowering salt, saturated fat, total sugar and calories. Recently, joint ambitions of the government and the private sector have been formulated in an 'agreement on improvements in product composition' (In Dutch: Akkoord Verbetering Productsamenstelling) (14), in which mandatory reformulation targets have been defined to be reached before the end of 2020. For meat products, agreed targets are defined: sodium levels have to be reduced with 10% for the product categories single heated, combined heated and combined raw meat products and saturated fat levels have to be reduced with 5% for (a part of) the product category combined heated, by the beginning of 2015 (15). Simultaneously a new European Regulation on the provision of food information to consumers has been implemented. According to this Regulation (EU) No. 1169/2011 (16), the nutrient content of manufactured foods, which are sold online, also have to be made available in electronic format (online). Mandatory on pack nutrition declaration includes energy value, the amount of fat, saturated fat, carbohydrate, sugars, protein and salt (Art. 30.1. a,b.)(16). This Regulation offers consistency and transparency in food labelling. The National Institute for Public Health and the Environment (RIVM) is responsible for monitoring and research on food composition and consumption. Traditionally, monitoring relies on national food composition tables (FCT's). However, there is a difference between the wide array of products available to consumers and what is captured in the national databases due to limited resources to update them (8). The current FCT's do not capture the total width of the overall products being offered for sale, consist of aggregated average compositions and are relatively expensive and time-consuming to update (17). New information technology brings opportunities to collect a wide range of data and in order to improve and/or simplify the accuracy and adequacy of the monitoring progress at product level. 1.1 Global Food Monitoring Group An international collaborative project that can offer such an opportunity to collect data on a detailed product level is being led by The George Institute for Global Health. The Global Food Monitoring Group (GFMG) is an international collaboration which has established a new (online) global branded food composition database (1), with a standardized and sustainable methodology to collect a large amount of food label data at low cost (8). It aims to collate data on nutrient information (or lack thereof) from labels of packaged foods. This can be used to drive national and international category-wide changes in the composition of processed food products, which even if they are small, can deliver potential health gains. More extensive information can be found in Appendix 2. 1.2 Overall goal and objective The overall goal of the current project was to collate and analyse data on nutrient information (or lack thereof) for packaged foods in the Netherlands, using a standardized methodology from the GFMG, that can be used to drive national improvements in the (processed) food supply. Focus of the present study was on meat products. The first objective was to analyse and compare new data (2015) to data available from reformulation monitoring reports (2011 & 2014) to assess changes over time (18, 19). The second objective is the evaluation of reformulation targets in the Netherlands that have been agreed for sodium and saturated fat levels of meat products and that should be met by January 1, 2015 (15). 2. METHODOLOGY The used methodology was based on an existing protocol for processed foods from the Global Food Monitoring Group (20). 2.1 Data sources and data collection The food category to be collected, in this case meat products, was defined in collaboration with the stakeholders involved in monitoring product improvement. The Ministry of Health, Welfare and Sports (VWS), The National Institute for Public Health and the Environment (RIVM) and the Choices Programme are all involved in the current food composition databases available in the Netherlands to monitor food and play an important role in the populations health. Data were obtained with a barcode scanning technology and photographing the Nutritional Information Panels (NIPs) on product labels from processed foods in-store with an IPhone (iOS), using the data collection app (version 1.1) of the George Institute (figure 1). Fig. 1 Data collection process of the George Institute for Global Health, Australia. During 6 days in March 2015, all relevant labelling information from the selected product group was collected. To ensure that the collected data were a good representation of the offered supply, permission to collect the data was asked for the two largest supermarket chains in the Netherlands (together the two chains cover 41% of the Grocery Retailers Company Shares in The Netherlands (2014) (21)). All available data of premium brand meat products were collected. Available data from 'supermarkets own label' (private label) were collected for approximately 99%. After data collection, the photos were uploaded into the online, password-protected, central branded food composition database of The George Institute (22) and entered manually. When photographs in the database turned out to be unreadable (<1%), data was obtained from websites were the supermarkets provide nutritional information for their products. The key variables that have been selected are divided into primary and secondary variables. These are the outlines of minimal information that should be collected for each product in order to monitor changes. For each product, all variables are sought, as shown in table 1. All missing data were recorded as such. Products which only display product and company name without nutritional values, were still included to highlight the absence of detailed data. Table 1. Variables to be collected per product and their format. Variable Format Primary outcome measures Brand name Product name GTIN (barcode) Saturated fat Sodium Salt Food category (major, minor, sub-minor levels) Type of Nutrition Information Panel (NIP) Choices logo* Presence Guideline Daily Amounts (GDA) Secondary outcome measures Serving size and unit Servings per pack Unprepared form of food (concentrate) Energy Protein Total fat Unsaturated fat Mono-saturated fat Poly-saturated fat Trans-fat Carbohydrate Total sugars Fibre Notes Allergen statements Gluten status Vegetarian symbol Ingredients list As per product label As per product label Barcode number per product Grams/100 grams or 100 millilitres Milligrams/100 grams or 100 millilitres Grams/100 grams or 100 millilitres As defined NIP, No NIP, US NIP, Multiple NIP's Present or absent: Choices thick on packaging Present or absent: GDA on packaging Grams or millilitres Amount of servings Kilojoules or Kilocalories/100 grams or 100 millilitres True or false: NIP for unprepared form of food Grams/100 grams or 100 millilitres Grams/100 grams or 100 millilitres Grams/100 grams or 100 millilitres Grams/100 grams or 100 millilitres Grams/100 grams or 100 millilitres Grams/100 grams or 100 millilitres Grams/100 grams or 100 millilitres Grams/100 grams or 100 millilitres Grams/100 grams or 100 millilitres Product comments, symbols, health or nutrition claims Listing of allergens on label Present or absent: gluten-free declaration on packaging Present or absent: vegetarian declaration on packaging Listing of ingredients on the label * Front-of-pack healthy choice logo from the international Choices programme (23). After entering the products, all data were screened by three researchers to identify outliers by comparing the data against the original source. Products were checked for correct product categorisation, typing errors and correct nutritional values. Some products (n=48) were removed from the dataset during this process for various reasons such as: duplicates, products with unclear or incomplete data or products that appeared not to belong to one of the selected product categories. When all data were checked, an excel-data spreadsheet with all entered data was extracted from the database, by the central management of The George Institute for further analysis. 2.2 Categorization of foods Data were categorized into a hierarchical structure of food groups, categories and subcategories as suggested by the Global Food Monitoring Group protocol. However, for the Netherlands this system was adapted to better enable meaningful comparisons with data from previous surveys. This analysis included only meat products and their alternatives, meat preparations (fresh processed meat products e.g. ' fresh raw beef burgers', 'fried fresh sausages') were not included in this category. The meat products were evaluated in the following manner for classification: prepared products underwent a heat treatment (e.g. cooked, boiled, grilled, fried, baked, hot smoked or through hot steam). Raw meat products did not have a heat treatment, they could be smoked (cold), dried, salted, fermented or a combination of the preceding. Within products where a muscular structure could be seen, the product was singular, regardless the meat content or added additives. Products without visual muscular structure belonged to the combined meat products, regardless of the meat content or added additives. In table 2 the categorization is shown. Table 2. Classification meat products. Examples Meat products and meat preparations Meat products Heated (prepared) meat products Single, heated For consumption bread: Ham, chicken breast, york ham, grilled bacon Combined, heated (Hausmacher) liver sausage, p??t??, Berliner, luncheon meat Raw meat products Single, raw Raw ham, carpaccio, bacon Combined, raw Filet americain, salami, chorizo Meat alternatives Vegetarian liver sausage, vegetarian filet americain, vegetarian smoked bacon Meat preparations * - - * This category is not included in this study. 2.3 Analysing and reporting data 2.3.1 Data cleaning Before all data were analysed, a second cleaning step was performed. Data were checked for possible errors and corrected where needed. This included odd averages, high standard deviations and wide ranges. To detect these inaccurate records from the dataset, all product groups were analysed separately (SPSS) into descriptive statistics with mean levels, standard deviations (SD) and ranges. Outliers and missing values were identified for each subcategory and amended with a verification against the original NIP or information provided on the supermarket's or manufacturer's website. This process was repeated until no errors were found. In total, eleven meat products were identified as erroneous and corrected. During this step, no meat products have been removed from the dataset. 2.3.2 Data analysis Analyses initially focused on the primary outcome measures: salt/ sodium, saturated fat and the front-of-pack (FOP) communications GDA and the Choices logo. All numeric values were determined by calculating the mean level, standard deviation, median and range with the statistical software SPSS (IBM statistics 20.0). The baseline and progress measurements for sodium and saturated fat of the meat products were collected from the RIVM reports in 2012 and 2014 (18, 19). These data were based on the data from the Netherlands food composition database (NEVO) 2011, 2013 and up on to 2014. Significant differences (P<0.05) between the composition of the new data (GFM 2015) compared to the already available data were tested with an ANOVA test. These differences between the data could only be tested statistically when a sufficient amount (n '10) of comparable products were present. Reformulation targets that have been agreed for sodium and saturated fat levels of meat product in the Netherlands can be found in Appendix 3. Non-numeric values (GDA and Choices logo) from the GFM database 2015 were evaluated as well. GDA was assessed as the presence on the products packaging and for the Choices logo, all products were specifically checked on the nutrient criteria for meat products according to the Choices programme (in the Netherlands). These criteria are as following: sodium ('900 mg/100g), trans fat (not added), saturated fat ('13 en%) and added sugars ('2.5 g/100g). For meat alternatives equivalence criteria should also be met and contain at least two of the following nutrients in sufficient amount: Retinol, Vitamin B1, Vitamin D, iron, Vitamin B12 (12, 24) (see Appendix 4). 3. RESULTS 3.1 General results Data were collected for 911 meat products and after data checking, data were available for 863 meat products, of which 86% (n=745) with an Nutritional Information Panel (NIP). Salt/sodium, saturated fat levels and sugars were not included in 3% (n=22) of all cases with an NIP. Only energy content, protein- total fat- and carbohydrate levels were available for these products (n= 723). The descriptive statistics of all nutritional values can be found in Appendix 5. For analysing meat products on salt/sodium, data were available for 723 products, of which 11% (n=78) with a sodium level and 83% (n=599) with a salt level on their NIP. About 6% (n=46) of the meat products has both salt and sodium levels displayed. The percentage varies by product category (table 3). For analysing sodium levels (in subparagraphs 3.2 - 3.4): the salt levels on the product packaging were converted into sodium levels using a conversion factor of 2.5. Table 3. Presence of sodium and/or salt declaration on the NIP* of meat products (n) in 2015. Product category Only salt declaration Only sodium declaration Both salt and sodium declaration (%) (n) (%) (n) (%) (n) Single heated (n= 100) Combined heated (n= 350) Single raw (n= 76) Combined raw (n= 186) 83 85 79 80 83 297 60 149 9 11 12 11 9 39 9 20 8 4 9 9 8 11 7 17 Meat alternatives (n= 11) 91 10 9 1 0 0 Total meat products with NIP (n= 723) 83 599 11 78 6 46 * Nutritional information panel. ' For all analysed meat products (n=863), about 15% (n=126) had a GDA label on their product packaging. The percentage varied by product category (table 4). Table 4. Presence of GDA* displayed on the product label of meat products (n) in 2015. GDA (%) (n) Single heated (n= 120) Combined heated (n= 401) Single raw (n= 103) Combined raw (n= 227) Meat alternatives (n= 12) 3 23 2 10 50 4 92 2 22 6 Total meat products with NIP (n= 863) 15 126 * Guideline Daily Amounts. Table 5 shows the amount of products that complied with the Choices criteria (in the Netherlands) and the amount of products which carried a health logo. Of all analysed products (n=723), 5.3% (n=38) meet the Choices criteria to carry such a health logo, only 2.4% (n=17) actually carries the health logo on their packaging. There are no products which carry the Choices health logo, but do not meet the criterion. Table 5. Number (n) and proportion (%) of meat products in 2015 that complied with the Choices criteria.* Product category Products that met Choices criteria (%) (n) Products that met criteria and have a Choices logo (%) (n) Products that had a Choices logo but did not meet the Choices criteria (%) (n) Single heated (n= 100) 21 21 8 8 0 0 Combined heated (n= 350) 2 8 2 6 0 0 Single raw (n= 76) 1 1 0 0 0 0 Combined raw (n= 186) 3 5 0 0 0 0 Meat alternatives (n= 11) 27 3 27 3 0 0 Total meat products with NIP (n= 723) 5 38 2 17 0 0 * The Choices programme uses specific nutrient criteria for meat products: sodium('900mg/100g), trans fat1 (not added), saturated fat ('13 en%) and added sugars2 ('2.5g/100g). For meat alternatives equivalence criteria should also be included. 1 Trans-fat: no values available, not used in analysis. 2 Assumption: added sugars are equal to 'Sugars' on labelling. No distinction can be made between naturally occurring or added sugars. ' 3.2 Results: sodium and saturated fat levels in meat products Box plots are given for sodium levels in milligrams per 100 grams of product (figure 2) and saturated fat content expressed as grams per 100 grams of product (figure 3), or as energy percentage (figure 4). Fig. 2 Sodium levels (mg/100g) of meat products in 2015. The variation in compositions: the 25th percentile (bottom of the box), 75th percentile (top of the box), lowest value in data (bottom of the whisker) and highest value of data (top of the whisker). Sodium levels of the meat products are very wide. Only combined heated meat products (m= 865 mg/100g) and meat alternatives (m=875 mg/100g) were below the Choices criterion (900mg/100g) with their mean levels. The mean level of the single heated meat products (m=1044 mg/100g) was just above the Choices criterion, mean levels of combined raw meat products (m=1317 mg/100g) and single raw meat products (m=1564 mg/100g) were widely above the Choices criterion. Fig. 3 Saturated fat levels (g/100 g) of meat products in 2015. The variation in compositions: the 25th percentile (bottom of the box), 75th percentile (top of the box), lowest value in data (bottom of the whisker) and highest value of data (top of the whisker). The saturated fat levels (in grams) of the meat products were also very wide. The mean levels of the products were as following: single heated meat products (m=4 g/100g), combined heated meat products (m=8 g/100g), singe raw meat products (m=6 g/100g) and a peak in the mean level of combined raw meat products (m=13 g/100g). The meat alternatives had the lowest mean level (m=3 g/100g). ' Fig. 4 Saturated fat (en%) of meat products from in 2015. The variation in compositions: the 25th percentile (bottom of the box), 75th percentile (top of the box), lowest value in data (bottom of the whisker) and highest value of data (top of the whisker). The distribution of the saturated fat levels in energy percent was substantially equal to the distribution of saturated fat in grams, as displayed in figure 3. Mean levels did not fall under the choices criteria level (13 en%) except for the meat alternatives (m=9 en%). In contrast, single heated meat products (m=18 en%), combined heated meat products (m=26 en%), singe raw meat products (m=21 en%) and combined raw meat products (m=29 en%) were all above the Choices criterion. ' 3.3 Comparison of 2015 sodium and saturated fat levels to previous years In tables 6 and 7, the mean levels and their corresponding standard deviations of sodium and saturated fat were compared to the data available from previous monitoring reports. Results are visualized in figures 5 and 6. Meat alternatives were not captured in previous monitoring reports so they were excluded in the comparison. NEVO 2011 NEVO 2013 NEVO 2014 GFM NL 2015 Product category N Mean (SD) N Mean (SD) N Mean (SD) N Mean (SD) Meat products, single heated 13 999 (228) 13 1024 (218) 58 951 (196) 100 1044 (491) Meat products, combined heated 23 867 (98) 23 881 (122) 98 892 (177) 350 865 (175) Meat products, single raw 5 1542 (344) 5 1600 (286) 26 1418 (274) 76 1564 (435) Meat products, combined raw 9 1242 (230) 9 1245 (231) 18 1323 (390) 186 1317 (393) Table 6. Comparison of mean sodium levels (mg/100 g) of meat products in 2015 with NEVO tables (2011, 2013 & 2014). Note: Different methods were used for assessing these data (NEVO and GFM). Figure 5 shows mean (?? SD) sodium levels in meat products (mg/100 g) based on table 6. (There were no significant differences found between sodium levels from NEVO 2011, 2013, 2014 and GFM 2015 per product category). Fig. 5 Comparison mean sodium levels (mg/100 g) of meat products in 2015 with NEVO tables (2011, 2013 & 2014). NEVO 2011 NEVO 2013 NEVO 2014 GFM NL 2015 Product category N Mean (SD) N Mean (SD) N Mean (SD) N Mean (SD) Meat products, single heated 2 2 (0) 2 2 (0) 20 3 (3) 100 4 (5) Meat products, combined heated 13 9 (2) 13 9 (3) 37 10 (3) 350 8 (4) Meat products, single raw 4 7 (4) 4 6 (5) 12 7 (6) 76 6 (5) Meat products, combined raw 6 10 (5) Insufficient numbers provided Insufficient numbers provided 186 13 (6) Table 7. Comparison mean saturated fat levels (g/100 g) of meat products in 2015 with NEVO tables (2011, 2013 & 2014). Note: Different methods were used for assessing these data (NEVO and GFM). Figure 6 shows mean (?? SD) saturated fat levels in meat products (g/100 g) based on table 7. Within the combined heated category, a significant difference was found between mean levels of NEVO 2014 and GFM 2015 (P <0.001). Fig. 6 Comparison mean saturated fat levels (g/100 g) of meat products in 2015 with NEVO tables (2011, 2013 & 2014). * P<0.05 for difference. ** Insufficient amount of data delivered by NEVO 2013 & 2014. ' 3.4 Evaluation reformulation targets of meat products To allow comparison, subcategories were created in accordance with the reformulation categorization (15). In the tables below, sodium levels (table 8) and saturated fat levels (table 9), as found in the meat product analysis in 2015 (GFM database), were compared with the reformulation targets. (Accompanying scatter plots can be found in Appendix 6). 14%-93% of the meat products had sodium levels that complied with the reformulation targets (table 8). Saturated fat percentages of complying meat products varied between 25% and 88% (table 9). Table 8. Reformulation targets meat products: Sodium levels (mg/100 g). Product categorization reformulation covenant* Agreed maximum per 100 gram of product Number of products analysed in 2015 GFM database Percentage of products that achieved the reformulation targets in 2015 Single heated meat products Grilled bacon 1120 mg N= 7 14% Remaining single heated products 1015 mg N= 93 82% Combined, heated meat products 945 mg N= 350 75 % Combined, raw meat products Filet Americain 900 mg N= 27 93% Remaining combined raw products 1280 mg N= 159 35% * Single raw meat products were not included in the reformulations monitor. Table 9. Reformulation targets of (a part of) combined heated meat products: Saturated fat levels (g/100 g). Product categorization reformulation covenant Agreed maximum per 100 gram of product Number of products analysed in 2015 GFM database Percentage of products that achieved the reformulation targets in 2015 Roasted meatloaf 9,75 g N= 5 80% Grilled sausage 9,70 g N= 14 57% Liver cheese/ Berliner 11,10 g N= 8 88% P??t?? 11,85 g N= 36 78% Smoked sausage* 10,55 g N= 20 25% Luncheon meat 10,20 g N= 23 78% Cooked sausage 10,80 g N= 13 54% Liver sausage/ Hausmacher 9,00 g N= 46 39% Liver sausage spread 10,35 g N= 10 70% * Lean products were excluded from this comparison: no reformulation targets were defined for saturated fat levels of lean products. 4. DISCUSSION Our analysis showed that there has been no progress in reducing sodium and little progress in reducing saturated fat levels in foods in the Netherlands, despite reformulation targets being in place to try to lower the levels of these adverse nutrients. We also found that a small proportion of meat products (16% with no NIP or an incomplete NIP) did not display the required nutrients on products labels as set out by the EU Regulation NO. 1169/2011 which states that manufactures must include energy value, the amount of fat, saturated fat, carbohydrates, sugars, proteins and salt on the packaging label (Art. 30.1. a,b.) (16). Products with no NIP can be explained with Regulation NO. 1169/2011, which includes a list of foods that are exempted from the requirement of the mandatory nutrition declaration. It excludes fresh meat products (Annex V regulation: point 19) (16). In addition, manufactures have still time to put a (complete) NIP on their packaging until December 2016, before then it is still a subject to change. Past experiences with trans fat has learned that obligated declaration leads to innovation and/ or reformulation, because food companies are forced to think about the health effects of their products and how to apply that knowledge (25). Also FOP labelling (which increases consumer awareness) proved to be important in successful salt reduction strategies (26, 27). Only a very small proportion (5%) of meat products complied to the Dutch Choices criteria. This could imply that these criteria are challenging for meat products. Even less (2%) actually carried a Choices health logo on the FOP, indicating the healthier choice within a product group. Unlike the required NIP, health logos are voluntary front-of pack labelling options. It is up to the producer or supermarket to decide whether to use them. According to Vyth et al (2010), the Choices logo mainly plays a role in the actual food purchases of people who are health conscious and weight conscious (28) and influences food manufactures to reformulate their products (10, 13). The increased availability of healthier products, such as those carrying the logo, can be an efficient way to improve the diets of all consumer groups, whether or not they identify as health-conscious consumers. According to the intake modelling of Roodenburg et al. (2013): when consuming more Choices-compliant foods, nutrient intake shifts towards nutrition population intake goals for which criteria were defined (29). While writing this article, new international Choices criteria were published. The sodium criteria became stricter and went from '900mg/100 grams of product to '820mg/100 grams of product. The criteria for SAFA (13 en% or 1.1g/100 grams of product) were removed and replaced by '5.0g/100 gram of product. Assumed is that also the Dutch criteria will be adjusted in the near future. No changes in sodium level were found over the past 4 years in the Netherlands. Similar results have been reported for other countries. For example a study in New Zealand, which explored a 10-year change in sodium contents of nine processed food categories and showed no significant differences in matched products over time (30). Also, because of its preservative effect, it is difficult to decrease salt/sodium levels in (single) raw meat products. For the SAFA only one significant different was found in the subcategory of combined, heated meat products between 2014 and 2015. This may be explained by the reformulation targets set, only for the combined, heated meat products. All other subcategories were excluded for reformulation on SAFA. This means that reformulation targets for SAFA in meat products are achieved for a small fraction. If, in a later stage, reformulation targets can be achieved (in multiple product categories), a reduction in sodium and saturated fat may lower populations blood pressure and cholesterol, respectively, both leading risk factors of death (2). This will lead to major improvements of public health and will save costs in healthcare (31). Since this is the first year data is collected for meat products in the Netherlands with the methodology of the Global food monitoring group, there were no other data available to see whether there has been some change in nutrient levels. Therefore data is used from previous monitoring reports. The data obtained in 2015 (with the GFM database) will apply as a baseline for future years. The comparison of sodium and SAFA with prior monitoring reports is difficult. The new EU labelling requirements, in combination with the methodology of the GFMG, has offered the possibility to collect data for a larger amount of products (within a short time period), as compared with the number of products used in the monitoring reports of the RIVM (2012 & 2014). In addition, the nutritional values of the products in the GFM database (in the Netherlands) are directly obtained from the products packaging. These are the direct values shown to consumers and believed to be reliable, because a products label is the interface of a the industry towards the consumer. The reports of the RIVM, uses data directly obtained from manufactures (voluntarily provided) and may differ from the data on the actual product packaging. These reports are dependent on what the manufactures provide (e.g. insufficient amount of data provided for SAFA in 2013 and 2014 for combined raw meat products). The illegibility of some products was a limitation of the GFM database, as data had to be obtained from the supermarkets website. This data was not always up-to-date: information on packaging was sometimes non-corresponding in relation to the information on the supermarket's website, depending on the supermarket, especially with respect to data of non-private label products. Next to the Global Food Monitoring database, other initiatives have started in the Netherlands. Some databases have already collected a lot of data. This might be interesting to join forces , by merging the databases. This is a possible manner to create one large (and open) nutritional and health platform, instead of many small initiatives. 5. CONCLUSION In conclusion, the project provides new information about the composition of meat products in the Netherlands, after reformulation targets are set by the government and industry. Yet, no progress is made with sodium and little progress with saturated fat reductions over the past 4 years in processed meat products. This information created more detailed insight about the specific subcategories of meat products and provided evidence on the reformulation progress. In the light of the reformulation covenant of 2014, more focus is needed if the Netherlands is to meet its national commitment to reduce sodium and saturated fat in meat products. The results can be used to make more effective use of resources and identify new (specific) strategies. In the short term: this method of monitoring and evaluation should also be used for other common product categories to (transparently) monitor progress, motivate the industry and to ultimately decrease the burden of nutrition-associated disease in the population in the Netherlands, because small improvements of the food supply will ultimately lead to significant advances in public health and health care budgets. More focus should be on reducing reformulation barriers to strengthen the commitment of the industry, when implementing targets. In the long term: the effectiveness of the food reformulation strategies could be measured in terms of health outcomes. ' REFERENCES 1. The George Institute for Global Health: Global Food Monitoring Group 2014. Available from: http://www.georgeinstitute.org/projects/global-food-monitoring-group.
2. WHO. Global action plan for the prevention and control of noncommunicable diseases 2013-2020. Geneva2013. Available from: http://apps.who.int/iris/bitstream/10665/94384/1/9789241506236_eng.pdf?ua=1.
3. WHO. Factsheet: Noncommunicable Diseases (NCD) Country Profiles: Netherlands 2014. Available from: http://www.who.int/nmh/countries/nld_en.pdf.
4. WHO. Diet, nutrition and the prevention of chronic diseases: report of a joint WHO/FAO expert consultation. Geneva: 2003.
5. RIVM. Volksgezondheid Toekomst Verkenning 2014 (VTV 2014) 2014 [Accessed: 28-04-2015]. Available from: http://www.eengezondernederland.nl/Heden_en_verleden/Determinanten.
6. WHO. The World Health Report 2002 – Reducing risks, Promoting health life 2002. Available from: http://www.who.int/whr/2002/en/whr02_en.pdf?ua=1.
7. van Raaij J, Hendriksen M, Verhagen H. Potential for improvement of population diet through reformulation of commonly eaten foods. Public health nutrition. 2009;12(3):325-30.
8. Ng SW, Dunford E. Complexities and opportunities in monitoring and evaluating US and global changes by the food industry. Obesity reviews : an official journal of the International Association for the Study of Obesity. 2013;14 Suppl 2:29-41.
9. Progress with a global branded food composition database. Food chemistry. 2013;140(3):451-7.
10. Vyth EL, Steenhuis IH, Roodenburg AJ, Brug J, Seidell JC. Front-of-pack nutrition label stimulates healthier product development: a quantitative analysis. Int J Behav Nutr Phys Act. 2010;7:65.
11. Roodenburg AJ, Popkin BM, Seidell JC. Development of international criteria for a front of package food labelling system: the International Choices Programme. Eur J Clin Nutr. 2011;65(11):1190-200.
12. International Scientific Committee Choices Foundation (2007) Qualifying criteria. [Accessed: 02-02-2015]. Available from: http://www.choicesprogramme.org/about/product-criteria.
13. Verhagen H, Milder I, Hendriksen M, Ock?? M, van Raaij J, Temme L, et al. Het Vinkje ‘ Inventarisatie van de ontwikkeling in de criteria voor toekenning. RIVM briefrapport 2015-0045; 2015.
14. RIVM. Akkoord verbetering productsamenstelling 2014 [Accessed January 2015]. Available from: http://www.rijksoverheid.nl/documenten-en-publicaties/convenanten/2014/01/23/akkoord-verbetering-productsamenstelling-zout-verzadigd-vet-suiker.html.
15. CBL. Convenant herformulering vleeswaren 2013-2015: Vereniging voor de Nederlandse Vleeswaren Industrie (VNV); 2013 [Accessed: January 2015]. Available from: http://www.cbl.nl/fileadmin/user_upload/formulieren__factsheets_etc/052813_Convenant_herformulering_vleeswaren_def.pdf.
16. EU. Regulation(EU) 2011/1169 of the European Parliament and of the Council of 25 October 2011 on the provision of food information to consumers: Official Journal of the European Union; 2011. Available from: http://eur-lex.europa.eu/legal-content/EN/TXT/?qid=1420635313914&uri=CELEX:32011R1169.
17. Greenfield H, Southgate, D.A.T. Food composition data: production, management, and use. Rome: Food and Agriculture Organization of the United Nations; 2003.
18. Temme E.H.M. ea. Natrium en verzadigd vet in beeld: Verandering in samenstelling van voedingsmiddelen in 2012. Bilthoven: RIVM, 2013.
19. Temme E.H.M. ea. Monitor Productsamenstelling voor zout, verzadigd vet en suiker: RIVM Herformulerinsmonitor 2014. Bilthoven: RIVM, 2015.
20. Dunford E, Webster J, Metzler AB, Czernichow S, Ni Mhurchu C, Wolmarans P, et al. International collaborative project to compare and monitor the nutritional composition of processed foods. European journal of preventive cardiology. 2012;19(6):1326-32.
21. Euromonitor. Grocery Retailers in the Netherlands 2015 [April 2015]. Available from: http://www.euromonitor.com/grocery-retailers-in-the-netherlands/report.
22. The George Institute for Global Health (2008): Food Composition Database. Sydney: The George Institute for Global Health.
23. Choices programme 2014 [Accessed April 2015]. Available from: http://www.choicesprogramme.org/about/the-programme
24. Het vinkje – Productcriteria Stichting Ik Kies Bewust 2012 [Accessed April 2015]. Available from: http://www.hetvinkje.nl/site/assets/files/1082/productcriteria_versie_4_4.pdf.
25. Korver O, Katan MB. The elimination of trans fats from spreads: how science helped to turn an industry around. Nutr Rev. 2006;64(6):275-9.
26. Webster J. Reformulating food products for health: context and key issues for moving forward in Europe. The George Institute for International Health, 2009.
27. Bruce Traill W, Bech’Larsen T, Gennaro L, Koziol’Kozakowska A, Kuhn S, Wills J. Reformulation for healthier food: a qualitative assessment of alternative approaches [Accessed June 2015]. Available from: http://ageconsearch.umn.edu/bitstream/123511/2/Traill_ReformulationforHealthierFoodAQualitativeAssessmentofAlternativeApproches.pdf.
28. Vyth EL, Steenhuis IH, Vlot JA, Wulp A, Hogenes MG, Looije DH, et al. Actual use of a front-of-pack nutrition logo in the supermarket: consumers’ motives in food choice. Public health nutrition. 2010;13(11):1882-9.
29. Roodenburg AJ, van Ballegooijen AJ, Dotsch-Klerk M, van der Voet H, Seidell JC. Modelling of usual nutrient intakes: potential impact of the choices programme on nutrient intakes in young dutch adults. PLoS One. 2013;8(8):e72378.
30. Monro D, Mhurchu CN, Jiang Y, Gorton D, Eyles H. Changes in the Sodium Content of New Zealand Processed Foods: 2003-2013. Nutrients. 2015;7(6):4054-67.
31. MacGregor GA, He FJ, Pombo-Rodrigues S. Food and the responsibility deal: how the salt reduction strategy was derailed. BMJ. 2015;350:h1936.
32. Hendriksen MA, van Raaij JM, Geleijnse JM, Breda J, Boshuizen HC. Health gain by salt reduction in europe: a modelling study. PLoS One. 2015;10(3):e0118873.
33. Lawrence M. Do food regulatory systems protect public health? Public health nutrition. 2009;12(11):2247-9.
34. Euromonitor. Euromonitor Global Market Information Database: Packaged food in the Netherlands 2015 [Accessed April 2015]. Available from: http://www.euromonitor.com/packaged-food-in-the-netherlands/report.
35. Foodwatch. Kindermarketing: Onverantwoord en ongereguleerd: Stichting Foodwatch; 2014 [Accessed 04-05-2015]. Available from: https://www.foodwatch.org/fileadmin/foodwatch.nl/Onze_campagnes/Kindermarketing/Documents/20150220_FW_Kindermarketing_rapport_feb15_2w.pdf.
36. Dunford E, Trevena H, Goodsell C, Ng KH, Webster J, Millis A, et al. FoodSwitch: A Mobile Phone App to Enable Consumers to Make Healthier Food Choices and Crowdsourcing of National Food Composition Data. JMIR mHealth and uHealth. 2014;2(3):e37.
37. Neal B, Sacks G, Swinburn B, Vandevijvere S, Dunford E, Snowdon W, et al. Monitoring the levels of important nutrients in the food supply. Obesity reviews : an official journal of the International Association for the Study of Obesity. 2013;14 Suppl 1:49-58.
APPENDIX – 1
Sales of packaged food in the Netherlands
As mentioned in the article, the majority of the food consumed in the Netherlands is being processed by the food industry (7). Hence, they have enabled an constant supply of affordable foods, so mass population is exposed to foods that are energy dense, high in saturated fat (SAFA), sugar and sodium, which emphasizes the important role the food industry plays in public health (4, 8). Table 1 displays the Dutch sales of packaged foods by category over the past 6 years and shows the extensive quantities of processed foods sold. Despite the recession, the trend has remained fairly stable.
Table 1 Sales of Packaged Food by Category in the Netherlands: Volume (in ‘000 tonnes) 2009-2014 (34).
‘000 tonnes 2009 2010 2011 2012 2013 2014
Baby Food 27.78 28.15 28.44 27.64 26.78 26.32
Bakery 1,342.61 1,339.31 1,339.51 1,334.61 1,328.97 1,323.10
Canned/Preserved Food 216.42 215.03 214.59 215.45 215.28 216.73
Chilled Processed Food* 193.91 192.57 182.57 183.63 181.47 179.96
Confectionery 168.89 170.41 171.31 167.42 166.05 165.63
Dairy 1,836.45 1,839.88 1,844.72 1,828.51 1,801.45 1,777.67
Dried Processed Food* 112.29 107.27 106.79 107.25 107.01 107.18
Frozen Processed Food 213.77 214.31 216.41 217.95 219.14 221.37
Ice Cream 54.95 56.18 54.77 54.16 54.56 55.41
Meal Replacement 1.05 1.35 1.60 1.62 1.62 1.61
Noodles 7.72 7.60 7.51 7.55 7.62 7.70
Oils and Fats 201.68 199.36 197.56 197.08 196.75 196.33
Pasta 50.83 49.82 49.35 49.60 49.90 50.30
Ready Meals 111.43 110.47 99.83 100.71 100.62 101.76
Sauces, Dressings and Condiments 131.89 133.12 135.38 136.86 138.31 139.73
Snack Bars 4.31 4.50 4.68 4.86 5.01 5.15
Soup 87.62 90.91 93.04 94.44 95.85 96.66
Spreads 60.75 61.77 63.02 63.72 64.48 65.27
Sweet and Savoury Snacks 100.33 104.56 103.40 106.65 108.62 110.73
Impulse and Indulgence Products 665.75 675.11 675.75 674.74 676.39 679.73
Nutrition/Staples 3,234.28 3,226.58 3,229.51 3,208.41 3,174.64 3,144.42
Meal Solutions 960.54 961.97 946.96 954.09 955.59 961.04
Packaged Food ** 4,698.11 4,702.92 4,702.73 4,688.05 4,658.44 4,636.69
* Meat products fall under ‘Chilled processed foods’ and dried sausages (e.g. salami) are covered by the ‘Dried processed foods’.
** Note: Sum of categories does not equal total packaged food because of double counting (e.g. canned soup is included in soups and canned foods).
All of these processed food products play an important role in the prevention and causes of diseases. Thereupon, the government, consumer organisations and other advocacy groups pointed out the need for the food industry to improve their products on the field of SAFA, TFA, total sugar and sodium levels (14, 23, 35).
APPENDIX – 2
The Global Food Monitoring Group
The George Institute for Global Health (figure 1) in Sydney, Australia, initiated in 2010 the Global Food Monitoring Group (GFMG). The GFMG is an international organization which brings together countries to support them to monitor the nutrition composition of foods. It established a new (online) global branded food composition database, with a standardized and sustainable methodology to collect accurate data at low cost, so robust comparisons can be made between countries and over time. Collaborators in each country can upload data into the database, using an agreed protocol for data collection. It originally focussed on salt reduction only and was funded by the British insurance company Bupa. Now, the project aims to collate data on all nutrient information (or lack thereof) from labels of packaged foods. Bringing transparency and objectivity to the monitoring process and having data that describes the progress of individual countries and companies over the world can be used to drive national and international category-wide changes in the composition of processed food products, which even if they are small, have an opportunity to deliver potential health gains. The project also helps to identify where reformulation strategies would have the greatest impact and support food companies/manufactures to improve their products on health. The information that evolves from the project will provide the food industry, governments, health professionals and other advocacy groups with new evidence about progress that is being made or not. By using this information, appropriate strategies could be identified and resources could be direct more effectively (1, 9).
Foodswitch app
Adjoining to the database, the GFMG collaborated with developers from Xyris Software Pty Ltd and launched the smartphone application FoodSwitch, which can help consumers to make a healthier food choice by providing easy-to-understand nutritional information about packaged foods. With a barcode scanning technology, an optimal mechanism for interaction of the mobile phone with the food database was identified. The application allows consumers to scan the barcodes of processed foods with their phone and then receive a traffic light label which shows the level of SAFA, sugar, salt and energy content of the product. Alongside, the app scans through the database for similar foods and gives recommendations for healthier alternatives. They also created special filters for the application, for consumers with specific diets: the SaltSwitch (preferred for people with high blood pressure) and the GlutenSwitch, which helps to identify gluten-free products.
In addition, the power of crowd-sourcing data collection is used for FoodSwitch to help further complete the database of branded food data by continuous data collection. Consumers are asked to collect the data from the label, if the information on the scanned product is not available. They can do this by taking pictures with their own smartphone (iOs and Android platforms), which are uploaded into the database. Within two days of release of the FoodSwitch application in Australia (February 2012), 26.000 photos of missing items were uploaded. The George Institute continues receiving approximately 500 photos every day of new items launched into the market. FoodSwitch started as a tool for consumers but has now become a method for data collection. The application was already successfully introduced in Australia, New Zealand and the United Kingdom and has gained international recognition. Currently, there are plans to introduce FoodSwitch in China, India and North America as well (36, 37).
‘
APPENDIX – 3
Agreement on improvements in product composition & covenant reformulation meat products
Choosing healthier products should be made easier for consumers. A healthy diet is important for good health. Therefore an agreement is signed for improving the composition of products and total product supply as a whole (14) by the Dutch government and the private sector (In Dutch: Akkoord Verbetering Productsamenstelling). This agreement aims at reducing the levels of salt, saturated fat, sugar and energy content. Ultimately, this will lead to a healthier product offering.
The parties signed on the following joint ambitions:
‘ Reducing salt in the product range. Making it easier for consumers not to exceed the maximum values of 6 gram salt per day. To be achieved by 2020.
‘ Reducing the saturated fat content in the product range. Making it easier for consumers not to exceed the maximum of 10 energy percent per day. To be achieved by 2020.
‘ Making it easier for consumers to consume less energy. To be achieved by 2020 and where possible reduce energy content through the reduction of sugar and/or (saturated) fat and/ or reduce portion size and to continue promoting fruit and vegetables.
When working on these ambitions, the priority of product categories will be based on the relevance of public health. Products intended for children will get a high priority.
Agreements for bread, canned vegetables, meat products and Gouda cheese have already been made.
Covenant reformulation meat products
The covenant reformulation meat products (15), which is signed by the Association of the Dutch meat industry (known as VNV) and the Central food retail office (Dutch: CBL), focuses on the following specific objectives:
1) The reduction of the mean sodium levels in the product groups ‘single heated meat products’ (subdivided in ‘grilled bacon’ and ‘other single heated meat products’), ‘combined heated meat products’ and ‘combined raw meat products’ (subdivided in ‘filet americain’ and ‘other raw combined meat products’) with 10% (table 4).
2) The reduction of SAFA in the product group combined heated meat products (subdivided to multiple products) with 5% (table 5).
Table 4. Maximum sodium levels (mg/100 g): agreed on in reformulation covenant meat products.
Product (group) Maximum per 100 gram of product
Single heated meat products
Grilled bacon 1120 mg
Other single heated meat products 1015 mg
Combined heated meat products 945 mg
Combined raw meat products
Filet Americain 900 mg
Other combined raw meat products 1280 mg
Table 5. Maximum saturated fat levels (g/100 g): agreed on in reformulation covenant meat products.
Product (group) Maximum per 100 gram of product
Roasted meatloaf 9,75 g
Grilled sausage 9,70 g
Liver cheese/ Berliner 11,10 g
P??t?? 11,85 g
Smoked sausage* 10,55 g
Luncheon meat 10,20 g
Cooked sausage 10,80 g
Liver sausage/ Hausmacher 9,00 g
Liver sausage spread 10,35 g
* Lean products are excluded from the reformulation in SAFA.
‘
APPENDIX – 4
Choices programme in the Netherlands
In 2006, a global programme named ‘The Choices programme’ was launched in the Netherlands (‘Ik Kies Bewust’). This is a multi-stakeholder cooperation which aims to fight obesity and other diet-related health issues (23, 24). It encourages the food industry to participate and evaluate the existing or reformulated nutritional content of their products against a set of qualifying criteria (12, 13). These criteria rely on transparent, science-based international dietary guidelines and are reviewed periodically by an independent international scientific committee. The criteria take product levels of saturated fat, trans fat, added sugars, salt and dietary fibre into account to determine its eligibility to carry the front-of-pack Choices logo (figure 2). This logo will help consumers to quickly distinguish healthy choices and stimulates food producers to develop healthier products.
Choices criteria
The Choices criteria for meat products are shown in table 2. Subsequently to these criteria, meat alternatives must also meet the equivalence criteria (table 3).
Table 2. Choices criteria meat products and meat substitutes (12).
Meat, fish, poultry, eggs and meat substitutes
Product group Criteria Definition
Processed meat, meat products and meat substitutes##. ‘ SAFA: ’13 en% or ‘ 1,1g/100 g
‘ TFA: not added
‘ Sodium: ‘ 900mg/100 g
‘ Added Sugars: ‘ 2,5g/100 g All kinds of processed meat/poultry, meat products and (vegetable) meat substitutes. Ready-to-eat meatballs, spiced or salted meat (fresh or deep-frozen), salami, smoke-dried beef, grilled ham, tempeh, tofu, dairy-based meat substitute, fungus-based meat substitute.
## Meat substitutes must meet the equivalence criteria.
Equivalence criteria meat substitutes:
A product should at least contain two of the following nutrients in sufficient amount: Retinol, Vitamin B1, Vitamin D, iron, Vitamin B12. These nutrients must include a minimum amount per 100 gram, shown in table 3.
Table 3 Minimum amounts of nutrients per 100 gram of product (12).
Nutrient Amount per 100 gram
Vitamin A (Retinol-equivalent) 70 ??g
Vitamin D 0,5 ??g
Vitamin B1 0,11 mg
Vitamin B12 0,24 ??g
Iron 0,8 mg
‘
APPENDIX – 7
Methods on cookies intended for children and sauces
The focus of the article is on meat products. In this section the two other groups will be described: cookies intended for children and sauces. The primary outcome measures of these product groups will be presented in Appendix 8.
The research of the product groups cookies and sauces is carried out exactly as is described in the article. Only new information is given below.
Cookies intended for children
To get acquainted with the methodology of the GFM in the Netherlands and at the request of the National Institute for Public Health and the Environment, data of a small product group was collected first in 2014: cookies intended for children (kids marketing). Due to marketing specifically aimed at children and most products being unhealthy (e.g. high in sugars) , this category received loads of negative attention from consumer organisations (e.g. Foodwatch) (35) and was on the priority list of the reformulations agreement (14) (Appendix 3).
Cookies intended for children is small product group and already a subcategory of ‘biscuits’ of the product group ‘bread and bakery’ (table 12). However, there were no data available for this category from previous monitoring reports.
During one day in May 2014, all relevant labelling information was collected in the same supermarket chains as described in the article with the data collection app of the George Institute of Global Health, Sydney. This product group was determined by the shelf location in the supermarket, usually on the second shelf, at the eye level of children. Packages contain many colours, pictures and cartoons. Information of 63 products was collected. After data entering and data checking only one product had to be removed from the dataset (duplicate product). During the data cleaning step, two products needed to be improved (typing errors). All recorded cookies had a NIP (table 12).
Sauces
At the request of the National Institute for Public Health, sauces have been selected as a second food group of interest. Especially the sodium levels of sauces were an important health aspect. This group included dressings, snack sauces, cooking sauces and seasonings.
A new classification system was created for the product group of sauces. This classification system combined the GFM-format with the Choices- and RIVM-format. This format was chosen because it was difficult to make a clear divisions based on the Choices and RIVM format only. For that reason the format was combined with the GFM classification, which had more transparency (see table 11: GFM HAS). In addition to the new classification, a new subcategory was added: seasonings (table 12).
Further, for the category of sauces, an additional format was included: ‘prepared form of food’ or ‘unprepared form of food’. This should be indicated during the data entry process. A products NIP makes a distinction between ‘per 100 grams of prepared product’ and ‘per 100 grams of unprepared product’. Sauces with a declaration for an unprepared (concentrated) product give much higher nutritional values when analysing then a declaration for an prepared (diluted) product. It is therefore important that they are separately analysed (table 12).
During eleven days in May 2014, all relevant labelling information was collected in the same two supermarket chains as described in the article. Information of 1209 products was collected. After data entering and data checking twelve products had to be removed from the dataset, because they were duplicated or misclassified (e.g. an spread instead of sauce). Of all recorded sauces, approximately 29% (n= 344) did not have a nutritional information panel (table 12). During the data cleaning analysis, four products needed to be improved (typing errors and misclassification).
APPENDIX – 8
Overall results
In totality data were collected for 2183 products (meat products, sauces and cookies). After data checking, data was available for 2122 products, of which 86% (n= 1820) with an NIP (table 12).
Brand shares in GFM database
About 25% (n= 530) of all the products in the Dutch database (n= 2122) belonged to the (private) brand of supermarket X, followed by 16% (n= 330) of the products from supermarkets Y (private) brand (figure 3). Due to lack of time, fewer products were collected at supermarket Y (<1%) and resulted in a smaller proportion of supermarket Y private label products. However, the largest share of brands in the database came from other premium brands with approximately 59% (n= 1262). As shown in figure 4, 56% (n=186) of supermarkets Y private brand had no NIP on the packaging, against 4% (n= 22) without NIP from supermarket X privates brand. Fig. 3 Brand shares of all products Fig. 4 Ratio of brand without NIP on in Dutch GFM Database in 2014-2015. products packaging in 2014-2015. ' Results cookies intended for children DATA FOR COOKIES WITH KIDS MARKETING WERE AVAILABLE FOR 62 PRODUCTS, OF WHICH 62 PRODUCTS WITH AN NUTRITIONAL INFORMATION PANEL (NIP). THE DATA WAS EXCLUSIVELY OBTAINED FROM FOOD LABELS. TABLE 13 SHOWS THE DESCRIPTIVE STATISTICS OF THE COOKIES FOR THE MOST IMPORTANT NUTRITIONAL VALUES. Table 13. Descriptive statistics per 100 gram & per serve: cookies with kids marketing (in 2014). MEAN PER 100 GRAM MEAN PER SERVING Energy (kcal/100g) 454,6 KCAL 89,5 KCAL Saturated fat (g/100g) 8,6 G 1,7 G Sugars (g/100g) 31,4 G 6,4 G Sodium (mg/100g) 239,9 MG 55,1 MG THE MEAN OF THE SUGAR LEVEL IS HIGH COMPARED TO THE CRITERIA OF THE CHOICES PROGRAMME IN THE NETHERLANDS (31.4 GRAM/100 GRAM AGAINST ' 20 GRAM/100 GRAM). AS SEEN IN FIGURE 5, ONLY FIVE COOKIES WERE IN LINE WITH THE SUGAR THRESHOLD OF THE CHOICES CRITERIA. HOWEVER, THERE WERE ONLY THREE COOKIES WHO ARE ELIGIBLE WITH THE CHOICES CRITERIA, DEPENDING FROM THE TRANS-FAT LEVEL (TABLE 14), BUT NONE OF THE COOKIES HAD A CHOICES THICK ON THE PRODUCTS PACKAGING. ' Table 14. Cookies eligible with Choices Criteria (in 2014). CHOICES CRITERIA Goodies Mini koeke mannetjes Goodies Beestenboel koekjes Liga Baby biscuits 4-6 maanden ENERGY PER PORTION (KCAL/PORTION) '110 108 51 61 SATURATED FAT (EN%) '13 8,125* 8,45** 6,07** TRANS-FAT (EN%) 1 '1,3 - - - SODIUM (MG/100G) '400 200 200 70 ADDED SUGARS (G/100G) 2 '20 18 16 16,5 1 TRANS-FAT: NO VALUES AVAILABLE. 2 ASSUMPTION ADDED SUGARS: same as 'Sugars' on labelling. No distinction can be made between naturally occurring or added sugars. * SATURATED FAT: 3,9 GRAM X 9 KCAL (ENERGY VALUE FAT) = 35,1 / 432 KCAL (PER 100 GRAM OF PRODUCT) = 0,08125 X 100% = 8,125 ENERGY PERCENT. ** SATURATED FAT: 4,0 GRAM X 9 KCAL = 36 / 426 KCAL = 0,0845 X 100% = 8,45 ENERGY PERCENT. *** SATURATED FAT: 2,8 GRAM X 9 KCAL = 25,2 / 415 = 0,0607 X 100% = 6,07 ENERGY PERCENT. Other label information: 63% (n= 39) of all cookies carried a GDA on the FOP label and 39 COOKIES (63%) CONTAINED GLUTEN, ACCORDING TO THE ALLERGEN INFORMATION ON THE PRODUCT LABELS. OF ALL COOKIES WITH KIDS MARKETING IN THE GFM DATABASE (NL), THERE WAS NOT ONE COOKIE DECLARED GLUTEN FREE. NEXT TO GLUTEN, THE MOST COMMON ALLERGENS WERE, MILK (52%), SOY (44%) AND EGG (31%). NEXT TO THE HIGH NUTRITIONAL VALUES, A STRIKING REMARK WAS THAT PORTION SIZES OF COOKIES WITH KIDS MARKETING WERE OFTEN MISLEADING (IN 35% OF THE CASES). A PORTION SIZE WAS OFTEN DESCRIBED AS ONE COOKIE, WHILE THERE WERE MOSTLY MULTIPLE COOKIES IN A TAKEOUT PACKAGE (RANGING FROM 2 TO 6 BISCUITS). THIS SUGGESTS, WHEN NOT PAYING ATTENTION CAREFULLY, THAT SUCH A TAKEOUT PACKAGING (WITH MOSTLY MULTIPLE COOKIES) IS EQUAL TO 1 PORTION SIZE AND THEREFORE SEEMS TO HAVE A LOWER NUTRITIONAL VALUE. IN REALITY, PORTION SIZES HAD MUCH HIGHER NUTRITIONAL VALUE. ' Results sauces DATA FOR SAUCES WERE AVAILABLE FOR 1197 PRODUCTS, OF WHICH 85% (N= 1013) WITH AN NUTRITIONAL INFORMATION PANEL (NIP). DATA COULD NOT BE COMPARED TO PREVIOUS MONITORING REPORTS, DUE TO THE DIFFERENT CATEGORIZATION. THEREFORE ONLY THE MOST IMPORTANT NUTRITIONAL VALUES THAT RESULTED FROM THE (DUTCH) GFM DATABASE WILL BE SHOWN IN TABLE 15 AS MEAN LEVELS. SALT LEVELS ARE CONVERTED (WITH A FACTOR OF 2.5) INTO SODIUM LEVELS. Table 15. Descriptive statistics per 100 gram of the sauces. Only mean levels are displayed. PRODUCT CATEGORY (AMOUNT OF PRODUCTS WITH NIP)* ENERGY (KCAL) SAFA (G/100G) SUGARS (G/100G) SODIUM (MG/100G) DRESSING WATER-BASED: PREPARED (N=10) 88 0.2 15 642 DRESSING EMULSION BASED: PREPARED (N=70) 271 2 9 578 SNACK SAUCE WATER-BASED: PREPARED (N=104) 116 0.2 20 852 SNACK SAUCE WATER-BASED: UNPREPARED (N=9) 311 0,2 49 2633 SNACK SAUCE EMULSION-BASED: PREPARED (N=186) 403 3 7 635 GARNISH COOKING SAUCE: PREPARED (N=93) 99 3 2 477 GARNISH COOKING SAUCE: UNPREPARED (N=41) 405 7 11 4790 ITALIAN COOKING SAUCE: PREPARED (N=115) 97 2 6 501 ITALIAN COOKING SAUCE: UNPREPARED (N=12) 346 3 12 4756 MEXICAN COOKING SAUCE: PREPARED (N=10) 69 1 6 648 ORIENTAL COOKING SAUCE: PREPARED (N=155) 164 1 21 1331 ORIENTAL COOKING SAUCE: UNPREPARED (N=21) 318 6 19 4157 SEASONING: PREPARED (N=175) 219 3 15 2730 SEASONING: UNPREPARED (N=10) 264 4 6 4298 * PRODUCT CATEGORIES WITH LESS THAN 10 PRODUCTS WERE EXCLUDED FROM THE ANALYSIS. THE FOLLOWING FIGURES VISUALIZE THE MEAN LEVELS OF SODIUM (FIGURE 6), SATURATED FAT IN GRAMS (FIGURE 7) AND IN ENERGY PERCENT (FIGURE 8). ' APPENDIX 9 Other initiatives capturing nutritional data Thanks to modern ICT-applications, consumers can, now and in the future, make food choices that fits with their personal needs and wishes. The Global Food Monitoring database (George Institute for Global Health, Australia) is not the only initiative in the Netherlands which tries to create a new nutritional and health database. There are other initiatives which try to create a nutritional database and ultimately present these data to the consumers in understandable language, in order to improve overall health (figure 12). Fig. 12 Digitization of communication towards consumers. Boodschapp (2012) Boodschapp is a Dutch phone application where consumers have to make a picture of the products barcode. The app contains (at this moment) 12.000 products from the Albert Heijn, Jumbo and C1000. Consumers are informed about the price and nutritional values through a traffic light system. Products get compared to each other, so consumers can make the best choice. A team of nutritionists and dietitians analyse, photograph and enter label information manually, against a set of self-set criteria. Questionmark (2012) Questionmark is an independent neutral organisation that examines the sustainability of consumer products. Each product has a score on the following themes: health, environment, human rights and animal welfare. The average of these scores form the final score of the product. Questionmark makes those scores and underlying information accessible and transparent for consumers through a website and an app (by scanning the barcode). Questionmarks goal is to facilitate as many people as possible and motivate conscious shopping. By creating transparency within production and towards consumers, Questionmark wants to stimulate the supply and demand for sustainable products. In 2012, Question Mark has been allowed to take over the data from 'SuperWijzer', which had 15.000 products in the database at that point. Currently, Questionmark has 26.000 products in its database, which covers about 80% of the Dutch product supply in the following product categories: meat, meat substitutes, dairy, eggs, butter, cheese, desserts, chocolate, peanut butter, beer, soda and water. Foodhouse (2012) Foodhouse is an independent information platform. It is a phone and web application about food products where consumers can compare products and find nutritional information (about energy, sugars, fat, saturated fat and salt). In addition, consumers can give rankings and write reviews about each product. The application is meant as a source of information not as an advisory. Foodhouse is officially made for consumers, but offers also opportunities for a cooperation to manufactures, supermarkets and dieticians. At the same time, Foodhouse offers its database to other phone and web applications. Foodhouse's goal is to increase transparency within the food industry and provide consumers with new insights about food. Through the application consumers can search on products in the database or scan the barcode for direct information. Data is entered by the food house team, but can be supplemented by manufacturers when they log in. GS1 Netherlands (1970) GS1 is an independent non-profit organisation which makes industry-wide agreements on identifying, capturing and sharing data. This involves logistical, administrative and non-competitive commercial processes between chain parties. Within the food industry they label information in the GS1 database. This allows them to capture standardized information on products and processes. In this way food sectors can save costs and meet the demand of consumers. Within the food labelling: the manufacturer or proprietor can enter data into the GS1 Data Source and indicates who can receive this data. Customers can then read the data in their systems and make them online available to consumers. The data in question that can be uploaded is: Ingredients, allergens, origin of the product, RDA, health claims, nutritional values, preparation and storage instructions. GS1 Trustbox (not available in the Netherlands) Through trustbox, retailers and web shops can efficiently meet the requirements of the EU regulation 1169/2011. Brand owners remain in control over the information that consumers getting to see through a 'single point of entry'. In addition, consumers can rely for 100% on the information they get to see on websites and applications. GS1 Trustbox is a user-friendly tool which allows brand owners, distributors and importers to make (mandatory) product information available (according to EU Regulation 1169/2001) towards dealers, web shops and application developers through a 'single point of entry'. Companies have three options to upload product information on trustbox: via the web interface, via an Excel template or through publication from a GDSN data pool. After uploading all the mandatory label information into trustbox, companies have automatic access to the GS1 Data Checker. This is an online tool which allows companies to determine the quality of the data that has been uploaded. The data checker then helps to improve the data quality. Through a barcode scanning technology, information is also made directly available to consumers and the following information will appear: ' Product Details: GS1 article number, name and description, product pictures, net content, instructions for use and storage instructions, etc. ' Nutritional information: energy, sugar, fat, protein, salt, etc. ' Ingredients list ' Allergens criteria: "contains", "may contain", "contains no" or "unknown" ' Labels (adjustable by means of a list: e.g.: MSC, bio) ' Health claims: nutrition and health claims ' Releases: other information such as origin of the data, sales promotion. Important for companies is dealing with different product variants within the supply chain: when a new 'variant' of a food product will appear on the market, the product labels will change to. After such a change, a transition period will follow, where there are multiple versions of the same product in the supply chain. At that point it will be very difficult for companies to ensure that the information which is offered online to the consumer is in line with the information that is made available via the physical label. Adjusting the GTIN (barcode) of the food product is in these cases a possible solution. GS1 has set out rules that establish, at which type of product change, the GTIN should change or not. Crucially, if a product change can affect consumer health, or its purchase, it is best to assign a new GTIN to the new product variant. However, it is de decision and responsibility of the food company to determine whether the GTIN should be changed. If the company decides not to take a new GTIN, GS1 believes that in some cases it may still be necessary to inform consumers of the existence of different product variants. Therefore, GS1 developed a number of procedures for communicating product changes between trading partners, together with consultation experts from the EU food industry and traders. Currently trustbox does not exist in the Netherlands. The information linked to GS1 has no open access and is only available for manufactures. ' APPENDIX 10 Valorisation What value does this add to the business and to the consumer (in the near future or even in the long run). Currently there are a number of methods and food composition databases available in the Netherlands to monitor food data, i.e. LEDA database, NEVO database and the Choices database. Each database varies in purpose, method of data entry, the completeness, ease to use and availability for others. As a value to the business, the GFM database can monitor the food supply on a detailed product level and can provide nutritional composition information over years ' where other databases do not have the resources to update them. This can enable the evaluation of policy towards a healthier supply of foods. It helps to identify where reformulation strategies (would likely) have the greatest impact and supports food companies/manufacturer's specific monitoring. The GFM database not only monitors nutritional values but also other relevant labelling information or the lack thereof (where other databases cannot). It identifies direct areas for improvement. For the data collection, the dataset is independent from the food industry, unlike other databases. Because data is directly obtained from the physical packaging, the database can work as a stimulants towards the food industry regarding product labelling and reformulation. In the long run this will also be valuable to the consumer, because improving the food environment is crucial. Individual behaviour change is not the solution. But a combination of product reformulation, consumer awareness and education, seem to be the most effective reduction strategy, linked to appropriate monitoring systems (32). However, success can only occur when food reformulation strategies include transparency, measurable, strict target setting and monitoring (33). By undertaking this: people ultimately can't help but make healthier choices. The majority of the Dutch population will benefit from a healthier product supply, but especially people with a high risk on NCD's. According to Gyles et al.(2013): 'Nutrition can be a powerful force in improving both the health and the economic status of society.' For example: the costs saved through the health benefits may outweigh the costs of reformulation: the UK FSA estimated the annual health benefits afforded by a half a percentage point reduction in saturated fat (i.e. from current intakes of 13.3% of food energy to 12.8%) as 217,500 Quality Adjusted Life Years (QALYs). http://ec.europa.eu/health/nutrition_physical_activity/docs/ev20090714_wp_en.pdf An unfavourable diet and obesity lead each to 40,000 new cases of the above diet-related diseases each year. An epidemiological model was used to estimate the effect of a healthy diet in future health care costs and health benefits. However, the results of this economic evaluation prove to be very sensitive to changes in the long-term effects of the intervention. http://kosteneffectiviteit-preventie.rivm.nl/interventies/factoren-met-invloed-op-gezondheid/personene-met-gezondheidsrisico/gedragsinterventie-ter-promotie-van-een-gezond-voedingspatroon/
The benefits from a 1gram reduction in salt intake is equivalent to the benefits from 25% of the population who are overweight or obese shedding all their surplus kilos.
http://ageconsearch.umn.edu/bitstream/123511/2/Traill_ReformulationforHealthierFoodAQualitativeAssessmentofAlternativeApproches.pdf