* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Misconceptions and misinformation: The problems with GDAs
Survey
Document related concepts
Malnutrition wikipedia , lookup
Diet-induced obesity model wikipedia , lookup
Food safety wikipedia , lookup
Obesity and the environment wikipedia , lookup
Academy of Nutrition and Dietetics wikipedia , lookup
Food studies wikipedia , lookup
Food coloring wikipedia , lookup
Food politics wikipedia , lookup
Food choice wikipedia , lookup
Saturated fat and cardiovascular disease wikipedia , lookup
Human nutrition wikipedia , lookup
Transcript
Misconceptions and misinformation: The problems with Guideline Daily Amounts (GDAs) A review of GDAs and their use for signalling nutritional information on food and drink labels Report written by: Dr Tim Lobstein, Jane Landon and Paul Lincoln with contributions from: Dr Ruth Ash and Dr Vivienne Press This report assesses the use of Guideline Daily Amounts (GDAs) on food and drink packages as a means of conveying nutritional information. It finds several areas of concern. This use of percentage GDA signals on front-of-pack labelling has been promoted by some sections of the food industry as an alternative to a ‘traffic-light’ signposting system recommended by the Food Standards Agency (FSA). This report concludes that GDA signals are not the optimum method for helping consumers make quick, informed choices for at least six good reasons: 1. The GDA values do not distinguish maximum, minimum and average recommended amounts. 2. GDA values for adults and for children are used inconsistently, and adult GDAs are sometimes used on child-targeted products. 3. The GDAs used for labelling are based on values which are not the most suitable either for public health policy or for individuals. 4. The GDA displays are based on arbitrary portion sizes. 5. GDA signals for different nutrients are sometimes included or left out in an arbitrary and confusing manner. 6. The standard GDA signals lack colour coding for quick consumer appraisal and interpretation. These problems in the construction and use of GDA front-of-pack signals undermine the ability to provide clear and consistent signals to consumers as a means of promoting better public health. The FSA traffic-light signposting system is better able to provide a clear signal to consumers at the point of purchase. Disclaimer: Commercial product formulations and package labelling are continually changing. Products and labels referred to in this report were available for sale, or shown on company websites, at some point during the drafting of this report (from summer 2006 to February 2007) but they may not necessarily continue to be available. Acknowledgements: The National Heart Forum would like to thank Dr Susan Jebb (MRC Human Nutrition Research), and Dr Mike Rayner for their comments on this report. Edited by Wordworks, London W4 4DB. Text © National Heart Forum, February 2007 National Heart Forum Tavistock House South Tavistock Square London WC1H 9LG Phone: 020 7383 7638 Email: [email protected] Website: www.heartforum.org.uk Registered charity number: 803286 Page 2 of 46 Contents 1 Summary 4 2 The development of traffic-light labelling 6 3 The development of Guideline Daily Amounts (GDAs) 10 4 Research on consumer understanding: traffic-lights vs GDA signals 13 5 Six problems with GDAs 15 6 Conclusion 35 Annex 1 Criteria for defining ‘a little’, ‘moderate’ and ‘a lot’ for specified nutrients in food 37 Annex 2 The IGD’s Guideline Daily Amounts (GDAs) for children 39 Annex 3 The need for a technical review of the Guideline Daily Amounts (GDAs) Page 3 of 46 41 1. Summary In March 2006, after extensive research and consultation with stakeholders in the food industry, the Food Standards Agency (FSA) Board announced its proposals for front-of-pack labelling. While allowing supermarkets and manufacturers to develop their own label designs with an individual look and feel, the FSA recommended that all schemes should comply with four core principles: • • • • provide separate information on fat, saturated fat, sugar and salt use red, amber or green colour coding [traffic lights] to indicate whether levels of these nutrients are high, medium or low use nutritional criteria developed by the FSA to determine the colour code give information on the levels of nutrients per portion of product. In February 2006, a consortium of food companies had announced their determination to proceed with a labelling scheme based on Guideline Daily Amount (GDA) values. This has led to two competing labelling schemes being presented to shoppers simultaneously. (See Figures 1 and 7 for examples.) Both these schemes have been supported by advertising campaigns broadcast in early 2007. In a survey carried out by Which? in 2006, 73% of consumers felt that having a variety of different labelling schemes was confusing.1 To ensure that consumers can use nutritional information wherever they shop and whatever the brand, a consistent, industry-wide scheme is needed. Manufacturers and retailers will always seek to differentiate their brands and to use the front of packages to promote particular product attributes, but we believe that nutritional signpost labelling should not be used in the potentially misleading ways described in this report. The FSA is undertaking a review of the two schemes during 2007 and has publicly committed to adopt the system which is shown to be the best. We believe it is vital that the success of these systems be measured against the following criteria: 1. The system can be easily used by all social and ethnic groups to help them make healthy choices within and between food categories. 2. It is not likely to cause any widening of dietary health inequalities. 3. It is quick and easy to use within the 4 to 10 seconds in which consumers make decisions about food products in shops and supermarkets. 4. It is based on the FSA's or government’s expert advisory groups' dietary guidelines. 1 Which? 2006. Healthy Signs? Which? Campaign Report. London: Which? Page 4 of 46 For the reasons stated in this report, we believe that the traffic-light signposting system recommended by the FSA gives consumers a clearer, more consistent set of signals for making food choices than GDA signals. We believe that information on Guideline Daily Amounts can be a useful supplement – on back of pack – to traffic-light signpost labels. However, we would encourage the FSA to consider, as part of its review of the different schemes, specific recommendations outlined in this report that address some of the weaknesses of the current GDA scheme. Our recommendations are shown below. Recommendations • The Food Standards Agency’s traffic-light signposting system should be used for front-of-pack labelling on all composite, processed food and drink products, to help consumers make healthier choices, easily and quickly. All schemes adopted by manufacturers and retailers should comply with the FSA’s four core principles (shown on page 4). • An independent scientific authority such as the Scientific Advisory Committee on Nutrition (SACN) should review the UK dietary guidelines for fat, saturated fat, sugar and salt. The Guideline Daily Amounts (GDAs) it recommends, and on which the FSA should base its labelling criteria, should be consistent with other initiatives concerned with protecting and promoting public health. • Manufacturers and retailers should apply the FSA’s recommended ‘per portion’ criteria to foods with a serving size of between 100g and 250g. (The criteria are shown in Annex 1.) • The Food Standards Agency should consider recommending that information on NME sugars* be provided on signpost labels. • If manufacturers and retailers wish to use GDAs on back-of-pack in addition to having front-of-pack traffic-light signals: – All schemes should clearly state if the Guideline Daily Amounts are a maximum, minimum or average value. – Age-appropriate GDAs should be used as the basis for calculating GDA values and be shown on products aimed at children and on those of special appeal to children. * NME sugars are non-milk, extrinsic sugars (similar to ‘free’ or ‘added’ sugars), in contrast to intrinsic sugars which are found naturally in milk and milk products and embedded in the cells of fruits, vegetables and seeds before extraction. Page 5 of 46 2. The development of traffic-light labelling For over two decades health and consumer organisations have been urging food companies and policy-makers to introduce a common standard for the presentation of nutritional information about food and drink products, designed to be quickly and readily understood by shoppers. An approach developed by the Coronary Prevention Group in the 1980s defined ‘high’, ‘medium’ and ‘low’ levels of key nutrients in food – principally fat, sugar and salt – and the concept was adopted by a few food companies – notably the Co-operative chain of supermarkets for their ownbrand products. This form of simplified nutritional labelling was not taken up by the majority of manufacturers or supermarkets, and for much of the 1990s the issue remained unresolved. However, with growing concern about the rising prevalence of obesity and other diet-related diseases, and given the amount of ready-prepared, processed foods being consumed in the UK, there is now an urgent need to ensure that consumers are fully informed about the nature of the foods they are buying and can make rapid assessments of the nutritional value of a product. This need was acknowledged in the government’s Choosing Health white paper (2004), which called for clear front-of-pack nutritional labelling.2 In response, the Food Standards Agency has developed proposals for a traffic-light labelling system, using colour coding to identify foods with low, medium or high amounts of four key nutrients – fat, saturated fat, sugar and salt. The nutritional criteria which underpin the FSA system represent a pragmatic expert interpretation of dietary recommendations by the Committee on Medical Aspects of Food and Nutrition and its successor, the Scientific Advisory Committee on Nutrition. It is understood that the FSA intends to keep the nutritional criteria under scientific review. Details of the FSA criteria are given in Annex 1. The aim of the system is to increase clarity for consumers about the comparative benefits of different foods and drinks, but it may also provide an incentive to manufacturers to reformulate their products. The FSA Board agreed a series of four ‘core principles’: manufacturers and supermarkets can develop their own labelling with an individual look and feel, as long as their schemes: provide separate information on fat, saturated fat, sugar and salt use red, amber or green colour coding to indicate whether levels of these nutrients are high, medium or low 2 Department of Health. 2004. Choosing Health: Making Healthy Choices Easier. London: The Stationery Office. Page 6 of 46 use nutritional criteria developed by the FSA to determine the colour code give information on the levels of nutrients per portion of product. Based on research to determine which products consumers had the most difficulty assessing the nutritional content of, the FSA has recommended front-ofpack nutritional signpost labelling for the following products: Sandwiches Prepared or ready meals Burgers and sausages Pies, pasties and quiches Breaded, coated or formed meat, meat alternative, poultry, fish or similar products (for example chicken nuggets, fish fingers or meat balls) Pizzas Breakfast cereals. An example of the FSA traffic-light labelling system is shown in Figure 1. Figure 1. Example of the traffic-light labelling system recommended by the Food Standards Agency The traffic-light system has been, or will be, adopted by several food retailers and manufacturers, notably the Co-operative Group, Sainsbury’s, Waitrose, Marks & Spencer, Asda, New Covent Garden Food Group and McCain’s. Each of these companies has designed its own version of the colour-coded approach. Some examples are shown in Figures 2-5. Figure 2. Example of Waitrose’s traffic-light labelling scheme This scheme is currently featured on Waitrose’s sandwiches and some ready meals, and is due to be extended to other own-brand products. Page 7 of 46 Figure 3. Example of Sainsbury’s ‘Wheel of Health’ colour-coded scheme This scheme has been introduced across a wide range of Sainsbury’s products. Figure 4. Example of traffic-light labelling on New Covent Garden Food Company’s soups Page 8 of 46 Figure 5. Example of McCain’s traffic-light labelling scheme This example, which appears on McCain’s Original Oven Chips, has GDA signals as well as traffic-light labelling. Experience from Sainsbury’s indicates that their traffic-light labelling system – the Wheel of Health – is not only driving customer demand towards healthier products, but also stimulating the retailer’s product technologists to reformulate products to achieve a healthier profile – and colour code – to meet this demand. For example, the product profile of Sainsbury’s Chicken and Bacon Pasta Bake shifted from three ‘red’ signals to only one after reformulation that increased the amount of chicken and reduced the amount of sauce, which in turn reduced the amount of fat and salt in the product (see Figure 6). Energy Total fat Saturated fat Total sugars Salt Chicken and Bacon Chicken and Bacon Pasta Bake Pasta Bake BEFORE AFTER 666kcal 563kcal 26.4g 18.8g 16.4g 12g 6g 1.6g 2.3g 2.0g Figure 6. Sainsbury’s Chicken and Bacon Pasta Bake before and after reformulation Information courtesy of J Sainsbury plc. Page 9 of 46 3. The development of Guideline Daily Amounts (GDAs) Food companies expressed concern that traffic-light labelling would lead to many products showing ‘red’ signals on them – including products labelled as healthy. (For example, Tesco’s ‘Healthy Living’ sunflower spread would get red signals for fat and saturated fat and an amber signal for salt.) The problem highlights the industry’s concerns about differentiating similar products where, for example, one formulation may have lower levels of fat than another but there is little or no gain to public health because even the lower fat version of the product has a considerable amount of fat. It may be argued that in relative terms it is better for consumers to choose one product compared to another, but in absolute terms both products should be consumed less. Traffic-light signalling could potentially label large sections of the food market as ‘consume less’. An alternative to a traffic-light signpost labelling scheme is to provide numerical details of the quantities of nutrients present in a product, along with the quantities needed to ensure adequate health for an average person, and leave consumers to calculate what they should buy in order to meet their own optimal dietary pattern. In 2005 an industry body, the Institute of Grocery Distribution (IGD), published proposals for values that extended the already existing set of Recommended Daily Allowance figures (defined under an EU directive for vitamins and minerals on food labels3) to cover the major nutritional components already being put on many food labels: energy, fats, sugars, salt and dietary fibre. These were called Guideline Daily Amounts, and the IGD recommended that a set of summary figures could be put on the back of packs to help consumers appraise the nutritional value of the foods they were buying.4 The GDAs developed in 2005 are mainly derived from the 1991 report of the Committee on Medical Aspects of Food Policy (COMA) on Dietary Reference Values (DRVs) and on a report on salt published by its successor, the Scientific Advisory Committee on Nutrition (SACN). Both COMA and SACN explicitly stated that the reference values were to be used as pragmatic population targets for public health policies, and were not to be used as advice for individuals. However, this advice appears not to have been followed. The IGD states that GDAs can “help consumers put the nutrition information they read on a food label into the context of their overall diet” and add that consumers should be encouraged to “Know your GDA” for each nutrient. Furthermore on-pack labelling 3 4 Council Directive of 24 September 1990 on nutrition labelling for foodstuffs (90/496/EEC) OJL 276, 6.10.1990, p40. See also: Discussion Paper on Revision of Technical Issues, DG SANCO, May 2006. Institute of Grocery Distribution. 2006. Best Practice Guidance on the Presentation of Guideline Daily Amounts. Institute of Grocery Distribution. Page 10 of 46 often refers to the quantities in a portion of food in terms of a percentage “of your Guideline Daily Amount” (our emphasis). It appears that consumers are being encouraged to consider GDAs as if they were personal targets, contrary to scientific advice. Under the IGD proposals, the on-pack displays do not use colour coding to identify high, medium or low levels of nutrients, and indeed the GDA approach does not define high, medium or low levels of any nutrient. The most recent version of the GDA front-of-pack labelling approach consists of five tabs in a uniform colour showing quantities per serving and an indication of the amount this provides as a percentage of the GDA (see Figure 7). Figure 7. The most recent version of the GDA front-of-pack labelling approach During 2005, GDA signalling was adopted by several leading food companies and retailers including Nestlé, Kellogg’s and Tesco. It soon became apparent that each company had its own way of displaying the information. For example, Tesco initially adopted a format similar to the one shown in Figure 7 but with colour coding in which the different types of nutrient were identified by a colour – for example, pink for sugar, and gold for salt (see Figure 8). Meanwhile, the cereal manufacturers offered widely differing styles of information display (see Figures 9 and 10), including a variety of nutrients. Figure 8. Example of GDA labelling as used by Tesco Page 11 of 46 Figure 9. Example of Kellogg’s use of a bar-chart GDA scheme Figure 10. Example of GDA labelling as used on a Nestlé cereal pack The differing schemes led to accusations that the industry was attempting to confuse consumers rather than help them, and in early 2006 a number of major companies announced that they had adopted a harmonised version, based on the ‘all blue’ version shown in Figure 7. However, in section 5 of this report, we show that manufacturers have not been using GDAs in a harmonised manner, and the problems for consumers are likely to continue. Page 12 of 46 4. Research on consumer understanding: traffic lights vs GDAs Independent surveys, commissioned by the FSA, on the use of the different labelling approaches have shown consumer preferences for a colour-coding system using criteria for high, medium and low levels of the four key nutrients. This approach has been found to be the most acceptable, readily understood and practical approach – especially for consumers with lower educational attainment.5 A further survey undertaken by the consumer organisation Which? involving over 600 consumers also confirmed consumer preference for traffic-light labelling.6 The FSA research showed that a third of consumers were unable to apply the GDA signalling information to food choices, and that in particular consumers from lower socio-economic and ethnic minority groups found it hard to identify the nutrient content of foods using the GDA approach. This is hardly surprising as almost a half (47%) of adults have difficulty using simple percentages,7 and it is estimated that most shoppers spend just 4 to 10 seconds choosing each product. Tesco reports that its own research on labelling preferences, carried out in 2004, showed that customers did not want colour coding. It reported that: consumers wanted straight nutritional facts; traffic lights did not give customers all the information that they needed; customers felt that they should never eat a ‘red’ food such as cheese; and customers did not trust the hidden criteria used for traffic lights.8 Tesco has reported that the GDA system is preferred by ethnic minority customers, and marginally preferred by social class DE customers and the over-55s. However, their research found that the traffic-light coloured label performed better when customers were asked which of the two labels shown in Figure 11 helped them to see the healthiest product ‘at a glance’.9 Note that this research is not in the public domain and has not been independently scrutinised. Tesco traffic-light label Tesco GDA signals Figure 11. Labels tested with customers for Tesco’s research 5 6 7 8 9 Food Standards Agency. Signpost Labelling Research summary. http://www.food.gov.uk/foodlabelling/signposting/siognpostlabelresearch/, accessed 12 August 2006. Which? 2006. Healthy Signs? Which? Campaign Report. London: Which? Department for Education and Skills. 2003. The Skills for Life Survey. London: The Stationery Office. North D. 2006. Nutritional labelling and health: an update. Presentation on behalf of Tesco. June 2006. Marketing Sciences Limited. 2006. Nutritional Signpost Research Findings 2. Prepared for Tesco Stores Ltd. February 2006. Marketing Sciences Limited. Page 13 of 46 A survey by YouGov on behalf of Sainsbury’s among a representative sample of 2,465 adults during September 2006 showed that 75% of people understood the red, amber and green traffic-lights concept, but that 64% had no idea what the GDA system was.10 YouGov found that among customers of each of the four nationwide supermarkets – Tesco, Sainsbury’s, Asda and Morrisons – there was universal preference for a traffic-light system. Tesco and Sainsbury’s customers were those most likely to believe that their supermarket already offered a trafficlight system. To attempt to resolve differences of opinion about the effectiveness of the FSA traffic-light signposting approach and the GDA approach, the FSA and the Department of Health have convened a multi-stakeholder Nutrition Strategy Steering Group to oversee a programme of research that will “evaluate the impact of ‘front-of-pack’ signpost labelling schemes on purchasing behaviour and consumer knowledge”. This programme of research will be undertaken in 2007. A pure baseline data collection will not be possible due to the many schemes rolling out in stores, but sales data will be retrospectively reviewed. 10 YouGov. Press release, 9 October 2006. Page 14 of 46 5. Six problems with GDAs Any examination of the GDA approach needs to consider two aspects: 1 the numerical values representing the dietary guidelines, and 2 how they are applied and represented on front-of-pack signalling schemes. The GDA numerical values for adults are shown in Figure 12. There are different figures for adult women, adult men, and all adults. (The figures for ‘all adults’ are the same as the figures for adult women.) There are also separate GDA values for children in different age groups: 4-6 years, 7-10 years, 5-10 years, 11-14 years and 15-18 years (see Annex 2). Examples of the application of these GDAs to front-of-pack GDA signalling schemes have been given in section 3 of this report. Figure 12. Guideline Daily Amounts for adults, for use on product labels Source: See reference 11. There are six main problems with GDAs. We look at each of these in turn. 11 Institute of Grocery Distribution. 2005. Report of the IGD/PIC Industry Nutrition Strategy Group Technical Working Group on Guideline Daily Amounts (GDAs). Watford: Institute of Grocery Distribution. Page 15 of 46 Problem 1: The GDA values do not distinguish maximum, minimum and average recommended amounts. The first problem with the GDA values is that the numbers do not distinguish between upper limits (i.e. levels of a nutrient, such as salt, which consumers should try to stay within and certainly not to exceed), and lower limits (i.e. levels of a nutrient, such as dietary fibre, which consumers should try to reach and exceed). A good example is the amount of saturated fat recommended for healthy diets. Most reports indicate that no saturated fat is necessary in a healthy diet, and that no more than 10% of total dietary energy should come from saturated fats – i.e. a healthy diet has anything between 0% and 10%, with 10% as the upper limit. The GDA reference level is given in grams based on 10% of energy. This implies that the target is the full 10%, so consumers currently eating less than this amount might be wrongly encouraged to increase their intake. The US Food and Drug Administration nutrition labelling guidance12 requires food companies in the US to include a table in their nutrition panel where room permits. An example is shown in Figure 13. This table is intended to make it clearer to consumers that certain figures – namely those for fats and salt – should be regarded as upper levels, and that consumers should aim to eat less than the amount stated. Figure 13. Example of information shown on nutritional labels in the US Food companies in the US are required to add this table to their nutrition panels, where space permits, to indicate that the daily values for nutrients such as fats and salt are to be regarded as upper maximum limits. 12 US Food and Drug Administration. 2002. Code of Federal Regulations, Title 21, Part 101. Food Labelling. Washington: Government Printing Office. http://www.access.gpo.gov/nara/cfr/waisidx_02/21cfr101_02.html Page 16 of 46 In the US and Canada, new Dietary Reference Intakes published in 2003 by the Food and Nutrition Board of the Institute of Medicine13 are being used with weighted census data to develop population benchmark daily values (DVs) for food labelling purposes. The recommendation of the advisory committee was that values for saturated fatty acids, trans fatty acids and cholesterol should be set at levels which are “as low as possible while consuming a nutritionally adequate diet”. This did not appear to be a consideration in the recommendations of the IGD’s GDA working group, and we believe that the working group’s approach needs to be reviewed, preferably by an independent scientific committee such as the Scientific Advisory Committee on Nutrition. The ambiguity on whether nutrients are to be regarded as maximum, minimum or average levels is compounded when the front-of-pack displays describe the amount present in the product as representing a percentage of the total Guideline Daily Amount (for example, ‘contains 19% of your guideline daily amount’). The effect is magnified further when the information is designed in a bar chart – such as that used initially by Kellogg’s (see Figure 9) – which can give consumers the impression that the ‘ideal’ diet contains all the nutrients in the full 100% quantity, and indeed that ‘more equals better’ so a healthier diet might include even larger amounts still. Consumers comparing products might choose those with the greater proportions, or longer bars, because the labelling implies that the product would meet a greater part of the consumer’s daily needs, when in fact the product supplies high levels of fats, sugar or salt. Recommendations If manufacturers and retailers wish to use GDAs on back-of-pack in addition to having front-of-pack traffic-light signals, all schemes should clearly state if the Guideline Daily Amounts are a maximum, minimum or average value. The GDA values should be established by an independent scientific authority such as the Scientific Advisory Committee on Nutrition (SACN). (See recommendation on page 23.) 13 Food and Nutrition Board of the Institute of Medicine. 2003. Dietary Reference Intakes: Guiding Principles for Nutrition Labeling and Fortification. Washington: The National Academies Press. http://www.nap.edu/catalog/10872.html, accessed 12 August 2006. Page 17 of 46 Problem 2: GDA values for adults and for children are used inconsistently, and adult GDAs are sometimes used on child-targeted products. The confusion about the apparent amounts of nutrients consumers need is further exacerbated by the use of several different sets of GDA reference values. As explained in section 3, there are different sets of GDA values for adult women, adult men, all adults and also for different age groups of children. Manufacturers can choose whichever set of values they want to use as their reference GDAs. The choice affects the proportion of GDAs in a portion. For example, if a manufacturer chooses to use the adult GDA value for sugar as the reference value for a childtargeted product, this can imply that the product supplies a smaller proportion of the Guideline Daily Amount of sugar than if the GDA value for children aged 5-10 had been used. Some examples of products which are specifically targeted at children and which use GDA values for adults include: KP Snacks Space Raiders, Kellogg’s Frosties, Kellogg’s Rice Krispies, Kellogg’s Ricicles (see Figure 14), Kellogg’s Coco Pops, Nestlé Shreddies ‘School Fuel’, Nestlé Cheerios and Honey Nut Cheerios, and Weetabix Chocolate Weetos. Figure 14. Kellogg’s Ricicles: An example of a breakfast cereal aimed at children, where the GDA signals are based on GDA reference values for adults Page 18 of 46 In some cases, a mixture of different sets of GDA values is used. For example, on Quaker Oats’ Sugar Puffs the GDA signals on the pack are based on a mixture of different sets of GDA values, with the adult values for calories, fat and salt, but the company has chosen a higher value, 100g (the value for boys aged 7-10), for their GDA for total sugars (see Figure 15). Using different bases for the GDA values (a practice which is not recommended by the IGD) adds to consumers’ confusion about how to interpret GDAs. Figure 15. Example of a product which uses different sets of GDA values as the basis for their labelling Quaker Oats’ Sugar Puffs uses adult GDA values for calories, fat and salt, but not for total sugars, thus adding to consumer confusion. Recommendation If manufacturers and retailers wish to use GDAs on back-of-pack in addition to having front-of-pack traffic-light signals, age-appropriate GDAs should be used as the basis for calculating GDA values and be shown on products aimed at children and on those of special appeal to children. Page 19 of 46 Problem 3: The GDAs used for labelling are based on values which are not the most suitable either for public health policy or for individuals. There are concerns that the industry has chosen values for the GDAs which may not be the best for improving public health. For example, the values for salt used in the GDAs for individual adults and children are those proposed by the government advisory body, the Scientific Advisory Committee on Nutrition (SACN), as population goals. SACN stated that the salt levels they were recommending as targets for the population were “substantially greater than the salt intake required” and “do not represent ideal or optimum consumption levels, but achievable population goals”.14 A comparison of the IGD’s GDA recommendations for adults with those of various authoritative reviews undertaken since the early 1980s is shown in Table 1. It can be seen that in many respects the GDA values have been selected in favour of encouraging the consumption of greater quantities of fats, saturated fat, sugar and salt and smaller amounts of fibre, running against the trends in dietary advice being given in the authoritative reviews. For example, the trend in the value for the proportion of dietary energy derived from total fat recommended in recent reports has been consistently under 30% of dietary energy, but the GDA value is 33% of energy, based on the 15-year-old COMA report recommendation for total fat. Similarly, the figures for saturated fatty acids recommended in authoritative reviews have progressively decreased to under 10% in the Eurodiet (2000)15 and WHO (2003)16 expert reports, but the GDA value is set at 10% for adult women, and at a higher percentage for men and for children in some of the age groups. As noted above, the GDA value for salt is fixed at the highest level of all the reports listed in Table 1, while the figures for dietary fibre and protein are as low, or lower than those given in any of the reports. 14 15 16 Scientific Advisory Committee on Nutrition. 2003. Salt and Health. London: The Stationery Office. Kafartos A, Codrington CA. 2001. Nutrition and diet for healthy lifestyles in Europe: the EURODIET evidence. Public Health Nutrition; 4; 2(A) and 2(B). See also: Eurodiet Core Report, available at http://eurodiet.med.uoc.gr/ World Health Organization. 2003. Diet, Nutrition and the Prevention of Chronic Diseases. Report of a Joint WHO/FAO Expert Consultation. Technical Report Series No 916. Geneva: WHO. Page 20 of 46 Table 1 Comparison of GDA values with the recommendations of scientific reports 1984-2004 %E AOAC = = percentage of dietary energy Association of Official Analytical Chemists Component UK COMA (Diet and 17 CVD), 1984 WHO, 199018 NSP RNI = non-starch polysaccharides = Reference Nutrient Intake UK COMA (DRV), WCRF, 199119 199720 EURODIET, 200121 WHO, 200322 US FDA, 2000-0423 2,500kcal and 2,000kcal (examples) 2,550kcal men 1,940kcal women Energy at age 20-50, average kcal/d Dietary fat %E 31–35% (77-87g/d) 15–30% Saturated fatty acids %E 15% 37g/d 0–10% 33% 15–30% less than 30% 15–30% less than 10% less than 10% less than 10% over 55% 55–75% Carbohydrates %E 55–75% 47% 55–75% Free/Non-milk extrinsic sugars %E 0–10% 10–11% (60g/d adults) less than 10% less than 10% Total sugars %E Dietary fibre (AOAC) g/d 27–40g/d Dietary fibre (NSP) g/d [16–24g/d] Protein %E or g/d Sodium (as salt NaCl) (g/d) 17 18 19 20 21 22 23 24 less than 6g/d IGD Guideline Daily Amounts (GDAs), 200524 over 25g/d (or 3g/MJ) 18g/d 20–35g/d 56g/d men 45g/d women 9–12% RNI = 4g/d less than 6g/d less than 6g/d over 25g/d from foods over 20g/d from foods 2,500kcal men 2,000kcal women or adults 33% less than 30% (95g/d men, 70g/d women (80g/d, 65g/d examples) or adults) 10% less than 10% (30g/d men, 20g/d women (25g/d, 20g/d examples) or adults) 47% 60% (300g/d men, 230g/d (375g/d, 300g/d examples) women or adults) 10% (65g/d men, 50g/d women or adults) 11% for children 19% (120g/d men, 90g/d women or adults) 25g/d and 30g/d examples 24g/d 18g/d 10–15% 8–9% (55g/d men, 45g/d women or adults) less than 5/d less than 6g/d 6g/d Committee on Medical Aspects of Food Policy. 1984. Diet and Cardiovascular Disease. Report on Health and Social Subjects 28. London: Department of Health. World Health Organization. 1990. Diet, Nutrition and the Prevention of Chronic Diseases. Technical Report Series No 797. Geneva: WHO. Committee on Medical Aspects of Food Policy. 1991. Dietary Reference Values for Food Energy and Nutrients for the United Kingdom. Report on Health and Social Subjects 41. London: Department of Health. World Cancer Research Fund. 1997. Food, Nutrition and the Prevention of Cancer: A Global Perspective. London: WCRF. Kafartos A, Codrington CA. 2001. Nutrition and diet for healthy lifestyles in Europe: the EURODIET Evidence. Public Health Nutrition; 4: 2(A) and 2(B). World Health Organization. 2003. Diet, Nutrition and the Prevention of Chronic Diseases. Report of a Joint WHO/FAO Expert Consultation. Technical Report Series No 916. Geneva: WHO. US Food and Drug Administration. How to Understand and Use the Nutrition Facts Label. June 2000; updated July 2003 and November 2004. http://www.cfsan.fda.gov/~dms/foodlab.html#see6, accessed 5 August 2006. Institute of Grocery Distribution. 2005. Report of the IGD/PIC Industry Nutrition Strategy Group Technical Working Group on Guideline Daily Amounts (GDAs). Watford: Institute of Grocery Distribution. Page 21 of 46 There are also particular concerns about dietary fibre and forms of sugar. The IGD recommends that all companies avoid using NSP dietary fibre, despite the recommendation from the UK COMA report25 that the health effects of fibre are most strongly related to NSP measures of fibre Also, the IGD endorses the declaration of total sugars rather than added sugars or NME sugars on GDA signals. NME sugars are non-milk extrinsic sugars (similar to ‘free’ or ‘added’ sugars), in contrast to intrinsic sugars which are found naturally in milk and milk products and embedded in the cells of fruits, vegetables and seeds before extraction. Most sugar in manufactured foods is NME sugar. The presentation of ‘total sugars’ as a GDA can be significantly misleading to consumers as it can imply that the total sugar value can be used in place of the NME sugar value. For example, while the IGD acknowledges that adults should consume only 50g of NME sugar per day, they suggest that labels should refer to the quantity of total sugars that could be consumed in a healthy diet which, for adults, is a GDA of 90g per day. By using the value for total sugar as the basis for labelling, consumers may be led to believe that sugar consumption of, say, 90g would be a reasonable daily level of consumption for an adult, even if all the sugar was NME sugar. In fact, the consumption of 90g sugar – for example from about three cans of soft drink – would exceed the Guideline Daily Amount for NME sugar by 44%. The problem with the two different forms of sugar can be illustrated using the example shown in Figure 16, taken from a pack of Tesco chocolate muffins, where the sugar content of a single muffin is nearly 50% of an adult’s recommended NME sugars. Virtually all the sugar in the product is NME sugar, yet the label is able to suggest that the product contains only 26% of a Guideline Daily Amount. If the same pack had been labelled using the FSA traffic-light signposting system, the sugar would be labelled with a ‘red’ signal. Figure 16. Example of potentially misleading labelling of sugar content: Tesco chocolate muffin One Tesco chocolate muffin contains nearly half of an adult’s GDA for NME sugars. However, as the label refers only to total sugar, the GDA signal can state that one muffin contains only 26% of the GDA for sugar. 25 Committee on Medical Aspects of Food Policy. 1991. Dietary Reference Values for Food Energy and Nutrients for the United Kingdom. Report on Health and Social Subjects 41. London: Department of Health. Page 22 of 46 Although many consumer organisations would prefer to have NME sugars labelled separately and not combined with intrinsic sugars, food manufacturers have objected strongly to this on the grounds of cost and commercial confidentiality. The Food Standards Agency has accepted these objections and has made proposals for classifying total sugars within their recommended traffic-light signposting system. On the principle that the sugars in most processed foods are NME sugars, the FSA has suggested that this be formalised on the basis of there being 20g of intrinsic or milk sugars for every 100g of NME sugars present,26 based on data from the National Diet and Nutrition Survey.27 This leads to a ‘red’ classification for total sugars if a product contains more than 15g sugars per 100g (or more than 7.5g per 100ml for liquids). These criteria for total sugars ensure that fresh milk and most fresh fruit fall below the threshold for a ‘red’ signal. In addition, under the FSA scheme manufacturers can make a statement about the sources of sugar contained in a product to facilitate consumer choice. Recommendations 26 27 An independent scientific authority such as the Scientific Advisory Committee on Nutrition (SACN) should review the UK dietary guidelines for fat, saturated fat, sugar and salt. The Guideline Daily Amounts (GDAs) it recommends, and on which the FSA should base its labelling criteria, should be consistent with other initiatives concerned with protecting and promoting public health. . The Food Standards Agency should consider recommending that information on NME sugars be provided on signpost labels. Rationale and discussion of the 15g high sugars criterion with the FSA signposting scheme. Discussion document prepared for FSA consultation meeting, London, 14 December 2006. Henderson L et al. 2003. The National Diet and Nutrition Survey: Adults Aged 19 to 64. Volume 2: Energy, Protein, Carbohydrate, Fat and Alcohol Intake. London: The Stationery Office. Page 23 of 46 Problem 4: The GDA displays are based on arbitrary portion sizes. When consumers make comparisons between products, they need to know that the figures being quoted are truly comparable, and that the information properly compares like with like. One of the FSA core principles is the requirement to use red, amber or green colour coding according to nutrient levels per 100g or 100ml of the product, with further labelling rules applying only if portion sizes are larger and hence likely to supply a considerable part of the day’s dietary intake. Portion sizes determined by the manufacturer can vary considerably and the consumer can be faced with significant problems in assessing the content of a product and in making comparative judgements even within a product category. For example, cereal manufacturers often use 30g portions for their nutrition information labels, but sometimes 35g, and sometimes even 45g. Furthermore, the cereal manufacturers’ individually boxed portions, often used in catering services, provide yet another set of portion sizes which adds to the confusion. (For example, on Kellogg’s Corn Flakes packs a portion size is 30g, but their one-pack portion is 17g. A standard portion of Nestlé Clusters is 30g, but their one-pack portion is 40g.) In addition, cereal companies are particularly keen to use ‘portion with a serving of semi-skimmed milk’ rather than a portion of the product contained in the box. The effect of this is to narrow the differences between products by adding to all of them a standard amount of fat, saturated fat and sugar (the intrinsic sugar found in milk in this case). Leaving manufacturers to determine portion sizes may lead to some clearly unrealistic quantities being stated as a portion, allowing the product to gain a very favourable profile. For example, the packaging from a pack of Cheese Singles from the Tesco ‘Healthy Living’ range shown in Figure 17 on the next page, gives nutritional values for a single slice (equivalent to 5% of the pack) indicating that it provides very small amounts of a consumer’s Guideline Daily Amount of saturates and salt. Under the FSA traffic light signposting scheme, this cheese would be colour-coded red for saturates and salt. The serving suggestion pictured on the package shows five slices, containing a total of 3g of salt – equivalent to half an adult’s maximum daily salt intake. Page 24 of 46 Figure 17. Example of potential confusion over what constitutes a portion size: Tesco’s ‘Healthy Living’ Cheese Singles On this pack of Cheese Singles, GDA signals are given per slice, the picture shows a serving suggestion with 5 slices, and on the back of pack Tesco states that the product is equivalent to one Weight Watchers point per 20g serving. 20g is a single slice. A similar problem is seen on packs of biscuits, where a single biscuit is often reported as a portion, even though biscuits vary considerably in size and weight, and differ in the likelihood that an adult would consume a single item. The issue is highlighted in the case of the cream crackers package shown in Figure 18, in which a single cracker is deemed to be a portion even though a picture, with the caption ‘Serving suggestion’, shows two crackers. Under the FSA traffic-light Page 25 of 46 signposting system, these crackers would be flagged as amber for total fat and for salt, and red for saturated fat. Figure 18. Tesco Cream Crackers: an example of disparity between the portion size and serving suggestion A portion for GDA purposes is a single cracker, even though the ‘serving suggestion’ shows two crackers. Figure 19. Tesco Original Cola (2-litre bottle) The labelling for this soft drink states the proportion of the GDAs provided by 100ml (20 teaspoons) of the drink. Information for this product on the Tesco website gives GDAs by 200ml serving,28 whereas a typical serving (can) of cola would be 330ml. Arbitrarily switching between ‘per 100 ml’ and ‘per serving’ signposting can lead to confusion among consumers. 28 www.tesco.com/superstore/xpi/9/xpi54807749.htm. Accessed 1.2.2007 Page 26 of 46 These anomalies in the definitions of a portion, and the confusing employment by manufacturers of different sizes (and additions such as milk to a cereal serving) make it very difficult for consumers to make reliable comparisons. Colour coding of portion sizes brings its own concerns. Sainsbury’s has opted to undertake both ‘per 100g’ and ‘per portion’ analyses and then ascribes a colour according to whichever gives the higher-value result. This approach puts public health ahead of marketing benefits. Initially the FSA proposed basing all colour coding on 100g quantities but, recognising that some products were consumed in large quantities (for example, ready meals and soft drinks) and that they should be flagged as potentially contributing a large amount of a nutrient, the FSA suggested that colour coding would also apply to a portion of a product if that portion size exceeded 250g. (For more details on the FSA criteria, see Annex 1.) The FSA has now suggested that, by 2008, colour coding should be based on portion size if the portion exceeds 100g.29 This moves it more into line with the approach taken by Sainsbury’s. We believe that using the FSA ‘per portion’ criteria for foods with a serving size of between 100g and 250g will enable consumers to make reliable judgements about foods sold in larger portion sizes. The use of portion sizes below 100g can mislead consumers. An example has been highlighted by Professor Tom Sanders in a presentation made to the cereal manufacturers.30 He showed that using a ‘per 100g’ colour coding gave far more differentiation between 21 examples of different breakfast cereal products than did a ‘per portion’ colour coding which used the same threshold values as the FSA but applied to the smaller quantities found in typical breakfast servings (see Figure 20 on the next page).* * Professor Sanders concluded that the FSA’s ‘per 100g’ coding “unfairly misclassified breakfast cereals.” 29 30 Food Standards Agency. 2007. Front of Pack Nutritional Signpost Labelling Technical Guidance. Issue 1: January 2007. London: Food Standards Agency. Sanders T. 2006. Breakfast Cereals Up Front – Debating the Issue of Food Labelling. Presentation to the Breakfast Cereal Information Service, London, 7 March 2006. http://www.breakfastcereal.org/BreakfastCerealsUpFront.ppt, accessed 4 August 2006. Page 27 of 46 Cereal 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 sample no. Per 100g Fat Saturated fat Sugar Salt Per portion Fat Saturated fat Sugar Salt Figure 20. Comparison of colour coding of labels of nutrients in cereals on the basis of per 100g and per portion The colour coding was calculated for 21 breakfast cereals on a per 100g and per portion basis. The use of portions for colour coding reduced the differentiation between products because the quantities of fat, sugar and salt were too low to reach significant threshold values. Recommendation • Manufacturers and retailers should apply the FSA’s recommended ‘per portion’ criteria to foods with a serving size of between 100g and 250g. (The criteria are shown in Annex 1.) Page 28 of 46 Problem 5: GDA signals for different nutrients are sometimes included or left out in an arbitrary and confusing manner. Although good practice might dictate that consumers need to find matching forms of information on different products so that they may assess the products and compare them rapidly, manufacturers have not necessarily put this into practice. The three problems are: 1 manufacturers adding extra GDA signals onto the front of packs 2 manufacturers occasionally adding GDAs for which no agreed value has been recommended by the IGD; and 3 the selection of only some of the principal GDA signalling tabs, on the grounds that there is insufficient room on the package to show them all. Some examples of extra GDA tabs are shown in Figure 21. Kellogg’s introduces a GDA for iron in its front-of-pack signalling for Corn Flakes (which contains iron as a fortificant), and for fibre and iron (a fortificant) for All Bran; and for both iron and calcium for Coco Pops (where both minerals are added as fortificants). It would be more helpful to consumers if a standard set of GDA signals were added to all products. Note that the IGD has not agreed a GDA for iron or calcium. Corn Flakes All Bran Figure 21. Examples of ‘extra’ GDA signals: Kellogg’s Corn Flakes and All Bran Kellogg’s adds iron GDA signalling to its Corn Flakes pack, and both fibre and iron GDA signalling to its All Bran. Page 29 of 46 Another example of an extra GDA tab can be found with Nestlé, who have created a new GDA entitled ‘whole grain’ and determined that 48g of whole grain constitutes 100% of an adult’s Guideline Daily Amount, apparently based on US Food and Drug Administration guidance. Forty eight grams of whole grain is the amount contained in a portion of the company’s Shredded Wheat (which, coincidentally, weighs 45g). The result is that the front-of-pack signalling on Shredded Wheat includes a tab for ‘whole grain’ claiming that a portion contains 100% of the Guideline Daily Amount for most adults (see Figure 22). 2005-2006 version 2006+ version Figure 22. Examples of an ‘extra’ GDA signal for which there is no agreed value: iron in Nestlé’s Shredded Wheat Nestlé has added a new GDA – for ‘whole grain’ – to the front-of-label signalling for Shredded Wheat. Another type of breach of the IGD’s GDA signalling scheme can be demonstrated when non-GDA nutrients are highlighted in a similar format to GDA nutrients, but without stating any Guideline Daily Amount or portion amount. It is a real concern that these additional ‘GDAs’ appear to function as a generic – but unregulated – health claim logo. In the example shown in Figure 23, the phrase ‘High in Omega 3’ is included alongside the GDA tabs and is associated with the word ‘Benefit’, and the word ‘High’ is used even though the definition of ‘high’ is not referred to on the label. Interestingly, very few other products – even other fish from the same manufacturer – have an omega 3 signal like this, with which to compare the product. Page 30 of 46 Figure 23. Example of the inclusion of a non-GDA nutrient in the GDA signal format Tesco states, in the GDA signal format, that its mackerel is ‘High in Omega 3’ and that it has a ‘Benefit’ but does not state a GDA or the proportion of GDA contained in a serving. Further confusion may occur when various types of nutritional claims, allergy advice, information about vegetarian suitability and cooking instructions are included in a similar format and in the same location as the GDA signals. See Figure 24. Muesli Instant Oats Figure 24. Examples of other signals being shown in the same location as GDA signals: Tesco’s Muesli and Instant Oats Tesco’s cereal packs add a range of other signals in the same location as the GDA signals, including information on suitability for vegetarians and cooking instructions. Note also that different portion sizes are used (50g of Muesli, 30g of Instant Oats) and the calculations include different quantities of added milk (125ml with the Muesli, 150ml with the Instant Oats). As consumers are likely to read information from left to right, it is unfortunate that the most significant health-related information is not given first. From the examples shown in Figure 24, it might be concluded that these irregular signals are being added to the standard set of GDA signals for product marketing purposes rather than the promotion of public health. Another problem occurs when relevant pieces of information are excluded from the GDA front-of-pack signalling. According to Nestlé, smaller food items may have only one GDA signal on the front, showing the energy content of the food. This is a Page 31 of 46 huge disadvantage to consumers who, for example, may wish to compare confectionery with other forms of snack food which may, for example, be much lower in saturated fat or sugar, than a chocolate bar. Nestlé gives as an example their four-bar KitKat pack which, they claim, would not be able to show the full set of five GDA signals. The company has also introduced single GDA labelling (for calories) for the front-of-pack labels of their individual portion packs of breakfast cereals. Figure 25. Example of a food item with only one GDA signal: Nestlé’s KitKat According to the manufacturer there is room for only one GDA signal (for calories) on the front of a KitKat bar. The signposting inconsistencies described above would not arise if manufacturers and retailers followed the four core principles recommended by the FSA (shown on page 4). Page 32 of 46 Problem 6: The standard GDA signals lack colour coding for quick consumer appraisal and interpretation. Most manufacturers and retailers now place detailed nutritional information on back-of-pack for customers. The key value of a front-of-pack signalling scheme is to allow rapid appraisal of key facts. The most suitable method, as determined by the FSA’s research into consumer preferences and confirmed by an independent survey from the consumer organisation Which?, is to use a traffic-light signalling system in which green is used to indicate ‘go for it’ or ‘eat freely’, amber indicates ‘OK most of the time’ or ‘eat in moderation’, and red indicates ‘enjoy it once in a while’ or ‘eat sparingly’. (See section 4.) The industry-preferred GDA signalling scheme does not use colour coding, and most companies have used a single-colour format (see Figure 26), with the exception of the initial Tesco version which, confusingly, uses colours to differentiate the types of nutrient rather than the levels of those nutrients in a product. Figure 26. The most recent version of the GDA front-of-pack labelling approach that is commonly accepted by industry The numerical values are shown in pale blue tabs. One potential compromise between the conflicting signalling schemes is to have a composite signalling scheme which combines traffic-light colour coding with the GDA numerical information. However, evaluation of combination labels by the FSA has indicated that this may not perform well in terms of helping consumers to make rapid choices of healthier foods.31 Also, a consumer survey by Which? indicated that traffic-light labels without the percentage information performed better than numerical labels with added traffic-light colours.32 A composite signalling scheme would also be dependent on agreement over the GDA values, the criteria for colour coding and the use of per 100g or per portion. 31 32 Food Standards Agency. Signpost Labelling Research summary. http://www.food.gov.uk/foodlabelling/signposting/siognpostlabelresearch/, accessed 12 August 2006. Which? 2006. Healthy Signs? Which? Campaign Report. London: Which? Page 33 of 46 Sainsbury’s interprets the FSA core principles in its Wheel of Health, front-of-pack scheme (see Figure 27), although it does not use the words ‘high’, ’medium’ or ‘low’. In addition to the FSA’s recommendations, Sainsbury’s include the calorie content of one portion of the food. Where the Wheel of Health does not currently fulfil the FSA core principles is in the use of non-FSA threshold criteria for colour coding of sugar. This could give rise to some inconsistencies between Sainsbury’s traffic-light-based labels and those adopted by Waitrose, the Co-op and others. Figure 27. An example of Sainsbury’s Wheel of Health scheme This provides colour-coding with portion information. Recommendation The Food Standards Agency’s traffic-light signposting system should be used for front-of-pack labelling on all composite, processed food and drink products, to help consumers make healthier choices, easily and quickly. All schemes adopted by manufacturers and retailers should comply with the FSA’s four core principles (shown on page 4). Page 34 of 46 6. Conclusion The primary purpose of front-of-pack nutritional signpost labelling on processed foods should be to help all consumers make healthier choices and encourage the food industry to make healthier food products through reformulating their products in response to the stimulated consumer demand. This will be an important influence in shaping a more healthy food culture and economy. Dietary inequalities closely mirror social and health inequalities in the UK. Nutritional labelling schemes should not widen health inequalities by being useful only to nutritionally and numerically literate consumers: clear and simple nutritional labelling should be a consumer’s right, not an optional extra. Effective signpost labelling is especially helpful for consumers who are poorer, time-pressured, or with lower educational attainment, to enable them to choose healthier products. Published research shows that the FSA’s traffic-light approach performs better than the numerical GDA signalling scheme in enabling the most disadvantaged consumers to discriminate between high, medium and low levels of key nutrients and to make healthier choices quickly and easily. This finding is unsurprising when we consider the specific problems with the GDA signalling approach outlined in this report. Given the current epidemics of chronic disease in the UK, the immediate priority for nutritional signposting has to be fat, saturated fat, sugar and salt. The high levels of these nutrients in our diet have contributed significantly to the high prevalence of avoidable premature deaths and disability brought about by the current epidemics of heart disease, diabetes, obesity, stroke and diet-related cancers. Given the high and increasing prevalence of obesity, further consideration should be given to the inclusion of calorie content in the FSA’s core principles for front-of pack labelling. It is a consumer’s right to have easy-to-understand information about what has been added to processed foods and the impact this may have on their own and their family's health. Until recently, the food industry has not shown any appetite to provide the public with additional clearer information about the nutrient value of foods. Meaningful front-ofpack labelling should have been introduced decades ago. If it had been, the UK and other countries might have avoided much of the current dietary chronic disease crisis. However, in the wake of major concerns about obesity, the EU and the UK government have signalled their wish to see the food industry voluntarily introduce consumer-friendly front-of-pack nutrition labelling under the threat of statutory regulation. In the UK, some in the food industry have broken away from the government-industry partnership on signpost labelling and we now have the unfortunate situation of two competing front-of-pack labelling schemes in the UK: the FSA's traffic-light signposting system and a GDA system developed by the IGD and a number of food manufacturers and retailers. Public opinion research has shown that the public would prefer one scheme, and current independent published research shows that the FSA's traffic-light system performs better in enabling consumers to make healthier food choices. Page 35 of 46 The government, in a renewed spirit of partnership with the food industry, has commissioned an evaluation of the two systems in 2007. It has established a high level nutrition group, involving industry and consumer organisations, and has publicly committed to adopt the system which is shown to be the best. We believe that, to be successful, it is vital that any system is measured against the following criteria: 1. The system can be easily used by all social and ethnic groups to help them make healthy choices within and between food categories. 2. It is not likely to cause any widening of dietary health inequalities. 3. It is quick and easy to use within the 4 to 10 seconds in which consumers make decisions about food products in shops and supermarkets. 4. It is based on the FSA's or government’s expert advisory groups' dietary guidelines. We do not believe that the current GDA scheme developed by the IGD and a group of 21 food producers and retailers meets any of these criteria. There may be merits in labelling systems based on GDAs, but only if they are designed against the criteria outlined above. This will necessitate further independent peerreviewed research. In the meantime the government, through the Department of Health and the Food Standards Agency, should promote the use of the traffic-light signposting system on front-of-pack and continue to evaluate its use and further development. The European Commission should take a similar line in its review of the European food and drinks labelling directive. The European Commission should not introduce any measures that impede the important progress being made on traffic-light front-of-pack signpost labelling in the UK which provides a useful testing ground for the development of front-of-pack labelling across Europe. Page 36 of 46 Annex 1 Criteria for defining ‘a little’, ‘moderate’ and ‘a lot’ for specified nutrients in food Current advice from Food Standards Agency (FSA) publications and website33 All products, per 100g Fat Saturates Total sugars Salt A little below 3g below 1g below 2g below 0.25g Moderate 3 – 20g 1 – 5g 2 – 10g 0.25 – 1.25g A lot over 20g over 5g over 10g over 1.25g FSA criteria proposed for traffic-light labelling signals34 Foods, per 100 grams Fat Saturates Total sugars Salt Low GREEN below 3g below 1.5g below 5g below 0.3g Medium AMBER 3 – 20g 1.5 – 5g 5 – 15g 0.3 – 1.5g High RED over 20g over 5g over 15g over 1.5g Low GREEN below 1.5g below 0.75g below 2.5g below 0.3g Medium AMBER 1.5 – 10g 0.75 – 2.5g 2.5 – 7.5g 0.3 – 1.5g High RED over 10g over 2.5g over 7.5g over 1.5g Liquids, per 100 ml Fat Saturates Sugar Salt For portions above 250g the following criteria apply per portion: High RED Fat 21g or more Saturates 6g or more Sugar 18g or more Salt 2.4g or more 33 34 A series of publications on label reading and dietary advice from MAFF and the FSA, 1998 onwards, and FSA advisory website: http://www.eatwell.gov.uk/healthydiet/nutritionessentials/fatssugarssalt/fats/#cat225126, accessed 4 August 2006. Food Standards Agency. 2007. Front of Pack Nutritional Signpost Labelling Technical Guidance. Issue 1: January 2007. London: Food Standards Agency. Page 37 of 46 Labelling criteria for low and high levels of nutrients per serving in the US35 Calories: 40 Calories is low, 100 Calories is moderate, 400 Calories or more is high. Guide for Dietary Values: 5% DV or less is low and 20% DV or more is high. For a 2,000kcal diet, this equates to the following (in grams per serving): Fat Saturated fat Salt Carbohydrate Fibre 35 Low up to 3.25 up to 1.0 up to 0.3 up to 15 up to 1.25 High 13 or more 4.0 or more 1.2 or more 60 or more 5 or more US Food and Drug Administration. 2004. How to Understand and Use the Nutrition Facts Label. http://www.cfsan.fda.gov/~dms/foodlab.html, accessed 4 August 2006. Page 38 of 46 Annex 2 The IGD’s Guideline Daily Amounts (GDAs) for children This annex contains details of the GDA values for children, developed by the IGD.36 GDAs for adults are shown in Figure 12. Figure 28. Guideline Daily Amounts for boys aged 4-18 years 36 Institute of Grocery Distribution. 2005. Report of the IGD/PIC Industry Nutrition Strategy Group Technical Working Group on Guideline Daily Amounts (GDAs). Watford: Institute of Grocery Distribution. Page 39 of 46 Figure 29. Guideline Daily Amounts for girls aged 4-18 years Figure 30. Guideline Daily Amounts for children aged 5-10 years Page 40 of 46 Annex 3 The need for a technical review of the Guideline Daily Amounts (GDAs) As suggested in the main text of this report, there are good reasons to suggest that the GDA values recommended by the IGD should be reviewed. In this Annex we cite in further detail some of the concerns which need to be addressed, and which lead us to believe that an independent agency, such as the Scientific Advisory Committee on Nutrition, should be asked to provide a new set of GDA values which can be used to help promote better public health. Background The Institute of Grocery Distribution (IGD) is a trade body for UK food retailers and manufacturers and is concerned with consumer issues. The IGD’s Policy Issues Council (PIC) first developed GDAs for energy, fat and saturated fat in 1998.37 These were based on the 1996 Guideline Daily Intakes (GDIs) published in the Ministry of Agriculture, Fisheries and Food (MAFF) leaflet Use your label. The IGD/PIC Industry Nutrition Strategy Group (INSG) was established in 2003 to “examine how the industry can play its part in encouraging healthy eating”, and in 2005 the IGD/PIC technical working group reviewed and expanded the GDAs. GDAs were intended for back-of-pack38 and were to be independent of, but where possible compatible with, FSA signposting. Most of the 2005 GDAs are derived from the 1991 report of the Committee on Medical Aspects of Food Policy (COMA) on Dietary Reference Values (DRVs).39 The GDA for salt was derived from the recommendations of the Scientific Advisory Committee on Nutrition (SACN)40 and the GDA for sugar is based on calculations described in a paper by Rayner et al.41 The GDAs have been calculated for men, women and all adults, and for children in defined age groups (see Annex 2). The GDAs for energy for adults The GDAs for energy for adults are based on the 1991 COMA Estimated Average Requirements (EARs) but rounded to convenient figures, which also match the figures used in the USA: EAR (kcals) GDA (kcals) Women 1,940 2,000 Men 2,550 2,500 The 1991 EAR values for energy (for children over 10 years and adults) were calculated on the basis of energy expenditure, which is more reliable than the measurement of energy intake (which is often mis- or under-reported and liable to subject biases). Energy expenditure is made up of basal metabolic rate (BMR, the metabolic rate at rest), thermic effect of the food eaten, and energy used in 37 38 39 40 41 www.IGD.com Institute of Grocery Distribution. 2005. Report of the IGD/PIC Industry Nutrition Strategy Group Technical Working Group on Guideline Daily Amounts (GDAs). Institute of Grocery Distribution. Committee on Medical Aspects of Food Policy. 1991. Dietary Reference Values for Food Energy and Nutrients for the United Kingdom. Report on Health and Social Subjects 41. London: Department of Health. Scientific Advisory Committee on Nutrition. 2003. Salt and Health. London: The Stationery Office. Rayner M, Scarborough P, Williams C. 2003. The origin of Guideline Daily Amounts and the Food Standards Agency’s guidance on what counts as ‘a lot’ and ’a little’. Public Health Nutrition: 7 (4): 549-556. Page 41 of 46 physical activities. BMR is principally determined by body weight (although both mass and composition are important) and varies by gender and with age. It was estimated using age/sex specific regression equations based on data collected in 1985 from 1,300 10-17 year olds, 3,500 men and 1,200 women42 and additional measurements in 451 elderly people (over 60 years). In the calculation of the EARs for energy, COMA used a PAL of 1.4 (i.e. a Physical Activity Level of 1.4 times BMR) and average bodyweight (from data collected in 1984) of 74kg for men aged 19-59, 60kg for women aged 19-49, and 63kg for those aged 50-59 years). The EARs for energy are based on population weight data and assumptions about physical activity levels which are over 20 years old. The mean bodyweight of the population has increased significantly, with obesity prevalence trebling over this period.43 Studies also point to a steady decline in population physical activity levels.44 Additionally, the EARs for energy do not adequately reflect the increasingly aging and multi-ethnic nature of the UK population. The FAO/WHO/UNU have published an Expert Consultation Report on Human Energy Requirements45 and, with these issues in mind, SACN is conducting a review of the COMA EARs.46 The GDAs for energy for children The original 1991 EAR values for children aged 3-10 years were based on estimates of dietary energy intake rather than the more reliable EE methods. These estimates need to be reviewed, especially given the recommendations of the FAO/WHO/UNU Expert Consultation Report on Human Energy Requirements (mentioned above) which identified significantly lower energy needs for younger children than those previously assumed. Furthermore, the recommendations of the World Health Organization concerning standards for child growth based on a set of ‘gold standards’ (derived from a set of children exclusively breastfed and optimally nourished) indicate possible further reductions in the recommended energy requirements for younger children, which may eventually apply to older children also.47 The GDAs for macronutrients GDAs for the macronutrients are based on 1991 COMA recommendations, which in turn were based on their estimations for total dietary energy, including alcohol. However, COMA noted that: “… apart from the essential fatty acids, there is no absolute requirement for fats, sugars or starches, so that it was not possible to derive useful reference figures for them based on a range of requirements. Thus, the panel made pragmatic judgements based on changes from current intakes which would be expected to result in certain changes in physiological and or health outcome ...... and should not be taken to represent ‘ideal’ figures.” As indicated in Table 1, which compares recent reviews of dietary recommendations for public health, the IGD’s GDAs do not adequately take account of current trends in the goals for certain 42 43 44 45 46 47 Schofield WN, Schofield C, James WPT. Basal metabolic rate – review and prediction. Human Nutrition – Clinical Nutrition; 30 (suppl): 1-96. National Audit Office. 2001. Tackling Obesity in England: pp. 1-66. London: National Audit Office. Department of Health. 2004. Health Survey for England 2003, trends. http://www.dh.gov.uk/PublicationsAndStatistics/PublishedSurveys/HealthSurvyForEngland/fs/en FAO/WHO/UNU. 2004. Human Energy Requirements. Report of a Joint FAO/WHO/UNU Expert Consultation. FAO Food and Nutrition Technical Report Series No. 1. Rome: FAO. Energy Requirements Working Group SACN/energy/05/03 See details at http://www.sacn.gov.uk/meetings/workinggroups/energy/2006_12_05.html WHO Child Growth Standards, from the Multicentre Growth Reference Study. World Health Organization, April 2006. http://www.who.int/childgrowth/en/, accessed 15 August 2006. Page 42 of 46 nutrients. For example, the GDA for fat has been set at the very upper end of the range, while that for total carbohydrate is set at the very lower end, and the figures do not reflect the most recent analyses available. The GDAs for fats The GDAs for fats are largely based on the 1991 COMA report and do not adequately take account of current UK macronutrient intakes and international population nutrient goals. Both WHO and the US/Canada have recently published recommendations which give a new range of recommendations. Total fat The GDA for total fat has been set at the upper end of the range. In 1990, UK fat intakes were averaging over 40% of total energy,48 and the COMA total fat target – based on their ‘pragmatic judgement’ – was set at a level of 35% of dietary energy. This was intended as an interim dietary target at a time of high population fat intakes. The earlier report of the National Advisory Committee on Nutrition Education (NACNE)49 acknowledged the importance of current intakes as well as the demands of farming policy, the EU CAP and the influence of the food industry by setting both short and longer term dietary recommendations for the macronutrients. For example, in the case of fat, NACNE recommended a short-term goal of 34% of dietary energy (intakes were then 42%), with a longer term goal, more in line with population public health, of 30%. Recently, in the US and Canada, Acceptable Macronutrient Distribution Ranges (AMDRs) were set in 2002 and a range of 20-35% of dietary energy adopted. The FDA labelling requirements stipulate that the guideline reference amount should be ‘less than’ the equivalent of around 29% total dietary energy. Not only are the GDAs for fats higher than many recommendations, they are also higher than the levels currently being consumed by the majority of adults.50 Although there are reporting problems with the dietary surveys, the GDAs for the proportion of fat in the diet are significantly higher than those reported and open up the alarming possibility that adults may believe they should consume more, rather than less, fat to achieve their ‘guideline’ figure. Total fat UK adult average intake (2001-02) GDA Women 61g 70g Men 86g 95g Saturated fat The upper limit of 10% of total dietary energy for saturated fats in the diet proposed by COMA was, like total fat, a pragmatic recommendation. There is no absolute requirement for saturated fat in the diet and WHO has proposed a lower limit of 0% although such a target would be at present unrealistic in the UK. COMA was informed by the 1990 National Diet and Nutrition Survey which showed that the mean intake of saturated fat in the UK adult diet was 36.5g per day (over 16% of dietary energy). A target of 10% saturated fat was set, to achieve a decrease in population serum cholesterol, to reduce the 48 49 50 Gregory J, Foster K, Tyler H, Wiseman M. 1990. The National Diet and Nutrition Survey of British Adults. London HMSO. NACNE. 1983. A Discussion Paper on Proposals for Nutritional Guidelines for Health Education in Britain. London: Health Education Council. Henderson L, Gregory J, Irving K, Swan G. 2003. The National Diet and Nutrition Survey: Adults Aged 19 to 64 Years. Vol 2: Energy, Protein, Carbohydrate, Fat and Alcohol Intake. London: The Stationery Office. Page 43 of 46 risk of coronary heart disease. Intakes of saturated fat appear to have fallen (the 2001-02 dietary survey reported that saturated fat formed 13.3% of an adult’s average dietary energy). However, a significant proportion of the population continues to have raised levels of blood cholesterol levels. A lower population target may therefore be beneficial for public health policy.51, 52 The GDAs for carbohydrate Total carbohydrate COMA recommended that total carbohydrate should contribute 47% of total dietary energy, of which non-milk extrinsic (NME) sugars should not exceed 10% of total dietary energy. Goals for carbohydrate are obtained by subtracting specified fat and protein DRV values from total energy intake or basing recommendations on levels in countries with a low incidence of chronic disease, as there are few criteria for specifying how much starch is required for health. The IGD’s GDAs for carbohydrate are based on the original COMA figures. In comparison with the recommendations of subsequent reviews (see Table 1), the COMA figures appear low for total carbohydrates. All other reviews recommend higher levels. Sugars At the time of the 1991 COMA report, dietary surveys indicated that total sugars constituted 18% of dietary energy, with NME sugars providing 8-13% of total dietary energy. The more recent dietary survey53 indicates that mean intakes of sugar are over 20% of total dietary energy, with NME sugar intake of 12% (women) and 14% (men) of dietary energy, suggesting a relative rise in the consumption of sugar-rich foods. This would be concordant with industry data showing significant per capita increases in the volume of soft drinks and confectionery purchased in the UK over the last two decades. NME sugars provided over 16% of dietary energy for school-age children,54 and consumption of NME sugar for some sub-groups is particularly high: for example 10% of young men (aged 19-34) obtain over 25% of their total dietary energy from NME sugar. COMA recommended that NME sugar should not exceed 60g per day or 10% of dietary energy. Under the GDA system, the value for NME sugar is 65g for men and 50g for women, but these values are not used for labelling purposes; instead a total sugar GDA is used (120g for men and 90g for women) with the justification that as there is no agreed definition for NME sugars and these are technically difficult to measure, this makes it difficult to derive an objective method for contentmonitoring purposes. It should be noted that NME sugar levels were calculated for the National Diet and Nutrition Surveys and are presumably known to manufacturers who are in control of their recipe formulations. The problem of communicating an effective signal to consumers about daily intake recommendations has remained unresolved and urgently needs attention. In the US review of Dietary Reference 51 52 53 54 Gregory J, Foster K, Tyler H, Wiseman M. 1990. The National Diet and Nutrition Survey of British Adults. London HMSO. Henderson L, Gregory J, Irving K, Swan G. 2003. The National Diet and Nutrition Survey: Adults Aged 19 to 64 Years. Vol 2: Energy, Protein, Carbohydrate, Fat and Alcohol Intake. London: The Stationery Office. Henderson L, Gregory J, Irving K, Swan G. 2003. The National Diet and Nutrition Survey: Adults Aged 19 to 64 Years. Vol 2: Energy, Protein, Carbohydrate, Fat and Alcohol Intake. London: The Stationery Office. Gregory J, Lowe S, Bates CJ, Prentice A, et al. 2000. National Diet and Nutrition Survey: Young People Aged 4-18 Years. Volume 1: Report of the Diet and Nutrition Survey. London: The Stationery Office. Page 44 of 46 Intakes55 (DRIs) the responsible committee acknowledged the problem and set an upper absolute limit of 25% dietary energy from sugars above which the balance of other nutrients was likely to be threatened, but added that this level should not be used for labelling purposes as it sent the wrong type of message about curbing intakes. They failed to reach agreement, noting only that “the committee is unable to recommend an approach for developing a reference value for sugars or added sugars for nutrition labelling based on the DRIs”. The GDAs for fibre Definitions of dietary fibre have been controversial and recommended amounts have been subject to change, potentially confusing consumers. Dietary fibre was originally ‘plant polysaccharides and lignin resistant to hydrolysis by the digestive enzymes of man’.56 However, this was a physiological definition and the dietary fibre fraction was difficult to analyse in foodstuffs. The COMA panel used an enzymatic, chemical method (Englyst)57 which precisely measures fibre defined as non-starch polysaccharide (NSP). The COMA recommendation of 18g per day NSP for men and women was set at a level associated with a reduced risk of colon cancer. In the US and in Europe the Association of Official Analytical Chemists (AOAC) procedure is used to quantify the dietary fibre content of foods.58 This method gives a fibre value about 30% higher than NSP because it includes a variety of unspecified polymers such as resistant, retrograded and synthetic starch. The argument is that all carbohydrates, including these latter components, reach the large intestine and have ‘fibre-like’ properties (increased stool weight and butyrate production), are beneficial to health and therefore should be included in the dietary fibre classification. However, the scientific evidence to support this view remains under question,59 and there is some evidence from animal studies of enhanced tumour formation with retrograded starches.60 There are thus criticisms of the use of the AOAC method in food labelling since AOAC ‘fibre’ values may represent largely novel material, such as retrograded starch, formed as the result of food processing, or synthetic starch derived from high amylose corn starch which has been added to foods as a thickener.61 Consumers will be unable to distinguish between foods that are naturally fibre-rich (and so recommended for healthy diets by COMA62) and those that have little NSP but have high levels of added processed starches. 55 56 57 58 59 60 61 62 Food and Nutrition Board of the Institute of Medicine. 2003. Dietary Reference Intakes: Guiding Principles for Nutrition Labeling and Fortification. Washington: The National Academies Press. http://www.nap.edu/catalog/10872.html, accessed 12 August 2006. Trowell H, Burkitt D, Heaton K (eds). 1985. Dietary Fibre, Fibre-depleted Foods and Disease. London: Academic Press. Englyst HN, Cummings JH. 1988. Improved method for measurement of dietary fibre as non-starch polysaccharides in plant foods. Journal of the Association of Official Analytical Chemists; 71: 808814. Association of Official Analytical Chemists (AOAC). 1990. Official Methods of Analysis. 15th edition, Vol II, section 985.29. Arlington, Virginia: AOAC. pp1105-1106. Johnson IT, Southgate DAT. 1993. Dietary Fibre and Related Substances. London: Chapman and Hall. Burn J et al. 1996. Intestinal tumours in the Apc mouse: aspirin not protective and resistant starch increases small bowel tumours. European Journal of Human Genetics; 4 (suppl 1): 13. See also: Young GP et al. 1996. Wheat bran suppresses potato starch-potentiated colorectal tumorigenesis at the aberrant crypt stage in a rat model. Gastroenterology; 110: 508-514. Englyst N, Hudson GJ. In: Garrow JS, James WPT and Ralph A (eds). 2000. Human Nutrition and Dietetics. 10th edition. Churchill Livingstone. Committee on Medical Aspects of Food Policy. 1984. Diet and Cardiovascular Disease. Report on Health and Social Subjects 28. London: Department of Health. Page 45 of 46 In conformity with European food legislation requirements, the FSA recommended that the AOAC method be used to analyse fibre for labelling proposes, but has continued to use NSP for dietary surveys such as the NDNS. The FSA is now reviewing this recommendation and considering a number of alternative definitions of dietary fibre. The GDAs for salt The salt GDA (6g per day for adults) is derived from a recent SACN report.63 However, the SACN report was concerned with reducing UK salt intakes from a high level of 9g per day and the recommendation of 6g per day (like that for dietary fat in 1991) was set as a pragmatic, interim target. (“… the targets for adults, infants and children at these levels do not represent optimal or ideal levels of salt intake but represent achievable goals”.)64 The Reference Nutrient Intake for sodium is 1.6g per day which gives a salt value of 4g per day and evidence suggests that a figure even lower than this is realistic average for a healthy population. The use of 6g as a GDA for salt may be understood by the consumer as a target to be attained, rather than a maximum recommended amount. The more complex message – that salt intakes are far too high, that this is an avoidable cause of hypertension, therefore salt intakes need to be reduced, with an initial target of 6g but a longer term target which is lower still – may help the public to be better informed and better able to make healthier choices. Conclusions The use of the IGD’s GDAs is not well advised, as many of the nutritional benchmarks used are outof-date and in need of review. In particular, the energy EAR is under review, and any changes to the DRVs for food energy will have implications for consumer guidance in relation to energy and the related guidance for levels of many of the macronutrients. The GDAs for fibre and sugar are particularly controversial. The use of a total sugar rather than a GDA based on a measure of NME sugars will be misleading, as the sugar content of processed foods containing only NME sugar will be benchmarked against the total sugar GDA and might thereby mislead consumers into believing they should consume the product. There is also a perverse incentive for manufacturers to reduce total sugars by reducing the fruit and vegetable content of ready meals.65 There is no dietary recommendation for total sugars in the UK, although the FSA is considering 60g per day as a labelling benchmark. For fibre, the AOAC definition may result in many processed foods appearing to have a high fibre content despite having low levels of non-starch polysaccharides. Population dietary recommendations must be subject to wide scrutiny and should be arrived at as part of a broad, consultative and scientific review of UK dietary targets, taking into account: current intake and activity levels and the overall macronutrient balance of the diet; widely accepted definitions for sugar and fibre; as well as the most current evidence about relationships between nutrient intakes and the risks of chronic disease. The IGD’s GDAs should not be taken as UK dietary guidelines and a proper review needs to be conducted by an authoritative, independent body such as the Scientific Advisory Committee on Nutrition. 63 64 65 Scientific Advisory Committee on Nutrition. 2003. Salt and Health. London: The Stationery Office. Paper FSA 03/06/02. Agenda Item 4, 12 June 2003 – report to the FSA Board on the Scientific Advisory Committee on Nutrition: Report on Salt and Health. http://www.food.gov.uk/multimedia/pdfs/fsa03602.pdf MRC response to the FSA signpost labelling consultation, 2006. Page 46 of 46