Front-of-package food labels: A narrative review

Appetite 144 (2020) 104485

Contents lists available at ScienceDirect

Appetite journal homepage: www.elsevier.com/locate/appet

Front-of-package food labels: A narrative review

T

Norman J. Temple Centre for Science, Athabasca University, Alberta, T9S 3A3, Canada

A R T I C LE I N FO

A B S T R A C T

Keywords: Food labels Front-of-package labels Food corporations Warning labels Traffic lights labels

Front-of-package (FOP) labels may help shoppers make healthier food choices. The objectives of this review are, first, to establish the effectiveness of different FOP labels at enabling shoppers to identify which foods are healthy and which are not healthy, and, second, to assess whether different FOP labels induce shoppers to buy healthier foods. Some labels are nutrient-specific, such as Multiple Traffic Lights (MTL) and Guideline Daily Amounts (GDA). These labels state the content per serving of energy and of several substances, most commonly saturated fat, sugar, and sodium (or salt). Warning labels are another type of nutrient-specific FOP label (e.g., for food high in added sugar). Summary labels, such as Nutri-Score and labels with stars, translate the components of the food into a single value that indicates how healthy it is. Studies on FOP labels lack consistency. The majority of such studies indicate that they help shoppers to distinguish between healthy and less healthy foods. The designs that appear to be most successful in this regard are MTL, warning labels, and Nutri-Score. Labels based on GDA or that included stars were much less successful. Many studies using a simulated shopping situation reported that shoppers exposed to FOP labels had an increased intent to purchase healthier foods. Warning labels were the most consistently successful FOP design followed by MTL, Nutri-Score, and labels that included stars, while GDA failed in almost every study. Very few studies have been carried out in real-world supermarkets; the findings indicate that FOP labels or shelf labels may achieve a small degree of success (< 2.0%) at persuading shoppers to buy healthier foods. Those advocating for effective FOP labels must resist opposition from food corporations.

1. Introduction Food labels are of much importance as they assist millions of people in making healthier food choices. This was recently demonstrated in a systematic review of various types of food labels, including front-ofpackage (FOP) labels, back-of-package (BOP) labels, labels on restaurant menus, and labels in grocery stores (Shangguan et al., 2019). The findings strongly indicate the food labels do indeed lead to a significant improvement in the healthiness of the diets selected by consumers. For many years the only labels in common use were BOP labels. These are usually in the form of a small chart with numeric information. Evidence indicates that most consumers have a poor ability to accurately interpret these labels (Cowburn; Stockley, 2005). This has led to ongoing efforts to develop designs for BOP labels that are more user friendly (Temple & Fraser, 2014). In recent years many countries have adopted much simpler FOP labels. Several different designs are now in use around the world. Some designs are nutrient-specific and display information on individual substances, most commonly saturated fat, sugar, and sodium (or salt). A recent development is warning labels which are only present on a food package if it has a relatively high content of an undesirable substance

such as sugar or salt. An alternative design strategy is a summary label. Here the healthiness of the food is indicated by a single value and expressed in a very simple manner such as stars or a tick. Many studies have investigated the effectiveness of these different FOP designs but there is great controversy regarding the ideal design of FOP labels (Temple & Fraser, 2014). Most shoppers browsing the aisles at a supermarket will typically spend no more than a few seconds examining the labels on food items before making a selection (Sanjari, Jahn, & Boztug, 2017). As a result it is the simple FOP label that plays the dominant role for most shoppers rather than the BOP label which is much more detailed. It is therefore important that FOP labels are well designed so that they are maximally effective at enabling shoppers to select a healthier diet. Much research in recent years has investigated how FOP labels can be better designed. This paper reviews research studies that have been carried out on different types of FOP labels. The paper examines the effectiveness of different FOP labels at enabling shoppers to accurately identify which foods are healthy and which are not healthy. The paper also examines the impact of different FOP labels on the likelihood that shoppers will buy healthier foods. Another key issue that is discussed is opposition from the food industry to FOP labels that may negatively affect sales.

E-mail addresses: [email protected], [email protected]. https://doi.org/10.1016/j.appet.2019.104485 Received 22 March 2019; Received in revised form 29 September 2019; Accepted 7 October 2019 Available online 09 October 2019 0195-6663/ © 2019 Elsevier Ltd. All rights reserved.

Appetite 144 (2020) 104485

N.J. Temple

This can take the form of lobbying activities to politicians (Hickman, 2010; Julia & Hercberg, 2016).

2. Design of FOP labels This section briefly describes the most common designs of FOP labels. They fall into two groups. One group is nutrient-specific. The label highlights the content of energy and of several substances, most commonly saturated fat, sugar, and sodium (or salt). Total fat is often included. Amounts are stated per serving. Summary labels, by contrast, use an algorithm so as to translate the components of the food into a single value that denotes how healthy or unhealthy it is. A major difference between the two designs is that while nutrient-specific labels focus on substances that are harmful when consumed in excess, summary labels cover a spectrum from the most healthy foods to the most unhealthy foods.

Fig. 2. An example of a Multiple Traffic Lights food label.

2.1. Nutrient-specific labels

Fig. 3. Examples from Chile of warning labels for foods high in sodium and sugar. At the bottom of each label it states Ministry of Health.

One such system is the Guideline Daily Amounts (GDA) label (Food and Drink Federation, 2019). The label states the amount of these substances per serving as well as of energy (see Fig. 1). The amounts are also stated as a percentage of an adult's reference intake. The label is used in Mexico. The GDA, or slight variations of it, are also known as Facts Up Front (anon, nd), Reference Intake label, and Daily Intake Guide. Another nutrient-specific label design is Traffic Lights or Multiple Traffic Lights (MTL; the abbreviation used here) (Food Standards Agency, 2016). The design is quite similar to GDA except that whereas GDA is monochrome, MTL incorporates traffic lights that indicate if the food has a high (red), medium (orange), or low (green) content of saturated fat, sugar, and sodium (and sometimes total fat). The system is used in several countries, including the UK, where it was developed. An example of a MTL label is shown in Fig. 2. A recently developed FOP label that is attracting much interest is health warnings (see Fig. 3). The label is usually a simple text statement such as “high in sodium”. As food components now fall into only two categories (warning or no warning) rather than three, health warnings are therefore less informative than MTL. Chile was the first country to introduce health warnings and in 2016 mandated their use for foods high in sugar, fats, salt, or calories (Carreño, 2015). This type of FOP label has been adopted by several other countries (Reyes et al., 2019). Some countries include saturated fat. Most interest has been shown for mandating the use of warning labels on containers of sugar-sweetened beverages. Brazil, Uruguay, Canada, and several US states and cities have taken steps to implement this public health strategy.

2.2. Summary labels Several designs for FOP labels have been developed that employ a single summary value. NuVal is one such system. This calculates the overall nutritional score for the food on a scale of 1–100 (Katz, Njike, Rhee, Reingold, & Ayoob, 2010; The NuVal, 2017). In order to achieve a high score the food must be below a maximum content of fat, saturated fat, and sodium, and contain a minimum amount of protein, dietary fibre, and of various vitamins and minerals. Nutri-Score, also known as 5-CNL (5-Colour Nutrition Label), presents its estimation as a letter from A (healthiest) to E (least healthy). As shown in Fig. 4 the label shows all five letters, each in a different colour, but the one indicating the value of the food is expanded. The system uses a nutrient profiling system based on the one developed by the Food Standards Agency in the UK (Julia et al., 2014). Components of the food that are included in the FSA score include saturated fat, sugar, sodium and other minerals, vitamins, protein, dietary fibre, n-3 fatty acids, fruits and vegetables, and whole grains. Nutri-Score was developed in France where it is used on many food products. It has also been adopted by Belgium and Spain. Another system is Guiding Stars (Fischer, 2011; Guiding Stars, 2019). The FOP label denotes the health value of the food with zero, 1, 2, or 3 stars. Shoppers can deduce a zero star rating only from the omission of star information. Components of the food that are included in the score include energy, saturated fat, sugar, sodium, fruits and vegetables, protein, and fibre. Other food components may also be included. Guiding Stars originated from a system proposed by the Institute of Medicine (now the National Academy of Medicine) in 2011 (Institute of Medicine, 2012). Use of this FOP label is mainly confined to the USA, though it was also adopted by a chain of supermarkets in Canada. The Health Star Rating system (HSR) is used in Australia and New Zealand (Australian Government Department of Health, 2016). The label displays from half a star to five stars. HSR can appear on labels in

Fig. 1. An example of a Guideline Daily Amounts (GDA) food label.

Fig. 4. Example of a Nutri-Score food label. 2

Appetite 144 (2020) 104485

N.J. Temple

In the current review a search was made for studies published after 2011. This date was chosen so as to include only studies published after the time period covered by the above two reviews. Papers were included if they studied the effect of FOP labels on the ability of adult shoppers to identify which foods are healthiest or on food choices made by shoppers. Only experimental studies were included while observational studies were excluded. A search was made in February 2019 as follows. PubMed was searched using the following key words or phrases: “front-of-package” OR “front-of-pack” AND “label” OR “labels” and was restricted to English. In addition, two other databases (Scopus and CINAHL) were searched in August 2019. Online searches were complemented by searching reference lists. Additional papers were identified by carrying out searches of the table of contents pages of many journals. No exclusions were made based on country. Appendix Table 1 presents the findings from 17 studies published after 2011 that investigated the effectiveness of different FOP labels at enabling shoppers to identify which foods are healthiest. In most studies the subjects were adults and were shown images of the food package on a computer. However, in one study the labels were on bottles or containers of beverages (Acton, Vanderlee, & Hammond, 2018a), in two studies the images were printed (Aschemann-Witzel et al., 2013; Roseman, Joung, & Littlejohn, 2018), while in another study the subjects included both adults and children (Talati et al., 2017a). There are major differences between the studies, both in the methodology used and the way the results are presented. For these reasons it is impossible to make an exact quantitative comparison of the results of the different studies. Instead, terms such as “successful”, “moderate success”, and “weak success” are used. The dominant finding from these studies is their lack of consistency. Twelve of the 17 studies found clear evidence that FOP labels can improve the ability of shoppers to identify which foods are healthier (i.e., at least one FOP label generated a clearly positive result) (Arrúa et al., 2017; Ducrot et al., 2015; Egnell, Talati, Hercberg, Pettigrew, & Julia, 2018; Goodman, Vanderlee, Acton, Mahamad, & Hammond, 2018; Gorski et al., 2018; Khandpur et al., 2019, 2018; Maubach, Hoek, & Mather, 2014; Roberto et al., 2012b; Roseman et al., 2018; Talati et al., 2017a; Watson et al., 2014). Of the other studies one produced inconsistent results (Acton et al., 2018a) while four failed to observe an improvement (Aschemann-Witzel et al., 2013; Edge, Toner, Kapsak, & Geiger, 2014; Hodgkins et al., 2015; Roberto et al., 2012a). GDA worked well four times (Ducrot et al., 2015; Gorski et al., 2018; Roseman et al., 2018; Watson et al., 2014) but did poorly eight times (Arrúa et al., 2017; Aschemann-Witzel et al., 2013; Edge et al., 2014; Egnell et al., 2018; Hodgkins et al., 2015; Maubach et al., 2014; Roberto et al., 2012a; Talati et al., 2017a). MTL performed better than GDA; it worked well in eight studies (Arrúa et al., 2017; Ducrot et al., 2015; Egnell et al., 2018; Gorski et al., 2018; Khandpur et al., 2018; Maubach et al., 2014; Roberto et al., 2012b; Watson et al., 2014), showed a small improvement in one study (Talati et al., 2017a), but three studies reported that it was no better than no label (AschemannWitzel et al., 2013; Hodgkins et al., 2015; Roberto et al., 2012a). The higher scores for MTL than with GDA likely reflects the fact that although they provide the same numerical information, MTL has colours and text to make it clear to shoppers whether the quantities are high, medium, or low, whereas GDA is monochrome and has no text. A simplified traffic lights label was tested in two studies. It achieved limited success in one study (Acton et al., 2018a) but failed to work in a second (Roberto et al., 2012b). The FOP labels that have stars (most commonly Guiding Stars and HSR) worked well in two studies (Gorski et al., 2018; Talati et al., 2017a), was mediocre three times (Acton et al., 2018a; Egnell et al., 2018; Watson et al., 2014), and failed to work once (Maubach et al., 2014). A tick label worked well in two studies (Roberto et al., 2012b; Roseman et al., 2018), was mediocre once (Ducrot et al., 2015), and failed to work once (Hodgkins et al., 2015). Nutri-Score was tested

Fig. 5. Examples of Health Star Rating food labels.

different ways. Some foods only carry the overall HSR of the product (Fig. 5, image on left). Other foods include information such as the quantity of energy, saturated fat, sugar, and sodium (Fig. 5, image on right). Extra substances, such as calcium and fibre, can be added if the amount is high or low. The simplest summary FOP label displays only a symbol such as a tick. The symbol is present on the food package if the food meets certain criteria that indicate that the food is healthy. Several different versions of this symbol have been developed. One version was developed in Australia and New Zealand. Another version is Choices (Choices International Foundation, 2019). It was developed in the Netherlands in 2006 and has been adopted in many other countries. Another variation of this design is the Keyhole, also known as the Nordic Keyhole (Swedish National Food Agency, 2019). As the name indicates the symbol depicts a keyhole. It was originally developed in Sweden in 1989 and is now used in the Nordic countries (Sweden, Denmark, and Norway). The USA developed its own version of a tick symbol, known as Smart Choices (Lupton et al., 2010). It was used for a short time around 2009. The tick symbol has also been adopted for heart-healthy foods. An Australian version of this is shown in Fig. 6.

3. Effectiveness of different FOP labels The focus of this paper is the effectiveness of different types of FOP labels. This review excludes discussion of demographic factors that may affect findings, such as gender, age, and socio-economic status. The paper also excludes findings regarding which FOP label is most preferred by shoppers. Studies of the effect of nutrition or health claims were excluded. The rationale for this narrow focus on the effectiveness of FOP labels is as follows. FOP labels have their own distinct design, intended use, and impact on consumers. By excluding discussion of other related labels, such as BOP labels and nutrition or health claims, the paper is much shorter and easier to read. Two systematic reviews were carried out that included papers published up to late 2010/early 2011 (Hawley et al., 2013; Hersey, Wohlgenant, Arsenault, Kosa, & Muth, 2013). Those reviews included both FOP labels and labels on the shelves in supermarkets that were adjacent to food. In general, MTL labels were found to be more effective than either the GDA label or summary systems at helping shoppers to identify which foods are healthier. However, a notable finding was the lack of consistency in the results from different studies.

Fig. 6. Heart foundation tick. © state of new south wales through the NSW food authority www.foodauthority.nsw.gov.au. 3

Appetite 144 (2020) 104485

N.J. Temple

2018), moderately successful in one (Appendix Table 2: Acton & Hammond, 2018b), and weakly successful in one (Appendix Table 3: Ang, Agrawal, & Finkelstein, 2019). Taking the findings from Appendix Tables 2 and 3 together, nine of the 18 studies found clear evidence that FOP labels can lead to shoppers displaying an increased intent to purchase healthier foods (i.e., at least one FOP label generated a clearly positive result) (Billich et al., 2018; Egnell et al., 2019; Goodman et al., 2013; Khandpur et al., 2018; Machín et al., 2017; Mantzari et al., 2018; Maubach et al., 2014; Roberto et al., 2016; Talati et al., 2017b). Two other studies saw a modest increased intent (Acton & Hammond, 2018b; Ducrot et al., 2016), one saw a weak positive response (Ang et al., 2019), while six failed to observe an improvement (Aschemann-Witzel et al., 2013; Borgmeier & Westenhoefer, 2009; Gorski et al., 2018; Julia et al., 2016; Roberto et al., 2012b, 2012a). The combined results for Appendix Tables 2 and 3 do not indicate whether nutrient-specific labels (MTL, GDA, and warning labels) outperform summary labels (Nutri-Score and those with stars or a tick). A few studies have been carried out in real-world supermarkets. In one such study a chain of supermarkets in the UK introduced MTL labels on some products (Sacks, Rayner, & Swinburn, 2009). An analysis of sales of “ready meals” and sandwiches revealed no shift to healthier varieties over the course of four weeks. A similar study was carried out in an online grocery store in Australia (Sacks, Tikellis, Millar, & Swinburn, 2011). MTL labels were added to a range of products. No evidence was seen that the FOP labels led to an increase in sales of healthier foods over ten weeks. Two studies were carried out in supermarket chains. In each case Guiding Stars labels were attached to the shelves adjacent to foods. The first study was carried out in Canada and included a wide variety of foods (Hobin et al., 2017). Over the course of six months there was an increase in mean star rating of 1.4% (comparing supermarkets that used the labeling system versus those that did not). The other study was in the USA and focused on the sale of readyto-eat cereals (Rahkovsky, Lin, Lin, & Lee, 2013). Over the course of seven months sales of cereals with no stars fell by 2.6% (comparing stores carrying the logo with control stores) while sales of cereals with one, two, and three stars increased by 1.15%, 0.89%, and 0.54%, respectively. This indicates that Guiding Stars caused a shift to healthier varieties by about 1.5–2%. These studies carried out in real-world supermarkets generate results that have much more credibility than do studies in a simulated shopping situation. This is because, first, the shoppers are not part of an experiment (which might influence their behavior), and, second, the shoppers have several weeks or months to adapt to the FOP labels rather than a single experiment that has a duration of less than 1 h. Taken as a whole, these studies indicate that FOP labels or shelf labels may achieve a small degree of success in persuading shoppers to buy healthier foods at the expense of less healthy foods (maximum of 2.0% shift in healthiness of food purchases). However, these findings must be viewed with much caution as they are based on only four studies.

twice (Ducrot et al., 2015; Egnell et al., 2018) and NuVal once (Gorski et al., 2018); they were the best performing labels each time. The design of Nutri-Score (Egnell et al., 2018) is comparable with MTL in that the key information is summarized using colours and is easy to understand. Warning labels performed well in four studies (Arrúa et al., 2017; Goodman et al., 2018; Khandpur et al., 2019, 2018) and had limited success once (Egnell et al., 2018). The findings can be summarized as follows:

• There is a lack of consistency. • FOP labels are generally helpful at enabling shoppers to identify which foods are more healthy and which are less healthy. • FOP labels that are easy for shoppers to understand, namely MTL, •

warning labels, and Nutri-Score, are generally most effective. The presence of colour in MTL and Nutri-Score labels is probably an additional reason for their effectiveness (Lohse, 1997). The results do not permit a verdict regarding whether nutrientspecific labels (MTL, GDA, and warning labels) outperform summary labels (Nutri-Score and those with stars or a tick).

Appendix Table 2 presents the findings from fourteen studies that investigated whether different FOP labels affect the intention of shoppers to buy healthier foods. The subjects were adults except for one study that tested both adults and children (Talati et al., 2017b). Each of the studies tested subjects in a simulated shopping situation, mostly on a computer. The investigators in two studies attempted to create a more realistic situation by using an imitation marketplace (Acton & Hammond, 2018b) or a supermarket (Julia et al., 2016). There were many other methodological differences between the studies. For example, in some studies healthier foods were more expensive than less healthy foods. As was the case in Appendix Table 1, the dominant finding from the studies shown in Appendix Table 2 is their lack of consistency. MTL manifested limited success: out of ten studies that tested the FOP label, it clearly worked in only three (Khandpur et al., 2018; Machín et al., 2017; Maubach et al., 2014), was weakly effective in two studies (Ducrot et al., 2016; Talati et al., 2017b), but failed to work in five others (Aschemann-Witzel et al., 2013; Borgmeier & Westenhoefer, 2009; Gorski et al., 2018; Roberto et al., 2012b, 2012a). One study found that a single traffic light was effective for inducing shoppers to prefer low-sodium crackers over a brand that is high in sodium (Goodman, Hammond, Hanningm, & Sheeshka, 2013). In another study a single traffic light failed to achieve a positive change (Roberto et al., 2012b). GDA produced a weak positive effect in one study (Maubach et al., 2014) but failed to work in seven other studies (AschemannWitzel et al., 2013; Borgmeier & Westenhoefer, 2009; Ducrot et al., 2016; Egnell et al., 2019; Gorski et al., 2018; Roberto et al., 2012a; Talati et al., 2017b). The FOP labels that have stars (mainly Guiding Stars and HSR) worked well in two studies (Maubach et al., 2014; Talati et al., 2017b) but failed to work in two others (Acton & Hammond, 2018b; Gorski et al., 2018). A tick label produced a weak positive effect in one study (Ducrot et al., 2016) but failed to work in two others (Borgmeier & Westenhoefer, 2009; Roberto et al., 2012b). Nutri-Score was tested in three studies; it was successful in one (Egnell et al., 2019), was moderately successful in a second (Ducrot et al., 2016), but was not successful in a third (Julia et al., 2016). NuVal failed in the one study that tested it (Gorski et al., 2018). Some of the studies also tested the effectiveness of warning labels at dissuading shoppers from buying unhealthy foods. This was tested in three studies shown in Appendix Table 2. It was also tested in the four studies shown in Appendix Table 3. The design of the studies in Table 3 is similar to those shown in Appendix Table 2. The combined results reveal that warning labels were clearly successful in five studies (Appendix Table 2: Khandpur et al., 2018; Machín et al., 2017; Appendix Table 3: Billich et al., 2018; Roberto, Wong, Musicus, & Hammond, 2016; Mantzari, Vasiljevic, Turney, Pilling, & Marteau,

4. Issues in the design of FOP labels Should summary labels include other information in addition to the summary score? The cardinal rule should be that information is only added to FOP labels where it informs consumers more than it confuses. The only food components where there is a case for their inclusion are those where there is solid evidence demonstrating that the substance (in the quantity present in many foods) is related to risk of common health disorders. One prominent example is sodium (or salt) (Cook et al., 2007; He & MacGregor, 2009). This information is especially important for persons with elevated blood pressure. Likewise, sugar should be added to summary labels for foods where sugar is the dominant source of calories, especially sugar-sweetened beverages and candies (de Koning, Malik, Rimm, Willett, & Hu, 2011; Narain, Kwok, & Mamas, 2016; Te Morenga, Howatson, Jones, & Mann, 2014). The most 4

Appetite 144 (2020) 104485

N.J. Temple

(Magnusson, 2010). But rather than fighting and risk losing, they took a pragmatic approach: in 2006 they introduced their own voluntary FOP system. This allowed the food industry to be in control. How effective was this scheme? According to findings from a survey carried out in 2012, food manufacturers often failed to state the quantity of saturated fat and sugar in energy-dense, nutrient-poor snacks (Carter, Mills, Lloyd, & Phan, 2013). In other words, the food industry formulated a code of practice which was then ignored when clear labels might negatively affect sales. In 2019 the World Cancer Research Fund International published a report on FOP labels (World Cancer Research Fund International, 2019). The report includes a discussion of common tactics used by industry in an effort to thwart attempts by governments to introduce FOP labels. This evidence is consistent with much other evidence in demonstrating that industries that make profits from the sale of unhealthy products cannot be trusted to engage in self-regulation. This extends to lobbying governments. This has been well documented in the USA where food corporations have had a major influence on all aspects of government policy (Nestle, 2007). These issues were analyzed by an international group of health advocates: The Lancet NCD (non-communicable diseases) Action Group (Moodie et al., 2013). Their focus was on the sale and promotion of tobacco, alcohol, and ultra-processed food and drink. They concluded as follows: (1) industries that produce unhealthy products should play no role in the development of national or international policies related to non-communicable diseases, and (2) there is no evidence that either self-regulation by industry or public–private partnerships is of any real value for improving public health.

effective way to present sodium and sugar in FOP labels is probably as a warning label. There is little point in including sugar in summary FOP labels for those sugar-rich foods where sugar is not the dominant source of calories, such as many brands of breakfast cereals, cakes, and cookies, as the information is already captured by the summary score. Similarly, such substances as potassium, vitamin C, saturated fat, and dietary fibre should be excluded as they are best integrated with the summary score. Their inclusion as separate substances would generally add more noise than signal. There are two reasons for this. One reason is that most of these substances are closely associated with healthy foods. In particular, fruit, vegetables, and whole grains are generally good sources of potassium, vitamin C, folate, and dietary fibre. Conversely, unhealthy foods are generally poor sources of these substances but often have a high content of sugar and salt. The other reason is that recent developments in nutrition place a strong emphasis on whether foods are healthy or unhealthy (based on their association with disease risk) while the content of individual substances is considered much less informative (Mozaffarian, 2016). There is a case for excluding the energy value of foods on all FOP labels as its inclusion implies that foods with a high energy content are unhealthy. In reality, many healthy foods have a high energy content. For example, the presence of oils in food increases their energy density. However, much evidence indicates that vegetable oils rich in n-6 fatty acids are protective against CHD (Maki, Eren, Cassens, Dicklin, & Davidson, 2018). Similarly, the PREDIMED study demonstrated that the addition of extra-virgin olive oil to the diet is strongly protective against cardiovascular disease (Temple et al., 2019). Most types of nuts have a much higher energy density than cornflakes. However, nuts have a strong protective association with risk of CHD and all-cause mortality (Aune et al., 2016) while cornflakes are highly refined and have little nutritional value. More research is required to determine whether the energy value of foods should be stated on FOP labels. A potential benefit of improved food labels is that food manufacturers may then reformulate some products so as to improve their nutritional quality. Studies carried out in the Netherlands and New Zealand have provided some evidence for this (Thomson, McLean, Ning, & Mainvil, 2016; Vyth, Steenhuis, Roodenburg, Brug, & Seidell, 2010; Young & Swinburn, 2002).

6. Summary of findings and discussion A limitation of this study is that the literature search was less rigorous than a systematic review. Three databases were used, as well as some additional searching. It is quite possible some relevant studies were missed. However, this is unlikely to materially affect the overall findings. The focus of the review is the effectiveness of different FOP labels. A limitation of the review is that related areas were excluded including discussion of which FOP label is most preferred by shoppers and of demographic factors that may affect findings, such as socioeconomic status. The majority of studies have reported that FOP labels improve the ability of shoppers to distinguish between more healthy and less healthy foods. However, a notable feature of these studies is their lack of consistency. The designs for FOP labels that appear to be most successful are MTL, warning labels, and Nutri-Score. The most likely reason for this is that these designs are fairly easy for shoppers to understand. An additional reason, in the case of MTL and Nutri-Score, is probably the use of colour (Lohse, 1997). Labels based on GDA or that included stars were much less successful. It was not possible to determine whether nutrient-specific labels (MTL, GDA, and warning labels) are more successful than summary labels (Nutri-Score and those with stars). Other studies have investigated whether FOP labels result in shoppers having an increased intent to buy healthier foods. The most commonly used experimental design was a simulated shopping situation on a computer. Many studies with this design reported positive findings. Warning labels were the only FOP design to display success in most studies. By contrast, MTL, Nutri-Score, and labels that included stars had a much lower frequency of success, while GDA failed in almost every study. Very few studies have been carried out in real-world supermarkets; the findings from those studies indicate that FOP labels or shelf labels may achieve a small degree of success at persuading shoppers to switch from less healthy foods to healthier foods (maximum of 2.0% change). The above studies lead to some tentative conclusions. FOP labels, especially MTL, warning labels, and Nutri-Score, assist shoppers in

5. Corporate and political opposition to improved food labels Attempts to implement an effective FOP labeling system run the risk of being met with stiff opposition from food corporations and their lobbyists. This is because such labels as MTL and warning labels give consumers an explicit warning that a food is not healthy. About 60% of the American diet is comprised of ultra-processed food, i.e., a mixture of such ingredients as sugar, refined flour, and added fats, plus liberal amounts of salt (Baraldi, Martinez Steele, Canella, & Monteiro, 2018). The inevitable result of a well-designed FOP labeling system, therefore, is that vast numbers of packaged foods will display prominent warnings signs. With MTL labels that means a red light adjacent to sugar and sodium. Likewise, Nutri-Score labels would inform shoppers that the food is unhealthy. Not surprisingly, food corporations have vigorously opposed attempts to implement FOP labeling systems as the following examples illustrate. In 2010 the parliament of the European Union (EU) debated making it mandatory for MTL food labels to be used across the EU. The food industry responded with a massive lobbying campaign that successfully defeated the proposed law (Anon, 2010; Hickman, 2010). In 2014–2015 the food industry in France strongly resisted attempts to introduce Nutri-Score FOP labels (Julia & Hercberg, 2016). Australia provides another example. Around 2006 the government of that country was openly discussing the implementation of a new system for FOP food labels. Much like the previous examples from France and the EU, the food industry in Australia was strongly opposed to the plan 5

Appetite 144 (2020) 104485

N.J. Temple

FOP labels are resolute. A closely related question is whether FOP labels should be mandatory or voluntary. In many countries that have adopted FOP labels, their use on food packages is voluntary. This allows the food industry to “have its cake and eat it”: they can endorse the use of food labels, use them for large numbers of products, but decline to use them when this might damage sales, as was the case in Australia (Carter et al., 2013; Magnusson, 2010). It follows, therefore, that for FOP labels to achieve their full value, their use must be mandatory. But even where the use of FOP labels is voluntary, they can still be of much value as, first, they direct shoppers to healthier food choices, and second, they may incentivize food manufacturers to make many foods healthier by reformulating them. FOP labels should be viewed in a broader context of public health policies carried out by governments. Such policies have a long history. Prominent examples include the provision of safe drinking water, the removal of lead from gasoline, restrictions on smoking in public places, and bans on advertising of tobacco products. Increasingly, this approach is being directed to the area of nutrition (Temple, 2016).

recognizing which foods are healthier. But there is little hard evidence that this enhanced knowledge has a significant impact on actual shopping behaviour. There is clearly a need for more research into FOP labeling systems. The highest priority should be studies carried out in real-world supermarkets that investigate which FOP label design is most effective at persuading shoppers to buy healthier foods. Warning labels show much promise and should therefore be included in such studies. It must be borne in mind, however, that warning labels advice shoppers what not to buy rather than what foods are healthiest. A roadblock that can appear between the development of improved FOP labels and their translation into government policy that mandate their use is opposition from the food industry (World Cancer Research Fund International, 2019). In some cases this can take the form of total rejection of the proposed policy (as occurred in the EU) (Anon, 2010; Hickman, 2010) and France (Julia & Hercberg, 2016), while in other cases the food industry may insist on implementing its own (flawed) FOP labeling scheme (as occurred in Australia) (Magnusson, 2010). It can be predicted that the degree of opposition from the food industry will be in proportion to the probability that the food label will have a negative impact on sales. For that reason, warning labels are likely to be met with the strongest opposition, MTL labels will face somewhat less opposition, while GDA labels will probably face the least opposition. It is therefore of utmost importance that those advocating for effective

Declaration of competing interest None.

Appendix Table 1 Findings from recent studies that tested the ability of shoppers to identify which foods are healthiest using different FOP labels. Country/countries

Number of subjects

12 countries (mainly Europe 12,015 and North America) Germany, UK, Poland, Turkey 2068 Australia

2058

USA

703

Australia

4357

Germany, Poland

1000

USA New Zealand

1247 768

France

13,600

Brazil

1607

Uruguay

387

USA

161

USA

480

USA, Canada, UK, Australia

11,617

USA Brazil

7363 2419

Canada

686

Findings

Reference

OR of improvement in ability to correctly rank nutritional quality of foods (vs GDA). Nutri-Score, 3.1; MTL, 1.8; HSR, 1.4; warning label, 1.3. All differences are significant. GDA scored slightly higher than no label. Little difference seen between different labels: MTL, GDA, hybrid (MTL combined with GDA), basic label with a tick, basic label (showing amount of energy and substances included in the other labels) HSR, clear improvement; MTL, small improvement; GDA, no improvement (vs no label).

Egnell et al. (2018)

Hodgkins et al. (2015) Talati et al. (2017a) Little difference seen between different labels: MTL, MTL plus protein/fibre, GDA, GDA plus extra nutrients, no Roberto et al. label (2012a) Watson et al. Score out of 9. MTL, 7.5; GDA, 7.5; GDA plus traffic lights colours, 7.6; Guiding Stars, 6.5; no label, 5.6. Guiding Stars was significantly less than first 3 and significantly more than no (2014) label. No consistent differences between different labels: MTL, GDA, no label Aschemann-Witzel et al. (2013) Quiz score. NuVal, 67.4; MTL, 63.2; GDA, 56.9; Guiding Stars, 57.6; no label, 50.6. Overall p < 0.001. Gorski et al. (2018) Rating of healthiness of food on a 10-point scale. Range healthy vs less healthy foods: MTL, 4.5; GDA, 2.3; Stars Maubach et al. (system with 7 stars), 2.1; no label, 2.3. (2014) % correct answers. Nutri-Score, 64.6; MTL, 56.4; GDA, 50.2; tick, 29.4; no label, 14.6. All differences are Ducrot et al. significant. Overall p < 0.0001. (2015) Percentage improvement in ability to correctly rank which foods are healthier (vs no label). Warning label, Khandpur et al. 13.7%; MTL, 6.9%. Overall p < 0.001. (2018) For identifying if a food is unhealthy: warning label was better than MTL or GDA. % correct responses for Arrúa et al. (2017) identifying the most healthy food: warning label, 82; MTL, 83; GDA, 67. With snack foods higher number able to identify healthier food from less healthy food. % correct: GDA, 92.3; Roseman et al. GDA plus extra nutrients, 92.7; tick, 95.1; no label, 12.5. Overall p < 0.001. Little difference seen when the test (2018) food was dairy or cereal. Higher number able to identify healthier food from less healthy food. % correct: MTL, 71.0; MTL plus calories, Roberto et al. 73.3; tick, 72.5; no label, 67.8. This was not seen for single traffic light plus specific nutrients to limit (65.8%). (2012b) Overall p < 0.001. Warning labels (several different designs) improved ability to identify if cereals have a high content of saturated Goodman et al. fat and sugar (vs no label). OR: 1.0–2.9. Most differences are significant. (2018) Little difference between GDA, GDA plus extra nutrients, no label Edge et al. (2014) Warning labels (3 different designs) improved ability to identify least nutritious foods (vs no label). Most Khandpur et al. differences are significant. (2019) HSR and single traffic light improved ability to identify if a beverage is healthy (vs no label) but results are not Acton et al. consistent. Numeric rating of beverage (similar to NuVal) had little effect. (2018a)

GDA: Guideline Daily Amounts. HSR: Health Star Rating system. MTL: Multiple Traffic Lights. OR: odds ratio. In cases where Reference Intake labels, Facts Up Front, and Daily Intake Guide labels were studied, this is stated as GDA as these labels are virtually the same.

Table 2 6

Appetite 144 (2020) 104485

N.J. Temple

Findings from recent studies that tested whether different FOP labels increase the intention of shoppers to buy healthier foods (simulated shopping). Country/ countries

Number of Experimental design subjects

Findings

Reference

USA

703

Foods selected on a computer

Germany, Poland

1000

Images shown on coloured printouts

Little difference seen between different labels: MTL, MTL plus protein/fibre, GDA, GDA plus extra nutrients, no label No consistent differences between different labels: MTL, GDA, no label

USA

1247

Foods selected on a computer

Little difference seen between different labels: NuVal, MTL, GDA, Guiding Stars, no label

New Zealand 768

Foods selected on a computer

MTL scored highest, then Stars (system with 7 stars), then GDA, then no label

France

901

No overall difference between Nutri-Score and no label

France

11,980

Germany

420

Subjects shopped at an imitation supermarket Subjects shopped at a virtual supermarket Images shown on coloured printouts

Roberto et al. (2012a) AschemannWitzel et al. (2013) Gorski et al. (2018) Maubach et al. (2014) Julia et al. (2016)

Australia

2069

Foods selected on a computer

Brazil

1607

Foods selected on a computer

Canada

430

Subjects selected box of crackers with high or low sodium content

USA

480

Foods selected on a computer

France

1866

Foods selected on a computer

Uruguay

1182

Foods selected on a computer

Canada

686

Imitation marketplace

Score for healthiness of diet (low score indicates a healthier diet). Nutri-Score, 8.7, MTL, 9.0; Ducrot et al. tick, 9.0; GDA, 9.2; no label, 9.3. Overall p < 0.0001. (2016) Little difference seen between different labels: MTL, GDA, tick, no label Borgmeier and Westenhoefer (2009) With HSR, subjects significantly more likely to prefer to buy healthy than unhealthy foods Talati et al. (40% vs 23%). Weak trend for MTL. No difference for GDA. (2017b) Percentage increase in number of subjects choosing to buy healthier foods (vs no label). Khandpur et al. Warning label, 16.1%; MTL, 9.8%. Overall p < 0.001. (2018) Label increase odds of selecting low-sodium box of crackers (vs no label). Simple traffic light Goodman et al. (2013) with text only (high/low sodium), OR = 4.2; as previous plus text stating amount sodium, OR = 2.4; text only (amount sodium and states high/low sodium), OR = 3.8. All differences are significant. No overall difference between MTL, MTL plus calories, tick, single traffic light plus specific Roberto et al. nutrients to limit, and no label. (2012b) Nutri-Score improved healthiness of diet (vs no label; not significant). GDA did not improve Egnell et al. diet. (2019) MTL and warning labels both improved healthiness of diet (vs no label). Intake of sugar, Machín et al. saturated fat, and sodium reduced by ~20%. All differences are significant. (2017) Selection of beverages high in sugar reduced by 7% by warning sign and by 2% by text warning Acton and (vs no label; not significant). HSR had no effect. Hammond (2018b)

Details as caption to Table 1.

Table 3 Findings from recent studies that tested whether FOP labels based on warning labels decrease the intention of shoppers to buy unhealthy foods (simulated shopping). Country

Number of subjects

Experimental design

Findings

Singapore

512

Australia

994

Subjects shopped at a virtual grocery store Food selected on a computer

USA

2381

UK

2002

Selection of foods high in sugar reduced by 2% by stop sign and by 4% by text warning than with no label. Ang et al. (2019) Label decreases odds of selecting sugar-sweetened beverage. Text warning OR = 0.49; as previous plus photo Billich et al. of rotten teeth, OR = 0.22; simple text only (amount sugar), OR = 0.47; HSR, OR = 0.45. All differences are (2018) significant. Parents 33% less likely to select sugar-sweetened beverage for child if it has a text warning than with no label Roberto (p < 0.001). et al. (2016) Compared with no label, parents are 51% less likely to select sugar-sweetened beverage for child if it has a Mantzari et al. (2018) text warning plus photo of rotten teeth (p < 0.001); 30% less likely if it has text warning plus image of teaspoon of sugar (p < 0.05).

Food selected on a computer Food selected on a computer

Reference

OR: odds ratio.

the Guideline Daily Amount and traffic-light systems. Public Health Nutrition, 20, 2308–2317. Aschemann-Witzel, J., Grunert, K. G., van Trijp, H. C. M., Bialkova, S., Raats, M. M., Hodgkins, C., et al. (2013). Effects of nutrition label format and product assortment on the healthfulness of food choice. Appetite, 71, 63–74. Aune, D., Keum, N., Giovannucci, E., Fadnes, L. T., Boffetta, P., Greenwood, D. C., et al. (2016). Nut consumption and risk of cardiovascular disease, total cancer, all-cause and cause-specific mortality: A systematic review and dose-response meta-analysis of prospective studies. BMC Medicine, 14, 207. Australian Government Department of Health (2016). Front-of-pack labelling updates. Australian Government Department of Healthhttp://foodregulation.gov.au/internet/ fr/publishing.nsf/Content/front-of-pack-labelling-1, Accessed date: 21 January 2019. Baraldi, L. G., Martinez Steele, E., Canella, D. S., & Monteiro, C. A. (2018). Consumption of ultra-processed foods and associated sociodemographic factors in the USA between 2007 and 2012: Evidence from a nationally representative cross-sectional study. BMJ Open, 8, e020574. Billich, N., Blake, M. R., Backholer, K., Cobcroft, M., Li, V., & Peeters, A. (2018). The effect of sugar-sweetened beverage front-of-pack labels on drink selection, health knowledge and awareness: An online randomised controlled trial. Appetite, 128,

References Acton, R. B., & Hammond, D. (2018b). The impact of price and nutrition labelling on sugary drink purchases: Results from an experimental marketplace study. Appetite, 121, 129–137. Acton, R. B., Vanderlee, L., & Hammond, D. (2018a). Influence of front-of-package nutrition labels on beverage healthiness perceptions: Results from a randomized experiment. Preventive Medicine, 115, 83–89. Ang, F. J. L., Agrawal, S., & Finkelstein, E. A. (2019). Pilot randomized controlled trial testing the influence of front-of-pack sugar warning labels on food demand. BMC Public Health, 19, 164. Anon About facts up front. http://www.factsupfront.org, Accessed date: 21 January 2019. Anon (2010). Food industry wins battle on “traffic light” labels. EurActiv 17 June 2010 http://www.euractiv.com/food-industry-wins-battle- traffic-light-labels-news495324, Accessed date: 9 July 2018. Arrúa, A., Machín, L., Curutchet, M. R., Martínez, J., Antúnez, L., Alcaire, F., et al. (2017). Warnings as a directive front-of-pack nutrition labelling scheme: Comparison with

7

Appetite 144 (2020) 104485

N.J. Temple

Journal of Clinical Nutrition, 91, 1102S–1108S. Khandpur, N., de Morais Sato, P., Mais, L. A., Bortoletto Martins, A. P., Spinillo, C. G., Garcia, M. T., et al. (2018). Are front-of-package warning labels more effective at communicating nutrition information than traffic-light labels? A randomized controlled experiment in a Brazilian sample. Nutrients, 10, E688 pii. Khandpur, N., Mais, L. A., de Morais Sato, P., Martins, A. P. B., Spinillo, C. G., & Rojas, C. F. U. (2019). Choosing a front-of-package warning label for Brazil: A randomized, controlled comparison of three different label designs. Food Research International, 121, 854–861. de Koning, L., Malik, V. S., Rimm, E. B., Willett, W. C., & Hu, F. B. (2011). Sugar-sweetened and artificially sweetened beverage consumption and risk of type 2 diabetes in men. American Journal of Clinical Nutrition, 93, 1321–1327. Lohse, G. L. (1997). Consumer eye movement patterns on yellow pages Advertising. Journal of Advertising, 26, 61–73. Lupton, J. R., Balentine, D. A., Black, R. M., Hildwine, R., Ivens, B. J., Kennedy, E. T., et al. (2010). The Smart choices front-of-package nutrition labeling program: Rationale and development of the nutrition criteria. American Journal of Clinical Nutrition, 91, 1078S–1089S. Machín, L., Cabrera, M., Curutchet, M. R., Martínez, J., Giménez, A., & Ares, G. (2017). Consumer perception of the healthfulness of ultra-processed products featuring different front-of-pack nutrition labeling scheme. Journal of Nutrition Education and Behavior, 49, 330–338 e1. Magnusson, R. S. (2010). Obesity prevention and personal responsibility: The case of front-of-pack food labelling in Australia. BMC Public Health, 10, 662. Maki, K. C., Eren, F., Cassens, M. E., Dicklin, M. R., & Davidson, M. H. (2018). ω-6 polyunsaturated fatty acids and cardiometabolic health: Current evidence, controversies, and research gaps. Advances in Nutrition, 9, 688–700. Mantzari, E., Vasiljevic, M., Turney, I., Pilling, M., & Marteau, T. (2018). Impact of warning labels on sugar-sweetened beverages on parental selection: An online experimental study. Preventive Medicine Reports, 12, 259–267. Maubach, N., Hoek, J., & Mather, D. (2014). Interpretive front-of-pack nutrition labels. Comparing competing recommendations. Appetite, 82, 67–77. Moodie, R., Stuckler, D., Monteiro, C., Sheron, N., Neal, B., Thamarangsi, T., et al. (2013). Profits and pandemics: Prevention of harmful effects of tobacco, alcohol, and ultraprocessed food and drink industries. Lancet, 381, 670–679. Mozaffarian, D. (2016). Dietary and policy priorities for cardiovascular disease, diabetes, and obesity: A comprehensive review. Circulation, 133, 187–225. Narain, A., Kwok, C. S., & Mamas, M. A. (2016). Soft drinks and sweetened beverages and the risk of cardiovascular disease and mortality: A systematic review and metaanalysis. International Journal of Clinical Practice, 70, 791–805. Nestle, M. (2007). Food politics: How the food industry influences nutrition and health. Berkeley: University of California Press. Rahkovsky, I., Lin, B. H., Lin, C. T. J., & Lee, J. Y. (2013). Effects of the Guiding Stars Program on purchases of ready-to-eat cereals with different nutritional attributes. Food Policy, 43, 100–107. Reyes, M., Garmendia, M. L., Olivares, S., Aqueveque, C., Zacarías, I., & Corvalán, C. (2019). Development of the Chilean front-of-package food warning label. BMC Public Health, 19, 906. Roberto, C. A., Bragg, M. A., Schwartz, M. B., Seamans, M. J., Musicus, A., & Novak, N. (2012a). Facts up front versus traffic light food labels a randomized controlled trial. American Journal of Preventive Medicine, 43, 134–141. Roberto, C. A., Bragg, M. A., Seamans, M. J., Mechulan, R. L., Novak, N., & Brownell, K. D. (2012b). Evaluation of consumer understanding of different front-of-package nutrition labels, 2010-2011. Preventing Chronic Disease, 9, E149. Roberto, C. A., Wong, D., Musicus, A., & Hammond, D. (2016). The influence of sugarsweetened beverage health warning labels on parents' choices. Pediatrics, 137, e20153185. Roseman, M. G., Joung, H. W., & Littlejohn, E. I. (2018). Attitude and behavior factors associated with front-of-package label use with label users making accurate product nutrition assessments. Journal of the Academy of Nutrition and Dietetics, 118, 904–912. Sacks, G., Rayner, M., & Swinburn, B. (2009). Impact of front-of-pack ‘traffic-light’ nutrition labelling on consumer food purchases in the UK. Health Promotion International, 24, 344–352. Sacks, G., Tikellis, K., Millar, L., & Swinburn, B. (2011). Impact of ‘traffic-light’ nutrition information on online food purchases in Australia. Australian & New Zealand Journal of Public Health, 35, 122–126. Sanjari, S. S., Jahn, S., & Boztug, Y. (2017). Dual-process theory and consumer response to front-of-package nutrition label formats. Nutrition Reviews, 75, 871–882. Shangguan, S., Afshin, A., Shulkin, M., Ma, W., Marsden, D., Smith, J., et al. (2019). A meta-analysis of food labeling effects on consumer diet behaviors and industry practices. American Journal of Preventive Medicine, 56, 300–314. Swedish National Food Agency (2019). https://www.livsmedelsverket.se/en/food-andcontent/labelling/nyckelhalet, Accessed date: 13 July 2019. Talati, Z., Norman, R., Pettigrew, S., Neal, B., Kelly, B., & Dixon, H. (2017b). The impact of interpretive and reductive front-of-pack labels on food choice and willingness to pay. International Journal of Behavioral Nutrition and Physical Activity, 14, 171. Talati, Z., Pettigrew, S., Ball, K., Hughes, C., Kelly, B., Neal, B., et al. (2017a). The relative ability of different front-of-pack labels to assist consumers discriminate between healthy, moderately healthy, and unhealthy foods. Food Quality and Preference, 59, 109–113. Te Morenga, L. A., Howatson, A. J., Jones, R. M., & Mann, J. (2014). Dietary sugars and cardiometabolic risk: Systematic review and meta-analyses of randomized controlled trials of the effects on blood pressure and lipids. American Journal of Clinical Nutrition, 100, 65–79. Temple, N. J. (2016). Strategic nutrition: A vision for the twenty-first century. Public Health Nutrition, 19, 164–175.

233–241. Borgmeier, I., & Westenhoefer, J. (2009). Impact of different food label formats on healthiness evaluation and food choice of consumers: A randomized-controlled study. BMC Public Health, 9, 184. Carreño, I. (2015). Chile's black STOP sign for foods high in fat, salt or sugar. European Journal of Risk Regulation, 6, 622–628. Carter, O. B., Mills, B. W., Lloyd, E., & Phan, T. (2013). An independent audit of the Australian food industry's voluntary front-of-pack nutrition labelling scheme for energy-dense nutrition-poor foods. European Journal of Clinical Nutrition, 67, 31–35. Choices International Foundation (2019). Choices programme. https://www. choicesprogramme.org, Accessed date: 13 July 2019. Cook, N. R., Cutler, J. A., Obarzanek, E., Buring, J. E., Rexrode, K. M., Kumanyika, S. K., et al. (2007). Long term effects of dietary sodium reduction on cardiovascular disease outcomes: Observational follow-up of the trials of hypertension prevention (TOHP). BMJ, 334, 885–888. Cowburn, G., & Stockley, L. (2005). Consumer understanding and use of nutrition labelling: A systematic review. Public Health Nutrition, 8, 21–28. Ducrot, P., Julia, C., Méjean, C., Kesse-Guyot, E., Touvier, M., Fezeu, L. K., et al. (2016). Impact of different front-of-pack nutrition labels on consumer purchasing intentions: A randomized controlled trial. American Journal of Preventive Medicine, 50, 627–636. Ducrot, P., Mejean, C., Julia, C., Kesse-Guyot, E., Touvier, M., Fezeu, L., et al. (2015). Effectiveness of front-of-pack nutrition labels in French adults: Results from the nutrinet-sante cohort study. PLoS One, 10, e0140898. Edge, M. S., Toner, C., Kapsak, W. R., & Geiger, C. J. (2014). The impact of variations in a fact-based front-of-package nutrition labeling system on consumer comprehension. Journal of the Academy of Nutrition and Dietetics, 114, 843–854 e8. Egnell, M., Boutron, I., Péneau, S., Ducrot, P., Touvier, M., Galan, P., et al. (2019). Frontof-pack labeling and the nutritional quality of students' food purchases: A 3-arm randomized controlled trial. American Journal of Public Health, 109, 1122–1129. Egnell, M., Talati, Z., Hercberg, S., Pettigrew, S., & Julia, C. (2018). Objective understanding of front-of-package nutrition labels: An international comparative experimental study across 12 countries. Nutrients, 10, E1542. Fischer, L. M., Sutherland, L. A., Kaley, L. A., Fox, T. A., Hasler, C. M., Nobel, J., et al. (2011). Development and implementation of the guiding stars nutrition guidance program. American Journal of Health Promotion, 26, e55–e63. Food Standards Agency (Department of Health). (2016). Guide to creating a front of pack (fop) nutrition label for pre-packed products sold through retail outlets. https://www.gov. uk/government/publications/front-of-pack-nutrition-labelling-guidance, Accessed date: 21 January 2019. Food and Drink Federation (2019). Food labelling. http://www.foodlabel.org.uk/label/ gda_values.aspx, Accessed date: 21 January 2019. Goodman, S., Hammond, D., Hanningm, R., & Sheeshka, J. (2013). The impact of adding front-of-package sodium content labels to grocery products: An experimental study. Public Health Nutrition, 16, 383–391. Goodman, S., Vanderlee, L., Acton, R., Mahamad, S., & Hammond, D. (2018). The impact of front-of-package label design on consumer understanding of nutrient amounts. Nutrients, 10, E1624 pii. Gorski, F. M. T., Werth, P. M., Musicus, A. A., Bragg, M. A., Graham, D. J., & Elbel, B. (2018). Comparing five front-of-pack nutrition labels' influence on consumers' perceptions and purchase intentions. Preventive Medicine, 106, 114–121. Guiding Stars (2019). About. https://guidingstars.ca/about, Accessed date: 21 January 2019. Hawley, K. L., Roberto, C. A., Bragg, M. A., Liu, P. J., Schwartz, M. B., & Brownell, K. D. (2013). The science on front-of-package food labels. Public Health Nutrition, 16, 430–439. He, F. J., & MacGregor, G. A. (2009). A comprehensive review on salt and health and current experience of worldwide salt reduction programmes. Journal of Human Hypertension, 23, 363–384. Hersey, J. C., Wohlgenant, K. C., Arsenault, J. E., Kosa, K. M., & Muth, M. K. (2013). Effects of front-of-package and shelf nutrition labeling systems on consumers. Nutrition Reviews, 71, 1–14. Hickman, M. (2010). Laid bare, the lobbying campaign that won the food labelling battle. The Independent 18 June 2010 http://www.independent.co.uk/life-style/food-anddrink/news/laid-bare-the-lobbying-campaign-thatwon-the-food-labelling-battle2003686.html, Accessed date: 18 July 2018. Hobin, E., Bollinger, B., Sacco, J., Liebman, E., Vanderlee, L., Zuo, F., et al. (2017). Consumers' response to an on-shelf nutrition labelling system in supermarkets: Evidence to inform policy and practice. The Milbank Quarterly, 95, 494–534. Hodgkins, C. E., Raats, M. M., Fife-Schaw, C., Peacock, M., Gröppel-Klein, A., Koenigstorfer, J., et al. (2015). Guiding healthier food choice: Systematic comparison of four front-of-pack labelling systems and their effect on judgements of product healthiness. British Journal of Nutrition, 113, 1652–1663. Institute of Medicine (2012). Front-of-Package nutrition rating systems and symbols: Promoting healthier choices. Washington, DC: The National Academies Press, Accessed date: 21 January 2019. Julia, C., Blanchet, O., Méjean, C., Péneau, S., Ducrot, P., Allès, B., et al. (2016). Impact of the front-of-pack 5-colour nutrition label (5-CNL) on the nutritional quality of purchases: An experimental study. International Journal of Behavioral Nutrition and Physical Activity, 13, 101. Julia, C., & Hercberg, S. (2016). Research and lobbying conflicting on the issue of a frontof-pack nutrition labelling in France. Archives of Public Health, 74, 51. Julia, C., Kesse-Guyot, E., Touvier, M., Mejean, C., Fezeu, L., & Hercberg, S. (2014). Application of the British Food Standards Agency nutrient profiling system in a French food composition database. British Journal of Nutrition, 112, 1699–1705. Katz, D. L., Njike, V. Y., Rhee, L. Q., Reingold, A., & Ayoob, K. T. (2010). Performance characteristics of NuVal and the overall nutrition quality index (ONQI). American

8

Appetite 144 (2020) 104485

N.J. Temple

pack nutrition label stimulates healthier product development: A quantitative analysis. International Journal of Behavioral Nutrition and Physical Activity, 7, 65. Watson, W. L., Kelly, B., Hector, D., Hughes, C., King, L., Crawford, J., et al. (2014). Can front-of-pack labelling schemes guide healthier food choices? Australian shoppers' responses to seven labelling formats. Appetite, 72, 90–97. World Cancer Research Fund International (2019). Building momentum: Lessons on implementing a robust front-of-pack food label. , Accessed date: 21 March 2019 wcrf.org/ frontofpack. Young, L., & Swinburn, B. (2002). Impact of the Pick the Tick food information programme on the salt content of food in New Zealand. Health Promotion International, 17, 13–19.

Temple, N. J., & Fraser, J. (2014). Food labels: A critical assessment. Nutrition, 30, 257–260. Temple, N. J., Guercio, V., & Tavani, A. (2019). The mediterranean diet and cardiovascular disease: Gaps in the evidence and research challenges. Cardiology in Review, 27, 127–130. The NuVal Attribute Program (2017). https://www.nuval.com, Accessed date: 21 January 2019. Thomson, R. K., McLean, R. M., Ning, S. X., & Mainvil, L. A. (2016). Tick front-of-pack label has a positive nutritional impact on foods sold in New Zealand. Public Health Nutrition, 19, 2949–2958. Vyth, E. L., Steenhuis, I. H., Roodenburg, A. J., Brug, J., & Seidell, J. C. (2010). Front-of-

9