FRONT-OF-PACK NUTRITION LABELLING: THE IMPACT OF THE MTL LABEL ON BEVERAGES

FRONT-OF-PACK NUTRITION LABELLING: THE IMPACT OF THE MTL LABEL ON BEVERAGES

Master thesis, MSc Marketing Management

University of Groningen, Faculty of Economics and Business

January 15, 2018

BOUDEWIJN VAN DEN BERG Studentnumber: 2192063

Vrolikstraat 276-4 1092 TZ Amsterdam

tel.: +316 11772989

e-mail: bjevandenberg@gmail.com

Supervisor

Prof. Dr. Ir. Koert van Ittersum

FRONT-OF-PACK NUTRITION LABELLING: THE IMPACT OF THE MTL LABEL ON BEVERAGES

ABSTRACT

Obesity is a global pandemic, currently the number of obese adults outweigh those who are underweight. The rising consumption of sugary beverages has been a major contributor to this global obesity pandemic. In the US, soda is considered ''the'' number one energy contributor, with children and adolescents nowadays deriving 10% to 15% of their total calorie intake from sugar- sweetened beverages (SSBs) and 100% fruit juices. To tackle the global obesity pandemic, improve diet quality and encourage healthy behaviour, several parties designed and implemented front-of-pack (FOP) nutrition labels to help the consumer with regard to healthy food choices. A multitude of FOP labels have been designed and implemented, but the Multiple-Traffic-Light (MTL) label is by many scholars promoted as the most effective format. This study wants to examine the impact of the most effective and widely promoted FOP format on the number one energy contributor. This study was interested in the impact of the MTL label on consumers' perceived healthiness and behavioural intention and made a distinction between the impact of the MTL label on the less healthy and healthier variants of the beverages. The results indicated that only in some instances did the MTL label influence the consumers' behavioural intention. While the MTL label did in all cases impact the consumers' perceived healthiness, the observed impact was not always the desired response.

Keywords: MTL label; FOP nutrition labelling; Beverages; Soda; Fruit juice

Table of Contents

1. INTRODUCTION ... 4

1.1 Promoting healthy choices ...4

1.2 Nutrition labelling ...4

1.3 Traffic light labelling and beverages ...5

1.4 Similarities ...6

2. LITERATURE REVIEW ... 6

2.1 Sugary beverages ...6

2.2 Nutrition labels ...7

2.2.1 Front-of-pack labels ...8

2.2.2 Traffic light labels ...8

2.3 Health literacy ...10

3. CONCEPTUAL MODEL AND HYPOTHESES ... 11

Hypothesis 1a ...12

Hypothesis 1b ...12

Hypothesis 2a ...13

Hypothesis 2b ...14

Hypothesis 3a ...14

Hypothesis 3b ...14

Hypothesis 4a ...15

Hypothesis 4b ...15

4. METHODS ... 15

4.1 Participants ...15

4.2 Materials ...16

4.3 Products ...16

4.4 Procedure ...16

4.5 Measures ...17

4.6 Control variables ...18

5. RESULTS ... 18

6. DISCUSSION ... 23

7. LIMITATIONS ... 25

REFERENCES ... 26

APPENDICES ... 32

Appendix A. Labelling criteria as determined by the FSA ...32

Appendix B. Survey questions ...32

1. INTRODUCTION

Obesity is a global pandemic and is now widely recognized as one of today's biggest threats to global public health (Swinburn et al., 2011). By the year 2000 a historical turning point was reached, for the first time in human history the number of adults with excess weight surpassed the number of those who were underweight (Caballero, 2007). ''The world has transitioned from an era when underweight prevalence was more than double that of obesity, to one in which more people are obese than underweight.'' From 1975 to 2014 the global prevalence of underweight decreased from 13.8% to 8.8% for men and from 14.6% to 9.7% for women. The number of obese people, however, increased from 3.2% to 10.8% for men and from 6.4% to 14.9% for women during the same period (Cesare, Bentham, Stevens, Geleijnse, & Kromhout, 2016). Underweight prevalence decreased, but has been substituted by an even bigger increase in obesity and ''if present trends continue, by 2025, global obesity prevalence will reach 18% in men and surpass 21% in women'' (Cesare et al., 2016).

1.1 Promoting healthy choices

To reverse this alarming trend and promote health, ''the food selection environment has to be more conducive to healthy choices'' (Cowburn & Stockley, 2005). Nutrition labelling is one of the means applied to make the food environment more supportive and is considered a crucial component of strategies tackling obesity (World Health Organization., 2004).

Promoting healthy food choices by providing nutritional information is not new. Already by the year 1990, the Nutritional Labelling and Education Act (NLEA) was established with the goal to increase public health (Jordan Lin, Lee, & Yen, 2004; Sung Yong Kim, Nayga, & Capps, 2000; Kurtzweil, 1993). It required mandatory nutrition labelling for most food products and replaced the voluntary label system established by the Food and Drug Administration (FDA) in 1973. Also in 1990, the Council Directive (90/496/EEC) on nutrition labelling rules for foodstuffs was established for all members of the European Union (EU).

1.2 Nutrition labelling

When applied correctly, the mandatory back-of-pack labels introduced in 1990 are known

to lower intakes of total fat, saturated fat and cholesterol and have a positive effect on diet quality

(Kim, Nayga, & Capps, 2001; Kim et al., 2000) ^. Despite its positive effect, good intentions and

broad appliance, the actual use of back-of-pack labels appears to be lower than expected (Cowburn

& Stockley, 2005; Rayner, Boaz, & Higginson, 2001). This is often caused by a lack of motivation or understanding (Byrd-Bredbenner, Wong, & Cottee, 2000; Shine, O’Reilly, & O’Sullivan, 1997;

Wandel, 1997). Interpreting nutrition labels requires high literacy and numeracy skills and is therefore by many people seen as a barrier (Rothman et al., 2006). Consumers often lack the ability to comprehend the presented information and convert this mainly numerical information into healthy food choices (Cowburn & Stockley, 2005). Even consumers who do possess the literacy and numeracy skills can have difficulties with interpreting nutrition labels (Rothman et al., 2006).

To simplify the decision making process for both low and highly literate consumers, several parties advocated the use of a uniform and easy to understand front-of-pack label that indicates the overall healthiness of a product (van Kleef, van Trijp, Paeps, & Fernández-Celemín, 2008).

Through the years, several front-of-pack labels were designed and implemented by retailers, governments, non-governmental organisations and food manufacturers. Debate on which format is the most effective remains (Lobstein, Landon, & Lincoln, 2007). The systematic review and meta-analysis performed by Cecchini and Warin (2016) indicates that all three most implemented food labelling systems are successful in steering the consumer towards the healthier option, with traffic light systems being the most effective.

1.3 Traffic light labelling and beverages

In line with the findings of Cecchini and Warin (2016) and despite some mixed findings such as those shown by Sacks, Rayner, & Swinburn (2009) in which traffic-light labels did not substantially influence supermarket sales, most research does indeed confirm the effectiveness of traffic light systems (Roberto et al., 2012; Sonnenberg et al., 2013; Thorndike, Sonnenberg, Riis, Barraclough, & Levy, 2012)

Thorndike et al., (2012) investigated the influence of colour-coding on both food and beverages and found that colour-coding was especially effective for beverages. The sales of beverages with a red colour decreased with 16.5%, while the sales of the green coloured beverages increased with 9.6%.

Because of the striking findings of Thorndike et al., (2012), this study also wants to

examine the influence of colour-coding on beverages. This study is however mainly interested in

the influence of colour-coding on unhealthy sugary "red beverages'' of which the consumption

should be discouraged, moreover it also examines what colour-coding does for the healthier light variants of these specific unhealthy beverages, since these are the most likely to be chosen as alternatives.

Unlike Thorndike et al., (2012), this study will consider fruit juices as red unhealthy beverage. In their study red beverages were sugar-sweetened beverages (SSBs) with 200 or more kilocalories per container (Thorndike et al., 2012). Fruit juices were not considered as red beverage, since they are not artificially sweetened with fructose corn syrup or sucrose by manufacturers, individuals or establishments (Malik & Hu, 2015). As fruit juices are often perceived as healthy, yet have a high sugar and calorie content, it was considered important to include this category of beverages in the study.

1.4 Similarities

Fruit juices might not contain added sugar, however this does not mean that they are low in sugar. ''Contrary to the general perception of the public, and of many health-care professionals, that drinking fruit juice is a positive health behaviour, their consumption might not be substantially different in health terms from consumption of SSBs'' (Gill & Sattar, 2014). The caloric content of fruit juices is similar to that of SSBs, they contain limited nutrients, and like SSBs, fruit juices have been associated with increased risks of weight gain (Dennison, Rockwell, & Baker, 1997;

Faith, Dennison, Edmunds, & Stratton, 2006). Despite these similarities in terms of health, SSBs are perceived as unhealthy, while fruit juices are perceived as healthy (Dennison et al., 1997;

Wang, Bleich, & Gortmaker, 2008). This misperception in healthiness makes fruit juice a silent killer and therefore, just as or even more suitable for front-of-pack nutrition labelling.

2. LITERATURE REVIEW 2.1 Sugary beverages

Soda plays without question a major role in the alarming obesity pandemic. In his study on

foods contributing to energy intake in the US, Block (2004) found that soda is ''the number one''

energy contributor and contributes 7.1% to the total energy intake of the US population in 1999-

2000. Nielsen & Popkin (2004) investigated the changes in beverage intake between 1977 and

2001 and found similar percentages. Even more alarming is the contribution of sugar-sweetened

beverages and 100% fruit juices to the calorie intake of children and adolescents, who nowadays derive 10% to 15% of their total energy from these sugary beverages (Wang et al., 2008).

The evidence for a role of SSBs in the development of obesity is becoming increasingly convincing (Gill & Sattar, 2014; Wang et al., 2008; Welsh, 2005), while the role of fruit juices however remains relatively unexplored. Studies often focus on merely SSBs and therefore only include: ''soda, sports drinks, energy drinks, lemonade, fruit drinks'' (Wang et al., 2008). Fruit drinks are considered SSBs, however fruit juices are not. This is because fruit drinks are juices and nectars with added sugar and fruit juices are unsweetened and made from 100% fruit juice (Wang et al., 2008).

Unsweetened and 100% fruit juice, however, does not mean healthy. Despite the lack of added sugar, the consumption of these fruit juices is also associated with weight gain (Dennison et al., 1997; Faith et al., 2006). Among consumers there seems to be a misperception that fruit juices and smoothies are low-sugar alternatives to SSBs (Gill & Sattar, 2014) and this is not true since fruit juices and sodas have a similar energy density. Apple juice typically contains 110 kcal, 26g of sugar per 250ml and coke regular 105 kcal and 26.5g of sugar per 250ml (Gill & Sattar, 2014). Despite the high sugar content of fruit juices, companies often market them as healthy (Institute of Medicine, 2006) and often refer to fruit juices as sources of vitamins and minerals, while ignoring their high sugar content.

2.2 Nutrition labels

Since 1990, in both the US and EU, back-of-pack nutrition labels have been required on

most food products. Despite broad appliance and good intentions, the actual use of the back-of-

pack labels appears to be lower than expected (Cowburn & Stockley, 2005). Especially older

consumers and consumers with a low level of education indicate that they find back-of-pack

nutrition labels confusing and complex (Cowburn & Stockley, 2005). Part of the reported

complexity is caused by the fact that during the decision making process, multiple nutrients have

to be taken into account simultaneously and comparing these is difficult (Feunekes, Gortemaker,

Willems, Lion, & van den Kommer, 2008). A more intuitive practical tool felt needed that would

improve consumers' decision making process, but not required detailed nutritional knowledge. As

an answer to this problem, several parties designed and implemented countless front-of-pack

labels, that were both easier to understand and use.

In today’s shopping environment this is not a frivolous luxury since consumers are bombarded with information. Due to limited cognitive capabilities and a lack of motivation, consumers do not devote too much cognitive effort to the low-involvement products and have switched to ''peripheral route information processing'' (Fiske & Neuberg, 1990; Petty, Cacioppo,

& Schumann, 1983). As consumers only glance at the nutrition information (Higginson, Rayner, Draper, & Kirk, 2002), a simple front-of-pack label which reduces process load (Scott & Worsley, 1994) to provide an overall interpretation of the healthiness of the product can help to make the decision making process with regard to healthy foods easier.

2.2.1 Front-of-pack labels

Within the realm of simple front-of-pack labels a distinction can be made between fact and criteria based systems (Pereira, 2010; Schor, Maniscalco, Tuttle, Alligood, & Kapsak, 2010). Fact based labelling systems simply summarize the quantitative nutrition information provided by the back-of-pack label in a more concise way, Guidelines daily amounts (GDA) is such a system and shows the nutritional content as a percentage of the daily recommended intake.

The second group, criteria based labelling systems base their categorization on a set of criteria, either categorical or continuous. Categorical systems use specifically defined criteria for each category. If products meet these criteria they will receive a label, for example a health tick.

Continuous labelling systems on the other hand use the same nutrition criteria ''across'' product categories (Kleef & Dagevos, 2015). A well known and often implemented continuous labelling system is the NuVal system, which weighs more than thirty nutrients and calculates a single score for each food item that can be compared across all product categories (Kleef &

Dagevos, 2015).

2.2.2 Traffic light labels

Other well known continuous labelling systems are the simple-traffic-light label (STL) and

the multiple-traffic-light label (MTL). The multiple-traffic-label recommended by the UK Food

Safety Association being the most well known. This labelling system is based on a model that

calculates scores for all foods based on the same set of nutrients and mathematical formulae

(Foltran et al., 2010). Subsequently score threshold criteria are set along with colour code

classification to identify the relative nutrient level in a food (Kleef & Dagevos, 2015). Red

indicates the presence of an especially high amount of a nutrient in a product, yellow a moderate amount and green a low amount. The use of these colours makes traffic-light-labels self explanatory, which is often seen as a key advantage of traffic light systems (Grunert & Wills, 2007;

Kleef & Dagevos, 2015).

A broad body of literature even claims the multiple-traffic-light label to be the most effective FOP format (Balcombe, Fraser, & Falco, 2010; Borgmeier & Westenhoefer, 2009;

Cecchini & Warin, 2016; Hawley et al., 2013; Roberto et al., 2012). This claim is however not supported by all scholars, ''when comparing multiple traffic lights and GDA-based systems, the results are not clear-cut'' (Grunert & Wills, 2007), ''traffic-light labels did not substantially influence supermarket sales'' (Sacks et al., 2009).

Despite the ongoing verdict on which FOP format is most effective, there seems to be consensus on which traffic light system is most effective. The simple traffic light label appears to be liked because it is easy to understand, but is often judged for being “too didactic” (Food Standards Agency, 2004a), “paternalistic” (McDonald’s, 2005) or ''simplistic'' (BEUC., 2006).

''Consumers still would like to know what this simplified information stands for and how the red light or the health logo has been arrived'' (Grunert, Wills, & Fernández-Celemín, 2010). The multiple-traffic-light label does indicate the nutritional information in a more sophisticated way, but also the multiple-traffic-light label has its disadvantages.

Two frequently heard criticisms of the MTL label is it ignores that foods have more constituents than the four listed nutrients on the label and, in some instances, the presence of preliminary green lights can create a false sense of security (Hagen, 2010). Coke regular for example carries a green light for three out of four nutrients, if consumers do not inspect the multiple-traffic-light label thoroughly enough, they might just base its healthiness on the presence preliminary green lights, corresponding with nutrients that should not play a decisive role for this type of product.

''The more green lights on the product, the healthier the choice'' (Kleef & Dagevos, 2015),

more green than red lights, does however not mean a healthy choice. Another problem associated

with the MTL label is ambiguity and confusion when multiple lights vary in colour or when a

product carries a similar number of green and red lights (Kleef & Dagevos, 2015). It could be even

more confusing if a product displays one red light for a relevant and problematic nutrient within

the product category, but three green lights for nutrients that are in general not very prevalent in

the category but are below the across category threshold score. ''Displaying multiple lights simultaneously could cause ambiguity and does not provide the consumer with a clear buying strategy'' (Hagen, 2010).

2.3 Health literacy

As has been mentioned previously, especially older consumers and people with a low level of education indicate that they find back-of-pack nutrition labels confusing and complex (Cowburn

& Stockley, 2005). This is not surprising since interpreting nutrition labels requires high literacy and numeracy skills (Rothman et al., 2006) and in contrast with what one might expect, poor literacy skills among adults in developed countries is surprisingly common, ''estimates of the proportion of the population in individual Organization for Economic Co-operation and Development (OECD) countries lacking functional literacy skills range from 7% to 47%'' (UN Development Program, 2007).

In its simplest form (functional) literacy skills can be defined as: ''sufficient basic skills in reading and writing to be able to function effectively in everyday situations'' (Nutbeam, 2000).

People who at least possess functional literacy skills are better able to participate in society and are better able to exert a higher degree of control over everyday events (Nutbeam, 2008). It is therefore not surprising that low literacy is associated directly and indirectly with poor health outcomes (Ratzan & Parker, 2000).

The term health literacy is highly compatible with basic functional literacy and can be defined as ''the degree to which individuals have the capacity to obtain, process, and understand basic health information and services needed to make appropriate health decisions'' (Ratzan &

Parker, 2000). The result of being health literate is improved knowledge and understanding of health determinants (Nutbeam, 2000).

Health literacy is context specific and the field has acknowledged distinct health literacy

forms, of which nutrition literacy explicitly focuses on literacy skills in a food context (Velardo,

2015). ''Nutrition literacy reflects the ability to access, interpret, and use nutrition information''

(Carbone & Zoellner, 2012). Having functional health literacy skills in the nutrition context leads

to improved knowledge of health risks, components of a healthy diet, and the benefits of good

nutrition (Velardo, 2015). People with ''interactive literacy'' should be able to translate the

declarative knowledge into positive dietary choices. The third level ''Critical literacy'', demands

''critical awareness and emancipatory action to address barriers to good nutrition'' (Velardo, 2015).

Hence, people differ in their level of (health) literacy and this has a major influence on their diet quality.

3. CONCEPTUAL MODEL AND HYPOTHESES

It has been made clear that the group of sugary beverages is a troublesome category and a major contributor to the global obesity pandemic. In the US, soda is considered ''the'' number one energy contributor (Block, 2004), with children and adolescents nowadays deriving 10% to 15%

of their total calorie intake from sugar-sweetened beverages (SSBs) and 100% fruit juices (Wang et al., 2008). One of the means implemented to tackle the obesity pandemic is FOP nutrition labelling. Thorndike et al., (2012) showed that the simple-traffic-Light format was especially effective for beverages.

Due to the striking results shown by Thorndike et al., (2012), this study also wants to investigate the influence of colour-coding on beverages. Unlike Thorndike's work which focuses on the effectiveness of the simple-traffic light label, this study will examine the influence of the multiple-traffic light label, considered by many as the most effective labelling format.

The aim of this study is to investigate the impact of the MTL label on consumers' perceived healthiness and behavioural intention through consumers’ perceived healthiness. A distinction is made between the impact of the MTL label on the less healthy and healthier variants of the beverages. This study is furthermore interested in whether the effect of the MTL label differs based on the type of beverage (e.g. soda, fruit juice) or consumers' level of health literacy.

Fig. 1. Model of the effect of displaying a Multiple-Traffic-Light-Label on consumers' perceived healthiness and

behavioural intention.

The model states that differences in consumers' perceived healthiness of beverages and changes in behavioural intention through consumers' perceived healthiness, can be explained from whether a multiple-traffic-light label is displayed and that the strength and direction of its effect can differ based on the type of beverage (e.g. soda, fruit juice) or consumers' level of health literacy. As mentioned, a distinction is made between the impact of the MTL label on the less healthy and healthier variants of the beverages. Sodas and fruit juices both have variants that are high and low in sugar. In this study, the variants that are high in sugar are considered the ''less healthy'' variants and the variants that are low in or contain no sugar are considered ''the healthier'' variants.

Previous literature on nutrition labelling by Steenhuis et al., 2010, indicates that participants significantly increased their perceived healthiness of a product when a nutrition logo indicating a healthier choice within the product group was displayed. In the study by Feunekes et al., (2008) on the effectiveness of front-of-pack labels, participants after being exposed to FOP labels increased their perceived healthiness of the healthier products and slightly decreased their perceived healthiness of the less healthy products. In this study, similar results are expected for beverages, this leads to the following hypotheses:

Hypothesis 1a. Exposure to a MTL label will decrease consumers' perceived healthiness of the less healthy variants.

Hypothesis 1b. Exposure to a MTL label will increase consumers' perceived healthiness of the healthier variants.

In line with the findings of Feunekes et al., (2008), it is expected that after exposure to a MTL label, participants will decrease their perceived healthiness of the less healthy variants (H1a).

Displaying a MTL label will thus decrease consumers' perceived healthiness of the less healthy

soda and the less healthy fruit juice. The expected decrease in perceived healthiness of the less

healthy variants is however not expected to be the same for both beverage types. Despite the

similarities between the less healthy variants of both beverage types in terms of health (Dennison

et al., 1997; Faith et al., 2006; Gill & Sattar, 2014), the decrease in perceived healthiness is

expected to be stronger for fruit juices than for sodas. The expected cause of this difference in

decrease of consumers' perceived healthiness between the beverage types, is the often misperceived sugar content of fruit juices, which people tend to underestimate by 48%, and to a lesser extent the overestimation of the sugar content of sodas by 18% (Gill & Sattar, 2014).

Because people seem to have an especially inaccurate health perception of fruit juices, the distance or adjustment from the initial health perception of the beverage without MTL label to the health perception of the beverage with MTL label is expected to be bigger for fruit juices than for sodas. This leads to the following hypothesis:

Hypothesis 2a. The decrease in consumers' perceived healthiness of the less healthy variants will be bigger for fruit juices than for sodas.

In line with the findings of Feunekes et al., (2008) and Steenhuis et al., 2010, it is expected that after exposure to a MTL label, participants will increase their perceived healthiness of the healthier variants (H1b). This is again expected for both beverage types, but the increase in perceived healthiness of the healthier variants is expected to be smaller for fruit juices than for sodas. The cause of this difference is again expected to be in the health perceptions of the beverages when the MTL label is not displayed.

Unlike the ''less healthy'' variants of the beverages, the ''healthier'' light variants do not have the same nutritional value and thus their MTL label slightly differs. The ''healthier'' light variant of soda is sugar free and the ''healthier'' light variant of fruit juice is low in sugar. Despite their minor difference in objective value, there is expected that both will be perceived as healthy when the MTL label is displayed.

A bigger difference in health perception between the beverage types is expected when the MTL label is not displayed. The ''less healthy'' variant of fruit juice without MTL label is already perceived as healthy (Dennison et al., 1997) and the same is expected for the ''healthier'' variant of fruit juice. For the ''healthier'' soda however, still a low perceived healthiness is expected, since a ''healthier'' soda remains a soda.

A more accurate initial health perception is thus expected for the ''healthier'' fruit juice than

for the ''healthier'' soda and therefore the distance or adjustment from the initial health perception

without MTL label to the health perception with MTL label is expected to be smaller for fruit

juices than for sodas. This leads to the following hypothesis:

Hypothesis 2b. The increase in consumers' perceived healthiness of the healthier variants will be will smaller for fruit juices for than for sodas.

Not only the type of beverage, but also the type of consumer in terms of health literacy is expected to have influence on the effect of the MTL label on consumers' perceived healthiness. As has been mentioned before, the result of being health literate is improved knowledge and understanding of health determinants (Nutbeam, 2000).

Low health literate consumers thus do ''not'' have improved knowledge, these consumers find it difficult to make informed health decisions because interpreting back-of-pack labels requires high literacy skills. It is expected low health literate consumers would benefit most from a simple FOP label which provides easy to understand nutritional information.

From high health literate consumers, no big adjustments in perceived healthiness after being exposed to a MTL label are expected, since the factual information provided by the MTL label is not likely to be new for these consumers. For the low health literate consumers however, the information provided by the MTL ''is'' likely to provide new insights and bigger differences in perceived healthiness between baseline and post measure are expected from these consumers. This leads to the following hypotheses:

Hypothesis 3a. The decrease in consumers' perceived healthiness of the less healthy variants will be smaller for high health literate consumers.

Hypothesis 3b. The increase in consumers' perceived healthiness of healthier variants will be smaller for high health literate consumers.

If as result of displaying a MTL label and by H1a and H1b the suggested changes in perceived healthiness do take place, the question whether these changes in perceived healthiness consequently lead to actual intended changes in consumer behaviour comes up. Borgmeier &

Westenhoefer (2009) found that the MTL labelling does indeed induce changes in perceived

healthiness, but that these changes in perceived healthiness are unlikely to influence consumer

behaviour in terms of food choice or consumption.

In the study performed by Feunekes et al., (2008), participants tended to slightly increase their consumption of healthier products and decrease their consumption of less healthy products after being exposed to FOP labels. Whether these changes were the result of changes in health perceptions was not investigated. Based on the findings of Borgmeier & Westenhoefer (2009), these changes in behavioural intention shown by Feunekes et al., (2008) are unlikely to be caused by the changes in perceived healthiness, but again this was not investigated. This study does want to investigate this possible effect, which leads to the following hypotheses:

Hypothesis 4a. Consumers will decrease their intended usage frequency of the less healthy variants as a result of a decrease in their perceived healthiness.

Hypothesis 4b. Consumers will increase their intended usage frequency of the healthier variants as a result of an increase in their perceived healthiness.

4. METHODS 4.1 Participants

To acquire data for the statistic analysis, a survey was developed and distributed online. In total, 168 people responded to the survey. For socio-demographic characteristics see (Table 1).

Table 1.

Socio-demographic characteristics

(n = 168) Gender % Age % Education level %

Male 64.9 18-24 years 39.3 University 70.2

Female 35.1 25-34 years 14.3 Higher vocational education 28.0

35-44 years 4.2 Intermediate vocational education 1.2

45-54 years 10.7 High school 0.6

55-64 years 20.2

65+ years 11.3

4.2 Materials

In this study, there is chosen for the most well-known traffic-light system, the UK Food Standard Agency (FSA) multiple-traffic-light system (Fig. 2. a-c). See appendix A. for labelling criteria as determined by the FSA.

a. b. c.

Fig 2. Food standard agency (FSA) multiple-traffic-light label used in the experiment: (a) Example of MTL label indicating green, yellow and red; (b) English version of the label used in the survey for the less healthy beverages; (c) Dutch version of label (b).

4.3 Products

Four product categories were included in the experiment. The critical target categories were: soda and fruit juice, the other two categories (e.g. bread and cheese) were added as filler categories. For each category a less healthy and a healthier variant was selected. For soda, the less healthy variant was ''coke regular'' and the healthier variant was ''diet coke''. For fruit juice, the less healthy variant was ''orange juice'' and the healthier variant was ''orange juice light''. For bread, the less healthy variant was ''brown bread'' and the healthier variant was ''whole grain bread'', this might not be what one expects, however, based on the labelling criteria by the (FSA), whole grain bread is the healthier choice. For cheese, the less healthy variant was ''full-fat cheese'' and the healthier variant was ''low-fat cheese''. To avoid the effect of brand preference all products were presented unbranded.

4.4 Procedure

The study was conducted online and the data were collected through a survey. A similar

procedure to that of Feunekes et al., (2008) was performed. All participants were exposed to all

four product categories. Each category consisted of two products, a less healthy and a healthier

product (product pair). The order of presentation of the product categories was randomized.

To obtain baseline measures, participants were first exposed to four pairs of pictures of products, one pair after the other. Each pair consisted of a picture of a less healthy variant and a picture of a healthier variant, side by side. Below the picture of each product, standard back-of- pack information was displayed, but no front-of-pack MTL label was presented. Below the picture of each product, participants were asked to indicate their current usage frequency and perceived healthiness of the products.

They then had to answer questions regarding demographics (e.g. age, gender and education level) and whether they were allergic to certain foods or have a colour vision deficiency. After this, the participants were again exposed to the same four pairs of pictures, again in a randomized order, with standard back-of-pack information, but now, also an enlarged MTL label was displayed. While being exposed to the same products, but now with MTL label displayed, participants were asked for their ''intended'' usage frequency and again for their perceived healthiness. This was followed by some questions regarding health motivation and at the end, the participants were exposed to the (NVS-D) to assess their level of health literacy.

4.5 Measures

The current usage frequency (baseline measure) of the products without MTL label was measured by the question ''How often a month do you usually eat or drink this product?'' The current perceived healthiness (baseline measure) of the products without MTL label was measured by the question "How healthy is this product for you?'', with answers ranging from 1 (not healthy at all) to 5 (very healthy).

The intended usage frequency (post measure) of the products with MTL label was measured by the question ''Seeing this product with the health indicator, how often a month do you intend to use this product now? The perceived healthiness of products with MTL label (post measure) was measured by the question ''Seeing this product with health indicator, how healthy is this product for you now?'', with answers ranging from 1 (not healthy at all) to 5 (very healthy).

Intended change in usage frequency (behavioural intention) was measured by calculating the

difference between baseline usage frequency and intended usage frequency after being exposed to

the MTL label. The impact of the MTL label on consumers' perceived healthiness, was measured

by calculating the differences in perceived healthiness between baseline and post measure.

To assess participants’ level of health literacy, the (NVS-D) was used, this is the Dutch version of the Newest Vital Sign (NVS), ''a 6-question tool to assess an individual’s ability to find and interpret information on an ice cream nutrition label'' (Weiss et al., 2005). The (NVS-D) was translated, tested and validated by (Fransen, van Schaik, Twickler, & Essink-Bot, 2011). See appendix B for specific questions. Participants who correctly answered 0-2 questions were categorized as health literacy level 1, participants who correctly answered 3-4 questions were categorized as level 2 and participants who correctly answered 5-6 questions were categorized as health literacy level 3.

4.6 Control variables

The demographic characteristics (e.g. age, gender and education level), were added as control variables.

5. RESULTS

A 2 (MTL Label: without MTL label, with MTL label) x 4 (Product category: Soda, Fruit juice, Bread, Cheese) x 2 (Healthiness of product: Less healthy variant, Healthier variant) within subject factorial design was used. For the ANOVAs the significance level was set at .05. The control variables were added as independent variables.

To determine the effect of displaying a MTL label on consumers' perceived healthiness, a repeated measures ANOVA was conducted with consumers' level of health literacy as between subjects factor. The main effect of displaying a MTL label was significant for all beverages (Table 2.) For the less healthy soda F(1, 165) = 15.25, p < .001, the less healthy fruit juice F(1, 165) = 2.85, p < .10, the healthier soda F(1, 165) = 14.80, p < .001 and for the healthier fruit juice F(1, 165) = 10.63, p < .01.

The participants rated the less healthy variant of soda without label as slightly unhealthy (M = 1.53) and the less healthy variant of fruit juice without label as neutral (M = 3.18). Displaying the exact same MTL label (Fig 2. b) decreased the perceived healthiness of the less healthy fruit juice (M

= -0.35), but increased the perceived healthiness of the less healthy soda (M

= 0.13).

The healthier variant of soda without label was rated as slightly unhealthy (M = 1.77), the

healthier variant of fruit juice without label as neutral (M = 2.89). Displaying a MTL label

increased consumers' perceived healthiness of the healthier variants, for the healthier soda (M

= 0.43) and the healthier fruit juice (M

= 0.31).

Fig 3. Mean difference scores for perceived healthiness per beverage type between baseline (without MTL label) and post measure (with MTL label).

Hence, based on the results H1a cannot be confirmed, displaying a MTL label decreased consumers' perceived healthiness of the less healthy of fruit juice, but increased consumers' perceived healthiness of the less healthy soda. Therefore, H2a can also not be confirmed, a bigger decrease in perceived healthiness was expected for the less healthy fruit juice than for the less healthy soda, but consumers' perceived healthiness of the less healthy soda did not decrease but increase (Fig 3).

H1b however can be confirmed, displaying a MTL label increased consumers' perceived healthiness of the healthier variants. H2b can also be confirmed, the increase in consumers' perceived healthiness of the healthier variants was stronger for the healthier soda than for the healthier fruit juice (Fig 3).

The main effect for consumers' level of health literacy was only significant for the less

healthy variant of soda F(2, 165) = 3.65, p < .05, (Table 2). The interaction effect between the

MTL label and consumers' level of health literacy was not significant for the healthier variants,

but was significant for the less healthy variants (Table 2). For the less healthy soda, F(2, 165) =

4.83, p < .01 and for the less healthy fruit juice, F(2, 165) = 3.39, p < .05.

Table 2.

Mean scores and ANOVAs for perceived healthiness without and with displaying a MTL label and consumers' level of health literacy.

Beverages ANOVAs

Without MTL label

With MTL label

MTL Label (df =1)

Health literacy (df = 2)

MTL label x Health literacy

(df = 2) (n = 168) (n = 168) (n = 168)

M M M

F(p) F(p) F(p)

Less healthy variant

Soda 1.53 1.66 0.13 15.25 (**) 3.65 (*) 4.83 (**)

Fruit juice 3.18 2.83 -0.35 2.85 ( †) ^{0.67 (ns) 3.39 (*)}

Healthier variant

Soda 1.77 2.20 0.43 14.80 (**) 2.07 (ns) 0.62 (ns)

Fruit juice 2.89 3.20 0.31 10.63 (**) 0.57 (ns) 0.08 (ns)

† p < .10; * p < .05; ** p < .01.

M

mean difference in perceived healthiness between baseline and post measure.

In contrast with the expectations, participants overall increased their perceived healthiness of the less healthy soda after being exposed to the MTL label (M

= 0.13). For the high health literate participants (M

= 0.11), the medium health literate participants (M

= 0.06) and the low health literate participants (M

= 0.75) (Table 3).

Table 3.

Mean scores for perceived healthiness of the significant interaction effects between MTL label x Health literacy.

Level of health literacy

Less healthy soda Less healthy fruit juice

Without MTL label

With MTL label

Without MTL label

With MTL label

M M M

M M M

Low (n = 8) 1.75 2.50 0.75 2.50 2.75 0.25

Medium (n = 31) 1.52 1.58 0.06 3.16 2.84 -0.32

High (n = 129) Overall (n = 168)

1.52 1.53

1.63 1.66

0.11 0.13

3.23 3.18

2.83 2.83

-0.40

-0.35

M ^diff mean difference in perceived healthiness between baseline and post measure.

In line with the expectations, participants overall decreased their perceived healthiness of the less healthy fruit juice after being exposed a MTL label (M

= -0.35). For the high health literate participants (M

= -0.40), for the medium health literate participants (M

= -0.32). The low health literate participants however, increased their perceived healthiness (M

= 0.25) (Table 3).

Hence, based on the results H3a cannot be confirmed. The expectation was a smaller decrease in the perceived healthiness of the less healthy variants for the high health literate participants, but in contrast with the expectations, the expected decrease in perceived healthiness of the less healthy soda appeared to be an increase (H1a). The increase was however smaller for high health literate participants than for low health literate participants (Table 3). Consumers' perceived healthiness of the less healthy fruit juice did decrease, however, not for the low health literate participants who increased their perceived healthiness after being exposed to the MTL label (Table 3).

H3b can also not be confirmed, for the healthier variants, no significant interaction effect between the MTL label and health literacy was found (Table 2).

For the control variables age and education level no significant results were found, for the variable gender however, some differences were found. The main effect for gender on the perceived healthiness of the less healthy soda was significant, F(1, 165) = 5.96, p < .05, for the less healthy fruit juice this was, F(1, 165) = 13.13, p < .001, the healthier soda not significant and the healthier fruit juice was, F(1, 165) = 11.26, p < .01. The perceived healthiness of every beverage, less healthy or healthy, with or without MTL label was higher for men than for women, however apart from this, no consistent and interpretable pattern was found.

To determine whether displaying a MTL label had effect on behavioural intention through perceived healthiness, MEMORE was used. MEMORE is a macro for SPSS that estimates the total, direct and indirect effects of X on Y through one or more mediators M in a repeated measures design (Montoya & Hayes, 2017).

For the less healthy soda, the total effect of displaying the MTL label on behavioural

intention was not significant t(167) = -1.11, p = .27. For the less healthy fruit juice however, a

significant total effect was found. The current usage frequency of the less healthy fruit juice

without MTL label (M=4.41, SD=6.72) and the intended usage frequency with MTL label

(M=3.84, SD=6.09) conditions; t(167) = 2.01, p < .05. The direct effect however, was not

significant t(165) = 0.94, p = .35 and neither was the indirect effect, b = 0.29, SE = 0.24, 95% CI [-0.11, 0.83]. Here the CI includes 0, indicating that the indirect effect is not significant and thus that the MTL label not influenced behavioural intention through consumers’ perceived healthiness.

Also for the healthier soda, the total effect was not significant t(167) = -1.51, p = .13. The total effect for the healthier fruit juice however was significant, the current usage frequency of the healthier fruit juice without MTL label (M=0.29, SD=1.18) and the intended usage frequency with MTL label (M=1.16, SD=2.87) conditions; t(167) = -4.15, p < .001, the direct effect was also significant t(165) = -3.043, p < .01, but the indirect effect was not, b = -0.20, SE = 0.14, 95% CI [-0.53, 0.03]. Again the CI includes 0, meaning a not significant indirect effect and thus, that the MTL label not influenced behavioural intention through consumers’ perceived healthiness.

Hence, based on the above results, H4a and H4b cannot be confirmed. Displaying a MTL label did not induce significant intended changes in the usage frequency of the less healthy or healthier soda. Displaying a MTL label did induce intended changes in the usage frequency of the less healthy and healthier fruit juice, but the changes in behavioural intention cannot be directly attributed to changes in perceived healthiness.

Table 4. Overview results

Hypothesis Accepted / Rejected

1a Exposure to a MTL label will decrease consumers' perceived healthiness of the less healthy variants.

Rejected

1b Exposure to a MTL label will increase consumers' perceived

healthiness of the healthier variants. Accepted

2a The decrease in consumers' perceived healthiness of the less healthy

variant will be bigger for fruit juice than for soda. Rejected 2b The increase in consumers' perceived healthiness of the healthier

variant will be will bigger for soda than for fruit juice.

Accepted

3a The decrease in consumers' perceived healthiness of the less healthy

variants will be smaller for high health literate consumers. Rejected 3b The increase in consumers' perceived healthiness of the healthier

variants will be smaller for high health literate consumers.

Rejected

4a Consumers will decrease their intended usage frequency of the less

healthy variants as a result of a decrease in their perceived healthiness. Rejected 4b Consumers will increase their intended usage frequency of the

healthier variants as a result of an increase in their perceived healthiness.

Rejected

6. DISCUSSION

The main objective of this study was to investigate the impact of the MTL label on consumers' perceived healthiness and behavioural intention. While not all hypotheses have been confirmed conclusively, the investigation has led to better insights on the effectiveness of MTL labelling.

The results suggest that in some instances displaying a MTL label can have significant effects on consumers' behavioural intention, however, none of the changes in behavioural intention could be directly attributed to changes in consumers' perceived healthiness.

More interesting were the results on the influence of the MTL label on consumers' perceived healthiness. Displaying a MTL label had a significant effect on consumers' perceived healthiness of all beverages, but its impact was not always desirable. Displaying a MTL label ''did'' have a desirable effect on the healthier variants and the less healthy fruit juice, but had an undesirable effect on the less healthy soda. Displaying the exact same MTL label (Fig. 2b) led to a desired decrease in the perceived healthiness of the less healthy fruit juice, but to an undesired increase in the perceived healthiness of the less healthy soda.

The results also suggest that especially the low health literate participants, those who should benefit the most from FOP labels, had difficulties with interpreting the MTL label. These participants tended to misinterpret the MTL label on the less healthy varieties of soda and fruit juice, increasing their perceived healthiness when the desired response to the MTL label was a decrease. These undesired responses from especially the low health literate participants are however explainable. Even on the MTL labels of the less healthy beverages, preliminary green lights are displayed and as Kleef & Dagevos (2015) stated: ''The more green lights on the product, the healthier the choice'', more green than red lights does however ''not'' mean a healthy choice.

The high health literate consumers ''do'' have improved health knowledge (Nutbeam, 2000) and might therefore be better able to acknowledge that some of the green lights on the MTL labels of unhealthy products are not relevant for certain product categories and that these green lights therefore should be ignored when trying to make a healthy food choice. Low health literate consumers on the other hand might not be able to do this, and might therefore respond in a different and undesired way to the same MTL label.

Based on these findings, one could indeed say that in some instances, when multiple in

colour varying lights are displayed simultaneously, the MTL label could lead to confusion and

ambiguity (Hagen, 2010) and that this is especially a problem for the low health literate consumers.

Using labelling formats such as the NuVal system, the choices label or even the STL label would reduce ambiguity, but then the high health literate consumers might complain that they want to know what the simplified information stands for and how the single light or rating has been arrived (Grunert et al., 2010). Clearly the parties that design and implement the labels should make a decision on which group of consumers they want to serve best.

Another important question that comes up with the results, is how desirable it actually is that the MTL label increases the perceived healthiness of the healthier variants. This is because the healthier variants might be ''healthier'', but not per se healthy. Previous literature has indicated that people overeat more easily when products are labelled as low fat or healthy (Provencher, Polivy, & Herman, 2009; Wansink & Chandon, 2006). This negative side effect of nutrition labelling might however be the most relevant for the ''healthier'' sodas, where the sugar is often substituted for artificial sweeteners, such as aspartame. On the MTL labels of these beverages only green lights are displayed and some consumers might therefore think that the consumption of these ''healthier'' light variants is harmless, but, accumulating evidence now suggests that frequent consumption of these artificial sweeteners also leads to increased of weight gain, metabolic syndrome, type 2 diabetes and cardiovascular disease (Swithers, 2013).

In conclusion, the results indicated that in some instances displaying a MTL can have a

significant effect on consumers' behavioural intention, but the influence of the MTL label on

behavioural intention was not through consumers' perceived healthiness. More interesting were

the findings on the influence of the MTL label on consumers' perceived healthiness. Displaying a

MTL label did have influence on consumers' perceived healthiness of beverages, but its impact is

not always desirable and sometimes even its desirable impact is questionable. The impact of the

MTL does differ between the beverage types and also between the types of consumers in terms of

health literacy. When the MTL label displays multiple varying coloured lights, which is not

uncommon, it can lead to ambiguity and confusion. This ambiguity appears to be mainly

problematic for the low health literate consumers. While the MTL label might be promoted by

many as the most effective FOP format, it might not be the most suitable FOP format for the

beverage category. To determine what is the most suitable FOP format for the number one energy

contributor, further research is needed. Since it appears to be difficult to serve all types of

consumers, managers and future researchers should start with making a decision on which group of consumers they want to serve best.

7. LIMITATIONS

Although this research has provided some valuable insights into the impact of the MTL

label on consumers' perceived healthiness and behavioural intention, there are some limitations

that should be addressed. First, for each beverage type, less healthy or healthier, only one product

has been tested. To generalize the results, more products should be tested. Another limitation that

should be addressed is the sample size and its diversity, as a total of 168 participants responded to

the survey of which most were university educated. To be able to generalize the results, a bigger

and more diverse sample should be used.

REFERENCES

Balcombe, K., Fraser, I., & Falco, S. Di. (2010). Traffic lights and food choice: A choice experiment examining the relationship between nutritional food labels and price. Food Policy, 35(3), 211–220. https://doi.org/10.1016/j.foodpol.2009.12.005

BEUC. (2006). Discussion group on simplified labelling: Final report.

Block, G. (2004). Foods contributing to energy intake in the US: Data from NHANES III and NHANES 1999-2000. Journal of Food Composition and Analysis, 17(3–4), 439–447.

https://doi.org/10.1016/j.jfca.2004.02.007

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(1), 184. https://doi.org/10.1186/1471-2458-9-184

Byrd-Bredbenner, C., Wong, A., & Cottee, P. (2000). Consumer Understanding of US and EU nutrition labels. British Food Journal, 102(April), 615–629.

https://doi.org/10.1108/00070700010348505

Carbone, E. T., & Zoellner, J. M. (2012). Nutrition and health literacy: A systematic review to inform nutrition research and practice. Journal of the Academy of Nutrition and Dietetics.

https://doi.org/10.1016/j.jada.2011.08.042

Cecchini, M., & Warin, L. (2016). Impact of food labelling systems on food choices and eating behaviours: A systematic review and meta-analysis of randomized studies. Obesity Reviews, 17(3), 201–210. https://doi.org/10.1111/obr.12364

Cesare, M. Di, Bentham, J., Stevens, G. A., Geleijnse, J. ., & Kromhout, D. (2016). Trends in adult body-mass index in 200 countries from 1975 to 2014: A pooled analysis of 1698 population based measurement studies with 192 million participants (The Lancet (2016) 387 (1377-96)). The Lancet. https://doi.org/10.1016/S0140-6736(16)30469-X

Cowburn, G., & Stockley, L. (2005). Consumer understanding and use of nutrition labeling: a systematic review. Public Health Nutr, 8(1), 21–28. https://doi.org/10.1079/PHN2004666 Dennison, B. a, Rockwell, H. L., & Baker, S. L. (1997). Excess fruit juice consumption by

preschool-aged children is associated with short stature and obesity. Pediatrics, 99(1), 15–

22.

Faith, M. S., Dennison, B. A., Edmunds, L. S., & Stratton, H. H. (2006). Fruit Juice Intake Predicts Increased Adiposity Gain in Children From Low-Income Families: Weight Status- by-Environment Interaction. PEDIATRICS, 118(5), 2066–2075.

https://doi.org/10.1542/peds.2006-1117

Feunekes, G. I. J., Gortemaker, I. A., Willems, A. A., Lion, R., & van den Kommer, M. (2008).

Front-of-pack nutrition labelling: Testing effectiveness of different nutrition labelling formats front-of-pack in four European countries. Appetite, 50(1), 57–70.

Fiske, S. T., & Neuberg, S. L. (1990). A Continuum of Impression Formation, from Category- Based to Individuating Processes: Influences of Information and Motivation on Attention and Interpretation. Advances in Experimental Social Psychology, 23(C), 1–74.

https://doi.org/10.1016/S0065-2601(08)60317-2

Foltran, F., Verduci, E., Ghidina, M., Campoy, C., Jany, K.-D., Widhalm, K., … Gregori, D.

(2010). Nutritional profiles in a public health perspective: A critical review. The Journal of International Medical Research, 38(2), 318–85.

https://doi.org/10.1177/147323001003800202

Gill, J. M. R., & Sattar, N. (2014). Fruit juice: Just another sugary drink? The Lancet Diabetes and Endocrinology. https://doi.org/10.1016/S2213-8587(14)70013-0

Grunert, K. G., & Wills, J. M. (2007). A review of European research on consumer response to nutrition information on food labels. Journal of Public Health.

https://doi.org/10.1007/s10389-007-0101-9

Grunert, K. G., Wills, J. M., & Fernández-Celemín, L. (2010). Nutrition knowledge, and use and understanding of nutrition information on food labels among consumers in the UK.

Appetite, 55(2), 177–189. https://doi.org/10.1016/j.appet.2010.05.045

Hagen, K. (2010). Nutritional Information : Traffic Light Labelling is the Best Way to Reach Consumers. DIW Weekly Report, 6(19), 141–151.

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(3), 430–

439. https://doi.org/10.1017/S1368980012000754

Higginson, C. S., Rayner, M. J., Draper, S., & Kirk, T. R. (2002). The nutrition label – which information is looked at? Nutrition & Food Science, 32(3), 92–99.

https://doi.org/10.1108/00346650210423392

Institute of Medicine, C. on F. M. and the D. of C. and Y. (2006). Food Marketing to Children and Youth: Threat or Opportunity? National Academies Press.

Jordan Lin, C. T., Lee, J. Y., & Yen, S. T. (2004). Do dietary intakes affect search for nutrient information on food labels? Social Science and Medicine.

https://doi.org/10.1016/j.socscimed.2004.02.030

Kim, S. Y., Nayga, R. M., & Capps, O. (2000). The Effect of Food Label Use on Nutrient Intakes: An Endogenous Switching Regression Analysis. Journal of Agricultural and Resource Economics, 25(1), 215–231.

Kim, S. Y., Nayga, R. M., & Capps, O. (2001). Food label use, self-selectivity, and diet quality.

Journal of Consumer Affairs, 35(2), 346–363. https://doi.org/10.1111/j.1745- 6606.2001.tb00118.x

Kleef, E. Van, & Dagevos, H. (2015). The Growing Role of Front-of-Pack Nutrition Profile Labeling: A Consumer Perspective on Key Issues and Controversies. Critical Reviews in Food Science and Nutrition, 55(3), 291–303.

https://doi.org/10.1080/10408398.2011.653018

Kurtzweil, P. (1993). New food label: Good reading for good eating. FDA Consumer, 27(4).

Lobstein, T., Landon, J., & Lincoln, P. (2007). Misconceptions and misinformation: the

problems with Guideline Daily Amounts (GDAs). In A Review of GDAs and Their Use for Signalling Nutritional Information on Food and Drink Labels. London: National Heart Forum.

Malik, V. S., & Hu, F. B. (2015). Fructose and Cardiometabolic Health. Journal of the American College of Cardiology, 66(14), 1615–1624. https://doi.org/10.1016/j.jacc.2015.08.025 Montoya, A. K., & Hayes, A. F. (2017). Two-condition within-participant statistical mediation

analysis: A path-analytic framework. Psychological Methods, 22(1), 6–27.

https://doi.org/10.1037/met0000086

Nielsen, S. J., & Popkin, B. M. (2004). Changes in beverage intake between 1977 and 2001.

American Journal of Preventive Medicine. https://doi.org/10.1016/j.amepre.2004.05.005 Nutbeam, D. (2000). Health literacy as a public goal: a challenge for contemporary health

education and communication strategies into the 21st century. Health Promot Int, 15(3), 259–267. https://doi.org/10.1093/heapro/15.3.259

Nutbeam, D. (2008). The evolving concept of health literacy. Social Science & Medicine, 67(12), 2072–2078. https://doi.org/10.1016/j.socscimed.2008.09.050

Pereira, R. (2010). Understanding the maze of nutrition symbols and nutrition labeling systems.

Diabetes Spectrum. https://doi.org/10.2337/diaspect.23.1.57

Petty, R. E., Cacioppo, J. T., & Schumann, D. (1983). Central to The and Peripheral Role of Routes Involvement Advertising Moderating Effectiveness : Journal of Consumer Research, 10(2), 135–146. https://doi.org/10.2307/2488919

Provencher, V., Polivy, J., & Herman, C. P. (2009). Perceived healthiness of food. If it’s healthy, you can eat more! Appetite, 52(2), 340–344. https://doi.org/10.1016/j.appet.2008.11.005 Ratzan, S. C., & Parker, R. M. (2000). Introduction. National Library of Medicine Current

Bibliographies in Medicine: Health Literacy., 2000–1, v–vi.

https://doi.org/10.1017/CBO9781107415324.004

Rayner, M., Boaz, A., & Higginson, C. (2001). Consumer Use of Health-Related Endorsements On Food Labels in the United Kingdom and Australia. Journal of Nutrition Education, 33(1), 24–30. https://doi.org/10.1016/S1499-4046(06)60006-7

Roberto, C. A., Bragg, M. A., Schwartz, M. B., Seamans, M. J., Musicus, A., Novak, N., &

Brownell, K. D. (2012). Facts up front versus traffic light food labels: A randomized controlled trial. American Journal of Preventive Medicine, 43(2), 134–141.

https://doi.org/10.1016/j.amepre.2012.04.022

Rothman, R. L., Housam, R., Weiss, H., Davis, D., Gregory, R., Gebretsadik, T., … Elasy, T. A.

(2006). Patient Understanding of Food Labels. The Role of Literacy and Numeracy.

American Journal of Preventive Medicine, 31(5), 391–398.

https://doi.org/10.1016/j.amepre.2006.07.025

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(4),

344–352. https://doi.org/10.1093/heapro/dap032

Schor, D., Maniscalco, S., Tuttle, M. M., Alligood, S., & Kapsak, W. R. (2010). Nutrition facts you can’t miss: The evolution of front-of-pack labeling. Nutrition Today, 45(1), 22–32.

Scott, V., & Worsley, A. F. (1994). Ticks, claims, tables and food groups: A comparison for nutrition labelling. Health Promotion International, 9(1), 27–37.

https://doi.org/10.1093/heapro/9.1.27

Shine, A., O’Reilly, S., & O’Sullivan, K. (1997). Consumer attitudes to nutrition labelling.

British Food Journal, 99(8), 283–289. https://doi.org/10.1108/00070709710188381

Sonnenberg, L., Gelsomin, E., Levy, D. E., Riis, J., Barraclough, S., & Thorndike, A. N. (2013).

A traffic light food labeling intervention increases consumer awareness of health and healthy choices at the point-of-purchase. Preventive Medicine, 57(4), 253–257.

https://doi.org/10.1016/j.ypmed.2013.07.001

Swinburn, B. A., Sacks, G., Hall, K. D., McPherson, K., Finegood, D. T., Moodie, M. L., &

Gortmaker, S. L. (2011). The global obesity pandemic: Shaped by global drivers and local environments. The Lancet, 378(9793), 804–814. https://doi.org/10.1016/S0140-

6736(11)60813-1

Swithers, S. E. (2013). Artificial sweeteners produce the counterintuitive effect of inducing metabolic derangements. Trends in Endocrinology and Metabolism.

https://doi.org/10.1016/j.tem.2013.05.005

Steenhuis, I. H. M., Kroeze, W., Vyth, E. L., Valk, S., Verbauwen, R., & Seidell, J. C. (2010).

The effects of using a nutrition logo on consumption and product evaluation of a sweet pastry. Appetite, 55(3), 707–709. https://doi.org/10.1016/j.appet.2010.07.013

Thorndike, A. N., Sonnenberg, L., Riis, J., Barraclough, S., & Levy, D. E. (2012). A 2-phase labeling and choice architecture intervention to improve healthy food and beverage choices.

American Journal of Public Health, 102(3), 527–533.

https://doi.org/10.2105/AJPH.2011.300391

van Kleef, E., van Trijp, H., Paeps, F., & Fernández-Celemín, L. (2008). Consumer preferences for front-of-pack calories labelling. Public Health Nutrition, 11(2).

https://doi.org/10.1017/S1368980007000304

Velardo, S. (2015). The Nuances of Health Literacy, Nutrition Literacy, and Food Literacy.

Journal of Nutrition Education and Behavior, 47(4), 385–389.

https://doi.org/10.1016/j.jneb.2015.04.328

Wandel, M. (1997). Food labelling from a consumer perspective. British Food Journal.

https://doi.org/10.1108/00070709710181559

Wang, Y. C., Bleich, S. N., & Gortmaker, S. L. (2008). Increasing Caloric Contribution From Sugar-Sweetened Beverages and 100% Fruit Juices Among US Children and Adolescents, 1988-2004. PEDIATRICS, 121(6), e1604–e1614. https://doi.org/10.1542/peds.2007-2834 Wansink, B., & Chandon, P. (2006). Can “Low-Fat” Nutrition Labels Lead to Obesity? Journal

of Marketing Research, 43(4), 605–617. https://doi.org/10.1509/jmkr.43.4.605

Welsh, J. A. (2005). Overweight Among Low-Income Preschool Children Associated With the Consumption of Sweet Drinks: Missouri, 1999-2002. PEDIATRICS, 115(2), e223–e229.

https://doi.org/10.1542/peds.2004-1148

World Health Organization. (2004). Global Strategy on Diet, Physical Activity and Health.

Geneva, Switzerland, WHO.

APPENDICES Appendix A. Labelling criteria as determined by the FSA

Criteria for drinks per 100ml

Criteria for Food per 100g

Appendix B. Survey questions

Intro Beste deelnemer, Allereerst wil ik u hartelijk danken voor uw deelname aan dit onderzoek.

Mijn naam is Boudewijn van den Berg. Voor mijn MSc Marketing aan de Rijksuniversiteit Groningen doe ik onderzoek naar consumentengedrag. Om hier inzicht in te krijgen vraag ik u daarom een aantal vragen te beantwoorden.

Het onderzoek zal ongeveer 5 minuten van uw tijd in beslag nemen. Er zal betrouwbaar met uw gegevens worden omgegaan en de resultaten worden geheel anoniem verwerkt. Nogmaals hartelijk dank voor uw deelname aan dit onderzoek. Met vriendelijke groet,

Boudewijn van den Berg

Cola regular/light

Q1 Hoeveel keer per maand eet of drinkt u dit product normaal gesproken?