Inverting the pyramid! Extent and quality of food advertised on Austrian television

Food marketing is intended to have persuasive effects on children’s food preferences, food purchasing behavior and short-term food consumption [13, 14]. For instance, children who watched more TV responded to frequently advertised food items more readily than those who watched less [15, 16]. Even brief exposure to food advertising influenced children’s food preferences [17]. As shown by Harris, Bargh and Brownell [1], elementary-school-aged children consumed 45 % more snacks when exposed to food advertisements. In adults, this effect was dissociated of the participants’ reported hunger, indicating that this particular snacking behavior is guided by automatic and habitual processes [18]. Repetitive exposure to food products as food primes can enhance the effect on shaping eating habits (mere-exposure effect [19]) through the Pavlovian stimulus–outcome (S-O) conditioning, therefore eliciting short-term food cravings [20].

Children as young as three years old are capable to identify advertisements during TV programs but, unlike adults, are not able to understand the persuasiveness of advertisements and their short- and long-term effects on health [21]. This makes food marketing targeted at children a deceptive way to propose commercial transactions to children and should therefore be restricted [22]. Consquently, international authoritites, including World Health Organization (WHO) member states, called for a general agreement to harmonize food and beverage advertisements within national dietary guidelines of the Global Strategy on Diet, Physical Activity and Health (2004). A systematic review by Galbraith-Emami and Lobstein [23] showed that worldwide initiatives are effective in restricting advertisement on-air time for children, but the effectiveness depends highly on audience definition and nutrient profile criteria. In some countries, limiting advertisement of energy-dense, nutrient-poor food and beverages is implemented in their national action plans on nutrition [24]. Inspite of measures taken, global food marketing, especially unhealthy food for children, appear to be on the rise since first regulatory actions have been initiated [25, 26].

In the absence of complete bans of food advertisements, there is a need aimed at a comprehensive and consistent approach across countries to regulate the quantity and quality of the advertised content. International regulatory guidelines have engaged different models, such as nutrient or category-based approaches. In Europe, the EU Pledge Nutrition Criteria are conceptualized to restrict food and beverage advertisements to children younger than 12 years old on TV, print media and on the internet [27]. It is designed as a self-regulatory and voluntary intitative including 20 companies being responsible for at least 80 % of the TV food marketing in the EU. The self-regulatory pledged criteria use a dual strategy to encourage the advertisement of healthy food components (e.g. high fiber and protein content) and simultaneously aim to restrict unhealthy food components (e.g. high amounts of sugar, saturated fatty acids). The voluntary commitment started in 2012 and participating members pledged to implement the suggested guidelines by 31st December 2014.

The EU Pledge Nutrition Criteria use 9 different food categories to arrange food advertisements. In short, food groups, such as soft drinks, or sugar and sugar-based products, are fully restricted for advertisement. Food products that promote a particular food within a certain category have to contain category-specific components per defintion (e.g. for dairy products to contain at least 50 % dairy), otherwise they should not be advertised. Additionally, depending on the presented food category, nutrient-based thresholds are defined for energy (kcal/portion), sodium (mg/100 g or 100 ml), saturated fats (mg/100 g or 100 ml) and total sugars (g/100 g or 100 ml).

Additional to nutrient-based thresholds, certain food components should be encouraged (nutrients and food groups). Desirable food components are qualified as positive and should be encouraged: such as fibre and whole grain in cereal-based products; protein and calcium in milk and dairy products; protein in meat, fishery products; polyunsaturated fatty acids (PUFAs) in vegetable oils, spreads and fats.

The present study was conducted before the EU Pledge Nutrition Criteria came into full effect (prior to 12/2014) to provide a status quo analysis regarding food advertisements aired on Austrian TV. In detail, we present data from advertisements, displaying frequency and the nutritional quality of the advertised food products. The aim of this study is threefold: we (i) analyze a set of advertisements broadcasted for children on six different TV stations. We split the advertised food into eight food categories based on the Austrian dietary guidelines as defined by the Austrian Food Guide Pyramid [28]; (ii) we provide data regarding the nutritional quality of food advertisements targeted at children by comparing the displayed food with the pledged EU Nutrition Criteria, and (iii) we discuss the results and their implications in a broader context of public health nutrition.

Data collection took place in February and March 2014. TV programs were recorded from six seperate TV channels over four days (two non-consecutive weekdays and two independent days on the weekend). TV channels were selected based on media analysis of children’s TV habits in Austria, Germany and Switzerland [29], choosing the six most popular TV channels for children (ATV, ORF1, Pro7, RTL, Sat1 and SuperRTL). Recording period for each dedicated day was from 6 am to 9 pm. Recording time was further split into three time periods: morning hours (6 am – 11 am), midday hours (11 am – 4 pm) and evening hours (4 pm – 9 pm).

In total, 360 h of broadcast material was recorded, spread equally over all six TV stations. We recorded full days of TV screening (6 am – 9 pm) to get a complete and consistent dataset of recordings. Previous studies reported segmented recording times (e.g. morning, after-school hours) [30] or solely recorded kids programs [31, 32] which may have led to incomplete datasets for the analysis. In this study, full days from 6 am to 9 pm and all types of programs were recorded to reduce the chance of missing any data. All stations broadcasted their program in German.

First, the content of the TV program was divided into non-programs and programs (e.g. TV shows, news updates). Non-programs included promotion of station programs, station identification, and product advertisements. To identify advertisements, we used a coding protocol established by Thompson et al. [33]. Brief sponsorship messages, such as “this program was brought to you by product X”, brand display or buying recommendations, were used to identify advertisements [33]. Hence, product advertisements were divided into food or non-food advertisements. As we only focus on food advertisements in this study, we did not record the percentage of food advertisements compared to non-food advertisements. Food advertisements were considered as such when they promoted the purchase or consumption of food or beverages. In a second step, we identified if food advertisements were targeted at children. This was conducted according to a dichotomous categorization scheme adapted from Chapman et al. [34] (Table 1).

The video material was reviewed by two coders fluent in German (author 2 and 3). Initial coding of the video material was performed by one coder, a second coder was given the same coding form and instructions to code a 10 % sample of the total duration of recordings (36 h). Inter-coder reliability was calculated using the following formula: number of agreements*100 / number of disagreements. This reliability check was only performed for audience orientation and not for the food categorization procedure.

Inter-coder reliability was 95.8 % in the present study. All discrepancies emerging during the process were discussed within the research team.

Initially, to identify food categories depicted in the advertisements, we categorized food according to the national dietary recommendations. The Austrian Food Guide Pyramid is described in seven categories [28]. According to the recommendations, items from each category should be eaten at different frequencies per day (Table 2). In addition, we added one extra category called ‘other food not further specified’ to complement the categories of the Austrian Food Guide Pyramid. All discrepancies emerging during the categorization process were discussed within the research team and conflicting food categorizations were resolved in consensus. Hence, we analyzed the video material according to following eight food categories:

We assessed the on-air frequency of the advertisements by counting the total number of advertisements. We interposed this step to analyze the dataset according to the Austrian Food Guide Pyramid to assess how the foods depicted on TV match with the national dietary recommendations.

In a second step, the displayed products were analyzed according to the EU Pledge Nutrition Criteria. This was conducted to assess how the foods depicted on TV match with the EU Pledge Nutrition Criteria. According to the EU Pledge Nutrition Criteria 9 different food categories are defined. For this we re-coded the displayed advertisements accordingly and conducted nutrient profiling analysis with those foods initially passing the category criteria (exclusion criteria: food groups such as soft drinks, sugar and sugar-based products and misleadingly declarated foods; see supplementary material, Additional file 1).

For further analysis, the nutritional information of the food products was obtained directly from the nutrition information on the label of the promoted products during supermarket visits or online. When nutrition information was not readily available, manufacturers were contacted. The data was then compared with the nutrient threshold limits and the food components that should be encouraged as defined by the EU Pledge Nutrition Criteria by the first and second author of this study.

In our study we show that one in seven food advertisement was targeted at children, with 92.4 % displaying fatty, sweet and salty snacks. In Austria, the average TV viewing time, from the age of three years on, is approximately 158 min per day [39]. TV watching in excess of 120 min and longer is associated with reduced physical and psychosocial health, and a large body of evidence suggests that especially decreasing sedentary time in youth aged 5–17 years leads to reductions in Body-Mass Index and health risks [40]. The reduction of sedentary lifestyles and promotion of early life nutritional education to strengthen overall self-regulatory resources may be key to responsible eating behavior. This has been addressed in a dual-model on a population-based childhood obesity prevention program provided by the WHO [41]. Regulating food marketing for children may be one effective measure to decrease food exposure to children [23]. Our results support the idea that food advertisement seem to distort national dietary recommendations, especially when the recipients of the advertisements are children (e.g. 17 times more likely for fatty, sweet and salty snacks). In Austria, for instance, five servings of fruits or vegetables per day are recommended, none of them were displayed in our study, although the National Action Plan for nutrition explicitly stresses the promotion of healthy food, such as fruits [42].

In the present study we investigated food advertisements by means of their eligibility within the EU Pledge Nutrition Criteria. There is a political alignment accompanied by voluntary codes of conduct by some food companies [43]. A preliminary and post-implementation study evaluating the self-regulatory Children’s Food and Beverage Advertising Initiative in Canada (2006–2011) came to the conclusion, that while the volume of aired advertisements for children decreased by 24 %, the advertisement quality decreased likewise. They showed that compared to 2006, food classified as less healthy were increasingly targeted at children (+47 %) and teenagers (+161 %) after implementing the initiative. The authors concluded that, in Canada, the initiated self-regulatory system designed to protect Canadian children from food advertising showed clear weaknesses and advertisement should be regulated more strictly [44]. A recent systematic review of evidence by Ronit and Jensen [45] provides an overview concerning research on industry self-regulation regarding food and beverage marketing and nutrition labeling. The authors conclude, that although methodological heterogenity was prevalent in the 22 reviewed articles, the ineffectiveness of existing self-regulation schemes is univocal, calling for more legislative guidelines. In addition, the use of commonly persuasive techniques, such as premium offers, promotional characters, nutrition and health-related claims, the theme of taste, and the emotional element of fun, are frequently displayed in the endorsement of food for children [46]. Especially promotional characters, like cartoon characters, have shown to influence food preferences in children, but branding has mainly been used for energy-dense and nutrient-poor foods (e.g. cookies, candy or chocolate), as compared to fruits or vegetables [47]. In the EU Pledge Nutrition Criteria, no restrictions for persuasive marketing techniques (mascots, cartoon characters) are implemented. Considering this and the findings of our study, that 65.9 % of the advertised food for children originate from member companies of the EU Pledge Nutrition Criteria and 97.9 % of their advertised foods showed nonconformity aspects, an argument for stricter regulations may be put forward.

As part of a new wave in public health improvements, the promotion of healthy food and nutrient components as proposed by the EU Pledge Nutrition Criteria would fit well into the new’cultural turn’ in the field of public health as proposed by Davies et al. [48]. Inarguably, to further minimize influences towards unhealthy behavior, it is necessary to both maximize the value of health and the promotion of healthy choices as default and to minimize factors that create an environment of unhealthy behaviors [49].

One major limitation of the present study is that we did not screen for seasonal differences and only presented data from a 2 month time frame (February and March 2014). Nevertheless, our results showed comparable results and displaying frequencies, which have been shown in other studies as well. A second limitation is more general and aimed at criticizing the definition of nutritional criteria. Different nutrient profiling strategies have been developed to ban advertisements targeted at children. Throughout this process, the question what is healthy food and what is not, proved to be difficult. To support this argument, Scarborough et al. [12] showed that when using eight different nutrient profile models on a dataset of food advertisements, the percentage of permitted food advertisements varied from 2.1 to 47.4 %. Although the authors used diverging models, the study shows that advertisment restraint highly depends on the applied restriction model. From this point of view, defining nutrient criteria as basis for marketing restrictions have to be grounded on solid scientific findings, taking all available evidence into consideration. In contrast, Lobstein and Davies [11] argued that nutrient profiling as a method to categorize food according to nutritional quality is both feasible and practical and can support a number of public health-related initiatives. A third limitation of the study is that we cannot provide data about causal effects of TV advertisements on food habits and eating behavior. Our study design is therefore limited to a mere descriptive level.