Over the past two decades numerous studies have shown that obesity, and particularly childhood obesity, is increasing globally(Reference Lobstein, Baur and Uauy1). South Africa is a middle-income country which has one of the highest levels of obesity in the world. A recent national study found that 14·0 % of children aged 1–9 years, and 51·5 % of women aged 16–35 years, had a BMI≥85th percentile or a BMI≥25kg/m2, respectively(Reference Labadarios, Swart and Maunder2). This is much higher in any other sub-Saharan African country(Reference Lobstein, Baur and Uauy1). This problem has been attributed to numerous causes, including sedentary lifestyles, diet, socio-economic status, the physical environment and marketing associated with the media(Reference Guran and Bekeret3).
Television (TV) is a powerful medium for marketing and advertising products(Reference Guran and Bekeret3, Reference Dijkstra, Heidi and Buijtels4). There is evidence to suggest that greater TV watching is associated with increased consumption of snacks, sweetened beverages and fast foods(Reference Andreyeva, Kelly and Harris5–Reference Halford, Gillespie and Brown7). Andreyeva et al.(Reference Andreyeva, Kelly and Harris5) found that fast food and soft drink TV advertising was associated with increased consumption of both these items among elementary-school children. Furthermore, fast food advertising was significantly associated with BMI of overweight and obese children. Zimmerman and Bell(Reference Zimmerman and Bell6) and Halford et al.(Reference Halford, Gillespie and Brown7) found evidence to support the theory that TV viewing does not contribute to obesity because it is a sedentary activity, but because of children's responses to the advertisements themselves. Halford et al.(Reference Halford, Gillespie and Brown7) further suggested that exposure to food advertisements promotes consumption. To substantiate these studies, a local study by Da Fonseca(8) that aimed at exploring South African parents’ perceptions of TV food advertising directed at children revealed that parents perceive food advertising as a strong influence on their children's food preferences and eating habits. Parents believed that although the Advertising Standards Authority of South Africa's (ASA) food and beverage code Section II, Clause 4·1(9) discourages less healthful food promotional activity on TV directed at children younger than 12 years, it is ineffective in protecting children against advertisers. Moreover, numerous studies have documented the association between the number of hours/food advertisements on TV and the prevalence of obesity(Reference Andreyeva, Kelly and Harris5, Reference Coon and Tucker10, Reference Goris, Petersen and Stamatakis11). TV advertising contributes to childhood obesity by an estimated 16–40 % in the USA, 10–28 % in Australia and Italy and 4–18 % in Britain(Reference Goris, Petersen and Stamatakis11).
In the USA in 2009, children aged 2–11 years watched an average of 10·9 to 12·7 food-related TV advertisements daily, and of these advertisements 86 % were for products high in saturated fat, sugar or sodium(Reference Coon and Tucker10). In 2008 a global study of thirteen research groups in Australia, Western Europe, North and South America and Asia evaluated TV advertisements shown during children's viewing times(Reference Goris, Petersen and Stamatakis11). They found that food advertisements comprised 11–29 % of all advertisements and that 53–87 % of the food advertisements were for foods high in undesirable nutrients and energy(Reference Goris, Petersen and Stamatakis11). Similarly, in Turkey, 32·1 % of all TV advertisements were related to foods of which 81 % were classified as being unhealthy(Reference Guran and Bekeret3, Reference Guran, Turan and Akcay12). The findings of these and numerous similar studies reveal that people, especially children, are exposed to many TV food advertisements, most of which are for unhealthy foods.
There has been virtually no research into food advertising on TV in Africa. In a small study carried out in South Africa, Temple et al.(Reference Temple, Steyn and Ndamane13) reported that approximately 55 % of advertisements were for fast-food restaurants or foods of poor nutritional value. Moreover, the only published study supporting this is by Da Fonseca(8), which showed that South African parents would like to see a reduction of food advertising on TV and stronger restrictions being applied to TV food advertising during children's viewing hours. Moreover, the stricter ASA(9) regulations have mandated at least thirty-one leading fast-moving consumer goods companies, such as Cadbury, Coca-Cola, Kellogg, McCain, Pioneer Foods, Roger Brands, Unilever, Pick and Pay, Nando's and Parmalat, to sign a pledge(14) endorsed by the Consumer Goods Council of South Africa based on the South African guidelines for healthy living(Reference Gibney and Vorster15).
The current study sought to investigate whether there have been any significant changes in the status of commercial TV food-related advertising in South Africa since the study by Temple et al.(Reference Temple, Steyn and Ndamane13). The aims were to determine the frequency of food-related TV advertisements appearing on four different non-paying domestic TV channels viewed by the largest audiences. Furthermore, content analysis of these food-related TV advertisements was undertaken to determine the marketing approach used by advertisers to persuade both children and adults to buy their products.
Methods
Television broadcasting in South Africa
The South African TV viewing statistics by ethnicity in May/June 2011 indicated that 76·7 % of blacks, 8·6 % of coloureds (Afr-Eur-Malay), 2·7 % of Asians and 12·0 % of whites watched the national South African Broadcasting Corporation (SABC) TV channels(16). TV in South Africa is broadcast in all eleven official languages; it receives its funding from both licence fees and advertising, and broadcasts on four non-paying domestic channels (SABC 1, 2, 3 and e-TV) with a mixed entertainment and public service mandate(16, 17). On all four TV channels the time interval 15.00–17.00 hours is dominated by child-focused programmes such as infomercials, educational programmes and cartoons(17). The 17.00–19.00 hours time interval is dedicated to whole family viewing and includes talk shows and soap operas. The time interval 19.00–21.00 hours is mostly dedicated to adults and includes news broadcasts as well as soap operas and movies that are sometimes not suitable for persons younger than the age of 18 years (i.e. semi- to full-restriction soap operas and movies; Table 1)(17).
Procedures
The current study focused on food-related TV advertisements. Multivitamins and slimming products were included due to their relevance to nutrition and weight maintenance. TV advertisements were recorded for 6 h each day for seven consecutive days within a period of 4 weeks, from 16 April 2011 to 13 May 2011. All advertisements were divided into three groups (Tables 1 and 2): (i) those aimed at children (appearing during the time slot 15.00–17.00 hours within and between the educational, infomercial and cartoon programmes); (ii) those aimed at both children and adults (i.e. the whole family, appearing during the time slot 17.00–19.00 hours within and between non-restriction soap operas (family drama), talk shows and early news broadcasts); and (iii) those aimed at adults (appearing during the time slot 19.00–21.00 hours within and between the late news and semi- to full-restriction movies).
Programme diversity by number of viewers, time of viewing and TV channel(17), South Africa
TV, television; SABC South African Broadcasting Corporation.
*Viewership numbers are approximations since people often switch between channels, thus the total TV viewership seems to be higher than the total population of 48·1 million.
Advertisements appearing on South African TV channels by time of viewing, 16 April–13 May 2011
TV, television.
The advertisements were viewed and coded by two researchers. The following information was collected: (i) TV channel; (ii) language of broadcast; (iii) name and type of programme; (iv) date and time of day; (v) target audience; (vi) company placing the advertisement; (vii) description of the product advertised; (viii) description of the health claim; (ix) personality presenting the advertisement; and (x) the inclusion of a web address in the advertisement. Health claims were recorded as being either implicit (having an implied function that is understood though not directly expressed by the advertisement) or explicit (having a clearly and fully defined function that is directly expressed by the advertisement). To test for reliability of coding, data from the first coder were compared with data of the second coder. When there was not 100 % agreement between the two coders, re-coding was done until agreement between the two coders was 100 %. All advertisements were then classified into five categories, namely: (i) food products; (ii) alcohol; (iii) supermarket and pharmacy promotions; (iv) slimming products; and (v) dietary supplements (Table 2). Supermarket and pharmacy promotions were grouped separately from food products, slimming products and dietary supplements as they advertised special offers on food and on dietary supplements, food and slimming products.
Statistical analysis
Analyses were done using IBM SPSS Statistics 19 (2010). Descriptive statistics and cross-tabulations were used to analyse data. As such, data are presented as numbers and percentages.
Results
Out of the total of 1512 TV advertisements, 665 (44 %) were related to food, 22 % were related to clothing and furniture, and 34 % were related to political party election campaigns. In the current study we dealt only with the 665 food-related advertisements. These advertisements were subsequently categorised into the five different groups, namely: 63·2 % were for food products; 21·1 % were for alcohol; 0·6 % were for slimming products; 2·3 % were for dietary supplements (multivitamins); and 12·9 % were for promotions by supermarkets (defined as a large retail markets that sell food and other household goods and are usually operated on a self-service basis) or pharmacies. Table 2 shows that 140 alcohol advertisements were shown on TV at the time the research was undertaken. There were ten different categories under the food group, with the majority of advertisements being for desserts and sweets (22 %), fast foods (20 %), starchy foods (16 %), condiments (14 %), sweetened drinks (10 %) and dairy products (8 %). Meat, chicken and fish, margarine and spreads and infant formula had the least advertisements. Of the 665 food-related TV advertisements, only seven (2 %) were for fruit and vegetables.
Table 3 indicates that the highest number of advertisements was recorded during family viewing time (17.00–19.00 hours; 326 (49·0 %) advertisements). During children (15.00–17.00 hours) and adult (19.00–21.00 hours) viewing times there were 164 (24·7 %) and 175 (26·3 %) advertisements, respectively. The largest number of advertisements was shown on TV channels SABC 2 and 3 (197 and 201 advertisements, respectively) compared with SABC 1 and e-TV (120 and 147 advertisements, respectively; Table 3). Table 2 shows that the majority of advertisements that were presented during children's viewing time were for desserts and sweets (n 39, 28 %), starchy foods and fast foods (each n 24, 17 %), sweetened drinks (n 15, 11 %), coffee, tea and condiments (n 15, 11 %), alcohol (n 12, 7 %) and milk and dairy products (n 11, 8 %). Moreover, out of the total of 140 alcohol advertisements, the large majority (n 93, 67 %) were shown during the children and family viewing time slots (15.00–17.00 and 17.00–19.00 hours, respectively; Table 2).
Numbers of advertisements by time of viewing and TV channel in South Africa, 16 April–13 May 2011
TV, television; SABC South African Broadcasting Corporation.
Of the 665 food-related TV advertisements, seventy-six advertisements (11 %) had health claims that were also explicit (Table 4). These advertisements were for starchy foods (n 38, 50 %), multivitamins (n 14, 18 %), supermarket and pharmacy promotions (n 7, 9 %), margarines and spreads (n 7, 9 %), infant formula (n 5, 7 %), slimming products (n 4, 5 %) and desserts and sweets (‘boosting energy’; n 1, 1 %). All infant formula and slimming product advertisements had health claims. However, out of all the advertisements for multivitamins, for margarines and spreads, for starchy foods, for supermarket and pharmacy promotions and for desserts and sweets, 93 %, 58 %, 57 %, 8 % and 1 %, respectively, had health claims (Table 4). The only advertisement under the desserts and sweets group that had a health claim was for a ‘Yogurt Lollie Pop’, claiming it to be fashionable (‘cool’ as per the advertisement) as it is ‘fruit-based, full of energy and better than a glass of milk’. All slimming product advertisements recorded in the current study had claims that promised weight loss and maintenance of the newly found body size. Seventy-four per cent of the multivitamin advertisements promised balanced nutrition, growth and energy, improvement of performance, relief of constipation and a boost for the immune system. Of the fifty-two food products that carried health claims, thirty-seven (71 %) promised enhancement of well-being.
The proportion of advertisements that used health claims on South African TV channels, 16 April–13 May 2011
TV, television.
Table 5 shows that the majority of advertisements that carried claims for enhanced well-being and improved performance were for food products (95 % and 56 %, respectively). Supermarket and pharmacy promotions, on the other hand, carried the majority (47 %) of claims for improved health, improved immune function and balanced nutrition.
Benefits claimed according to group of TV advertisements in South Africa, 16 April–13 May 2011
TV, television.
Six of the advertisements included information regarding the company's web address. These advertisements were for supermarkets and pharmacies, breakfast cereals, energy drinks, infant formula and dairy products. Twenty-five per cent of supermarket and pharmacy advertisements endorsed best offers, including Easter holiday specials.
Table 6 shows that out of 663 different advertisements the majority (63 %) were presented using ordinary people such as adult men and women, children and the whole family. Twenty-two per cent were presented using a mixture of personalities (i.e. celebrities, professionals such as pharmacists and chefs, as well as ordinary people). Ten per cent of advertisements (alcohol advertisements, in particular) were presented using celebrities only (i.e. actors, sportsmen and TV personalities). The rest of the advertisements used professionals only and popular cartoon characters. For instance, all of the McDonald's children's meal advertisements promised a free toy if one buys a meal. Some examples of the free toys were in the form of popular cartoon characters such as ‘Hello Kitty’, or TV characters such as ‘TransFormers’, or ‘RoboCop’ based on the science fiction action films.
Advertisements by type of presenter on South African TV, 16 April–13 May 2011
TV, television.
Discussion
In the current study we recorded a total of 665 advertisements for food, slimming, multivitamin products, alcohol and supermarket/pharmaceutical promotions that were shown on the non-pay SABC television channels 1, 2, 3 and e-TV over a period of 4 weeks. This is a conservative number of advertisements. The SABC allows at least 10–12 min of advertising each hour(18). According to previous TV advertising records(19) this allows at least six to ten food-related advertisements per hour. This adds up to 1344 advertisements over a period of 4 weeks. The explanation for the low number that we recorded is that at the time of the study the bulk of the TV time (airtime) was dedicated to political parties campaigning for the local elections held in May 2011.
The channels targeted by most SA advertisers were SABC 2 and 3, which are channels viewed by the majority of middle-class South Africans who are also ethnically diverse(17). Moreover, the appearance of nearly 50 % of food advertisements during child and family viewing time and the fact that the most shown advertisements are for desserts and sweets, fast foods and sweetened drinks are a cause for concern. There is strong evidence that suggests TV advertising to be the most powerful medium that encourages the consumption of high-energy food products and beverages(Reference Guran and Bekeret3, Reference Guran, Turan and Akcay12), increases meal frequency(Reference Guran and Bekeret3), promotes fast-food restaurant use(Reference Guran, Turan and Akcay12) and lowers the consumption of fruit and vegetables(Reference Coon and Tucker10). Previous South African research has suggested that parents believe that TV viewing does affect their children's nutritional choices as children tend to prefer unhealthy foods, such as snack foods(8). This therefore makes it less likely that the children will eat nutritious meals.
The strategies used by advertisers in the South African TV media are similar to those seen in other countries(Reference Dijkstra, Heidi and Buijtels4, Reference Lear, Runyan and Whitaker20) in that they include persuasive appeals that target more than one sensory mode, namely vision and auditory modes. In the current study the advertisers used a mixture of personalities, such as celebrities, ordinary people, professionals and cartoon characters, to advertise food-related products. The majority of advertisements that used celebrities, such as popular and attractive TV presenters, soap opera actors and sportsmen, were shown during the children and family viewing times. Reinhard et al.(Reference Reinhard, Messner and Sporer21) found attractive and popular male and female salespersons to induce more positive attitudes and stronger intentions of consumers to purchase a product compared with unattractive salespersons. Moreover, Ülger(Reference Ülger22) found that both children and adults appreciate seeing age-appropriate celebrities on TV as they trust and respect them. International studies also suggest that celebrities appearing in advertising enhance a product's worth and increase sales because they heighten attention to advertisements by virtue of the visual and aural cues associated with celebrity endorsement and the celebrities’ fame extends to the product/brand they are endorsing(Reference Erdogan23, Reference Gantz, Schwartz and Angelini24).
Of concern is that South African celebrities are used in advertisements directed at promoting alcoholic beverages, even during children and family viewing times (15.00–19.00 hours). According to the ASA's code on alcohol(25), alcohol advertisements may not be shown between 14.00 and 17.00 hours on Monday to Friday or before noon on Saturday and Sunday. They may not be directed at persons under the age of 18 years. Moreover, commercial communication may not employ images or icons that have unique appeal to children. In the current study, although advertisements on alcohol had the fine print ‘not suitable for persons under the age of 18 years’, they also emphasised fun, sophistication, flexibility, peer status and championship, and contravened the ASA(25) code.
In the current study popular cartoon characters were also used to endorse sweets such as yoghurt lollipops, sugar-added cereals as well as sugar-concentrated beverages, such as Coke® and Fanta® from the Coca-Cola Company. Brand recognition is thought to be enhanced in young children when cartoon-related characters are used in advertising or on packaging(Reference Connor26). Moreover, children who recognise characters, logos and slogans from advertisements have been shown to be more likely to select those products and brands when they see them(Reference Batada and Borzekowski27). Advertisements for desserts, sweets and sugar-concentrated beverages in the current study also contained depictions of exaggerated pleasure sensations and dependency or addiction: for instance, advertisements for sugar-concentrated beverages (Coke and Fanta from the Coca-Cola Company, in particular) depicted lovely taste, fun and addiction sensations by adolescent endorsers. It is significant that the TV slogan for the South African flavoured water (Aquelle) is ‘taste it, crave it’.
Explicit health claims by advertisers are also rife, with the most common claims on South African TV being for food products such as starchy foods, infant formula and fat spreads. Some of these advertisements comply with the health claim in that they mentioned the actual component (vitamin or mineral) that brings about the health benefit. However, some of the health claims were somewhat exaggerated. These claims suggested that the whole product brings about the health benefit, such that, if the user consumes these products, the user rapidly becomes smarter and gets a surge of mental/brain power, resulting in a sudden flow of bright ideas. Moreover, more than half of the advertisements that had health claims promised enhancement of well-being. This is misleading as well-being is a broad area that involves the balance of mental, social and physical being(Reference Goldstein, Pejchar and Daily28).
Marketing techniques (such as premium offers) or other attributes of advertising (such as the use of web addresses) have previously been identified in TV food advertising(Reference Hastings, Stead and McDermott29). Such endorsements have been shown to increase recall rates for marketing communications as well as positively impacting upon consumers’ attitudes towards the brand. Similarly, in the current study, best offers including Easter holiday specials were emphasised by supermarkets and pharmacies who also specified the usefulness of their Smart Card that rewards points when used to purchase the advertised food-related products at the local stores. Moreover, including reference to a company's web address was used to encourage consumers learn more about other products and offers. Lastly, in the current study it was observed that all of the McDonald's meal advertisements included a reference to a free toy, which is one of the features that make McDonald's advertisements and products so successful internationally(Reference Sahud, Binns and Meadow30).
Conclusion
TV food-related advertisements in South Africa continue to promote less healthful food products. Misleading health claims are rife. Only a handful of advertisements shown are for food with a high nutrient content, such as vegetables, fruit and low-fat milk and milk products. This persistence of advertisements for less healthful products is a cause for concern. The findings reported here are broadly similar to the only other study available of food advertising on South African TV(Reference Temple, Steyn and Ndamane13). That study reported that on SABC 1 and 2 advertising of foods of poor nutritional value to children was prevalent, with less than 50 % of advertisements on good foods. Taken as a whole, the findings of the present study and the previous one(Reference Temple, Steyn and Ndamane13) suggest that government intervention is needed to reduce the advertising of unhealthy food-related products, especially to children. At the same time, efforts could be made to encourage more advertising of healthy foods. There is good reason to believe that such steps would modestly improve the national diet and might therefore help combat the obesity epidemic(Reference Fox31).
Acknowledgements
Sources of funding: The authors gratefully acknowledge the financial support provided by Athabasca University in making this research possible. Conflicts of interest: The authors declare that there were no conflicts of interest. Ethics: Ethical approval was not required for the research. Authors’ contributions: Z.J.M. recorded, entered and analysed the data, conceptualised the information in the paper, produced the first draft of the paper and made all changes resulting from editing by co-authors, reviewers and editors; N.J.T. suggested the research idea, outlined the variables to be considered for the study and edited the first and second last drafts of the paper; N.P.S. helped in the conceptualisation of the paper and edited the first and last drafts of the paper; Z.A. helped in the analysis of data and edited the second draft of the paper; M.C. recorded, entered and conducted the first analysis of data, helped in the literature search for the background of the paper and edited the first draft of the paper. Acknowledgements: Gratitude is extended to the Medical Research Council (MRC)'s Chronic Diseases of Lifestyle Unit and the MRC Library for providing facilities and equipment during the transcription of the taped/recorded data.
