Sugar-sweetened beverage (SSB) consumption has been increasing globally in recent decades( Reference Singh, Micha and Khatibzadeh 1 ), as has SSB marketing( 2 ), with about $US 2 billion spent on SSB marketing worldwide in 2006( 2 ). These coexisting trends are alarming, as SSB consumption relates to overweight and obesity both directly through excess energy intake( Reference Hu 3 ) and indirectly through overall diet patterns( Reference Swinburn, Sacks and Vandevijvere 4 ) across ages( Reference Malik, Schulze and Hu 5 ) from pre-school( Reference de Ruyter, Olthof and Seidell 6 ) to adolescence( Reference Forshee, Anderson and Storey 7 ) and adulthood( Reference Bray, Nielsen and Popkin 8 ), while advertising and front-of-package (FOP) marketing( Reference Cousté, Martos-Partal and Martínez-Ros 9 ) contribute to positive attitudes, preferences and consumption of these products( Reference Cairns, Angus and Hastings 10 – Reference Skaczkowski, Durkin and Kashima 14 ).
Chile leads the world in SSB sales at 787 kJ (188 kcal)/capita per d based on 2014 sales data( Reference Popkin and Hawkes 15 ) and has among the highest overweight and obesity prevalence rates worldwide. About 11 % of Chilean children <6 years of age are obese; 60 % of people aged >15 years are overweight or obese( 16 – 18 ). Guided by WHO( 19 ) and Organisation for Economic Co-operation and Development( Reference Sassi 20 ) recommendations, as well as other governments’ efforts to reduce unhealthy food/beverage marketing( Reference Hawkes and Lobstein 21 ), the Chilean Government passed a series of regulations to improve the local food environment and reduce obesity-related disease in the population, especially in children. A prior 13 % beverage tax (beyond the existing value-added tax) was raised in October 2014 to 18 % for SSB with >6·25 g sugars/100 ml and lowered to 10 % for drinks with ≤6·25 g sugars/100 ml. Unflavoured waters, milks and milk-based beverages, and 100 % fruit juices are tax-exempt( 22 ). Then, a regulation( 23 ) was implemented in July 2016 which requires FOP warning labels (black octagonal stop signs) to identify products high in saturated fat, total sugar, sodium and energy, and restricts child-directed (≤14 years) marketing for those products, with plans for an additional regulation banning advertisements of unhealthy food/beverages across daytime television.
Chile’s regulations represent the most comprehensive attempt by any country to restrict marketing for the purpose of reducing obesity-related disease. Yet, little is known about the FOP marketing used on beverage packaging, or whether those strategies differ by beverage type or nutritional content. According to extant literature on food marketing, unhealthy products (e.g. candies) are often marketed with child-directed strategies( Reference Cairns, Angus and Hastings 10 , Reference Hawkes and Lobstein 21 , Reference Harris, LoDolce and Dembek 24 ), including cartoon and youth characters, references to school or childhood experiences, and appeals suggesting fun or play( Reference Harris, LoDolce and Dembek 24 – Reference Roberto, Baik and Harris 27 ). Thus, we expect that child-directed strategies will be more commonly used on unhealthy beverages (e.g. SSB) that are higher in sugars and energy than on the healthier beverages exemplifying national healthy diet promotion initiatives( 23 , 28 , Reference Olivares Cortés, Zacarías Hasbún and González González 29 ), i.e. beverages such as the plain waters and milk recommended by Chilean and global nutritional guidelines( 28 , Reference Albala, Ebbeling and Cifuentes 30 – 34 ).
We also expect to find ample presence of health-oriented appeals in products contrary to dietary recommendations. Product packaging across food categories has also been shown to feature implicit and explicit health and nutrition claims, especially for products high in sodium, saturated fat or sugars, e.g. labelling a high-sugar product as low in fat( Reference Colby, Johnson and Scheett 35 , Reference Taillie, Ng and Xue 36 ). Also problematic, products using multiple marketing strategies (i.e. child-directed and health-oriented strategies) are often lower in nutritional value than products that do not combine those strategies( Reference Lapierre, Brown and Houtzer 37 ). We therefore anticipate a greater number of different marketing strategies observed on packages of unhealthy beverages, compared with plain waters and milk. However, the extant literature does not address what specific combinations of appeals will be most prevalent, based on the type of beverage.
The present study tests the following hypotheses using a comprehensive sample of beverage packages sold in Chilean supermarkets in 2015, one year after the new SSB tax and before the marketing regulation was implemented. FOP marketing strategies are evaluated according to beverage type, sugars and energy. Addressing these hypotheses provides a deeper understanding of the marketing of different beverage options, as well as a baseline for the longitudinal evaluation of the most comprehensive food labelling and marketing regulation to date.
H1. Use of child-directed FOP marketing strategies will be least common among plain waters and plain milks, compared with other beverage categories.
H2. Use of health-related FOP marketing strategies will be least common among plain waters and plain milks, compared with other beverage categories.
H3. Beverages that use child-directed FOP marketing strategies are more likely to be higher in total sugars, compared with beverages that do not use these strategies.
H4. Beverages that use health-related FOP marketing strategies are more likely to be higher in total sugars, compared with beverages that do not use these strategies.
H5. Beverages that use child-directed FOP marketing strategies are more likely to be higher in energy, compared with beverages that do not use these strategies.
H6. Beverages that use health-related FOP marketing strategies are more likely to be higher in energy, compared with beverages that do not use these strategies.
H7. Compared with plain waters and plain milks, other beverage categories will use a greater number of different FOP marketing strategies on their packaging.
In addition to these hypotheses, the following research question is addressed regarding the combined use of different marketing strategies on beverage packages:
RQ1. Which FOP marketing strategies most commonly appear together on the packaging of different beverage categories?
Photographs of beverage packages (n 1005) were obtained from a larger food and beverage sample collected from five different Chilean supermarket chains between February and April 2015 (plain waters and milks were gathered during February–March 2016) as part of the food environment monitoring activities of the International Network for Food and Obesity/Non Communicable Diseases Research, Monitoring and Action Support (INFORMAS)( Reference Swinburn, Sacks and Vandevijvere 4 ). Six supermarkets (one each from four chains, two from one chain) were visited, all of which were located in a high-income Santiago neighbourhood. Market selection was based on an agreement between the Association for Supermarkets and the University of Chile’s Nutrition and Food Technology Institute( Reference Kanter, Reyes and Corvalán 38 ).
Beverage categories encompassed the scope of the Chilean food labelling and marketing regulation (alcohol is excluded). Over two days of data collection, fieldworkers photographed as many sub-categories of products as possible in each market, avoiding duplication. Photographs were continually logged so that fieldworkers would know which packages had already been captured in another market and therefore did not need to be photographed again. Photographs of the largest package size were taken when more than one package size of a given product was available, provided the same marketing strategies were used on these sizes. If different marketing strategies were deployed in the different package sizes, photographs were taken of each size.
Packages were grouped according to nutrients and product type by a team of trained Chilean and American nutritionists. Groups included: (i) ‘plain waters’ (plain and sparkling unflavoured water); (ii) ‘sports drinks and flavoured waters’ (flavoured water with and without sugar, liquid tea, isotonic sports drinks); (iii) ‘soft drinks with sugar’; (iv) ‘soft drinks without sugar’; (v) ‘100 % fruit juice’ (including 100 % fruit pulp for reconstituting into 100 % fruit juice); (vi) ‘fruit-flavoured drinks’ (sugary and diet fruit-flavoured drinks, nectars, frozen and non-frozen pulps, and concentrated juices); (vii) ‘powder drinks’ (sugary and diet/light); (viii) ‘plain milks’ (liquid and powdered skimmed, reduced-fat and whole unflavoured milk); and (ix) ‘milk-based beverages’ (liquid and powdered drinks containing milk, including skimmed, reduced-fat and whole flavoured milks and liquid yoghurt).
Coding of marketing strategies
FOP marketing appeals, promotions and incentives on the sampled packages were quantitatively assessed using categories of text and images that a team of trained mass communication scholars, including one Chilean, derived from previous marketing studies( Reference Cousté, Martos-Partal and Martínez-Ros 9 – Reference Enax, Weber and Ahlers 11 , Reference Skaczkowski, Durkin and Kashima 14 , Reference Harris, LoDolce and Dembek 24 , Reference Roberto, Baik and Harris 27 , Reference Colby, Johnson and Scheett 35 , Reference Eldesouky, Pulido and Mesias 39 – Reference Page and Brewster 53 ). Chile’s food labelling and marketing regulation( 23 ) was used as a reference. Two coders assessed the presence or absence of marketing appeals, which were combined into seven overarching categories of marketing strategies based on their thematic similarity (Table 1 shows specific appeals within each category): (i) nutrition and health claims; (ii) nature and fruit depictions; (iii) emotion and fun appeals; (iv) child-directed characters; (v) childhood and family situations; (vi) sports references and physical activity portrayals; and (vii) product-oriented appeals.
Untaxed products include plain and sparkling water, milk and milk-based beverages, and 100 % fruit juices.
a,b,cProportions within a row with unlike superscript letters were significantly different at P < 0·05 according to post hoc Fisher z tests. Comparisons made within overarching marketing strategies only. Proportions for all beverages are excluded from these comparisons.
Presence/absence assessments were made irrespective of the quantity of specific iterations of any one strategy. For example, within the child-directed characters category, the sub-category of personified animal was coded as present regardless of whether there were one or more personified animals on the FOP. Then, the overall strategy of use of characters was assessed as present if at least one sub-category was coded as present. Presence of gifts (e.g. stickers, toys), games and contests was also initially coded, but their frequency of occurrence was minimal. Therefore, gifts, games and toys as marketing strategies were excluded from the reported analyses. The codebook is available as online supplementary material.
A random selection of 20 % of the sample (n 202) was analysed by each of the two trained coders to assess inter-rater reliability using ReCal software( Reference Freelon 54 ) prior to splitting the remaining 80 % of the sample between the coders( Reference Neuendorff 55 ). Coders achieved between 92·87 and 99·67 % agreement on each marketing strategy (e.g. nutritional and health appeals) and between 88·12 and 100 % agreement on the individual codes comprising the strategy (Scott’s π=0·81, 96·99 % agreement across all codes).
Product names, brands, product descriptions and ingredients were used to place each beverage into one of the nine categories noted above (plain waters, sports/flavoured waters, soft drinks with or without sugar, 100 % fruit juice, fruit-flavoured drinks, powders, plain milks, or milk-based beverages). Total sugars (g/100 ml) and energy (kcal/100 ml; 1 kcal=4·184 kJ) were also recorded for each beverage using the nutrition facts panel data provided on the packages. Although mandatory in Chile to declare sugar content, thirty-two packages did not provide sugar information and so sugars were imputed by a trained dietitian based on nutrition information on packages sharing the same product sub-category, brand, tax status and ingredients as the package in question. For powders and concentrated formats, sugars and energy were computed based on the reconstituted product.
Tax status was also considered, as it reflects total sugars, yet notably excludes 100 % juice and milk-based beverages. Based on Chilean law, the following products were assigned an 18 % tax rate if they contained >6·25 g total sugars/100 ml and a 10 % tax rate if they contained ≤6·25 g total sugars/100 ml: natural or artificially flavoured non-alcoholic beverages; energy or hypertonic drinks; syrups and any product that uses syrups to make a beverage; and mineral or thermal waters with added colour, flavour or sweeteners. Plain waters, milks and milk-based beverages, and 100 % fruit juice were assigned a 0 % beverage tax rate.
For each FOP marketing strategy (e.g. emotion and fun appeals, use of child-directed characters), a χ 2 analysis was used to evaluate whether the prevalence of that strategy differed according to beverage taxation status (untaxed, 10 %, 18 %). Post hoc Fisher’s exact probability z tests provided pairwise comparisons between tax rates within any strategy that significantly differed in its prevalence based on tax status. The same procedure was used to test H1 and H2 by detecting differences in the prevalence of each marketing strategy across product category (e.g. soft drinks with sugar), with post hoc pairwise comparisons performed between each of the taxed product categories and the tax-exempt plain waters, plain milks and 100 % fruit juice categories.
Descriptive and inferential statistics (t tests) were used to compare the average amount of sugars (H3 and H4) and energy (H5 and H6) in products using (v. not using) each marketing strategy. An adjusted t test was used in the few cases when variances between groups were unequal (noted in the relevant tables). H7 was tested with an ANOVA (and post hoc pairwise mean comparisons) to compare the average number of different types of marketing strategies used across the different product categories (range=0 to 7 types of strategies). Frequencies were used to address RQ1, identifying the marketing strategies most likely to be used in tandem, depending on product category. All analyses were conducted with the statistical software package IBM SPSS Statistics version 24.
According to χ 2 tests of independence, the prevalence of use of six of the seven overarching marketing strategies on the front of beverage packages varied according to the beverages’ taxation status, χ 2(2) ranging from 6·80 for product appeals to 189·32 for nutrition/health strategies, P<0·05. The one exception to this finding was that emotion and fun appeals were independent of taxation status, χ 2(2)=4·10, P=0·13. The proportions of products using these appeals according to product category and taxation status are shown in Table 1. Beverages taxed at 10 % were more likely to feature nutrition/health appeals and, to a much lesser extent, sports/physical activity references (93 and 17 % of packages, respectively) than either untaxed products (80 %; 10 %) or products with the 18 % tax (49 %; 8 %). Also, significantly more of the untaxed products used nutrition/health appeals, compared with products at the 18 % tax rate. Products taxed with either a 10 % tax or a 18 % tax were more likely to use nature/fruit depictions (80 and 82 % of packages, respectively), compared with untaxed products (59 %).
Likely driven by the marketing strategies of milks and milk-based beverages, a greater number of untaxed products featured child-directed characters (19 %) and childhood and family references (5 %), compared with character use or child/family references among the 10 % taxed or 18 % taxed products (Table 1). Recall that milks and milk-based beverages are not taxed in Chile.
Mean energy and amount of total sugars for each FOP marketing strategy are shown in Tables 2 and 3, respectively, with results of the t tests comparing means based on use (v. non-use) of each strategy as follows. Products that deployed a nutrition and health strategy were lower on average in both total energy (kcal) and sugars (g) per 100 ml, compared with products that did not use a nutrition/health strategy (energy: t(463·86) = −8·45, P<0·05; sugars: t(1003) = −11·47, P<0·05). A similar pattern was seen for sports and physical activity appeals, in that use of this strategy was indicative of products with lower amounts of energy and sugars rather than higher amounts of energy and sugars (energy: t(1003) = −2·53, P<0·05; sugars: t(1003) = −2·31, P<0·05).
* P < 0·05.
† Adjusted t test due to statistically significant Levene’s test for equality of variances.
* P < 0·05.
† Adjusted t test due to statistically significant Levene’s test for equality of variances.
Conversely, products that used FOP child-directed characters were significantly higher in both energy and sugars, compared with products that did not use them (energy: t(103·32) = 6·02, P<0·05; sugars: t(119·99)= 2·56, P<0·05). Also, packages featuring a childhood or family reference had more energy per 100 ml, on average, than packages without this reference (t(1003)=2·33, P<0·05), and packages with a nature or fruit representation tended to have more sugars than packages that did not use this type of strategy (t(497·44) = 2·36, P<0·05). In the light of innate differences in the composition of milks (i.e. fats, lactose) compared with other beverages, the above analyses were repeated within milks and milk-based beverages combined, as well as within the remaining sample minus milks and milk-based beverages (data not shown). Findings noted above were, for the most part, robust across these subgroups. Qualifying the differences in FOP use of childhood/family references and child-directed characters, statistical significance was found only within milks and milk-based beverages but not within the remaining sample.
Detailed in Table 4, the prevalence of each type of FOP marketing strategy (e.g. nutrition and health, emotion and fun) depended on the type of product being marketed; χ 2 results comparing the prevalence of each strategy across the product categories ranged from χ 2(8)=12·92 (emotion/fun) to 375·45 (nature/fruit), P<0·05. Nutrition/health strategies were on nearly all powder drinks and on all soft drinks without sugar, whereas these strategies appeared on significantly fewer plain waters, plain milks or 100 % fruit juice packages (Table 4). Nutrition/health strategies were also more likely to exist on milk-based beverage packages (89 %), when compared with plain waters (63 %) or 100 % fruit juice (56 %). More sports drinks/flavoured waters also contained nutrition/health appeals, compared with plain waters and 100 % fruit juice. Worthy of note, a similar percentage of soft drinks with sugar (48 %) employed nutrition/health strategies as plain waters (63 %) and 100 % fruit juice (56 %). Fewer nutrition/health appeals existed among soft drinks with sugars than among plain milks (87 %).
A,B,CPercentages within a row with superscript letters were significantly different from that for Aplain waters, Bplain milks and/or C100 % fruit juice at P<0·05 based on post hoc Fisher’s exact probability tests (z tests).
Depictions of nature or fruit were widely used on soft drinks without sugar, powder drinks, fruit-flavoured drinks, milk-based beverages, 100 % juice, and on nearly half of all soft drinks with sugar. These depictions were rare among plain milks. Powders, fruit-flavoured drinks and 100 % juices were also more likely to use a nature or fruit appeal, compared with plain waters. Sports/physical activity references were on half of the sports drinks/flavoured waters and one-fifth of the milk-based beverages, but largely absent for plain waters, plain milks and 100 % fruit juice.
The use of child-directed characters was seen on one-third of milk-based beverages, yet rarely occurred on plain waters, plain milks or 100 % fruit juice packages, although plain milks did feature FOP characters more often than either plain waters or 100 % fruit juices. Emotion/fun appeals were more widely used on plain waters, compared with many other product categories, including plain milks (Table 4). Product-based appeals appeared on one-third of powder drinks, whereas they were rarely used on plain waters, plain milks or 100 % fruit juice packages. Soft drinks with sugar were more likely to feature product-based appeals than plain waters.
Overall, relatively few plain waters, plain milks and 100 % fruit juice packages employed the examined marketing strategies, compared with the other product categories. The only exception to this statement concerns the rarely used childhood/family appeals; these appeals were employed on 6 % of plain waters and 10 % of plain milks but were essentially absent for sports/flavoured waters, fruit-flavoured drinks and powder drinks (Table 4).
Shown in Table 5, nutrition/health and nature/fruit appeals were most commonly used together on packages for all but one product category; 56 % of the overall sample contained this pairing. Sports/flavoured waters were the exception, where half (50 %) of those packages featured the combination of nutrition/health and sports/physical activity. With few exceptions, products that used multiple strategies featured nutrition and health appeals as one of the strategies alongside a variety of other tactics, including product-oriented appeals (e.g. powders, soft drinks), emotion and fun appeals (e.g. soft drinks, plain waters, 100 % fruit juice), and use of child-directed characters (e.g. plain milks and milk-based beverages).
The average number of different marketing strategies used on FOP (mean (sd)) varied according to the type of product, F(8, 996) = 22·49, P<0·01. According to pairwise post hoc comparisons (Table 6), powder drinks (2·51 (0·65)) and milk-based beverages (2·46 (1·10)) used more than two types of marketing strategies on average, which is a significantly greater number of strategies compared with the number of strategies used by other product categories, at P<0·05. Soft drinks with sugar (1·22 (0·94)) and without sugar (1·60 (0·64)), plain waters (1·63 (0·89)) and plain milks (1·61 (1·01)) used the fewest number of strategies on their FOP on average, significantly fewer than the above categories, but also fewer than sports/flavoured waters (2·13 (0·82)) and fruit-flavoured drinks (1·95 (0·81)), based on pairwise comparisons evaluated at P<0·05. The number of FOP strategies used by 100 % fruit juices (1·79 (0·73)) was similar to that of fruit-flavoured drinks.
a,b,c,dMean values with unlike superscript letters were significantly different at P<0·05 according to Fisher’s Least Significant Difference post hoc pairwise comparison. Mean for all beverages is excluded from these pairwise comparisons.
FOP marketing is a key component of integrated marketing campaigns that include television and Internet advertising( Reference Grigsby-Toussaint, Moise and Geiger 56 , Reference Hawkes 57 ), and a focus of Chile’s regulatory efforts to decrease obesity rates. The present study revealed a number of differences in FOP marketing strategies on beverages available in Chilean supermarkets in 2015, according to the nutritional quality of the products. As hypothesized, beverages with child-directed characters had more sugars (H3) and energy (H5) than beverages without FOP characters, and beverages using childhood/family references were also higher in energy than beverages without these references (H5). Partly supporting H1 and qualifying the above findings, flavoured milk-based beverages more often featured child-friendly characters than other beverages, including plain waters or plain milk. However, no such differences were found for use of childhood references or fun appeals; few packages in any category used these appeals. Thus, whereas the present study generally corroborates research finding that child-oriented marketing strategies, e.g. use of characters, are likely to be used with a variety of foods less healthy for children( Reference Harris, LoDolce and Dembek 24 – Reference Roberto, Baik and Harris 27 ), the primary finding for beverage packages centres on use of characters by flavoured milks, which suggests a strategic capitalization on children’s information processing, as children are attracted to characters( Reference Ares, Arrua and Antunez 25 ) or suggestions of fun( Reference Elliott 26 , Reference Roberto, Baik and Harris 27 , Reference Hawkes 57 , Reference Chacon, Letona and Barnoya 58 ) in part because these appeals are easier to comprehend than nutritional information( Reference Livingstone and Helsper 59 , Reference Roedder 60 ).
Yet, the present study also notes widespread use of health-oriented appeals. Almost half of the beverages taxed at the highest rate (>6·25 g sugars/100 ml) featured nutrition and health appeals and over 80 % of these products featured nature and fruit references, suggesting a disconnect between policy and marketing messages. Addressing H2, powder drinks and soft drinks without sugar used nutrition and health claims more often that both plain waters and plain milks. Although H4 and H6 were not supported with a general difference in total sugar or energy between products (categories collapsed) using v. not using nutrition and health claims, beverages with nature or fruit references were shown to have significantly more sugars than beverages without these references. The frequency of nutrition, health, nature and fruit appeals found in the present study is consistent with other findings that packages across food categories often feature health-oriented marketing strategies, even if the product’s overall nutrient profile is problematic (e.g. low in fat but high in sugar)( Reference Colby, Johnson and Scheett 35 , Reference Taillie, Ng and Xue 36 ). These strategies are problematic because they can appeal to adults and youth alike( Reference Ares, Arrua and Antunez 25 , Reference Elliott 26 , Reference Colby, Johnson and Scheett 35 , Reference Lapierre, Brown and Houtzer 37 , Reference Chandon and Wansink 61 – Reference Mayhew, Lock and Kelishadi 64 ) and create a ‘health halo’ around the product, whereby consumers generalize the positive health attribute and overestimate the nutritional quality of the whole product( Reference Abrams, Evans and Duff 65 – Reference Sundar and Kardes 67 ). Thus, a sugary product with a nature( Reference Schuldt 68 ) or fruit image( Reference Heller, Martin-Biggers and Berhaupt-Glickstein 69 ) might capitalize on beliefs that real fruit, and real fruit sugar, is healthy( Reference Abrams, Evans and Duff 65 , Reference Gill and Sattar 70 , Reference Sütterlin and Siegrist 71 ).
Potentially enhancing effects of health-oriented marketing strategies, the present study found that health-oriented appeals often co-occurred with other marketing strategies, such as use of characters, emotion/fun, sports and product appeals (addressing RQ1). For example, in support of H7, powder drinks, flavoured milk-based beverages and sports drinks/flavoured waters employed more FOP marketing strategies in tandem, compared with plain waters and plain milk. Soft drinks without sugar were similar to plain waters and plain milk in using few concurrent strategies. Using multiple strategies increases the information density of a package, which hinders consumers’ ability to identify and process all of the information provided about a product( Reference Bialkova, Grunert and van Trijp 72 ). Younger children are particularly vulnerable to multiple appeals( Reference Gunter 73 ) because of limitations in literacy and capacity to process and retain more than one message at a time( Reference Roedder 60 , Reference Schneider and Pressley 74 ). This understudied issue of information density might become particularly important when assessing the addition of the new mandatory ‘high in’ warning labels to product packages that also contain other marketing strategies, especially if those messages contain health-oriented appeals that seem to contradict the warnings. In any case, future research is needed to understand whether the use of health-oriented appeals alone and in combination with other appeals result in higher purchases and intake of unhealthy beverages, especially among children.
The present study inherits sample limitations that prevent analyses of the prevalence of marketing strategies based on interactions between product categories and other nutritional qualifications, as there were not enough observations of each strategy in the sample to support this type of analysis. The five supermarket chains selected for sampling also reduces the generalizability of these findings to the markets that are part of the INFORMAS’ food system monitoring project in urban Santiago. Moreover, photographs were based on the availability of products in those markets rather than beverage consumption patterns in the Chilean population. Although the sample of products obtained from these markets represents 76 % of all beverage products purchased in Chile, according to Kantar Worldpanel’s (2013–2015) nationally representative consumer panel, this sample might still exclude beverages and branding strategies (e.g. affordability appeals) that target supermarkets in demographically different neighbourhoods. Thus, even though many national brands exist across chains, regardless of the composition of the surrounding neighbourhood, future research beyond the current baseline study will include a wider sampling of supermarkets. Furthermore, the decision to photograph the largest packaging of a given product (e.g. an eight-pack bundle), unless other marketing tactics were seen in smaller packages, introduces the possibility that different tactics on individual containers might have been obscured in the few cases when bulk packaging was not transparent or translucent.
The present study also focused on FOP marketing strategies, as FOP marketing is a key promotional medium that influences attitudes towards products and point-of-sale purchase decisions( Reference Harris, Pomeranz and Lobstein 13 , Reference Hawkes 57 , 75 ). However, our coders noted that some products containing child-directed FOP strategies (e.g. characters) also featured strategies on the sides and/or backs of the products (data not shown). Thus, the decision to only examine FOP marketing might under-represent the amount of various marketing appeals used on the total surface of beverage packaging.
The present study also does not include FOP features related to visual design aspects, such as attractive colours or unconventional fonts. Such features have been argued as targeting children, e.g. an unconventional font suggesting fun( Reference Elliott 26 , Reference Abrams, Evans and Duff 65 ). However, these aspects are not included in the Chilean food marketing regulation( 23 ) and were ultimately excluded from the present study because of the difficulty in defining child-targeted (v. other-targeted) visual design in the concrete terms needed for inclusion in a rigorous quantitative content analysis (or regulation). Still, visual design elements constitute part of an overall marketing strategy and exclusion of such aspects likely leads to underestimation of the prevalence of strategies that would appeal to children.
The June 2016 Chilean regulations banning food marketing aimed at children (≤14 years) are designed to eliminate the characters, fun appeals and childhood references seen in the present study and other research( Reference Cairns, Angus and Hastings 10 , Reference Hawkes and Lobstein 21 , Reference Harris, LoDolce and Dembek 24 ) on the packaging of beverages high in total sugar, saturated fat, sodium or energy. First, given that SSB are a major contributor to excess intake of added sugars and energy, it is crucial that countries monitor, longitudinally, the marketing strategies on food and beverage packages and in other venues, e.g. corporate sponsorships, advertising or product placement( Reference Gunter 73 ), to evaluate how these strategies evolve in response to these regulations. The current study provides a baseline description of the types and prevalence of various marketing strategies, including child-directed strategies, that can be used to test new hypotheses about shifts in marketing efforts. Then, data on marketing shifts can be used in combination with both longitudinal survey data assessing changes in diet and longitudinal household-level purchase data to directly assess whether reducing these child-directed marketing strategies translates into a reduction of SSB consumption (or substitution to plain water), despite any marketing shifts.
Future research is also needed to determine whether including the ‘high in’ FOP warning labels implemented by the June 2016 regulation will provide consumers with an accurate impression of products’ nutritional profiles, given these warnings will likely exist alongside health-oriented and other marketing appeals allowed by the regulation. The regulation currently forbids products with a ‘high in’ label from making claims about the identified excessive nutrient; thus, a ‘high in saturated fat’ product cannot claim to be low fat. There is also a regulation in process to ban certain specific nutrition claims in ‘high in’ products (other health appeals, e.g. use of the Guideline Daily Amount, are not covered). However, products with a ‘high in sugar’ warning, for example, can claim high calcium, wholesome ingredients or low fat content. Given the prevalence of health-oriented appeals found in the present study, we expect marketing efforts will shift to include even more of these appeals (e.g. more nature or fruit claims) as allowed by the regulation.
Perhaps the most important need for research is the assessment of how FOP marketing shifts couple with FOP warning labels and SSB taxation to impact children’s intake of high-sugar beverages( Reference Nestle and Ludwig 76 ). Based on the present study’s findings, there is real concern about the information density on packages, especially if the information might suggest both healthy and unhealthy qualities of a product. At minimum, research is needed to examine the coexistence of tax and warning labels on products that currently use health-oriented appeals, as half of the beverages taxed for sugars content were found to feature such appeals. This line of research will also be useful for evaluating trends in consumption of flavoured milks and other milk-based beverages as potential substitutes of SSB in Chile, given both flavoured and plain milks are exempt from taxation and, as the present study found, flavoured milks used more characters and nature depictions than plain milk. Longitudinal data collection of products will be needed to note any reformulations of nutrients within beverages from this baseline collection, in addition to tracking changes in marketing strategies, to fully understand how the combination of marketing, warnings and taxation affects SSB purchases and intake in the Chilean youth population.
Acknowledgements: The authors thank Bloomberg Philanthropies, the International Development Research Centre, the University of North Carolina Institute for Global Health & Infectious Diseases, and the Carolina Population Center for financial support. They also thank Natalia Rebolledo Fuentealba and Donna Miles for their exceptional assistance with categorizing beverages based on nutritional databases, Frances Dancy for administrative assistance, Dr Barry Popkin for his leadership and support, and the Carolina Population Center for general support. The authors also wish to thank the ‘super chicas’ who collected the photographs of food packages and the International Network for Food and Obesity/Non Communicable Diseases Research, Monitoring and Action Support (INFORMAS) group at the University of Chile’s Institute of Nutrition and Food Technology for their continual support throughout the data collection. Financial support: Support comes primarily from Bloomberg Philanthropies, with additional support from International Development Research Centre (IDRC) grants 107731-002 (INFORMAS) and 108180-001 (INTA-UNC), CONICYT (FONDECYT #1161436, FONDECYT Postdoctorado Proyecto #3150183), the Carolina Population Center (P2C HD050924), and the University of North Carolina Institute for Global Health & Infectious Diseases (Explorations in Global Health grant). Beyond financial support, funders had no role in the study design, data collection, analysis or interpretation. Conflict of interest: None of the authors have conflicts of interest of any type with respect to this manuscript. Authorship: F.M.S. led the content analysis, statistical analysis and writing. J.O.B. contributed to the content analysis. R.K. led the data collection. C.C. and M.R. supervised the data collection and contributed an initial codebook. L.S.T. contributed to the research design. F.R.D.C. supervised the study execution, analysis and writing. All authors contributed to manuscript edits. Ethics of human subject participation: No human subjects were involved in this study.
To view supplementary material for this article, please visit https://doi.org/10.1017/S1368980017002671