Associations between exposure to digital food marketing and food consumption in adolescence: A cross-sectional study in an emerging country

doi:10.21203/rs.3.rs-5011928/v1

Background

Evidence regarding the link between digital food marketing and eating habits is lacking in the majority world, i.e., the world regions where most people live. This study sought to investigate i) self-reported exposure to digital food marketing, ii) associations between such exposure and socio-demographic characteristics, and iii) associations between said exposure and food consumption frequency among adolescents in a Latin American country (Uruguay).

Methods

A sample of adolescents in grades 1–6 attending 29 public and 10 private high schools (n = 1542) was obtained through a cross-sectional survey using a stratified, two-stage cluster probability-based sampling approach. Participants filled out closed and open-ended questions about exposure to digital food marketing, food consumption frequency, social media usage, and socio-demographic characteristics. The data were analyzed through descriptive statistics and ordinal logistic regressions.

Results

Almost 90% of participants reported having seen a food and beverage advertisement on digital media in the week prior to the survey, with more than 70% of participants recalling advertisements of fast food, soft drinks, and savory snacks. Age, socio-economic status, or total social media use did not predict exposure frequency of digital food marketing but females (vs. males) reported higher exposure. Exposure to advertisements of fast food or ultra-processed products on social media or websites and total social media use typically predicted higher reported consumption frequency of such categories. However, exposure to digital food marketing did not predict consumption frequency of fruits, vegetables, meats, or fish, although total social media use predicted lower consumption frequency of fruits and vegetables.

Conclusions

Mere exposure to digital food marketing and total social media use were associated with higher consumption frequency of ultra-processed products. These findings underscore the need to reduce adolescent exposure to digital food marketing featuring unhealthy foods, possibly through mandatory policies including total bans of certain foods in such marketing.

marketing

digital marketing

social media

food environment

adolescence

Marketing of ultra-processed products is one of the commercial determinants of health [1, 2]. Through a hierarchy of sequential effects, exposure to food marketing influences food-related attitudes, increases purchases and consumption, and, in the long term, impacts health-related outcomes [3]. With the advent of the internet, digital marketing has become one of the most popular components of the marketing mix of food companies [4–6]. It can be defined as any "promotional activity, delivered through a digital medium, that seeks to maximize impact through creative and/or analytical methods" [7].

Digital marketing has a series of characteristics that make it potentially more persuasive than other forms of off-line marketing. Firstly, digital media are ubiquitous and not confined to specific settings, providing opportunities to reach audiences anytime and anywhere [8]. It can be tailored to the specific characteristics of users through audience segmentation strategies based on data analytics [9]. Digital media also enables interactions with users, endorsement of brand messages, as well as embedding advertisements within entertainment content, which increases persuasiveness [8, 10–12].

Exposure to digital food marketing is highly relevant for adolescents as they are avid users of digital media [13, 14]. Susceptibility to the persuasive effects of unhealthy food marketing is particularly high in adolescence due to neurodevelopmental mechanisms that heighten sensitivity to rewards, reduce inhibitory control, and encourage risky behaviors [15, 16]. In addition, adolescents are in a key stage of their identity development and are highly susceptible to social pressure, wanting to conform to social norms around food consumption [17, 18]. Increased autonomy and agency and reduced parental control and guidance during this life period may also increase vulnerability to the persuasive effects of marketing [19].

Adolescents across the globe have been reported to be frequently exposed to digital marketing of unhealthy foods, especially fast food and ultra-processed products such as sweetened beverages, chocolates and confectionary, savory snacks and breakfast cereals [20–28]. However, knowledge on the effect of exposure to digital food marketing on the eating habits of adolescents is still limited [29, 30]. A handful of observational studies have associated self-reported exposure to digital marketing with increased consumption of unhealthy foods and drinks [22, 23, 29, 31]. In addition, a recent experimental study conducted in a Latin American country has shown that exposure to digital marketing of unhealthy foods leads to increased advertisement and brand recall [32].

Additional information on the nature, extent, and impact of digital food marketing is needed to strengthen public policies to protect adolescents [9]. In particular, evidence on the associations between exposure to digital food marketing and eating habits is still lacking in the majority world, i.e., the regions of the world where most of the population live [33]. Therefore, this study aimed at expanding current knowledge by examining exposure to digital marketing of ultra-processed products among adolescents in Uruguay, a high-income Latin American country. The objectives were to: i) estimate self-reported exposure to digital food marketing among Uruguayan adolescents, ii) analyze associations between such exposure and socio-demographic characteristics, and iii) analyze associations between said exposure and food consumption frequency.

The present cross-sectional observational study was part of a larger project on Uruguayan adolescents' exposure to digital food marketing. Uruguay has one of the highest prevalence of overweight and obesity among adolescents in the region: 33.6% among 13–17 years old adolescents [34]. Unhealthy diets, and particularly consumption of ultra-processed products, has been identified as a relevant behavioral risk factor contributing to adolescent overweight and obesity [34–37]. Internet access is widespread among Uruguayan adolescents. According to a nationally representative survey conducted in 2019, 94% of the adolescents between 14 and 19 years had daily access to the internet, and they all had at least one social media account [38]. At the time of the study, the country did not have any regulations or guidelines on food marketing on mass media, including digital food marketing.

Participants

A stratified, two-stage cluster probability-based sampling approach was used to obtain a sample of adolescents in grades 1–6 attending public and private high schools in Montevideo, the capital city of Uruguay. The sampling frame included all the public (n = 77) and private (n = 94) high schools in Montevideo, registered at the National Administration of Public Education.

The first stage consisted of selecting high-schools as primary sampling units using probability proportional to size. High schools were stratified by type (public vs. private) to ensure the sample reflected the local distribution of students. The probability of selection of each public/private high-school was proportional to the proportion of students in the corresponding type of high-school. A sample of 39 high schools was obtained, composed of 29 public and 10 private schools. The research team contacted each school to obtain authorization from the principal and to request information about the number of classes in each grade and the corresponding number of students. All chosen schools agreed to participate.

The second sampling stage consisted of selecting classes within each school. One class within each grade was randomly selected. All students attending school the day the questionnaires were distributed were eligible for participation.

Questionnaire

A questionnaire was developed by the research team based on a previous qualitative study on adolescents' experiences with digital food marketing [27], and questionnaires from published studies [22, 25, 39]. It included four main sections: i) exposure to digital food marketing, ii) food consumption frequency, iii) social media usage, and iv) socio-demographic characteristics.

Frequency of exposure to digital food marketing was assessed using the question: Have you seen advertisements of foods and drinks on social media or websites in the last week? The following response options were available: 'Yes, more than once a day', 'Yes, once a day', 'Yes, several times a week', 'Yes, once in the week' and 'No, I haven't seen any advertisement'. Frequency of exposure was then recoded into the following three categories: 'No exposure', 'Less than once a day', 'Once a day or more'. Then, participants indicated their spontaneous recall of food and beverage advertisements on social media by answering the following open-ended question: What advertisements of food and beverages do you recall seeing on social media? (data not analyzed in the present work). Exposure to advertisements on specific social media was assessed through the following question: In which of the following media do you remember seeing any food or beverage advertisements in the last week? For each of the following social media, participants had to indicate Yes or No: Instagram, TikTok, Facebook, YouTube, Snapchat, Twitter, Twitch, Website browsing. The final question of the section explored prompted recall of advertisements of specific ultra-processed products: Do you remember seeing any advertisement of the following food or beverages on social media or websites in the last week? Participants had to answer 'Yes' or 'No' for each of the following products: i) soft drinks, ii) energy drinks, iii) flavored water, iv) bottled juices or powdered drinks, v) cookies or crackers, vi) alfajores [traditional product composed of two soft biscuits, joined by a sweet filling, usually covered in chocolate], vii) flavored milk, yogurt or milk desserts, viii) chocolates or confectionary, ix) ice-cream, x) cold cuts or sausages (e.g., ham, salami, sausages, hot dogs), xi) breakfast cereals or cereal bars, xii) bakery products (e.g., croissants, donuts, pastries), xiii) savory snacks (e.g., potato chips, Doritos, Cheetos), xiv) Condiments (e.g., ketchup, mustard, or mayonnaise), xv) soups or broths, xvi) marmalades or dulce de leche [local type of sweetened condensed milk], xvii) food eaten in fast food restaurants (e.g., burgers, pizza, French fries).

In the second section, participants were asked to indicate how often they ate a variety of natural foods and ultra-processed products the week prior to the survey ('How many day last week did you eat (name of the food item)?): i) fruits, ii) vegetables, iii) beef, chicken or pork, iv) fish, v) milk or cheese, vi) soft drinks, vii) energy drinks, viii) flavored water, iv) bottled juices or powdered drinks, x) cookies or crackers, xi) alfajores, xii) flavored milk, yogurt or milk desserts, xiii) chocolates or candies, xiv) ice-cream, xv) cold cuts or sausages (e.g., ham, salami, sausages, hot dogs), xvi) breakfast cereals or cereal bars, xvii) bakery products (e.g., croissants, donuts, pastries), xviii) savory snacks (e.g., potato chips, Doritos, Cheetos), xix) ketchup, mustard, or mayonnaise, xx) at a fast food restaurant (e.g., burgers, pizza, French fries). The response options for all items were: '0 day', '1 day', '2–3 times', '4–6 times', '7 days'.

Self-reported daily social media time was measured using the question: On a normal weekday, how much time do you use (name of the social media)? Participants were asked to answer this question for eight media: Instagram, TikTok, Facebook, Twitter, YouTube, Snapchat, Twitch, Website browsing (e.g., Google). Participants were also asked about their time spent watching TV: On a normal weekday, how much time do you watch TV (including cable TV, Netflix, Disney + or other platforms)? For both questions the response options were: 'I don't use', 'Less than 15 minutes', '15 to 30 minutes', '30 minutes to 1 hour', '1 to 2 hours', '2 to 3 hours', '3 to 4 hours', 'More than 4 hours'. Responses were recoded considering the middle point of the range (e.g., 15 to 30 minutes was recoded to 22.5 minutes). The response option 'More than 4 hours' was recoded as 240 minutes (4 hours). Participants' total exposure to digital media combined (in hours) was calculated by summing up the time spent on the seven digital media (Instagram, TikTok, Facebook, Twitter, YouTube, Snapchat, Twitch, Website browsing).

The last section focused on socio-demographic information. Participants were asked to indicate their gender (male, female, other), age, and neighborhood of residence. Socio-economic status was estimated using the score of the neighborhood of residence in the national socio-economic status index, which ranges from 0 to 14 (Centro de Investigaciones Económicas, 2023). The scores associated to each socio-economic status were: <4 low socio-economic status, ≥ 4 and < 11 medium socio-economic status, ≥ 11 high socio-economic status.

The questionnaire was pilot tested with a convenience sample of 17 adolescents of different socio-economic status. After they completed the questionnaire, they were asked if they had faced any challenge or difficulty. As they all answered negatively, no changes were introduced before data collection.

Data collection

Paper-based questionnaires were delivered to the high schools by members of the research team. The staff of each high school oversaw distributing the questionnaires to all the students in the selected classes. Participants not attending classes the day the questionnaires were distributed did not participate in the study.

The questionnaires were self-administered and completed at participants' homes. They included an introductory letter to parents about the survey and provided contact information if they had any questions. Informed written consent from one of the parents (or an adult in charge of the adolescent) and written assent from adolescents were required. Parents and adolescents were informed that the study was related to social media. No specific information about digital marketing or food was provided in the information sheet to avoid potential response biases. No compensation for participation was provided.

After the questionnaires had been collected from participants, school staff contacted the research team to pick them up. Data collection was performed between April and July 2024. After data analysis was finalized, high-school principals, participants, and their parents were sent information sheets summarizing the key results of the study.

Data analyses

All data analyses were performed in R software [40]. Descriptive statistics were used to summarize results from multiple choice questions. Average, median and standard deviation were calculated for total exposure to digital media. Participants with missing data (n = 24) or whose self-reported gender identities (n = 9) differed from males or females were not considered in the analyses given the small sample size and, hence, insufficient statistical power to detect 'typical' effect sizes in the published literature.

An ordinal logistic regression model was used to analyze associations between self-reported exposure to food digital marketing and socio-demographic variables. The model was run considering self-reported frequency of exposure in the week prior to the survey ('None', 'Less than once a week', 'Once a week or more') as the dependent variable, whereas the following individual characteristics were considered as independent variables: gender (male vs. female), age (range 11–14 years old vs. 15–19 years old), socio-economic status (low vs. high and low vs. medium), total social media use (in hours), and TV watching time (in hours).

The association between consumption frequency of ultra-processed products and self-reported exposure to digital marketing was assessed using an ordinal logistic regression model. For each category of ultra-processed products included in the questionnaire, a separate ordinal logistic regression model was run considering consumption frequency ('0 days', '1 day' and '2–3 days', '4–6 days' and '7 days') as the dependent variable and the following independent variables: self-reported exposure to an advertisement of the category on social media (Yes vs. No) and total social media use (in hours). The models controlled for gender, age range, socio-economic status, and TV watching time (in hours). Similar models were run to analyze the relationship between self-reported exposure to digital food marketing (Yes vs. No) on consumption frequency of natural foods (fruits, vegetables, meat, and fish).

The proportional odds assumption of the ordinal logistic regression models was assessed using Lipsitz goodness of fit test and Pulkstenis-Robinson chi-squared test [41]. These tests were conducted using the package generalhoslem in R [42].

A total of 1542 adolescents, aged between 11 and 19 years (M = 14.3, SD = 1.8), completed the questionnaire. The number of completed questionnaires corresponded to 38.8% of the total number of students in the selected classes. However, the actual response rate was most likely somewhat higher as high-school staff did not report the number of students who were sick or absent for other reasons. As such, our effective response rate is likely an underestimation of the actual response rate. Considering that studies with response rates as low as 20% have been shown to yield equally or even more accurate results than studies with response rates of 60–70% [43], the current response rate was deemed to be satisfactorily high.

The sample comprised 56.2% of females, 42.5% of males, and 0.6% who identified with other gender identities. Approximately 6 out of 10 participants were 11–14 years old and lived in neighborhoods of medium socio-economic status (Table 1). The great majority of participants (99.5%) reported using at least one digital media. Website browsing, YouTube, Instagram, and TikTok were the most popular media, being used by more than 80% of the participants. Although the minimum age for signing up to Instagram and TikTok is 13 years [44, 45], 72.9% and 65.6% of the participants younger than this age reported using these platforms, respectively. Total daily use of digital media accounted for an average of 5.5 hours and a median of 5.0 hours (SD = 3.2).

Table 1. Characteristics of the participants (n=1542).

Characteristic

Number of participants

Percentage of participants (%)

Gender

Female

Male

Other

No response

866

655

9

12

56.2

42.5

0.6

-

Age

11-14

15-19

No response

909

619

14

58.9

40.1

-

Socio-economic status

Low

Medium

High

No response

214

938

364

26

13.9

60.8

23.6

-

Digital media use (*)

Any digital media

Website browsing

YouTube

Instagram

TikTok

Twitter

Twitch

Facebook

Snapchat

1535

1438

1335

1273

1226

434

322

148

113

99.5

99.3

87.3

84.0

80.1

28.3

21.0

9.7

7.4

Note: Participants with missing data were not considered in the calculation of percentages. Socio-economic status was estimated based on the score of the neighborhood of residence in the national socio-economic status index [79].

Self-reported exposure to digital food marketing

The percentage of participants who reported seeing a food or beverage advertisement on digital media in the week prior to the survey corresponded to 87.6%, with 39.1% reporting daily exposure (Table 2). Instagram, YouTube, and TikTok were the platforms where participants reported being most frequently exposed to food and beverage advertisements. As shown in Table 2, the percentage of participants who reported seeing an advertisement on these platforms ranged from 63.5–72.6%. Less than 10% of the participants recalled seeing food and beverage advertisements on Twitch, Facebook, or Snapchat.

Table 2. Exposure to food and beverage advertisements in digital media. Number and percentage of participants reporting having seen different types of food and beverage advertisements the week prior to the survey (n=1509).

Type of exposure to food and beverage advertisements in digital media

Number of participants

Percentage of participants (%)

Frequency of exposure to advertisements of food and beverages on social media or websites in the week prior to the survey

No exposure

At least once a week

At least once per day

187

590

732

12.4

48.5

39.1

Exposure to advertisements of food and beverages in the previous week on specific digital media in the week prior to the survey

Instagram

YouTube

TikTok

Website browsing

Twitter

Twitch

Facebook

Snapchat

1095

1011

958

456

196

136

74

10

72.6

67.0

63.5

30.2

13.0

9.0

4.9

0.7

Exposure to advertisements of specific products on social media or websites in the week prior to the survey

Fast food

Soft drinks

Savory snacks

Alfajores

Ice-cream

Energy drinks

Flavored water

Chocolates and confectionary

Bakery products

Condiments

Cold cuts or sausages

Cookies or crackers

Flavored milk, yogurt or dairy desserts

Bottled juices or powdered drinks

Breakfast cereals or cereal bars

Marmalades or dulce de leche

Soups or broths

1218

1087

1063

703

636

635

565

559

494

480

449

408

364

324

263

222

184

80.7

72.0

70.4

46.6

42.1

37.4

37.0

32.7

31.8

29.8

27.0

24.1

21.5

17.4

14.7

12.2

Note: Participants with missing data (n=24) or whose self-reported gender identities (n=9) differed from males or females were not included.

When participants were prompted to recall advertisements of specific product categories on digital media, more than 7 of 10 participants reported seeing advertisements of fast food, soft drinks, and savory snacks (Table 2). On the contrary, less than 2 of 10 recalled seeing advertisements of breakfast cereals and cereal bars, marmalade or dulce de leche, and soups or broths.

Associations of self-reported frequency of exposure to digital food marketing with socio-demographic characteristics

As shown in Table 3, self-reported frequency of exposure to digital food marketing was not significantly associated with age, socio-economic status, total social media use (in hours) or TV watching time (hours). The only variable with a significant association was gender. The odds of reporting a higher frequency of exposure to digital food marketing was 21% higher for female than for males, holding constant all other variables.

Table 3. Results of the ordinal logistic regression analyzing the association between socio-demographic characteristics and self-reported frequency of exposure to digital food marketing, expressed as odd-ratios with 95% confidence intervals and p-values.

Characteristic	Odd-ratio	p-value
Gender
Female	1	-
Male	0.79 (0.64-0.95)	0.013
Age range
11-14	1	-
15-19	0.85 (0.70-1.04)	0.125
Socio-economic status
Low	0.94 (0.60-1.17)	0.297
Medium	1.00 (0.79-1.27)	0.997
High	1	-
Total social media use (hours)	0.98 (0.95-1.01)	0.146
TV watching time (hours)	0.93 (0.85-1.01)	0.079

Note: Participants with missing data (n=24) or whose self-reported gender identities (n=9) differed from males or females were not included in the model. The reference in the model was 'No exposure to digital food marketing' compared to 'Less than once a week' and 'At least once a week'. Odd-ratios highlighted in bold were significant at 0.05. The proportional odds assumption was accepted according to Lipsitz goodness of fit test (0. 271) and Pulkstenis-Robinson chi-squared test (p-value=0.537).

Associations between exposure to digital food marketing and food consumption frequency

Table 4 shows results of the ordinal logistic regression models analyzing the associations between exposure to advertisements of specific product categories and consumption frequency. Self-reported exposure to an advertisement of fast food or ultra-processed products on social media or websites was significantly associated with increased likelihood of reporting higher consumption frequency of the category, while controlling for all other variables. The only exception was the category cookies and crackers, whose consumption frequency was not significantly associated with exposure to advertisements on social media or websites in the week prior to the survey (p = 0.988). For the rest of the food categories, the increase in the odds of reporting a higher consumption frequency ranged from 31% for fast food to 114% for energy drinks (Table 4).

Table 4

Results of the ordinal logistic regressions analyzing the association between consumption frequency of specific ultra-processed product categories and self-reported exposure to an advertisement of the food category on digital media in the week prior to the survey and total social media use (hours). Results are expressed as odd-ratios with 95% confidence intervals and p-values.
Product category	Exposure to an advertisement in the week prior to survey (Yes vs. No)	Total social media use (hours)
Soft drinks	1.53 (1.24–1.88)	1.07 (1.05–1.11)
Energy drinks	2.14 (1.67–2.76)	1.09 (1.05–1.13)
Flavored water	1.81 (1.46–2.24)	1.02 (0.99–1.05)
Bottled juices or powdered drinks	1.32 (1.05–1.67)	1.05 (1.02–1.08)
Savory snacks	1.82 (1-48-2.26)	1.07 (1.04–1.10)
Bakery products	1.41 (1.16–1.72)	1.05 (1.02–1.08)
Cookies or crackers	1.18 (0.95–1.45)	1.02 (0.99–1.05)
Alfajores	1.64 (1.35–1.98)	1.07 (1.04–1.10)
Breakfast cereals or cereal bars	1.71 (1.30–2.23)	1.01 (0.98–1.05)
Flavored milk, yogurt or dairy deserts	1.90 (1.53–2.36)	1.05 (1.02–1.08)
Ice-cream	1.35 (1.09–1.68)	1.00 (0.91–1.26)
Chocolates or confectionary	1.86 (1.53–2.27)	1.05 (1.00-1.08)
Cold cuts and sausages	1.37 (1.11–1.69)	1.03 (1.00-1.06)
Ketchup or mayonnaise	1.49 (1.22–1.81)	1.05 (1.02–1.09)
Fast food	1.31 (1.00-1.73)	1.07 (1.03–1.10)
Notes: Participants with missing data (n = 24) or whose self-reported gender identities (n = 9) differed from males or females were not included in the model. The reference in the model for consumption frequency was '0 days' compared to '1 day', and '2–3 days', '4–6 days' and '7 days'. Odd-ratios highlighted in bold were significant at 0.05. The models controlled for gender, age range, socio-economic status and TV watching time. The p-values of the full models and results of the Lipsitz goodness of fit and Pulkstenis-Robinson chi-squared tests are shown in the Supplementary Material Table 1.

For 12 of the 16 categories, total social media use was significantly associated with an increased likelihood of reporting higher consumption frequency. For every hour increase in social media use, the odds of being more likely to report high consumption frequency of fast food or ultra-processed products increased by 3–9%, holding all other variables constant (Table 4). The four categories that did not follow this trend were flavored water (p = 0.241), cookies or crackers (p = 0.150), breakfast cereals and cereal bars (p = 0.487), and ice-cream (p = 0.098).

As shown in Table 5, self-reported exposure to digital food marketing in the week prior to the survey was not significantly associated with consumption frequency of fruits, vegetables, meats, or fish. However, total social media use was significantly associated with a reduced likelihood of reporting higher consumption frequency of fruits and vegetables. For every hour increase in social media use, the odds of being more likely to report high consumption frequency of fruits decreased by 6%, whereas the likelihood of reporting high consumption frequency of vegetables decreased by 4%. For meats and fish, consumption frequency was not significantly associated with total social media use (Table 5).

Table 5

Results of the ordinal logistic regressions analyzing the association between consumption frequency of specific food groups and self-reported exposure to digital food marketing in the week prior to the survey and total social media use (hours). Results are expressed as odd-ratios with 95% confidence intervals and p-values.
Product category	Exposure to digital food marketing in the week prior to the survey (Yes vs. No)	Total social media use (hours)
Fruits	1.27 (0.80-2.00)	0.94 (0.91–0.96)
Vegetables	1.27 (0.81–2.01)	0.96 (0.93–0.99)
Meats	1.11 (0.70–1.76)	1.02 (0.99–1.05)
Fish	1.02 (0.63–1.70)	0.98 (0.95–1.01)
Notes: Participants with missing data (n = 24) or whose self-reported gender identities (n = 9) differed from males or females were not included in the model. The reference in the model for consumption frequency was '0 days' compared to '1 day', and '2–3 days', '4–6 days' and '7 days'. Odd-ratios highlighted in bold were significant at 0.05. The models controlled for gender, age range, socio-economic status and TV watching time. The p-values of the full models and results of the Lipsitz goodness of fit and Pulkstenis-Robinson chi-squared tests are shown in the Supplementary Material Table 2.

The present study explored the role of exposure to digital food marketing among adolescents in Uruguay, a high-income Latin American country. Results showed frequent exposure to digital food marketing: almost 9 of 10 participants reported having seen a food and beverage advertisement on digital media in the week prior to the survey. This percentage is in the upper range of self-reported exposure reported in previous studies [22, 28].

Instagram, YouTube, and TikTok emerged as the media where participants most frequently reported seeing food and beverage advertisements. An in-depth analysis of marketing activities in these platforms can inform the development of regulations to reduce the persuasiveness of digital food marketing [9, 46]. In this sense, previous studies have reported that Instagram posts include a wide range of cues that trigger systematic and heuristic processing in an attempt to maximize persuasiveness, including interactions with users, videos and music, contests and raffles, price promotions, memes or pranks as well as references to specific characteristics of the products [47–52]. Recent studies have also explored how influencers promote unhealthy foods and beverages on YouTube and TikTok [53–55]. However, an in-depth characterization of the prevalence and power of food and beverage advertisements on these platforms is lacking.

Advertisements of fast food, soft drinks, and savory snacks were the most frequently recalled by participants, consistent with results from a previous qualitative study with Uruguayan adolescents [27]. Several studies conducted in different countries around the globe have also identified these categories as the most frequently advertised across different media, including social media platforms [21, 24, 26, 28, 56–58] and TV [59]. Further, these categories have been found to be frequently marketed to adolescents [60], and to contribute to the largest proportion of the expenditure on food marketing targeted at children and adolescents [61].

Self-reported frequency of exposure to food digital marketing was not significantly associated with participants' age and socio-economic status. The lack of age differences is particularly concerning given that several social media platforms have an age requirement of 13 years [62]. Results from the present study indicate that many platforms lack adequate protection for children and adolescents, as participants below the minimum age requirement frequently reported using social media and seeing unhealthy food advertisements.

Total social media use was not significantly associated with self-reported frequency of exposure to digital food marketing. Although this result may seem counterintuitive, it can be explained considering the high prevalence of food and beverage advertising in digital media. Studies have shown that users are exposed to several advertisements of unhealthy foods and beverages per hour across social media platforms and websites [20, 26, 58].

Gender was the only socio-demographic variable with a significant association with self-reported frequency of exposure to digital food marketing. Females had greater odds of being frequently exposed to digital food marketing compared to males. This result differs from a previous study reporting similar exposure to digital food marketing for male and female adolescents in Canada [24]. Considering that gender has been reported to mediate the effect of marketing on food preferences [63], further research is needed to investigate gender differences in adolescents' experiences with digital food marketing across countries and cultures.

Results from the present work contribute to the growing body of literature reporting associations between exposure to digital food marketing and eating habits [22, 23, 25, 31, 39]. Self-reported exposure to advertisements of fast food and ultra-processed products was associated with increased likelihood of reporting high consumption frequency. As far as can be ascertained, this is the first study to report associations between exposure to advertisements and consumption frequency of specific product categories. Previous studies have found general associations between exposure to digital food marketing and consumption of unhealthy foods and beverages [22, 23, 25, 31, 39]. In the present work, mere self-reported exposure to digital food marketing was associated with higher consumption frequency of ultra-processed products, as postulated by the Hierarchy of Unhealthy Food Promotion Effects [3, 64]. This contradicts results from the study conducted by Baldwin et al., who reported that intake of unhealthy foods was only associated with interaction with advertisements and not with passive exposure [31].

The present study cannot establish causality between exposure to digital food marketing and eating habits. Higher consumption frequency of a specific product category can lead to attentional bias, making adolescents more likely to attend to and recall advertisements of products they like and frequently consume [65]. In addition, targeting algorithms in social media can trigger more frequent exposure to advertisements of specific products because of higher user interest [9]. Reverse causality (i.e., higher consumption frequency triggering higher exposure to advertisements) is also worrying from a public health perspective, as it can contribute to reinforce unhealthy food preferences and social norms around consumption of fast food [17].

The present work shows an association between total social media use and consumption frequency of fast food and several categories of ultra-processed products, while controlling for self-reported exposure to advertisements of the specific category. This result may be explained by exposure to food advertisements without conscious awareness, as suggested by the Reactivity to Embedded Food Cues in Advertising Model [66]. Adolescents may process advertisements implicitly, creating positive emotional and symbolic associations with products [46]. In this sense, exposure to digital food marketing and user-generated content of ultra-processed products may contribute to create social norms around food, reinforcing adolescents' unhealthy dietary habits [25]. Results from the current research also show that total social media use is associated with reduced consumption frequency of fruits and vegetables. Further research should delve into how social norms around food are created in adolescence and whether social media use may make adolescents prioritize heavily advertised unhealthy food options characterized by a high reward value (e.g. ultra-processed products), while downplaying the importance of healthier alternatives.

Together, the findings from this study highlight the need to reduce adolescent exposure to digital food marketing featuring unhealthy foods. Considering that social media platforms have not adopted restrictions on unhealthy food marketing [67], mandatory policies are needed. This regulatory approach is recommended by the World Health Organization [9, 68], and supported by a systematic review suggesting that marketing regulation has the potential to reduce purchases of foods high in sugars, saturated fat, and sodium [69]. Given that adolescents are influenced by digital food marketing [70], even when it is not specifically targeted at them, the focus should arguably be placed on the general population instead of a sole focus on vulnerable consumer segments. A total ban of digital food marketing featuring unhealthy food seems warranted [71]. Empowering adolescents on the topic can also contribute to improve their ability to resist the persuasiveness of digital food marketing [72], and may increase political will for action [73, 74]. An interesting avenue for further research, therefore, is the use of co-creation for developing strategies to achieve this objective [75].

Strengths and limitations

The main strength of this study is its novelty, as it addressed an emerging topic on an underrepresented population in the food marketing literature [29, 30]. From a methodological point of view, the study has several strengths related to the large sample size, the recruitment method, and the specificity of the questions on digital food marketing recall.

A number of limitations should also be noted. First, the study was based on self-reported data of exposure to food and beverage advertisements. Thus, the study only captured the advertisements adolescents attended to and recalled. Actual exposure to advertisements may be underreported considering that advertisements may be embedded in entertainment or media content generated by celebrities or influencers, which can be difficult to distinguish for adolescents [76]. Adolescents also self-reported their consumption frequency, which may be susceptible to social desirability bias. Indeed, a relevant share of adolescents tends to underreport their energy intake [77, 78]. Finally, the study did not control for exposure to other types of marketing, such as TV or out-of-home advertising.

The present work sheds light on the extent to which adolescents in an emerging Latin American country report being exposed to digital food marketing featuring unhealthy foods and beverages. It makes a significant contribution to the literature by reporting associations between self-reported exposure to advertisements on social media and websites, social media use, and consumption frequency of fast food and ultra-processed products. The study is the first to report associations between exposure to advertisements and consumption of specific product categories. Although causality cannot be established, results are consistent with the Hierarchy of Unhealthy Food Promotions Effects and Reactivity to Embedded Food Cues in Advertising Models. The findings stress the need to implement regulatory approaches to reduce exposure to digital marketing of unhealthy foods and beverages and enable adolescents to adopt and maintain healthy eating habits.

Ethics approval and consent to participate

The full study protocol was approved by the Ethics Committee of the School of Chemistry of Universidad de la República, Uruguay (Protocol 101900-000608-20). Approval was also obtained from the National Administration of Public Education, as well as from the principal of each of the participating high schools. Written informed consent was obtained from an adult in charge and participants provided informed assent to participate in the study.

Consent for publication

Not applicable.

Availability of data and materials

The datasets used during the current study are available from the corresponding author on reasonable request.

Competing Interests

The authors declare no competing interests.

Funding

Financial support was obtained from Comisión Sectorial de Investigación Científica (Universidad de la República, Uruguay) and Espacio Interdisciplinario (Universidad de la República, Uruguay). The funders were not involved in the conceptualization, design, data collection, analysis, decision to publish, or preparation of the manuscript.

Authors' contributions

Gastón Ares: Conceptualization, Methodology, Investigation, Formal analysis, Writing – Original draft, Writing – Review & Editing, Supervision, Funding acquisition.

Lucía Antúnez: Conceptualization, Methodology, Investigation, Writing – Review & Editing.

Florencia Alcaire: Conceptualization, Methodology, Investigation, Writing – Review & Editing.

Virginia Natero: Conceptualization, Methodology, Investigation, Writing – Review & Editing.

Leandro Machín: Conceptualization, Methodology, Investigation, Writing – Review & Editing.

Vanessa Gugliucci: Conceptualization, Methodology, Investigation, Writing – Review & Editing.

Carolina de León: Conceptualization, Methodology, Writing – Review & Editing.

Tobias Otterbring: Conceptualization, Methodology, Writing – Original draft, Writing – Review & Editing.

Acknowledgements

The authors are thankful to school principals and staff, as well as to Gloria Canclini and Leticia Bentancor from Administración Nacional de Educación Pública, without whom the study would not have been possible. Participants and their parents are also thanked for the essential and generous contributions to the study.

Mialon M. An overview of the commercial determinants of health. Global Health. 2020;16:1–7.
Kickbusch I, Allen L, Franz C. The commercial determinants of health. Lancet Glob Health. 2016;4:e895–6.
Kelly B, King L, Chapman K, Boyland E, Bauman AE, Baur LA. A hierarchy of unhealthy food promotion effects: Identifying methodological approaches and knowledge gaps. Am J Public Health. 2015;105:e86–95.
Felix R, Rauschnabel PA, Hinsch C. Elements of strategic social media marketing: A holistic framework. J Bus Res. 2017;70:118–26.
Nielsen. Global annual marketing report 2022. New York; 2022.
Vollero A, Valentini C. Social media and consumer power. Opportunities and challenges for digital marketing activities. In: Niininen O, editor. Contemporary Issues in Digital Marketing. New York, NY: Routledge; 2022. p. 105–16.
World Cancer Research Fund International. Building momentum: Lessons on implementing robust restrictions of food and non-alcoholic beverage marketing to children. World Cancer Research Fund International; 2020.
Appel G, Grewal L, Hadi R, Stephen AT. The future of social media in marketing. J Acad Mark Sci. 2020;48:79–95.
World Health Organization. Tackling food marketing to children in a digital world: trans-disciplinary perspectives Children’s rights, evidence of impact, methodological challenges, regulatory options and policy implications for the WHO European Region. 2016.
Kannan PK, Li H. Digital marketing: A framework, review and research agenda. International Journal of Research in Marketing. 2017;34:22–45.
Buchanan L, Kelly B, Yeatman H, Kariippanon K. The effects of digital marketing of unhealthy commodities on young people: A systematic review. Nutrients. 2018;10:1–19.
Montgomery K, Chester J, Nixon L, Levy L, Dorfman L. Big Data and the transformation of food and beverage marketing: undermining efforts to reduce obesity? Crit Public Health. 2019;29:110–7.
Valkenburg PM, Piotrowski JT. Plugged in: How media attract and affect youth. New Haven, CT: Yale University Press; 2017.
Shankleman M, Hammond L, Jones FW. Adolescent Social Media Use and Well-Being: A Systematic Review and Thematic Meta-synthesis. Adolesc Res Rev. 2021;6:471–92.
Lowe CJ, Morton JB, Reichelt AC. Adolescent obesity and dietary decision making—a brain-health perspective. Lancet Child Adolesc Health. 2020;4:388–96.
Van Leijenhorst L, Zanolie K, Van Meel CS, Westenberg PM, Rombouts SARB, Crone EA. What Motivates the Adolescent? Brain Regions Mediating Reward Sensitivity across Adolescence. Cerebral Cortex. 2010;20:61–9.
Buijzen M, Van Reijmersdal EA, Owen LH. Introducing the PCMC Model: An Investigative Framework for Young People’s Processing of Commercialized Media Content. Communication Theory. 2010;20:427–50.
Pechmann C, Levine L, Loughlin S, Leslie F. Impulsive and Self-Conscious: Adolescents’ Vulnerability to Advertising and Promotion. Journal of Public Policy & Marketing. 2005;24:202–21.
Neufeld LM, Andrade EB, Ballonoff Suleiman A, Barker M, Beal T, Blum LS, et al. Food choice in transition: adolescent autonomy, agency, and the food environment. The Lancet. 2022;399:185–97.
van der Bend DLM, Jakstas T, van Kleef E, Shrewsbury VA, Bucher T. Adolescents’ exposure to and evaluation of food promotions on social media: a multi-method approach. International Journal of Behavioral Nutrition and Physical Activity. 2022;19:74.
Elliott C, Truman E, Black JE. Tracking teen food marketing: Participatory research to examine persuasive power and platforms of exposure. Appetite. 2023;186:106550.
Gascoyne C, Scully M, Wakefield M, Morley B. Food and drink marketing on social media and dietary intake in Australian adolescents: Findings from a cross-sectional survey. Appetite. 2021;166:105431.
Evans R, Christiansen P, Masterson T, Pollack C, Albadri S, Boyland E. Recall of food marketing on videogame livestreaming platforms: Associations with adolescent diet-related behaviours and health. Appetite. 2023;186:106584.
Amson A, Pauzé E, Remedios L, Pritchard M, Potvin Kent M. Adolescent exposure to food and beverage marketing on social media by gender: a pilot study. Public Health Nutr. 2023;26:33–45.
Qutteina Y, Hallez L, Raedschelders M, De Backer C, Smits T. Food for teens: how social media is associated with adolescent eating outcomes. Public Health Nutr. 2021;:1–13.
Potvin Kent M, Bagnato M, Remedios L, Soares Guimarães J, Gillis G, Soto C, et al. Child and adolescent exposure to unhealthy food marketing across digital platforms in Canada. BMC Public Health. 2024;24:1740.
Ares G, Antúnez L, de León C, Alcaire F, Vidal L, Natero V, et al. ‘Even if you don’t pay attention to it, you know it’s there’: A qualitative exploration of adolescents’ experiences with digital food marketing. Appetite. 2022;176:106128.
Demers-Potvin É, White M, Potvin Kent M, Nieto C, White CM, Zheng X, et al. Adolescents’ media usage and self-reported exposure to advertising across six countries: implications for less healthy food and beverage marketing. BMJ Open. 2022;12:e058913.
Qutteina Y, De Backer C, Smits T. Media food marketing and eating outcomes among pre‐adolescents and adolescents: A systematic review and meta‐analysis. Obesity Reviews. 2019;20:1708–19.
Boyland E, McGale L, Maden M, Hounsome J, Boland A, Angus K, et al. Association of Food and Nonalcoholic Beverage Marketing with Children and Adolescents’ Eating Behaviors and Health: A Systematic Review and Meta-analysis. JAMA Pediatr. 2022;176.
Baldwin HJ, Freeman B, Kelly B. Like and share: associations between social media engagement and dietary choices in children. Public Health Nutr. 2018;21:3210–5.
Ares G, Alcaire F, Antúnez L, Natero V, de León C, Gugliucci V, et al. Exposure effects to unfamiliar food advertisements on YouTube: A randomized controlled trial among adolescents. Food Qual Prefer. 2023;111:104983.
Khan T, Abimbola S, Kyobutungi C, Pai M. How we classify countries and people—and why it matters. BMJ Glob Health. 2022;7:e009704.
WHO. Global school-based student health survey. Uruguay 2019 fact sheet. Geneva; 2019.
OPS. Alimentos y bebidas ultraprocesados en América Latina: tendencias, efecto sobre la obesidad e implicaciones para las políticas públicas. Washington, D.C.; 2015.
Köncke F, Berón C, Toledo C, Ceriani F, Iervolino A, Klaczko I, et al. Consumo aparente de alimentos y bebidas en los hogares uruguayos: Una mirada a la realidad nacional y en hogares donde viven niños menores de 5 años. Montevideo: Ministerio de Salud Pública; 2023.
Ares G, Antúnez L, Alcaire F, Vidal L, Bove I. Listening to the voices of adolescents for the design of strategies to promote healthy eating: an exploratory study in a Latin American country. Public Health Nutr. 2021;24:5953–62.
INE. Encuesta de Usos de Tecnologías de la Información. Montevideo: INE; 2019.
Scully M, Wakefield M, Niven P, Chapman K, Crawford D, Pratt IS, et al. Association between food marketing exposure and adolescents’ food choices and eating behaviors. Appetite. 2012;58:1–5.
R Core Team. R: A language and environment for statistical computing. 2022.
Fagerland MW, Hosmer DW. How to test for goodness of fit in ordinal logistic regression models. Stata J. 2017;17:668–86.
Jay M. Package ‘generalhoslem.’ https://cran.r-project.org/web/packages/generalhoslem/generalhoslem.pdf. 2024.
Visser PS, Krosnick JA, Marquette J, Curtin M. Mail surveys for election forecasting? An evaluation of the Columbus Dispatch poll. Public Opin Q. 1996;60:181–227.
TikTok. Guardian’s guide. https://www.tiktok.com/safety/en/guardians-guide#. 2024.
Instagram. About Instagram teen privacy and safety settings. https://help.instagram.com/3237561506542117#. 2024.
van der Bend DLM, Jakstas T, van Kleef E, Shrewsbury VA, Bucher T. Making sense of adolescent-targeted social media food marketing: A qualitative study of expert views on key definitions, priorities and challenges. Appetite. 2022;168 January 2021:105691.
Gugliucci V, Machín L, Alcaire F, Otterbring T, León C de, Natero V, et al. The content of Instagram posts featuring ultra-processed products through the lens of the heuristic-systematic model. Appetite. 2023;181:106393.
Vassallo AJ, Kelly B, Zhang L, Wang Z, Young S, Freeman B. Junk food marketing on instagram: Content analysis. J Med Internet Res. 2018;20:1–11.
Bleakley A, Ellithorpe ME, Jordan AB, Hennessy M, Stevens R. A content analysis of sports and energy drink advertising. Appetite. 2022;174:106010.
Buchanan L, Yeatman H, Kelly B, Kariippanon K. A thematic content analysis of how marketers promote energy drinks on digital platforms to young Australians. Aust N Z J Public Health. 2018;42:530–1.
Horta PM, Rodrigues FT, Dos Santos LC. Ultra-processed food product brands on Facebook pages: Highly accessed by Brazilians through their marketing techniques. Public Health Nutr. 2018;21:1515–9.
Freeman B, Kelly B, Baur L, Chapman K, Chapman S, Gill T, et al. Digital junk: Food and beverage marketing on facebook. Am J Public Health. 2014;104:e56–64.
Potvin Kent M, Bagnato M, Amson A, Remedios L, Pritchard M, Sabir S, et al. #junkfluenced: the marketing of unhealthy food and beverages by social media influencers popular with Canadian children on YouTube, Instagram and TikTok. International Journal of Behavioral Nutrition and Physical Activity. 2024;21:37.
Potvin Kent M, Bagnato M, Amson A, Remedios L, Pritchard M, Sabir S, et al. #junkfluenced: the marketing of unhealthy food and beverages by social media influencers popular with Canadian children on YouTube, Instagram and TikTok. International Journal of Behavioral Nutrition and Physical Activity. 2024;21:37.
Winzer E, Naderer B, Klein S, Lercher L, Wakolbinger M. Promotion of Food and Beverages by German-Speaking Influencers Popular with Adolescents on TikTok, YouTube and Instagram. Int J Environ Res Public Health. 2022;19:10911.
Elliott C, Truman E, Aponte-Hao S. Food marketing to teenagers: Examining the power and platforms of food and beverage marketing in Canada. Appetite. 2022;173:105999.
Murphy G, Corcoran C, Tatlow-Golden M, Boyland E, Rooney B. See, Like, Share, Remember: Adolescents’ Responses to Unhealthy-, Healthy- and Non-Food Advertising in Social Media. Int J Environ Res Public Health. 2020;17:2181.
Kelly B, Bosward R, Freeman B. Australian Children’s Exposure to, and Engagement With, Web-Based Marketing of Food and Drink Brands: Cross-sectional Observational Study. J Med Internet Res. 2021;23:e28144.
Kelly B, Vandevijvere S, Ng S, Adams J, Allemandi L, Bahena‐Espina L, et al. Global benchmarking of children’s exposure to television advertising of unhealthy foods and beverages across 22 countries. Obesity Reviews. 2019;20:116–28.
Ares G, Alcaire F, Gugliucci V, Machín L, de León C, Natero V, et al. Colorful candy, teen vibes and cool memes: prevalence and content of Instagram posts featuring ultra-processed products targeted at adolescents. Eur J Mark. 2023. https://doi.org/10.1108/EJM-12-2022-0899.
Federal Trade Commission. A review of food marketing to children and adolescentes. Follow-up report. 2012.
Schneble CO, Favaretto M, Elger BS, Shaw DM. Social Media Terms and Conditions and Informed Consent From Children: Ethical Analysis. JMIR Pediatr Parent. 2021;4:e22281.
Castronuovo L, Guarnieri L, Tiscornia MV, Allemandi L. Food marketing and gender among children and adolescents: a scoping review. Nutr J. 2021;20:52.
Kelly B, Boyland E, Tatlow-Golden M, Christiansen P. Testing a conceptual Hierarchy of Effects model of food marketing exposure and associations with children and adolescents’ diet-related outcomes. Public Health Nutr. 2024;27:e10.
Hendrikse JJ, Cachia RL, Kothe EJ, McPhie S, Skouteris H, Hayden MJ. Attentional biases for food cues in overweight and individuals with obesity: a systematic review of the literature. Obesity Reviews. 2015;16:424–32.
Folkvord F, Veling H, Hoeken H. Targeting implicit approach reactions to snack food in children: Effects on intake. Health Psychology. 2016;35:919–22.
Sacks G, Looi ESY. The advertising policies of major social media platforms overlook the imperative to restrict the exposure of children and adolescents to the promotion of unhealthy foods and beverages. Int J Environ Res Public Health. 2020;17:1–11.
WHO. Policies to protect children from the harmful impact of food marketing: WHO guideline. Geneva: WHO; 2023.
Boyland E, McGale L, Maden M, Hounsome J, Boland A, Jones A. Systematic review of the effect of policies to restrict the marketing of foods and non‐alcoholic beverages to which children are exposed. Obesity Reviews. 2022;23:e13447.
Ares G, Antúnez L, Alcaire F, Natero V, Otterbring T. Is this advertisement designed to appeal to you? Adolescents’ views about Instagram advertisements promoting ultra-processed products. Public Health Nutr. 2024;27:e96.
Powell T, Gheera M, Foster D, Balogun B, Conway L. Restrictions on TV and online advertising of less healthy food and drink. The Health and Care Bill [Bill 140 of 2021-22]. London: Department of Health & Social Care; 2021.
Bulger M, Davison P. The Promises, Challenges, and Futures of Media Literacy. Journal of Media Literacy Education. 2018;10:1–21.
Cullerton K, Donnet T, Lee A, Gallegos D. Effective advocacy strategies for influencing government nutrition policy: a conceptual model. International Journal of Behavioral Nutrition and Physical Activity. 2018;15:83.
Russell C, Lawrence M, Cullerton K, Baker P. The political construction of public health nutrition problems: a framing analysis of parliamentary debates on junk-food marketing to children in Australia. Public Health Nutr. 2020;23:2041–52.
Voorberg WH, Bekkers VJJM, Tummers LG. A Systematic Review of Co-Creation and Co-Production: Embarking on the social innovation journey. Public Management Review. 2015;17:1333–57.
Packer J, Croker H, Goddings A-L, Boyland EJ, Stansfield C, Russell SJ, et al. Advertising and Young People’s Critical Reasoning Abilities: Systematic Review and Meta-analysis. Pediatrics. 2022;150.
Savage JS, Mitchell DC, Smiciklas-Wright H, Symons Downs D, Birch LL. Plausible Reports of Energy Intake May Predict Body Mass Index in Pre-Adolescent Girls. J Am Diet Assoc. 2008;108:131–5.
Jones L, Ness A, Emmett P. Misreporting of Energy Intake From Food Records Completed by Adolescents: Associations With Sex, Body Image, Nutrient, and Food Group Intake. Front Nutr. 2021;8.
Centro de Investigaciones Económicas. Índice de nivel socioeconómico. Montevideo: Centro de Investigaciones Económicas; 2023.

No competing interests reported.

Supplementarymaterial.pdf

Associations between exposure to digital food marketing and food consumption in adolescence: A cross-sectional study in an emerging country

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Background

Methods

Participants

Questionnaire

Data collection

Data analyses

Results

Self-reported exposure to digital food marketing

Associations of self-reported frequency of exposure to digital food marketing with socio-demographic characteristics

Associations between exposure to digital food marketing and food consumption frequency

Discussion

Strengths and limitations

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1