Restricting Advertisements for High-Fat,High-Sugar Foods during Children's Television Programs: Attitudes in a US Population-Based Sample

Abstract

Background:

Children in the United States (US) are frequently exposed to advertisements for high-fat, high-sugar (HFHS) foods, which is linked to greater demand for and consumption of those foods. Restricting advertisements for HFHS foods may be a viable obesity prevention strategy—however, public support for policy change is unclear.

Methods:

A secondary analysis of the 2012 Annenberg National Health Communication Survey was conducted. Respondents (N = 1838) were 53.2% female, mean age 50.0 ± 16.5 years. Race/ethnic composition was 76.8% white, 7.4% black, 9.2% Hispanic, and 6.6% other. The percentage of respondents supporting and opposing the regulation was calculated and logistic regression models identified predictors of support. Potential predictors included sociodemographic variables, attitudes towards other health regulations (e.g., smoking bans in public places), and various health behaviors (e.g., fruit and vegetable intake).

Results:

A total of 56.3% of respondents supported or strongly supported advertisement restrictions, while only 8.2% strongly opposed. Approximately 20% had no opinion. Greatest support was found among respondents who supported smoking bans in public settings (OR = 4.3), who supported banning trans fats in restaurants (OR = 1.7), and who were older (OR = 1.7).

Conclusion:

The US adult population appears to have an appetite for restricting HFHS advertising to children, with more than half the populace supporting such a policy in 2012. This may be an opportune time to implement and rigorously evaluate such childhood obesity prevention strategies.

Introduction

Obesity is linked to numerous comorbidities, poorer quality of life, diminished worker productivity, and increased mortality.¹ These health and economic costs begin in childhood, making pediatric obesity prevention a global priority. Almost one in three children in the United States (US) is overweight or obese, with adiposity and comorbidities tracking into adulthood.² For these reasons there is mounting interest in policy and macro-environmental changes for obesity prevention. Potential strategies include food price manipulations, removing vending machines from schools, and posting calorie information in restaurants.³ One strategy receiving heightened attention, and the focus of this report, is the regulation of high-fat, high-sugar (HFHS) product commercials during children's television (TV) programs.

The average child in the United States views 15 TV food advertisements each day, the majority of which are for foods high in fat, sugar, and/or sodium.⁴ Exposure to food marketing prompts children to prefer, request, and consume targeted foods,⁵ and this may contribute to obesity.⁶ In 2012, the fast food industry spent 4.6 billion dollars on advertising to children, outspending the 367 million dollars spent by advertisers of fruit, vegetables, bottled water, and milk combined.⁷ TV was the most common advertising medium, accounting for 88% of all advertising to children in 2012. Developmentally, children four to five years and younger are not able to distinguish commercial content from TV programming, and children seven to eight years and younger are not able to recognize the persuasive nature of advertisements.⁸ For these reasons, restricting food advertisements during children's TV programs has received great scrutiny.

The aim of the current study was to determine the extent to which the US population is supportive of greater regulation of advertisements for HFHS foods during children's TV programs. Knowing the public will regarding such a policy could help to forecast its acceptability and potential traction, and may be a critical contextual factor in implementing American Heart Association nutrition guidelines.⁹ We examined these questions for the year 2012, in a secondary data analysis of a national dataset. A secondary aim was to identify subgroups that were more favorably predisposed than others to this policy.

Methods

Participants

The Annenberg National Health Communication Survey (ANHCS), a nationally representative cross-sectional survey of adults 18 years and older residing in the United States, was analyzed. The ANHCS was conducted annually from 2005 to 2012 and has seven sections that address health behaviors and knowledge, health communication sources, policy preferences and beliefs, and social support, among other topics. We focused on the most recent data from 2012 for this report. The online survey of individuals with landline telephones used list-assisted random digit dialing to assemble a probability-based sample respondent pool. To avoid potential bias that might result from only including Internet users, participants without Internet were provided with a WebTV box, Internet connection, and televisions to complete the online surveys. Questions were made available to household participants to answer on a monthly basis. After assembling a pool of potential phone numbers and verifying postal addresses, those that had been disconnected or belonged to nonresidents were excluded from the sample. Additional details on the ANHCS are provided elsewhere.¹⁰ The total sample size for the analyses in this report was 1838.

Measures

Outcomes

The primary outcome was a single-item question assessing respondents' attitudes towards regulation of TV advertisements for HFHS foods during children's programs. The question read: “Some people say it is time to restrict ads for high-fat and high-sugar foods on children's television shows. Others oppose such restrictions, saying companies should be free to advertise such foods. Do you support or oppose restricting ads for high-fat and high-sugar foods on children's TV shows?” Response options were “Strongly Support,” “Support,” “Oppose,” “Strongly Oppose,” and “No Opinion.” For logistic regression analyses, we recoded these responses as two categories: “Strongly Oppose; Oppose; No Opinion” or “Support; Strongly Support.” These were coded as 0 and 1, respectively. Those with no opinion were categorized along with those expressing opposition, so the “support” category exclusively reflected those in favor of such restrictions.

Predictors

Seventeen variables were examined as potential predictors of the primary outcome. These variables included attitudinal measures (two questions); dietary intake, physical activity, and TV viewing practices (five questions); progeny status (four questions); obesity status (one question, derived from BMI); recent weight loss efforts (one question); and demographic measures (four questions). These measures are described next.

Attitudinal measures

Two questions assessed respondents' attitudes towards regulation of public health policies. These questions were: “Some states make it illegal to smoke in all workplaces, restaurants, and bars. Do you support or oppose banning smoking in all workplaces, restaurants, and bars in your state?” and, “At least one city has stopped restaurants from serving foods made with trans fats. Trans fats are thought to be related to heart disease. Do you support or oppose banning the use of trans fats in restaurants in your state?” Response options for both questions were “Strongly Support,” “Support,” “Oppose,” “Strongly Oppose,” and “No Opinion.” For the purpose of our logistic regression analyses, these variables were dichotomized as 0 = “Strongly Oppose; Oppose; No Opinion” vs. 1 = “Support; Strongly Support.”

Fruit and vegetable intake, physical activity, and TV viewing practices

Two open-ended questions addressed fruit and vegetable consumption: “In the past week, on average, how many servings of fruit did you eat or drink per day? Please include 100% fruit juice, and fresh, frozen, or canned fruits.” The second question asked, “In the past week, on average, how many servings of vegetables did you eat or drink per day, not counting potatoes? Please include green salad, 100% vegetable juice, and fresh, frozen, or canned vegetables.” In order to capture physical activity behavior, the following question was asked: “During an average week, how many days do you exercise?” Regarding screen time, the following two open-ended questions were asked: “On a typical weekday, Monday through Friday, about how many hours do you watch television each day?” and “On a typical weekend, including both Saturday and Sunday combined, about how many total hours do you watch television?” For our logistic regression analyses, scores were dichotomized based on a median split. Fruit intake scores were dichotomized as 0 = ≤1 servings/day vs. 1 = >1 serving per day; vegetable intake scores were dichotomized as 0 = ≤2 servings/day vs. 1 = >2 servings/day; exercise was dichotomized as 0 = ≤2 days per week vs. 1 = >2 days per week; weekday TV viewing time was dichotomized as 0 = ≤4 hours/day vs. 1 = >4 hours per day; and weekend TV viewing time was dichotomized as 0 = ≤6 hours/day vs. 1 = >6 hours/day.

Progeny status

Four questions asked whether the respondent had children in distinct age strata: younger than 2 years, 2 to 5 years, 6 to 12 years, and 13 to 17 years. Response options were “Yes” or “No.”

Obesity status and weight loss attempts

Respondents self-reported their weights and heights, from which BMI (kg/m²) was computed. We classified respondents as nonobese (BMI <30) or obese (BMI ≥30). One question was used to assess recent weight loss history: “During the past 30 days, have you tried to lose weight?” Response options were “Yes” or “No.”

Demographic measures

The following demographic measures were obtained: sex, age, race/ethnicity (white, black, Hispanic, other), marital status (married, single, divorced, widowed, separated), educational attainment (< high school, high school, some college, bachelor's degree or higher), and income status (yearly household income). For the logistic regression analyses, age was dichotomized as 0 = <52 years vs. 1 = ≥52 years, based on median split. Race was categorized as 0 = white vs. 1 = nonwhite, education was categorized as 0 = ≤ high school vs. 1 = > high school, and income was categorized as 0 = <$25,000 vs. 1 = ≥$25,000. This latter categorization was based on cutoffs for federal poverty level for a family of four.

Data Analytic Plan

Descriptive statistics are presented as means and standard deviations for continuous measures and percentages for categorical measures. For the primary question, we computed the percentage of respondents who reported that they strongly oppose, oppose, support, or strongly support regulation of advertisements for HFHS foods during children's programs as well as those with no opinion. For all response items, less than 1% of respondents refused to answer, so those refusals were removed from the sample.

Binary logistic regression models were used to identify predictors of respondent support (vs. no support/no opinion) of restricting HFHS food advertisements. We tested each of the 17 predictors described previously. First, we tested separate models for each predictor variable, reporting the β coefficient, standard error of β (β_SE), the Wald statistics, the p-value, the odds ratio (OR), and the 95% confidence interval (CI). Second, we applied a backward elimination logistic regression model, with alpha = 0.05, that initially included all predictors in the full model. This latter approach was used as a “stay criterion” to identify significant predictors when all variables were modeled together in a multivariate manner. All significance tests were two-tailed and with alpha = 0.05, using SPSS software.

Results

Descriptive characteristics are presented in Table 1. Respondents (N = 1838) were 53.2% female with a mean age of 50.0 ± 16.5 years. Race/ethnic composition was 76.8% white, 7.4% black, 9.2% Hispanic, and 6.6% other. Responses indicate 22.7% strongly support, 33.6% support, 16.1% oppose, and 8.2% strongly oppose restricting HFHS advertisement to children, while 19.5% had no opinion. Thus, 56.3% of respondents supported or strongly supported advertisement restrictions.

Table 1.

Participant Characteristics 2012, N = 1838

Sex (% female)	53.2
Race/ethnicity (%)
White	76.8
Black	7.4
Hispanic	9.2
Other	6.6
Education level (%)
Less than high school	8.2
High school	28.0
Some college	31.1
College degree or higher	32.8
Marital status (%)
Married	59.7
Single^a	29.5
Divorced	9.3
Separated	1.5
Income status (%)
Less than $25,000	17.2
$25,000 or higher	82.8
Parent status (% Yes)
Has a child younger than 2 years	3.8
Has a 2–5 year-old child	8.9
Has a 6–12 year-old child	15.4
Has a 13–17 year-old child	15.8
Mean (SD) age (yrs)	50.0 (16.5)

Single indicates those who are never married, widowed, or living with partner.

A total of 1261 participants were included in the regression model. The main sources of missingness were due to two variables: the single question assessing weekly exercise habits, and respondent BMI. With respect to the former, 369 participants were not asked the question because they had replied “no” to a screener question that asked if the respondent exercises at least once per week. With respect to BMI, 216 respondents were missing data for reasons that were not specified. For all other variables in our regression models, less than 1% of the responses were missing.

Bivariate logistic regression analyses indicated that three predictor variables showed significant associations with support for regulating TV advertisements (Table 2). The odds of supporting restrictions for HFHS advertisements is 4.3 times higher for respondents who supported smoking bans in public settings compared to those who did not (95% CI: 3.1, 6.0) and 1.7 times higher for respondents who supported banning trans fats in restaurants compared to those who did not (95% CI: 1.3, 2.1). Odds were also higher for older adults compared to younger adults (OR = 1.7, 95% CI: 1.3, 2.2).

Table 2.

Bivariate Associations of Demographic and Attitudinal Variables and Support for Children's TV Advertising Restriction, N = 1261

							95% CI for EXP(B)
Variables	B	SE	Wald	df	Sig	OR	Lower	Upper
Sex	0.022	0.126	0.030	1	0.863	1.0	0.8	1.3
Age	0.513	0.137	14.100	1	0.000	1.7^*	1.3	2.2
Education	0.178	0.141	1.595	1	0.207	1.2	0.9	1.6
Ethnicity	0.087	0.153	0.325	1	0.569	1.1	0.8	1.5
Income	−0.269	0.183	2.157	1	0.142	0.8	0.5	1.1
Children 0–1 years	−0.489	0.367	1.777	1	0.183	0.6	0.3	1.3
Children 2–5 years	−0.071	0.238	0.088	1	0.767	0.9	0.6	1.5
Children 6–12 years	0.241	0.189	1.625	1	0.202	1.3	0.9	1.8
Children 13–17 years	−0.295	0.176	2.818	1	0.093	0.7	0.5	1.1
Obesity status	−0.041	0.137	0.091	1	0.763	1.0	0.7	1.3
Supports smoking bans	1.456	0.172	71.741	1	0.000	4.3^*	3.1	6.0
Supports trans fat bans	0.512	0.126	16.515	1	0.000	1.7^*	1.3	2.1
Fruit intake	0.069	0.132	0.270	1	0.604	1.1	0.8	1.4
Vegetable intake	0.112	0.142	0.622	1	0.430	1.1	0.8	1.5
TV time weekdays	0.003	0.141	0.000	1	0.986	1.0	0.8	1.3
TV time weekends	−0.050	0.138	0.133	1	0.715	1.0	0.7	1.2
Physical activity	0.127	0.125	1.041	1	0.308	1.1	0.9	1.4
Weight loss efforts	0.166	0.127	1.716	1	0.190	1.2	0.9	1.5

B, beta coefficient; CI, confidence interval; df, degrees of freedom; EXP(B), exponentiated beta coefficient; OR, odds ratio; SE, standard error; Sig, significance; Wald, Wald statistic.

p ≤ 0.0001

In backwards elimination logistic regression analysis, the same three predictors of greater support emerged: support for banning smoking in public places (95% CI: 1.9, 2.8), support for trans fat bans in restaurants (95% CI: 1.4, 1.9), and being older than 52 years old (95% CI: 1.1, 1.5). Additionally, there was a marginally significant effect for daily servings of vegetables (OR = 1.2, 95% CI: 1.0,1.4).

Discussion

We found that 56.3% of respondents in this population-based survey favored or strongly favored restricting advertisements for HFHS foods during children's TV shows in 2012; only a quarter were opposed. This more recent sample yields findings consistent with and perhaps even more favorable than previous studies. A 2004 population-based survey of US households found that 47.9% of adult respondents said they would support “Prohibiting the advertising and promotion of fast food and less healthy foods marketed to children.”^5,11 In a 2005 population-based phone interview of 1000 Germans aged 14 years and older, 47.7% indicated support for “Restricting advertisements for unhealthy foods, e.g., sweets or chips.”¹² Hence, as of 2012, attitudes towards this policy are relatively positive and may even have shifted more favorably over the past decade. In comparison, less support among US adults was reported for taxing less healthy foods (39.1%), increasing the cost of less healthy foods in schools (44%), and removing all school vending machines (35.9%).¹¹

An increasing number of countries have proposed or implemented legislative measures to restrict food marketing to children, including Brazil, Thailand, Korea, Malaysia, Columbia, and Chile.¹³ In 2006, the United Kingdom adopted a legal definition of junk food and banned advertisements for certain types of foods during children's TV programs. The impact of these initiatives requires monitoring and evaluation in order to inform effective policies.¹⁴

We found greater support for restricting advertisements for HFHS foods among adults who also endorse smoking bans and banning the use of trans fats in restaurants. This may reflect an underlying belief about the importance of external environmental influences on health, or a belief that the most effective solutions to health issues require regulation. Ours is one of the few studies to date to examine predictors of public support for specific obesity prevention policies.^11,12,15

One group that is particularly interesting to focus on is those who were undecided on this topic (19.5% in this study). This group may be most amenable to changing their view. Advocates for greater advertisement restrictions could benefit from understanding more about this audience and could guide efforts to shift their opinion on this issue.

Our survey results were agnostic with respect to who should take leadership over advertisement regulation: government or the food industry. In fact, the food industry has taken steps towards self-regulation, although these have been challenged.^4,6,17 The Children's Food and Beverage Advertising Initiative (CFBAI) was formed in 2006, as part of an industry-sponsored National Review Council, in which 16 major food companies pledged to devote at least half of advertising time to healthy food and lifestyle choices, restrict the use of licensed characters to advertise foods high in sugar and fat, and avoid food/beverage advertising in elementary schools.¹⁷ Only 45% to 48% of food ads viewed by children align with CFBAI criteria.¹⁸ The industry-sponsored Children's Advertising Review Unit (CARU) oversees compliance to advertising pledges—the effectiveness of this effort has been debated.^19–21

Despite important policy implications, there are limitations to this study. First, this was a self-report survey study, and responses may have been influenced by social desirability. Second, while our sample was nationally representative, blacks and Hispanics only made up small proportions of respondents. Future research should explore attitudes of racial minorities in greater depth, as these populations are disproportionately exposed to such advertisements.²² Third, only a subset of participants were included in the logistic regression analysis due to missing data. This potentially challenges the extent to which these analyses can generalize back to the 1838 participants included in the primary analysis. To probe this, we examined how the demographic statistics presented in Table 1 would change when restricting analyses to participants with valid scores for the aforementioned exercise question, as well as respondent BMI (i.e., the two variables with the greatest missing data). In fact, the percentages reported in Table 1 changed by no more than ±3.5% and most were within ±1.0%. This gives at least some assurance that the demographic characteristics of those contributing to the regression analysis are representative of the fuller sample. Additionally, this study is limited to opinions on food advertisement exposure via TV programming. Other media such as print and Internet sites were not examined.²³ Lastly, the extent to which the public favors government- versus industry-sponsored regulation of food advertisements could not be determined in this study. Support for innovative strategies, such as requiring equal advertising time for “healthy” and “unhealthy” foods to children should also be explored.

In summary, the US populace appears to have an appetite for restricting advertisements for less healthy foods during children's TV programs. Less than one-quarter of US adults were opposed to restricting HFHS food advertisements during children's TV programs. Approximately 56% were supportive or very supportive, especially in certain audience segments. Advertising restrictions have been highly effective in addressing other behaviors, such as cigarette smoking,^24–26 and restricting HFHS advertisements to children may be an effective strategy for obesity prevention by mediating changes in child eating behavior. However, it is imperative that these strategies be evaluated empirically, whether government or industry sponsored.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Biro

, Wien

. Childhood obesity and adult morbidities. Am J Clin Nutr, 2010; 91:1499S–1505S.

Ogden

, et al. Prevalence of childhood and adult obesity in the United States, 2011–2012. JAMA, 2014; 311:806–814.

Nestle

, Jacobson

. Halting the obesity epidemic: A public health policy approach. Public Health Rep, 2000; 115:12–24.

Harris

, et al. A crisis in the marketplace: How food marketing contributes to childhood obesity and what can be done. Annu Rev Public Health, 2009; 30:211–225.

Dorfman

, Wootan

. The nation needs to do more to address food marketing to children. Am J Prev Med, 2012; 42:334–335.

Guran

, Bereket

. International epidemic of childhood obesity and television viewing. Minerva Pediatr, 2011; 63:483–490.

Yale RUDD Center. Fast Food FACTS. 2013. Available at www.fastfoodmarketing.org. (Last accessed July 1, 2015.)

Wilcox

, Kunkel

, Cantor

, et al. Report of the APA Task Force on Advertising and Children. Available at www.apa.org/pi/families/resources/advertising-children.pdf. (Last accessed July 1, 2015.)

Gidding

, et al. Implementing American Heart Association pediatric and adult nutrition guidelines: A scientific statement from the American Heart Association Nutrition Committee of the Council on Nutrition, Physical Activity and Metabolism, Council on Cardiovascular Disease in the Young, Council on Arteriosclerosis, Thrombosis and Vascular Biology, Council on Cardiovascular Nursing, Council on Epidemiology and Prevention, and Council for High Blood Pressure Research. Circulation, 2009; 119:1161–1175.

10.

Annenberg National Health Communication Survey. ANHCS 2005 User's Guide & Codebook. 2007.

11.

Evans

, et al. Public perceptions of childhood obesity. Am J Prev Med, 2005; 28:26–32.

12.

Hilbert

, Rief

, Braehler

. What determines public support of obesity prevention?. J Epidemiol Community Health, 2007; 61:585–590.

13.

The PolMark Project: Policies an marketing food and beverages to children. Executive Report, July 2010. Available at www.worldobesity.org. (Last accessed March 4, 2014.)

14.

Fussenegger

, Pietrobelli

, Widhalm

. Childhood obesity: Political developments in Europe and related perspectives for future action on prevention. Obes Rev, 2008; 9:76–82.

15.

Goren

, et al. Predicting support for restricting food marketing to youth. Health Aff (Millwood), 2010; 29:419–424.

16.

Harris

, Brownell

, Bargh

. The food marketing defense model: Integrating psychological research to protect youth and inform public policy. Soc Issues Policy Rev, 2009; 3:211–271.

17.

Sharma

, Teret

, Brownell

. The food industry and self-regulation: Standards to promote success and to avoid public health failures. Am J Public Health, 2010; 100:240–246.

18.

Harris

, et al. Redefining “child-directed advertising” to reduce unhealthy television food advertising. Am J Prev Med, 2013; 44:358–364.

19.

Dembek

, Goren

, Schwartz

. Trends in Television Food Advertising to Young People: 2011 Update. Rudd Center for Food Policy and Obesity: New Haven, CT, 2012.

20.

Speers

, Goren

, Schwartz

, et al. Public Perceptions of Food Marketing to Youth: Results of the Rudd Center Public Opinion Poll, May 2008. Rudd Center for Food Policy and Obesity: New Haven, CT, 2008.

21.

Ustjanauskas

, Goren

, Schwartz

, et al. Focus Group with Parents: What Do They Think About Food Marketing to Their Kids?. Rudd Center for Food Policy and Obesity: New Haven, CT, 2010.

22.

Powell

, Wada

, Kumanyika

. Racial/ethnic and income disparities in child and adolescent exposure to food and beverage television ads across the U.S. media markets. Health Place, 2014; 29C:124–131.

23.

Cowburn

, Boxer

. Magazines for children and young people and the links to Internet food marketing: A review of the extent and type of food advertising. Public Health Nutr, 2007; 10:1024–1031.

24.

Harris

, et al. Effects of the 2003 advertising/promotion ban in the United Kingdom on awareness of tobacco marketing: Findings from the International Tobacco Control (ITC) Four Country Survey. Tob Control, 2006; 15:iii26–iii33.

25.

Saffer

, Chaloupka

. The effect of tobacco advertising bans on tobacco consumption. J Health Econ, 2000; 19:1117–1137.

26.

Quentin

, et al. Advertising bans as a means of tobacco control policy: A systematic literature review of time-series analyses. Int J Public Health, 2007; 52:295–307.