Customize your JAMA Network experience by selecting one or more topics from the list below.
Copyright 2007 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2007
To examine exposure to food advertising on television (TV) among children aged 2 through 11 years.
Weighted examination of the distribution of national advertisements (ads) using TV ratings data.
National ads from 170 top-rated TV shows viewed by children aged 2 through 11 years from September 1, 2003, through May 31, 2004.
Sample of 224 083 ads.
Main Outcome Measures
Television nonprogram content time was assessed across 6 mutually exclusive categories that included food products, non–fast food restaurants, fast food restaurants, other products, public service announcements, and TV promotions. Food advertising was assessed according to 7 food categories—cereal, snacks, sweets, beverages, fast food restaurants, non–fast food restaurants, and other food products—and then examined across more detailed categories.
In 2003-2004, 27.2% and 36.4% of children's exposure to total nonprogram content time and product advertising, respectively, was for food-related products. Similar distributions were found by race. Cereal was the most frequently seen food product, making up 27.6% of all food ads. Comparisons with previous studies suggest that, over time, food ads account for a smaller share of the product ads seen by US children.
Children aged 2 through 11 years are exposed to a substantial amount of food advertising through TV, but the dramatic increase in childhood obesity rates during the past few decades was not mirrored by similar changes in food advertising exposure. However, we found evidence of a very recent (2000-2005) upward trend in the amount of exposure to food advertising on TV among US children.
Overweight among US children has increased dramatically over the past few decades, reaching 13.9% for children aged 2 through 5 years and 18.8% for children 6 through 11 years in 2003-2004.1 Overweight children are at increased risk for developing type 2 diabetes mellitus2 and cardiovascular disease risk factors.3 Studies have shown that cardiovascular risk factors present in children are likely to persist into adulthood4-7 and that overweight itself tracks from childhood to adulthood.8,9 The long-term negative health risks associated with obesity are extensive and have been well documented.10
Data suggest that US children consume too much dietary fat and sugar and that fruit and vegetable consumption and micronutrient intake are low compared with dietary recommendations.11-14 Trend data also show that children are increasingly consuming food away from home, fast food, salty snacks, soft drinks, and larger portion sizes.15
Poor dietary practices and overweight among children are related to many factors at the individual, family, and environmental levels, including television (TV) media.16-18 In 2005, American children aged 2 through 11 years watched, on average, 3 hours 19 minutes of TV per day.19 Black children watch 29% and 35% more TV during daytime and prime time hours, respectively, compared with white children.20 Television viewing is hypothesized to contribute to child overweight by displacing physical activity with sedentary activity, because of snacking while viewing, and through the influence of advertisements (ads) for unhealthy food products.21 Television viewing itself is found to be associated generally with poor food consumption patterns,18 particularly when the TV is in use during mealtimes.22 Several studies have shown significant positive associations between increased levels of TV watching and overweight among children,23-27 although some studies found such associations to be inconsistent28 or statistically insignificant.29
Previous content analyses of advertising on children's programming (as will be discussed later in greater detail) have found that food is the single largest product category advertised and often makes up most of the TV ads directed to children. Furthermore, most food products advertised to children are of poor nutritional content, being high in sugar or fat.22,30 Randomized trials have shown that exposure to food advertising significantly influences children's food preferences/choices31-33 and food intake.34,35 Survey data similarly have shown that advertising exposure affects children's food intake36-38 and food purchase requests.38-41
The Institute of Medicine report on food marketing to children and youth42 drew the following conclusions for children aged 2 through 11 years: (1) there is strong evidence that TV advertising influences children's food and beverage preferences and purchase requests; (2) there is strong evidence that TV advertising influences short-term food consumption patterns and moderate evidence that it influences usual dietary intake; and (3) there is strong evidence that exposure to TV advertising is associated with adiposity in children.42
Advertisers of food products clearly recognize the purchasing power and importance of early brand recognition among children.43-45 Children aged 4 through 12 years were estimated to have a personal purchasing power of $30 billion in 2002 (up 400% since 1989) and to have directly influenced $330 billion of adult spending.45 Children are also estimated to have substantial passive influence on parents' purchases.43 The items most frequently purchased by children are sweets, snacks, and beverages,45 and approximately one third of candy, cookies, cereal, juice, and fast food household purchases are influenced by children.46
The present study examines exposure to food advertising among children aged 2 through 11 years based on TV ratings for programs viewed on 170 top-rated shows. Using ratings data to weight each ad, we categorized the distribution of exposure to nonprogram content time by food products, non–fast food restaurants, fast food restaurants, other products, public service announcements (PSAs), and TV promotions. Food advertising was assessed according to 7 categories—cereal, snacks, sweets, beverages, fast food restaurants, non–fast food restaurants, and other food products—and then was examined across 25 detailed food product categories. Finally, we summarized study findings on content analyses from the past 35 years to highlight trends in food advertising directed to children.
Television show ratings from Nielsen Media Research for children aged 2 through 11 years were used to select the following top-rated children's programming: the top 60 broadcast network series, top 60 cable network series, top 30 syndicated series, top 10 network specials, and top 10 cable specials. We acquired monthly ratings from Nielsen Media Research on every nationally aired ad that appeared on these shows from September 1, 2003, through May 31, 2004. Standardized to a 30-second ad, the advertising data totaled 224 083 ads (including product ads, PSAs, and TV promotions) in our sample of top-rated shows viewed by children. Separate ratings data by race were used to examine differences in exposure among black and white children.
The selection of shows yielded a sample drawn from 6 broadcast networks and 4 cable networks covering all time slots and days of the week. Because of higher ratings on cable shows and a higher frequency of TV show airings per week, 86.5% of the sample came from cable networks and 13.5% from broadcast networks. Cable TV exposure occurred predominantly while viewing Nickelodeon (54.1%) and the Cartoon Network (45.7%), with Toon Disney and the Sci Fi Channel making up less than 0.2% combined. The distribution of TV viewing across broadcast networks was as follows: WB (41.5%), ABC (20.2%), FOX (13.6%), UPN (12.0%), NBC (10.6%), and CBS (2.1%). Cable and broadcast shows that fell below the 60-show cutoff were viewed, on average, by less than 1.4% and 1.0%, respectively, of children.
The use of ratings data allowed us to distinguish actual exposure rather than give all ads equal weight based simply on the airing of the ad. Ratings provide an estimate of the percentage of households with TVs watching a program or ad during a specified time slot. This study used targeted ratings points, which estimate the reach and frequency of advertising to 2- through 11-year-old children. For example, a commercial with 80 targeted ratings points per month was estimated to have been seen an average of one time by 80% of 2- through 11-year-olds during that month. The monthly targeted ratings points were aggregated across the 9-month sampling period and were used to weight each ad for our full sample and by race.
Nielsen Media Research assigns each product ad a product classification code (PCC) that identifies it with a product category. Each of the product categories has a corresponding aggregated major group PCC and industry group PCC, along with a PCC description. The structure of PCCs used by Nielsen Media Research is based on that used by the Publishers Information Bureau.47 In our sample, the food ads fell into 85 product categories made up of 544 distinct food product brands. The product restaurants and fast food restaurants (defined by quick-service restaurants such as McDonald's and Burger King) are single categories at both the product and group levels. Several distinct product categories have common product groups. For example, the single product categories of candy, candy bars, and chewing gum all have a corresponding product group category of “candy and gum.” The PCC codes, descriptions, and major group categories provide a clear characterization of items that are cereal, sweets, snacks, or beverages.
We examined the distribution of exposure to total nonprogram content time across 6 mutually exclusive categories that included food products, non–fast food restaurants, fast food restaurants, other products, PSAs, and TV promotions. We then examined exposure to total food-related advertising according to 7 broad categories that included cereal, snacks, sweets, beverages, fast food restaurants, non–fast food restaurants, and other food products. The distinct product categories that fell into the 7 food categories are listed in the legend to the Figure. We also examined the distribution of food advertising across the top 25 of the 85 disaggregated food product categories; these accounted for 97.5% of all food product advertising. Finally, we compared our findings with previous study findings on content analyses of advertising on children's programming from the past 35 years.
Distribution of food product advertising exposure among children aged 2 through 11 years. Cereal includes cereal and oatmeal; sweets include breath mints candy, candy bar, chewing gum, cookie dough, cookies, cupcakes, frozen novelties, prepared gelatin, gelatin mix, ice cream, ice cream novelties, pastry, prepared pudding, pudding mix, and snack cakes; snacks include crackers, nuts, popping corn, potato chips, snack bars, snacks, and tortilla chips; beverages include bottled water, cocoa mix, coffee, regular and diet soft drinks, drink mix, isotonic drinks, noncarbonated drinks, fruit drinks, fruit juices, iced coffee and tea, milk, and vegetable juice; and other includes artificial sweetener, baby foods, baking mix, beans, beef, bread, buns, cheese, dough, entrees (prepared), entrees (frozen), food products, French toast (frozen), hot dogs, infant formula, luncheon meat, mayonnaise, mustard, nondairy creamer, pasta dinners, pasta sauce, peanut butter, pickles, pizza (frozen), preserves, rice mix, salad dressings (bottled and mixed), salsa sauce, barbeque sauce, steak sauce, Worcestershire sauce, sausage, soup (condensed), soup, sour cream, syrup, canned vegetables, waffles (frozen), whipped topping, and yogurt. Fast food restaurants and non–fast food restaurants are single categories.
Table 1 shows that, based on our national sample, 27.2% of all nonprogram content time exposure among children aged 2 through 11 years was for food-related products. Roughly one quarter (24.3%) of total nonprogram content time exposure was for TV promotions, whereas just 0.8% consisted of PSAs. Excluding TV promotions and PSAs, total food-related advertising made up 36.4% of total product advertising seen by children. Based on our separate ratings data by race (data not shown), we found a similar distribution of exposure to food advertising by race; food advertising made up 27.2% of nonprogram content time among both black and white children and 36.2% and 36.4% of total product advertising, respectively, by race. However, given that black children watched, on average, more hours of TV per day, they viewed correspondingly more food ads.
The Figure, an examination of the distribution of total food advertising, shows that cereal was the most frequently viewed food product category, accounting for 27.6% of all food advertising exposure, while 17.7% of ads viewed were for sweets; 12.2%, for snacks; and 8.8%, for beverages. Ads for fast food restaurants also represented a substantial portion of food-related advertising, accounting for 12.0% of all food ads seen by children.
Table 2 gives the distribution of food advertising at a more disaggregated level, highlighting the distribution rates for the top 25 food product categories. Cereal was the top food product category (making up 27.6% of food ads), followed by fast food restaurants (12.0%), and the single snack category (8.3%). Among the top sweets, the candy product category made up 7.0% of all food advertising, followed by cookies (3.2%), chewing gum (2.2%), candy bars (2.1%), and pastry (1.6%). When we examined beverages at a disaggregated level, 4.9% of all food ads seen by children were for fruit drinks. Advertising of regular soft drinks and sports (isotonic) drinks did not appear to be targeted toward children because they each represented less than 1% of food ads seen by children aged 2 through 11 years. Only 1.1% of the food ads were for milk; however, yogurt made up 5.8%. Fruits and vegetables constituted less than 0.01% of food advertising. The combination of prepared and frozen entrees plus frozen waffles and French toast totaled 6.9% of all food products and made up almost half of the food items in the “other” food category (Figure).
Table 3 provides a comparison of study findings from content analyses of advertising on children's programming during the past 35 years. The table describes the study sampling methods, sample size and time frame, percentage of food ads of the total nonprogram content time and of the product advertising, and percentage of cereal and sweets and snack ads of the total food commercials. To provide comparable estimates across studies, we often drew on data provided in the respective studies to calculate the information presented in Table 3. For the cases in which the information was not reported in the paper and in which there were insufficient data to derive the estimates ourselves, that information is indicated as not available.
Studies from the early 1970s until 2000 consistently found that food advertising made up roughly one half or more (48%-87%) of all product advertising to children—an amount higher than the 36% found in the present study. The studies with samples drawn exclusively from Saturday morning programming tended to find higher rates of food advertising among the product ads (57%-87%).
Advertising data by network showed that food advertising on children's programming in 1990 made up 77% of product advertising on broadcast networks vs only 40% on cable networks.53 Just over a decade later, our study has found the opposite pattern: food ads made up 37% of product ads on cable TV, but only 29% on broadcast networks. In the 1970s, children's programming was available primarily on broadcast TV. With the emergence of cable TV and channels such as Nickelodeon offering full-day children's programming, we have witnessed a shift in the share of child viewership toward cable TV.58 By the mid-1990s, cable TV accounted for approximately 80% of children's network viewing.59
Over time, cereal has consistently remained the single food product most heavily advertised to children; however, its share of total food advertising appears to have fallen. Prior to 2000, cereal accounted for 31% to 49% of all food advertising. The present study found that cereal made up 28% of all food ads, and another recent study57 found that cereal constituted only 15% of food ads, although the latter study's target sample was children aged 6 through 11 years. Our finding that ads for sweets and snacks accounted for 30% of all food ads is in the midrange of the study findings spanning the past 35 years (Table 3).
Although our study found that food ads made up a smaller proportion of ads compared with previous study findings, we also assessed changes in children's TV viewing time and changes in the number and length of ads per hour on children's programming to fully determine whether the absolute amount of exposure to food advertising among children has changed over time. Table 4 shows that children's daily hours of TV viewing trended downward from 3 hours 49 minutes a day in the mid-1980s to just under 3 h/d in 1998-1999; since then, however, the amount has risen to a mean of 3 hours 19 minutes a day, which is roughly equal to the mean daily viewing time in 1990.
The number of ads per hour has increased during the past few decades because of an increasing number of shorter ad segments. While 95% of network commercials were 30 seconds long in 1980, 54% of the commercials on network TV in 2004 were 30 seconds long, 39% were 15 seconds, 6% were 60 seconds, and 1% were 90 seconds.60 Although the absolute number of TV ads has increased, it is not clear whether total commercial time per hour has changed.
During the 1970s through 1984, the Federal Communications Commission adopted policies to restrict advertising during children's programming to 12 min/h on weekdays and 9.5 min/h on weekends.61 During the 1980s, these limits were deregulated. The Children's Television Act of 1990 now limits commercial time on children's programming to 12 min/h and 10.5 min/h on weekdays and weekends, respectively.61
Several studies have provided evidence of the extent of nonprogram content time on children's TV. Early studies showed that approximately 20% (12 min/h) of children's programming time during the 1970s consisted of nonprogram content time, with commercial product advertising time averaging 9 to 10 min/h, depending on the network and programming time slot.49,62-64 Generally, commercial time trended downward during the 1970s under regulation of the National Associations of Broadcasters code.65 During the 1980s, total nonprogram time on children's Saturday morning programming was reported to be 11.5 to 11.8 min/h,51,66 with commercials on Saturday morning broadcast networks accounting for 8.16 min/h, although this value was higher on weekdays.51 Study results from 1990 found that advertising time varied substantially by network type: total nonprogram content (and commercial ad time) of 13 minutes 26 seconds (9 minutes 37 seconds) per hour on independent networks, 12 minutes 9 seconds (10 minutes 5 seconds) per hour on broadcast networks, and just 10 minutes 38 seconds (6 minutes 48 seconds) per hour on cable networks.53 An additional study based on 1993 data found that nonprogram content time constituted 13 minutes 30 seconds per hour, with 16% (9 minutes 36 seconds) of children's viewing time made up of commercials.55
Table 5 presents a breakdown of total nonprogram content time by commercial ads, PSAs, and TV promotions for the years 2000 and 2005 during Saturday morning and after-school and early evening weekday programming on selected cable networks. In 2000, total nonprogram content time on these stations on Saturday morning and from 4 to 8 PM on weekdays averaged about 11.5 minutes and increased to just under 13 min/h in 2005, an increase of 11% to 12%. In 2000, total commercial ad time averaged 8 minutes 2 seconds per hour during Saturday morning programming and 8 minutes 12 seconds per hour from 4 to 8 PM weekdays. By 2005, total commercial ad time increased by 15% to 9 minutes 14 seconds on Saturday mornings and by 21% to 9 minutes 56 seconds per hour from 4 to 8 PM weekdays. These data indicate that the amount of TV advertising time on children's cable TV network programs is catching up to the amount of advertising time traditionally aired on the broadcast networks.
Our findings that food advertising constitutes a lower proportion of product advertising compared with findings from earlier content analyses, taken together with the relatively modest fluctuations in total hours of children's TV viewing and the comparable number of minutes per hour of advertising on children's programming during the past few decades, suggest that children do not see more food ads on TV today than they did 20 or 30 years ago.
A key question at the heart of the recent Institute of Medicine report on food marketing to children and youth42 is: To what extent does food advertising affect children's diet and contribute to obesity? A related question is: How might changes in exposure to food advertising have contributed to the increasing rates of childhood obesity witnessed in the United States during the past several decades?
Assessing a causal relationship between exposure to food advertising and obesity, however, is difficult. Existing empirical evidence is unable to disentangle the effects of total viewing time and food ads. It is difficult to quantify the effect of advertising on obesity because of its interrelatedness with sedentary activity and snacking while watching TV. Television viewing itself also may be endogenous to child overweight; that is, an observed correlation may be due to unobserved characteristics that are related to both outcomes.
The present study provides new evidence on content analyses of TV advertising viewed by children aged 2 through 11 years. The results showed that 27.2% of total nonprogram content time and 36.4% of total product advertising seen by children were for food products. Cereal was found to be the most frequently seen food product type, making up 27.6% of all food ads.
The analyses from our study showed that food advertising accounts for a substantial amount of the total TV ad exposure among children but an amount lower than what generally has been reported in previous studies, including comparable studies that examined weekday and cable viewing. To fully assess whether the absolute amount of exposure to food advertising among children has changed over time, we examined changes in children's TV viewing time and changes in the number and length of ads per hour on children's programming. We found modest fluctuations in TV viewing and a comparable number of minutes per hour of advertising time; taken together with the lower rates of food advertising found in the present study, these results suggest that children do not currently see more food ads on TV than they did during the past 35 years. Our study findings, combined with the data presented on average viewing and commercial times, suggest that children aged 2 through 11 years see, on average, 23 (30-second–equivalent) food-related TV ads per day (calculated on the basis of no commercial-free program viewing). Thus, while American children aged 2 through 11 years are exposed to a substantial amount of food advertising, the dramatic increase in childhood obesity rates in the past several decades does not appear to be mirrored by similar changes in food advertising exposure. Nevertheless, although total exposure may not have increased, the use of more sophisticated marketing strategies (including tie-ins to TV characters) may have changed the effectiveness of food advertising to children over time.
If one holds the recent distribution of food vs non-food ads constant, the very recent trends of increased commercial time on children's cable TV programming (a 15%-21% increase during 2000-2005) coupled with the recent upward trend in TV viewing (from 2 hours 58 minutes per day in 1999 to 3 hours 19 minutes per day in 2005) suggest that children's exposure to food ads may be increasing. This short-term trend raises concerns, especially when combined with the potentially increased effectiveness of food product ads and the proliferation of cable TV. These most recent trends are not to be taken lightly given the recently released national estimates on the prevalence of obesity that show overweight among children aged 2 through 5 years to have increased from 10.3% in 1999-2000 to 13.9% in 2003-2004 and among children aged 6 through 11 years to have increased from 15.1% to 18.8% during the same period.1
The results presented in this study are not without limitations. First, the results are based on national advertising and do not include local spot advertising. To address the scope of this limitation, TV advertising expenditure data show that 28.8% of advertising expenditures are for spot TV advertising.67 However, by product classification, 23.1% of total spot advertising is for automotive products.68 Among the top 25 product classification categories, roughly 70% of the spot advertising is for products not relevant to the 2- to 11-year-old child market (eg, automotive, telecommunications, furniture, insurance, real estate, and financial).68 Thus, spot advertising is estimated to make up a much smaller proportion of advertising directed to children. Hence, among the spot advertising that is directed to children, the product distribution would have to be significantly different from the product distribution in national advertising to bias our findings. Previous studies that have drawn their samples from tapings of children's programming on a given set of channels and therefore captured the local spot ads have not sampled all markets in the country; hence, the findings from any one of those studies are subject to bias in the direction of the distribution of the spot advertising for their particular regional sample. A second limitation is that we assessed a subset of top-rated shows viewed by children aged 2 through 11 years. However, our sample was so extensive that it was strongly representative of children's viewing: shows that fell below the sampling cutoff were viewed by less than approximately 1% of children.
Despite these limitations, there are several major methodological advances in this study over previous studies. First, the reported findings are weighted on the basis of TV ratings and hence provide an actual exposure measure. Second, all advertising segments in this study were standardized to a 30-second ad, by which exposure across different advertising segments was equalized. Third, given that the sample was drawn on the basis of ratings, the sample itself was representative of the shows that children aged 2 through 11 years actually watch on TV and covered all time slots and days of the week. While the audience of Saturday morning children's programming is predominantly children, most of children's TV viewing does not take place during those hours. Data show that, for children aged 2 through 11 years, only 6% of total weekly viewing time takes place on Saturdays between 7 AM and 1 PM.20
Further research is needed to assess the relationship between child overweight and differences in exposure to the content of TV advertising while holding TV viewing time constant and controlling for sociodemographic and other contextual factors. In addition, research is warranted to examine the extent to which the effectiveness of food advertising has increased through the use of improved media design and the association of media characters with food products and to examine how this may have changed children's preferences and related weight outcomes over time. Evolving empirical evidence in this field will help policymakers to assess the potential effectiveness of various policy instruments that range from calls for complete bans on food advertising during children's programming to varying degrees of regulations such as restrictions on character licensing/merchandising and media endorsements of unhealthful food product advertising directed to children. Finally, in addition to regulatory policies for food advertising, researchers and policymakers should also assess the effectiveness of PSAs for healthy eating as a potential policy instrument to improve children's perceptions of nutrition, food consumption patterns, and weight outcomes.
Correspondence: Lisa M. Powell, PhD, Institute for Health Research and Policy, University of Illinois at Chicago, 1747 W Roosevelt Ave, Room 558, Mail Code 275, Chicago, IL 60608 (firstname.lastname@example.org).
Accepted for Publication: December 14, 2006.
Author Contributions: The authors had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Powell, Szczypka, and Chaloupka. Acquisition of data: Powell, Szczypka, and Chaloupka. Analysis and interpretation of data: Powell and Szczypka. Drafting of the manuscript: Powell and Szczypka. Critical revision of the manuscript for important intellectual content: Powell, Szczypka, and Chaloupka. Statistical analysis: Powell and Szczypka. Obtained funding: Chaloupka. Administrative, technical, and material support: Szczypka. Study supervision: Powell and Chaloupka.
Financial Disclosure: None reported.
Funding/Support: This study was supported by the Robert Wood Johnson Foundation through the Bridging the Gap program for the ImpacTeen project.
Role of the Sponsor: The funding organization had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or preparation, review, or approval of the manuscript.
Powell LM, Szczypka G, Chaloupka FJ. Exposure to Food Advertising on Television Among US Children. Arch Pediatr Adolesc Med. 2007;161(6):553–560. doi:10.1001/archpedi.161.6.553