Consumption of Low-Nutrient, Energy-Dense Foods and Beverages at

RESEARCH Current Research

Consumption of Low-Nutrient, Energy-Dense Foods and Beverages at School, Home, and Other Locations among School Lunch Participants and Nonparticipants RONETTE R. BRIEFEL, DrPH, RD; ANDER WILSON; PHILIP M. GLEASON, PhD

ABSTRACT Background Access to foods and beverages on school campuses, at home, and other locations affects children’s diet quality, energy intake, and risk of obesity. Objectives To describe patterns of consumption of “empty calories”—low-nutrient, energy-dense foods, including sugar-sweetened beverages— by eating location among National School Lunch Program (NSLP) participants and nonparticipants. Design Cross-sectional study using 24-hour dietary recall data from the 2004-2005 third School Nutrition Dietary Assessment Study. Subjects/setting A nationally representative sample of 2,314 children in grades one through 12, including 1,386 NSLP participants. Statistical analyses performed Comparisons, using t tests, of the proportion of children consuming low-nutrient, energydense foods and beverages, mean daily energy and energy from low-nutrient, energy-dense foods, and energy density by NSLP participation status. Results On a typical school day, children consumed 527 “empty calories” during a 24-hour period. Eating at home provided the highest mean amount of energy from lownutrient, energy-dense foods (276 kcal vs 174 kcal at school and 78 kcal at other locations). NSLP participants consumed less energy from sugar-sweetened beverages at

R. R. Briefel is a senior fellow, Mathematica Policy Research, Inc, Washington, DC. A. Wilson is a senior programming analyst, Mathematica Policy Research, Inc, Cambridge, MA. P. M. Gleason is a senior fellow, Mathematica Policy Research, Inc, Geneva, NY. STATEMENT OF CONFLICT OF INTEREST: See page S89. Address correspondence to: Ronette R. Briefel, DrPH, RD, Mathematica Policy Research, Inc, 600 Maryland Ave, Ste 550, Washington, DC 20024-2512. E-mail: [email protected] Address reprint requests to: Jackie Allen, Mathematica Policy Research, Inc, PO Box 2393, Princeton, NJ 085432393. E-mail: [email protected] Published by Elsevier Inc. on behalf of the American Dietetic Association. 0002-8223/09/10902-1003$0.00/0 doi: 10.1016/j.jada.2008.10.064

school than nonparticipants (11 kcal vs 39 kcal in elementary schools and 45 kcal vs 61 kcal in secondary schools, P⬍0.01), but more energy from low-nutrient, energydense solid foods such as french fries and higher-fat baked goods in secondary schools (157 kcal vs 127 kcal, P⬍0.01). Participants were not more likely to consume sugar-sweetened beverages or low-nutrient, energydense foods at home or other locations. School lunch participants’ consumption at school was less energy-dense than nonparticipants’ consumption at school (P⬍0.01). Energy density was highest for consumption at locations away from home and school. Conclusions Improving home eating behaviors, where the largest proportion of total daily and energy from lownutrient, energy-dense foods are consumed (especially from sugar-sweetened beverages, chips, and baked goods) is warranted. At schools, consumption of energy from low-nutrient, energy-dense foods may be reduced by limiting access to competitive foods and beverages, enforcing strong school wellness policies, and minimizing the frequency of offering french fries and similar potato products and higher-fat baked goods in school meals or à la carte. J Am Diet Assoc. 2009;109:S79-S90.

U

nderstanding the role of eating behaviors and school meal program participation in children’s food consumption patterns is critical to addressing and improving the weight status of America’s children (1-4). During the period between 2003 and 2006, one in three school-aged children in the United States (33% of 6- to 11-year-olds and 34% of 12- to 19-year-olds) was overweight or obese, accounting for 25 million American children (5,6). Children from low-income or minority families are more likely to be overweight and to receive free or reduced-price school meals than children from higherincome non-Hispanic white families (7-9). Obesity is a multifactorial, complex issue, and a child’s weight status is affected by energy intake and expenditure over the course of a day and over time (1,4,6). Schools play an important role in shaping the dietary and physical activity behaviors of children (1-3). School food policies affect children’s access to foods available in vending machines or served à la carte as alternatives to eating a school meal. These “competitive foods” are often low-nutrient and energy-dense, and thus provide excess energy

Supplement to the Journal of the AMERICAN DIETETIC ASSOCIATION

S79

relative to their nutrient value (10-13). Examples of lownutrient and energy-dense items are sugar-sweetened beverages, salty/high-fat chips, high-fat baked goods, and desserts. Some low-nutrient and energy-dense items, such as breakfast pastries, brownies, cookies, and french fries, may also be available to children as part of a school meal. Research has shown that limiting access to lownutrient and energy-dense foods at school reduces the consumption of these items in schools (2,3,12). Further, access to such “junk foods” or “empty calories” has been related to increased energy intake at school and higher body mass index (BMI) among middle school children (14). Beyond the campus, children’s access to healthful or less healthful foods and beverages in the home and nearschool or community affects their overall diet. There is strong evidence linking particular dietary behaviors to the risk of childhood overweight. Consumption of fast foods and low-nutrient and energy-dense foods and beverages, particularly sugar-sweetened beverages, is prevalent among US children and adolescents and can lead to increased energy intake, increased BMI, or obesity, depending upon energy expenditure (15-25). Equally important is that these low-nutrient and energy-dense items displace more healthful alternatives such as fresh fruits and vegetables and low-fat milk and can reduce children’s diet quality and intake of essential nutrients (13,15,17,19,24). Several other factors are associated with childhood overweight and obesity, including social influences, family resources, knowledge and attitudes about diet, the frequency of eating and snacking, and genetics (1,17,2631). Increases in children’s BMI and overweight in the 1980s and 1990s have been accompanied by increases in the frequency of eating away from home and by shifts in the types of foods and beverages and portion sizes consumed (1,17,32-35). Studies of breakfast and adiposity have found varying results across age and sex subgroups (27). Skipping breakfast may be a risk factor for increased adiposity among older children or adolescents and appears to be more important for girls than for boys (27). There is limited information on eating frequency and adiposity among children. Cross-sectional studies have found no significant relationship, but a longitudinal study of girls aged 9 to 19 years found that meal frequency was significantly and inversely related to BMI (28,30). Snacking frequency does not appear to be related to childhood adiposity, but the evidence is clouded by the lack of a consistent definition of “snack” across research studies (29). In 2007 more than 30 million children participated in the National School Lunch Program (NSLP) (9). The evidence linking participation in school meals to obesity has been weak or mixed. A 2004 review by Fox and colleagues (36) found no definitive evidence linking NSLP participation and overweight. Studies published since that 2004 review found either no relationship or suggestive evidence among only the youngest elementary school students (37-39). Research using national data collected in the 2004-2005 third School Nutrition Dietary Assessment Study (SNDA-III) found that usual participation in the School Breakfast Program (SBP), but not the NSLP, was associated with a lower BMI (39). Gleason and Dodd (39) suggest that eating breakfast (ie, not skipping breakfast)

S80

February 2009 Suppl 1 Volume 109 Number 2

and energy distributions across the day are possible explanations. A child’s food environment can be viewed as consisting of school, home, and other locations away-from-home/ school—“away” locations. One of the nutritional benefits of school meal programs is that certain low-nutrient, energy-dense foods and beverages (eg, soft drinks and candy) cannot be part of the meals. Participants presumably consume less of these kinds of foods and beverages at school than do nonparticipants. However, school meals may contain some types of low-nutrient, energy-dense foods (eg, french fries, pastries, and high-fat baked goods), and frequent consumption of these items by participants may diminish the nutritional benefits of school meal participation. Alternatively, if school meal participants consume fewer low-nutrient, energy-dense foods and beverages at school, it may be that they make up for these “at school” differences by consuming more low-nutrient, energy-dense items off campus, either at home or at other locations. Understanding the dietary patterns of school meal participants and nonparticipants will inform the policy debate about how best to maximize the benefits of school meal programs and identify areas to target to improve children’s eating behaviors. This article uses nationally representative, cross-sectional data from the 2004-2005 SNDA-III to describe the eating patterns of public school lunch program participants and nonparticipants. We focus on three components of children’s diets: meal and snack patterns; consumption of low-nutrient, energy-dense foods and beverages (eg, sugarsweetened beverages); and location (ie, school, home, away). The fundamental motivation for this descriptive analysis is to test the research hypothesis that children who participate in a school lunch program are likely to consume proportionately fewer sugar-sweetened beverages and low-nutrient, energy-dense items at school, and compensate for their school behavior by eating proportionately more of these items at home and away locations. Because any observed relationship between school lunch participation and dietary patterns could be driven by differences in the characteristics of participants and nonparticipants, such as socioeconomic status, no causal inferences can be drawn. However, exploring the relationship between school lunch participation and dietary patterns away from school is a first step in understanding the dietary patterns of school lunch participants and nonparticipants leading to more in-depth analysis in the future. We also compare children’s energy intake and energy density across food environments (ie, school, home, or away) to gain a better understanding of consumption patterns in different food environments and their potential association with risk of obesity. METHODS Sample Design The SNDA-III data set is based on a nationally representative, cross-sectional sample of students at US public schools participating in the NSLP. Data were collected during spring 2005. The sample design included a stratified sample of school districts, schools within districts, and children (or students) within schools. The final sample includes 2,314 children in grades one through 12,

distributed among 287 schools. Twenty-four– hour dietary recalls, child surveys, and parent surveys were collected for each selected child. About one third of school children sampled were in elementary school (most in grade five or six); two thirds were in middle school or high school. Because overall meal and snack patterns did not differ significantly between middle and high school, these groups were combined as secondary schools to simplify data presentation. Overall, 62% of children participated in the NSLP on a typical school day and 18% in the SBP (40). Most students who participate in the SBP on a given day also participate in the NSLP, though the reverse is not necessarily true. About 54% of children were non-Hispanic white, 22% Hispanic, 17% non-Hispanic African American, and 7% other race/ethnicity (40). Data Collection All data collection instruments and procedures were reviewed and approved by the US Department of Agriculture (USDA), Food and Nutrition Service, the 2004 Education Information Advisory Committee of the Council of Chief State School Officers, and the Office of Management and Budget. In addition, the study worked with any institutional review process a school district required. One 24-hour recall was collected for each child and a second day’s intake on a subsample (with the help of parents for elementary school children) using the USDA Automated Multiple Pass Method software (version 2.3, 2003, USDA, Agricultural Research Service, Beltsville, MD). Dietary recalls were processed with the SurveyNet coding system (version 3.14, 2004, USDA, Agricultural Research Service) and the Food Nutrient Database for Dietary Studies (version 1.0, 2004) (41). The response rate among children selected for data collection was 63% (40). Children and their parents responded to separate surveys to obtain information on household and other student characteristics. The response rate for the parent interview, given that the child had completed the inschool interview, was 89% (40). Additional details on data collection are described elsewhere (40). The dietary findings in this paper are group means based on a single day’s intake per sampled child. Analytic Methods A key aspect of the analysis presented here compares children who participate in the NSLP to those who do not. Participation was defined using target day participation; that is, whether the child consumed a school lunch on the day in which the 24-hour dietary recall was completed. Children in food-based menu-planning schools were counted as participating in NSLP if either the child consumed at least three of the required five food groups (ie, one grain, one mean/meat alternate, two fruits and/or vegetables, and one milk) and all three were on the school menu for the target day, or if the child reported consuming at least one of the required five groups and reported consuming a school lunch on the target day. Children in nutrient standard menu-planning schools were counted as participating in NSLP if either the child consumed at least one entrée and one side, both of which were on the

school menu for the target day, or the child reported consuming one entrée or one side that was on the school menu for the target day and also reported consuming a school lunch for that day (42). We classified foods and beverages as low-nutrient, energy-dense items if they were energy-dense and low in nutrients or were of minimal nutritional value, as defined by USDA school meal regulations (43). All beverages reported in the study were grouped into seven mutually exclusive categories, including sugar-sweetened beverages, flavored milk, unflavored milk (whole/2%, skim/1%), 100% fruit juice, diet/low-energy drinks, and bottled water (44). For this analysis, we considered only sugar-sweetened beverages (including soda, fruit-flavored sweetened beverages, “energy” and sports drinks, and sweetened iced teas) as meeting our criteria of a low-nutrient, energy-dense beverage. We classified low-nutrient, energy-dense solid foods into five mutually exclusive categories: higher-fat baked goods, including muffins and desserts such as cakes, cookies, and brownies; candy (all types) and sweetened gum; dairy-based desserts (eg, ice cream); french fries and similar potato products; and chips and salty snacks (eg, potato chips, corn chips, and “buttered” popcorn). Foods in all five groups were included as low-nutrient, energy-dense items for this analysis. To determine eating patterns, we used child- (or respondent-) defined eating occasions. Breakfasts, lunches, and suppers/dinners were reported as such. Snacks included foods reported as snacks, drinks, and extended consumption (ie, typically a drink consumed over a long period of time). Some children reported more than one breakfast, lunch, or supper/dinner (eg, a breakfast at home and a breakfast at school). Children were allowed multiple meals of the same meal name, but each child was counted once in the population estimates. If a child reported eating a school lunch they were considered a school lunch participant even if they reported a second lunch at another location. The 24-hour dietary recall interviews captured the time and name of each eating occasion, the foods and beverages reported at each eating occasion, and the source from which each item was obtained. We used this information to deduce where each eating occasion was consumed and to assign the location of eating occasions as school, home, or away in the following manner: ●

●

●

For each food or beverage reported in the 24-hour recall, a child was asked if the item was consumed at school. All foods reported as consumed at school, and all other items consumed at the same eating occasion as a food a child reported as consumed at school, were considered to be consumed at school. Foods that were not eaten at school and were obtained from home, a friend/classmate (excluding entire classes) or neighbor, or a relative were considered to have been consumed in the home environment. In addition, any food consumed at the same eating occasion time as a food obtained from one of these sources was also considered to have been eaten in the home environment. Other foods were considered to have been consumed at locations away from home and school. These included foods obtained from restaurants, non-school vending

February 2009 ● Supplement to the Journal of the AMERICAN DIETETIC ASSOCIATION

S81

machines, churches, YMCAs, Boys’ & Girls’ clubs, other community sites, sporting events, and ice cream trucks, among others. Across all sample children, 31,108 food items were reported (unweighted count); 28% of reported items were obtained and consumed at school; 8% were brought from home and consumed at school; 2% were obtained from an away location and consumed at school; and ⬍1% were obtained from school and consumed elsewhere. About half (51%) of reported items were obtained and consumed at home and 10% obtained and consumed at away locations. Energy density is the amount of energy in foods consumed by individuals per gram of weight of those foods (35,45). Energy density was calculated by summing the total number of kilocalories consumed by individuals at a location and dividing by the total number of grams consumed at that location. Most children consumed only one beverage at a meal or snack; it was important to include all beverages in the energy density calculation because excluding low- and non– energy-containing beverages in the energy density calculation would provide an artificially high estimate of the energy density of a meal or snack. Therefore, energy density was calculated based on total consumption of all foods and beverages. For each child, energy density was calculated for each location. Means were taken across this child-location-level variable. Mean daily energy density was calculated for the full sample, whereas mean energy density by location was calculated only for the sample that consumed at least one item at that location. Statistical Methods We conducted descriptive analyses of the meal and snack patterns of children; mean energy density; energy for low-nutrient, energy-dense foods and sugar-sweetened beverages; and total energy by location. All statistical procedures were completed using Statistical Analysis Software (SAS) (version 9.1, 2004, SAS Institute, Cary, NC), and SUDAAN (release 9, 2005, Research Triangle Institute, Research Triangle Park, NC), incorporating appropriate sampling weights for school children and design effects caused by the SNDA-III complex sample design. We conducted t tests to determine whether there were statistically significant differences in eating patterns, energy intake, and energy density between school lunch program participants and nonparticipants. Differences were considered statistically significant at P⬍0.05; P values were not adjusted for multiple comparisons because the analysis was exploratory rather than confirmatory (46). None of the analyses controlled for family or child characteristic other than school meal participation. RESULTS Meal and Snack Patterns Table 1 shows the proportion of children reporting eating breakfast, lunch, supper/dinner, and snacks by location consumed (note that breakfast and lunch include all meals, not only reimbursable school meals). The majority of children reported eating breakfast, supper/dinner, and

S82


at least one snack at home. Nearly all children (91%) consumed lunch at school, whereas 23% consumed breakfast, and 40% reported at least one snack at school. NSLP participants consumed breakfast and lunch at school more frequently than nonparticipants (P⬍0.01). Breakfast was the meal most commonly skipped by children in both elementary and secondary school. On an average school day, the most common eating occasion away from school or home was a snack (17%), followed by supper/ dinner (13%). Patterns of Consumption of Sugar-Sweetened Beverages and Low-Nutrient, Energy-Dense Solid Foods Table 2 shows the proportion of children consuming sugarsweetened beverages; low-nutrient, energy-dense solid foods; and any low-nutrient, energy-dense solid food or sugar-sweetened beverage, by location consumed. Overall, 68% of children consumed sugar-sweetened beverages at some location during the day, about half (50%) consumed sugar-sweetened beverages at home, and one fourth (25%) at school. Most children (88%) consumed some amount of a low-nutrient, energy-dense solid food, and nearly all (95%) consumed “empty calories” from a low-nutrient, energy-dense item over the course of a 24hour period (on a Monday through Friday). In elementary schools, NSLP nonparticipants were more than four times as likely as participants to consume sugar-sweetened beverages at school (38% vs 9%, P⬍0.01); however, there were no significant differences between the proportion of participants and nonparticipants who reported consumption of low-nutrient, energy-dense solid foods, or low-nutrient, energy-dense solid foods and sugar-sweetened beverages combined, at school, home, or away (Table 2). The consumption patterns of secondary school NSLP participants and nonparticipants at home were similar, although nonparticipants were more likely to consume sugar-sweetened beverages at school (38% vs 29%, P⬍0.01) and away (21% vs 15%, P⬍0.01). Secondary school NSLP participants were more likely to consume any low-nutrient, energy-dense solid food at school (67% vs 55%, P⬍0.01), contributing to being more likely to consume any low-nutrient, energy-dense item at school (74% vs 67%, P⬍0.05). Secondary school lunch participants were less likely to consume any low-nutrient, energy-dense item at other locations away from school and home (23% vs 30%, P⬍0.05). Mean Daily Energy Intake, Energy Density, and Energy from Low-Nutrient, Energy-Dense Items The previous section describes the proportion of children consuming various types of low-nutrient, energy-dense items, but it is also important to consider the amounts of these items consumed, because both contribute to the population estimates of average energy consumed from low-nutrient, energy-dense items. Table 3 shows mean energy and mean energy density for total diet (ie, all foods and beverages) and for dietary components (eg, sugarsweetened beverages; low-nutrient, energy-dense solid foods; and any low-nutrient, energy-dense item) among NSLP participants and nonparticipants by the location in which the foods were consumed. Table 3 also shows the

Table 1. Meal and snack patterns among participants and nonparticipants in the National School Lunch Program (NSLP), school year 2004-2005a Elementary School

Eating occasion

NSLP participantsb (nⴝ531)

Secondary School

NSLP nonparticipants (nⴝ201)

NSLP participantsb (nⴝ855)


All (nⴝ2,314)

4™™™™™™™™™™™™™™™™™™™™™™™ %⫾standard error ™™™™™™™™™™™™™™™™™™3 Total daily consumption (% eating) Breakfast Lunch Supper/dinner Snacksc Consumption at school (% eating) Breakfast Lunch Supper/dinner Snacksc Consumption at home (% eating) Breakfast Lunch Supper/dinner Snacksc Consumption away from school/ home (% eating) Breakfast Lunch Supper/dinner Snacksc

90⫾1.6 100⫾0* 96⫾1.0 95⫾1.4

92⫾2.6 95⫾2.0 95⫾2.2 93⫾2.6

78⫾1.8 100⫾0.0** 92⫾1.2 94⫾0.8

77⫾2.0 88⫾1.6 92⫾1.2 94⫾1.2

84⫾1.0 96⫾0.4 94⫾0.6 94⫾0.8

34⫾3.6** 100⫾0** 3⫾1.0d 40⫾3.2**

8⫾2.4 90⫾2.8 2⫾1.0d 56⫾6

24⫾2** 100⫾0** 1⫾0.4d 36⫾2.4

13⫾1.6 68⫾2.8 2⫾0.6d 37⫾2.6

23⫾1.8 91⫾0.8 2⫾0.4 40⫾2.0

57⫾3.4** 0.7⫾0.6* 82⫾1.8 84⫾2.0

84⫾3.6 6⫾2.0d 83⫾3.2 79⫾3.2

54⫾2.6** 2⫾0.8** 79⫾1.8 85⫾1.4

62⫾2.0 9⫾1.4 75⫾2.0 82⫾1.4

61⫾2.0 4⫾0.6 80⫾1.0 83⫾1.0

2⫾0.8d 0.1⫾⬍0.1* 11⫾1.4 15⫾1.8

1⫾0.8d 4⫾1.4 12⫾2.8 16⫾3.2

1⫾0.4d** 0.8⫾0.4d** 13⫾1.4 20⫾2.0

3⫾0.6 11⫾1.4 16⫾2.0 20⫾1.8

2⫾0.4 3⫾0.4 13⫾0.8 17⫾1.0

a Data are from the third School Nutrition Dietary Assessment Study, 24-hour Dietary Recall Interview, school year 2004-2005. Meals and snacks are defined by child or respondent in the 24-hour dietary recall interview. Tabulations are weighted to be nationally representative of children in public NSLP schools. Sample sizes are unweighted. b NSLP participation is participation on the target recall day. c Includes eating occasions reported by the child (or respondent) as a snack or a drink. d Statistic is potentially unreliable due to a coefficient of variation ⬎30%. *Significantly different from nonparticipants at 0.05 level. **Significantly different from nonparticipants at 0.01 level.

proportion of energy consumed at school, home, or other locations over the course of a typical school day. On average across all children, 35% of daily energy was consumed at school. Among NSLP participants, the proportion was 40% in elementary school and 38% in secondary school (Table 3). Among children who participate in both the SBP and the NSLP, up to 51% of daily energy was consumed at school; the majority of daily energy (47%) was also obtained at school (data not shown). Among elementary school children, there were no significant differences in the amount of energy consumed at school, away, or over the entire day; however, at home, elementary school participants consumed an average of 119 kcal fewer than nonparticipants (1,105 kcal vs 1,224 kcal, respectively, P⬍0.05). Among secondary school children, NSLP participants consumed more energy at school (808 kcal vs 533 kcal, P⬍0.01) and over the entire day (2,250 kcal vs 2,076 kcal, P⬍0.01), but consumed less energy away (208 kcal vs 309 kcal, P⬍0.01). Mean energy consumed at home was not different between secondary school participant groups. Overall, children consumed an average of 527 kcal from all low-nutrient, energy-dense items over the course of a

day (Table 3). In elementary schools, NSLP participants consumed less than one third the mean amount of energy from sugar-sweetened beverages as nonparticipants (11 kcal vs 39 kcal, P⬍0.01). In the other locations, there are no significant differences in sugar-sweetened beverage consumption for elementary school participants vs nonparticipants. In secondary schools, NSLP participants consumed significantly less energy from sugar-sweetened beverages at school (45 kcal vs 61 kcal, P⬍0.01) and significantly fewer away (34 kcal vs 52 kcal, P⬍0.05). However, NSLP participants in secondary schools consumed more energy from low-nutrient, energy-dense solid foods at school (157 kcal vs 127 kcal, P⬍0.01), but less energy from such foods away (44 kcal vs 70 kcal, P⬍0.05). Figures 1 and 2 show the mean energy intake from specific categories of low-nutrient, energy-dense items at school. On average, NSLP participants consumed significantly more energy from french fries and similar potato products in elementary school and significantly less from sugar-sweetened beverages, candy, and chips/salty snacks compared with nonparticipants (Figure 1). Elementary school participants’ had a higher intake of chips/ salty snacks at home compared with nonparticipants


S83

Table 2. Consumption of sugar-sweetened beverages (SSBs) and low-nutrient, energy-dense (LNED) foods among participants and nonparticipants in the National School Lunch Program (NSLP) (% eating)ab Elementary School

Eating occasion

NSLP participantsc (nⴝ531)

Secondary School




All (nⴝ2,314)

4™™™™™™™™™™™™™™™™™™™™™™™ %⫾standard error ™™™™™™™™™™™™™™™™™™™™™3 All locations LNED beverages (ie, SSBs) LNED solid foods Total LNED items At school LNED beverages (ie, SSBs) LNED solid foods Total LNED items At home LNED beverages (ie, SSBs) LNED solid foods Total LNED items Away from school and homed LNED beverages (ie, SSBs) LNED solid foods Total LNED items

60⫾2.8 90⫾2.2 95⫾1.4

65⫾4.9 87⫾2.9 95⫾2.0

72⫾2.0* 90⫾1.3 96⫾0.7

78⫾1.5 85⫾1.8 95⫾1.0

68⫾1.4 88⫾1.1 95⫾0.7

9⫾1.5** 63⫾4.1 65⫾4.2

38⫾4.6 66⫾4.8 76⫾3.8

29⫾2.6** 67⫾2.3** 74⫾1.8*

38⫾2.4 55⫾2.4 67⫾2.5

25⫾1.5 62⫾1.9 69⫾1.9

48⫾2.7 61⫾2.7 76⫾2.0

45⫾4.0 62⫾3.6 77⫾3.2

55⫾2.0 60⫾1.8 79⫾1.6

52⫾1.7 57⫾2.1 77⫾1.3

50⫾1.3 60⫾1.4 77⫾1.0

12⫾1.5 14⫾1.7 19⫾1.9

9.3⫾2.1e 18.1⫾3.2 20.6⫾3.2

15⫾1.4** 15⫾1.6* 23⫾1.9*

21⫾2.0 22⫾1.9 30⫾2.2

15⫾0.9 17⫾1.0 23⫾1.1

a Data are from the third School Nutrition Dietary Assessment Study, 24-hour Dietary Recall Interview, school year 2004-2005. Tabulations are weighted to be nationally representative of children in public NSLP schools. Sample sizes are unweighted. b LNED items include SSBs (eg, carbonated soft drinks, fruit-flavored juice drinks, lemonades, sweetened teas, and “energy” or “sports” drinks), chips (eg, regular, not lower-/reduced-fat), cookies, ice cream, cake-type desserts, muffins (eg, regular, not lower-fat), pastries, donuts, crispy rice bars, candy, energy bars, fruit snacks, and french fries/similar potato products. c NSLP participation is participation on the target recall day. d Consumed at other locations away from school and home. e Statistic is potentially unreliable due to a small sample size or a coefficient of variation ⬎30%. *Significantly different from nonparticipants at 0.05 level. **Significantly different from nonparticipants at 0.01 level.

(31⫾4.2 kcal vs 19⫾4.3 kcal, P⬍0.05, data not shown), partially offsetting an observed difference at school (20⫾3.1 kcal vs 51⫾7.8 kcal, P⬍0.01). No other differences were observed for elementary school children’s consumption of any other low-nutrient, energy-dense solid food categories at home or at away locations. On average, NSLP participants consumed significantly more energy from french fries/similar potato products, baked goods, and dairy-based desserts in secondary school and significantly less from sugar-sweetened beverages and salty snacks compared with nonparticipants (Figure 2). Among secondary school children, there were no differences in mean energy intake from low-nutrient, energy-dense solid foods at home between NSLP participants and nonparticipants; however, energy consumed from french fries/similar potato products at away locations were significantly lower for NSLP participants compared to nonparticipants (15⫾3.6 kcal vs 37⫾6.1 kcal, P⬍0.01, data not shown), partially offsetting an observed difference at school (45⫾6.1 kcal vs 14⫾3.4 kcal, P⬍0.01). Overall, NSLP participants in secondary schools consumed less energy from low-nutrient, energy-dense items at away locations (78 kcal vs 122 kcal, P⬍0.05), but showed no significant differences at school, at home, or for the total day (Table 3). Mean energy density for consumption at school was lower among NSLP participants

S84


than nonparticipants (1.32 vs 1.67 in elementary school and 1.30 vs 1.56 in secondary school, both P⬍0.01). Mean energy density over the entire day was lower for NSLP participants in elementary school than nonparticipants (1.21 vs 1.33, P⬍0.01), but higher for NSLP participants in secondary school (1.14 vs 1.08, P⬍0.05). Differences between Children Who Consume Sugar-Sweetened Beverages at School and Those Who Do Not In addition to describing consumption patterns comparing school lunch participants and nonparticipants (in Tables 1 through 3) we investigated whether at school behavior with respect to sugar-sweetened beverages was associated with the consumption of energy outside of school, either at home or away locations. Table 4 shows the proportion of children consuming sugar-sweetened beverages, mean energy intake from sugar-sweetened beverages, and mean total 24-hour energy intake from all foods and beverages, stratified by whether or not any amount of sugar-sweetened beverage was consumed at school. Elementary school children had no differences in the proportion consuming sugar-sweetened beverages at home or away from home/school based on their school consumption of sugar-sweetened beverages. Secondary

Table 3. Mean daily energy, energy density, and mean energy from sugar-sweetened beverages (SSBs) and low-nutrient, energy-dense (LNED) foods consumed by participants and nonparticipants in the National School Lunch Program (NSLP)ab Intake by Location and NSLP Participation Status Elementary School

Eating location


Secondary School




All (nⴝ2,314)

4™™™™™™™™™™™™™™™™™™™™ mean⫾standard error ™™™™™™™™™™™™™™™™™™3 All locations LNED beverages (ie, SSBs) (kcal) LNED solid foods (kcal) All LNEDs (kcal) Total energy (kcal) Energy densityd At school LNED beverages (ie, SSBs) (kcal) LNED solid foods (kcal) All LNEDs (kcal) Total energy (kcal) Percentage of daily energy consumed at school (% kcal) Energy densitye At home LNED beverages (ie, SSBs) (kcal) LNED solid foods (kcal) All LNEDs (kcal) Total energy (kcal) Percentage of daily energy consumed at home (% kcal) Energy densitye Away from home or schoolf LNED beverages (ie, SSBs) (kcal) LNED solid foods (kcal) All LNEDs (kcal) Percentage of daily energy consumed away from home and school (% kcal) Total energy (kcal) Energy densitye

107⫾7.4 334⫾18.2 441⫾18.1 2,048⫾37.1 1.21⫾0.013**

118⫾11.2 378⫾27.3 495⫾31.2 2,075⫾67.4 1.33⫾0.042

196⫾9.8 399⫾17.5 596⫾22.1 2,250⫾44.0** 1.14⫾0.014*

224⫾9.9 380⫾16.6 604⫾22.4 2,076⫾43.5 1.08⫾0.021

159⫾5.2 368⫾10.0 527⫾12.0 2,109⫾24.5 1.18⫾0.012

11⫾2.0** 121⫾10.6 132⫾11.4** 790⫾24.2

39⫾5.2 166⫾20.5 206⫾21.6 710⫾45.5

45⫾4.4** 157⫾8.5** 203⫾10.1 808⫾17.4**

61⫾4.8 127⫾8.7 189⫾11.3 533⫾24.2

36⫾2.5 138⫾6.1 174⫾7.4 720⫾12.7

40⫾1.1 1.32⫾0.034**

35⫾1.8 1.67⫾0.091

38⫾0.7** 1.30⫾0.031**

26⫾1.0 1.56⫾0.071

35⫾0.6 1.42⫾0.030

77⫾7.2 177⫾12.1 254⫾11.9 1,105⫾32.0*

61⫾7.2 171⫾16.9 232⫾19.8 1,224⫾52.0

116⫾6.5 198⫾13.4 315⫾15.9 1,234⫾35.8

110⫾6.6 183⫾12.9 293⫾16.6 1,233⫾33.8

93⫾3.7 183⫾7.0 276⫾7.5 1,185⫾20.0

53⫾1.1 1.23⫾0.027

58⫾1.8 1.28⫾0.048

53⫾1.0** 1.16⫾0.027

60⫾1.1 1.12⫾0.020

56⫾0.6 1.19⫾0.015

20⫾2.9 36⫾6.8 56⫾9.2

17⫾4.2 40⫾8.8 57⫾11.2

34⫾3.7* 44⫾6.4* 78⫾9.0*

52⫾6.3 70⫾8.1 122⫾13.3

31⫾2.1 47⫾3.6 78⫾5.1

7⫾0.7 153⫾17.3 1.89⫾0.115

7⫾1.2 142⫾26.7 2.15⫾0.327

9⫾0.7** 208⫾19.8** 1.77⫾0.122

14⫾1.1 309⫾25.2 1.57⫾0.092

9⫾0.4 203⫾10.5 1.79⫾0.063

a Data are from the third School Nutrition Dietary Assessment Study, 24-hour Dietary Recall Interview, school year 2004-2005. Meals and snacks are defined by child (or respondent) in the 24-hour dietary recall interview. Tabulations are weighted to be nationally representative of children in public NSLP schools. Sample sizes are unweighted. b LNED items include SSBs (eg, carbonated soft drinks, fruit-flavored juice drinks, lemonades, sweetened teas, and “energy” or “sports” drinks), chips (eg, regular, not lower-/reduced-fat), cookies, ice cream, cake-type desserts, muffins (eg, regular, not lower-fat), pastries, donuts, crispy rice bars, candy, energy bars, fruit snacks, and french fries/similar potato products. c NSLP participation is participation on the target recall day. d Mean daily energy density calculated for 24-hour period among all persons. e Mean daily energy density calculated only among those who consumed any food or beverage at the location (ie, sample sizes vary for location estimates). f Consumed at other locations away from school and home. *Significantly different from nonparticipants at 0.05 level. **Significantly different from nonparticipants at 0.01 level.

school children who consumed a sugar-sweetened beverage at school were more likely to consume a sugar-sweetened beverage at home (59% vs 51%, P⬍0.05), but less likely to consume a sugar-sweetened beverage away (15% vs 20%, P⬍0.05). Consumers of sugar-sweetened beverages at school obtained a mean of 112 kcal from sugar-sweetened beverages in elementary school and 157 kcal in secondary school. Energy from sugar-sweetened beverages account for the

entire energy differential between sugar-sweetened beverages consumers and nonconsumers in elementary school, and more than half the differential in secondary school. Mean total daily energy intake for at-school consumers of sugar-sweetened beverages was 114 kcal greater in elementary school and 260 kcal greater in secondary school than the respective means for nonconsumers of sugar-sweetened beverages at school (both P⬍0.01). There were no significant differences in the energy from sugar-sweet-


S85

Figure 1. Consumption of low-nutrient, energy-dense items at school, by NSLPa participation status (elementary schools). aNSLP⫽National School Lunch Program. bSSBs⫽sugar-sweetened beverages. cIncludes similar potato products. *Significantly different from participants at P⬍0.01.

Figure 2. Consumption of low-nutrient, energy-dense items at school, by NSLPa participation status (secondary schools). aNSLP⫽National School Lunch Program. bSSBs⫽sugar-sweetened beverages. cIncludes similar potato products. *Significantly different from participants at P⬍0.05. **Significantly different from participants at P⬍0.01.

ened beverages consumed at home or away locations between the at-school consumers and the nonconsumers of sugar-sweetened beverages at school. There were no significant differences in total energy at home, away, or over the entire day between the two groups at the elementary school level. Among secondary school children who consumed sugar-sweetened beverages at school, energy intake over the entire day averaged 229 kcal higher than those who consumed no sugar-sweetened beverages at school (P⬍0.01). There were no differences between consumers and nonconsumers in mean energy from all items consumed at home and away between the two groups. Consumers of any sugar-sweetened beverages at school were significantly less likely to be participants in the NSLP or both the school lunch and breakfast programs (both P⬍0.01). For example, participation in the NSLP was 38% among consumers of sugar-sweetened beverages at school compared to 80% among nonconsumers in elementary school, and 44% to 54% in secondary school, respectively (data not shown, P⬍0.01).

S86


DISCUSSION The meals a school offers, as well as its overall food environment, are important factors in children’s daily energy intake and diet quality, but consumption patterns at home and other locations are also key to diet and to risk of overweight and obesity (1,2,4,24,25,35). We analyzed nationally representative data from SNDA-III from spring 2005 to improve our understanding of school-aged children’s consumption patterns across a typical 24-hour period that included attending public school. A key feature of this analysis involves examining consumption of low-nutrient, energy-dense foods and beverages over the course of the day and across eating locations. Although the analysis examines the relationship between school meal participation and dietary patterns, it is descriptive and should not be interpreted as capturing the effects of participation on dietary patterns. Rather, exploring patterns of intake among school lunch participants and nonparticipants may identify areas for further research to address the question of whether school lunch causes these observed differences in children’s patterns of consumption. Determining eating behaviors associated with consuming “empty calories” could lead to improved nutrition education and health promotion messages for children and their parents, as well as identify areas for school wellness policies to target. Dietary Patterns among All School-Aged Children On an average day that US children attended public schools offering the NSLP, they received more than one third (35%) of their daily energy from foods consumed at school and more than half (56%) from foods consumed at home. Less than 10% of their daily energy came from foods consumed at other locations. Whereas proportionately less energy was consumed away from school and home, these meals and snacks consumed away from school and home were the most energy-dense. Nearly all children consume lunch and dinner/supper on a typical school day, but a nontrivial percentage (16%) fail to consume any breakfast, including nearly one in four secondary school students. This is important given that previous studies have found that skipping breakfast is associated with increased weight gain from adolescence to adulthood and that eating breakfast is associated with lower BMI (23,27,30,39). Participation in the school breakfast program is one avenue by which children who would typically skip breakfast could access breakfast at school. SNDA-III data show that snacking in childhood and adolescence is prevalent across all ages and school meal participation groups and most prevalent at home. Overall, 94% of children reported consuming at least one snack during the day, and eight of 10 children consumed a snack at home, after school and/or before bedtime. Snacking at school was somewhat less common, but still widespread as four of 10 children consumed a snack at school, a behavior associated with the number of snack machines and school policies (11,12,14). The extent of snacking among children is of concern given that prior research has shown that the number of eating occasions is a significant positive predictor of consumption of lownutrient, energy-dense items, and that snacking at home

Table 4. Mean daily energy and mean energy from sugar-sweetened beverages (SSBs) based on consumption of SSBs at schoolab Elementary School Intake

Secondary School

Consumers of SSBs Nonconsumers (no SSBs Consumers of SSBs Nonconsumers (no SSBs at school (nⴝ115) at school) (nⴝ617) at school (nⴝ512) at school) (nⴝ1,070) All (nⴝ2,314) 4™™™™™™™™™™™™™™™™™™™™™™™™™™™™ %⫾standard error ™™™™™™™™™™™™™™™™™™™™™™™™™™™™3

SSBs (% consuming) Daily total At school At home Awayc Energy from SSBs (kcal) Daily total At school At home Awayc Energy (kcal) Daily total At school At home Awayc

100⫾0.0** 54⫾2.6 100⫾0.0** 62⫾2.0 100⫾0.0** 0⫾0.0 100⫾0.0 0⫾0.0 49⫾4.3 47⫾2.6 59⫾2.7* 51⫾2.0 14⫾3.3 11⫾1.4 15⫾1.6* 20⫾1.5 4™™™™™™™™™™™™™™™™™™™™™™™™™™™™ mean⫾standard error ™™™™™™™™™™™™™™™™™™™™™™™™™™3 202⫾12.2** 112⫾4.2** 64⫾8.3 25⫾7.1 2,079⫾52.3 863⫾40.2** 1,040⫾54.0 176⫾35.0

92⫾6.9 0⫾0.0 74⫾6.6 18⫾2.6 2,050⫾37.8 749⫾21.5 1,157⫾31.7 145⫾16.9

320⫾11.7** 157⫾6.0** 125⫾7.7 38⫾5.6 2,315⫾53.4** 844⫾23.7** 1,223⫾39.3 248⫾24.4

68⫾1.4 25⫾1.5 50⫾1.3 15⫾0.9

153⫾7.7 0⫾0.0 107⫾6.5 46⫾3.8

159⫾5.2 36⫾2.5 93⫾3.7 31⫾2.1

2,086⫾31.8 584⫾18.5 1,239⫾27.5 263⫾18.0

2,109⫾24.5 720⫾12.7 1,185⫾20.0 203⫾10.5

a Data are from the third School Nutrition Dietary Assessment Study, 24-hour Dietary Recall Interview, school year 2004-2005. Tabulations are weighted to be nationally representative of children in public National School Lunch Program schools. Sample sizes are unweighted. b SSBs include carbonated soft drinks, fruit-flavored juice drinks, lemonades, sweetened teas, and “energy” or “sports” drinks. c Consumed at other locations away from school and home. *Significantly different from children who did not consume SSBs at school at 0.05 level. **Significantly different from children who did not consume SSBs at school at 0.01 level.

is positively associated with consumption of sugar-sweetened beverages and salty snacks (12,23,24,29). With such widespread snacking, results from SNDA-III and other studies suggest the need to improve children’s and adolescents’ snack choices both at home and at school to reduce consumption of energy-dense foods and beverages that provide significant energy from low-nutrient, energydense foods over the course of the day. Low-nutrient, energy-dense foods often contain both fat and sugar, whose tastes are immediately appealing even to young children (47,48). As with snacking, consumption of low-nutrient, energy-dense foods and beverages is common among children. On a typical school day, 1 kcal out of every 4 kcal they consume comes from a low-nutrient, energy-dense food or beverage, with about two-thirds of this amount from low-nutrient, energy-dense foods and the remaining one-third from sugar-sweetened beverages. Part of the reason for the high level of consumption of low-nutrient, energy-dense foods and beverages may be from the easy access to these items by most children. Increased access to low-nutrient, energy-dense items is associated with consumption of high-fat and high-sugar foods and beverages among children and adolescents (16,23,35). Competitive food sources such as school stores and snack bars, vending machines, and à la carte offerings provide increased access to low-nutrient, energydense items at school, and school meals are also a source of higher-fat baked goods and french fries (11,12,44,49). In another analysis using SNDA-III data, we found that school food practices, including children’s access to school stores and snack bars and “pouring rights” contracts were

significantly related to consumption of sugar-sweetened beverages in secondary schools (44). However, consumption of low-nutrient, energy-dense items is not limited to those consumed at school. Among secondary students, both NSLP participants and nonparticipants consumed about 300 kcal from low-nutrient, energy-dense foods on average at home; among elementary school children, the figure ranged from 230 to 250 kcal from low-nutrient, energy-dense foods consumed at home. In addition, whereas the total amount of energy from low-nutrient, energy-dense items consumed away from home and school was not great, the foods consumed at these other locations were likely to include a large proportion of low-nutrient, energy-dense items, as 38% of energy consumed at these locations was from low-nutrient, energy-dense items. This led to the energy density of foods consumed away from school and home being higher than the energy density of foods consumed at school or home. Dietary Patterns of School Lunch Participants and Nonparticipants NSLP participants get a greater percentage of their energy from foods consumed at school than do nonparticipants, though the difference is modest. Among children in elementary school, participants get 40% of their energy from foods consumed at school compared with 35% among nonparticipants. Among secondary school students, the comparable figures are 38% and 26%. Similarly, there are differences in the proportion of NSLP participants and


S87

nonparticipants who consume breakfast and lunch at school. For example, a third of elementary school NSLP participants and only 8% of nonparticipants eat breakfast at school, whereas nonparticipants are more likely to eat breakfast at home. At the secondary school level, only two thirds of nonparticipants eat lunch at school compared with all NSLP participants. Among nonparticipants who eat lunch away from school, about half eat at home and half eat at some other location. There were clear differences in the consumption of low-nutrient, energy-dense items at school. Even though NSLP participants consumed a larger amount of total energy at school than nonparticipants, they consumed significantly less energy from sugar-sweetened beverages. This was especially true at the elementary school level, where participants consumed an average of 11 kcal and nonparticipants consumed 39 kcal from sugar-sweetened beverages. There was no evidence that participants made up for consuming fewer sugar-sweetened beverages at school by consuming more sugar-sweetened beverages outside of school. This is encouraging because research has shown that consumption of sugar-sweetened beverages is associated with higher energy intake and higher BMI and overweight among children and adolescents (14,18-21). The results with respect to low-nutrient, energy-dense foods was less clear. At the elementary school level, the difference between energy from low-nutrient, energydense foods among participants and nonparticipants was not statistically significant. At the secondary school level, energy from low-nutrient, energy-dense foods were significantly higher among participants than nonparticipants; however, this followed directly from the fact that participants’ overall energy intake from foods consumed at school was higher than that of nonparticipants. The specific low-nutrient, energy-dense foods that participants were more likely to consume at school included higher-fat baked goods and french fries. One of the key research questions we addressed was whether school meal participation is associated with higher or lower consumption of low-nutrient, energydense items away from school. The underlying issue here was whether participants, because they were consuming school meals that presumably did not include certain types of low-nutrient, energy-dense items (such as soft drinks or potato chips), were more likely to seek out these items once they left school for the day. We found no significant differences between NSLP participants’ and nonparticipants’ consumption patterns of low-nutrient, energy-dense foods and beverages at home, with one small exception: elementary school participants consumed more energy from salty snacks at home than nonparticipants. At locations away from home and school, there were no significant differences in consumption of low-nutrient, energy-dense items at the elementary school level, and participants consumed significantly less energy from low-nutrient, energy-dense items at the secondary school level. We also examined the relationship between children’s consumption of sugar-sweetened beverages at school and their low-nutrient, energy-dense item consumption over the remainder of the day. Children who did not consume sugar-sweetened beverages at school consumed less low-

S88


nutrient, energy-dense energy from sugar-sweetened beverages all day. In addition, consumption of a school breakfast or school lunch was associated with consuming less energy from sugar-sweetened beverages at school. This suggests that reduced access to sugar-sweetened beverages in the school environment was not associated with increased consumption away from campus. In fact, another analysis of SNDA-III data found that for elementary school children, most sugar-sweetened beverages consumed at school were brought from home (44). Data Limitations This analysis is based on cross-sectional, self-reported 1-day dietary recall data that may be subject to over- or underreporting of intake. The data reflect consumption on school days (Mondays through Fridays) and not Saturdays and Sundays, when consumption in general or at away locations may be higher. Finally, comparisons of the eating patterns of school lunch participants and nonparticipants are descriptive and do not control for differences in the background characteristics of the two groups. Therefore, they should not be interpreted as indicating the causal effects of the school meal programs. CONCLUSIONS SNDA-III findings, as well as prior research on children’s and adolescents’ eating behaviors, suggest that both family eating practices, at home and away from home, and the school food environment are important factors to address in the prevention of childhood overweight and obesity. SNDA-III data provide a recent view of US public school children’s consumption of low-nutrient, energydense foods and beverages and their associated energy. A high proportion—about one fourth— of average daily energy was categorized as from low-nutrient, energy-dense foods or beverages. An important and challenging opportunity to improve children’s diets and reduce low-nutrient, energy-dense food consumption is increased emphasis on behaviors at home, where the largest proportion of total energy and energy from low-nutrient, energy-dense foods are consumed— especially from sugar-sweetened beverages, chips/salty snacks, and baked goods. Food and nutrition professionals should encourage parents to serve more healthful beverages, such as low-fat milks, 100% fruit juices (in moderation), and water in place of sugar-sweetened beverages for meals and snacks at home and in bag lunches for school. To reduce consumption of low-nutrient, energy-dense foods at home, parents should serve reduced- or lower-fat baked goods and dairy-based desserts; offer more healthful alternatives such as fresh fruit; and minimize the frequency of fast foods high in fat, salt, and sugar. Parents also play a key role in their modeling of healthful eating behaviors at home and in their selection of foods at away locations. Parents can let adolescents know that eating healthful breakfasts and lunches is important and that skipping meals can be detrimental to maintaining a healthful weight. Avenues for more healthful eating away from home and school include improving food and beverages choices, selecting smaller portion sizes, and eating less frequently at places that offer mostly energy-dense selections.

The school food environment provides an important and feasible opportunity to improve children’s diet and health behavior. SNDA-III data show that areas to improve include eliminating sugar-sweetened beverages available in schools, limiting or eliminating access to other competitive food sources, minimizing the frequency of offering french fries and higher-fat baked goods, instituting strong school wellness policies, providing nutrition education to parents and children, and encouraging the bringing of more healthful bag lunches from home (especially for nonparticipants in elementary school). It is encouraging that school lunch participants do not seem to make up for lower consumption of low-nutrient, energydense foods and beverages when they go home, for the most part. These study findings are relevant to school nutrition and wellness policies and efforts to improve the healthfulness of school meals, and highlight the need for continued emphasis on nutrition education and health promotion for families and parents of children of all ages and at all income levels. STATEMENT OF CONFLICT OF INTEREST: The authors have no conflict of interest to report with the sponsor of this supplement article or products discussed in this article. This research was supported by a grant from the US Department of Agriculture, Economic Research Service, contract no. 59-5000-6-0076. The opinions expressed are those of the authors and do not necessarily represent the views or recommendations of Mathematica Policy Research, Inc, the Economic Research Service, or the US Department of Agriculture. The authors thank Allison Hedley Dodd, PhD, for input on the study analysis plan and Mary Kay Crepinsek, MS, RD, and Liz Condon, MS, RD, for assistance with categorization of the food and beverage variables used for this analysis. The authors also thank Michael Ponza, PhD, and Katherine Ralston, PhD, for their review of an earlier draft of the manuscript.

10.

11.

12. 13.

14. 15.

16. 17. 18.

19. 20. 21. 22.

23.

24.

References 1. Koplan JP, Liverman CT, Kraak VI, eds. Preventing Childhood Obesity: Health in the Balance. Washington, DC: National Academies Press; 2005. 2. Story M, Kaphingst KM, French S. The role of schools in obesity prevention. Future Child. 2006;16:109-142. 3. Story M, Kaphingst KM, Robinson-O’Brien R, Glanz K. Creating healthy food and eating environments: Policy and environmental approaches. Annu Rev Public Health. 2008:29:253-272. 4. The Wingspread Conference on Childhood Obesity, Healthy Eating, and Agricultural Policy. Conference summary. http://www. agobservatory.org/library.cfm?RefID⫽99597. Accessed November 10, 2008. 5. Ogden CL, Carroll MD, Flegal KM. High body mass index for age among US children and adolescents, 2003-2006. JAMA. 2008;299: 2401-2405. 6. Childhood obesity. Robert Wood Johnson Foundation Web site. http:// www.rwjf.org/programareas/ChildhoodObesityFramingDoc.pdf. May 7, 2008. 7. Meich RA, Kumanyika SK, Stettler N, Link BG, Phelan JC, Chang VW. Trends in the association of poverty with overweight among US adolescents, 1971-2004. JAMA. 2006;295:2385-2393. 8. Gordon-Larsen P, Adair LS, Popkin BM. The relationship of ethnicity, socioeconomic factors, and overweight in US adolescents. Obes Res. 2003;11:121-129. 9. National School Lunch Program: Participation and lunches served. US Dept of Agriculture, Food and Nutrition Service Web site.

25.

26. 27. 28. 29. 30.

31. 32.

http://www.fns.usda.gov/pd/slsummar.htm. Accessed September 10, 2008. O’Toole T, Anderson S, Miller C, Guthrie J. Nutrition services and foods and beverages available at school: Results from the School Health Policies and Programs Study 2006. J Sch Health. 2007;77:500521. Wiecha JL, Finkelstein D, Troped PJ, Fragala M, Peterson KE. School vending machine use and fast-food restaurant use are associated with sugar-sweetened beverage intake in youth. J Am Diet Assoc. 2006; 106:1624-1630. Cullen KW, Zakeri I. Fruits, vegetables, milk, and sweetened beverages consumption and access to à la carte/snack bar meals at school. Am J Public Health. 2004;94:463-467. Templeton SB, Marlette MA, Panemangalore M. Competitive foods increase the intake of energy and decrease the intake of certain nutrients by adolescents consuming school lunch. J Am Diet Assoc. 2005;105:215-220. Kubik MY, Lytle LA, Story M. Schoolwide food practices are associated with body mass index in middle school students. Arch Pediatr Adolesc Med. 2005;159:1111-1114. Kant AK. Reported consumption of low-nutrient-density foods by American children and adolescents: Nutritional and health correlates, NHANES III, 1988 to 1994. Arch Pediatr Adolesc Med. 2003; 157:7899-7896. Kant AK, Graubard BI. Predictors of reported consumption of lownutrient-density foods in a 24-h recall by 8-16 year old US children and adolescents. Appetite. 2003;41:175-180. Nicklas TA, Baranowski T, Cullen KW, Berenson G. Eating patterns, dietary quality, and obesity. J Nutr. 2001;20:599-608. Calorically sweetened beverage intake and childhood overweight. American Dietetic Association Evidence Analysis Library. http:// www.adaevidencelibrary.com/evidence.cfm?evidence_summary_id⫽ 28&auth⫽1. Accessed May 7, 2008. Dietz WH. Sugar-sweetened beverages, milk, and obesity in children and adolescents. J Pediatr. 2006;148:152-154. Vartanian LR, Schwartz MB, Brownell KD. Effects of soft drink consumption on nutrition and health: A systematic review and metaanalysis. Am J Public Health. 2007;97:667-675. Ludwig DS, Peterson KE, Gortmaker SL. Relation between consumption of sugar-sweetened drinks and childhood obesity: A prospective, observational analysis. Lancet. 2001;357:505-508. Boutelle KN, Fulkerson JA, Neumark-Sztainer D, Story M, French SA. Fast food for family meals: Relationships with parent and adolescent food intake, home food availability and weight status. Public Health Nutr. 2007;10:16-23. Niemeier HM, Raynor HA, Lloyd-Richardson EE, Rogers ML, Wing RR. Fast food consumption and breakfast skipping: Predictors of weight gain from adolescence to adulthood in a nationally representative sample. J Adolesc Health. 2006;39:842-849. Traveras EM, Berkey CS, Rifas-Shiman SL, Ludwig DS, Rockett HR, Field AE, Colditz GA, Gillman MW. Association of consumption of fried food away from home with body mass index and diet quality in older children and adolescents. Pediatrics. 2005;116:e518-e524. Duerksen SC, Elder JP, Arredondo EM, Ayala GX, Slymen DJ, Campbell NR, Baquero B. Family restaurant choices are associated with child and adult overweight status. J Am Diet Assoc. 2007;107:849853. Factors associated with childhood overweight. American Dietetic Association Evidence Analysis Library. http://www.adaevidencelibrary. com/topic.cfm?cat⫽2792. Accessed June 13, 2008. Breakfast skipping and childhood overweight. American Dietetic Association Evidence Analysis Library. http://www.adaevidencelibrary. com/topic.cfm?cat⫽1047. Accessed June 13, 2008. Meal or eating frequency and overweight in children. American Dietetic Association Evidence Analysis Library. http://www.adaevidencelibrary. com/evidence.cfm?evidence_summary_id⫽21. Accessed June 13, 2008. Snacking and childhood overweight. American Dietetic Association Evidence Analysis Library. http://www.adaevidencelibrary.com/topic. cfm?cat⫽1056. Accessed June 13, 2008. Franko DL, Streigel-Moore RH, Thompson D, Affenito SG, Daniels SR, Crawford PB. The relationship between meal frequency and body mass index in black and white adolescent girls: More is less. Int J Obes (Lond). 2008;32:23-29. Perusse L, Rankinen T, Zuberi A, Chagnon YC, Weisnagel SJ, Argyropoulos G, Walts B, Snyder EE, Bouchard C. The human obesity gene map: The 2004 update. Obes Res. 2005;13:381-490. Huang T-K, Howarth NC, Lin B-H, Roberts SB, McCrory MA. Energy intake and meal portions: Associations with BMI percentile in U.S. children. Obes Res. 2004;12:1875-1885.


S89

33. Briefel RR, Johnson CL. Secular trends in dietary intake in the United States. Annu Rev Nutr. 2004;24:401-431. 34. Briefel RR. The changing consumption patterns and health and nutritional status in the United States: Evidence from national surveys. In: Kennedy E, Deckelbaum R, eds. The Nation’s Nutrition. Washington, DC: ILSI Press; 2007:11-27. 35. The Keystone forum on away-from-home foods: Opportunities for preventing weight gain and obesity. Keystone Center Web site. http://www. keystone.org/Public_Policy/Obesity.html. Accessed June 12, 2008. 36. Fox MK, Hamilton W, Lin B. Evaluating the Impact of Food and Nutrition Assistance Programs, Volume III: Review of the Literature. Cambridge, MA: Abt Associates; 2004. 37. Hofferth SL, Curtin S. Poverty, food programs, and childhood obesity. J Pol Anal Manage. 2005;24:703-726. 38. Schanzenbach DW. Do school lunches contribute to childhood obesity? Harris School Working Paper Series 05.13. http://harrisschool. uchicago.edu/About/publications/working-papers/pdf/wp_05_13.pdf. Accessed July 14, 2008. 39. Gleason PM, Dodd AH. School breakfast program but not school lunch program participation is associated with lower body mass index. J Am Diet Assoc. 2009;109(suppl 1):S118-S128. 40. Gordon AR, Cohen R, Crepinsek MK, Fox MK, Hall J, Zeidman E. The third School Nutrition Dietary Assessment Study: Background and study design. J Am Diet Assoc. 2009;109(suppl 1):S20-S30. 41. Raper N, Perloff B, Ingwersen L, Steinfeldt L, Anand J. An overview of USDA’s dietary intake data system. J Food Comp Anal. 2004; 17;545-555.

S90


42. Clark MA, Fox MK. Nutritional quality of the diets of US public school children and the role of the school meal programs. J Am Diet Assoc. 2009;109(suppl 1):S44-S56. 43. National School Lunch Program and School Breakfast Program: School Meals Initiative for Healthy Children. Final Rule. 60 Federal Register 31188-31222 (1995) (codified at 7 CFR § 210 and 220). 44. Briefel RR, Crepinsek MK, Cabili C, Wilson A, Gleason PM. School food environments and practices affect dietary behaviors of US public school children. J Am Diet Assoc. 2009;109(suppl 1):S91-S107. 45. Ledikwe JH, Blanck HM, Khan LK, Serdula MK, Seymour JD, Tohill BC, Rolls BJ. Dietary energy density determined by eight calculation methods in a nationally representative United States population. J Nutr. 2005;135:273-278. 46. Schochet PZ. Guidelines for multiple testing in experimental evaluations of education programs. Mathematica Policy Research Web site. http://www.mathematica-mpr.com/publications/PDFs/Educational Interventions.pdf. Accessed September 19, 2008. 47. Drewnowski A, Bellisle F. Liquid calories, sugar, and body weight. Am J Clin Nutr. 2007;85:651-661. 48. Drewnowski A. The real contribution of added sugars and fats to obesity. Epidemiol Rev. 2007;29:160-171. 49. Cullen K, Watson K, Zakeri I. Middle school student school lunch consumption: Impact of National School Lunch Program meal and competitive foods. US Dept of Agriculture, Economic Research Service Web site. http://www.ers.usda.gov/publications/ccr30/ccr30.pdf. Accessed November 10, 2008.