Effects of Changes in Lunch-Time Competitive Foods,Nutrition Practices,and Nutrition Policies on Low-Income Middle-School Children's Diets

Abstract

Background:

The School Nutrition Advances Kids project tested the effectiveness of school-initiated and state-recommended school nutrition practice and policy changes on student dietary intake in low-income middle schools.

Methods:

Schools recruited by an application for grant funding were randomly assigned to (1) complete an assessment of nutrition education, policies, and environments using the Healthy School Action Tools (HSAT) and implement an action plan, (2) complete the HSAT, implement an action plan, and convene a student nutrition action team, (3) complete the HSAT and implement an action plan and a Michigan State Board of Education nutrition policy in their cafeteria à la carte, or (4) a control group. All intervention schools were provided with funding and assistance to make self-selected nutrition practice, policy, or education changes. Block Youth Food Frequency Questionnaires were completed by 1176 seventh-grade students from 55 schools at baseline and during eighth-grade follow-up. Nutrient density and food group changes for the intervention groups were compared to the control group, controlling for baseline dietary intake values, gender, race/ethnicity, school kitchen type, urbanization, and percent of students eligible for free or reduced-price meals. Analyses were conducted by randomization and based on changes the schools self-selected.

Results:

Improvements in students' nutrient density and food group intake were found when schools implemented at least three new nutrition practice changes and established at least three new nutrition policies. Students in schools that introduced mostly healthful foods in competitive venues at lunch demonstrated the most dietary improvements.

Conclusions:

New USDA nutrition standards for à la carte and vending will likely increase the healthfulness of middle school children's diets.

Introduction

Foods offered at school can play a major role in the overall quality of children's diets.¹ Passage of the Healthy, Hunger-Free Kids Act of 2010 and the new USDA Nutrition Standards in the National School Lunch and Breakfast Program highlight a national commitment to enhance the nutritional quality of foods available to children in schools, including competitive foods outside of federal school meals programs, such as those available in vending machines and à la carte.^2,3 Recent national estimates suggest that virtually all high school students and 71–81% of middle school students have access to competitive foods of some kind during the school day, which nearly always have included unhealthy options.^4–6 This will soon change because of the 2013 USDA interim final rule, Nutrition Standards for All Foods Sold in School as Required by the Healthy, Hunger-Free Kids Act of 2010, which was designed to improve the healthfulness of competitive foods in most school venues.⁷

Many states and schools did not wait for federal regulations, but established their own competitive foods policies. In 2003, the Michigan State Board of Education (MSBE) adopted a nutrition policy recommending that schools offer and promote healthy foods and beverages in all competitive venues. That policy was superseded in 2010 by more comprehensive Michigan standards.⁸ Such state policies have resulted in reductions in junk food available in schools and have been associated with healthier student dietary intake.^9–11 Using data from the most recent School Health Policies and Practices Study conducted in 2006, researchers found that elementary and middle schools in states with firm policies to prohibit junk foods had less junk food than schools that were not under such policies.¹² Taber and colleagues found that high school students in California with standards for competitive foods reported consuming less fat, sugar, and total calories at school than students in states with no competitive standards.¹¹

Other school nutrition practices and policies also shape the nutrition environment and hence what children consume.^1,13 These include practices such as marketing healthful foods in the cafeteria, hosting health fairs, and classroom nutrition education as well as policies such as eliminating the use of foods as rewards. These types of practices and policies are not mandated under the Healthy, Hunger-Free Kids Act and are up to individual schools to adopt and implement on their own. Since 2004, hundreds of Michigan schools have utilized the online Healthy School Action Tools (HSAT)¹⁴ to assess their current policies and practices and create healthier school action plans. Many Michigan schools have also received small grant funding from various state organizations to implement their action plans.

Few controlled studies have evaluated the effects of implementing competitive foods policies and other school nutrition practices and policy changes on children's total diets. The School Nutrition Advances Kids (SNAK) project evaluated the effectiveness of various nutrition interventions, based on the HSAT, on low-income middle school children's nutrient density and healthy food group intake.

Methods

The SNAK project was a 2-year study conducted in low-income middle schools in Michigan from 2007 to 2010. Design elements included both a modified experimental intervention and a natural experiment, whereby schools were encouraged to make self-selected nutrition policy and practice changes and followed over time. Schools were recruited through an application for small grant funding with award values ranging from $2,000 to $4,600. Eligibility criteria included having 50% or more of students eligible for free or reduced-price meals and having seventh and eighth grades within the same building (for follow-up purposes). Recruitment methods included direct mailings, e-mails, and phone calls to eligible schools, as well as a posting on the Michigan Team Nutrition website. All study procedures and instruments were approved by the Michigan State University (MSU) Institutional Review Board. Collaborators included MSU Departments of Food Science and Human Nutrition and Epidemiology and Biostatistics, Michigan Department of Community Health, Michigan Department of Education (MDE), MSU Extension, Grand Rapids Public Schools, several partner agencies of the Michigan Action for Healthy Kids Coalition, and the University of Colorado Rocky Mountain Prevention Research Center.

The design involved two overlapping cohorts with schools recruited in 2007 and 2008 and follow-up assessments measured the next school year. Seventy-five schools were randomized into the four intervention groups. The 10 schools that did not initiate the project (usually because of misunderstanding the project scope or a change in personnel) were similar to schools that did in terms of percent of students eligible for free and reduced-price lunch (both 68%) and size [average enrollment was 370 students (standard deviation [SD], 130) vs. 490 students (SD, 211) for those who initiated the study]. The first cohort included 32 schools in 30 school districts participating from October 2007 to June 2009, and the second cohort included 33 schools in 20 districts participating from September 2008 to June 2010. Of the 65 schools that participated, 55 had both pre- and postintervention diet assessment measures from food frequency questionnaires (FFQs).

There were four study groups, including the control group (see Table 1). On their application, schools selected which of the study groups to which they were willing to be assigned. Fifty-six percent of schools selected all groups, 27% selected at least two groups, and 17% selected only one group. Schools that met the eligibility criteria and selected more than one group were randomized within the nutritional intervention groups they had selected, as well as within strata of school size and district. School size was stratified by schools with more than 200 seventh graders versus all others as well as school district by the largest two Michigan school districts versus all others. We randomized within these strata of schools to ensure balance. Schools that selected only one study group were assigned that group.

Table 1.

Summary of Interventions in the SNAK Project

	Study groups
	Control	HSAT Only	Student SNAK Team	MSBE Nutrition Policy
Participating schools
Schools randomized to group	21	24	5	25
Schools that did not initiate project	1	6	0	3
Schools that completed some aspect of project	20	18	5	22
Study sample of schools^a	17	16	4	18
Funding to schools ($)
Evaluation funding amount	2000	2000	2000	2000
Funding for HSAT participation, nutrition education, and marketing^b		1100	1100	1100
Funding for Student SNAK Team			400
Funding to cover potential losses to à la carte line				1500
Total funding to school	2000	3100	3500	4600
Intervention elements
CSHT		X	X	X
HSAT and Action Plan		X	X	X
Make ≥1 nutrition education/marketing change		X	X	X
Student SNAK Team			X
Implement MSBE nutrition policy				X

Schools with pre- and post-FFQ data.

HSAT facilitator, substitute teacher pay to enable classroom teachers to participate in HSAT meeting, and funding for school nutrition education and/or marketing nutritious choices in cafeteria.

CSHT, Coordinated School Health Team; HSAT, Healthy School Action Tools; MSBE, Michigan State Board of Education; SNAK, School Nutrition Advances Kids.

Description of the Interventions

The HSAT, an online assessment and action planning process similar in content to the CDC's School Health Index, was a key component of all of the interventions.¹⁵ With the HSAT intervention, Coordinated School Health Teams (CSHT)—with representatives from various sectors of the school, including administration, faculty, food service, healthcare, and students—completed a number of questions about their school regarding nutrition education, practices, and policies, were given ideas for improvement, and developed an action plan based on the assessment and their own priorities. The four intervention groups included the following:

1. Control Group schools participated only in data collection during the study period and were offered the HSAT intervention after the last data collection point.

2. HSAT Only intervention schools were asked to convene a CSHT. A trained facilitator met with the CSHTs one time to complete the HSAT nutrition module, and schools received $1000 funding to implement nutrition education or marketing aspects of their action plan.

3. Student SNAK Team intervention schools built upon the HSAT Only intervention by additionally providing a facilitator to work with a seventh-grade student team to assess school nutrition environments and policies, participate in the HSAT with the CSHT, and also provide $400 for the students to implement the nutrition education/marketing aspect of their own student action plan. A standardized curriculum was developed for the facilitator/student meetings, and each student team met with the facilitator three times. This intervention was intended as a pilot and had a reduced sample size (n=5 schools).

4. MSBE Nutrition Policy intervention schools, also built upon the HSAT Only intervention; in addition to developing their own action plan with funding for nutrition education/marketing, schools were asked to implement the 2003 MSBE Healthy Food and Beverage Policy¹⁶ in their cafeteria à la carte lines during the study's second year and were provided with $1500 to compensate for any losses to their food service revenue. The SNAK research partner team reviewed current recommendations from the Institute of Medicine,¹⁷ the Alliance for a Healthier Generation,¹⁸ and the Michigan Action for Healthy Kids¹⁹ and created nutrient and food standards for each food group. These standards were virtually identical to the new proposed standards issued by USDA, with sugars being limited to less than 35% of calories; differences include SNAK requirement for greater than 1 g of fiber for grain products, rather than that they be primarily whole grain, and less than 230 mg of sodium, rather than 200 mg.²⁰ Guidance documents and assistance were provided to the schools' food service professionals. Schools were asked to implement their plans before the second-year fall data collection.

Student-Level Dietary Intake

Active parental consent and student assent were obtained for participating students. The Block Kids FFQ 2004 (ages 8–17 years) was administered to assess usual student dietary intake.^21,22 Online surveys were originally attempted, but discontinued because of lack of computer access and low participation by students. Students in cohort 1 completed the baseline survey between November 2007 and March 2008 (either online or paper) and the follow-up survey (paper only) October–December 2008. Cohort 2 students completed baseline surveys between November 2008 and February 2009 (paper only) and the follow-up survey between October 2009 and January 2010 (paper only). Students who returned consent, assent, and surveys received small prizes and were entered into a raffle for prizes, such as a bicycle.

Of 8626 seventh-grade students eligible for the study, 1777 responded with a completed consent form, assent form, and baseline surveys. Among those, 1274 completed follow-up surveys during eighth grade. Compared with schools with higher response rates (>= 40%), schools with lower response rates had larger student enrollments [average 531 students (SD, 212) vs. average 331 (SD, 108)], but a similar percentage of children eligible for free or reduced-price school lunch (71 vs. 70%). Surveys were removed from analysis if they had incomplete or “playful” patterns (i.e., always responded “every day”) upon visual inspection (baseline, n=14; follow-up, n=19), physiologically implausible values for energy intake (<500 or >5000 kcal/day; baseline, n=120; follow-up, n=69), or missing ethnicity data (n=17). Schools with fewer than five completed surveys were removed from analyses to conduct the hierarchical modeling (eight schools; baseline, n=28 students; follow-up, n=10 students). Two schools closed during the project and were not included in the final analysis. The 10 schools not included in the analysis (eight with too few completed surveys and two that closed) were similar to the remaining schools in terms of enrollment [average enrollment 420 students (SD, 189) vs. 490 (SD, 211)] and percent of children eligible for free and reduced-price school lunch [average percent eligible was 80% (SD, 13) vs. 68% (SD, 15)]. The final sample used to analyze student dietary intake consisted of 55 schools with 1176 students with both baseline and follow-up surveys.

Data on the gender and race/ethnicity of all children in each of the schools in the final sample were obtained from MDE²³ and compared to the SNAK sample using a test for proportion and chi-square goodness of fit. The SNAK sample had more females (58.7 vs. 49.3%; p<0.05) and fewer white children (48.3 vs. 55.2%; p<0.05) than the school data. However, students' race/ethnicity was identified by parents or school personnel in the MDE data, but self-identified in the SNAK data. There were differences in reporting categories between the MDE data (parents selected among African American, White, Hispanic, Asian, Native American, Native Hawaiian, and multiracial) and those available to students in the FFQ survey (students selected among African American, White, Hispanic/Latino, Asian, American Indian/Alaskan native, and other).

School-Level Measures

School characteristics

The percent of students eligible for free or reduced-price school meals at each school was obtained through the MDE. School setting (urban, rural, or suburban) was determined using 2000 US Census data. Type of food service program was determined through interactions with each school. A variable was created distinguishing kitchen types: traditional (a full-service or satellite kitchen in the district with food prepared on site); satellite kitchen district (large district with a central kitchen with limited preparation done on site); vendor operated (a vendor brings in prepared foods to the school); operations run by a local food service management company; smaller districts with heat-and-serve kitchens (without full cooking capacity); or a large district with a heat-and-serve operation.

Cafeteria à la carte and vending offerings

Standardized data collection forms were created, requesting information over a 5-day period on school breakfast and lunch, student-accessible vending machines, and à la carte foods at baseline and follow-up. Food service directors or other food service personnel were asked to complete these forms during the fall and spring of each study year. School food service programs received $325 for each round of data forms returned. All data were entered into Access databases for processing. Response rates ranged from 82 to 98% for each of the four data collection periods.

À la carte and vending

Schools were categorized at baseline and at follow-up as selling à la carte or vending during lunch if any additional foods or beverages were sold in the cafeteria, not including milk. Data collected for à la carte included items offered, visible nutrition information, amount in package/item, cost to school per item, selling price, meal time offered, and vendor code. Data collected on each vending machine available to students during meal times included location, hours turned on, number of slots, items available, size of package/item, selling price, and vendor code.

To verify whether schools had implemented the MSBE policy, analysis was conducted of the schools' à la carte line before the intervention and at follow-up. Some schools had vending machines on in the cafeteria during lunch, and thus changes made to vending machines available to students during lunch in the cafeteria (but not outside the cafeteria or during other times) were also included in the analysis. À la carte and vending items were categorized according to whether they met criteria for “healthy” as outlined by the MSBE policy. Percentage of healthy à la carte and vending items at each data collection period were calculated. Foods were considered to meet the MSBE policy if their nutrient profiles fell within 10% of the SNAK nutrient criteria provided to schools (e.g., SNAK guidance suggested less than 200 calories; we allowed up to 220). Entrée items that could be sold as part of a reimbursable school meal were not included. Schools were considered to have implemented the MSBE policy if 50% of the foods they served in vending or à la carte met the criteria or there was a substantial increase in the percentage of healthy foods offered. The verification process revealed that schools made a range of changes to their à la carte and vending during the intervention period (see Table 2).

Table 2.

Completion of Intervention Components by Study Group

	Study group
	Total (n=55)	Control (n=17)	HSAT Only (n=16)	Student SNAK Team (n=4)	MSBE Nutrition Policy (n=18)
Completion of the HSAT (%)
Never	10 (18)	8 (47)	1 (6)	0 (0)	1 (6)
Before initiation of SNAK project only	6 (11)	6 (35)	0 (0)	0 (0)	0 (0)
During SNAK intervention	39 (71)	3 (18)	15 (94)	4 (100)	17 (94)
Changes in à la carte/vending during lunch (%)
No changes	21 (38)	10 (59)	3 (19)	2 (50)	6 (33)
No à la carte/vending program during lunch	8 (15)	2 (12)	4 (25)	1 (25)	1 (6)
Implemented the MSBE policy	5 (9)	1 (6)	0 (0)	1 (25)	3 (17)
Added healthy à la carte/vending	6 (11)	1 (6)	3 (19)	0 (0)	2 (11)
Removed à la carte/vending	2 (4)	0 (0)	1 (6)	0 (0)	1 (6)
Continued to implement MSBE policy	3 (5)	0 (0)	1 (6)	0 (0)	2 (11)
Missing or unreliable data	10 (18)	3 (18)	4 (25)	0 (0)	3 (17)
Number of school-initiated nutrition practice changes^a (%)
0–2	10 (18)	6 (35)	2 (13)	0 (0)	2 (11)
3–6	21 (38)	5 (29)	5 (31)	1 (25)	10 (56)
7–14	21 (38)	5 (29)	7 (44)	3 (75)	6 (33)
Missing data	3 (5)	1 (6)	2 (13)	0 (0)	0 (0)
Number of school-initiated nutrition policy changes^b (%)
None	23 (42)	7 (41)	6 (38)	0 (0)	10 (56)
1–2	18 (33)	7 (41)	5 (31)	1 (25)	5 (28)
3–6	11 (20)	2 (12)	3 (19)	3 (75)	3 (17)
Missing data	3 (5)	1 (6)	2 (13)	0 (0)	0 (0)

Number of school-initiated nutrition practice changes made during the study period reported by the school administrator or food service professional, of 15 possible changes (see Table 3).

Number of school-initiated nutrition policies adopted during the study period reported by the school administrator or food service professional, of six possible polices (see Table 3).

HSAT, Healthy School Action Tools; MSBE, Michigan State Board of Education; SNAK, School Nutrition Advances Kids.

Nutrition practice and policy changes

The Middle-School School Environment and Policy Survey (SEPS) for administrators/principals and food service directors/kitchen managers was used to assess building-level nutrition education, practices, and policies at baseline and follow-up. This instrument was adapted from an Elementary-School SEPS developed for use in Colorado elementary schools²⁴ and was reviewed by several school food service and administration representatives to establish face and content validity. A $25 gift card was provided for each completed survey. Response rates ranged from 86 to 94%.

The follow-up SEPS ended with an item asking respondents to report on whether improvements had been made in 15 nutrition practices or whether they had adopted any of six nutrition policies since the beginning of the study (see Table 3). A new practice or policy was considered to have been adopted if either the administrator or the food service professional reported it. Two scale scores were created: the new nutrition practices (0–14; Cronbach's alpha=0.74) and new policies (0–6; Cronbach's alpha=0.76). Schools were then categorized by the number of new practices changes: 0–2, 3–6, or 7–14 changes; and the number of new policies adopted: 0, 1–2, and 3–6 (see Table 2).

Table 3.

Nutrition Practices and Policies

Nutrition practices

1. Added nutrition information to service line/menus

2. Added nutrition information to newsletters or other materials sent home with students

3. Offered taste tests of healthy foods and beverages to students

4. Marketed healthy foods and beverages in the cafeteria/school

5. Removed advertisements for unhealthy foods and beverages

6. Changed placement of foods and beverages online to highlight healthy choices

7. Added/increased nutrition or healthy eating posters in cafeteria

8. Required that more than 50% of foods and beverages offered at classroom parties/celebrations be healthy

9. Added/increased amount of nutrition education classes

10. Coordinated nutrition education with classroom teachers in other subject areas

11. Coordinated nutrition education with the food service department

12. Provided after-school nutrition-related activities for students

13. Hosted nutrition-related events for parents

14. Hosted a health fair to promote nutrition

15. Eliminated use of food as a reward (e.g., teachers are no longer allowed to give any food for good behavior)

Nutrition polices

1. Policy to promote healthy foods and beverages in vending machines

2. Policy to promote healthy foods and beverages in à la carte in the cafeteria

3. Policy to promote healthy foods and beverages in concessions at school events

4. Policy to promote healthy foods and beverages served during classroom parties/celebrations

5. Policy to eliminate use of food as a reward

6. Policy to eliminate sale of unhealthy foods and beverages as fundraisers

Statistical Analyses

To assess intervention effects on student dietary intake, nutrient and food-level intake assessed by the FFQ at follow-up were modeled as outcomes using baseline dietary intake as a control. For all models, hierarchical linear (SAS PROC MIXED; SAS Institute Inc., Cary, NC) and logistic (SAS PROC GLIMMIX; SAS Institute) regression analyses with the random intercept specified was used, taking into account variation within and between schools, following the protocol outlined by Singer.^25,26 In all analyses, a p value<0.05 was considered statistically significant.

All nutrient and food-level FFQ outcomes were energy adjusted by dividing by total calories. Because most student nutrient and food intake variables were not normally distributed, all outcomes were log-transformed, except for percent calories from fat, saturated fat, sodium, intake of orange and leafy vegetables, and percent calories from sugar-sweetened beverages. Categorical outcomes were defined for two variables with extreme skewing: intake of orange and green leafy vegetables/1000 kcals was categorized into 0, >0–<5, 5–<25, and >=25 g, and percent calories from sugar-sweetened beverages were categorized into cut points: 0, >0–<5, 5–<10, 10–< 15, and 15–71.9%.

Analyses were conducted to determine the effect of study group assignment on dietary outcomes (experimental design) and also based on actual changes made to nutrition practice and policies: (1) changes made to the à la carte line and/or vending offerings to students during lunch; (2) school changes in nutrition practices excluding à la carte and vending; and (3) school changes in nutrition policies. Because the school nutrition interventions could affect all students in a school, this article reports analyses of all students combined, rather than specific subgroups. Analyses were adjusted for baseline dietary intake values (student-level variable), school setting (school-level variable), school kitchen type (school-level variable), percent of students eligible for free or reduced-price school meals (school-level variable), gender (student-level variable), and student-selected race/ethnicity (student-level variable). The interaction of student gender and race/ethnicity were found to be associated with multiple outcomes, and this interaction was also included in the models. Models were assessed controlling for school breakfast and lunch nutrients (e.g., analysis of changes in dietary intake of vitamin C controlled for changes in breakfast and lunch vitamin C from baseline to follow-up), but school meal changes were not statistically significant and they were left out of the models.

To interpret regression coefficients, the following calculations were made. For log-transformed outcomes, the exponents of the coefficients were calculated to determine the percentage additional change for the treatment group, compared to the control group (controlling for the baseline dietary intake value and other control variables). For the two categorical outcomes, exponents of the coefficients were calculated to determine the odds of being in the higher category for the treatment group, compared to the control group (controlling for the baseline dietary intake value and other control variables). For variables that were not log-transformed, coefficients are interpreted as the additional change in value for the treatment group, compared to the control group (controlling for the baseline dietary value and other control variables).

Results

Characteristics of schools and students included in the study and student dietary intake at baseline are shown in Tables 4 and 5. There were no significant differences among intervention groups with regard to school characteristics at baseline. The mean percentage of students eligible for free/reduced-price meals for schools in the SNAK project was 68% (range, 50–98%), which was similar to all Michigan low-income middle schools in 2007 (72%; n=514). There were some significant differences at baseline with regard to student dietary intake and all subsequent analyses adjusted for baseline dietary values.

Table 4.

Baseline School and Student Characteristics

	Study group
	Total		Control Group		HSAT Only		HSAT+Student Team		HSAT+MSBE Policy
	N	% or mean (SD)	N	% or mean (SD)	N	% or mean (SD)	N	% or mean (SD)	N	% or mean (SD)
School characteristics
Percent of children eligible for free/reduced-price school lunch (mean)	55	68.0 (14.7)	17	65.6 (14.2)	16	71.5 (15.8)	4	70 (17.1)	18	69.6 (14.3)
Seventh-grade enrollment (mean)	55	136.9 (85.1)	17	153.2 (105.3)	16	126.1 (71.6)	4	142.8 (158.7)	18	129.7 (56.1)
Setting (%)
Urban	34	61.8	11	64.7	9	56.3	3	75.0	11	61.1
Rural	13	23.6	5	29.4	4	25.0	0	0.0	4	22.2
Urban cluster/suburban	8	14.6	1	5.9	3	18.8	1	25.0	3	16.7
School type (%)
Elementary and middle school	16	29.1	6	35.3	5	31.3	1	25.0	4	22.2
Middle school only	33	60.0	11	64.7	8	50.0	2	50.0	12	66.7
Middle and high school	4	7.3	0	0.0	2	12.5	0	0.0	2	11.1
Elementary, middle, and high school	2	3.6	0	0.0	1	6.3	1	25.0	0	0.0
Kitchen type (%)
Traditional	24	43.6	9	52.9	6	37.5	0	0.0	9	50.0
Satellite kitchen	6	10.9	1	5.9	3	18.8	1	25.0	1	5.6
Vendor operated	4	7.3	0	0.0	2	12.5	1	25.0	1	5.6
Local food service company	12	21.8	5	29.4	2	12.5	1	25.0	4	22.2
Small district with heat-and-serve kitchen	2	3.6	0	0.0	0	0.0	1	25.0	1	5.6
Large district with heat-and-serve kitchen	7	12.7	2	11.8	3	18.8	0	0.0	2	11.1
Percent of schools with lunch-time cafeteria à la carte	55	74.6	17	88.2	16	56.3	4	75.0	18	77.8
Percent of schools with lunch-time cafeteria vending	52	25.0	16	43.8	14	14.3	4	0.0	18	22.2
Student characteristics
Age (years)	1237	12.3 (0.6)	448	12.4 (0.6)	308	12.3 (0.6)	57	12.4 (0.9)	424	12.3 (0.6)
Gender (%)
Male	510	40.6	192	42.2	121	39.2	26	45.6	171	39.3
Female	746	59.4	263	57.8	188	60.8	31	54.4	264	60.7
Race/ethnicity (%)
White	605	49.2	251	55.2	128	41.4	26	45.6	200	46.0
African American	289	23.0	80	17.6	88	28.5	16	28.1	105	24.1
Hispanic	199	15.8	55	12.1	53	17.2	8	14.0	83	19.1
Native American	60	4.8	26	5.7	19	6.2	2	3.5	13	3.0
Asian and other	101	8.0	43	9.5	21	6.8	5	8.8	32	7.4

HSAT, Healthy School Action Tools; MSBE, Michigan State Board of Education; SD, standard deviation.

Table 5.

Baseline Student Dietary Intake

	Study group
	Total		Control Group		HSAT Only		HSAT+Student Team		HSAT+MSBE Policy
	Female	Male	Female	Male	Female	Male	Female	Male	Female	Male
Student dietary intake	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)
Sample size	746	510	263	192	188	121	31	26	264	171
% kcal from fat	31.6 (5.4)	32 (5.2)	32.1 (5.3)	32.3 (5.6)	30.9 (5.5)^*	31.9 (5.5)	31.1 (5.9)^*	31.7 (4.4)	31.6 (5.5)	31.8 (4.7)
% kcal from saturated fat	10.9 (2.1)	11.2 (2.1)	11.2 (2)	11.2 (2.2)	10.8 (2.2)	11 (2.1)	11 (2)	11.3 (2.1)	10.8 (2)^*	11.2 (2)
Cholesterol, mg	180 (113.4)	206.9 (118.6)	185.8 (114.3)	210.2 (128.6)	181.5 (122.6)	210.4 (113.4)	168 (97.1)	201.2 (121.2)	174.6 (107.7)	201.7 (110.5)
Fiber, g	15.2 (9.1)	16.2 (9.7)	15.4 (9.3)	15.9 (9.6)	14.2 (8.5)	17.1 (9)^*	11.7 (5.3)^*	16.6 (10.6)	16.2 (9.5)^*	15.9 (9.9)
Sodium, mg	2676.9 (1386.5)	2947.6 (1526.7)	2761.3 (1440.5)	2951.7 (1547.9)	2542.9 (1318.3)	3052.2 (1573.7)	2469.9 (1181.2)	3119.8 (1590.8)	2712.5 (1398.9)	2842.9 (1463.7)
Calcium, mg	855.5 (422.1)	952.1 (421.6)	874.1 (423)	945 (441.1)	829.5 (389.2)	953.6 (408.7)	767.3 (326.1)	1076 (442.2)^*	865.7 (452.3)	940 (405.5)
Vitamin A, μg	573.5 (348.6)	614.7 (350.5)	564 (324.9)	591.6 (320.7)	561.9 (336.6)	647.2 (358.5)	455.1 (204.4)	645.7 (290.6)	605.2 (388.4)	612.9 (384.2)
Vitamin C, mg	143.3 (117.6)	138.1 (110.9)	135.5 (116.4)	135.1 (107.9)	142.8 (109.9)^*	152.8 (115.1)	132.5 (128.8)	124.1 (118.9)	152.6 (122.5)^*	133.3 (110.1)
Fruit, cups	2 (1.6)	1.9 (1.6)	1.8 (1.6)	1.8 (1.5)	2 (1.6)	2.2 (1.8)	1.5 (1.3)	1.7 (1.8)	2.1 (1.7)^*	1.9 (1.4)
Fruit and fruit juice, cups	2.7 (2.4)	2.6 (2.3)	2.5 (2.4)	2.5 (2.2)	2.8 (2.4)	3 (2.7)	2.2 (2.1)	2.2 (2.5)	2.9 (2.5)^*	2.6 (2.1)
Total vegetables, cups	1.2 (0.9)	1.2 (1)	1.1 (0.9)	1.1 (1)	1.1 (0.9)	1.2 (0.9)	1 (0.6)^*	1.3 (1.2)	1.3 (1)	1.2 (1.1)
Vegetables without French fries, cups	1.1 (0.9)	1.1 (1)	1.1 (0.9)	1.1 (1)	1 (0.9)	1.2 (0.9)	0.9 (0.5)^*	1.1 (1.2)	1.2 (1)	1.1 (1.1)
Orange/green leafy vegetables, cups	0.5 (0.4)	0.3 (0.4)	0.3 (0.4)	0.3 (0.4)	0.3 (0.4)	0.4 (0.4)	0.2 (0.3)	0.4 (0.5)	0.4 (0.5)^*	0.3 (0.5)
% calories from sugar-sweetened beverages	8.9 (8.6)	8.5 (7.5)	9.6 (8.6)	8.7 (7.5)	9.4 (9.6)	7.6 (6.8)	12.2 (10.4)	8.6 (6.8)	7.6 (7.2)^*	8.8 (8.1)
Added sugars, g	18.8 (14.3)	18.8 (12.9)	20.2 (15.2)	18.9 (12.4)	18.7 (15.2)	18.6 (13.2)	20.9 (16.1)^*	19.2 (13.5)	17.3 (12.1)	18.9 (13.2)
Whole grains, oz. equivalents	0.5 (0.4)	0.6 (0.4)	0.5 (0.4)	0.6 (0.5)	0.6 (0.4)	0.6 (0.4)	0.4 (0.3)	0.6 (0.3)	0.5 (0.4)	0.5 (0.4)

Significantly different from the control group within the same gender after controlling for race/ethnicity, urbanization, kitchen type, and percent of children eligible for free/reduced-price meals.

HSAT, Healthy School Action Tools; MSBE, Michigan State Board of Education; SD, standard deviation.

Nutrient and food group changes (adjusted for total caloric intake) between baseline and follow-up for the three intervention groups were compared to the control group (experimental design; Table 6). When compared to the control group, students in the HSAT Only group significantly increased their fiber intake by 5.2%, fruit by 17.3%, and fruit plus juice by 13.8%; students in the Student SNAK Team group decreased their intake of vitamin C by 11.8% and increased their added sugars by 14.6%; and students in the MSBE Nutrition Policy group decreased their cholesterol intake by 6.7%, increased their fiber by 4.9%, and increased their fruit by 18.3%. When the Student SNAK Team and MSBE Nutrition Policy intervention groups were compared to the HSAT Only group, students in the Student SNAK Team also decreased their vitamin C intake (by 17.7%) and increased their intake of added sugars (by 21.8%; data not shown). Changes in student intake between baseline and follow-up were not significantly different between the MSBE Nutrition Policy group and the HSAT Only group (data not shown).

Table 6.

Results of Hierarchical Multivariate Regression Models to Examine Student Nutrient and Food Group Intake Changes Comparing Intervention Conditions Versus Control Group

	HSAT Only (n=16 schools)			Student SNAK Team (n=4 schools)			MSBE Nutrition Policy (n=18 schools)
FFQ outcome	Coef (95% CI)	p value	y^a or OR	Coef (95% CI)	p value	y^a or OR	Coef (95% CI)	p value	y^a or OR
% kcal from fat	0.24 (−0.67, 1.15)	0.60	0.2%	−0.15 (−1.51, 1.22)	0.83	−0.2%	−0.66 (−1.73, 0.41)	0.22	−0.7%
% kcal from saturated fat	0.05 (−0.22, 0.32)	0.71	0.1%	−0.09 (−0.68, 0.49)	0.75	−0.1%	−0.3 (−0.73, 0.13)	0.16	−3.0%
Cholesterol, mg	−0.03 (−0.08, 0.02)	0.31	−2.6%	−0.05 (−0.19, 0.09)	0.46	−4.9%	−0.07 (−0.13, −0.01)	0.02	−6.7%
Fiber, g	0.05 (0, 0.1)	0.03	5.2%	0 (−0.09, 0.09)	0.95	0.3%	0.05 (0, 0.09)	0.03	4.9%
Sodium, mg^b	−4.85 (−47.22, 37.51)	0.82		−18.77 (−96.17, 58.63)	0.63		1.73 (−46.71, 50.17)	0.94
Calcium, mg	0.03 (−0.02, 0.07)	0.25	2.6%	0 (−0.12, 0.11)	0.95	−0.4%	0 (−0.05, 0.05)	0.86	0.4%
Vitamin A, μg	0.01 (−0.06, 0.08)	0.73	1.2%	−0.03 (−0.12, 0.06)	0.47	−3.1%	−0.02 (−0.08, 0.03)	0.39	−2.5%
Vitamin C, mg	0.07 (−0.03, 0.17)	0.17	7.2%	−0.13 (−0.23, −0.02)	0.02	−11.8%	0.05 (−0.06, 0.16)	0.34	5.4%
Fruit, cups	0.16 (0.02, 0.3)	0.03	17.3%	0.07 (−0.23, 0.38)	0.64	7.6%	0.17 (0.02, 0.32)	0.03	18.3%
Fruit and fruit juice, cups	0.13 (0.01, 0.25)	0.04	13.8%	−0.03 (−0.23, 0.18)	0.80	−2.6%	0.11 (−0.03, 0.24)	0.11	11.1%
Total vegetables, cups	0.02 (−0.09, 0.13)	0.69	2.2%	−0.11 (−0.34, 0.11)	0.32	−10.8%	−0.03 (−0.14, 0.07)	0.53	−3.2%
Vegetables without French fries, cups	0.02 (−0.08, 0.13)	0.65	2.5%	−0.13 (−0.38, 0.12)	0.30	−12.0%	−0.04 (−0.14, 0.07)	0.50	−3.5%
Orange/green leafy vegetables^c	0.26 (−0.19, 0.70)	0.25	1.3	−0.09 (−0.92, 0.74)	0.82	0.91	−0.21 (−0.62, 0.20)	0.31	0.81
Added sugars, g	−0.06 (−0.15, 0.03)	0.17	−6.0%	0.14 (0.02, 0.25)	0.03	14.6%	−0.02 (−0.11, 0.07)	0.62	−2.2%
Whole grains, oz. equivalents	0 (−0.12, 0.11)	0.94	−0.4%	0.04 (−0.12, 0.2)	0.63	3.9%	0.08 (−0.02, 0.17)	0.11	7.9%

All analyses controlled for baseline value, interaction of race^*gender, kitchen type, urbanization, and percent of children eligible for free/reduced-price meals. FFQ outcome % calories from sugar-sweetened beverages did not converge.

y=(1 – exp(Coeff)) and is interpreted as percentage change for the treatment group, compared to the control group (n=17), controlling for the baseline value.

Coefficient=additional change for the treatment group, compared to the control group, controlling for the baseline value in milligrams.

y=exp(Coeff) and is interpreted as the odds of being in the higher category for the treatment group, compared to the control group, controlling for the baseline value. Categories include: 0, >0–<5, 5–<25, and >=25 g.

FFQ, Food Frequency Questionnaire; HSAT, Healthy School Action Tools; MSBE, Michigan State Board of Education; SNAK, School Nutrition Advances Kids; Coef, coefficient; CI, confidence interval; OR, odds ratio.

Changes to Lunch-Time à la Carte and Vending

Students in schools that started a healthy lunch-time à la carte or vending program increased their fiber by 12.6%, calcium by 7.9%, vitamin A by 15.4%, vitamin C by 18.4%, fruit by 26.7%, total vegetables by 14.5%, and whole grains by 30.5% and increased the odds of being in the next-higher category of orange and green leafy vegetables by 2.2 (see Table 7). There were no dietary improvements detected for the group that implemented the MSBE policy.

Table 7.

Results from the Natural Experiment—Cafeteria à la Carte/Vending Changes: Student Nutrient and Food Group Intake Changes Comparing Five Groups to the Group That Made No à la Carte/Vending Changes (n=994 Students)

	No ALC, no vend (n=8 schools)			Implemented MSBE policy (n=5 schools)			Started healthy ALC/vend (n=6 schools)			Removed ALC (n=2 schools)			Already implementing MSBE policy (n=3 schools)
FFQ outcome	est (CI)	p value	y^a or OR	est (CI)	p value	y^a or OR	est (CI)	p value	y^a or OR	est (CI)	p value	y^a or OR	est (CI)	p value	y^a or OR
% kcal from fat	1.27 (−018, 2.72)	0.08	1.3%	−0.12 (−1.74, 1.51)	0.88	−0.1%	−0.18 (−1.29, 0.93)	0.74	−0.2%	1.74 (−0.09, 3.58)	0.06	1.7%	−0.27 (−2.12, 1.58)	0.77	−0.3%
% kcal from saturated fat	0.28 (−0.23, 0.78)	0.27	0.3%	−0.12 (−1.00, 0.75)	0.77	−0.1%	−0.02 (−0.41, −0.37)	0.92	0.0%	−0.24 (−0.64, 1.11)	0.59	0.2%	0.08 (−0.70, 0.86)	0.83	0.1%
Cholesterol, mg	−0.01 (−0.08, 0.06)	0.76	−1.1%	−0.03 (−0.12, −0.07)	0.59	−2.6%	0 (−0.07, 0.06)	0.92	0.0%	−0.14 (−0.24, −0.04)	0.008	−13.2%	0 (−0.11, 0.11)	1.00	0.0%
Fiber, g	0.01 (−0.08, 0.1)	0.82	1.0%	0.01 (−0.04, 0.07)	0.58	1.4%	0.12 (0.08, 0.16)	<0.0001	12.6%	0.04 (−0.06, 0.13)	0.42	3.9%	0.05 (−0.04, 0.13)	0.01	4.7%
Sodium, mg^b	−53.41 (−128.31, 21.49)	0.16		36.6 (−18.86, 92.06)	0.19		11.13 (−72.74, 95.0)	0.79		55.86 (−37.15, 148.88)	0.23		39.72 (−42.31, 121.74)	0.33
Calcium, mg	−0.03 (−0.1, 0.03)	0.33	−3.2%	0.01 (−0.1, 0.12)	0.82	1.2%	0.08 (0, 0.15)	0.04	7.9%	−0.09 (−0.18, 0)	0.06	−8.6%	0.02 (−0.06, 0.1)	0.60	2.2%
Vitamin A, μg	−0.05 (−0.13, 0.03)	0.19	−4.9%	0.03 (−0.02, 0.09)	0.22	3.4%	0.14 (0.06, 0.22)	0.001	15.4%	−0.04 (−0.18, 0.11)	0.61	−3.7%	0.08 (−0.02, 0.18)	0.13	8.3%
Vitamin C, mg	−0.06 (−0.17, 0.05)	0.29	−5.8%	−0.02 (−0.12, 0.08)	0.67	−2.0%	0.17 (0.06, 0.28)	0.003	18.4%	0.01 (−0.27, 0.29)	0.92	1.4%	0.02 (−0.17, 0.22)	0.80	2.5%
Fruit, cups	0 (−0.17, 0.17)	0.96	−0.4%	−0.04 (−0.24, 0.16)	0.67	−4.1%	0.24 (0.03, 0.44)	0.027	26.7%	0.08 (−0.19, 0.35)	0.55	8.3%	−0.10 (−0.34, 0.14)	0.40	−9.5%
Fruit and fruit juice, cups	−0.04 (−0.17, 0.09)	0.52	−4.1%	−0.9 (−0.21, 0.02)	0.11	−9.1%	0.13 (−0.03, 0.29)	0.11	14.0%	0.01 (−0.29, 0.32)	0.92	1.4%	−0.12 (−0.30, 0.06)	0.19	−11.2%
Total vegetables, cups	−0.03 (−0.15, 0.09)	0.63	−2.8%	0.03 (−0.11, 0.18)	0.64	3.4%	0.14 (0.03, 0.24)	0.02	14.5%	0.18 (−0.09, 0.44)	0.19	19.3%	0.02 (−0.13, 0.17)	0.80	1.9%
Vegetables without French fries, cups	−0.02 (−0.15, 0.1)	0.70	−2.3%	0.03 (−0.13, 0.18)	0.74	2.6%	0.12 (−0.01, 0.25)	0.07	12.9%	0.22 (−0.03, 0.48)	0.09	24.8%	0.02 (−0.13, 0.16)	0.83	1.6%
Orange/green leafy vegetables^c	0.12 (−0.52, 0.77)	0.70	1.1	0.26 (−0.36, 0.88)	0.39	1.3	0.78 (0.12, 1.43)	0.02	2.2	0.71 (−0.51, 1.93)	0.24	2.0	0.60 (−0.27, 1.47)	0.17	1.8
% calories from sugar-sweetened beverages^d	0.31 (−0.23, 0.84)	0.25	1.4	0.18 (−0.31, 0.66)	0.46	1.2	−0.37 (−0.89, 0.15)	0.16	1.4	−0.05 (−1.03, 0.94)	0.93	1.0	0.14 (−0.55, 0.83)	0.69	1.1
Added sugars, g	0.06 (−0.08, 0.20)	0.38	6.2%	−0.03 (−0.17, 0.18)	0.70	−2.7%	−0.07 (−0.16, 0.03)	0.17	−6.4%	−0.12 (−0.27, 0.03)	0.12	−11.3%	−0.01 (−0.11, 0.09)	0.82	−1.1%
Whole grains, oz. equivalents	−0.06 (−0.24, 0.11)	0.48	−6.0%	0.09 (−0.02, 0.2)	0.10	9.4%	0.27 (0.15, 0.38)	<0.0001	30.5%	−0.15 (−0.4, 0.09)	0.21	−14.3%	0.16 (−0.09, 0.42)	0.20	17.8%

All analyses controlled for baseline value, interaction of race/ethnicity^*gender, kitchen type, urbanization, and percent of children eligible for free/reduced-price meals.

y=(1 – exp(Coeff)) and is interpreted as percentage change for the treatment group, compared to schools that did not change ALC/vending (n=21), controlling for the baseline value.

Coefficient=additional change for the treatment group, compared to the control group, controlling for the baseline value in milligrams.

y=exp(Coeff) and is interpreted as the odds of being in the higher category for the treatment group, compared to schools that did not change ALC/vending, controlling for the baseline value. Categories include: 0, >0–<5, 5–<25, and >=25 g.

FFQ, Food Frequency Questionnaire; ALC, à la carte; vend, vending; MSBE, Michigan State Board of Education; est CI, estimated confidence interval; OR, odds ratio.

Nutrition Practice and Policy Changes

As compared to schools that made 0–2 nutrition practice changes, students in schools that made 3–6 practice changes significantly increased their intake of fiber by 10.5% and fruit by 30.1%; students in schools that made 7–14 practice changes significantly increased their intake of fiber by 6.3% and fruit by 24.4% (p=0.10; data not shown). As compared to schools that made no nutrition policy changes, students in schools that made 1–2 policy changes significantly decreased their intake of whole grains by 12.0%, whereas students in schools that made 3–6 policy changes significantly increased their fruit intake by 24.3% (data not shown).

Discussion

In our study, improvements in student dietary intake were noted in the schools that started new vending or à la carte programs with almost entirely or entirely healthy choices, such as salads, fruits, and whole grain snacks. Other studies have shown that implementation of school interventions or state-wide policies that call for an increase in healthy food options, without a corresponding elimination of unhealthy options, show weak results on student intake.^27–31 For example, one intervention eliminating certain snack foods from school snack bars, but not vending, showed a decrease in chip consumption from the snack bar, but the number of vending machines doubled during the intervention period, and consumption of snack chips and candy from vending, as well as ice cream sales, increased.²⁷ In contrast, implementation of new policies in Texas and California that called for the elimination of “unhealthy” items showed some positive effects on student dietary intake.^22,27,32,33

Our findings raise some cautions regarding a popular belief that students have better diets when no competitive foods are available. The presence of competitive foods at lunch is associated with a decrease in students' consumption of the Federal School Meals Program lunch, fruits, vegetables, and key nutrients, such as calcium and vitamin A, and an increase in consumption of sugar-sweetened beverages and low-nutrient-dense snacks.^1,34,35 However, these studies were conducted in the milieu that has existed up to this point, in which competitive foods consisted primarily of low-nutrient, energy-dense foods or beverages.³⁶ Few studies have examined the effect on student dietary intake when competitive foods offered are entirely or almost entirely healthy and offer students healthy options above and beyond the school meal programs.

Some observers question the value of improving the nutritional quality of competitive foods in schools, assuming that students who are denied popular snack foods in school venues will simply get them at home or buy them elsewhere. Our findings suggest otherwise. Unlike most studies of school nutrition environmental and policy changes, we measured students' total diets—not just foods eaten or purchased at schools—and showed dietary improvements with school-level changes.^1,32,37–39

We also found support for the importance of tackling issues beyond competitive foods. Students in schools that made improvements in nutrition practices, including nutrition education, and policies in general, improved intake of key nutrients and food groups more than students in schools that had made few policy and practice improvements. Students in schools that only made 1–2 policy changes, however, showed a decrease in whole grain intake; this could point to a need to make more significant modifications than 1–2 policy changes. When we tested practices and policies individually, we did not find any one that stood out (data not shown); rather, we observed a threshold effect whereby making changes to at least three practices and adopting at least three policies improved students' diets. This is consistent with the conclusions of reviews of school-based interventions—that multi-component interventions have the greatest likelihood of improving student behaviors and body weight.^40,41 The SNAK intervention schools were provided with funding for nutrition education improvements and/or marketing healthful foods in their cafeterias, and the majority of the SNAK intervention schools made 2–5 nutrition education or marketing-related changes (data not shown), which could have contributed to the interventions' successes.⁴²

In our study, results for the schools randomized to convene a student SNAK team on student dietary intake did not show improvements in diet and instead showed a decrease in vitamin C intake and increase in sugar intake. Process evaluation and exit interviews with the schools revealed that whereas some of the schools randomized to this group had some qualitative successes (e.g., the student SNAK team in one school convinced the school board to switch their food service vendor to a company more committed to serving healthy foods), they were not a homogenous group and each made quite disparate changes within their schools. Our sample size for the Student SNAK Team group (n=4 schools and n=94 students) was very limited for this intervention. The CDC has advised that student involvement is critical to the success of a nutrition policy, and students who become involved in school-based promotions have reported more healthful eating behaviors than less-involved students.^13,29,43 Other interventions have successfully involved student leadership in enacting school improvements. The TEENS study team, for example, found that having students taste low-fat items being considered for introduction into the food environment was critical for their acceptance.²⁹ The TACOS study worked with student teams to conduct nutrition promotions in a randomized study of 20 secondary schools; more promotions were associated with a higher percentage of lower-fat snack sales.³¹

There are many challenges when undertaking school-based nutrition interventions, and this study has several limitations. To maximize intervention implementation, the randomization process we used took into account the preferences of the schools. For example, schools that stated that they did not want to create a student SNAK team were excluded from the randomization to that group. Even so, we found that some schools did not complete the interventions as designed. When this occurred, reasons included competing priorities, low motivation, and/or inadequate guidance.

Likewise, we found it impossible to keep schools in the control group from making nutrition improvements. We believe that it would have been unethical to actively discourage schools from making improvements that they believed would improve kids' health. Many control schools were disappointed that they were not randomized into the intervention, actively told us so throughout the project, and went ahead and made nutrition-related changes. Three control schools even completed the HSAT during the intervention. Thus, our ability to detect effects when analyzing the data by the experimental design was greatly hindered. The planned natural experiment yielded fruitful results; however, our findings likely only apply to schools that are motivated to make nutrition policy and practice changes. Other school studies have discussed challenges with randomized designs and poor implementation.^29–31,44

An additional limitation was the low response rates for the student surveys, primarily the result of the difficulty in receiving signed consent forms from parents. Compared to schools with higher response rates, schools with lower response rates had similar student family income levels, as measured by percent of students eligible for free and reduced-price lunch, but larger student enrollment. Results are therefore best generalized to smaller schools. Our response rate of 72% for student completion of follow-up surveys compares well with nationally representative surveys, such as the YRBSS.⁴⁵ Our response rate for completion of school-based food offerings was high (82–98%). Because we wanted to involve school professionals as active participants, we trained and paid food service professionals to fill out standardized forms for data collection. Receiving complete data about school food service operations, foods and beverages offered, and school lunch recipes from busy food service professionals proved challenging. However, during our exit interviews, we learned that many food service professionals implemented improvements based on what they learned through participation in the study.

Conclusion

In summary, findings from the SNAK project demonstrate that implementation of school-initiated nutrition practices and policies resulted in improvements in student dietary intake. If implemented strictly, the new USDA standards for school competitive foods options will likely increase the nutrient density of middle school children's diets and increase intake of fruits, vegetables, and whole grains.

Footnotes

Acknowledgments

The SNAK project was funded by the Robert Wood Johnson Foundation's Healthy Eating Research Program, the Michigan Department of Community Health, and the USDA Supplemental Nutrition Assistance Program–Nutrition Education, supported by the Michigan Department of Human Services under contract numbers ADMIN-07-99010, ADMIN-08-99010, and ADMIN 09-99010. The authors are grateful to the students and staff of the schools that participated in the SNAK project. In addition, the authors thank: Richard Miles; Deb Bailey; Deanne Kelleher; Paul McConaughy; Ann Guyer; Paul Baumgartner; Donna Hensey; Larry Merx; Dru Szczerba; Elaine Belansky; and Julie Marshall. The authors also acknowledge Julie Orth, Hali Sobzcek, Tiffany Chritz, Elizabeth Weber, Caitlin Fischer, Ellen Mang, and Allison Krusky for data-processing support, as well as Karen Petersmarck and Caroline Crawford for assistance with manuscript preparation.

Author Disclosure Statement

No competing financial interests exist.

References

Briefel

, Crepinsek

, Cabili

, et al. School food environments and practices affect dietary behaviors of US public school children. J Am Diet Assoc, 2009; 109:S91–S107.

US Department of Agriculture Food and Nutrition Service. Nutrition standards in the National School Lunch and School Breakfast programs. 2012:4088–4167. http://www.gpo.gov/fdsys/pkg/FR-2012-01-26/pdf/2012-1010.pdf

Healthy, Hunger-Free Kids Act of 2010. Public Law No. 111-296, 124 Stat, 31832010.

O'Toole

, Anderson

, Miller

, et al. Nutrition services and foods and beverages available at school: Results from the school health policies and programs study 2006. J Sch Health, 2007; 77:500–521.

Johnston

, O'Malley

, Terry-McElrath

, et al. School policies and practices to improve health and prevent obesity: National secondary school survey results, school years 2006–07 and 2007–08. Bridging the Gap Program, Survey Research Center, Institute for Social Research: Ann Arbor, MI, 2011.

Gordon

, Crepinsek

, Nogales

, et al. School nutrition dietary assessment study-III, vol. I: School foodservice, school food environment, and meals offered and served. US Department of Agriculture, Food and Nutrition Service, Office of Research Nutrition and Analysis: Alexandria, VA, 2007.

US Department of Agriculture Food and Nutrition Service. National School Lunch Program and School Breakfast Program: Nutrition standards for all foods sold in school as required by the Healthy, Hunger-Free Kids Act of 2010, interim final rule, rule no. 39067-39120, 78 FR 39067: Federal Register, 2013.

Michigan Department of Community Health, Michigan Department of Education. Michigan Nutrition Standards Toolkit. 2010. Available at www.michigan.gov/mdch/0,4612,7-132-2940_2955_2959_58773-261516–,00.html Last accessed March 10, 2013.

Terry-McElrath

, O'Malley

, Delva

, et al. The school food environment and student body mass index and food consumption: 2004 to 2007 national data. J Adolesc Health, 2009; 45:S45–S56.

10.

Rudd Center for Food Policy and Obesity. School wellness policy report. 2008. Available at www.sde.ct.gov/sde/cwp/view.asp?a=2626&q=322168 Last accessed June 1, 2009.

11.

Taber

, Chriqui

, Chaloupka

FJ.

Differences in nutrient intake associated with state laws regarding fat, sugar, and caloric content of competitive foods. Arch Pediatr Adolesc Med, 2012; 166:452–458.

12.

Kubik

, Wall

, Shen

, et al. State but not distirict nutrition policies are associated with less junk food in vending machines and school stores in US public schools. J Am Diet Assoc, 2010; 110:1043–1048.

13.

Centers for Disease Control and Prevention. Guidelines for school health programs to promote lifelong healthy eating. Morbidity and Mortality Weekly Report recommendations and reports. Vol. 45: ERIC Clearinghouse, 1996. Available at www.eric.ed.gov/contentdelivery/servlet/ERICServlet?accno=ED400104 Last accessed September 3, 2013.

14.

Healthy School Action Tools. Available at www.mihealthtools.org/hsat Last accessed March 21, 2013.

15.

Centers for Disease Control and Prevention. School Health Index: A self-assessment and planning guide. 2000. Available at www.cdc.gov/healthyyouth/shi Last accessed October 1, 2012.

16.

Michigan State Board of Education. Policy on offering healthy food and beverages in venues outside of the federally regulated child nutrition programs. Michigan Department of Education: Lansing, MI, 2003.

17.

Institute of Medicine. Nutrition Standards for Foods in Schools: Leading the Way toward Healthier Youth. The National Academies Press: Washington, DC, 2007.

18.

Alliance for a Healthier Generation. Available at www.healthiergeneration.org Last accessed February 18, 2013.

19.

Michigan Action for Healthy Kids Coalition. Michigan Action for Healthy Kids Toolkit. 2005. Available at www.mihealthytools.org/hsat/default.asp?tab=resourceguide Last accessed October 2, 2013.

20.

US Department of Agriculture Food and Nutrition Service. National School Lunch Program and School Breakfast Program: Nutrition standards for all foods sold in school as required by the Healthy, Hunger-Free Kids Act of 2010, proposed rule. 2013. Available at www.fns.usda.gov/cga/020113-snacks.pdf Last accessed March 28, 2013.

21.

Smith

, Fila

Comparison of the Kid's Block Food Frequency Questionnaire to the 24-hour recall in urban Native American youth. Am J Hum Biol, 2006; 18:706–709.

22.

Cullen

, Watson

, Zakeri

Improvements in middle school student dietary intake after implementation of the Texas Public School Nutrition Policy. Am J Public Health, 2008; 98:111–117.

23.

Michigan Department of Education. Center for Educational Performance and Information: Public student counts (headcount data). Available at www.michigan.gov/cepi/0,4546,7-113-21423_30451_30460—,00.html Last accessed March 28, 2013.

24.

Belansky

, Chriqui

, Schwartz

MB.

Local school wellness policies: How are schools implementing the congressional mandate? Robert Wood Johnson Foundation: Princeton, NJ, 2009.

25.

SAS 9.3 [computer program]. SAS Institute Inc.: Cary, NC, 2012.

26.

Singer

JD.

Using SAS PROC MIXED to fit multilevel models, hierarchical models, and individual growth models. J Educ Behav Stat, 1998; 24:323–355.

27.

Cullen

, Watson

, Zakeri

, et al. Exploring changes in middle-school student lunch consumption after local school food service policy modifications. Public Health Nutr, 2006; 9:814–820.

28.

Moore

, Tapper

The impact of school fruit tuck shops and school food policies on children's fruit consumption: A cluster randomised trial of schools in deprived areas. J Epidemiol Community Health, 2008; 62:926–931.

29.

Lytle

, Murray

, Perry

, et al. School-based approaches to affect adolescents' diets: Results from the TEENS study. Health Educ Behav, 2004; 31:270–287.

30.

French

, Story

, Fulkerson

, et al. An environmental intervention to promote lower-fat food choices in secondary schools: Outcomes of the TACOS Study. Am J Public Health, 2004; 94:1507–1512.

31.

Fulkerson

, French

, Story

, et al. Promotions to increase lower-fat food choices among students in secondary schools: Description and outcomes of TACOS (Trying Alternative Cafeteria Options in Schools). Public Health Nutr, 2004; 7:665–674.

32.

Woodward-Lopez

, Gosliner

, Samuels

, et al. Lessons learned from evaluations of California's statewide school nutrition standards. Am J Public Health, 2010; 100:2137–2145.

33.

Cullen

, Watson

KB.

The impact of the Texas public school nutrition policy on student food selection and sales in Texas. Am J Public Health, 2009; 99:706–712.

34.

Gonzalez

, Jones

, Frongillo

EA.

Restricting snacks in US elementary schools is associated with higher frequency of fruit and vegetable consumption. J Nutr, 2009; 139:142–144.

35.

Templeton

, Marlette

, Panemangalore

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.

36.

Finkelstein

, Hill

, Whitaker

RC.

School food environments and policies in US public schools. Pediatrics, 2008; 122:e251–e259.

37.

Schwartz

, Novak

, Fiore

SS.

The impact of removing snacks of low nutritional value from middle schools. Health Educ Behav, 2009; 36:999–1011.

38.

Fletcher

, Frisvold

, Tefft

Taxing soft drinks and restricting access to vending machines to curb child obesity. Health Aff (Millwood), 2010; 29:1059–1066.

39.

Briefel

, Wilson

, Gleason

PM.

Consumption of low-nutrient, energy-dense foods and beverages at school, home, and other locations among school lunch participants and nonparticipants. J Am Diet Assoc, 2009; 109:S79–S90.

40.

Van Cauwenberghe

, Maes

, Spittaels

, et al. Effectiveness of school-based interventions in Europe to promote healthy nutrition in children and adolescents: Systematic review of published and ‘grey’ literature. Br J Nutr, 2004; 103:781–797.

41.

Waters

, de Silva-Sanigorski

, Hall

, et al. Interventions for preventing obesity in children. Cochrane Database Syst Rev, 2011:CD001871.

42.

Jones

, Chu

, Burke

, et al. A case for targeting marketing and availability in school food policy: Adolescents' food purchases at school and exposure to television, internet, and video games. J Hunger Environ Nutr, 2012; 7:1–10.

43.

Hamdan

, Story

, French

, et al. Perceptions of adolescents involved in promoting lower-fat foods in schools: Associations with level of involvement. J Am Diet Assoc, 2005; 105:247–251.

44.

Sallis

, McKenzie

, Conway

, et al. Environmental interventions for eating and physical activity: A randomized controlled trial in middle schools. Am J Prev Med, 2003; 24:209–217.

45.

Eaton

, Kann

, Kinchen

, et al. Youth risk behavior surveillance—United States, 2011. MMWR Surveill Summ, 2012; 61:1–162.