RESEARCH ARTICLE
State Farm-to-School Laws Influence the Availability of Fruits and Vegetables in School Lunches at US Public Elementary Schools LISA NICHOLSON, PhDa LINDSEY TURNER, PhDb LINDA SCHNEIDER, DC, MSc JAMIE CHRIQUI, PhDd FRANK CHALOUPKA, PhDe
ABSTRACT BACKGROUND: State laws and farm-to-school programs (FTSPs) have the potential to increase fruit and vegetable (FV) availability in school meals. This study examined whether FV were more available in public elementary school lunches in states with a law requiring/encouraging FTSPs or with a locally grown-related law, and whether the relationship between state laws and FV availability could be explained by schools opting for FTSPs. METHODS: A pooled, cross-sectional analysis linked a nationally representative sample of public elementary schools with state laws. A series of multivariate logistic regressions, controlling for school-level demographics were performed according to mediation analysis procedures for dichotomous outcomes. RESULTS Roughly 50% of schools reported FV availability in school lunches on most days of the week. Schools with the highest FV availability (70.6%) were in states with laws and schools with FTSPs. State laws requiring/encouraging FTSPs were significantly associated with increased FV availability in schools and a significant percentage (13%) of this relationship was mediated by schools having FTSPs. CONCLUSIONS: Because state farm-to-school laws are associated with significantly higher FV availability in schools—through FTSPs, as well as independently—enacting more state legislation may facilitate increased FTSP participation by schools and increased FV availability in school meals. Keywords: school meals; school lunch; farm-to-school programs; fruit and vegetable consumption. Citation: Nicholson L, Turner L, Schneider L, Chriqui J, Chaloupka F. State farm-to-school laws influence the availability of fruits and vegetables in school lunches at US public elementary schools. J Sch Health. 2014; 84: 310-316. Received on June 19, 2012 Accepted on June 4, 2013
xisting research1-3 has raised questions about the availability of fruits and vegetables (FV) in school meals served through the National School Lunch Program (NSLP) and School Breakfast Program. The School Nutrition and Dietary Assessment IV study, a nationally representative study of public schools in 2009-2010, revealed that as offered, on average NSLP lunches provided more than one-third of the recommended amounts of fruit (42% to 50%). As
E
served, average NSLP lunches provided considerably smaller amounts of recommended fruit (22% to 32%) specified by the Dietary Guidelines for Americans4 suggesting that many students did not select a serving of fruit for their lunch.3 As offered, average NSLP lunches provided about 30% (29% to 33%) of recommended amounts of vegetables.3,4 As served, average NSLP lunches provided slightly less (23% to 24%) of the recommended amounts of vegetables.
a Research Scientist, ([email protected]), University of Illinois at Chicago, Health Policy Center, Institute of Health Research and Policy, 1747 West Roosevelt Road, Chicago, IL 60608. bResearch Scientist, ([email protected]), University of Illinois at Chicago, Health Policy Center, Institute of Health Research and Policy, 1747 West Roosevelt Road, Chicago, IL 60608. c Visiting Research Specialist, ([email protected]), University of Illinois at Chicago, Health Policy Center, Institute of Health Research and Policy, 1747 West Roosevelt Road, Chicago, IL 60608. d Senior Research Scientist, ([email protected]), University of Illinois at Chicago, Health Policy Center, Institute of Health Research and Policy, 1747 West Roosevelt Road, Chicago, IL 60608. e Distinguished Professor of Economics, ([email protected]), University of Illinois at Chicago, Director Health Policy Center, Institute of Health Research and Policy, 1747 West Roosevelt Road, Chicago, IL 60608.
Address correspondence to: Lisa Nicholson, Research Scientist, ([email protected]), University of Illinois at Chicago, Institute of Health Research and Policy, 1747 West Roosevelt Road, Chicago, IL 60608. We would like to acknowledge financial support from the Robert Wood Johnson Foundation.
310 •
Journal of School Health
•
May 2014, Vol. 84, No. 5
•
© 2014, American School Health Association
Furthermore, in elementary schools, the average NSLP lunch, both offered and served, was low in dark green vegetables and legumes, providing only 7% to 15% of recommended amounts,4 and only 15% to 20% of the recommended amounts of red and orange vegetables.3 FVs were less available in schools with higher percentages of students eligible for free or reduced-price lunch.1 A nationally representative survey revealed that during the 2007-2008 school year, less than one-fourth of US public elementary school students had access to a salad bar at lunch on most days or every day of the week.5 State laws and school programs both have the potential to shape school meals, and in combination, may be even more powerful. A number of states have enacted farm-to-school (FTS) laws6-8 that can be grouped into 2 categories: (1) ‘‘formal,’’ codified laws that specifically require or encourage the state’s department of agriculture and/or education to establish a formal, FTS/cafeteria program in the state; and (2) other, related laws that address efforts to provide locally grown or locally produced FV to schools, without specifically mentioning a formal ‘‘FTS/cafeteria’’ program. The latter types of laws are often codified in a state’s board of education procurement laws and address giving procurement consideration to state- or locally based growers of fresh FV or agriculture. Another avenue by which state laws can impact FV availability in schools is by increasing the prevalence of farm-to-school programs (FTSPs). One study found that state FTS laws significantly increased the likelihood of a school having its own FTSP.9 Although no single definition of FTSPs exists, in general, the purpose of FTSPs is to connect schools with local or regional producers and to serve local or regionally produced foods in school cafeterias.10 The objective of FTSPs is to serve healthy meals in school cafeterias ´ including salad bars, hot entrees, and other meal items, educating students through classroom activities that can include snacks, taste-tests, and/or nutrition education programming, and improving student nutrition by reconnecting students with the origins of their food. Whereas virtually no studies to date examine the effect of FTSPs on school lunch menus, a review of school participation in FTSPs finds that 7 of 8 schools that incorporate a FTSP have a salad bar in the cafeteria, which increases students’ FV consumption between 25% and 84% per meal.11 Schools also report a 3% to 16% increase in school meal participation when farm-fresh food is served.10 This is the first study to examine whether state FTS laws or locally grown procurement laws have the potential to increase FV availability in schools independently. The purpose of this study is 2-fold. The first aim is to examine the extent to which US public elementary schools are more likely to incorporate FV into school meals, if they are in a state with a FTS Journal of School Health
•
law or in a state with a locally grown-related law. The second aim is to test the hypothesis that state FTS laws increase FTSPs, which in turn, increase FV availability in school meals.
METHODS Analyses used cross-sectional survey data from 1,792 public US elementary schools. Data were gathered during the 2006-2007, 2007-2008, and 20082009 school years via a larger study on health-related practices and policies in nationally representative samples of public elementary schools and districts, conducted by the Bridging the Gap Project.12 Elementary School Data Data were gathered with mail-back surveys conducted during the 2006-2007, 2007-2008, and 2008-2009 school years, with data collection occurring primarily between February and June of each school year. A survey was mailed to the school principal with a request that the school administrator module, regarding general school practices and policies pertaining to nutrition and physical activity, be completed by the administrator and that a separate part of the survey, pertaining to specific foods and beverages available on campus, be completed by food service personnel. A $100 incentive was offered to the principal or the school for completing and returning the survey. Surveys were processed and double-entered for quality assurance. Response rates were calculated using the American Association for Public Opinion Research method 2,13 counting partial responses as complete. Response rates and number of responding cases were as follows across the 3 years, respectively: 54.6% (578 schools), 70.6% (748 schools), and 61.8% (641 schools). The nationally representative samples were developed at the Institute for Social Research at the University of Michigan, with sampling frames based on datasets from the National Center for Education Statistics. Because elementary schools vary in grade composition, all schools in the sampling frame were required to include a third grade. Public schools from all contiguous US states were eligible for sampling. Instruments. Bridging the Gap researchers initially developed the surveys in 2006-2007 by identifying a variety of topics relevant to childhood obesity, including but not limited to recess, competitive food and beverage availability, and school meal characteristics. Survey items were selected by reviewing prior published work, adopting items from existing surveys already used by the project in secondary schools,14 and by seeking expert advice. The surveys were pretested with 3 school principals (not at schools in the sample) to verify comprehension, face validity, clarity and feasibility of completing the survey.
May 2014, Vol. 84, No. 5 •
© 2014, American School Health Association • 311
FTSPs were assessed with 1 item from the first school administrator module: ‘‘Does your school currently incorporate any locally grown foods (fruits, vegetables, meat, and/or dairy) into the meals offered at school (through, eg a ‘farm-to-cafeteria’, ‘farm-toschool’, or other program)?’’ Response options were 1 = yes, 0 = no. Availability of FV during lunches was measured with 2 items from the second food service personnel module, with instructions to indicate how often the following items were available to elementary students in the school lunch meal (not a` la carte): (1) fresh fruit and (2) vegetables. The survey items referred to ‘‘fresh fruit’’ but did not specify whether vegetables were fresh or not; in other words, cooked vegetables (whether previously frozen or canned) would have been included here, but cooked fruit would not. Considering that FTS programs generally supply schools with fresh produce, fruits are often served raw, whereas vegetable preparation often involves cooking or baking, as well as serving vegetables raw; this difference seems reasonable, and captures most of the likely categories for FV availability in lunches. Response options for both items were (1) never, (2) some days, and (3) most or every day. To create a dichotomous outcome that combined FV, data were recoded such that schools that offered FV on most days (coded = 1) were compared against schools that offered FV never or on some days (coded = 0). Any mention of this variable is referred to from this point forward as fruit and vegetable (FV). Preliminary analyses examined FV as 2 separate outcomes but results were similar, so for parsimony, the combined measure is used here. Weighting. Data were weighted to provide inference to all public elementary schools in the United States. The weights were adjusted for potential nonresponse bias by modeling every school’s propensity to respond. Variables used to model these adjustments included: student enrollment; percentage of black, white, and Latino students; percentage of students eligible for free or reduced-price lunch; US census region; and urbanicity. The school samples were designed to be nationally representative in each year; some schools participated in multiple years, but analyses were conducted as a stacked cross-sectional design. Contextual factors. To control for school-level factors that could confound the impact of the state laws on the school-reported practices, school-level demographic and socioeconomic data were obtained from the National Center for Education Statistics’ Common Core of Data.15 Data were obtained from the 2006-2007 file for the corresponding year of survey data, and from the 2007-2008 file for the 2007-2008 and 2008-2009 survey data (2008-2009 data were not available at the time of analysis). Data were obtained on the total number of students in the school, the 312
•
Journal of School Health
•
May 2014, Vol. 84, No. 5 •
percentage of students eligible for free or reducedprice lunch, race/ethnicity of the school students, and locale. Race or ethnicity of the school students was measured with 5 continuous variables that each reflected the proportion of the given school students that were white, black, Asian, and/or American-Indian (as compared to other), and the proportion of students that were Hispanic (as compared to non-Hispanic). State Laws State laws effective as of September 2006, 2007, and 2008 were collected by the authors using primary legal research techniques16 through searches of the Lexis and Westlaw state statutory and administrative law databases available by subscription. September of each year was chosen as reflective of the laws in place as of the beginning of each school year. The Lexis and Westlaw databases were searched through a combination of natural language and Boolean key word searches, as well as through reviews of the table of contents and indices, as available, for each state’s statutory and administrative law code for the years of interest. The state law information was also validated against secondary source compilations.6-8,16-18
Each state’s laws were reviewed and analyzed by 2 coders to create 2 dichotomous yes/no variables. The first variable was coded 1 = yes if the state law required or encouraged the state’s department of agriculture and/or education to establish a FTSP in the state or mentioned FTS activities, such as establishing a FTS website or income tax credit for local producers without specifically establishing a FTSP, and coded 0 = no if such a law did not exist. The second variable was coded 1 = yes if the state law addressed creating other initiatives to encourage providing locally grown or locally produced fresh FV to schools without specifically establishing or suggesting creation of a FTSP and coded 0 = no if such a law did not exist. Data Analysis This study is a pooled, cross-sectional analysis of the effect of state laws on school FTSPs and FV availability in public elementary schools. Due to the study design, whereby schools were clustered within districts, data were analyzed using the survey (svy) command in STATA SE v. 11.19 Analyses controlled for clustering within district, to account for the lack of independence between schools within the same district. Descriptive statistics were calculated for the predictor and outcome variables by year (Table 1). According to mediation analysis procedures for dichotomous outcomes,20 4 regression models were computed (Figure 1). These models all included control variables: school racial composition; student eligibility for free or reduced lunches (continuous); school size as indicated by number of students enrolled (continuous); locale, coded as city (referent), suburban, town, and rural;
© 2014, American School Health Association
Table 1. Weighted Percentages∗of Schools in States With Farm-to-School or Locally Grown-Related Laws, Schools With Farm-to-School Programs, and Schools With Fruit and Vegetable Availability, by School Year
Table 2. Weighted Percentages∗of Schools With Fruit and Vegetable Availability by State Farm-to-School Program Laws and School Farm-to-School Program Availability for the Full Sample of Schools (N = 1,792)
% of schools 2006-2007 2007-2008 2008-2009 (N = 528) (N = 687) (N = 577) School is in a state with a farm-to-school program law School is in a state with a locally grown-related law School has a farm-to-school program School has fruits and vegetables available in lunches on most days or every day
7.7
11.2
20.8
27.3
30.6
28.8
5.1
7.1
18.0
46.5
45.8
54.0
∗ Estimates are weighted at the school level.
Figure 1. Mediation Analysis Path Diagram With Separate Regression Coefficients (b1 , b2 , b3 , and b4 ) for Each Regression Model for the Full Sample of Schools (N = 1,792). School farm-to-school program (FTSP) 2 b = .09**
State farm-toschool (FTS) law
3 b = .13**
1 b = 0.10* 4 b = .08*
Fruit & vegetable (FV) availability in school meals
**p < .01; *p < .05
region, coded as South (referent), Northeast, South, and West. Two dummy variables were included as a control for year of data collection, to compare both 2007-2008 and 2008-2009 versus the 2006-2007 baseline. The Sobel Test statistic21 was used to evaluate statistical significance. Mediation analysis requires the evaluation and comparison of standardized regression coefficients; therefore, these coefficients are presented instead of odds ratios. After removing cases with missing data on school locally grown foods (N = 92) and percentage of students eligible for a free or reduced-price lunch (N = 3), 1,792 public elementary schools remained for analysis (N = 528, N = 687, and N = 577 schools, respectively across the 3 years). Schools were located in 729 school districts and 47 states, with an average of 2 to 1/2 schools per district (range: 1-25) and 39.8 schools per state (range: 5-220).
RESULTS Descriptive statistics for the dependent and independent variables over the school years are presented in Table 1. The percentage of schools with a FTSP more than tripled from the 2006-2007 to Journal of School Health
•
State farm-to School: farm to school law school program Yes No Yes No
Percentage of schools: fruits and vegetables available on most days or every day
Yes Yes No No
No
Yes
29.41% 35.90% 46.33% 51.49%
70.59% 64.10% 53.67% 48.51%
∗ Estimates are weighted at the school level.
2008-2009 school year (5.1% to 18.0%). This parallels the increase in schools located in states with FTS laws, which increased nearly 3-fold (7.7% to 20.8%). The percentage of schools located in states with locally grown procurement laws was slightly less than 30% over the 3-year period. Across time, roughly 50% of schools reported FV availability on most days or every day of the week. Table 2 presents a cross-tabulation of state laws, school FTSPs, and FV availability. FV availability was highest in schools that had a FTSP and were in a state with a FTS law. FV availability was more than 20 percentage points higher in these schools than for schools without FTSPs in states without FTS laws. Figure 1 depicts the results of the mediation analysis by displaying separate regression coefficients (b1 , b2 , b3 , b4 ) for each model and corresponding significance tests. Regression model b2 shows that states with laws requiring or encouraging FTSPs significantly and positively increase the number of schools with FTSPs (b2 = 0.09, p < .01). In turn, having a school FTSP significantly and positively increased FV availability (b3 = 0.13, p < .01). State FTS laws significantly increased FV availability in schools (b1 ) and 13% of this relationship was mediated by school FTSPs, as displayed by the reduction in coefficient b4 compared with b1 . The Sobel test (Sobel test statistic 2.5, p = .011) confirmed that a significant percentage of the relationship between state FTS laws and FV availability was explained by FTSPs. However, this relationship was not fully mediated, because state law remained a significant predictor of FV availability (p < .05). This is the first study utilizing a nationally representative sample of public elementary schools to examine the extent to which state laws and school FTSPs are associated with FV availability in school meals. Results suggest that state FTS laws increase FV availability in schools in 2 ways. First, state laws significantly increase the prevalence of school FTSPs, which in turn, increases FV availability in schools. Second, a residual effect of state laws remain, suggesting that
May 2014, Vol. 84, No. 5
•
© 2014, American School Health Association • 313
state laws significantly increase FV availability in schools, independent of FTSPs.
DISCUSSION Results from this study show that state laws have the potential to impact school meals both directly and indirectly through school-based FTSPs. One might expect that schools with FTSPs, which are more common in states with FTS laws, increase the availability of FV to their students. The direct impact of state FTS laws on FV availability, net of the school having a FTSP is more complex. This suggests that states with FTS laws significantly increase FV availability in schools, independent of school FTSPs. This may indicate that states with FTSP laws may be influencing schools to provide fresh FV to students, perhaps through other school food-related policies, even if the school does not have a FTSP. This research is critical to continue to encourage federal, state, and local legislation to continue to improve the quality of food provided in schools, especially with the recent focus specifically on the availability and consumption of FV. For example, recent legislation, The Healthy Hunger-Free Kids Act of 2010 (PUBLIC LAW 111-296; DEC 13, 2010) updated the requirements for school meals, aligning with the 2010 Dietary Guidelines for Americans,4 ensuring that no later than January 1, 2012, students are offered FV on all days of the week, including fresh, canned, dried, or frozen FV, in addition to whole grain products, lean meat products, and low-fat and nonfat dairy products. Furthermore, this act encourages schools, state and local agencies, Indian tribal organizations, nonprofits, and others that participate in the National School Lunch or Breakfast Programs to connect with local or regional farmers who can provide locally or regionally produced food for school cafeterias. Effective October 1, 2012, $5 million will be provided to the US Department of Agriculture (USDA) on an annual basis to support grants, including technical assistance and financial support to help schools build or improve their FTSPs; with the highest funding targeted at schools that serve a high proportion of children who are eligible for free or reduced-price meals. This study did not evaluate the impact of laws and FTSP programs on FV consumption, which is an important outcome variable. Additional research is needed to evaluate student-level measures of FV consumption, assessing both the amount and type of items consumed. Moreover, future research is needed to examine barriers to schools participating in FTSPs, to increase the prevalence of such programs. Limitations Results are subject to several limitations. As with any self-reported survey, there is potential 314
•
Journal of School Health
•
May 2014, Vol. 84, No. 5 •
for response bias or inaccurate responding based on incomplete knowledge or lack of information. These results pertain only to elementary schools, not secondary schools. In addition, the study did not examine the extent to which schools sourced locally grown foods, types of items, how many items were sourced locally, or how frequently local items were offered. Another potential limitation is that there are other state laws that could impact the availability of FV and some laws may have stronger associations with increasing access to FV than others. In addition to state laws, other factors such as participation in the Healthier US School Challenge and implementation of local school wellness policies could also be driving forces in increasing FV availability in schools. Much more research is needed to understand additional mechanisms by which state laws could impact school meals. Finally, the study design was cross-sectional and does not allow causality to be determined. Conclusions Although the number of states with FTS laws and schools incorporating locally grown foods into school meals has nearly tripled in recent years, and this evolution is paralleled by an increase in the availability of FV in school meals, more research is needed to understand the barriers schools face in incorporating FVs in school meals. Future research should investigate whether FV availability actually increases FV intake among students, including elementary, middle, and high school students.
IMPLICATIONS FOR SCHOOL HEALTH On the basis of data available through 2006, 6.6% of school districts required schools to make FV available whenever other foods are offered or served.22 One objective of Healthy People 202022 sets a goal to increase the percentage of school districts requiring schools to make fruits or vegetables available when other foods are served to 18.6%. Our results suggest that school-based FTSPs are one avenue by which schools have increased FV availability in school meals. Results also suggest that state FTS laws have the potential to directly increase FV availability in school meals. Encouragingly, the number of states with FTSP-relevant legislation has steadily increased in the past few years.9 In addition, recent federal legislation, The Healthy Hunger-Free Kids Act of 2010 (PUBLIC LAW 111-296; DEC 13, 2010), updated the requirements for school meals to ensure that students are offered FV (including fresh, canned, dried, or frozen) on all days of the week. Although such items may not necessarily be sourced from FTSP programs, the increased demand for such products may facilitate more connections between local farmers (eg, produce © 2014, American School Health Association
suppliers) and schools. As demonstrated here, such FTSP relationships effectively increase student access to fruit and vegetables in school meals. The recent federal and state legislative actions regarding FTSP programs and meal requirements are all likely to contribute to the goal of promoting increased FV consumption among children and adolescents. In addition, the following specific recommendations for state and local officials are offered. (1) Financial support is important for the development of such programs. States can assist schools to create FTSPs by providing funding, grants, and income-tax credits for schools that implement such programs. (2) Although the literature shows that the prevalence of states having FTSP laws has increased, many of these laws are weak.9 Thus, an area for improvement is in strengthening existing laws, as well as developing new laws in states where they do not already exist. Such laws might be more effective where they specifically address issues relevant to the implementation of FTSP programs, such as training staff and providing ongoing technical assistance for schools. (3) Facilitating coordination among multiple stakeholders is important because coordination and distribution of food between the farmer and the school is one of the biggest challenges for FTSPs. State-level advisory councils may be helpful in promoting district-level and school-level adoption of FTSPs, and such councils could also foster communication and cooperation among individuals from various government organizations (ie, State Departments of Education, Public Health, and Agriculture) as well as nonprofit community groups and the health-care sector. (4) At the school and district level, several resources already exist to support FTSP programming. In some states, the opportunities recommended above (ie, state funding or tax credits, advisory councils) are already in place; schools are encouraged to make use of such resources. In addition, at the school level, appointing a school-level ‘‘nutrition champion’’ or key person for FTSP and other nutrition-related efforts can help to coordinate efforts. Federal organizations such as the USDA offer technical assistance and small grants to FTSP programs, and other nonprofit organizations nationwide offer small grants for such programs. Human Subjects Approval Statement All research protocols and survey materials were approved by the Institutional Review Board at the University of Illinois at Chicago, with a waiver of written informed consent.
REFERENCES 1. Finkelstein DM, Hill EL, Whitaker R. School food environments and policies in U.S. public schools. Pediatrics. 2008;122(1):e251e259.
Journal of School Health
•
2. Condon EM, Crepinsek MK, Fox MK. School meals: the types of foods offered to and consumed by children at lunch and breakfast. J Am Diet Assoc. 2009;109(suppl 1):S67-S78. 3. US Department of Agriculture, Food and Nutrition Service, Office of Research and Analysis, School Nutrition Dietary Assessment Study IV, Vol. I: School Foodservice Operations, School Environments, and Meals Offered and Served, by Mary Kay Fox, Elizabeth Condon, Mary Kay Crepinsek, et al. Project Officer, Fred Lesnett Alexandria, VA: November 2012. Available at: http://www.fns.usda.gov/Ora/menu/Published/CNP/FILES/ SNDA-IV_Vol1Pt1.pdf. Accessed January 5, 2012. 4. US Department of Health and Human Services and US Department of Agriculture. Dietary Guidelines for Americans 2010, 7th ed. Available at: http://www.health.gov/dietary guidelines/dga2010/DietaryGuidelines2010.pdf/. Accessed January 23, 2013. 5. Turner L, Chaloupka FJ, Chriqui JF, Sandoval A. School Policies and Practices to Improve Health and Prevent Obesity: National Elementary School Survey Results: School Years 2006-07 and 2007-08. Vol I. Chicago, IL: Bridging the Gap Program, Health Policy Center, Institute for Health Research and Policy, University of Illinois at Chicago, 2010. Available at: www.bridgingthegapresearch.org. Accessed January 5, 2012. 6. Trust for American’s Health. F as in FAT: How Obesity Policies are Failing America, 2007. Washington, DC: Trust for American’s Health; 2007. 7. Trust for America’s Health. F as in FAT: How Obesity Policies are Failing America, 2008. Washington, DC: Trust for America’s Health; 2008. 8. Trust for America’s Health. F as in FAT: How Obesity Policies are Failing America, 2009. Washington, DC: Trust for America’s Health; 2009. 9. Schneider L, Chriqui JF, Nicholson L, Turner L, Gourdet C, Chaloupka FJ. Are farm-to-school programs more common in states with farm-to-school-related laws? J Sch Health. 2012;82(5):210-216. 10. Joshi A, Azuma AM. Bearing Fruit: Farm to School Program Evaluation Resources and Recommendations. Los Angeles, CA: National Farm to School Network, Center for Food & Urban & Environmental Policy Institute, Occidental College; 2009. 11. Joshi A, Azuma AM, Feenstra G. Do farm-to-school programs make a difference? Findings and future research needs. J Hunger Environ Nutr. 2008;3(2-3):229-246. 12. Turner L, Chaloupka FJ and Sandoval A. School Policies and Practices for Improving Children’s Health: National Elementary School Survey Results: School Years 2006-07 through 2009-10. Vol. II. Chicago, IL: Bridging the Gap Program, Health Policy Center, Institute for Health Research and Policy, University of Illinois at Chicago, 2012. Available at: www.bridgingthegapresearch.org. Accessed July 5, 2012. 13. The American Association for Public Opinion Research. Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys. 2009. Available at: http://www.aapor. org/AM/Template.cfm?Section=Standard_Definitions1&Templ ate=/CM/ContentDisplay.cfm&ContentID=1814. Accessed July 23, 2012. 14. Delva J, O’Malley PM, Johnston LD. Availability of morehealthy and less-healthy food choices in American schools: a national study of grade, racial/ethnic, and socioeconomic differences. Am J Prev Med. 2007;33(suppl 4):S226-S239. 15. National Center for Education Statistics. US Department of Education. Institute of Education Sciences, National Center for Education Statistics, 2006-2007. Available at: http://nces.ed.gov/ccd/pubschuniv.asp. Accessed January 10, 2011. 16. Mersky RM, Dunn DJ. Fundamentals of Legal Research. 8th ed. New York, NY: Foundation Press; 2002.
May 2014, Vol. 84, No. 5
•
© 2014, American School Health Association • 315
17. Coalition for Food Security. Farm to school legislation: a State by State listing, 2008. Available at: http://www.food security.org/policy/StateByStateFarmtoSchoolLegislation.pdf. Accessed August 1, 2010. 18. National Conference of State Legislatures’ Healthy community design and access to healthy food legislation database. Available at: www.ncsl.org. Accessed September 5, 2010. 19. Kreuter F, Valliant R. A survey on survey statistics: what is done and can be done in Stata. Stata J. 2007;7:1-21.
316
•
Journal of School Health
•
May 2014, Vol. 84, No. 5 •
20. Kenny DA. Mediation with dichotomous outcomes. 2008. Available at: http://davidkenny.net/doc/dihnmed.doc. Accessed January 11, 2011. 21. Sobel ME. Asymptotic confidence intervals for indirect effects in structural equation models. Sociol Methodol. 1982;13: 290-312. 22. United States Department of Health and Human Services. Office of Disease and Health Promotion. Healthy People 2020. Available at: www.healthypeople.gov. Accessed May 25, 2013.
© 2014, American School Health Association
State Farm-to-School Laws Influence the Availability of Fruits and Vegetables in School Lunches at US Public Elementary Schools LISA NICHOLSON, PhDa LINDSEY TURNER, PhDb LINDA SCHNEIDER, DC, MSc JAMIE CHRIQUI, PhDd FRANK CHALOUPKA, PhDe
ABSTRACT BACKGROUND: State laws and farm-to-school programs (FTSPs) have the potential to increase fruit and vegetable (FV) availability in school meals. This study examined whether FV were more available in public elementary school lunches in states with a law requiring/encouraging FTSPs or with a locally grown-related law, and whether the relationship between state laws and FV availability could be explained by schools opting for FTSPs. METHODS: A pooled, cross-sectional analysis linked a nationally representative sample of public elementary schools with state laws. A series of multivariate logistic regressions, controlling for school-level demographics were performed according to mediation analysis procedures for dichotomous outcomes. RESULTS Roughly 50% of schools reported FV availability in school lunches on most days of the week. Schools with the highest FV availability (70.6%) were in states with laws and schools with FTSPs. State laws requiring/encouraging FTSPs were significantly associated with increased FV availability in schools and a significant percentage (13%) of this relationship was mediated by schools having FTSPs. CONCLUSIONS: Because state farm-to-school laws are associated with significantly higher FV availability in schools—through FTSPs, as well as independently—enacting more state legislation may facilitate increased FTSP participation by schools and increased FV availability in school meals. Keywords: school meals; school lunch; farm-to-school programs; fruit and vegetable consumption. Citation: Nicholson L, Turner L, Schneider L, Chriqui J, Chaloupka F. State farm-to-school laws influence the availability of fruits and vegetables in school lunches at US public elementary schools. J Sch Health. 2014; 84: 310-316. Received on June 19, 2012 Accepted on June 4, 2013
xisting research1-3 has raised questions about the availability of fruits and vegetables (FV) in school meals served through the National School Lunch Program (NSLP) and School Breakfast Program. The School Nutrition and Dietary Assessment IV study, a nationally representative study of public schools in 2009-2010, revealed that as offered, on average NSLP lunches provided more than one-third of the recommended amounts of fruit (42% to 50%). As
E
served, average NSLP lunches provided considerably smaller amounts of recommended fruit (22% to 32%) specified by the Dietary Guidelines for Americans4 suggesting that many students did not select a serving of fruit for their lunch.3 As offered, average NSLP lunches provided about 30% (29% to 33%) of recommended amounts of vegetables.3,4 As served, average NSLP lunches provided slightly less (23% to 24%) of the recommended amounts of vegetables.
a Research Scientist, ([email protected]), University of Illinois at Chicago, Health Policy Center, Institute of Health Research and Policy, 1747 West Roosevelt Road, Chicago, IL 60608. bResearch Scientist, ([email protected]), University of Illinois at Chicago, Health Policy Center, Institute of Health Research and Policy, 1747 West Roosevelt Road, Chicago, IL 60608. c Visiting Research Specialist, ([email protected]), University of Illinois at Chicago, Health Policy Center, Institute of Health Research and Policy, 1747 West Roosevelt Road, Chicago, IL 60608. d Senior Research Scientist, ([email protected]), University of Illinois at Chicago, Health Policy Center, Institute of Health Research and Policy, 1747 West Roosevelt Road, Chicago, IL 60608. e Distinguished Professor of Economics, ([email protected]), University of Illinois at Chicago, Director Health Policy Center, Institute of Health Research and Policy, 1747 West Roosevelt Road, Chicago, IL 60608.
Address correspondence to: Lisa Nicholson, Research Scientist, ([email protected]), University of Illinois at Chicago, Institute of Health Research and Policy, 1747 West Roosevelt Road, Chicago, IL 60608. We would like to acknowledge financial support from the Robert Wood Johnson Foundation.
310 •
Journal of School Health
•
May 2014, Vol. 84, No. 5
•
© 2014, American School Health Association
Furthermore, in elementary schools, the average NSLP lunch, both offered and served, was low in dark green vegetables and legumes, providing only 7% to 15% of recommended amounts,4 and only 15% to 20% of the recommended amounts of red and orange vegetables.3 FVs were less available in schools with higher percentages of students eligible for free or reduced-price lunch.1 A nationally representative survey revealed that during the 2007-2008 school year, less than one-fourth of US public elementary school students had access to a salad bar at lunch on most days or every day of the week.5 State laws and school programs both have the potential to shape school meals, and in combination, may be even more powerful. A number of states have enacted farm-to-school (FTS) laws6-8 that can be grouped into 2 categories: (1) ‘‘formal,’’ codified laws that specifically require or encourage the state’s department of agriculture and/or education to establish a formal, FTS/cafeteria program in the state; and (2) other, related laws that address efforts to provide locally grown or locally produced FV to schools, without specifically mentioning a formal ‘‘FTS/cafeteria’’ program. The latter types of laws are often codified in a state’s board of education procurement laws and address giving procurement consideration to state- or locally based growers of fresh FV or agriculture. Another avenue by which state laws can impact FV availability in schools is by increasing the prevalence of farm-to-school programs (FTSPs). One study found that state FTS laws significantly increased the likelihood of a school having its own FTSP.9 Although no single definition of FTSPs exists, in general, the purpose of FTSPs is to connect schools with local or regional producers and to serve local or regionally produced foods in school cafeterias.10 The objective of FTSPs is to serve healthy meals in school cafeterias ´ including salad bars, hot entrees, and other meal items, educating students through classroom activities that can include snacks, taste-tests, and/or nutrition education programming, and improving student nutrition by reconnecting students with the origins of their food. Whereas virtually no studies to date examine the effect of FTSPs on school lunch menus, a review of school participation in FTSPs finds that 7 of 8 schools that incorporate a FTSP have a salad bar in the cafeteria, which increases students’ FV consumption between 25% and 84% per meal.11 Schools also report a 3% to 16% increase in school meal participation when farm-fresh food is served.10 This is the first study to examine whether state FTS laws or locally grown procurement laws have the potential to increase FV availability in schools independently. The purpose of this study is 2-fold. The first aim is to examine the extent to which US public elementary schools are more likely to incorporate FV into school meals, if they are in a state with a FTS Journal of School Health
•
law or in a state with a locally grown-related law. The second aim is to test the hypothesis that state FTS laws increase FTSPs, which in turn, increase FV availability in school meals.
METHODS Analyses used cross-sectional survey data from 1,792 public US elementary schools. Data were gathered during the 2006-2007, 2007-2008, and 20082009 school years via a larger study on health-related practices and policies in nationally representative samples of public elementary schools and districts, conducted by the Bridging the Gap Project.12 Elementary School Data Data were gathered with mail-back surveys conducted during the 2006-2007, 2007-2008, and 2008-2009 school years, with data collection occurring primarily between February and June of each school year. A survey was mailed to the school principal with a request that the school administrator module, regarding general school practices and policies pertaining to nutrition and physical activity, be completed by the administrator and that a separate part of the survey, pertaining to specific foods and beverages available on campus, be completed by food service personnel. A $100 incentive was offered to the principal or the school for completing and returning the survey. Surveys were processed and double-entered for quality assurance. Response rates were calculated using the American Association for Public Opinion Research method 2,13 counting partial responses as complete. Response rates and number of responding cases were as follows across the 3 years, respectively: 54.6% (578 schools), 70.6% (748 schools), and 61.8% (641 schools). The nationally representative samples were developed at the Institute for Social Research at the University of Michigan, with sampling frames based on datasets from the National Center for Education Statistics. Because elementary schools vary in grade composition, all schools in the sampling frame were required to include a third grade. Public schools from all contiguous US states were eligible for sampling. Instruments. Bridging the Gap researchers initially developed the surveys in 2006-2007 by identifying a variety of topics relevant to childhood obesity, including but not limited to recess, competitive food and beverage availability, and school meal characteristics. Survey items were selected by reviewing prior published work, adopting items from existing surveys already used by the project in secondary schools,14 and by seeking expert advice. The surveys were pretested with 3 school principals (not at schools in the sample) to verify comprehension, face validity, clarity and feasibility of completing the survey.
May 2014, Vol. 84, No. 5 •
© 2014, American School Health Association • 311
FTSPs were assessed with 1 item from the first school administrator module: ‘‘Does your school currently incorporate any locally grown foods (fruits, vegetables, meat, and/or dairy) into the meals offered at school (through, eg a ‘farm-to-cafeteria’, ‘farm-toschool’, or other program)?’’ Response options were 1 = yes, 0 = no. Availability of FV during lunches was measured with 2 items from the second food service personnel module, with instructions to indicate how often the following items were available to elementary students in the school lunch meal (not a` la carte): (1) fresh fruit and (2) vegetables. The survey items referred to ‘‘fresh fruit’’ but did not specify whether vegetables were fresh or not; in other words, cooked vegetables (whether previously frozen or canned) would have been included here, but cooked fruit would not. Considering that FTS programs generally supply schools with fresh produce, fruits are often served raw, whereas vegetable preparation often involves cooking or baking, as well as serving vegetables raw; this difference seems reasonable, and captures most of the likely categories for FV availability in lunches. Response options for both items were (1) never, (2) some days, and (3) most or every day. To create a dichotomous outcome that combined FV, data were recoded such that schools that offered FV on most days (coded = 1) were compared against schools that offered FV never or on some days (coded = 0). Any mention of this variable is referred to from this point forward as fruit and vegetable (FV). Preliminary analyses examined FV as 2 separate outcomes but results were similar, so for parsimony, the combined measure is used here. Weighting. Data were weighted to provide inference to all public elementary schools in the United States. The weights were adjusted for potential nonresponse bias by modeling every school’s propensity to respond. Variables used to model these adjustments included: student enrollment; percentage of black, white, and Latino students; percentage of students eligible for free or reduced-price lunch; US census region; and urbanicity. The school samples were designed to be nationally representative in each year; some schools participated in multiple years, but analyses were conducted as a stacked cross-sectional design. Contextual factors. To control for school-level factors that could confound the impact of the state laws on the school-reported practices, school-level demographic and socioeconomic data were obtained from the National Center for Education Statistics’ Common Core of Data.15 Data were obtained from the 2006-2007 file for the corresponding year of survey data, and from the 2007-2008 file for the 2007-2008 and 2008-2009 survey data (2008-2009 data were not available at the time of analysis). Data were obtained on the total number of students in the school, the 312
•
Journal of School Health
•
May 2014, Vol. 84, No. 5 •
percentage of students eligible for free or reducedprice lunch, race/ethnicity of the school students, and locale. Race or ethnicity of the school students was measured with 5 continuous variables that each reflected the proportion of the given school students that were white, black, Asian, and/or American-Indian (as compared to other), and the proportion of students that were Hispanic (as compared to non-Hispanic). State Laws State laws effective as of September 2006, 2007, and 2008 were collected by the authors using primary legal research techniques16 through searches of the Lexis and Westlaw state statutory and administrative law databases available by subscription. September of each year was chosen as reflective of the laws in place as of the beginning of each school year. The Lexis and Westlaw databases were searched through a combination of natural language and Boolean key word searches, as well as through reviews of the table of contents and indices, as available, for each state’s statutory and administrative law code for the years of interest. The state law information was also validated against secondary source compilations.6-8,16-18
Each state’s laws were reviewed and analyzed by 2 coders to create 2 dichotomous yes/no variables. The first variable was coded 1 = yes if the state law required or encouraged the state’s department of agriculture and/or education to establish a FTSP in the state or mentioned FTS activities, such as establishing a FTS website or income tax credit for local producers without specifically establishing a FTSP, and coded 0 = no if such a law did not exist. The second variable was coded 1 = yes if the state law addressed creating other initiatives to encourage providing locally grown or locally produced fresh FV to schools without specifically establishing or suggesting creation of a FTSP and coded 0 = no if such a law did not exist. Data Analysis This study is a pooled, cross-sectional analysis of the effect of state laws on school FTSPs and FV availability in public elementary schools. Due to the study design, whereby schools were clustered within districts, data were analyzed using the survey (svy) command in STATA SE v. 11.19 Analyses controlled for clustering within district, to account for the lack of independence between schools within the same district. Descriptive statistics were calculated for the predictor and outcome variables by year (Table 1). According to mediation analysis procedures for dichotomous outcomes,20 4 regression models were computed (Figure 1). These models all included control variables: school racial composition; student eligibility for free or reduced lunches (continuous); school size as indicated by number of students enrolled (continuous); locale, coded as city (referent), suburban, town, and rural;
© 2014, American School Health Association
Table 1. Weighted Percentages∗of Schools in States With Farm-to-School or Locally Grown-Related Laws, Schools With Farm-to-School Programs, and Schools With Fruit and Vegetable Availability, by School Year
Table 2. Weighted Percentages∗of Schools With Fruit and Vegetable Availability by State Farm-to-School Program Laws and School Farm-to-School Program Availability for the Full Sample of Schools (N = 1,792)
% of schools 2006-2007 2007-2008 2008-2009 (N = 528) (N = 687) (N = 577) School is in a state with a farm-to-school program law School is in a state with a locally grown-related law School has a farm-to-school program School has fruits and vegetables available in lunches on most days or every day
7.7
11.2
20.8
27.3
30.6
28.8
5.1
7.1
18.0
46.5
45.8
54.0
∗ Estimates are weighted at the school level.
Figure 1. Mediation Analysis Path Diagram With Separate Regression Coefficients (b1 , b2 , b3 , and b4 ) for Each Regression Model for the Full Sample of Schools (N = 1,792). School farm-to-school program (FTSP) 2 b = .09**
State farm-toschool (FTS) law
3 b = .13**
1 b = 0.10* 4 b = .08*
Fruit & vegetable (FV) availability in school meals
**p < .01; *p < .05
region, coded as South (referent), Northeast, South, and West. Two dummy variables were included as a control for year of data collection, to compare both 2007-2008 and 2008-2009 versus the 2006-2007 baseline. The Sobel Test statistic21 was used to evaluate statistical significance. Mediation analysis requires the evaluation and comparison of standardized regression coefficients; therefore, these coefficients are presented instead of odds ratios. After removing cases with missing data on school locally grown foods (N = 92) and percentage of students eligible for a free or reduced-price lunch (N = 3), 1,792 public elementary schools remained for analysis (N = 528, N = 687, and N = 577 schools, respectively across the 3 years). Schools were located in 729 school districts and 47 states, with an average of 2 to 1/2 schools per district (range: 1-25) and 39.8 schools per state (range: 5-220).
RESULTS Descriptive statistics for the dependent and independent variables over the school years are presented in Table 1. The percentage of schools with a FTSP more than tripled from the 2006-2007 to Journal of School Health
•
State farm-to School: farm to school law school program Yes No Yes No
Percentage of schools: fruits and vegetables available on most days or every day
Yes Yes No No
No
Yes
29.41% 35.90% 46.33% 51.49%
70.59% 64.10% 53.67% 48.51%
∗ Estimates are weighted at the school level.
2008-2009 school year (5.1% to 18.0%). This parallels the increase in schools located in states with FTS laws, which increased nearly 3-fold (7.7% to 20.8%). The percentage of schools located in states with locally grown procurement laws was slightly less than 30% over the 3-year period. Across time, roughly 50% of schools reported FV availability on most days or every day of the week. Table 2 presents a cross-tabulation of state laws, school FTSPs, and FV availability. FV availability was highest in schools that had a FTSP and were in a state with a FTS law. FV availability was more than 20 percentage points higher in these schools than for schools without FTSPs in states without FTS laws. Figure 1 depicts the results of the mediation analysis by displaying separate regression coefficients (b1 , b2 , b3 , b4 ) for each model and corresponding significance tests. Regression model b2 shows that states with laws requiring or encouraging FTSPs significantly and positively increase the number of schools with FTSPs (b2 = 0.09, p < .01). In turn, having a school FTSP significantly and positively increased FV availability (b3 = 0.13, p < .01). State FTS laws significantly increased FV availability in schools (b1 ) and 13% of this relationship was mediated by school FTSPs, as displayed by the reduction in coefficient b4 compared with b1 . The Sobel test (Sobel test statistic 2.5, p = .011) confirmed that a significant percentage of the relationship between state FTS laws and FV availability was explained by FTSPs. However, this relationship was not fully mediated, because state law remained a significant predictor of FV availability (p < .05). This is the first study utilizing a nationally representative sample of public elementary schools to examine the extent to which state laws and school FTSPs are associated with FV availability in school meals. Results suggest that state FTS laws increase FV availability in schools in 2 ways. First, state laws significantly increase the prevalence of school FTSPs, which in turn, increases FV availability in schools. Second, a residual effect of state laws remain, suggesting that
May 2014, Vol. 84, No. 5
•
© 2014, American School Health Association • 313
state laws significantly increase FV availability in schools, independent of FTSPs.
DISCUSSION Results from this study show that state laws have the potential to impact school meals both directly and indirectly through school-based FTSPs. One might expect that schools with FTSPs, which are more common in states with FTS laws, increase the availability of FV to their students. The direct impact of state FTS laws on FV availability, net of the school having a FTSP is more complex. This suggests that states with FTS laws significantly increase FV availability in schools, independent of school FTSPs. This may indicate that states with FTSP laws may be influencing schools to provide fresh FV to students, perhaps through other school food-related policies, even if the school does not have a FTSP. This research is critical to continue to encourage federal, state, and local legislation to continue to improve the quality of food provided in schools, especially with the recent focus specifically on the availability and consumption of FV. For example, recent legislation, The Healthy Hunger-Free Kids Act of 2010 (PUBLIC LAW 111-296; DEC 13, 2010) updated the requirements for school meals, aligning with the 2010 Dietary Guidelines for Americans,4 ensuring that no later than January 1, 2012, students are offered FV on all days of the week, including fresh, canned, dried, or frozen FV, in addition to whole grain products, lean meat products, and low-fat and nonfat dairy products. Furthermore, this act encourages schools, state and local agencies, Indian tribal organizations, nonprofits, and others that participate in the National School Lunch or Breakfast Programs to connect with local or regional farmers who can provide locally or regionally produced food for school cafeterias. Effective October 1, 2012, $5 million will be provided to the US Department of Agriculture (USDA) on an annual basis to support grants, including technical assistance and financial support to help schools build or improve their FTSPs; with the highest funding targeted at schools that serve a high proportion of children who are eligible for free or reduced-price meals. This study did not evaluate the impact of laws and FTSP programs on FV consumption, which is an important outcome variable. Additional research is needed to evaluate student-level measures of FV consumption, assessing both the amount and type of items consumed. Moreover, future research is needed to examine barriers to schools participating in FTSPs, to increase the prevalence of such programs. Limitations Results are subject to several limitations. As with any self-reported survey, there is potential 314
•
Journal of School Health
•
May 2014, Vol. 84, No. 5 •
for response bias or inaccurate responding based on incomplete knowledge or lack of information. These results pertain only to elementary schools, not secondary schools. In addition, the study did not examine the extent to which schools sourced locally grown foods, types of items, how many items were sourced locally, or how frequently local items were offered. Another potential limitation is that there are other state laws that could impact the availability of FV and some laws may have stronger associations with increasing access to FV than others. In addition to state laws, other factors such as participation in the Healthier US School Challenge and implementation of local school wellness policies could also be driving forces in increasing FV availability in schools. Much more research is needed to understand additional mechanisms by which state laws could impact school meals. Finally, the study design was cross-sectional and does not allow causality to be determined. Conclusions Although the number of states with FTS laws and schools incorporating locally grown foods into school meals has nearly tripled in recent years, and this evolution is paralleled by an increase in the availability of FV in school meals, more research is needed to understand the barriers schools face in incorporating FVs in school meals. Future research should investigate whether FV availability actually increases FV intake among students, including elementary, middle, and high school students.
IMPLICATIONS FOR SCHOOL HEALTH On the basis of data available through 2006, 6.6% of school districts required schools to make FV available whenever other foods are offered or served.22 One objective of Healthy People 202022 sets a goal to increase the percentage of school districts requiring schools to make fruits or vegetables available when other foods are served to 18.6%. Our results suggest that school-based FTSPs are one avenue by which schools have increased FV availability in school meals. Results also suggest that state FTS laws have the potential to directly increase FV availability in school meals. Encouragingly, the number of states with FTSP-relevant legislation has steadily increased in the past few years.9 In addition, recent federal legislation, The Healthy Hunger-Free Kids Act of 2010 (PUBLIC LAW 111-296; DEC 13, 2010), updated the requirements for school meals to ensure that students are offered FV (including fresh, canned, dried, or frozen) on all days of the week. Although such items may not necessarily be sourced from FTSP programs, the increased demand for such products may facilitate more connections between local farmers (eg, produce © 2014, American School Health Association
suppliers) and schools. As demonstrated here, such FTSP relationships effectively increase student access to fruit and vegetables in school meals. The recent federal and state legislative actions regarding FTSP programs and meal requirements are all likely to contribute to the goal of promoting increased FV consumption among children and adolescents. In addition, the following specific recommendations for state and local officials are offered. (1) Financial support is important for the development of such programs. States can assist schools to create FTSPs by providing funding, grants, and income-tax credits for schools that implement such programs. (2) Although the literature shows that the prevalence of states having FTSP laws has increased, many of these laws are weak.9 Thus, an area for improvement is in strengthening existing laws, as well as developing new laws in states where they do not already exist. Such laws might be more effective where they specifically address issues relevant to the implementation of FTSP programs, such as training staff and providing ongoing technical assistance for schools. (3) Facilitating coordination among multiple stakeholders is important because coordination and distribution of food between the farmer and the school is one of the biggest challenges for FTSPs. State-level advisory councils may be helpful in promoting district-level and school-level adoption of FTSPs, and such councils could also foster communication and cooperation among individuals from various government organizations (ie, State Departments of Education, Public Health, and Agriculture) as well as nonprofit community groups and the health-care sector. (4) At the school and district level, several resources already exist to support FTSP programming. In some states, the opportunities recommended above (ie, state funding or tax credits, advisory councils) are already in place; schools are encouraged to make use of such resources. In addition, at the school level, appointing a school-level ‘‘nutrition champion’’ or key person for FTSP and other nutrition-related efforts can help to coordinate efforts. Federal organizations such as the USDA offer technical assistance and small grants to FTSP programs, and other nonprofit organizations nationwide offer small grants for such programs. Human Subjects Approval Statement All research protocols and survey materials were approved by the Institutional Review Board at the University of Illinois at Chicago, with a waiver of written informed consent.
REFERENCES 1. Finkelstein DM, Hill EL, Whitaker R. School food environments and policies in U.S. public schools. Pediatrics. 2008;122(1):e251e259.
Journal of School Health
•
2. Condon EM, Crepinsek MK, Fox MK. School meals: the types of foods offered to and consumed by children at lunch and breakfast. J Am Diet Assoc. 2009;109(suppl 1):S67-S78. 3. US Department of Agriculture, Food and Nutrition Service, Office of Research and Analysis, School Nutrition Dietary Assessment Study IV, Vol. I: School Foodservice Operations, School Environments, and Meals Offered and Served, by Mary Kay Fox, Elizabeth Condon, Mary Kay Crepinsek, et al. Project Officer, Fred Lesnett Alexandria, VA: November 2012. Available at: http://www.fns.usda.gov/Ora/menu/Published/CNP/FILES/ SNDA-IV_Vol1Pt1.pdf. Accessed January 5, 2012. 4. US Department of Health and Human Services and US Department of Agriculture. Dietary Guidelines for Americans 2010, 7th ed. Available at: http://www.health.gov/dietary guidelines/dga2010/DietaryGuidelines2010.pdf/. Accessed January 23, 2013. 5. Turner L, Chaloupka FJ, Chriqui JF, Sandoval A. School Policies and Practices to Improve Health and Prevent Obesity: National Elementary School Survey Results: School Years 2006-07 and 2007-08. Vol I. Chicago, IL: Bridging the Gap Program, Health Policy Center, Institute for Health Research and Policy, University of Illinois at Chicago, 2010. Available at: www.bridgingthegapresearch.org. Accessed January 5, 2012. 6. Trust for American’s Health. F as in FAT: How Obesity Policies are Failing America, 2007. Washington, DC: Trust for American’s Health; 2007. 7. Trust for America’s Health. F as in FAT: How Obesity Policies are Failing America, 2008. Washington, DC: Trust for America’s Health; 2008. 8. Trust for America’s Health. F as in FAT: How Obesity Policies are Failing America, 2009. Washington, DC: Trust for America’s Health; 2009. 9. Schneider L, Chriqui JF, Nicholson L, Turner L, Gourdet C, Chaloupka FJ. Are farm-to-school programs more common in states with farm-to-school-related laws? J Sch Health. 2012;82(5):210-216. 10. Joshi A, Azuma AM. Bearing Fruit: Farm to School Program Evaluation Resources and Recommendations. Los Angeles, CA: National Farm to School Network, Center for Food & Urban & Environmental Policy Institute, Occidental College; 2009. 11. Joshi A, Azuma AM, Feenstra G. Do farm-to-school programs make a difference? Findings and future research needs. J Hunger Environ Nutr. 2008;3(2-3):229-246. 12. Turner L, Chaloupka FJ and Sandoval A. School Policies and Practices for Improving Children’s Health: National Elementary School Survey Results: School Years 2006-07 through 2009-10. Vol. II. Chicago, IL: Bridging the Gap Program, Health Policy Center, Institute for Health Research and Policy, University of Illinois at Chicago, 2012. Available at: www.bridgingthegapresearch.org. Accessed July 5, 2012. 13. The American Association for Public Opinion Research. Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys. 2009. Available at: http://www.aapor. org/AM/Template.cfm?Section=Standard_Definitions1&Templ ate=/CM/ContentDisplay.cfm&ContentID=1814. Accessed July 23, 2012. 14. Delva J, O’Malley PM, Johnston LD. Availability of morehealthy and less-healthy food choices in American schools: a national study of grade, racial/ethnic, and socioeconomic differences. Am J Prev Med. 2007;33(suppl 4):S226-S239. 15. National Center for Education Statistics. US Department of Education. Institute of Education Sciences, National Center for Education Statistics, 2006-2007. Available at: http://nces.ed.gov/ccd/pubschuniv.asp. Accessed January 10, 2011. 16. Mersky RM, Dunn DJ. Fundamentals of Legal Research. 8th ed. New York, NY: Foundation Press; 2002.
May 2014, Vol. 84, No. 5
•
© 2014, American School Health Association • 315
17. Coalition for Food Security. Farm to school legislation: a State by State listing, 2008. Available at: http://www.food security.org/policy/StateByStateFarmtoSchoolLegislation.pdf. Accessed August 1, 2010. 18. National Conference of State Legislatures’ Healthy community design and access to healthy food legislation database. Available at: www.ncsl.org. Accessed September 5, 2010. 19. Kreuter F, Valliant R. A survey on survey statistics: what is done and can be done in Stata. Stata J. 2007;7:1-21.
316
•
Journal of School Health
•
May 2014, Vol. 84, No. 5 •
20. Kenny DA. Mediation with dichotomous outcomes. 2008. Available at: http://davidkenny.net/doc/dihnmed.doc. Accessed January 11, 2011. 21. Sobel ME. Asymptotic confidence intervals for indirect effects in structural equation models. Sociol Methodol. 1982;13: 290-312. 22. United States Department of Health and Human Services. Office of Disease and Health Promotion. Healthy People 2020. Available at: www.healthypeople.gov. Accessed May 25, 2013.
© 2014, American School Health Association
