8: No. 6, November 2011
Myde Boles, PhD, MBA; Julia A. Dilley, PhD, MES; Clyde Dent, PhD; Miriam R. Elman, MPH; Susan C. Duncan, PhD; Donna B. Johnson, PhD
Suggested citation for this article: Boles M, Dilley JA, Dent C, Elman MR, Duncan SC, Johnson DB. Changes in local school policies and practices in Washington State after an unfunded physical activity and nutrition mandate. Prev Chronic Dis 2011;8(6):A129.
http://www.cdc.gov/pcd/issues/2011/nov/10_0191.htm. Accessed [date].
Policies and practices in schools may create environments that encourage and reinforce healthy behaviors and are thus a means for stemming the rising rates of childhood obesity.
We assessed the effect of a 2005 statewide school physical activity and nutrition mandate on policies and practices in middle and high schools in Washington State.
We used 2002, 2004, and 2006 statewide School Health Profiles survey data from Washington, with Oregon as a comparison group, to create longitudinal linear regression models to describe changes in relevant school policies after the Washington statewide mandate. Policy area composite measures were generated by principal component factor analysis from
survey questions about multiple binary measure policy and practice.
Relative to expected trends without the mandate, we found significant percentage-point increases in various policies, including restricted access to competitive foods in middle and high schools (increased by 18.8-20.0 percentage points);
school food practices (increased by 10.4 percentage points in middle schools); and eliminating exemptions from physical education (PE) for
sports (16.6 percentage-point increase for middle schools), exemptions
from PE for community activities (12.8 and 14.4 percentage-point increases for
middle and high schools, respectively) and exemptions from PE for academics (18.1 percentage-point increase for middle schools).
Our results suggest that a statewide mandate had a modest effect on increasing physical activity and nutrition policies and practices in schools. Government policy is potentially an effective tool for addressing the childhood obesity epidemic through improvements in school physical activity and nutrition environments.
Back to top
Since 1980, the prevalence of obesity among children and adolescents in the United States has tripled (1). From 1980 to 2008, the prevalence of obesity among children aged 6 to 11 years increased from 6.5% to 19.6%, and among adolescents aged 12 to 19 years obesity increased from 5.0% to 18.1% (2). Policies and practices in schools may create environments that encourage and reinforce healthy eating and regular physical activity and thus are promising means to stem the rising rates of childhood
To address the youth obesity problem, state and federal authorities have adopted obesity-prevention strategies such as legal mandates to improve school physical activity and nutrition policies. The federal Child Nutrition and WIC Reauthorization Act of 2004 mandated that all
US school districts participating in the federally reimbursed school meal programs develop a local school wellness policy by the beginning of the 2006-2007 school year (5). Washington’s legislature adopted
Washington Senate Bill 5436, which was similar to the federal mandate and required each of Washington’s 296 school districts to establish a nutrition and physical fitness policy by August 1, 2005,
one year before the federal deadline. No funding was authorized for the implementation of SB 5436.
To assist school districts, the Washington State School Directors’ Association developed a model policy regarding access to nutritious foods, opportunities for exercise, and classroom instruction related to nutrition and physical activity (6). The model policy required that middle school students have an average of 100 minutes per week
(20 minutes per day) of aerobic education activity. High school students were required to complete
2 credits of health and fitness. It further recommended
that districts adopt policies to hire certified physical education teachers; provide
after-hours access to school facilities for physical activity, fitness, sports, and recreation programs; and identify safe routes to school for walking and biking.
For nutrition, the model policy required that school breakfasts and lunches meet the nutritional standards of state and federal school breakfast and lunch programs (7). In addition, the model policy contained provisions for the availability of fresh fruit and safe drinking water, use of nonfood alternatives for rewards, competitive pricing for healthy food options, adequate time and places to eat lunch, and a nutrition education curriculum focused on knowledge, skills, and assessment of
personal eating habits.
Limited research has examined the relationship between state and federal legislation related to physical activity, nutrition, and school wellness and local physical activity, nutrition, and wellness policies. Research on the
effect of the federal wellness policy mandate on local wellness policies in a nationally representative sample of schools in the 2006-2007 or 2007-2008 school years found that policies were weak overall and varied greatly from district to district (8). Among the few
studies that specifically evaluated changes in local policies before and after the federal wellness mandate, 1 found that overall time available for physical activity did not change in a random sample of 45 rural elementary schools in Colorado after the policy went into effect (9). Another study of 847 medium and larger schools in the United States found that nutrition components increased significantly and were reported as the components most frequently implemented (10).
Previous studies examining the effect of a policy mandate on school policies have been limited by lack of both a comparison group and strong longitudinal data on policy trends. The primary objective of our study was to assess the
effect of Washington’s statewide mandate on physical activity and nutrition (PAN) policies and practices in schools relative to historical trends in Washington and to compare these with those in Oregon, a geographically and demographically similar state
without a statewide PAN mandate. We hypothesized that an unfunded policy mandate may not lead to successful policy adoption and
that its success depends on both components of PAN policies and characteristics of the school districts. Our secondary objective was to investigate whether the law’s
effect was associated with geographic area (urban or rural), school-level socioeconomic status
(SES), or the average standardized test scores of schools. Washington and Oregon’s demographic similarities make them ideal
for examining how statewide laws lead to local policy changes.
Back to top
We used public health surveillance data to conduct a secondary data analysis. We performed hierarchical longitudinal linear regression with schools nested in time to test whether the proportion of schools in Washington with PAN policies and practices changed after the implementation of the statewide PAN mandate
compared with schools in Oregon where there was no mandate.
Data and sample
The School Health Profiles survey (Profiles) is conducted biennially by the Centers for Disease Control and Prevention (CDC) in collaboration with state and local education and health agencies. It is a self-administered survey of public secondary school principals and lead health education teachers and is designed to assess school health programs, policies, and activities (11). CDC uses a random, systematic, equal-probability sampling strategy to produce representative samples of public
schools that serve students in grades 6 through 12. In 2004 and 2006, Washington modified this sampling procedure and invited all secondary schools, rather than just a sample, to participate. We obtained identical 2002, 2004, and 2006 data from principal and health educator surveys from the Oregon Public Health Division and the Washington State Department of Health. Study schools had an enrollment of at least 15 students per grade, had a
standardized health and physical education curriculum, and were not alternative schools or combined middle/high schools (Table 1).
We used the 30 nutrition-related questions and 26 physical activity-related questions in Profiles. Most item responses in Profiles were binary, indicating the presence or absence of a particular policy, practice, activity, or attribute at the school. We recoded items so that positive responses always indicated a better or desired condition, such as presence of a policy.
To create a manageable set of PAN outcome measures we performed a preliminary analysis of principal component factors to identify subsets of associated items within physical activity-related and nutrition-related domains. We began by grouping items in conceptual domains based on item content (eg, nutrition policy, nutrition curriculum). We then extracted the empirical principal components
and used structural equation modeling to create item groupings (12).
In the nutrition domain, the factor analysis procedure confirmed 7
independent sets of items — 3 factors relating to classroom educational content
and 4 factors relating to nutritional policy. In the physical activity domain,
we found 10 independent factors — 3 educational content factors and 7
policy-related factors (Appendix).
We combined items within a factor into composite outcome measures by summing. Measures with larger scores reflected a greater number of positive responses to component variables. Once combined, we recoded these measures to a percentage scale (ie, 0-100).
We used indicator variables for year, state, and introduction of the law (1 = presence of policy in Washington schools in year 2006, 0 = otherwise).
We considered schools with students in grades 6 to 8 as middle schools and schools with students in grades 9
through 12 as high schools. We examined middle and high schools separately because Washington’s model policy gave different recommendations for each and because of other differences such as availability of competitive foods.
As school descriptors, we used publicly available school-level data from the
2 states’ departments of education as covariates in the models. These data included SES (measured as the average percentage of students eligible for the free and reduced-price lunch program during study years; schools with more than one-third of students enrolled in the lunch program were considered low-SES schools); geographic area (urban and not urban using Rural Area Commuting Area codes;
schools located in urban and suburban areas were classified as urban; schools located in small towns and frontier areas were classified as
not urban) (13); and average percentage of students meeting state standards on statewide achievement tests (high performing schools were those with more than half of students meeting state standards).
To determine whether the Washington State mandate had an effect on the presence of PAN policies and practices in 2006, we created hierarchical longitudinal linear regression models with schools nested
in time by using the GLMMIX procedure in SAS version 9.2 (SAS Institute, Inc, Cary, North Carolina). In the base model, we tested whether the proportion of schools in Washington reported having changed each of the 17 PAN policy and practice composite outcome measures from 2002 and 2004 (before the statewide mandate) to 2006 (after the statewide mandate). We used Oregon as an implicit control to strengthen these estimates. We included indicator variables “state,” “year,” and “law” in models for each PAN outcome and stratified by school type (middle school vs high school):
Base Model: policyjst = β0 + β1states +
β2yeart + β3lawst +
with j = 1 to number of schools;
s = 0 (Oregon), 1 (Washington); and
t = 0 (2002), 2 (2004), 4 (2006).
The coefficient β0 indicated the baseline (2002) percentage value of policy in Oregon; the coefficient
β1 for the variable “state” measured the baseline difference between Washington and Oregon; the coefficient
β2 for the variable “year” assessed secular annual trends across the study years; the coefficient
β3 for the variable “law” determined the deviation from the expected trend in Washington in 2006; and
indicated random error. We stratified the base models by grade and tested the model coefficients for consistency across school type using contrast statements in SAS.
To determine whether the Washington State mandate had similar effects across various school subgroups, we added terms to adjust for the covariates SES, geographic area, and school test score and incorporated terms for interactions of these variables with the enactment of the law.
We used a logistic link function in all models to test for statistical
significance. Model coefficient estimates are presented in percentage-point
scale values in our discussion and in Tables
3 through 5 in this article. We considered P values of less than or equal to .05 (2-sided test) as significant and estimates of 5 percentage points or more as substantively meaningful.
The Washington State and Oregon Public Health institutional review boards declared this study to be exempt under 45 CFR 46.101(b)(4).
Back to top
Approximately half of all Washington State schools had high SES or high
test scores (Table 2) At
the 2002 baseline, only 31.6% to 41.6% of middle schools and 14.9% to 23.1%
of high schools had policies in place that restricted access to competitive foods (what types of foods or times of day for access),
and only 9.0% of middle schools and 17.7% of high schools had favorable school
food practices (adequate time for lunch and availability of fruits and
vegetables at school events) (Table 3).
In the absence of the Washington PAN mandate, the expected trends (based on 2002-2006 Oregon data and 2002-2004 Washington data) for the percentage of Washington schools with nutrition policies and practices generally did not change; however, the percentage of schools with restricted access to competitive foods (ie,
what foods and when accessible) increased significantly, and healthy food options
(ie, low-fat snacks, fruits, and vegetables) decreased significantly. The expected trends in physical activity policies and practices
did not change significantly, except for a decline in the percentage of schools requiring certification for
middle school physical education (PE) teachers.
Both middle schools and high schools showed a significant (18.8-20.0 percentage-point) increase in the number of schools with restricted access to
type of competitive foods (Table 4).
For restricted access to competitive foods (time of day), high schools increased
by 19.2 percentage points, which is significantly higher than in middle schools where the increase was
not statistically significant. Unexpectedly, healthy food options for middle and high schools
declined significantly, by 5.9 and 2.0 percentage points, respectively. Middle schools showed a
significant (10.4 percentage-point) increase in school food
practices (ie, adequate time for lunch and availability of fruits and vegetables at school events). There was no
significant increase for high schools.
There was a significant increase in the percentage of middle schools that did not allow exemptions from PE for sports, community activities, or academics, and a significant increase in high schools that did not allow such exemptions for community activities
(Table 5). Profiles found no other significant increases for other physical activity policies and practices.
When we examined the interactions of each of the 17 policy measures with urban/not urban, high/low SES, and high/low academic performance (middle and high schools combined), we found only
3 significant interactions: restricted access to competitive foods (type of
food), which was present in 14.0% more higher-performing than lower-performing
schools; restricted access to competitive foods (time of day), which was present
in 11.3% fewer urban vs not-urban schools; and facilitators for PA (ie, safe
routes to schools, community programs), which were available in 0.3% low- vs high-SES schools.
Back to top
We found significant increases in the percentage of middle and high schools reporting the presence of certain PAN policies and practices after the implementation of the Washington State PAN mandate.
This is the first published study to present a longitudinal analysis of changes
in PAN policies and practices in 1 state by using measures of PAN policies and
practices in another state for comparison to control for secular trend.
In 2002, few schools in Washington State had restricted access to competitive foods or nutrition-related policies in place, although
most schools did have healthy food options available. It is not surprising that restricted access to competitive foods and
school food practices were areas for growth, even in absence of the Washington mandate. We were surprised, however, to find a substantial decline in the percentage of middle and high schools offering healthy food options. Because
the Profiles questions about healthy food options focused on the availability of healthy foods in vending machines and school stores,
these schools may have been eliminating these venues for food purchases rather than reducing the availability of healthier food types
in vending machines or school stores. Another explanation for the decline may be changing perceptions of school principals about what constitutes a “healthy” option.
Because so many Washington State schools already had physical activity policies and practices in place
before the statewide mandate, opportunity for growth in this area was limited. The only area that changed after the implementation of the law was an increase in the number of schools that were not allowing student exemptions from PE because of participation in school sports and other school or community activities. Rather than trying to fund new provisions and programs for physical
activity, the elimination of exemptions from PE may have been a budget-neutral way for schools and districts to respond to the mandate.
Our study’s findings of a significant increase in nutrition policies and practices and only small improvement in physical activity policies are consistent with a recent study of trends in state-level school nutrition and physical activity policy environments (14). That study found that schools adopted more food service and nutrition policies than physical activity, education, or weight assessment policies. Similarly, a study of the effects of federal wellness legislation in school
districts throughout the United States found that nutrition components were the most frequently implemented (10). However, even in the presence of nutrition policies, improvements in school food environments are modest (15), and foods of minimal nutritional value remain available (16). In contrast to most nutrition policies, many physical activity policies have a direct
effect on instruction time (eg, requiring more PE classes or longer classes), and this may be a barrier
to school districts adopting such changes.
We saw very few differences in change associated with various school-level factors, such as urban/not urban setting, SES, and academic performance. This suggests that the effect of a policy mandate was similar among different school types.
Washington State did not allocate funding to support schools in implementing their PAN policies, nor was there meaningful quality assurance of adopted policies or clear punitive measures in place for school districts that failed to effectively implement a policy. Addition of any of these supportive measures might have resulted in different — perhaps greater — policy and practice improvements.
This study has several limitations. First, we did not see appreciable increases in physical activity policies and practices associated with the Washington State mandate.
This may indicate that Profiles did not ask about the features of the model policy that were most emphasized and acted on by local school districts.
A substantial portion of the language in the model policy described the amount of instructional time for physical education,
which was not asked about in the Profiles questionnaire. Second, we examined the changes in school policies and practices only
1 year after the Washington statewide PAN mandate went into effect. Arguably, 1 year is too short a time for schools to mobilize their efforts for PAN-related policy changes. However, we did see changes in restricted access to competitive foods, nutrition policy, and reduced exemptions from PE relative to trend, which
could confirm that PAN effected changes. Conversely, a federal wellness policy
requirement similar to the Washington policy requirement was scheduled for
implementation in both Washington and Oregon in the fall of 2006, and the Profiles
survey was administered in spring of that year. Thus, some Oregon schools may
have already responded to the federal requirement. If this was the case, our analysis of the Washington trends may have underestimated the real effect of the
Washington mandate. Finally, this study describes the policy environment in the Pacific Northwest, which may differ in other
Future studies should examine the relationship between state and federal laws and the quality of PAN policy implementation in schools and the association of PAN policy with youth PAN outcomes.
Our results suggest that a statewide mandate had a modest positive effect on PAN policies and practices in schools. Government policy is potentially an effective tool for addressing the childhood obesity epidemic through improvements in PAN environments in schools.
Back to top
This study was supported by a grant (no. 5R03CA129964) from the National Cancer Institute, National Institutes of Health. We
thank Susan Richardson, MPH, for her assistance with obtaining Washington State surveillance data and performing the crosswalk between Washington and Oregon for 2002, 2004, and 2006 data.
Back to top
Corresponding Author: Myde Boles, PhD, MBA, Program Design and Evaluation Services, Multnomah County Health Department and Oregon Public Health Division, 827 NE Oregon St, Ste 250, Portland, OR 97232. Telephone: 971-673-0595. E-mail:
Author Affiliations: Julia A. Dilley, Clyde Dent, Multnomah County Health Department and Oregon Public Health Division, Portland, Oregon; Miriam R. Elman, Oregon
State University and Oregon Health and Science University, Portland, Oregon; Susan C. Duncan, Oregon Research Institute, Eugene, Oregon; Donna B. Johnson, University of Washington, Seattle, Washington.
Back to top
- Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM.
Prevalence of overweight and obesity in the United States, 1999-2004. JAMA 2006;295(13):1549-55.
- Ogden CL, Carroll MD, Curtin LR, Lamb MM, Flegal KM.
Prevalence of high body
mass index in US children and adolescents, 2007-2008. JAMA 2010;303(3):242-9.
- Centers for Disease Control and Prevention.
Guidelines for school health programs to promote lifelong healthy eating. MMWR
Recomm Rep 1996;45(RR-9):1-41.
- Centers for Disease Control and Prevention.
Guidelines for school and community programs to promote lifelong physical activity among young people. MMWR
Recomm Rep 1997;46(RR-6):1-36.
- Pub L No. 108–265. http://www.fns.usda.gov/cnd/Governance/Legislation/Historical/PL_108-265.pdf. Accessed May 18, 2010.
- Superintendent of Public Instruction. Model policy on nutrition and physical fitness; 2005. http://www.k12.wa.us/BulletinsMemos/bulletins2005/B004-05.doc. Accessed May 18, 2010.
- US Department of Health and Human Services and US Department of Agriculture. Dietary
guidelines for Americans, 2005. 6th edition. Washington (DC): US Government Printing Office; 2005.
- Chriqui JF, Schneider L, Chaloupka FJ, Ide K, Pugach O. Local wellness policies: assessing school district strategies for improving children’s health. School years 2006-07 and 2007-08. Chicago
(IL): Bridging the Gap, University of Illinois at Chicago, Health Policy
Center, Institute for Health Research and Policy; 2009.
- Belansky ES, Cutforth N, Delong E, Ross C, Scarbro S, Gilbert L, et al.
Early impact of the federally mandated local wellness policy on physical activity in rural, low-income elementary schools in Colorado. J Public Health Policy 2009;30(Suppl 1):S141-60.
- Longley CH, Sneed J. Effects of federal legislation on wellness policy formation in school districts in the United States. J Am Diet Assoc 2009;109(1):95-101.
- Centers for Disease Control and Prevention. Data and Statistics School Health Profiles.
http://www.cdc.gov/HealthyYouth/profiles/. Accessed May 18, 2010.
- Bartholomew DJ, Knott M. Latent variable models and factor analysis: Kendall’s Library of Statistics 7.
2nd edition. London (GB): Edward Arnold, 1999.
- US Department of Agriculture, Economic Research Service. Rural-urban commuting
area codes. http://www.ers.usda.gov/Data/RuralUrbanCommutingAreaCodes/. Accessed February 14, 2011.
- Nanney MS, Nelson T, Wall M, Haddad T, Kubik M, Laska MN, et al.
State school nutrition and physical activity policy environments and youth obesity. Am J Prev Med 2010;38(1):9-16.
- Woodward-Lopez G, Gosliner W, Samuels SE, Craypo L, Crawford PB.
Lessons learned from evaluations of California’s statewide school nutrition standards. Am J Public Health 2010;100(11):2137-45.
- Whatley Blum JE, Beaudoin CM, O’Brien LM, Polacsek M, Harris DE, O’Rourke KA. Impact of Maine’s statewide nutrition policy on high school food environments. Prev Chronic Dis
http://www.cdc.gov/pcd/issues/2011/jan/09_0241.htm. Accessed February 14, 2011.
Back to top