Background
According to the 2008
Physical Activity Guidelines for Americans, children should engage in 60 min of moderate to vigorous physical activity (MVPA) each day and the Institute of Medicine (IOM)
1,
Educating the Student Body: Taking Physical Activity and Physical Education to School, advises that more than half of daily MVPA be obtained during the school day [
1,
2]. Research suggests that children who meet these requirements have a myriad of health benefits including stronger bones, better cardiovascular health, higher self-esteem, and a decreased risk of chronic diseases such as obesity, hypertension, and diabetes in adulthood [
2]. A systematic review by the Centers for Disease Control and Prevention (CDC) also suggests an association between school-time physical activity (PA) and improved academic performance, standardized test scores, and concentration; as well as better attention and classroom behavior [
3]. Despite these known benefits, recent evidence reveals that only 42% of children and 8% of adolescents are meeting daily requirements [
4]. As such, further research on strategies to increase activity levels and opportunities for children is critical to reduce physical inactivity and the associated health and economic burden.
The
Physical Activity Guidelines for Americans mid-course report posits schools as an optimal environment for PA intervention [
5]. School-based PA interventions have repeatedly been shown to increase activity levels in children and adolescents [
6‐
8] and present an ideal platform for intervention delivery given the time children spend in school and the potential to engage children from diverse socioeconomic backgrounds [
9]. The
Physical Activity Guidelines for Americans mid-course report demonstrates a broad range of school-based interventions as validated mechanisms to increase students’ PA levels, including classroom PA breaks, active transport to and from school, before and after school programs, and physical education (PE). While findings support the effectiveness of these standalone interventions, recent evidence suggests multi-component interventions that combine two or more programs are most successful and provide the greatest benefit to children’s PA levels, fitness, and cognitive ability [
5]. Given the multi-disciplinary nature of these interventions, however, additional research is needed to understand the leadership, funding, staffing, evaluation, and support required for implementation.
Despite the known benefits of providing PA opportunities for students, a majority of school districts are not meeting daily recommendations [
10]. In 2006, only 3.8% of elementary and 7.9% of middle schools met annual recommendations for PE minutes [
11]. In 2012, 58.9% of school districts required and 34.2% recommended that elementary students engage in some amount of daily recess; however, only 30.2% of these districts set targets at or above 30 min [
10]. These findings are cause for concern given recent recommendations by the IOM, and demonstrate the need to increase PA opportunities for students during the school day. In addition, while recent research highlights the potential for PA opportunities outside of recess or scheduled PE to increase activity, data reveal fewer than 15% of school districts required PA breaks for elementary and middle school students [
12]. Despite this apparent lack of PA opportunities in many school districts, some districts are successfully meeting recommendations and incorporating PA into the school day. However, little formal evaluation has been done with these district programs to identify common strategies for successful implementation. As such, research is needed to understand the programmatic efforts underway in districts meeting PA recommendations to inform strategies in districts that are underperforming.
In 2014, the Physical Activity in Youth Innovation Collaborative (now the Physical Activity and Health) (PA IC) was launched as an ad hoc activity associated with the Roundtable on Obesity Solutions [
13] (the Roundtable is an activity within the Health and Medicine Division [formerly the Institute of Medicine] of the National Academies of Sciences, Engineering, and Medicine [the National Academies]). This PA IC is currently comprised of 20 experts in physical activity, exercise and fitness, childhood obesity and health, school, and public policy. Members of the PA IC come from academia, government, non-profit organizations, associations, foundations, and industry. A primary objective of the PA IC was to use existing evidence from successful school- and community-based programs to influence public policies and to map longer-term strategies to increase PA in youth. Therefore, given these interests, this project was conceptualized by members of the PA IC, with the primary objectives to: 1) identify school districts across the country that demonstrated exemplary efforts to provide students with many PA opportunities, 2) understand the factors that facilitated their programmatic success, and 3) develop recommendations for school-based PA interventions and programs that can reach across district demographics and successfully increase PA among America’s youth. Given that this is a novel and under-researched area of study, we utilized a mixed methods approach, which provides more comprehensive data on the many factors that may influence districts ability to provide adequate amounts of PA in their schools. The common, yet unique, strategies identified by this research will fill an important knowledge gap and have the potential for widespread dissemination and impact on the PA and health of America’s youth.
Methods
This project was conducted by a collaboration between members of the PA IC, who designed the project, established the nomination criteria for identifying exemplary districts, helped with recruiting districts, and provided expert feedback and guidance, and researchers at Tufts University, who implemented the project and carried out the qualitative and quantitative analyses, with some assistance from collaborators at the CDC. Two of the study authors are members of the PA IC, one of whom is also affiliated with Tufts University.
Nomination criteria
All aspects of the nomination and selection process were carried out by members of the PA IC, an ad hoc activity associated with the Roundtable on Obesity Solutions at the National Academies of Sciences, Engineering, and Medicine. PA IC members identified a convenience sample of school districts in the United States that could best describe exemplary strategies for implementing and sustaining physical activity policies and programs to support students’ PA during the school day. Example exemplary efforts identified by PA IC members included district representative attendance at various physical activity conferences or conventions, partnership with external physical activity organizations, or application for federal funding. PA IC members also identified districts from published write-ups or publicly available data on ongoing physical activity efforts and activity levels among students within the district. The research team aimed to identify a convenience sample of districts that were diverse in geographic location and socioeconomic status (SES). For the nomination process, PA IC members submitted recommended district names to the PA IC Senior Program Officer based on their organizations’ involvement with and knowledge of the ongoing PA activity efforts in each of these districts. The PA IC Senior Program Officer created a master list and removed all duplicate districts. The PA IC discussed and revised the list by consensus, and approved the final nominations over in-person meetings, phone calls, and email. Based on these criteria, PA IC members identified 59 school districts in total.
Recruitment
Recruitment was conducted by study staff, with assistance from members of the PA IC, from June–October of 2015. Superintendents were contacted via email and asked to identify an individual(s) within the district who could best speak to the PE and PA programs and policies currently underway. Superintendents who did not respond to our initial recruitment emails were followed up with two additional times. When contact information was provided, we subsequently followed up with the individual(s) identified by the superintendent via email and/or phone and asked them to participate in a voluntary phone interview. We attempted to contact these potential interviewees up to three times. If we did not receive a response after the third phone call/email, we stopped recruitment for that district. The interviewees identified by district superintendents assumed various administrative roles within the district including, but not limited to, the: District Director for Health and PE; PE/Health Specialist; Coordinated School Health Administrator; PE Supervisor; Assistant Director of the Whole Student Initiative; and Superintendent.
Data collection
The interview script was developed by research study staff with input from PA IC members to gain an in-depth understanding of factors that may have contributed to the success of the district’s creation, implementation, and/or long-term maintenance of programmatic and policy efforts around PA, PE, and recess. The survey instrument included questions regarding the reach, initial development, and funding for all programs; the leaders engaged in the development, implementation, and/or continuation of PA programs; the sustainability of programmatic efforts; the implementation and evaluation of programs; equity in program reach; and policy impact. We piloted the initial version of the interview script in March and April 2015. These pilot interviews were conducted with three California districts that differed in size and geographic location to evaluate survey length and comprehensibility. We modified the initial instrument based on feedback from these pilot interviews (See Additional file
1: Final interview instrument).
Between June and October 2015, semi-structured interviews were conducted, via phone, with the final group of recruited districts. For each interview, at least two members of the research study staff were present and took notes on the interviewees responses, as the interviews were not audio recorded. When possible, district characteristics including number of schools, student demographics, and student-to-teacher ratio were collected from the district website to augment the time devoted to non-publicized information. Most interviews spanned 45–60 min depending on the knowledge of the interviewee, the detail provided, and the availability of district data. All protocols and study materials were reviewed by the Tufts University Social, Behavioral, and Educational Research Institutional Review Board. Written consent was not obtained, as data obtained for this study was regarding the school districts and not the individuals interviewed.
Immediately following each interview, research staff reconciled their notes and discussed any inconsistencies and/or points that required further clarification. Wherever necessary, contacts were provided with follow-up questions and clarifications were incorporated into the final dataset. All but one district correspondent responded to these requests for clarification. When clarification was not obtained, this was noted in the dataset. Data were entered into Microsoft Excel and coded. Quantitative data were double entered by two researchers and compared using SAS 9.3 (Cary, NC). Qualitative data were subsequently reviewed for quality by study staff.
Measures
Demographics
District student demographic data on race/ethnicity and SES were recoded based on whether the majority of students (> 50%) were non-white, Hispanic, and/or eligible for free and reduced price lunch (FRPL; y/n). The average number of students served in PRE-K-12th grade, average number of schools, and average student-to-teacher ratio in both PE and academic classes were also calculated and reported. For the purpose of these analyses, district size was determined using the National Center for Education Statistics classification system [
14] and modified based on our sample: districts with 1–5 schools were considered small, 6–19 schools were considered midsize, and 20+ schools were considered large. District SES was identified based on the percent of students that qualified for FRPL; districts with 76–100% of students eligible for FRPL were classified as low SES, 26–75% students eligible for FRPL were classified as medium SES, and 0–25% of students eligible for FRPL were classified as high SES [
15,
16].
Physical activity
Data on minutes of PA/PE provided were recategorized based on whether the district offered at or above IOM recommendations for minutes of school-time PA/PE (150 min/wk. for elementary students and 225 min/wk. for middle/high school students) [
1] and/or at or above 20 min/wk. for recess. Minutes of school-based PA could only be assessed based on estimates for PE, since respondents were asked to report on efforts at the district level and few districts had requirements for PA outside of PE classes. District offerings of programs during and after school (school-day PA programs, recess, and after-school PA programs) and district requirements for minutes of recess and/or PE were recorded. The number of different school day PA programs offered (i.e. GoNoodle© and PlayWorks©) were also coded and recorded. Whether districts required PE teacher professional development, hired paid versus volunteer staff for PA efforts, and whether hired staff oversaw PA efforts were calculated and recorded. When available, nationally representative district level data obtained from the CDC’s School Health Policies and Practices Study (SHPPS) were used as a national comparison [
10]. The most recent district-level SHPPS data were from 2012. We also evaluated whether there were differences in noted PE requirements reported by interview respondents and archived versions of district websites, wherever available (identified on WayBack Machine:
https://archive.org/).
Funding
Funding sources utilized by each district were recategorized based on whether they were district/school based, external grants, and/or funding from community organizations or individuals. Internal funding sources were further characterized based on whether they were part of the school department or school level budget, part of the PA budget, part of the district budget, internal grants, part of the state budget, Title I funding or Medicaid funding. We also noted when the districts reported that they required no funding for their PA efforts.
Champions
The number and type of champions, defined as an individual or group of individuals who are instrumental in the development, implementation and/or continuation of the PA program(s) in the district, were identified from the following sectors: PE teachers, Principals, Families/Parents, Board of Education members, Parent Teacher Organizations, Superintendents, Community Members/Organizations, Director of Health and PA, Mayor’s Office/County Supervisor/Town Manager, or other.
Data analysis
Descriptive statistics of quantitative data were conducted in Microsoft Excel and statistical analyses of quantitative data were conducted in RStudio (Version 1.0.153). All qualitative data analyses were carried out using NVivo Version 10. A preliminary codebook was developed based on an initial review of the data and the a priori codes inherent to the interview script. An inductive thematic approach was used for all qualitative analyses to identify explicit and implicit themes within the data [
17]. Six districts that represented the demographic diversity of the total interview sample were randomly selected and independently coded by two study staff. After independently coding the first two interviews, staff met to discuss new themes, any necessary adjustments in the existing codes, and any discrepancies in coding. This process was repeated in NVivo until the two researchers established inter-rater reliability (≥ 80% agreement) on all codes. Emergent themes were identified by study staff, discussed with members of the PA IC, and further developed based on feedback from PA IC members. Ultimately, three overarching factors were identified as integral to providing many PA opportunities to students: having champions, funding and support, and bi-directional partnerships with the community and external PA programs. Themes were evaluated overall and by district SES; results are presented below wherever appropriate. We were not able to discuss differences in our findings based on geographic location, as there was only one rural district and at least three participants are required in each group to evaluate qualitative differences [
18]. Quotes presented below were modified to de-identify participating districts throughout.
Discussion
Despite variability between the sample districts based on district size, the availability of resources, and SES, the results of this project demonstrate that school districts can provide substantive PA opportunities to students to increase the likelihood that 30 min of PA is acquired during the school day. We identified three overarching and broadly relevant themes that were associated with these districts’ successful efforts to enhance PA opportunities for students: champions, funding and tangible support, and bi-directional partnerships. Not only were the three themes key correlates of the development of PA opportunities for students, but each theme also remained important for the implementation, evaluation, and sustainability of programs. Furthermore, each theme supported the development of a culture of health within the schools and districts, which was frequently cited as a contributing factor to the success of enhancing PA opportunities for students.
A culture of health, which can be broadly defined as an environment that supports good health and well-being, was often associated with having a superintendent or upper level leadership member who identified PA for students as a district priority. In our sample, respondents who identified the superintendent as a champion consistently emphasized the importance of this upper level leadership in the success of their programs. Many of these respondents also mentioned that administrations who were supportive of PA often developed a funded position for a lead champion for PA programmatic efforts; and these lead champions were frequently involved in acquiring funding and tangible support for district programs, as well as the developing and maintaining bi-directional partnerships. As such, the synergistic nature of the three identified themes may be enhanced by district leadership that prioritizes PA for students and supports a lead PA champion.
In 2014, the CDC and SHAPE America released a guide for developing a Comprehensive School Physical Activity Program (CSPAP), which is a multi-component model for school districts and schools to enhance opportunities for students to be physically active. The model suggests five important components that contribute to PA opportunities for students: PE, PA before and after school, PA during school, staff involvement, and family and community engagement [
11]. While there is a lack of data to suggest whether schools and districts meeting PA recommendations are following the CSPAP guide, the efforts identified by respondents in this sample closely align with the CSPAP, with the majority of districts (74%) providing PA during school, before or after school, and PE, as well as noting staff involvement and champions at the family and community levels. The inclusion of champions at the staff, family and community levels, as recommended by the CSPAP model, supports the development and sustainability of bi-directional partnerships and funding and tangible support.
While there were limited differences in results across district SES, suggesting the potential for widespread and meaningful impact, this project was not met without limitations. First, the study design did not allow us to draw any conclusions on causality or to directly attribute the success of the included to any one of the identified factors. We also cannot determine whether the PA programs identified here were responsible for the success of these districts, if it is a combination of the identified factors, or if it is the result of unmeasured factors. However, given the limited prior work in this space, we aimed to generate hypotheses about how these factors interact to facilitate successful PA programming. Future controlled studies should evaluate the combined impacts of the factors uncovered here on student PA levels. Second, the school districts interviewed were identified by way of convenience sampling and agreed upon by group consensus through the nomination process. Some members of the PA IC were involved with youth PA funding agencies, which may have led to the identification of districts that have received funding, introducing the potential for bias among our results. Given the response rate, it is also possible that those districts that decided to participate were among those providing the most physical activity opportunities for their students, compared to those that did not respond. Third, interview data were not recorded and thus not transcribed verbatim. While this could have introduced the potential for missing data, having two research staff present during each interview mitigated this possibility. In addition, our sample did not include underperforming districts as a comparison group; however, we were able to compare our sample against the SHPPS data for national comparison. Although these data were collected three years prior to the start of this project, comparison of reports collected in 2006 [
12], 2012 [
10], and 2016 [
20] show a high degree of similarity at the district-level suggesting that these estimates may not have changed substantially over time (See Additional file
2: Table S2).
We also acknowledge the possibility of reporting bias with regard to weekly physical activity levels. It’s possible that social desirability biases lead respondents to over-report weekly minutes of physical activity, given their awareness of the purposes and underlying goals of this study. Additionally, given the variability in how districts reported their requirements, it’s also possible the varied timeframes (e.g., semester requirements versus requirements spanning the full school-year) might have led to over-reporting since we did not adjust for time in our analyses. Yet, in this sample of school districts, variability in PE requirements was reported almost exclusively at the high school level, and would likely not have impacted estimates at the lower grade levels. Moreover, in reviewing PE requirements listed on archived versions of district websites, we found that the respondents’ estimates were consistent with those listed on the district site, and many of these districts offered enough PE to meet these requirements. Alternatively, while interviewees were identified by superintendents as individuals most knowledgeable of PA programmatic efforts, district level contacts may not have been fully aware of all efforts taking place at the school level, which may have led to underreported PA efforts. We also have evidence that students likely engaged in additional physical activities outside of district-PE requirements, which would not have been captured in our analyses and would contribute an additional source of under-reporting. Despite identified limitations, this project fills an important research gap, provides a foundation for future research, and presents evidence to inform school district’s efforts to enhance PA opportunities to students nationwide.
Acknowledgements
The authors would like to thank Heather Cook from the Health and Medicine Division (formerly the Institute of Medicine), National Academies of Sciences, Engineering, and Medicine and the members of the Health and Medicine Division’s Roundtable on Obesity Solutions’ Physical Activity and Health Innovation Collaborative (PA IC): Howell Wechsler Alliance for a Healthier Generation, Sharon Adams-Taylor The School Administrators Association (AASA), Jim Whitehead American College of Sports Medicine, Anthony Wall and Todd Galati American Council on Exercise, Laurie Whitsel American Heart Association, Sarah Lee Centers for Disease Control and Prevention, Heidi Burke Greater Rochester Health Foundation, Catherine Kwik-Uribe Mars, Inc., Charlene Burgeson Partnership for a Healthier America, Kathleen Tullie and Ewunike Akpan BOKS (Build Our Kids’ Success) and Reebok International, Mary Lou Goeke United Way, Jim Sallis University of California-San Diego, John Jakicic University of Pittsburgh, and University of Texas Health Science Center–Houston School of Public Health and University of Texas at Austin, Harold W. (Bill) Kohl, III. We thank William Dietz The George Washington University for careful review of the manuscript during development.