The Association Between School-Based Physical Activity, Including Physical Education, and Academic Performance

OBJECTIVE
The purpose of this review is to synthesize the scientific literature that has examined the association between school-based physical activity (including physical education) and academic performance (including indicators of cognitive skills and attitudes, academic behaviors, and academic achievement).


METHOD
Relevant research was identified through a search of nine electronic databases using both physical activity and academic-related search terms. Forty-three articles (reporting a total of 50 unique studies) met the inclusion criteria and were read, abstracted, and coded for this synthesis. Findings of the 50 studies were then summarized.


RESULTS
Across all the studies, there were a total of 251 associations between physical activity and academic performance, representing measures of academic achievement, academic behavior, and cognitive skills and attitudes. Slightly more than half (50.5%) of all associations examined were positive, 48% were not significant, and 1.5% were negative. Examination of the findings by each physical activity context provides insights regarding specific relationships.


CONCLUSION
Results suggest physical activity is either positively related to academic performance or that there is not a demonstrated relationship between physical activity and academic performance. Results have important implications for both policy and schools.


EXECUTIVE SUMMARY
When children and adolescents participate in the recommended level of physical activity-at least 60 minutes daily-multiple health benefits accrue. Most youth, however, do not engage in recommended levels of physical activity. Schools provide a unique venue for youth to meet the activity recommendations, as they serve nearly 56 million youth. At the same time, schools face increasing challenges in allocating time for physical education and physical activity during the school day.
There is a growing body of research focused on the association between school-based physical activity, including physical education, and academic performance among school-aged youth. To better understand these connections, this review includes studies from a range of physical activity contexts, including school-based physical education, recess, classroom-based physical activity (outside of physical education and recess), and extracurricular physical activity. The purpose of this report is to synthesize the scientific literature that has examined the association between school-based physical activity, including physical education, and academic performance, including indicators of cognitive skills and attitudes, academic behaviors, and academic achievement.

Methods
For this review, relevant research articles and reports were identified through a search of nine electronic databases, using both physical activity and academicrelated search terms. The search yielded a total of 406 articles that were examined to determine their match with the inclusion criteria. Forty-three articles (reporting a total of 50 unique studies) met the inclusion criteria and were read, abstracted, and coded for this synthesis.
Coded data from the articles were used to categorize and organize studies first by their physical activity context (i.e., physical education, recess, classroom-based physical activity, and extracurricular physical activities), and then by type of academic performance outcome. Academic performance outcomes were grouped into three categories: 1) academic achievement (e.g., grades, test scores); 2) academic behavior (e.g., ontask behavior, attendance); and 3) cognitive skills and attitudes (e.g., attention/concentration, memory, mood). Findings of the 43 articles that explored the relationship between indicators of physical activity and academic performance were then summarized.

Results
Across all 50 studies (reported in 43 articles), there were a total of 251 associations between physical activity and academic performance, representing measures of academic achievement, academic behavior, and cognitive skills and attitudes. Measures of cognitive skills and attitudes were used most frequently (112 of the 251 associations tested). Of all the associations examined, slightly more than half (50.5%) were positive, 48% were not significant, and only 1.5% were negative. Examination of the findings by each physical activity context provided insights regarding specific relationships.

1) School-Based Physical Education Studies
School-based physical education as a context category encompassed 14 studies (reported in 14 articles) that examined physical education courses or physical activity conducted in physical education class. Typically, these studies examined the impact of increasing the amount of time students spent in physical education class or manipulating the activities during physical education class. Overall, increased time in physical education appears to have a positive relationship or no relationship with academic achievement. Increased time in physical education does not appear to have a negative relationship with academic achievement. Eleven of the 14 studies found one or more positive associations between school-based physical education and indicators of academic performance; the remaining three studies found no significant associations.

2) Recess Studies
Eight recess studies (reported in six articles) explored the relationship between academic performance and recess during the school day in elementary schools. Six studies tested an intervention to examine how recess impacts indicators of academic performance; the other two studies explored the relationships between recess and school adjustment or classroom behavior. Time spent in recess appears to have a positive relationship with, or no relationship with, children's attention, concentration, and/or on-task classroom behavior. All eight studies found one or more positive associations between recess and indicators of cognitive skills, attitudes, and academic behavior; none of the studies found negative associations.

3) Classroom Physical Activity Studies
Nine studies (reported in nine articles) explored physical activity that occurred in classrooms apart from physical education classes and recess. In general, these studies explored short physical activity breaks (5-20 minutes) or ways to introduce physical activity into learning activities that were either designed to promote learning through physical activity or provide students with a pure physical activity break. These studies examined how the introduction of brief physical activities in a classroom setting affected cognitive skills (aptitude, attention, memory) and attitudes (mood); academic behaviors (on-task behavior, concentration); and academic achievement (standardized test scores, reading literacy scores, or math fluency scores). Eight of the nine studies found positive associations between classroom-based physical activity and indicators of cognitive skills and attitudes, academic behavior, and academic achievement; none of the studies found negative associations.

4) Extracurricular Physical Activity Studies
Nineteen studies (reported in 14 articles) focused specifically on the relationship between academic performance and activities organized through school that occur outside of the regular school day. These activities included participation in school sports (interscholastic sports and other team or individual sports) as well as other after-school physical activity programs. All 19 studies examining the relationships between participation in extracurricular physical activities and academic performance found one or more positive associations.

Strengths and Limitations
This review has a number of strengths. It involved a systematic process for locating, reviewing, and coding the studies. Studies were obtained using an extensive array of search terms and international databases and were reviewed by multiple trained coders. The studies cover a broad array of contexts in which youth participate in school-based physical activities and span a period of 23 years. Furthermore, a majority (64%) of studies included in the review were intervention studies, and a majority (76%) were longitudinal.
The breadth of the review, however, is a limitation. All studies meeting the established review criteria were included and treated equally, regardless of the study characteristics (e.g., design, sample size). The studies were not ranked, weighted, or grouped according to their strengths and limitations. The breadth of the review, while revealing a variety of study designs, measures, and populations, often made comparisons and summaries difficult. As a result, conclusions are intentionally broad.

Implications for Schools
The results of this review support several strategies that schools can use to help students meet national physical activity recommendations without detracting from academic performance: • School-based physical education: To maximize the potential benefits of student participation in physical education class, schools and physical education teachers can consider increasing the amount of time students spend in physical education or adding components to increase the quality of physical education class. Articles in the review examined increased physical education time (achieved by increasing the number of days physical education was provided each week or lengthening class time) and/or improved quality of physical education (achieved through strategies such as using trained instructors and increasing the amount of active time during physical education class).
• Recess: School boards, superintendents, principals, and teachers can feel confident that providing recess to students on a regular basis may benefit academic behaviors, while also facilitating social development and contributing to overall physical activity and its associated health benefits. There was no evidence that time spent in recess had a negative association with cognitive skills, attitudes, or academic behavior.
• Classroom-based physical activity: Classroom teachers can incorporate movement activities and physical activity breaks into the classroom setting that may improve student performance and the classroom environment. Most interventions reviewed here used short breaks (5-20 minutes) that required little or no teacher preparation, special equipment, or resources.
• Extracurricular physical activities: The evidence suggests that superintendents, principals, and athletic directors can develop or continue school-based sports programs without concern that these activities have a detrimental impact on students' academic performance. School administrators and teachers also can encourage after-school organizations, clubs, student groups, and parent groups to incorporate physical activities into their programs and events.

INTRODUCTION
When children and adolescents participate in at least 60 minutes of physical activity every day, multiple health benefits accrue. 1,2 Regular physical activity builds healthy bones and muscles, improves muscular strength and endurance, reduces the risk for developing chronic disease risk factors, improves self-esteem, and reduces stress and anxiety. 1 Beyond these known health effects, physical activity may also have beneficial influences on academic performance.
Children and adolescents engage in different types of physical activity, depending on age and access to programs and equipment in their schools and communities. Elementary school-aged children typically engage in free play, running and chasing games, jumping rope, and age-appropriate sports-activities that are aligned with the development of fundamental motor skills. The development of complex motor skills enables adolescents to engage in active recreation (e.g., canoeing, skiing, rollerblading), resistance exercises with weights or weight machines, individual sports (e.g., running, bicycling), and team sports (e.g., basketball, baseball). 1,3 Most youth, however, do not engage in the recommended level of physical activity. For example,

Defining Academic Performance
In this review, academic performance is used broadly to describe different factors that may influence student success in school. These factors fall into three primary areas: • Cognitive Skills and Attitudes • (e.g., attention/concentration, memory, verbal ability).
only 17.1% of U.S. high school students meet current recommendations for physical activity (CDC, unpublished data, 2009).
Schools, which serve nearly 56 million youth in the United States, provide a unique venue for youth to meet the physical activity recommendations. 4 At the same time, schools face increasing challenges in allocating time for physical education and physical activity during the school day. Many schools are attempting to increase instructional time for mathematics, English, and science in an effort to improve standards-based test scores. 5 As a result, physical education classes, recess, and other physical activity breaks often are decreased or eliminated during the school day. This is evidenced by data from both students and schools. For example, in 2007 only 53.6% of U.S. high school students reported that they attended physical education class on 1 or more days in an average week at school, and fewer (30%) reported participating in physical education classes daily. 6 Similarly, in 2006 only 4% of elementary schools, 8% of middle schools, and 2% of high schools in the United States provided daily physical education or its equivalent for all students in all grades. 7 Furthermore, in 2006 only 57% of all school districts required that elementary schools provide students with regularly scheduled recess. As for physical activity outside of physical education and recess, during the school day, 16% of school districts required elementary schools, 10% required middle schools, and 4% required high schools to provide regular physical activity breaks. 7 In addition to school-day opportunities, youth also have opportunities to participate in physical activity through extracurricular physical activities (e.g., school sports, recreation, other teams), which may be available through schools, communities, and/or after-school programs. 8 Seventy-six percent of 6-to 12-year-olds reported participating in some sports in 1997, 9 and in 2007, 56% of high school students reported playing on one or more sports teams organized by their school or community in the previous 12 months. 6 There is a growing body of research focused on the association between school-based physical activity, including physical education, and academic -performance among school-aged youth. 3,[10][11][12][13][14][15][16] This developing literature suggests that physical activity may have an impact on academic performance through a variety of direct and indirect physiological, cognitive, emotional, and learning mechanisms. 12,17,18 Research on brain development indicates that cognitive development occurs in tandem with motor ability. 19 Several review articles also have examined the connections between physical activity and academic behavior and achievement. Sibley and Etnier 12 conducted a meta-analysis of published studies relating physical activity and cognition in youth. Two additional reviews described the evidence for relationships between physical activity, brain physiology, cognition, emotion, and academic achievement among children, drawing from studies of humans and other animals across the lifespan. 14,20 Finally, two other reviews summarized select peer-reviewed research on the relationship between physical activity and academic performance, with an emphasis on school settings and policies. 15,16 Research also has explored the relationships among physical education and physical activity, fitness levels and motor skill development, and academic performance. For example, several studies have shown a positive relationship between increased physical fitness levels and academic achievement 10,21-27 as well as fitness levels and measures of cognitive skills and attitudes. 28 In addition, other studies have shown that improved motor skill levels are positively related to improvements in academic achievement 29-31 and measures of cognitive skills and attitudes. [32][33][34] To extend the understanding of these connections, this review offers a broad examination of the literature on a range of physical activity contexts, including physical education classes, recess, classroom-based physical activity breaks outside of physical education class and recess, and extracurricular physical activity, thereby providing a tool to inform program and policy efforts for education and health professionals. The purpose of this report is to synthesize the scientific literature that has examined the association between schoolbased physical activity, including physical education, and academic performance, including indicators of cognitive skills and attitudes, academic behaviors, and academic achievement.
How Physical Activity Affects the Brain 16

Conceptual Definitions
The research on this topic suggests that physical activity can be related to many different aspects of academic performance (e.g., attention, on-task behavior, gradepoint average [GPA]), and as a result, the existing literature examines a wide range of variables. In this report, those variables are organized into three categories: 1) cognitive skills and attitudes, 2) academic behaviors, and 3) academic achievement. The three categories, as well as other important terms used in this report, are defined below.
Academic Performance: In this review, academic performance is used broadly to describe different factors that may influence student success in school. These factors are grouped into three primary areas:

1) Cognitive Skills and Attitudes
Cognitive skills and attitudes include both basic cognitive abilities, such as executive functioning, attention, memory, verbal comprehension, and information processing, as well as attitudes and beliefs that influence academic performance, such as motivation, self-concept, satisfaction, and school connectedness. Studies used a range of measures to define and describe these constructs.

2) Academic Behaviors
Academic behaviors include a range of behaviors that may have an impact on students' academic performance. Common indicators include on-task behavior, organization, planning, attendance, scheduling, and impulse control. Studies used a range of measures to define and describe these constructs.

3) Academic Achievement
Academic achievement includes standardized test scores in subject areas such as reading, math, and language arts; GPAs; classroom test scores; and other formal assessments.
Physical Education: Physical education, as defined by the National Association for Sport and Physical Education (NASPE), is a curricular area offered in K-12 schools that provides students with instruction on physical activity, health-related fitness, physical competence, and cognitive understanding about physical activity, thereby enabling students to adopt healthy and physically active lifestyles. 35 A high-quality physical education program enables students to develop motor skills, understand movement concepts, participate in regular physical activity, maintain healthy fitness levels, develop responsible personal and social behavior, and value physical activity. 35

Recess:
Recess is a time during the school day that provides children with the opportunity for active, unstructured or structured, free play.
Physical Activity: Physical activity is defined as any bodily movement produced by the contraction of skeletal muscle that increases energy expenditure above a resting level. 1 Physical activity can be repetitive, structured, and planned movement (e.g., a fitness class or recreational activity such as hiking); leisurely (e.g., gardening); sportsfocused (e.g., basketball, volleyball); work-related (e.g., lifting and moving boxes); or transportation-related (e.g., walking to school). The studies in this review included a range of ways to capture the frequency, intensity, duration, and type of students' physical activity.
Physiology: In this report, physiology includes indicators of structural or functional changes in the brain and body. Studies most often reported measures of physical fitness, motor skills, and body composition from this construct.

Inclusion Criteria
The following criteria were used to identify published studies for inclusion in this review. Studies had to • Be published in English.
• Present original data.
• Include clear measures of physical education and/or physical activity, such as -Physical education class. -Recess.
-Classroom-based physical activity (outside of physical education and recess). -Extracurricular physical activities (including school sports and other teams).
• Measure academic performance (cognitive skills and attitudes, academic behaviors, and academic achievement) using one or more educational or behavioral outcomes. Examples include -Graduation or dropout rates (n=2).
Studies were excluded if they did not meet the above criteria or if they focused solely on sedentary lifestyle variables, overweight status, or media use rather than physical activity. Studies also were excluded if they focused exclusively on the relationship between academic performance and fitness test scores rather than physical activity itself. Review articles, meta-analyses, and unpublished studies were excluded from the coding and analysis portion of this review, although their reference lists were used to identify original research to be reviewed for inclusion. * Articles published between October 2008 and the publication date that met the inclusion criteria and made a notable contribution to the field may have been included in the review based on expert recommendations. † School connectedness refers to students' belief that adults and peers in the school care about their learning as well as about them as individuals. 36

Identification of Studies that Met the Inclusion Criteria
Studies were identified through a search of nine electronic databases (ERIC, Expanded Academic Index ASAP, Google Scholar, PsycNET ® , PubMed, ScienceDirect ® , Sociological Abstracts, SportDiscus™), and the Cumulative Index to Nursing and Allied Health Literature (CINAHL ® ) using a pre-established set of search terms that included both physical activity and academicrelated terms (see Appendix A). Additional studies also were located from reference lists of the identified articles.

Classification of Studies
The search yielded 406 articles (see Figure 1). Two trained researchers examined each article to determine its match with the inclusion criteria; it was then classified as "included for review" or "excluded from review." When the match was unclear, articles were temporarily classified as "possible inclusion" before being reviewed by two additional researchers for final classification. Initially, 50 articles were identified for inclusion. Four of those articles were later excluded because they lacked clarity necessary to categorize them appropriately for the review. For example, one article examining movement lacked sufficient information to determine whether the movement should be classified as physical activity; another article lacked a clear academic performance variable. The other two articles lacked clarity in descriptions of analyses and testing of research questions that was necessary for categorization. A fifth article was excluded because of its focus on elite athletes rather than a general student population. Two additional articles that examined associations between participation in a sportsbased interdisciplinary curriculum and academic grades were excluded because of insufficient detail about the physical activity participation levels of students and the subsequent lack of fit into the review categories.
A total of 363 articles were excluded. Reasons for exclusion were failure to include an appropriate measure of physical activity (n=103), academic achievement (n=40), or both physical activity and academic achievement (n=25); classification as a review or metaanalysis (n=82); inclusion of participants outside the F I GU R E 1 : Article Classification System age range of interest (n=58); inability to obtain full text of the study (n=49); and a publication date outside the inclusion range (n=6).
Overall, 43 articles (describing 50 unique studies) met the inclusion criteria and were read, abstracted, and coded for this synthesis. Two articles in this review presented findings from more than one study that met inclusion criteria; one article described three studies, 37 and the other reported six. 2

Study Coding Process
The coding method for this report is similar to that of several prominent literature reviews in the public health field. 38-40 A team of eight trained reviewers read and coded the 43 articles using a standard coding protocol (see Appendix B). When multiple studies were presented in a single article, this information was noted in the coding, but the studies remained grouped by article. The coding protocol involved abstracting information from the studies and entering it into a Microsoft Access ® database. Whenever possible, information was abstracted directly from articles as stated by authors. The following information was abstracted: purpose, research questions, study design, sampling, sample characteristics, setting, theory, intervention, methods, analytic strategy, results, limitations, study focus, and additional comments. For this review, study designs were classified as experimental, quasi-experimental, descriptive, or case studies (study designs are defined in Appendix C); data collection methods and time points were noted as described. Studies that lacked details regarding any field of interest were coded as "information not provided." To ensure consistency in coding, approximately 17% of all articles were double-coded by a reviewer and a senior coder. A team of article reviewers met regularly during the coding process to discuss and resolve issues associated with coding. A system was established for handling coding questions and concerns. Senior team members resolved and verified issues as they arose.
A brief summary profile of each study was then created (see Appendices D-G). A list of the studies classified as using quasi-experimental or experimental designs is provided at the beginning of each of these appendices. These summaries were e-mailed to the studies' corresponding authors for review and verification. Authors not responding within the initial timeline received a second request for review. Seventy-two percent of the authors (31 of 43) reviewed their summaries. Author edits and suggestions were incorporated where applicable.

Data Analysis
Coded data from the articles were used to categorize and organize studies first by their physical activity context and then by outcome, cohort, sampling groups, and date published. The individual studies were identified (in the instances where articles described more than one study), and all reviewed studies were treated equally, regardless of study characteristics or design. Although meta-analysis was considered as a method to analyze data in this review, the small number and heterogeneity of studies precluded use of that method. Therefore, descriptive literature synthesis was conducted. In this report, the results describe the types of associations or relationships reported in the studies. When positive or negative associations are described in the Results section below, they refer to findings that the study authors reported as reaching statistical significance (p ≤ 0.05).

RESULTS
This review examines the findings of 43 articles (reflective of 50 studies total) that explored the relationship between physical activity and/or physical education and academic performance. Each study was categorized in one of four physical activity context areas: 1) schoolbased physical education; 2) recess; 3) classroom-based physical activity (outside of physical education and recess); and 4) extracurricular physical activity.
School-based physical education as a context category encompassed all studies that were explicitly set in physical education class or consisted of a schoolbased course or curriculum that addressed primary aspects of physical education. This category included activities conducted in physical education class but did not exclude curricula with components that extended beyond formal physical education. Typically, studies in this category examined the impact of increasing the amount of time students spent in physical education class or manipulating the types of activities conducted with students.
Recess studies explored the relationship between academic performance and recess during the school day in elementary schools. Recess is typically 10-15 minutes or longer of unstructured free play that may occur as a break during the school day or in association with lunch.
Classroom-based physical activity as a context category included studies that were classroom-based but were not physical education class or recess. In general, these studies explored short physical activity breaks (5-20 minutes) or ways to introduce physical activity into learning activities that were either designed to promote learning through physical activity or provide students with a pure physical activity break. These interventions are relatively easy and inexpensive for a teacher to incorporate into the classroom.
Extracurricular physical activity as a context category encompassed studies that focused specifically on the relationships between activities organized through school that occur outside of the regular school day. This category included participation in school sports (interscholastic sports and other teams) as well as other after-school physical activity programs. Of all 50 studies in the review, almost two-thirds (62%) focused on youth physical activity experiences through school-based physical education, during recess, or in the classroom; the remaining studies (38%) examined extracurricular physical activity (see Table 1). Slightly more than half (54%) of the articles focused exclusively on students in secondary school settings; 44% included studies conducted with elementary students; and 2% included both elementary and secondary grade levels.

Results at a Glance
The scope and research designs varied as well. Most studies were descriptive (44%) or quasi-experimental (34%) in nature, and the majority (76%) reported longitudinal data. Most studies (80%) were conducted during the school day, and about two-thirds (64%) included a physical education or physical activity intervention. Finally, the majority (68%) of studies were conducted in the United States; overall, studies were conducted in nine countries other than the United States.
For ease of review, the results are presented here by physical activity context. Within each context, results are described by study focus (intervention or nonintervention) and by the type of results.
Each results subsection also includes a summary table that shows the number of associations (total, positive, negative, and no association) for all the studies reviewed in that context area. Results with p values less than 0.05 are considered statistically significant in this report.
Qualitative and descriptive studies that did not include significance testing are described in the text of this report, but not in the outcome counts. Associations are displayed by type of academic performance outcome measured: cognitive skills and attitudes, academic behavior, or academic achievement. * Studies often included more than one type of measure; thus, the number of studies that include these different academic performance measures may exceed the total number of studies in any given category. † Some studies included more than one setting; therefore, the total number of studies by setting exceeds 50.

School-Based Physical Education Studies
Fourteen studies (reported in 14 articles) examined the relationship between school-based physical education and academic performance (see Figure 2, and Tables  2a and 2b). Most (n=10) described intervention studies and assessed the impact of an intervention on a range of outcomes. The remaining four were descriptive and examined the relationships between physical education and academic measures. Appendix D includes summary profiles for each of the articles reviewed in this section. Intervention Studies. In general, the intervention studies (three implemented in the United States and seven in other countries) examined how differences in physical education affected academic performance. Six studies 41-46 examined increasing the amount of physical education or the level of physical activity intensity in physical education class and comparing students' academic performance by intervention condition (e.g., physical education two times per week versus daily physical education, or physical education for 20 minutes versus physical education for 30 or 40 minutes, or the intensity of physical activity during physical education). Two studies examined strategies for improving the quality of physical education: one focused on implementation by trained instructors of a curriculum that promotes greater amounts of moderate and vigorous physical activity in students, and the other implemented multiple strategies such as focusing on students' personal goal setting, emphasizing opportunities for active participation by all students, and maximizing active use of class time. 47,48 The remaining two studies examined the relationship between increasing the emphasis on different types of activities (i.e., aerobic exercise, coordinative exercise) and aspects of academic performance. 49,50 Collectively, the studies were conducted across a broad range of grade levels, representing elementary, middle, and high schools. Seven studies employed an experimental design, and three reported data from quasi-experimental designs. Most studies involved shortterm follow-up (e.g., immediate or 3-month delay). Sallis and colleagues 48 and Ericsson 43 both followed youth for approximately 3 years.

School-Based Physical Education Studies: Highlights
Finally, the studies assessed a range of indicators of academic performance, including cognitive skills (e.g., concentration and creativity), attitudes (e.g., self-esteem and motivation), academic behaviors (e.g., conduct), and/or academic achievement (e.g., standardized test scores and GPA).
Results varied across the 10 intervention studies, with most (8 of 10) showing one or more positive associations. Two studies showed all or mostly positive associations between physical education and cognitive skills and attitudes or academic achievement. Specifically, Ericsson 43 found that extending physical education (from 2 days per week to daily) was associated positively with academic achievement (math, reading, and writing test scores). That study also noted positive associations for attention, an indicator of cognitive functioning, although the relationships dissipated over time. Budde and colleagues 49 found that coordination exercises (i.e., exercises that require the body to balance, react, adjust, and/or differentiate) were more beneficial than normal sport lessons in boosting cognitive functioning (specifically, concentration and attention).
Six studies reported more mixed conclusions. Five found a mixture of positive and nonsignificant associations. 41  (fitness group: 75 minutes of activity daily, with an emphasis on intensity of activities; skill group: 75 minutes of activity daily with no focus on intensity; and control group: three 30-minute periods of physical education per week). They found no differences in academic achievement across the three intervention conditions, despite the fact that students in the fitness and skill groups actually had less classroom teaching time to accommodate the increase in time for physical education. They also found that classroom behavior improved for students in the skill and fitness intervention conditions. The sixth study found four positive and three nonsignificant associations, as well as one negative relationship. 48 Sallis and colleagues 48 examined an intensive 2-year health-related physical education program that was taught by trained classroom teachers or physical education specialists and was designed to increase students' physical activity levels. They found that the SPARK program taught by trained teachers had a positive impact on reading, language, and basic battery standardized test scores, but had no significant impact on math. When taught by physical education specialists, students in the SPARK program scored better than students not enrolled in SPARK on reading, but lower on language and about the same in math.
Finally, two studies found no associations between physical education and indicators of academic performance. These studies examined the relationship between the frequency of physical education and either cognitive skills and attitudes 46 or academic achievement. 45 Raviv and Low 46 found that physical education did not reduce concentration, contrary to the beliefs of some teachers in their study. Pollatschek and O'Hagan 45 found that the frequency of physical education participation (daily versus twice a week) was not associated with students' standardized math and reading test scores or affect towards school; similar results were found for boys and girls.
Collectively, the results of these studies suggest that physical education may have favorable associations with students' cognitive skills and attitudes and their academic achievement, but the relationships are not universal and vary by outcome studied. Furthermore, increasing time for physical education does not appear to have negative associations with academic achievement.
Nonintervention Studies. The four nonintervention studies (two conducted in the United States and two in other countries) examined associations between physical education and academic performance using cross-sectional designs (n=3) or secondary analyses of an existing longitudinal data set (n=1). Three of the four studies were conducted at the elementary or middle school level; the fourth study was completed with high school students. All studies used standardized tests to assess academic achievement. Results were either positive or neutral. Three of the studies found positive associations between time spent in physical education or skills learned in physical education and indicators of academic achievement. As an example, one study 51 noted a positive association between standardized English language arts test scores and time spent in physical education but found no such association for math scores. Another study 52 found small but significant associations between physical education and academic achievement in math and reading for girls who had more physical education (70-300 minutes per week) compared with those getting lower amounts (0-35 minutes per week); none of the associations were significant for boys. Dexter 53 found a combination of positive associations and no association between performance on sports learned in physical education and an average of math and English test scores and grades, depending on the sport; results were similar for boys and girls. The remaining study found no significant associations between physical education and academic performance on state literacy and numeracy tests. 54 There were no negative associations between physical education and indicators of academic performance across these four studies. Consistent with the results of the physical education intervention studies, the data from these four studies suggest physical education has some positive associations with academic outcomes, but these results vary by outcome.

Strengths and Limitations of Methods.
This collection of studies has a number of strengths as well as limitations. The studies were conducted across a range of grade levels and used a broad array of indicators related to cognitive skills, attitudes, and academic achievement. Furthermore, nearly half featured experimental designs, and half explored associations by gender. Several limitations were noted by the authors of the studies, including small samples or samples with potential biases that may affect the generalizability of the results (e.g., university research/laboratory school populations or affluent populations). Several authors acknowledged measurement issues, such as limited follow-up, not assessing precursors of academic achievement (e.g., concentration, memory, or classroom behavior), or failing to collect data on socioeconomic status (SES) and other potentially important background variables. Finally, authors of intervention studies also noted implementation limits, such as unequal participation in the intervention or lack of data on implementation quality. Many of the studies did not report data on the racial/ethnic characteristics of their samples, and only one examined results by racial/ethnic subgroups.

Recess Studies: Highlights
• • All eight studies found one or more positive associations between recess and indicators of cognitive skills, attitudes, and academic behavior.
• • Time spent in recess appears to have a positive relationship or no relationship with children's attention, concentration, and/or on-task classroom behavior.

Recess Studies
Eight studies (reported in six articles) examined the relationship between school recess, cognitive skills, attitudes, and/or academic behavior (see Figure 3, and Tables 3a and 3b). Six of the studies tested an intervention to examine how recess impacts these indicators of academic performance. The other two descriptive, nonintervention studies explored the relationships between recess and school adjustment or classroom behavior. Appendix E includes summary profiles for each of the articles reviewed in this section.

FIGURE 3:
Type of Association Observed for Cognitive Skills and Attitudes, Academic Behaviors, and Academic Achievement Outcomes Across All Recess Studies (n=17 associations within 8 studies) Intervention Studies. The six intervention studies (all implemented in the United States) examined the relationship between recess, or increased physical activity during recess, and cognitive skills (attention or concentration) and academic behavior (on-task behavior). All of these studies were conducted in elementary schools with students in kindergarten through fourth grade, and all six employed an experimental or quasi-experimental design. 37,55-57 Most used trained observers to collect data on classroom and recess behaviors, with multiple observation points. The data collection follow-up period ranged from 0 to 4 months following baseline. The interventions involved the introduction of recess into the daily school schedule or manipulating the timing of recess (e.g., holding recess after varying lengths of class time).
Results across these six studies showed positive outcomes or no association. Three studies reported all positive associations between more physical activity at recess and classroom behaviors. 55-57 Specifically, Caterino and Polak 55 found that fourth-grade students who participated in directed physical activity during recess (stretching and aerobic walking) had significantly higher concentration scores than those students who sat quietly in the library during recess. Both studies that measured academic behavior found a positive relationship between recess and on-task behavior. Jarrett and colleagues 56 observed that children were less fidgety, less listless, more focused, and more on task when they had recess compared with when they did not have recess. Pellegrini and Davis 57 found that students who engaged in physical activity (as opposed to sedentary behavior) during recess fidgeted less in the classroom after recess.
The three intervention studies by Pellegrini and colleagues 37 reported both positive and nonsignificant associations. These studies examined the relationships between the timing of recess (i.e., recess after 2.5 hours versus 3 hours of classroom time) and students' behaviors during recess and students' classroom behaviors before and after recess. Investigators found that students' attention rates were lower after longer periods of classroom work without a break than after shorter periods. They also found that, in general, students' attention was better after recess than before. Finally, they found that the type of behavior during recess did not affect classroom attention after recess for any grade or gender groups.
Nonintervention Studies. One of the two nonintervention recess studies 58 explored the impact of the frequency of recess on teacher reports of classroom behavior in a very large sample (n=11,529); the other 59 explored the impact of recess on observations of individual students' cognitive and emotional adjustment to school within one school (n=77). Barros, Silver, and Stein 58 found that overall classroom behavior (based on * Studies may have measured the relationship between recess and academic performance in more than one way (e.g., measured the association between recess and attention and behavior). Individual studies in this section measured between 1 and 9 different outcomes and may be represented in multiple cells of the table. On-task behavior (not fidgeting) * Studies may have measured the relationship between recess and academic performance in more than one way (e.g., measured the association between recess and perceptions of school adjustment and on-task behavior). Individual studies in this section measured between 1 and 2 different outcomes and may be represented in multiple cells of the table.
teacher ratings) was significantly better for students who had recess every day for at least 15 minutes than for those who did not. Exploration of the impact of recess on individual students showed a positive association with end-of-year social competence and perceptions of school adjustment for boys, but not for girls. 59

Strengths and Limitations of Methods.
These studies feature several strengths as well as a few notable limitations. Six of the eight studies used experimental or quasi-experimental designs, and most involved observations of student behaviors with multiple observation points (e.g., 6, 12, or 32 observations). The studies focused on elementary-level students, primarily because recess is most common at the elementary grade levels. Study authors reported a number of limitations including small sample sizes (range of 23-77 students in seven of the eight studies), and the inability in most of the studies to analyze data by SES, race/ethnicity, or other subgroups. In addition, the authors noted that classroomlevel ratings of student behavior by the classroom teacher could be influenced by the teachers' perceptions of the benefits of recess.

Classroom Physical Activity Studies
Nine studies (reported in nine articles) examined the relationship between classroom-based physical activity and academic performance (four implemented in the United States and five in other countries) (see Figure 4 and Table 4). All nine of the studies were interventions. Appendix F includes summary profiles for each of the articles reviewed in this section.
These studies examined how the introduction of brief physical activities in a classroom setting affected cognitive skills (e.g., aptitude, attention, memory); attitudes (e.g., mood); academic behaviors (e.g., on-task behavior, concentration); and academic achievement (e.g., standardized test scores, reading literacy scores, math fluency scores). The interventions involved the introduction of physical activities by trained teachers or facilitators into the classroom setting, with activities lasting 5-20 minutes per session. Physical activity sessions or breaks typically were delivered on a daily or regular basis. Intervention implementation periods spanned from 1 day to 16 months, with most lasting 2-3 months.
All but two of these studies were conducted with elementary school students in first through fifth grades; the others were conducted in a primary and secondary school in Sweden 60 and an urban middle school in the United States. 61 Five studies employed quasiexperimental designs, 60-64 three used experimental designs, 32,65-67 and one used a qualitative case-study design. 68 The data collection follow-up period ranged from 0 to 12 months after the intervention. Outcome measures most often included standardized aptitude and achievement tests and teacher or trained observer ratings of classroom behavior.

FIGURE 4:
Type of Association Observed for Cognitive Skills and Attitudes, Academic Behaviors, and Academic Achievement Outcomes Across All Classroom Physical Activity Studies (n=20 associations within 9 quantitative studies)

Classroom Physical Activity Studies: Highlights
• • Eight of the nine studies found positive associations between classroombased physical activity and indicators of academic performance.
• • One study examined gender effects and found no differences in outcomes by gender.
Results across the nine intervention studies showed positive outcomes or no association. Four studies reported all positive associations between classroom physical activity and classroom behaviors and academic achievement. 60-63 Specifically, Della Valle and colleagues 61 found using movement with seventh-grade learners who had an active learning style enhanced their performance on a word recognition task. Maeda and Randall 62 reported that second-grade students exhibited greater concentration and demonstrated higher math fluency after engaging in brief movement breaks consisting of 5 minutes of vigorous exercise 1 hour after lunch. Similarly, Mahar et al. 63 observed greater frequency of verbal and motor behavior that followed class rules and was appropriate to the learning situation for third-and fourth-grade students whose teachers led them in daily 10-minute regimens of physical activities (e.g., jumping, rolling, hopping, twisting) during academic instruction. This relationship was especially strong among students who were least on task at baseline. Furthermore, Norlander and colleagues 60 found that teachers observed higher student concentration levels after daily stretching exercises.
Four intervention studies reported positive and nonsignificant associations. 64,66-68 Fredericks et al. 66 described improvements in spatial aptitude, reading skills, and math skills among first-grade students exposed to daily classroom exercises focused on the development of perceptual and sensory motor skills. However, there were no associations with other indicators of aptitude, such as perception, reasoning, memory, and verbal comprehension or emotional indicators. In their feasibility study, Lowden et al. 68 qualitatively described that students and teachers perceived that student exposure to The Class Moves! ® program was positively related to improvements in on-task classroom behaviors and concentration. Teachers, however, did not feel they could relate the program to academic or cognitive achievement. Molloy 64 observed that students exposed to 5 minutes, but not 10 minutes, of aerobic exercise improved their arithmetic performance. Exposure to aerobic exercise was unrelated to observed on-task behavior (attention) for all but a small sample of hyperactive students. Uhrich and Swalm 67 found that daily sessions to develop motor skills (bimanual coordination) through a sport cupstacking exercise were associated with improvements in reading comprehension but not reading decoding scores. These improvements were comparable for boys and girls.
The ninth intervention study found no relationship between an additional 15 minutes of daily classroombased physical activity (skipping, dancing, and resistance exercises) in the context of a schoolwide physical activity program and standardized achievement tests. 65 The classroom intervention lasted * One qualitative study (Lowden 68 ) and one quantitative study (Maeda and Randall 62 ) that did not include significance testing were not included in these results. Studies may have measured the relationship between classroom physical activity and academic performance in more than one way (e.g., measured the associations among classroom physical activities and ability, classroom behaviors, and standardized test scores). Individual studies in this section measured between 1 and 11 different outcomes.
16 months and was designed to complement 80 minutes of weekly physical education. Analyses by gender showed similar results.
Collectively, eight of the nine studies reviewed suggest that classroom-based physical activities may have favorable associations with indicators of cognitive functioning, academic behaviors, and/or academic achievement. Furthermore, there was no evidence that allotting classroom time for these activities was negatively associated with academic achievement.

Strengths and Limitations of Methods.
These studies feature both strengths and important limitations.
Eight of the nine studies employed either experimental or quasi-experimental designs, and most used standardized measures of cognitive functioning and academic achievement and standardized protocols for classroom observations. Several studies collected data at multiple follow-up dates. When reported, study populations represented an array of racial and ethnic backgrounds. Limitations reported by study authors include small sample sizes, with all but two studies having fewer than 100 students, and the inability to analyze data by SES, race/ethnicity, or other subgroups. Others noted that classroom observers typically were not blinded to study condition. Some authors also noted concerns about group comparability at baseline and its potential impact on determining an intervention effect.

Extracurricular Physical Activity Studies
Nineteen studies (reported in 14 articles) examined the relationship between involvement in extracurricular physical activity (such as interscholastic sports or other physical activities outside of the regular school day) and academic performance (see Figure 5, and Tables 5a and 5b). Nine studies focused on involvement in school interscholastic sport teams; the other 10 focused on other school-related extracurricular physical activities. Appendix G includes summary profiles for each of the 14 articles reviewed in this section.

Interscholastic School Sports
All nine of the studies assessing the relationship between school sports team participation and academic performance were descriptive in nature and focused on secondary school students. [69][70][71][72][73][74][75][76][77] Most studies (n=8) were implemented in the United States. Eight of the nine studies examined how students' participation on sports teams was related to test scores, grades, or teacher ratings of academic achievement; two 75,77 measured dropout rates.

Extracurricular Physical Activity Studies: Highlights
• • Nearly all the associations between extracurricular physical activity and indicators of academic performance were either positive (52%) or neutral (46%).
• • Grade point average was positively associated with extracurricular physical activity 12 of the 22 times it was measured.
• • The two different interventions aimed at improving academic performance through extracurricular physical activity had some positive impacts on students' grades and/or verbal and conceptual skills.
• • Two studies examined the relationship between extracurricular physical activity and dropout rates and found that participation was associated with decreased high school dropout rates.
Three of the nine studies were cross-sectional, collecting data at one time point; six were longitudinal and involved a secondary analysis of data collected at baseline and 3-5 years later. Two of the nine studies had small samples ; the remaining studies had larger sample sizes (883-14,249).
These studies varied in measurement of academic performance and participation in sports. Some used school records (test scores, GPAs, or dropout rates), 72,74,75,77 and one used a teacher rating of student academic ability 71 for students who participated in interscholastic sports. Others examined the relationship between student report of participation on sports teams (type of team was not specified) and students' selfreported grades. 69,70,73,78 The studies that explored the relationship between school records of academic achievement and student participation in interscholastic sports found mostly positive and neutral results. For example, one study 74 of eighth-grade students found that participation in interscholastic sports was associated with higher math grades, higher math standardized test scores, and higher overall GPAs; however, another study 72 of 12th-grade students found no relationship between varsity sports participation and grades. Two studies 75,77 examined the impact of sports on high school dropout rates and found that participants were significantly less likely to drop out of school compared with nonparticipants.
Two studies 75,76 found that the relationship between academic achievement and varsity sports participation was inconsistent, showing positive, negative, and no association, depending on the outcome measured and the grade level of the students. Fredricks and Eccles 76 found participation in seventh-grade school sports was associated with a decreased school value in eighth and 11th grades but was associated with increased resiliency in 11th grade. Yin and Moore 75 found that students who reported participation in interscholastic sports in eighth grade showed significantly lower test scores for that year compared with students who did not participate. However, as these same students moved through high school, those differences disappeared, showing no differences in test scores between sport participants and nonparticipants in the 10th or 12th grades. Hawkins and Mulkey's 71 exploration of the relationship between interscholastic sports participation and teacher ratings of students' academic ability showed no relationship between participation and academic ability; however, other measures of academic behavior and cognitive skills and attitudes showed positive relationships or no relationship, varying by gender. As an example, male athletes were more likely to plan to attend college than nonathletes, and both male and female athletes showed greater interest in class than nonathletes.
Three of the four studies that examined the relationship between student report of participation in sports teams and self-report of grades showed positive relationships. Fredricks and Eccles 70 also found a positive relationship between sports participation and students' educational expectations and school completion rates.
Seven of the nine studies examined gender effects [69][70][71][72]74,75,76 on academic performance; five of the seven studies found at least one significant difference by gender; however, overall, 68% of the associations by gender showed no relationship. One study 76 also examined results by race and SES. No other subgroup or demographic analyses were reported in the other studies.
Other School-Related Extracurricular Physical Activity Ten studies focused on extracurricular physical activities organized through the school but conducted outside the regular school day (e.g., after school). Seven studies examined the effects of an intervention, 2,34 and the other three were descriptive, 79,80,78 with sample sizes ranging from 35 to 4,264. Measures of academic performance included grades, math scores, homework completion, and attendance.
Intervention Studies. One intervention article 2 focused on physical activity opportunities after school using six different studies. This article, which focused on studies conducted in the United States, assessed the impact of a life skills program with an emphasis on improving physical fitness on students' self-reported grades, school attendance, and self-concept. The program was taught after school in a sample of middle schools, high schools, and community centers. As part of the program, students completed an individual exercise program as well as instruction about related topics such as self-assessment, goal setting, fitness, and exercise planning. Program impact was evaluated at six sites immediately after the program. Results showed positive associations between program participation and academic performance (grades and attendance) or no significant relationships. The positive findings (for selfconcept, school attendance, and self-reported grades) were concentrated in the community site, which had the largest sample size. Across all sites (middle schools, high schools, and community centers), self-concept improved significantly for program participants.
The other intervention study was conducted in the United Kingdom and examined participation in a schoolorganized, year-long exercise program completed at home and its relationship with cognitive skills and math * Studies may have measured the relationship between extracurricular physical activity and academic and cognitive performance in more than one way (e.g., measured the association between extracurricular physical activity and grade point average, self-concept, and attendance).
Individual studies in this section measured between 3 and 11 different academic measures. Consequently, the number of studies across the three academic performance areas exceeds 2.
outcomes of 7-to 10-year-old children diagnosed with, or at risk of, dyslexia or dyspraxia. 34 This study used a quasi-experimental design with immediate and longterm (3-year) follow-up. Little detail was provided on how the physical activity intervention was developed or implemented. Children showed improved verbal and cognitive skills following the individualized physical activity intervention, although there were no math improvements. 34 Nonintervention Studies. The three nonintervention studies, all conducted with secondary students in the United States, examined associations between participation in after-school physical activity and academic performance using existing data sets (one cross-sectional and two longitudinal). Two studies 78,79 explored the association between student reports of participation in extracurricular activities and student selfreported grades, including involvement in a combination of sports and nonsport activities. Those studies found consistently positive associations between extracurricular activity participation and self-reported grades as well as positive academic attitudes and higher academic aspirations. Harrison and Narayan 80 examined the impact of participation in after-school activities (including participating 1-2 hours per week or more in school sports) on homework completion and class attendance. The study showed that physical activity participation was positively related to homework completion and class attendance.

Strengths and Limitations of Methods.
These studies featured a number of strengths. Most had relatively large sample sizes. Most (n=16) of the studies focused on measures of academic achievement, such as standardized test scores or grades (record data or self-reported data). In addition, of the studies that examined sports participation compared with nonparticipation, more than half (n=6) specified the level of competitiveness of team participation; nonetheless, these studies did not compare outcomes by varying levels of sports competitiveness.
More than three-fourths of the studies were longitudinal in nature (n=15), allowing for an exploration of causality; the cross-sectional nature of the remaining studies (n=4) limited the ability to establish the temporal relationship between the variables. Relatively few studies examined data by race/ethnicity, and only two explored physical activity interventions. Reports were unclear whether sport participation required a minimal level of academic achievement, a requirement that could bias the samples.
Several authors also acknowledged limitations such as the need to include measures of social influence (e.g., parental support) in future research, the need to look more closely at how level of participation or competitiveness in sport might affect academic achievement, and the fact that some of the associations found were relatively weak.

SUMMARY
This report identified peer-reviewed studies and published reports addressing the association between physical activity, including physical education, and indicators of academic performance, including those related to cognitive skills and attitudes, academic behaviors, and academic achievement.

Overall Findings
Overall, what do the results of these studies say about the relationship between physical activity and academics, and what does it mean for schools?
• Collectively, the results suggest that physical activity is either positively related to academic performance (50.5% of the associations summarized) or that there is not a demonstrated relationship between physical activity and academic performance (48% of the associations summarized). In addition, increasing time during the school day for physical activity does not appear to take away from academic performance. This pattern of having positive relationships or no relationships, along with the lack of negative relationships, was consistent throughout the results, despite the heterogeneous nature of the included studies, and is consistent with other published reviews. 15,81 • School boards, school administrators, and principals can feel confident that maintaining or increasing time dedicated for physical activity during the school day will not have a negative impact on academic performance, and it may positively impact students' academic performance.
What kinds of academic outcomes were positively related to physical activity?
• Studies looked at a broad range of outcomes. Researchers reported that participating in physical activity was positively related to outcomes including academic achievement, academic behaviors, and indicators of cognitive skills and attitudes, such as concentration, memory, self-esteem, and verbal skills.
Which outcomes were most positive?
• Positive associations were found across measures of academic achievement, academic behavior, and cognitive skills and attitudes, but there are some interesting patterns for different outcomes within these categories. Seven articles describing intervention studies (three school-based physical education, two recess, one classroom-based physical activity, and one extracurricular activity) evaluated the relationship between physical activity and academic behaviors, such as classroom conduct. 2,41,42,50,56,57,63 The majority of these articles (86%) found at least one positive association with academic behavior outcomes. Given these findings, physical activity interventions may offer one approach to improving academic behaviors (e.g., classroom conduct) in some youth.

Does physical activity have any negative relationship with grades or test scores?
• Very few of the findings in the studies reviewed were negative (only 4 associations of 251 examined), a percentage small enough to reasonably be expected by chance. This pattern of results is consistent with other reports 15,16 that suggest that adding physical activity to the school day does not detract from academic performance. The evidence suggests that superintendents and principals can devote school time to physical activity without concern that it will lower student test scores.
Why are some of the study results positive whereas others show no relationship?
• There are a number of possible explanations. Some of the studies had relatively small samples, which can make it more difficult to find statistically significant results. Other studies measured a very broad range of student attitudes and behaviors to try to understand which factors may be related and which may not.
Other issues, such as the questionnaires used in the studies, may account for some of the differences. Finally, differences that may not have been discussed in the studies-such as the intensity or duration of the physical activity, the context in which the physical activity took place, individual student differences (e.g., in motivation), and levels and quality of implementation for intervention studies-may help explain the different results among the studies.

Findings for Physical Activity by Context
Is school-based physical education related to academic performance?
• The study results suggest that school-based physical education either leads to a positive result or is associated with no change in academic performance. Overall, 11 of 14 studies found one or more positive associations between physical education and indicators of cognitive skills and attitudes, academic behavior, and/or academic achievement. Nearly half the associations (49.5%) between physical education and academic performance were positive; nearly all remaining associations in this context area showed no relationship.
• The studies also suggest that increased time spent in physical education is not likely to detract from academic performance even when less time is devoted to subjects other than physical education. Across the nine studies that examined the relationship between time spent in physical education and academic performance, 16 outcomes were positive and 31 showed no association. No negative associations were found.

Is recess related to academic performance?
• Yes, for some outcomes. Eight studies meeting the criteria of this review looked at the impact of physical activity during recess on academic performance. Of all outcomes measured in this context area, 59% were positive. In addition, all eight studies found one or more positive findings suggesting that recess was associated with improvements in attention, concentration, and/or on-task classroom behavior. None of the studies looked directly at the association between recess and measures of academic achievement (e.g., test scores or grades).
• None of the studies reported negative relationships, which indicates that recess does not appear to detract from students' focus in the classroom.
Are physical activity breaks during class related to academic performance?
• Yes, for some outcomes. Nearly all studies (eight of nine) in this category found that offering physical activity breaks during standard classroom instruction may have favorable associations with some indicators of cognitive functioning (e.g., attention/ concentration); academic behaviors (e.g., classroom conduct); and/or academic achievement (e.g., test scores). Of the individual outcomes studied, 40% of associations between physical activity breaks and academic performance were positive and 60% showed no relationship.
• None of the studies found negative associations. Classroom physical activity breaks do not appear to have a negative relationship with academic performance. Indeed, classroom teachers can include physical activity breaks as one strategy to promote academic-related benefits for students. Furthermore, incorporating brief physical activity breaks into the classroom might contribute to students' overall levels of physical activity and health. 82 Is participation in extracurricular physical activities at school related to academic performance?
• Yes, for some outcomes. More than half of the associations examined in these studies were positive (52% overall), and almost none were negative (2%). Of note, GPA was positively associated with extracurricular physical activity 12 of the 22 times it was measured. Two studies also examined the association between extracurricular activities and dropout rates and found that participation was linked to decreased high school dropout rates.

Findings by Gender, Other Demographic Characteristics, and Research Design
Do the results vary by gender?
• Relatively few studies examined differences in associations by gender, and there were no distinct patterns. Of studies that did examine data by gender, a few found differences, but most did not. For example, eight of the studies on school-based physical education examined data by gender. Six found no differences by gender; one found effects favoring boys (higher-level motor skills were associated with greater increases in reading and math scores among boys than girls); and one found effects favoring girls (greater time spent in physical education was related to higher reading and math scores for girls but not for boys).

How do the studies and results differ by grade level?
• Studies of recess and classroom-based physical activity tended to be from elementary school settings, and studies of extracurricular physical activity tended to be from secondary school settings. • Overall, the pattern of results appeared slightly more positive in the secondary school setting. Of the associations examined among elementary youth only, 43% were positive, 56% were neutral, and 1% were negative. At the secondary level, 55% of the associations examined were positive, 43% were neutral, and 2% were negative.

Do the results vary by race/ethnicity?
• Very few studies examined the relationships between physical activity and academic performance by race or ethnicity, so it is difficult to make conclusions at this time. Of the seven studies that explored race/ ethnicity, most focused on how race/ethnicity affected participation in physical activity rather than on how it influenced the association between physical activity and academic achievement. One study examining classroom-based physical activity breaks by race found no differences in academic performance between Asian and Caucasian students. 65 A study of an 8-week movement intervention found that language spoken (Afrikaans, English, and Other), used as a proxy indicator for race/ethnicity, may have explained some differences in children's spatial aptitude. 66 Do the results vary by research design?
• Not much variation in results by research design was noted. Although many factors influence a study's quality, experimental or quasi-experimental research designs are generally regarded as more rigorous. The pattern of associations in studies with either of these types of design had very similar results. In the 29 studies using experimental or quasi-experimental designs, 50% (55 of 109) of associations were positive, and 49% (53 of 109) were not significant. Less than 1% (1 of 109) of the associations were negative.

Strengths and Limitations of Review
What are the strengths of this review?
• This review has a number of strengths. It covers 23 years of research; it involved a systematic process for locating, reviewing, and coding the studies; articles were obtained using an extensive array of search terms and international databases; articles were reviewed by multiple trained coders; and the articles cover a broad array of contexts in which youth participate in schoolbased physical activities. Furthermore, a majority (64%) of studies included in the review were intervention studies, and a majority (76%) were longitudinal.

What are the limitations of this review?
• This review summarizes all studies that met the established review criteria, regardless of the study characteristics. The studies were not ranked, weighted, or grouped according to their strengths and limitations; as a result, findings from studies with more rigorous research designs and larger sample sizes were given no more influence than findings from studies with weaker designs and smaller sample sizes. Instead, results were based on counts of statistical findings, and this, in essence, had the effect of allowing individual studies containing multiple comparisons to have a greater influence on the findings as a whole. The number of statistical findings in any given study ranged from 1 to 32; given those differences, it becomes clear that a single study with 32 comparisons would have influenced the overall results more than a study that included only 1 comparison.
In addition, the breadth of the review, while revealing a variety of study designs, measures, and populations, often made comparisons and summaries difficult. For example, similar constructs were often defined and/or measured differently across studies. Even something as seemingly consistent as standardized test scores can vary from state to state. Therefore, these inconsistencies limit the ability of this or any review to draw specific conclusions across all studies.
For the same reason, it was not appropriate to make broad statements about effect sizes. Although the studies in this review include examples of moderate and large effect sizes, 47,51,63 there were not enough studies analyzing the same variables in any given category to make summary statements about the magnitude of associations between physical activity and academic performance variables. As a result, conclusions do not summarize magnitudes of effect sizes and are intentionally broad.

Implications for Future Research or Evaluation
What other research or evaluation needs to be done to further the field in this area?
• Within the contexts reviewed, there were relatively few studies of the impact of recess and classroom physical activity on academic achievement. None of the reviewed studies examined the relationship of sports and academic achievement within the primary grades or the relationship of physically active breaks/recess and academic achievement within the secondary grades. Few studies conducted subgroup analyses beyond gender comparisons.
• Less than half of the studies described effect sizes or magnitudes of the associations observed. Reporting of effect sizes can guide researchers and practitioners towards interventions most likely to impact outcomes of interest.
• Although nearly all of the reviewed studies described a practical framework for the research, few of the studies articulated a theoretical basis for the work or explicitly described how the findings informed theory development. Theoretical specificity may enable researchers to more easily identify relevant bodies of work from other disciplines, consider new relationships and mechanisms of action, align and strengthen intervention design and measurement, and ultimately progress the field more effectively and efficiently.
• Improved understanding of the specific cognitive and behavioral impacts of particular physical activities could inform intervention developers and improve the match between interventions, populations, and educational goals. For example, compared with measures of cognitive skills and attitudes, academic behaviors such as on-task classroom behavior or following instructions were less likely to be examined as proximal outcomes of physical activity or potential mediators of academic achievement.
• Future research should further examine the relationship between school-based physical activity and academic performance in subpopulations of students (e.g., based on gender, race/ethnicity, or SES). Results from this type of research could help physical education teachers and physical activity coordinators apply findings of programs and interventions to meet the needs of particular groups of students.
• Future research should be developed in consultation with educators (e.g., school administrators and staff) and informed by research across disciplines, such as neurobiology, cognitive science, social psychology, and kinesiology. For instance, few studies placed the work within a neurobiology model to better understand the role of brain physiology, within an ecological framework to account for contextual variables, or within a developmental perspective to examine developmental differences in relationships between physical activity and academic achievement.
• Future research and evaluation would benefit from identifying uniform ways to measure key outcomes, including both physical activity and academic performance outcomes. Similarly, future studies would benefit from larger sample sizes and stronger research designs that include longitudinal follow-up, as appropriate. Adequate follow-up of interventions has been more limited in physical education compared with other contexts.

Implications for Schools
What are the policy and practice implications from this review?
• Schools should continue to offer or increase opportunities for physical activity. There is evidence that physical activity may help improve academic performance (including grades and standardized test scores) in some situations. Increasing or maintaining time dedicated to physical education does not adversely impact academic performance.
• The studies in this review also suggest that physical activity can impact cognitive skills and attitudes, important components of improved academic performance. This includes enhanced concentration and attention as well as improved classroom behavior.
• Taking all of the evidence into account, schools should strive to meet the National Association for Sport and Physical Education's recommendation of daily physical education and offer students a balanced academic program that includes opportunities for a variety of daily physical activities.
What are the current recommendations for students' physical activity?
• Recent recommendations indicate that 6-to 17-year olds should be participating in at least 60 minutes of physical activity daily, 83 and schools can and should provide opportunities for physical activity to help students meet this recommendation. In fact, the Institute of Medicine's Preventing Childhood Obesity: Health in the Balance report recommended that schools provide a significant portion of students' daily physical activity. 84 • To enable students to meet these recommended levels of physical activity, the National Association for Sport and Physical Education recommends that all pre-K through grade 12 schools implement a comprehensive school physical activity program, which includes quality physical education; physical activity before, during, and after school, including recess and other physical activity breaks; extracurricular, noncompetitive physical activity clubs; interscholastic sports; and walkand bike-to-school initiatives. 35 How can schools promote physical activity at school?
• Physical activity can be included in the school environment in a number of ways without detracting from academic performance. Studies highlight potential benefits of physical activity in physical education classes, during recess, in regular classrooms, and through extracurricular sports and other physical activity opportunities.
-School-based physical education: To maximize the potential benefits of student participation in physical education class, schools and physical education teachers can consider increasing the amount of time students spend in physical education class or adding components to increase the quality of physical education class.
Studies reviewed here showed that programs were able to increase physical education time by increasing the number of days per week or the length of class time, adding trained physical education instructors, supplementing programs with community resources, and using outside facilities (e.g., swimming pools). In addition, the studies reviewed here explored several different strategies for enhancing the quality of physical education class, requiring varying levels of resources. These range from implementing a standards-and research-based physical education curriculum to adding specific components to physical education.
-Recess: Studies reviewed here used structured or unstructured play during recess as a means to provide students with time for movement and play during the school day. School boards, superintendents, principals, and teachers can feel confident that providing recess to students on a regular basis may benefit academic behaviors (e.g., attention), facilitate social development, 85 and contribute to overall physical activity 86 and its associated health benefits.
-Classroom-based physical activity: Movement activities and physical activity breaks are simple ways for classroom teachers to enhance student physical activity and possibly academic performance. Most interventions described in this review used short breaks (5-20 minutes) that required little or no teacher preparation, special equipment, or resources. As an example, interventions such as speed (cup) stacking could be a center or activity station. Simple movement-based learning techniques (e.g., walking around the perimeter of the classroom while learning vocabulary or using music and rhythmic movement to enhance memory tasks) could be incorporated into large group lessons. Short exercise breaks (e.g., 5 minutes of walking or 10 minutes of prescribed exercise) could be introduced into the classroom routine prior to teaching subjects that require intense student concentration.
-Extracurricular physical activities: The evidence in this review suggests that superintendents, principals, and athletic directors can develop or continue school-based sports programs (e.g., intramurals or physical activity clubs and interscholastic sports programs), without concern that participation in such activities would have negative associations with academic performance. Increasing or maintaining time dedicated to physical activity does not adversely impact academic performance. Indeed, studies suggest there may be a range of possible benefits for some students, including developing a stronger sense of self, fostering educational aspirations, maintaining interest in class, encouraging homework completion, and reducing dropout rates. School administrators and teachers can also encourage after-school organizations, clubs, student groups, and parent groups to incorporate physical activities into their programs and events (e.g., fundraisers, special activities).
• Collectively, the findings from this review support the National Association for Sport and Physical Education's recommendations for a comprehensive school physical activity program. 35 The results also suggest that physical education and physical activity may help advance academic performance for many students and should not hinder academic progress.  Before school (on school grounds or on the way to school): After school (on school grounds). Specify grade levels served by school: Community-based organization: Other. Please describe: or Other) ‡ The Access database will display a drop down box with the following response options: paper-pencil survey, computer assisted survey, fitness test, skill assessment, measurement device (e.g., pedometer, accelerometer, heart rate monitor), diary or journal, observation, interview, focus group, and other. § The Access database will display a drop down box with the following options: student, teacher, parent, school administration, research staff, other adult, peer, and other. ** The Access database will display check boxes with the following response options: baseline, 1 month, 2 months, 3 months, 6 months, 12 months, 18

Appendix C: Glossary of Research Design Terms
Case-study design A case study is an in-depth examination (often over time) of one or a small number of cases believed to represent a broader phenomenon; 87 it is usually, but not always, observational. 88 In this report, all case studies reported only qualitative data.

Cross-sectional study
A cross-sectional study is conducted at a single time point (often through a survey), with a sample believed to represent a cross section of the population of interest on relevant variables such as sex, age, education levels, etc. Cross-sectional studies can be used to determine whether two variables are associated but do not allow for the direct examination of the impact of time on such associations, a condition necessary to establish casuality. 88,89 Descriptive design Descriptive studies have the purpose of describing activities, events, or behaviors that have occurred in a given situation; their goal is often to create a "profile" of a phenomenon, program, or population as it exists. 90 Descriptive and inferential statistics may be used. 87 These studies differ from most quasi-experimental and experimental designs in that they do not control environments or expose subjects to different treatments and typically lack a control or comparison group, making it more difficult to account for the influence of extraneous factors. 87,88 Experimental design Experimental design is often considered the most rigorous of research designs and is frequently referred to as the gold-standard for establishing causality; in order for a study to be classified as experimental, it must include a control group and use random assignment to intervention and control groups. 91 Results may not generalize beyond the sample or conditions of the experiment. 87 Intervention An intervention is "a specific activity (or set of related activities) intended to change the knowledge, attitudes, beliefs, behaviors, or practices of individuals and populations to reduce their health risk. An intervention has distinct process and outcome objectives and a protocol outlining the steps for implementation." 92

Longitudinal study
A longitudinal study is a study conducted over time of a variable or a group of subjects, 87 unlike a cross-sectional study. By collecting data at a minimum of two distinct points in time, 90 one advantage of longitudinal studies is that they allow for the direct observation of the impact of time on variable associations, a condition necessary to establish casuality. 88 The studies in this review had a wide range of time between the initial and final data collection points; in some, final data were collected immediately following interventions, and in others, final data were collected as many as 4 years after the initial data collection.

Quasi-experimental design
A quasi-experimental design is similar to an experimental design but lacks the important characteristic of random assignment to intervention and control or comparison groups. 93 Though not considered as rigorous as an experimental design, it is often considered the next best thing for establishing causality. --

Appendix D: School-Based Physical Education Summary Matrix
Articles describing quasi-experimental or experimental designs are highlighted in the table before the matrix for each setting. Not all studies used these designs.  2 times (baseline, immediate follow-up after 10-week intervention) Paper-pencil survey (self-report academic and non-academic competence, perception of physical and social performance during intervention as measured by the Self-Perception Profile for Learning Disabled Students) 2 times (baseline, 10 weeks)

School Based Physical Education Studies Using Quasi Experimental or Experimental Design (Authors and
Teacher observation (social behavior) 3 times (baseline, 10 weeks, 3 months) Conditions: Students with learning disabilities participated in intense physical education instruction or academic enrichment.

Structure:
Intervention: Closely supervised twiceweekly, 90-minute extracurricular sessions in the pool or gymnasium designed to combine vigorous physical activity with social skills training and problem solving over a 10-week period. 26 instructors received 17 hours of training and were assigned 2 students each for the duration of the project. Control: Students received the same amount of individual attention in academic instruction (twice weekly, 90 minutes, 10 weeks).

Topics covered:
Intervention: Gym sessions focused on upper-limb coordination (gross and fine motor). Pool sessions focused on strength and visual-motor control. Social skills were developed through modeling, practice problem-solving, role-play with feedback. Control: Sessions focused on deficit skills as identified by classroom teacher.
Methods: Sessions were based on pool and gymnasium activities (one of each every week). 70 minutes of the 90 minutes were activity based.
Does the intervention group perform significantly better on self-reported academic and teacher-observed social measures than the control group after 10week intervention, controlling for gender? a • Self-perception of general intellectual ability • Self-perception of spelling competence • Self-perception of mathematical competence • Self-perception of writing competence • Self-perception of reading competence • Global self-worth Does the intervention group perform sig nificantly better on motor skills (as mea sured by Bruininks-Oseretsky Test of Motor Proficiency) and self-reported nonacademic measures than the control group after 10 week intervention, controlling for gender? • Composite motor skill scores (gross, fine, battery) • Nonacademic scores (social acceptance, athletic competence, physical appearance, behavioral conduct) 0 0 § § § Results are coded as follows: + signifies a significant positive outcome; 0 signifies no significant outcome; -signifies a significant negative outcome. Matrices may not include all outcomes described in the article; shaded outcomes are outcomes of primary interest to (and were included in) this review; additional outcomes reported here may be of interest to readers. NR = Not reported by study authors.
Indicates data collection time points. 2 times (pretest week 2 after a regular class and posttest week 3 after exercise or sport class) Conditions: Students were assigned to the coordinated exercise condition or the normal sport lesson condition.
Coordinated exercise condition (CE): Exercises selected from special forms for soccer and Munich Fitness test; groups of 4 students spent 1.75 minutes at each of 5 exercise stations.

Normal sport lesson condition (NSL):
Students exercised for 10 minutes at same intensity as CE group but without any specification on motor coordination.
Did the students in the coordinated exercise group have higher concentration and attention scores on the d2 test than control groups from pretest to posttest?
• Overall concentration and attention score + Did the students in the coordinated exercise group have a greater number of correct responses on the d2 test than control groups from pretest to posttest? Structure: Students were divided into three groups (skill, fitness, and control) that focused on developing student skill and competence level in minor games. The intervention took place over 14 weeks and was overseen by the investigators to ensure adherence.
Implementation: In the skill group, the exercise (duration and frequency) was increased to 75 minutes daily, 15 minutes of which were in the morning. The fitness group had the same.
Does participation in a fitness program improve performance on arithmetic and reading tests?
• Academic performance 0 Does participation in a fitness program improve classroom behavior (as observed by teacher)?
• Classroom behavior + Paper-pencil survey using Conners´ questionnaire (teachers' and parents' conception of children's attention ability and impulse control)

Conditions:
Intervention: Students received physical education lessons 5 days per week. Comparison: Standard physical education lessons 2 days per week.

Methods:
Intervention: 3 regular school physical education lessons per week plus different local sports clubs gave physical activity lessons for 2 more lessons every week, for a total of 5 lessons of motor skills training and physical activity per week. If needed (for students deemed motor deficient), 1 extra lesson of MUGI motor training per week was provided.
Do students with good motor skills have better attention than students with deficits in motor skills (as observed by teachers and parents)?
• Attention • Impulse control + + Do students in intervention groups have better attention than students in comparison group (as observed by teachers and parents)? Structure: Two procedures: walking for a specific duration and receiving a timed mathematical computation test. Testing was conducted over a 5-week period.
Methods: Subjects walked around the perimeter of a regulation basketball court at a monitored moderate intensity (120 to 145 beats per minute). Walking duration was systematically ordered for the 20, 30, and 40 minutes. Walking occurred early morning (8:30 a.m.), before lunch (11:50 a.m.) or afternoon (2:20 p.m.). A mathematics test was given at the end of the specified duration of activity, and subjects had 90 seconds to complete the task.

No intervention
Do students who are exposed to recess during the school day have better classroom behavior, as reported by the teacher, than students who do not have recess?
• Classroom behavior (overall classroom) + Do students who participate in directed physical activity have significantly higher scores on the Woodcock-Johnson Test of Concentration than students who participate in typical classroom activity?
• Concentration (4th-grade students) + ***Results are coded as follows: + signifies a significant positive outcome; 0 signifies no significant outcome; -signifies a significant negative outcome. Matrices may not include all outcomes described in the article; shaded outcomes are outcomes of primary interest to (and were included in) this review; additional outcomes reported here may be of interest to readers. NR=Not reported by study authors.
Indicates data collection time points. Conditions: Students were exposed to 1 of 2 conditions, a shorter confinement period in the classroom (2.5 hours) and a longer confinement period in the classroom (3 hours) before recess.
Structure: Each child was exposed to both conditions by counterbalancing the order in which the whole class experienced confinement across the 14-week observation period.
Method: Students were observed before recess, during recess, and after recess. Conditions: Children in 2nd and 4th grades were presented with either a male-preferred or female-preferred task immediately before recess and immediately after recess. Children were expected to sit quietly in their seats while the teacher read a story with either a male or female main character.
Structure: 4 days per week recess timing was manipulated: 2 days per week students went to recess at 10:00 a.m. (short deprivation period) and 2 days per week students went to recess at 10:30 a.m. (long deprivation period).
Method: Students were observed before recess, during recess, and after recess.
Are children more attentive to classroom tasks after recess than before recess (as observed by research staff)?  Implementation: Within-subjects design, the same students learned word-pairs in each of the 2 conditons. † † † † Results are coded as follows: + signifies a significant positive outcome; 0 signifies no significant outcome; -signifies a significant negative outcome. Matrices may not include all outcomes described in the article; shaded outcomes are outcomes of primary interest to (and were included in) this review; additional outcomes reported here may be of interest to readers. NR = Not reported by study authors. NA = Not applicable.
Indicates data collection time points. Data collection method and time points: Structured and timed observation (on-task behavior of 3rd-and 4th-grade students only-defined as verbal or motor behavior that followed class rules and was appropriate to the learning situation) Daily over a 12-week period (pre-and postintervention) Measurement devicepedometers (number of steps taken in all K-4th-grade classes) 5 days in a week (all children in one grade per week) Structure: Students in all K-4th-grade classrooms (3 classes per grade) in one school participated in "Energizers." These are classroom-based physical activities that last approximately 10 minutes, integrate grade-appropriate learning materials, involve no equipment, and require little teacher preparation.
Training: Before the study, classroom teachers attended a 45-minute training session where they were taught how to lead students through Energizers activities. Training included information about the childhood obesity epidemic.
Implementation: Teachers were each asked to lead one 10-minute activity per day for 12 weeks.
Does participation in Energizers increase on-task behavior in school (based on researcher observation)?
• On-task behavior (3rd-and 4thgrade students) + Does participation in Energizers increase physical activity in school (based on pedometer counts)?
• Physical activity + Paper-pencil survey (student satisfaction with exercise and relaxation program, stress levels) 2 times (baseline, 1 month) Paper-pencil survey (teacher perception of student concentration and stress levels) 2 times (baseline, 1 month) Structure: Relaxation/exercise episodes occurred twice daily for 4 weeks, immediately following the morning break and after the lunch break. The program took 5 to 10 minutes and consisted of a combination of stretching exercises and relaxation exercises.
Does the intervention affect student ability to concentrate (as reported by teachers)?
• Concentration + Does the intervention reduce classroom noise level (as measured by research staff)?
• Noise levels + Does the intervention affect student selfreported stress level?
• Stress level 0 Paper-pencil survey (self-report of physical activity times per week and rate how compares to peers, self-report of school functioning and grades, selfconcept, school attendance, well-being, church participation, relationship with parents, friends' use of cigarettes, alcohol, and drugs) 2 times (baseline, 3 months)

Name: First Choice program
Training: Staff were trained as Fitness Leaders in a 40-hour course with a written and practicum certification exam. Content of the staff course focused on fitness assessment techniques, goal setting and exercise prescription, exercise leadership, teaching skills, safety, and foundations content.
Structure: Program was implemented in 22 settings across the state (including 1 juvenile correctional facility, 6 drug treatment facilities, 4 junior high schools, 2 senior high schools, and 9 neighborhood or community centers). This evaluation was conducted in 6 sites (1 high school, 2 junior high schools, and 3 community sites).
Method: There were 24 different modules to teach physical fitness as a life skill, focusing on self-assessment, goal setting, exercise and nutrition planning, and self-reward motivation through exercise classes, educational modules, discussion modules, and individual exercise program maintenance.
What were the effects of the First Choice fitness program on participating youth academic outcomes and risk factors?

Site 1
• Grades (self-report) • School attendance • Self-concept Site 2 • Grades (self-report) • School attendance • Self-concept Site 3 • Grades (self-report) • School attendance • Self-concept Site 4 • Grades (self-report) • School attendance • Self-concept Site 5 • Grades (self-report) • School attendance • Self-concept Site 6 • Grades (self-report) • School attendance • Self-concept 0 0 + 0 0 + 0 0 + 0 0 + 0 0 + + + + ‡ ‡ ‡ ‡ Results are coded as follows: + signifies a significant positive outcome; 0 signifies no significant outcome; -signifies a significant negative outcome. Matrices may not include all outcomes described in the article; shaded outcomes are outcomes of primary interest to (and were included in) this review; additional outcomes reported here may be of interest to readers. NR=Not reported by study authors.
Indicates data collection time points. Additional analyses were conducted and more years of participation were associated with higher grades, a more positive attitude toward school, and higher academic aspirations.
Is participation in school-based extracurricular activities associated with indicators of adolescent adjustment (after adjustment for demographics) across mutiple years of study?
• Data collection method and time points: Paper-pencil survey and face-to-face interview [self reported grades, school value, participation in school activities (e.g., clubs, student government, athletic or sports teams at school, and participation in Summer or after-school recreational programs), selfesteem, psychological resiliency, depression, prosocial peers, risky behaviors, race, and gender] 3 times (7th, 8th, and 11th grades)

No intervention
Did participation in 7th-grade school sports improve academic outcomes at 8th grade?
• Self-reported GPA • School value Additional analyses showed a significant effect in 8th grade for race on resiliency and SES on depression; no significant results by gender.
Did participation in 7th-grade out-of-school recreation improve academic outcomes at 8th grade?
Did participation in 7th-grade school clubs improve academic outcomes at 8th grade? (participation in clubs, sports or prosocial activities) 3 times (baseline, 3 years, and 5 years-8th grade, 11th grade, and 1 year out of high school) Paper-pencil survey (self-reported grades) 2 times (baseline and 3 years at 8th grade and 11th grade) Paper-pencil survey (depression and psychological health, educational expectations, alcohol and drug use, civic engagement) 3 (baseline, 3 years, and 5 years at 8th grade, 11th grade, and 1 year out of high school) • Externalizing behavior + Did participation in high school sports improve academic performance outcomes 1 year out of high school (controlling for demographics, motivations, and educational expectations)?
• School completion • Self-esteem • Depression + 0 0 Additional analyses showed no significant results by gender, race, or income for these outcomes. The Association Between School-Based Physical Activity, Including Physical Education, and Academic Performance | 79