Measuring intervention outcomes for children requires the use of both standardized and individualized measures. Goals are individualized ways of determining what is important for the child, guiding service delivery, and measuring responses to intervention. In a systematic review of interventions for children with cerebral palsy (CP), goal-directed training was effective for improving motor activities, function, and self-care.1 The Individuals with Disability Education Improvement Act2 requires the use of goals to measure student outcomes. In recent years, use of a dichotomous achieved or did not achieve a goal has been exchanged for more comprehensive systems to determine a range of goal achievement, such as the Canadian Occupational Performance Measure3 and goal attainment scaling (GAS),4,5 which include different levels of achievement.
In the United States, children with disabilities who receive special education are also eligible to receive related services, including physical therapy.2 The effectiveness and efficiency of physical therapy services in school-based practice has not been adequately addressed.6 Studies of school-based physical therapy practice have focused on retrospective record reviews of outcomes7; documentation of goal achievement and functional assessment8; consideration of social and learning opportunities9; surveys of service delivery and supports provided10–12; therapists' perceptions of school-based practices13; and goal development.14 These studies do not address what interventions were provided and the relationship of services to student outcomes.
The scarcity of research regarding school-based physical therapy practice and complexity of service delivery in schools impedes the execution of controlled trial studies. Therefore, to initially study school-based practice, a practice-based evidence (PBE) research design can offer insights into the effectiveness of natural service provision through systematic documentation and analysis of a comprehensive set of student, service, and outcome variables.15 The objective of this approach is to explore the relationships among student characteristics, the services provided, and change in the outcomes. Causal relationships between services and outcomes cannot be determined with a PBE study design; however, the identified relationships provide preliminary evidence to support key practices and future experimental designs.
Previous research using a PBE design investigated student individualized outcomes16 and the relationship of school-based physical therapy services to standardized outcomes on the School Function Assessment (SFA) measure.17 Students achieved individualized outcomes addressed by school-based physical therapy, and goal attainment did not differ based on students' gross motor ability.16 Younger students had higher goal attainment than older students in 2 goal categories: primary (the goal identified to have been most addressed by physical therapy services) and recreation/fitness.16 Services and service-related factors associated with better standardized outcomes on the SFA included mobility, sensory, motor learning, aerobic conditioning, functional strengthening, and playground access interventions, and positive behavioral involvement of the student during therapy.
As school-based physical therapy is primarily directed to supporting students' Individualized education program objectives,2 the purpose of this investigation was to determine the relationship of physical therapy services to goal achievement using GAS. We collected data on school-based physical therapy services (types, activities, and interventions), students' behavioral involvement during therapy sessions, and students' achievement of individualized goals during 6 months of a school year.
This study, Physical Therapy related Child Outcomes in the Schools (PT COUNTS), used a PBE research design.15,18 Institutional Review Board approvals were obtained from the investigators' universities and individual school districts across the United States. The measures and procedures for this study have been previously described.16,17,19
School-based physical therapists, with at least 1 year of school-based experience, were recruited from across the United States. The inclusion criteria for students denoted that the students were in kindergarten through the sixth grade (ages 5-12 years), received special education, and received physical therapy services at least monthly. Exclusion criteria were students who (1) had a progressive disability; (2) planned to move; (3) had major surgery planned; or (4) had school absences greater than 30% in the previous year. Therapists selected the students who met these criteria from their workloads and provided the investigators with a coded number list. If the list contained 6 or fewer students, all of the students were invited to participate. If the list was longer than 6, then 6 numbers were randomly selected and the students aligned with those numbers were invited to participate. Additional random selection of numbers continued, if needed, until each therapist had recruited a minimal of 1 student but not more than 6. Signed consent forms were obtained from therapists and parents.
The final number of participating physical therapists was 109 (105 females, 4 males), mean age 46 years (SD 9 years 2 months), and 296 students with disabilities (130 females, 166 males), and mean age 7 years 4 months (SD 2 years), from 59 school systems in 28 states. The number of participants exceeded sample size calculation for the study.18 Participant characteristics have been previously presented16 and are reproduced with permission from the publisher in Table 1.
The Gross Motor Function Classification System (GMFCS)20 is a 5-level system to classify children with CP. Classification is made based on current gross motor function in daily activities, with emphasis on mobility and sitting. Children in GMFCS level I have the highest motor function. The GMFCS has evidence of content, construct, and discriminative validity, and interrater reliability for children with CP.20,21 Although the GMFCS was developed for children with CP, we wanted an indication of functional motor ability level for all students. We believe that the GMFCS was the most appropriate option to describe the student's gross motor function.
GAS22 is an individualized, goal-based, criterion-referenced outcome measure of change in performance of a behavior. Criteria for each level of change must be measurable and meaningful. The individual's performance at baseline is assigned a value of −2 and the expected outcome following intervention is assigned a value of 0. A value of −1 represents progress toward the expected outcome and values of +1 and +2 represent outcomes that are possible but exceed expectations. A value of −3 can be used to reflect that the client has regressed in function on the goal below baseline. GAS has demonstrated content validity, interrater and intrarater reliability, and responsiveness in studies of children with a range of conditions,8,23 and has been more responsive to changes in performance than standardized measures.24 The GAS score was used to establish the outcome groups for this study.
The School-Physical Therapy Interventions for Pediatrics (S-PTIP)25 is a data collection system developed for students receiving school-based physical therapy services. Using a previous version developed for children in multiple practice settings (Pediatric Physical Therapy Intervention Activities form),26 we made edits to reflect school-based practice. The S-PTIP is used to report minutes provided in types of service delivery (eg, individual, group, and within a school activity) and time spent on behalf of the student (eg, consultation, collaboration, and documentation). Student participation in therapy, behavioral involvement, is rated on a 0-point (not at all conducive to session objectives) to 6-point (extremely conducive to session objectives) scale. The form includes 13 types of activities that might be used during therapy sessions and the time spent in each type is recorded in 5-minute increments. The form includes 79 possible interventions that might be used during therapy sessions and a count is documented for each intervention used. The interventions are grouped under the categories of neuromuscular, musculoskeletal, cardiopulmonary, integumentary, orthoses, mobility assistive devices, mobility, positioning, equipment, sensory, educational, assessment, and other. Face and content validity and intrarater reliability of the S-PTIP have been supported.27,28 The manual, including definitions of all activities and interventions, is available at https://www.uky.edu/chs/academic-programs/department-rehabilitation-sciences/physical-therapy/pt-counts.
Details of the procedures have been reported previously.18 Therapists completed online training modules on GAS and the S-PTIP and passed a posttraining assessment with a score of at least 80%. Therapists received individualized feedback and could repeat the assessments if they did not pass them the first time. Therapists then viewed 2 videos of students receiving physical therapy at school and completed S-PTIPs. Therapists had to meet a minimum criterion of 70% agreement with the investigator's scoring. This minimal level of agreement was established due to the difficulty in interpreting the purpose for an intervention from observation alone. All differences in scoring between the researcher and the therapists were discussed to verify understanding of completion of the S-PTIP.
During the study, therapists followed up 1 to 6 students. Therapists reported students' GMFCS level and parents reported their child's age. Therapists identified students' individualized education program (IEP) goals related to participation in school activities that were supported by their physical therapy services. Depending on the students' IEPs, therapists were asked to identify goals reflecting the following outcome areas of interest: posture/mobility, recreation/fitness, self-care, and academics. Using a checklist, 4 of the researchers worked in pairs to systematically review the GAS goals to ensure they met the criterion for GAS format regarding measurability, equidistance, difficulty, and clinical relevance.4 Within the pairs, 1 of the researchers, who was not the investigator for that region, completed the checklist. The other researcher, the investigator for the region, confirmed the findings and communicated the information to the therapists. Of the 498 GAS goals reviewed for this study, only 19 goals required substantial revisions by the therapists. Through a consensus process, the research team confirmed the classification of these goals into the 4 outcome areas. Within each outcome area, students only had 1 goal. If a student had goals across more than 1 outcome area identified for this study, therapists were asked to select the 1 goal that was most pertinent to the student's participation in school and physical therapy services addressed most often. This goal was referred to as the primary goal.
For 6 months (20 weeks excluding winter and spring breaks), therapists completed the S-PTIP weekly for each student. Mid-way through the study, researchers completed a procedural fidelity analysis, previously described in the PT COUNTS methodology article.18 Each researcher observed 30% of therapists in their region during a school-based physical therapy session. The researcher and the therapist independently completed the S-PTIP and discussed differences. Given that the researchers did not know the students and had to draw their own conclusions of the purpose of various observed activities, agreement between researchers and therapists was expected to be moderate. The number of discrepancies between researcher and therapist S-PTIP forms ranged from 1 to 9. There were a total of 13 possible activities and 79 possible interventions; therefore, this range of discrepancies was judged by the researchers to reflect acceptable agreement.
After 6 months of S-PTIP data collection, each therapist determined their students' goal attainment. Goal attainment was verified by at least 2 IEP team members for 78% of primary goals, 81% of posture/mobility goals, 71% of the recreation/fitness goals, 94% of self-care goals, and 89% of academic goals.
GAS scores and S-PTIP data were entered into REDCap (Research Electronic Data Capture, developed by Vanderbilt University, Nashville, Tennessee). Because it was previously reported that, on average, students achieved goal expectation (GAS score of 0),16 3 logistic regressions (primary, posture/mobility, and recreation/fitness goals) were used to determine the association of physical therapy services to exceeding goal expectation (GAS scores of +1 and +2), adjusting for students' GMFCS level (I, II-III, IV-V) and age group (5-7, 8-12). Physical therapy service variables tested in the model were first selected based on services that differed between students who exceeded expected goal attainment and students who did not exceed expected goal attainment (independent t test, α level of 0.15 due to exploratory nature of the study). When examining the types of activity variables, we grouped them based on similarities into prefunctional, sitting/standing/transitions, classroom activity, mobility, physical education/recreation, self-care, and communication. For the purpose of including types of intervention variables in the logistic regressions, counts of interventions were summed across several types and included as 1 variable (total counts for halls/stairs/doors training, and playground access). Prior to conducting the logistic regressions, linear regressions were fitted to discard predictors with high variance inflation factors. An α level of 0.05 was set for statistical significance of the logistic regression models. Nagelkerke R2 values were calculated.
We were unable to conduct regression analyses with self-care and academic goals because the number of goals was too small. We confirmed that exceeding goal attainment in these areas did not differ based on age nor GMFCS level. For these goal areas, differences in service variables between the students who exceeded and did not exceed expected goal attainment were examined with an independent t test, an α level of 0.05.
Differences in physical therapy between students who exceeded and did not exceed expected goal attainment for each goal area are shown in Table 2. The odds ratios for school physical therapy and student characteristic variables in the final regression models are given in Table 3.
Forty percent of students exceeded goal expectation for their primary goal. The overall regression model, including if group services were received; minutes of services on behalf of student and in self-care activities; total counts for balance, motor learning, and functional strength interventions; age group; and gross motor function group, was significant (χ2 = 22.66, P < .001). The model explained 10% of the variance for exceeding or not exceeding goal expectation. Even though the model was significant, no individual service variable was significantly associated with students' exceeding goal expectation.
Forty-one percent of students who had a posture/mobility goal exceeded goal expectation. The overall regression model including minutes of services on behalf of the student and self-care activities, total counts for cognitive and behavioral interventions, age group, and gross motor function group was significant (χ2 = 20.84, P < .001). The model explained 13% of the variance for exceeding or not exceeding goal expectation. More minutes of services on behalf of the student and self-care activities (P < .05) were associated with exceeding goal expectation. An increase in 20 minutes of services on behalf of the student (1 minute per week) increased the odds of exceeding goal expectation by 4%. An increase in 20 self-care activity minutes (1 minute per week) increased the odds of exceeding goal expectation by 37%. Use of cognitive and behavioral training interventions was associated with not exceeding goal expectation (P < .05). Every increase in the combined cognitive and behavioral interventions decreased the odds of exceeding goal expectation by 11%.
Forty percent of students who had a recreation goal exceeded goal expectation. The overall regression model including if group services were received, minutes of services on behalf of the student, total counts for playground access and functional strength interventions, total counts for cognitive and behavioral training interventions, age group, and gross motor function group was significant (χ2 = 21.09, P = .002). The model explained 16% of the variance for exceeding goal expectation. Greater use of functional strength, mobility for playground access, cognitive, and behavioral interventions was associated with exceeding goal expectation (P < .05). Every increase in the combined functional strength and mobility for playground access interventions increased the odds of exceeding goal expectation by 5.9%. Every increase in the combined cognitive and behavioral interventions increased the odds of exceeding goal expectation by 8.8%.
Forty percent of students who had a self-care goal exceeded goal expectation. Group comparative analyses revealed that, for self-care goals, the students who exceeded goal expectation received less physical therapy time with no other students present (P = .05), less documentation time (P < .02), and more neuromuscular interventions (P < .04) than students who did not exceed goal expectation.
Twenty-seven percent of students who had an academic goal exceeded goal expectation. Group comparative analyses revealed that, for academic goals, the students who exceeded goal expectation received more mobility interventions than students who did not exceed goal expectation (P < .03).
Only a few physical therapy services were related with more positive student individualized goal outcomes. Even though we examined 88 service variables and accounted for student age and gross motor function level in our analyses, the final models explained only 10% to 16% of the variance in exceeding or not exceeding goal expectation. Physical therapy services, student age, and gross motor function level explained a higher percentage of the variance (11%-33%) in standardized SFA outcomes.17 We were not surprised by these findings, as conceptually, we believe it is more difficult to explain individualized outcomes with group data. The finding that no single physical therapy service was significantly related with students' exceeding goal expectation on their primary goal was also not surprising, as the focus of the primary goals varied among students. Nevertheless, the findings of the physical therapy services related with exceeding goal achievement for posture/mobility and recreation/fitness goals suggest key service types, activities, and interventions for school-based physical therapy practice.
The finding that more minutes of services on behalf of the student was associated with exceeding goal expectation for posture/mobility goals provides support for a collaborative team approach. Through coordination, collaboration, and documentation, therapists share knowledge with other team members who support student goals throughout the day. Team collaboration has long been recognized as a key element in school-based practice29 and a necessary strategy to support inclusion of students in educational environments. This finding provides evidence for physical therapists to advocate for “indirect” time to educate and support other team members and staff in strategies to promote the student's postural stability and mobility throughout daily school routines.
The relationship between more minutes of self-care activities and exceeding goal expectation for posture/mobility goals suggests that practice of self-care activities affords effective motor learning challenges for postural stability and movement. This finding is particularly significant given that on average students only received less than 1 minute per week in self-care activities.19
To support students' recreation and fitness, the findings highlight the need to address functional strength and focus on mobility specific to playground access. Our study adds to the body of evidence on school-based physical therapy interventions6 by showcasing in an educational setting, the type of strengthening (ie, the student working against resistance/gravity or eccentrically within functional activity), associated with recreation and fitness outcomes. Noteworthy, mobility for playground access was related to both individual student outcomes of exceeding recreation/fitness goal expectation and standardized outcomes via the SFA recreational movement scores.17
The findings that sitting, standing, movement transition, and mobility activities were not related to exceeding goal expectation for posture/mobility goals, and PE and recreation activities were not related to exceeding goal expectation for recreation/fitness goals appears counterintuitive. However, these findings may be explained by the fact that therapists, for all students in the study, provided these activities for the greatest amount of time.19 These activities appear central to school-based physical therapy practice, but the time spent in these activities was not different between students who exceeded and did not exceed goal expectation. Perhaps other aspects of these activities, such as degree of challenge or specificity of alignment with the student's goal, might distinguish goal attainment.
Cognitive and Behavioral Training Interventions
Greater use of cognitive and behavioral training interventions was related with not exceeding goal expectation for posture/mobility goals but exceeding expectation for recreation/fitness goals. As posture/mobility goals relate to foundational gross motor function, greater use of cognitive and behavioral interventions and not exceeding goal expectation may signify that the students had greater needs related to planning and executing movement, including awareness, judgment, and behavioral regulation. However, recreation/fitness goals may represent more complex gross motor skills, often in the social context of PE and playground games, and greater use of these interventions was beneficial for skill acquisition.
The findings related to differences in services between students who exceeded expectation for self-care and academic goals need to be considered cautiously due to the small number of goals. For achievement of self-care goals, having other students present may provide positive peer modeling and observational learning opportunities and greater use of neuromuscular interventions highlights the importance of balance and motor learning strategies to optimize self-care performance. We are unable to reasonably interpret the finding that students who received lower amounts of documentation time exceeded self-care goal expectation. Perhaps this finding reflects a characteristic of the student or therapist. Unexpectedly, we did not find a difference in time spent on self-care activities between those who exceeded and did not exceed self-care goal expectation. Consistent with goal-directed service philosophy, on average, both groups of students with self-care goals received more minutes of self-care activities (approximately 60 minutes/6 months) than the average received by all students (15 minutes/6 months).19
The finding that students who exceeded academic goal expectation received more mobility interventions during physical therapy services adds to the body of knowledge on the influence of physical activity for cognitive outcomes.30 It is important for physical therapists to consider that providing opportunities for mobility might enhance students' ability to attend, focus, and learn classroom material.
Due to the limitations of the PBE design (ie, inability to determine causality), findings of this study need to be substantiated with further research that examines the intervention strategies under controlled conditions with consideration for quality of service provision. In this study, goals were dependent on what was written in the student's IEP and because on average, students achieved their goals (GAS score = 0), the study explored services associated with exceeding goal expectation (GAS scores +1, +2). For the examination of the students' primary goal, we acknowledge that therapists may have had bias in selecting the goal. The analyses accounted for student age, gross motor function, and behavioral involvement in therapy but did not control for other student characteristics such as cognitive function. The GMFCS was used to account for the gross motor function of students with varied diagnoses, only some of which had CP. Therapists in this study were experienced, with the majority employed by the schools, and services may not reflect those of therapists with less experience or in other employment arrangements. Since the S-PTIP was completed weekly, therapists might have forgotten some aspects of the interventions that were provided that week. As some therapists followed up more than 1 student, service data do not fully meet the statistical assumption of independence. How a therapist served 1 student may have influenced how they served another student.
This study adds to our knowledge about which physical therapy school-based practice interventions are related with better individualized student outcomes. A limited number of physical therapy services were related with exceeding goal expectation. For posture/mobility goals, providing interventions during self-care activities and services on behalf of students such as collaboration with team members may provide a meaningful motor learning challenge and assist others to support the student. For recreation/fitness goals, emphasizing the combination of functional strengthening and mobility interventions may offer students the opportunity to access the playground environment and the body function needed to participate in school games and sports. As physical therapy interventions related with better outcomes differed across the goal areas, we recommend that therapists plan interventions guided by the focus of student goals. Further research is warranted to understand collectively how child factors such as motivation, environmental factors such as teacher and family support, and therapy services influence child outcomes.
Additional statistical assistance was also provided by Heather Bush, PhD, and Catherine Starnes, PhD, from the Department of Biostatistics at the University of Kentucky. We thank the participating school districts, therapists, and students, whose involvement made this study possible.
1. Novak I, Mcintyre S, Morgan C, et al A systematic review of interventions for children with cerebral palsy: state of the evidence. Dev Med Child Neurol. 2013;55:885–910.
2. Public Law 108-446. Individuals with Disabilities Education Improvement Act of 2004, 20 U.S.C. http://www.copyright.gov/legislation/pl108-446.pdf
. Published 2004.
3. Law M, Baptisite S, Carswell A, McColl M, Polatajko H, Pollack N. Canadian Occupational Performance Measure. 3rd ed. Toronto, Canada: CAOT Publications ACE; 1998. http://www.thecopm.ca/
4. McDougall J, King G. Goal Attainment Scaling
: Description, Utility, and Applications in Pediatric Therapy Services. 2nd ed. London, Canada: Thames Valley Children's Centre; 2007.
5. Brewer K, Pollock N, Wright FV. Addressing the challenges of collaborative goal setting with children and their families. Phys Occup Ther Pediat. 2014;34(2):138–152.
6. Effgen SK, McEwen IR. Review of selected physical therapy interventions for school age children with disabilities. Phys Ther Rev. 2008;13(5):297–312.
7. Stuberg W, DeJong SL. Program evaluation of physical therapy as an early intervention and related service in special education. Pediatr Phys Ther. 2007;19:121–127.
8. King G, McDougall J, Palisano RJ, Gritzan J, Tucker MA. Goal attainment scaling
: its use in evaluating pediatric therapy programs. Phys Occup Ther Pediat. 1999;19(2):31–52.
9. Fallang B, Øien I, Østensjo S, Gulbrandsen L. Micro-processes in school and learning activities at school generate exclusion for children with disabilities. Scand J Disabil Res. 2017;19(3):1–12. doi:10.1080/15017419.2016.1276472.
10. Effgen SK, Kaminker MK. Nationwide survey of school-based physical therapy
practice. Pediatr Phys Ther. 2014;26(4):394–403.
11. Kaminker MK, Chiarello LA, O'Neil ME, Dichter CG. Decision making for physical therapy service delivery in schools: a nationwide survey of pediatric physical therapists. Phys Ther. 2004;84(10):919–933.
12. Soim A, Lamb M, Campbell K, et al A cross-sectional study of school experiences of boys with Duchenne and Becker muscular dystrophy. Phys Disabil: Ed Relat Serv. 2016;35(2):1–22.
13. Holt S, Kuperstein J, Effgen SK. School-based physical therapists' perceptions of school-based practices. Phys Occup Ther Pediatr. 2015;35(5):381–395.
14. Wynarczuk KD, Chiarello LA, Gohrband CL. Goal development practices of physical therapists working in educational environments. Phys Occup Ther Pediatr. 2017;37(4):425–443.
15. Horn SD, DeJong G, Deutscher D. Practice-based evidence
research in rehabilitation: an alternative to randomized controlled trials and traditional observational studies. Arch Phys Med Rehabil. 2012;93:S127–S137.
16. Chiarello LA, Effgen SK, Jeffries L, McCoy SW, Bush H. Student outcomes of school-based physical therapy
as measured by Goal Attainment Scaling
. Pediatr Phys Ther. 2016;28(3):277–284.
17. McCoy S, Effgen S, Chiarello L, Jeffries L, Tezanos AGV. Relationship of school-based physical therapy
services to student outcomes on the School Function Assessment. Dev Med Child Neurol. 2018;60(11):1140–1148.
18. Effgen SK, McCoy SW, Chiarello LA, Jeffries LM, Bush H. Physical therapy-related child outcomes in school: an example of practice-based evidence
methodology. Pediatr Phys Ther. 2016;28(1):47–56.
19. Jeffries L, McCoy S, Effgen S, Chiarello L, Tezanos AGV. Description of the services, activities and interventions within school-based physical therapy
practice across the United States. Phys Ther. 2019;99(1):98–108. doi:10.1093/ptj/pzy123.
20. Palisano RJ, Rosenbaum P, Bartlett D, Livingston MH. Content validity of the expanded and revised Gross Motor Function Classification System. Dev Med Child Neurol. 2008;50:744–750.
21. Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi B. Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol. 1997;39:214–223.
22. Kiresuk TJ, Smith A, Cardillo JE. Goal Attainment Scaling
: Applications, Theory, and Measurement. Hillsdale, NJ: Lawrence Erlbaum Associates; 1994.
23. Steenbeek D, Ketelar M, Galama K, Gorter JW. Goal attainment scaling
in paediatric rehabilitation: a critical review of the literature. Dev Med Child Neurol. 2007;49:550–556.
24. Steenbeek D, Gorter JW, Ketelar M, Galama K, Lindeman E. Responsiveness of goal attainment scaling
in comparison to 2 standardized measures in outcome evaluation of children with cerebral palsy. Clin Rehabil. 2011;25(12):1128–1139.
25. McCoy SW, Jeffries L, Effgen S, et al School Physical Therapy Interventions for Pediatrics (S-PTIP) Manual and Forms, Version 4. http://www.mc.uky.edu/healthsciences/grants/ptcounts
. Published February 2014.
26. Hashimoto M, McCoy SW. Validity of an activity-based data form developed to reflect the interventions used by pediatric physical therapists. Pediatr Phys Ther. 2009;21:53–61.
27. McCoy SW, Linn M. Validity of the school-physical therapy interventions for pediatrics data system for use in clinical improvement design studies. Pediatr Phys Ther. 2011;23:121–122.
28. Effgen SK, McCoy S, Jeffries L, Chiarello L, Smarr J, Bush H, Smith T. Reliability of the school-physical therapy interventions for pediatrics data system. Pediatr Phys Ther. 2013;26(1):118–119.
29. Rainforth B, York-Barr J. Collaborative Teams for Students With Severe Disabilities: Integrating Therapy and Educational Services. Baltimore, MD: Paul H. Brookes; 1997.
30. Fedewa AL, Ahn S. The effects of physical activity and physical fitness on children's achievement and cognitive outcomes: a meta-analysis. Res Q Exerc Sport. 2011;82(3):521–535.