Abstract
Background:
Adolescent athletes are experiencing an increased number of concussions. There is currently a debate within the medical community regarding the most effective and safe treatment approach for this population, post-concussion. Interventions currently range from cognitive and physical rest to various types of physical activity, including aerobic exercise. While there are systematic reviews that focus on rest as the main intervention, there are no other systematic reviews that focus on the effects of aerobic exercise on concussion recovery in adolescent athletes.
Purpose:
The aim of this systematic review and meta-analysis was to investigate the effectiveness of aerobic exercise on concussion recovery for adolescent athletes compared to an alternate intervention.
Study Design:
Systematic Review and Meta-Analysis
Methods:
A computer-based search (population: adolescent athletes with concussions, intervention: aerobic exercise, comparator: non-aerobic interventions, outcome: symptom severity and recovery time) was performed. Databases including PubMed, CINAHL, SPORTdiscus, ProQuest, and Scopus were searched up to December 2019 for randomized controlled trials published since 1965. A hand search of relevant articles and exploration of grey literature was also completed. Data were extracted for the following information: interventions prescribed, outcome measures, and overall results of the study. A meta-analysis was performed for aerobic interventions using standardized mean difference as the summary measure of effect.
Results:
Five studies, which all held a moderate to low risk of bias according to the PEDro scale, met the inclusion criteria for this systematic review and meta-analysis. Overall results favored aerobic exercise for both acute and prolonged recovery symptoms as demonstrated by a decrease in symptom severity and improved recovery time. The meta-analysis revealed a moderate effect size in favor of the intervention group (SMD: 0.51, CI: 0.02, 0.81, p=0.00) when looking at the three outcome measures combined: Post-Concussion Symptom Scale, Post-Concussion Symptom Inventory, and recovery time.
Conclusion:
The results of this systematic review and meta-analysis indicate that there is currently moderately significant evidence in support of implementing an aerobic exercise program for adolescent athletes with both acute and prolonged recovery concussion symptoms. Additional higher quality studies are needed to continue to study the effectiveness of aerobic exercise in post-concussion treatment of adolescents.
Level of Evidence:
1a
Keywords: adolescents, aerobic exercise, athletes, concussion, movement system
INTRODUCTION
A concussion or mild traumatic brain injury (mTBI) is a chemical change that occurs in the brain due to a direct blow to the head or a hit to the body that causes the head and brain to move rapidly back and forth.1 Signs and symptoms often appear soon after the injury and may include confusion, personality changes, brief loss of consciousness, dizziness, nausea, vomiting, light or noise sensitivity, headache, and feeling foggy.1 A common population afflicted by concussions are adolescent athletes. According to the Centers for Disease Control and Prevention (CDC), nearly 330,000 children, ages 19 and younger, were treated in emergency departments for sport and recreation-related concussions and traumatic brain injuries in 2012. The rate of emergency department visits for sport-related concussions more than doubled in the adolescent athlete population between 2001-2012.2
Current standard treatments for concussion include both physical and cognitive rest, as well as gradual progression to participation in activities.3 However, the activities and their dosage remain a debate within the medical community. While there are multiple guidelines regarding adult athletes and concussion recovery, there is a lack of standardization regarding interventions within the adolescent athlete post-concussive population. Physical activity, including aerobic exercise, is beginning to be explored as a standard intervention. Aerobic exercise has been shown to such as decreasing risk for Type II diabetes, cancer, hypertension, obesity, depression, and osteoporosis.4 Therefore, aerobic exercise should be explored for brain health, including its effects on concussion recovery.
Despite preventative measures being taken, concussions still occur. Therefore, it is imperative to explore evidenced-based interventions for post-concussive rehabilitation. Previous systematic reviews (SRs) have primarily focused on rest and cognitive testing as the main interventions to promote concussion recovery.5 However, there is evidence that suggests active treatment strategies are favorable in the process of concussive recovery for athletes.5
The adolescent athlete may be the most vulnerable to long-term consequences of concussions.6 Exploring active recovery as a strategy as for reducing post-concussive symptoms and accelerating return to sport-specific activities is warranted. As a result, the purpose of this systematic review (SR) and meta-analysis (MA) was to investigate the effectiveness of aerobic exercise on concussion recovery for adolescent athletes compared to an alternate intervention.
METHODS
Protocol and Registration
The Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) was utilized to guide this review. The PRISMA checklist is a 27-item list designed to improve reporting for authors of both systematic reviews and meta-analyses.7 Research topic, pilot title, and protocol were registered online on the PROSPERO International Prospective Register of Systematic Reviews database with registration number: CRD42019121764. PROSPERO provides an extensive list of registered systematic reviews to attempt to prevent duplication and reduce reporting bias.
Eligibility Criteria
Eligibility for this review included randomized controlled trials (RCTs) that compared aerobic exercise to alternate interventions (i.e. stretching, manual therapy, vestibular training, rest) for adolescent athletes with a clinically diagnosed concussion. Studies that compared aerobic exercise to control groups were also included. Included articles were limited to RCTs in an attempt to improve the quality of the data analyzed. Aerobic exercise was defined as activities that elevated heart rate above resting heart rate or increased Rate of Perceived Exertion (RPE) to a moderate level or higher. Concussion was defined using the Centers for Disease Control and Prevention statement, which classifies it as a type of traumatic brain injury caused by a blow to the head or a hit to the body, causing the head and brain to move rapidly back and forth.2 Studies published prior to 1965 were excluded since the MESH term “brain concussion” was first introduced in 1965. Non-English studies were excluded as well as studies that investigated the effectiveness of a one-time intervention.
Search Strategy, Databases Utilized, and Study Selection
The following databases were searched: PubMed, CINAHL, ProQuest, Scopus, and SPORTdiscus. The search strategy included three main classifications: concussions, adolescent athletes, and aerobic exercise. Synonymous terms were defined and used with the main classification terms to mark the search. Similar terms were used in each database while using database-specific terms, such as Medical Subject Heading (MESH) terms when relevant. An example of a specific search strategy can be found in Appendix 1. A hand search of relevant articles and inspection of grey literature, including clinicaltrials.gov and opengrey.eu, were also done for any studies that were not identified in the database search. The initial search was conducted in January 2019, and the final search was performed in December 2019.
Titles, abstracts, and full texts of studies retrieved using the search strategy were screened independently by two authors to identify studies that potentially met the inclusion criteria or belonged within the exclusion criteria. Any disagreement between the two authors over the eligibility of particular studies that could not be resolved by consensus was resolved with a third reviewer's vote. A Cohen's unweighted Kappa was calculated for agreement during both title and abstract selection and full-text selection. A Kappa of less than or equal to 0 is considered no agreement; 0.01 to 0.2, none to slight agreement; 0.21 to 0.40, fair; 0.41 to 0.60, moderate; 0.61 to 0.8, substantial; and more than 0.8, almost perfect agreement.8
Data Extraction and Analysis
Data from the final full text articles were extracted by two authors, which were then cross checked and agreed upon. Data extracted included sample size, mean age of participants, and interventions. Outcome data extracted from the articles included the following outcome measures: The Post-Concussion Symptom Scale (PCSS), the Post-Concussion Symptom Inventory (PCSI), and recovery time.
The PCSS is a 21-item patient-rated scale. Symptom severity is rated using a 7-point Likert scale with 0 meaning “no symptom” and 6 meaning “severe”. Moderate test-retest reliability has been reported with an intra-class correlation coefficient (ICC) of 0.62.9,10 The PCSI is a self-report form that focuses on symptoms in the cognitive, emotional, sleep, and physical domains. For individuals between the ages of 5-7, there are 13 items to rank. For individuals ages 8-12, there are 25 items to rank. For individuals ages 13-18, there are 26 items to rank.11 Moderate to high interrater reliability has been reported for the child report (two separate reports that assess ages 5 to 7 and 8 to 12), as well as moderate to strong test-retest reliability (ICC = 0.65-0.89).11 However, it is reported that the strongest data for adolescents with concussions supports using the PCSS over the PCSI.9 Recovery time is defined as being the number of days to recovery since the date of injury.12 Recovery is defined as symptom resolution to normal, confirmed by a normal physical examination.12
Means and standard deviations for aerobic exercise and alternate interventions were also synthesized in each study. Meta-analyses were made using Comprehensive Meta-Analysis Version 3 where applicable. All meta-analyses were done using standardized mean difference (SMD) as the summary measure of effect. SMD with 95% confidence intervals (CI) were used. I2 statistics were calculated in order to determine the level of heterogeneity between included studies. The I2 statistic is more useful than the Q test, which only indicates the presence versus absence of heterogeneity.13 Percentages used by Higgins and Thompson14 were utilized to quantify the magnitude of heterogeneity: 25% = low, 50% = medium, 75% = high heterogeneity. Utilizing the scale, if I2 was < 50%, a fixed effects model was used, and if the I2 was > 50%, a random effects model was used. The I2 value for this meta-analysis was 61.34%, therefore, a random effects model was used. Interpretation of effect size used Cohen's criteria for pooled estimates.15 An SMD of 0.2-0.4 = small effect; 0.5-0.7 = moderate effect; 0.8 and higher = large effect.16
Quality Assessment
Two authors independently assessed the risk of bias in all included studies using the PEDro Scale. Disagreements in the quality of the studies were discussed between the two authors. If a consensus was not determined, a third author made the ultimate decision. The PEDro Scale is a criteria-based scale used to the assess the quality of randomized controlled trials. There are 11 criteria used in this scale regarding topics such as eligibility criteria, randomization, blinding, and statistical reporting.17 A higher score represents a higher quality study.17
RESULTS
Study Selection
The results of the initial search produced 2,442 titles. After duplicates were removed, 2,175 titles were found. Titles and abstracts were then screened by two authors, and 2,133 articles were excluded. Hand search and grey literature searches were then completed and no additional randomized control trials were found. After title and abstract assessment, 42 articles were deemed appropriate for full-text review, with an observed к = 0.46, confidence interval (CI) = 0.30 - 0.62, indicating moderate agreement. Non-randomized control trials were excluded resulting in five articles being included in the systematic review and meta-analysis as they satisfied the inclusion criteria. After screening the accepted full text articles, the observed к = 0.90, CI = 0.69 - 1, exhibiting almost perfect agreement. After deciding on the full text articles to include к = 1, CI = 1 - 1, demonstrating perfect agreement. There was no need to call on a third author for a tie-breaker decision. Details of the study selection process can be found in Figure 1.
Figure 1.
PRISMA flow diagram.
n = number of studies, RCT = randomized controlled trial
Study Characteristics
Cycling was used as the aerobic intervention for three out of the five included studies while the remaining two used treadmill training. Rest was used as a part of usual care for the control groups in two out of the five included studies. Patient education and medication was used along with rest for one of those two usual care control groups. Stretching was used for the control groups in three of the five included studies. The intervention and control groups were matched by age in all studies. In total, 179 subjects were used as participants; 98 were males, and 82 were females. Acute symptoms versus prolonged recovery were defined by either less than 4 weeks (acute) or greater than four weeks (prolonged recovery).18 Other pertinent study characteristics can be found in Table 1.
Table 1.
Study characteristics. N = number of participants, SD = standard deviation, PCSS = post concussion symptom scale, PCSI = post-concussion symptom inventory.
| Intervention Group | Control Group | |||||||
|---|---|---|---|---|---|---|---|---|
| Study | Intervention | Control | Outcome Measure | Acute vs. Prolonged Recovery Symptoms | N | Mean Age (SD) | N | Mean Age (SD) |
| Kurowski et al 21 | Aerobic Training (cycling) | Stretching | PCSI | Prolonged Recovery | 12 (5 males, 7 females) | 15.22 (1.37) | 14 (8 males, 6 females) | 15.50 (1.80) |
| Yuan et al22 | Aerobic Training (cycling) | Stretching | PCSI | Prolonged Recovery | 8 (4 males, 4 females) | 15.04 (1.35) | 9 (6 males, 3 females) | 15.48 (2.08) |
| Chan et al19 | Active Rehabilitation (Submaximal Aerobic Training) | Treatment As Usual (Education, Medication) | PCSS | Prolonged Recovery | 10 (4 males, 6 females) | 15.90 (1.66) | 9 (1 male, 8 females) | 15.10 (1.42) |
| Micay et al20 | Exercise (cycling) | Usual Care (six-stage progression of activity) | PCSS, Recovery Time | Acute | 8 (8 males, 0 females) | 15.80 (1.20) | 7 (7 males, 0 females) | 15.60 (1.00) |
| Leddy et al12 | Treadmill | Stretching | PCSS, Recovery Time | Acute | 52 (28 males, 24 females) | 15.30 (1.60) | 51 (27 males, 24 females) | 15.40 (1.70) |
Quality Assessment
Two reviewers assessed the risk of bias of the five articles included using the PEDro scale. Each article received a PEDro Scale score (see Table 2). An unweighted kappa was calculated to evaluate the level of agreement between the two authors (к = 0.82, CI = 0.66 - 0.99, which is almost perfect agreement). None of the articles met Criterion 5 or Criterion 6 due to an inability to blind participants and researchers to specific interventions. All of the articles met Criterion 9 and Criterion 10 because outcomes were reported on at least one of the groups in all articles. Thus, the included articles were of good quality with regard to bias as they all scored 6 or 7 out of 10 items (Criterion 1 was excluded).17 These articles were deemed acceptable by all group members for use in this study.
Table 2.
Quality assessment using PEDro scale . 1. Eligibility criteria were specified. 2. Subjects were randomly allocated to groups. 3. Allocation was concealed. 4. The groups were similar at baseline regarding the most important prognostic indicators. 5. There was blinding of all subjects. 6. There was blinding of all therapists who administered the therapy. 7. There was blinding of all assessors who measured at least one key outcome. 8. Measures of at least one key outcome were obtained from more than 85% of the subjects initially allocated to groups. 9. All subjects for whom outcome measures were available received the treatment or control condition as allocated or, where this was not the case, data for at least one key outcomes was analyzed by “intention to treat.” 10. The results of between-group statistical comparisons are reported for at least one key outcome. 11. The study provides both point measures and measures of variability for at least one key outcome.
| Chan19 | Leddy12 | Micay20 | Kurowski21 | Yuan22 | |
|---|---|---|---|---|---|
| Criterion 1 | Yes | Yes | Yes | Yes | No |
| Criterion 2 | Yes | Yes | Yes | Yes | Yes |
| Criterion 3 | No | No | No | No | Yes |
| Criterion 4 | Yes | Yes | Yes | Yes | Yes |
| Criterion 5 | No | No | No | No | No |
| Criterion 6 | No | No | No | No | No |
| Criterion 7 | Yes | Yes | No | Yes | No |
| Criterion 8 | Yes | Yes | Yes | Yes | Yes |
| Criterion 9 | Yes | Yes | Yes | Yes | Yes |
| Criterion 10 | Yes | Yes | Yes | Yes | Yes |
| Criterion 11 | Yes | Yes | Yes | Yes | No |
| Score | 7 | 7 | 6 | 7 | 6 |
Outcomes
PCSS
Two of the five studies19,20 used in the systematic review evaluated symptom severity using the PCSS (see Table 3, Figure 2). However, only one of the studies that used the PCSS, Chan et al,19 was included in the meta-analysis as Micay et al20 reported results as a total point value of change for symptom severity rather than providing baseline and final measurement scores. Results demonstrated a decrease in symptom severity in prolonged recovery symptoms with aerobic exercise with a moderate effect size (SMD: 0.622, CI: -0.30,1.54, p=0.19).
Table 3.
Outcomes. PCSS = post-concussion symptom scale, PCSI = post-concussion symptom inventory, SD = standard deviation.
*The PCSS data for the Micay et al study is reported as a total point value of change for symptom severity rather than baseline and final measurement scores.
| Intervention Mean Baseline Outcome Measure (SD) | Intervention Mean Final Outcome Measure (SD) | Control Mean Baseline Outcome Measure (SD) | Control Mean Final Outcome Measure (SD) | Effect Size (Standardized Mean Difference) | 95% Confidence Interval | |
|---|---|---|---|---|---|---|
| PCSS | ||||||
| Chan | 51.50 (27.80) | 25.00 (19.40) | 56.90 (31.00) | 40.30 (29.40) | 0.62 | (−0.30-1.54) |
| Micay* | 18.80 (4.90) | 10.00 (6.10) | 1.58 | (0.42-2.74) | ||
| PCSI | ||||||
| Kurowski | 37.40 (25.01) | 4.17 (7.36) | 40.27 (27.25) | 15.93 (20.18) | 0.76 | (−0.05-1.57) |
| Yuan | 28.38 (20.94) | 5.00 (8.99) | 42.00 (27.84) | 19.00 (23.81) | 0.76 | (−0.23-1.75) |
| Recovery Time (Days) | ||||||
| Leddy | 0.01(p-value) | |||||
| Micay | 36.10 (18.50) | 29.60 (15.80) | −0.38 | (−1.40-0.65) |
Figure 2.
Standardized difference in means in individual studies for PCSS.
Squares represent study-specific findings, and diamond represents summary estimates of fixed/random effects meta-analysis.
PCSS = post-concussion symptom scale, Std diff in means = standardized difference in means, CI = confidence interval.
PCSI
Two of the five studies21,22 used in the systematic review and meta-analysis evaluated symptom severity using the PCSI (see Table 3, Figure 3). Results demonstrated a decrease in symptom severity with aerobic exercise with a moderate effect size (SMD: 0.75, CI: 0.13, 1.37, p=0.02). Both studies analyzed prolonged recovery symptoms.
Figure 3.
Standardized difference in means in individual studies for PCSI.
Squares represent study-specific findings, and diamond represents summary estimates of fixed/random effects meta-analysis.
PCSI = post-concussion symptom inventory, Std diff in means = standardized difference in means, CI = confidence interval.
Recovery Time
Two of the five studies12,20 used in the systematic review and meta-analysis evaluated recovery time (in days) (see Table 3, Figure 4). Leddy et al12 evaluated recovery time by the number of days from the time of injury to the third consecutive day in which the individual's PCSS score fell below 7. Micay et al20 determined recovery time by reviewing each participant's electronic medical record for their return to play status (measured by days). Both studies evaluated acute symptoms. Results demonstrated a decrease in recovery time with aerobic exercise with a small effect size (SMD: 0.20, CI: -0.64, 1.05, p=0.64).
Figure 4.
Standardized difference in means in individual studies for recovery time.
Squares represent study-specific findings, and diamond represents summary estimates of fixed/random effects meta-analysis.
Std diff in means = standardized difference in means, CI = confidence interval.
Overall
When looking at three outcome measures combined, PCSS, PCSI, and recovery time, there is a moderate effect size in favor of the aerobic exercise intervention group (SMD: 0.51, CI: 0.02, 0.81, p=0.00) (see Table 3, Figure 5).
Figure 5.
Standardized difference in means in individual studies for overall outcomes.
Squares represent study-specific findings, and diamond represents summary estimates of fixed/random effects meta-analysis.
PCSS = post-concussion symptom scale, PCSI = post-concussion symptom inventory, Std diff in means = standardized difference in means, CI = confidence interval.
DISCUSSION
The majority of the studies reported significant post-concussion improvements with aerobic exercise,19,21,22 suggesting that aerobic exercise may be beneficial for concussion recovery in adolescent athletes.
There were limitations with the included studies that have the potential to affect the overall conclusion of this review. Neither the participants nor researchers were blinded to the interventions being completed in the included studies for this SR/MA. This may have created a “placebo” effect, meaning that reductions in the patient's symptoms could have occurred due to the participants knowing which intervention they were receiving and having a positive outlook on the effects of that intervention. The studies selected targeted the population of concussed adolescent athletes, which produced a limited number of RCTs. Finally, while the sample sizes within the studies were small; the meta-analysis was completed on a sample of 90.
There was variability in the intervention that the control groups received in the individual studies. This makes between study comparisons less precise and more challenging to develop an understanding of the true influence of interventions. Finally, there was variability in reporting results. The majority of articles reported their data as baseline and final measurement scores. However, one study20 reported PCSS results as a total point value of change for symptom severity. This discrepancy in reporting resulted in the exclusion of this study's20 PCSS findings from the MA. If all studies uniformly reported results, then more studies could be compared objectively, which would strengthen the conclusions drawn from this data.
There were several strengths to the way in which the SR was conducted. First, the PRISMA guidelines were incorporated in the reporting of the findings to ensure transparency.7 Secondly, only RCTs were included. An RCT is an optimal research design to minimize bias and is well suited to answer research questions pertaining to the effect an intervention has on a population.23 Finally, the review process proved to be reliable. This ensured that the appropriate studies were included in this SR/MA.
While the Pedro scale is widely used for quality assessment, there are not published cut-off scores to classify a study as high, moderate or low quality. In the absence of this cut-off scored, the guideline was followed17 of a higher score indicating a higher quality study. This could be construed as a weakness.
While other systematic reviews have studied the effect of aerobic exercise on concussion, none have focused on an adolescent population. An SR published in 2017, which did not explore a specific age group, suggests that after a brief (24-48 hour) period of physical and cognitive rest, patients should be encouraged to gradually increase activity. In addition, the abovementioned SR also supports the use of cervical and vestibular rehab as indicated.24 Another SR/MA without an age-specific population of interest had similar findings that physical exercise (which included stretching as well as aerobic activity) improves symptoms in patients with concussion.25
One possible explanation for the significant results found in the current SR/MA is related to the physiology of the brain. Aerobic exercise may positively affect various aspects of brain healing, including improvement in cerebral blood flow, blood oxygen extraction, autonomic control pathway, and neuroplasticity. All of these may encourage structural reorganization of the brain that can positively impact tissue healing.26
A second prospective explanation for the post-concussive improvement seen in adolescents treated with aerobic exercise deals with the biopsychosocial model, which is a clinical model that suggests that pathological experiences are affected by multiple variables. One component of the model focuses on the relationship between mental and physical health, suggesting that subjective experiences are not confined to physiology.27 It may be that once adolescents are given permission to participate in physical activity after an injury, negative illness beliefs, such as fear avoidance behaviors, are countered.28 Adolescents can then reintegrate into their social and recreational activities, which can lead to both physical and mental health benefits.28
Implications for Further Research
Future research should focus on return to play or return to learn protocols to guide practice for efficient and safe returns. This could include time until symptoms subside, time until the athlete is medically cleared, and which interventions provided the fastest and safest return. These items should be researched using a randomized controlled trial design comparing an aerobic exercise experimental group to a resting or stretching control group. Future research needs to be completed on adolescent athletes directly after experiencing concussion symptoms. This could be within the first few days following diagnosis of the concussion. Randomized controlled trials are limited in this area as not many studies choose to target athletes in the acute phase after developing concussion-like symptoms.
CONCLUSION
The results of this systematic review and meta-analysis indicate that there is currently moderately significant evidence in support of implementing an aerobic exercise program for adolescent athletes with both acute and prolonged recovery concussion symptoms. Additional higher quality studies are needed to continue to study the effectiveness of aerobic exercise in post-concussion treatment of adolescents.
Appendix 1.
Search Strategy.
| Pubmed | (((((((“exercise”[MeSH Terms] OR exercise[Text Word]) OR “cardiovascular system”[MeSH Terms]) OR cardiovascular system[Text Word]) OR “cardiorespiratory fitness”[MeSH Terms]) OR cardiorespiratory fitness[Text Word]) OR “aerobic exercise”[Text Word]) AND ((((((“athletes”[MeSH Terms] OR athletes[Text Word]) OR “adolescent”[MeSH Terms]) OR adolescent[Text Word]) OR “adolescent athlete”[Text Word]) OR adolescence[Text Word]) OR teenagers[Text Word])) AND ((“brain concussion”[MeSH Terms] OR brain concussion[Text Word]) OR concussion[Text Word]) |
REFERENCES
- 1.What is a concussion? Centers for Disease Control and Prevention. https://www.cdc.gov/headsup/basics/concussion_whatis.html. Updated January 31, 2017.
- 2.Traumatic brain injury & concussion. Centers for Disease Control and Prevention. https://www.cdc.gov/traumaticbraininjury/get_the_facts.html. Published April 27, 2017.
- 3.Broglio SP Collins MW Williams RM Mucha A Kontos AP. Current and emerging rehabilitation for concussion. Clin Sports Med. 2015;34(2):213-231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Warburton DE. Health benefits of physical activity: the evidence. CMAJ. 2006;174(6):801-809. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.McLeod TC Lewis JH Whelihan K Bacon CE. Rest and return to activity after sport-related concussion: a systematic review of the literature. J Athl Train. 2017;52(3):262-287. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kimbler DE Murphy M Dhandapani KM. Concussion and the adolescent athlete. J Neurosci Nurs. 2011;43(6):286-290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.PRISMA. Preferred reporting items for systematic reviews and meta-analyses. http://www.prisma-statement.org/. Updated October 2015.
- 8.McHugh Mary. Interrater reliability: the kappa statistic. Biochemia Medica. October 2012:276-282. [PMC free article] [PubMed] [Google Scholar]
- 9.Graham R. Sports-related concussions in youth: improving the science, changing the culture. Washington, D.C.: The National Academies Press; 2014. [PubMed] [Google Scholar]
- 10.Merritt VC Bradson ML Meyer JE Arnett PA. Evaluating the test-retest reliability of symptom indices associated with the ImPACT post-concussion symptom scale (PCSS). J Clin Exp Neuropsychol. 2018;40(4):377-388. [DOI] [PubMed] [Google Scholar]
- 11.Sady MD Vaughan CG Gioia GA. Psychometric characteristics of the postconcussion symptom inventory in children and adolescents. Arch Clin Neuropsychol. 2014;29(4):348-363. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Leddy JJ Haider MN Ellis MJ, et al. Early subthreshold aerobic exercise for sport-related concussion. JAMA Pediatrics. 2019;173(4):319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Huedo-Medina T Sanchez-Meca J Martin-Martinez F. Assessing heterogeneity in meta-analysis: Q statistic or I2 index? The center for Health, Intervention, and Prevention (CHIP): Digital Commons; 2006. [DOI] [PubMed] [Google Scholar]
- 14.Higgins J Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002; 21(11):1539-58. [DOI] [PubMed] [Google Scholar]
- 15.Landis J Koch G. The measurement of observer agreement for categorical data. Biometrics. 1977;33(1):159-174. [PubMed] [Google Scholar]
- 16.Faraone SV. Interpreting estimates of treatment effects: implications for managed care. P & T: A Peer-Reviewed Journal for Formulary Management. December 2008. [PMC free article] [PubMed] [Google Scholar]
- 17.PEDro - Physiotherapy Evidence Database. PEDro Scale. https://www.pedro.org.au/wp-content/uploads/PEDro_scale.pdf.
- 18.McCrory P Meeuwisse W Dvorak J, et al. : Consensus statement on concussion in sport: the 5th international conference on concussion in sport held in Berlin, October 2016. Br J Sports Med. 2018;51:838–847. [DOI] [PubMed] [Google Scholar]
- 19.Chan C Iverson GL Purtzki J, et al. Safety of active rehabilitation for persistent symptoms after pediatric sport-related concussion: a randomized controlled trial. trial. Arch Phys Med Rehabil. 2018;99(2):242-249. [DOI] [PubMed] [Google Scholar]
- 20.Micay R Richards D Hutchison MG. Feasibility of a postacute structured aerobic exercise intervention following sport concussion in symptomatic adolescents: a randomised controlled study. BMJ Open Sport Exerc Med. 2018;4(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kurowski BG Hugentobler J Quatman-Yates C, et al. Aerobic exercise for adolescents with prolonged symptoms after mild traumatic brain injury. J Head Trauma Rehabil. 2017;32(2):79-89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Yuan W Wade SL Quatman-Yates C Hugentobler JA Gubanich PJ Kurowski BG. Structural connectivity related to persistent symptoms after mild TBI in adolescents and response to aerobic training. J Head Trauma Rehabil. 2017;32(6):378-384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Jewell DV. Guide to evidence-based physical therapist Practice 4th ed. Burlington, MA: Jones and Bartlett Learning; 2018. [Google Scholar]
- 24.Schneider KJ Leddy JL Guskiewicz KM, et al. Rest and treatment/rehabilitation following sport-related concussion: a systematic review. Br J Sports Med. 2017;51(12):930-934. [DOI] [PubMed] [Google Scholar]
- 25.Lal A Kolakowsky-Hayner SA Ghajar J, et al. The effect of physical exercise after concussion: a systematic review and meta-analysis. Am J Sports Med. 2018;46(3):743-752. [DOI] [PubMed] [Google Scholar]
- 26.Cotman C. Exercise: a behavioral intervention to enhance brain health and plasticity. Trends Neurosci. 2002;25(6):295-301. [DOI] [PubMed] [Google Scholar]
- 27.Borrell-Carrio F. The Biopsychosocial model 25 years later: principles, practice, and scientific inquiry. Ann Fam Med. 2004;2(6):576-582. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Difazio M Silverberg ND Kirkwood MW, et al. Prolonged activity restriction after concussion. Clin Pediat. 2015;55(5):443-451. [DOI] [PubMed] [Google Scholar]