Effectiveness of Non-Pharmacological Therapy on Physical Symptoms in Patients With Persistent Concussion Symptoms: A Systematic Review

Abstract

This systematic review provides a comprehensive overview on the effectiveness of rehabilitation on physical symptoms in patients of all ages with persistent concussion symptoms. PubMed, MEDLINE^®, Cochrane library, Cumulative Index to Nursing and Allied Health Literature (CINAHL), and Embase were searched from January 1, 2012 to September 1, 2023 using terms related to physical post-concussion symptoms. Eligible articles were critically appraised using the Scottish Intercollegiate Guidelines Network (SIGN) and the Quality Assessment Tool. The Grading of Recommendations Assessment, Development, and Evaluation system was applied to rate the quality of evidence. Thirty-two articles were included. Preliminary evidence suggests that transcranial magnetic stimulation improves symptoms in adults, specifically headaches. Young adults reported a significant decrease in physical symptoms following sub-symptom aerobic training as well as cervical spine manual therapy. Tentatively, adults demonstrated improvements in headache symptoms following neurofeedback sessions, and progressive muscle relaxation resulted in a decrease in monthly headaches. Multimodal therapy in adults produced significant change in physical symptoms when compared with usual care. However, no further reduction in physical symptoms was observed when adult patients received a program of care that afforded cervicovestibular rehabilitation with symptom-limited exercise compared with a symptom-limited exercise program alone. Cognitive behavioral therapy demonstrated inconsistent findings for its effects on physical symptoms, specifically headaches. Veterans had a significant change in post-concussive symptoms, specifically headaches, following 3-month use of an interactive smartphone application as compared with standard care. Finally, in a pediatric population, the use of melatonin did not produce any changes in physical persistent concussion symptoms as compared with placebo. Preliminary evidence suggests that various forms of rehabilitative therapies can improve persistent physical concussive symptoms. However, given the methodological limitations in the majority of trials, the results need to be interpreted with caution.

Introduction

Disabilities secondary to traumatic brain injury (TBI) are a major social and economic burden. Patients in Ontario seeking tertiary care for persisting symptoms following mild traumatic brain injury (mTBI) incur health costs of > $110,000,000 annually.¹ The majority of patients who sustain an mTBI recover; however, timelines are highly variable.² Additionally, recent findings suggest that up to 82% of patients report at least one symptom 6–12 months following an mTBI.³ The implications for this are significant for the person, the healthcare system, and society.

Two recent systematic reviews have examined the state of the literature on treatment of persistent concussion symptoms. Rytter and coworkers examined non-pharmacological treatments for persistent (> 4 weeks) symptoms in mTBI and concussion samples.⁴ They reported that there is weak evidence for the effectiveness of non-pharmacological interventions.⁴ The second review by Makdissi and coworkers examined various treatment modalities in patients with sport-related concussion with persistent symptoms (> 10 days).⁵ However, only two trials examined the efficacy of physical therapy on persistent symptoms.

Ultimately, there exists a gap regarding the state of literature on the effectiveness of rehabilitative interventions for physical dominant persistent concussion symptoms, such as headaches, dizziness, phonophobia, nausea, and visual problems. Prior well-designed reviews^4,5 have reported on rehabilitative treatments for persistent concussion symptoms; however, apart from exercise therapy, the reviews' discussion focuses on cognitive-based interventions. This is likely the result of the lack of well-designed clinical trials investigating interventions directed to patient's physical symptoms. Given that headaches are the most common symptom reported by patients following an mTBI, there is a major need to specifically examine the efficacy of interdisciplinary treatments specifically targeting physical symptoms.⁶

The present review was developed to better understand the effects of rehabilitation in concussion patients experiencing persistent symptoms. Specifically, the objective was to determine the effectiveness, as measured by clinical outcomes, of non-pharmacological rehabilitation in the management of patients of all ages experiencing persistent physically dominant concussion symptoms, as compared with other interventions such as placebo/sham interventions, or no interventions.

Methods

Registration

This review protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO) on June 12, 2022 (CRD42022341100).

Population

This review examined the best available evidence regarding the effectiveness of non-pharmacological conservative treatments on physical symptoms in patients of all ages experiencing persistent concussion symptoms. The definition used for persistent concussion symptoms was taken from Tator and coworkers, which indicates that persistent symptoms reflect a failure to recover within 1 month.⁷

Evaluation of assessment

We restricted our review to articles assessing the effectiveness of non-pharmacological rehabilitation on patients with persistent concussion symptoms. “Rehabilitation” was defined as per the World Health Organization (WHO) as a “set of interventions designed to optimize function and reduce disability in individuals with health conditions in interaction with their environment.”⁸

Outcomes

This review focused on either self-rated recovery outcomes (global perceived recovery) or changes in clinical outcomes (e.g., pain and symptom severity scores). Outcomes related to psychological variables (e.g., depression, anxiety, stress scores), administrative outcomes, or outcomes related to costs were excluded from this review.

Inclusion and exclusion criteria

To be included in our review, articles had to fulfill the following inclusion criteria: (1) be written in English, (2) be published from January 1, 2012 to September 1, 2023, (3) be published in a peer-reviewed journal, (4) describe clinical trials examining the effectiveness of treatment in patients of all ages experiencing persistent concussion symptoms, and (5) discuss a patient population who had symptoms beyond 4 weeks as per the definition of post-concussion syndrome by Tator and coworkers. Articles examining mild, moderate, and severe TBI were included only if it was possible to separate data on the populations.

Articles fulfilling any of the following criteria were excluded: (2) publication types including guidelines, letters, editorials, commentaries, unpublished manuscripts, dissertations, government reports, books and book chapters, conference proceedings, meeting abstracts, lectures and addresses, consensus development statements, and guideline statements; (2) study designs including systematic and non-systematic reviews, and case studies; (3) cadaveric or animal trials; and (4) articles solely targeting individuals with cognitive or emotional symptoms. Articles that examined pharmacological treatments or any trial solely examining the effectiveness of any treatments directed to mood/affect or cognitive post-concussion symptoms were excluded.

Data sources

We systematically searched the following electronic databases: MEDLINE^®, PubMed, Cochrane Central Register of Controlled Trials, and Cumulative Index to Nursing and Allied Health Literature (CINAHL). Search terms consisted of subject headings specific to each database (e.g., MeSH in MEDLINE and PubMed) and free text words relevant to concussion, post-concussion, headache, therapy, treatment, and rehabilitation (Supplementary Appendix I). In addition, the reference lists of included articles and related systematic reviews were screened to identify any article that we may have missed using our search strategy.

Study selection

Articles were screened for relevance in different phases. The first phase involved screening titles and abstracts for relevance and possibly relevant citations based on the aforementioned inclusion and exclusion criteria. The full text of potentially relevant articles was then independently screened by two reviewers (N.M. and S.G.) to determine final article selection. Discrepancies were resolved by consensus between the two reviewers. Disagreements were resolved by a third reviewer.

Risk of bias

Two independent reviewers assessed risk of bias with the Scottish Intercollegiate Guidelines Network (SIGN) criteria for randomized controlled trials (RCTs), cohort studies, and case-control studies (Table 1).⁹ The tool includes seven items divided into six domains of bias: selection, performance, detection, attrition, reporting, and any other bias. A summary assessment of risk of bias within a study was then made by grading a study as being of high, acceptable, or low quality using the SIGN criteria. For case series, single-armed observational studies, and retrospective observational studies the Effective Public Health Practice Project Quality Assessment Tool was utilized (Table 2). The tool evaluates studies under the headings of selection, design, confounders, blinding, data collection, and withdrawals/dropouts, generating a final global rating for quality assessment of weak, moderate, or strong.¹⁰ Consensus between reviewers was reached through discussion. An independent third reviewer (S.K.) was included to resolve disagreements when consensus could not be reached.

Table 1.

Risk of Bias Rating for Controlled Trials Based on the Scottish Intercollegiate Guidelines Network (SIGN) Criteria

Author, year	Research question	Randomization	Concealment	Blinding of patient / treatment provider	Blinding of outcome assessor or data analyst	Similarity at baseline	Similarities between arms	Outcome measurement	Percent drop-out	Intention to treat	Multiple sites
Moussavi et al. 2019¹⁹	Y	Y	Y	Y	Y	Y	Y	Y	19%	U	N/A
Leung et al. 2016¹⁷	Y	Y	Y	U	U	Y	Y	Y	18%	U	N/A
Choi et al. 2018¹⁸	Y	Y	Y	Y	Y	Y	Y	Y	0%	U	N/A
Leddy et al. 2013²⁶	Y	Y	N	N	N	N	N	N	20%	U	N/A
Kurowski et al. 2017²⁷	Y	Y	Y	N	Y	Y	Y	Y	14%	Y	N/A
Potter et al. 2016²⁹	Y	Y	U	N	N	Y	Y	Y	4%	Y	Y
Schneider et al. 2014³²	Y	Y	Y	Y	Y	Y	Y	Y	7%	Y	N/A
Leung et al. 2018²¹	Y	Y	Y	N	U	Y	N	Y	34%	U	N/A
Stilling et al. 2019²²	Y	Y	Y	Y	Y	Y	Y	Y	0%	U	N/A
Thastum et al. 2019⁴⁰	Y	Y	Y	N	N	Y	Y	Y	16%	Y	N/A
Rytter et al. 2019³⁹	Y	Y	Y	Y	Y	Y	Y	Y	19%	Y	N/A
Kjeldgaard et al. 2014³⁰	Y	Y	N	N	N	Y	Y	Y	23%	U	N/A
Barlow et al. 2020⁴⁸	Y	Y	Y	Y	Y	Y	Y	Y	8%	Y	N/A
McGeary et al. 2022³¹	Y	Y	Y	Y	Y	Y	Y	Y	58.6%	Y	N/A
Belanger et al. 2022⁴⁷	Y	Y	N	N	N	Y	Y	Y	23.2%	Y	N/A
Langevin et al. 2022⁴³	Y	Y	N	N	Y	Y	Y	Y	13.4%	Y	N/A

Y, yes; N, no; U, unclear; N/A: not applicable.

Table 2.

Risk of Bias of Non-Controlled Trials Based on Either Scottish Intercollegiate Guidelines Network (SIGN) Criteria or the Quality Assessment Tool for Quantitative Studies (QAT)

Author, year	Questionnaire utilized	Research question	Individuals representative / percentage selected	Study design	Confounders	Blinding of outcome assessor	Blinding of participants	Data collection reliable / valid	Percentage of drop-outs	Intention to treat	Consistency of intervention	Participants receiving unintended interventions
Jensen et al. 1990³³	SIGN	Y	Y	Cohort	U	N	U	U	17%	N	U	U
Walker et al. 2008⁴⁴	QAT	Y	Y	Single arm Cohort	U	N	N	Y	28%	U	Y	U
Leung et al. 2016¹⁷	QAT	Y	Y	Prospective case series	Y	N	N/A	Y	U	N/A	Y	U
Koski et al. 2015²³	QAT	Y	Y	Single arm Cohort	U	N	N	Y	20%	U	Y	U
Huang et al. 2017²⁵	QAT	Y	Y	Single arm Cohort	Y	N	N	Y	0%	Y	Y	Y
Hugentobler et al. 2015⁴²	QAT	Y	Y	Retrospective case series	Y	N	N/A	Y	0%	U	U	U
Leddy et al. 2010²⁸	QAT	Y	Y	Single arm cohort	Y	N	N	U	8%	U	Y	U
Gagnon et al. 2009⁴¹	QAT	Y	Y	Retrospective case series	Y	N	N/A	N	0%	N	U	U
Hammerle et al. 2019³⁴	QAT	Y	Y	Retrospective cohort	N	Y	N	U	0%	U	Y	U
Kennedy et al. 2017³⁵	QAT	Y	Y	Single arm cohort	U	Y	Y	N	0%	U	N	U
Marshall et al. 2015³⁶	QAT	Y	Y	Retrospective case series	Y	N	N/A	N	0%	N	U	U
Miller et al. 2020²⁴	QAT	Y	Y	Single arm cohort	U	N	N	N	12.5%	N	Y	U
Usmani et al. 2021⁴⁶	QAT	Y	Y	Single arm cohort	N	N	N	U	19%	Y	Y	U
Kennedy et al. 2021³⁷	QAT	Y	Y	Prospective case series	Y	N	N	Y	0%	N	U	U
Wetzler et al. 2017³⁸	QAT	Y	Y	Prospective case series	Y	Y	N/A	U	0%	U	N	U
Elbogen et al. 2021⁴⁵	QAT	Y	Y	Single arm cohort	U	N	U	Y	12%	U	U	Y

Y, yes; N, no; U, unclear; N/A, not applicable.

Data extraction

The reported data were extracted in a two-step process. Firstly, the lead author (N.M.) extracted the data from accepted articles to build the evidence tables. Author name, year of publication, number of participants, description of intervention, (where applicable) description of control/comparison interventions, subjects under investigation, primary outcomes, and key findings including: (1) summary statistics for each group, and (2) effect estimate and precision (when reported). Second, the two reviewers (N.M. and S.G.) independently reviewed the accuracy of the extracted data by referring to the original articles.

Quality of evidence assessment

In order to rate the quality of evidence, the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) system was applied to the articles' primary outcome measure.¹¹ The assessment was based on the five downgrading factors (risk of bias, inconsistency, indirectness, imprecision, and publication bias) and three upgrading factors (large effect, dose response, and all plausible confounding). In brief, the two reviewers independently assessed the quality of evidence level. Assessments were then compared during a formal meeting during which consensus with regard to the strength of evidence was reached (high, moderate, low, or very low). The quality of evidence level, coupled with the risk of bias score, was taken into consideration in the Results/Discussion section.¹²

Study homogeneity was assessed via visually examining the data extraction tables. Secondary to the considerable variability in subjects, therapeutic interventions, and definitions of post-concussion syndrome or persistent symptoms, meta-analysis was not possible. A narrative synthesis was conducted with articles grouped according to therapeutic intervention utilized.

Statistical analysis

The agreements among the two reviewers for the screening of articles were computed and the kappa coefficient (ĸ) reported.^13,14 The percentage agreement was calculated for classifying controlled trials as high quality, acceptable, or low quality and for classifying non-controlled trials as strong, moderate, or weak following independent critical appraisal. We considered conducting a random effect analysis if the trials were homogeneous. However, because of the clinical heterogeneity of the articles (population, treatment rendered, use of control groups or comparison, outcome measures used), no meta-analysis was performed.¹⁵

Reporting

This review complies with the Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (Supplementary Appendix II).¹⁶

Results

Study selection

Overall, 2015 articles were screened for eligibility and 241 duplicates were excluded. Titles and abstracts of 1773 articles were screened for eligibility and 52 articles were recovered for full text evaluation (Fig. 1). Following full-text review, a total of 20 articles were excluded for ineligible study design or publication type (7 citations), intervention other than rehabilitation (4 citations), inclusion of participants outside the scope of the review (5 citations), and primary outcome outside the scope of the review (3 citations). One RCT that was deemed to be at high risk of bias in accordance with the SIGN checklist (1 citation) was also excluded. Therefore, 32 articles were considered acceptable and were included in the analysis.

FIG. 1.

Identification and selection of articles on the effectiveness of non-pharmacological conservative treatments on somatic symptoms in patients with post-concussion syndrome.

The inter-rater agreement for screening of articles was k = 0.74. The agreement for classifying articles as high quality, acceptable, or low quality was 89%. Consensus was achieved without the need for a third reviewer.

Study characteristics

Various forms of rehabilitation were identified, including repeated transcranial magnetic stimulation (rTMS),^{17

–25} aerobic exercise,^26
–28 cognitive behavioral therapy (CBT),^29
–31 physical therapy directed to the cervical spine,^32

–38 multimodal rehabilitative interventions,^39

–43 neurofeedback training,^44,45 progressive muscle relaxation (PMR),⁴⁶ interactive smartphone application,⁴⁷ and the effects of melatonin supplementation.⁴⁸

Risk of bias assessment

All acceptable RCTs used clear research questions, and apart from one study,²⁶ achieved similarities at baseline between groups. The majority of articles reported on adequate concealment, methodology of randomization, and intention-to-treat analysis. The follow-up rate for all RCT articles, apart from two articles,^21,31 was <25%. All acceptable non-controlled trials used a clear research question. A minority of trials utilized blinding of outcome assessors and participants; however, the vast majority of trials did not report on intention-to-treat analysis, nor did the majority of trials report on consistency of intervention or on unintended interventions. The majority of trials did not report sample size calculations, although their studies were properly powered. Additionally, of the 32 included trials, only 18 utilized validated and reliable outcome measures for this population, such as the Rivermead Postconcussion Questionnaire.

Assessment of evidence

After reviewing all relevant trials, only a single RCT was rated “high” in regards to the primary outcome measure addressed (self-reported Post-concussion Symptom Inventory Score). Half of the reviewed RCT trials were rated as “moderate” (8/16) regarding primary outcome measures (Postconcussion Symptom Inventory [PCSI], Rivermead Postconcussion Questionnaire [RPQ], medical clearance, Numerical Rating Scale [NRS]), with downgrading resulting from issues with risk of bias and imprecision (Table 3). A minority of RCT trials were rated “low” (6/16) regarding primary outcomes (RPQ, change in headache scores, NRS) with downgrading resulting from issues with risk of bias and imprecision. One RCT trial was rated very low regarding primary outcome (fMRI activation and number of symptoms) with downgrading resulting from risk of bias and imprecision. The majority of non-controlled trials were rated “very low” (14/16) regarding primary outcomes (Global Improvement Scale [GIS], Numerical Pain Rating Scale [NPRS]/NRS, Postconcussion Symptoms Scale [PCSS], exercise tolerance test), with downgrading as a result of issues with risk of bias, imprecision, and inconsistency. Two non-controlled trials were rated as “low” regarding primary outcome measures (patient reported self-dizziness and headache frequency) with downgrading as a result of risk of bias and imprecision (Table 3).

Table 3.

Summary of Findings of Primary Outcomes and Quality Assessment

Study	Number of participants	Subjects	Interventions	Control intervention	Primary outcomes	Summary statistics	Adverse events	Level of certainty	Reasons for downgrading
Moussavi et al. 2019¹⁹	22	Age: 49.5 ± 12.4 (SD) PCS duration: Ranged from 4.5 months to 4.8 years	rTMS	Sham rTMS	RPQ3 and RPQ13	RPQ3: rTMS: Post-tx: 6.7 ± 2.9 FU1: 5.5 ± 2.9 FU2: 5.6 ± 2.9 Sham: Post-tx: 6.1 ± 3.1 FU1: 5.9 ± 3.1 FU2: 5.1 ± 3.2 RPQ13: rTMS Post-tx: 28.1 ± 9.6 FU1:21.1 ± 10.3^* FU2: 21.9 ± 11.3^* Sham: Post-tx: 29.6 ± 10.9 FU1: 29.8 ± 9.9 FU2: 28.9 ± 13.5	Not reported on	Low	Risk of bias and imprecision
Leung et al. 2016¹⁷	24	Adult veterans from the Veteran Administration San Diego Healthcare System Average age rTMS group: 41 ± 14 (SD) Average age Sham group: 21 ± 12 (SD) Average duration of symptoms rTMS group (months): 178 ± 76 (SD) Average duration of symptoms sham group (months): 163 ± 42 (SD)	Neuronavigational guided rTMS	Sham rTMS	Headache intensity	Change in persistent headache intensity as measured by NRS post-treatment one-week assessment: rTMS: 56.3 ± 48.3% (SD) Sham: 15.4 +/1 43.6% (SD) (p = 0.041; F = 4.73; df = 1)	Not reported on	Low	Risk of bias and imprecision
Choi et al. 2018¹⁸	12	Participants were though who received comprehensive rehabilitation management at Department of Physical Medicine and Rehabilitation at Union hospital Average age rTMS group: 43.2 ± 9.7 (SD) Average age Sham group: 42.0 ± 8.4 (SD) Average duration symptoms rTMS group: 17 ± 7.5 Average duration symptoms sham group: 14.3 ± 7.2 (SD)	rTMS	Sham	NRS	Pre-treatment NRS score: Sham: 5.8 (0.8) rTMS: 5.8 (0.8) Change in post- treatment 1.NRS between Sham and r-TMS: NRS score at fifth treatment: p = 0.003. 2.NRS score at tenth treatment: p < 0.001. 3.1 week post-tenth treatment: p < 0.001. 4. 2 weeks post tenth treatment: p < 0.001 5. 4 weeks post-tenth treatment: p < 0.001	No adverse events were reported by participants	Moderate	Imprecision
Leddy et al. 2013²⁶	12	Participants from University Concussion Clinic Average age Exercise Group: 24 Average age Stretching Group: 21 Average age healthy control: 21	Aerobic exercise	Placebo stretching and healthy control	fMRI activation patterns; change in number of symptoms;	Number of symptoms at baseline: 1.Aerobic exercise: 17.5. 2. Placebo stretching: 19 Number of symptoms at 12-weeks. 3. Aerobic exercise:2^. 4. Placebo stretching: 15 Functional magnetic resonance imaging activation patterns: At baseline: Healthy control showed greater activation in posterior cingulate gyrus, lingual gyrus, and cerebellum versus PCS groups^* At 12-weeks: Aerobic exercise group showed no significant differences from healthy control Placebo stretching showed less activation in the cerebellum^*	Not reported on	Very low	Risk of bias and imprecision
Kurowski et al. 2017²⁷	30	Adolescents recruited from community, brain injury clinics, and emergency departments across greater Cincinnati, OH, USA. Average age aerobic group: 15.2 ± 1.37 Average age stretching group: 15.5 ± 1.8 Average time since injury aerobic group: 52.3 ± 19.9 (SD) days Average time since injury aerobic group: 55.9 ± 22.1 (SD) days	Sub-symptom exacerbation aerobic training	Full body stretching	Post-concussion Symptom Inventory (PCSI)	Mean baseline self PCSI aerobic training (SD): 37.4 (25.01) Mean baseline parent PCSI aerobic training (SD): 38.93 (15.13) Mean baseline self PCSI stretching (SD): 40.27 (27.25) Mean baseline self PCSI stretching (SD): 46.93 (25.22) Week 8: Mean (SD) self and parent PCSI ratings aerobic training 6.00 (8.22) and 17.33 (23.01) respectively Mean (SD) self and parent PCSI ratings full-body stretching 16.31 (24.15) and 9.15 (12.57) respectively Final assessment: Mean (SD) self and parent PCSI ratings aerobic training 4.17 (7.36) and 9.50 (19.11) respectively Mean (SD) self and parent PCSI ratings full-body stretching 15.93 (20.18) and 10.79 (13.33) respectively Improved rate of recovery for aerobic training compared with full-body stretching: F value = 4.45, p value – 0.036	No issues related to study procedures in intervention group but 1 participant dropped out of stretching group because of worsening of symptoms after 6 visits.	Moderate	Risk of bias
Potter et al. 2016²⁹	46	Adults from outpatient referral to two National Health Service secondary/tertiary care brain injury clinics London, UK Average age CBT group: 40.1 ± 10.3 (SD) Average age waiting list group: 43.1 ± 13.1 (SD) Average time since injury CBT group (months): 42 ± 39 (SD) Average time since injury waiting list group (months): 34 ± 38 (SD)	CBT	Wait listed controls	RPQ	Mean baseline RPQ scores (SD): CBT – 34.1 (11.4) Control – 33.4 (9.2) Mean 12-Weeks post- RPQ score (SD) (95% CI): CBT – 26.3 (16.4) Control – 27.8 (9.0) (-3.85 to 8.03) (p = 0.481)	Not reported on	Low	Risk of bias and imprecision
Schneider et al. 2014³²	31	Adolescents and young adults recruited from University of Calgary Sports Medicine Centre Average age treatment plus group: 15 (12-27 range) Average age standard care group: 15 (13-30 range) Average time since injury treatment plus group (days): 53 (8-276 range) Average time since injury standard care group (days): 47 (31-142 range)	Standard of care protocol for sport related concussion plus cervical spine and vestibular rehabilitation	Standard of care protocol for sport related concussion	Medical clearance to return to sport	Standard care treatment – n = 1 of 15 cleared Treatment plus cervical and vestibular rehabilitation – n = 11 of 15 cleared^* (p = 0.002) Treatment plus group were 10.27 times (95% CI 1.51- 69.56) more likely to be medically cleared to return to sport within 8 weeks than the participants in the standard care group	Not reported on	Moderate	Imprecision
Leung et al. 2018²¹	29	Adult Veterans Average age rTMS group: 33 ± 8 (SD) Average age Sham group: 35 ± 8 (SD) Average duration of symptoms rTMS group (months): 95 ± 83 (SD) Average duration of symptoms sham group (months): 99 ± 58 (SD)	rTMS	Sham	1.Average daily persistent headache intensity. 2.Prevalence of persistent headaches. 3.Debilitating headache	Post-treatment 1 week: Significant improvement noted in rTMS – 25.3 +/1 16.8% reduction in average daily headache intensity (p < 0.0001) Post-treatment 4 weeks: Significant improvement noted in rTMS – 23.0 ± 17.7% reduction in average daily headache intensity (p < 0.01)	Not reported on.	Low	Risk of bias and inconsistency
Stilling et al. 2020²²	20	Adults recruited from Calgary Brain Injury Program Average age: 36 ± 11.4 (SD) Average time since injury (months): 32.5 ± 13.9 (SD)	rTMS	Sham	Change in headache frequency / severity at 1-month post-intervention	Headache severity (two-way mixed ANOVA) at primary end-point 1 month: F (df) – 3.214 (3,54) p = 0.03	Mild aggravation of headaches (4.23% SD = 14.32), scalp discomfort (0.96% SD = 8.9), toothache (0.675% SD = 4.14) and dizziness (0.3% SD = 9.7). No serious adverse events	Low	Risk of bias and imprecision
Thastum et al. 2019⁴⁰	112	Adults presenting to two University hospitals in Central Denmark, Copenhagen Average age: 22.9 ± 4.2 (SD) Average time since injury (months): 3.8 ± 1.7 (SD)	Individually tailored interdisciplinary intervention program based on principles of CBT (GAIN)	Usual care	RPQ	Mean adjusted difference on RPQ between GAIN and usual care: 7.6 (95% CI 2.0-13.1) p = 0.008 Effect size – Cohen's d = 0.5 (95%CI 0.1-0.9)	No adverse events described by participants	Moderate	Imprecision
Rytter et al. 2019³⁹	89	Adults presenting for treatment at specialized post-acute outpatient hospital, the Centre for Rehabilitation of Brain Injury in Denmark, Copenhagen Age range: 18-65 Average time since injury S-REHAB group (months): 29 ± 18.11 (SD) Average time since injury Usual care group (months): 27 ± 16.3 (SD)	S-REHAB (neuropsychological treatment with exercise therapy)	Usual care	RPQ	6-month follow-up: Usual Care RPQ mean (SD): 37.5 (8.48) S-REHAB RPQ mean (SD): 32.29 (14.18) p = 0.013 Effect size = 0.29	Not reported on	Moderate	Imprecision
Kjeldgaard et al. 2014³⁰	90	Adults referred to the Danish Headache Center Average age: 34 ± 11.3 (SD)	9-week CBT intervention program	Waiting list	Basic headache diary	Headache frequency (days/4 weeks): Baseline CBT mean (SD)– 26.4 (4.39) Baseline waiting list mean (SD) – 25.9 (4.87) Week 26 follow-up CBT mean (SD) – 26.0 (5.13) Week 26 follow-up waiting list mean (SD) – 25.1 (6.51) p = 0.551 Headache intensity (0-10 scale): Baseline CBT mean (SD)– 6.33 (1.78) Baseline waiting list mean (SD) – 5.45 (1.79) Week 26 follow-up CBT mean (SD) – 6.34 (1.69) Week 26 follow-up waiting list mean (SD) – 5.10 (2.27) p = 0.128	Not reported on	Low	Risk of bias and imprecision
Barlow et al. 2020⁴⁸	99	Children seen at Alberta Children's Hospital emergency department Average age 3 mg melatonin group: 13.3 ± 2.5 (SD) Average age 10 mg melatonin group 13.8 ± 2.6 (SD) Average age placebo group: 14.3 ± 2.5 (SD) Average symptom duration 3 mg melatonin group: 38.8 ± 6.2 (SD) Average symptom duration 10 mg melatonin group 38.0 ± 5.2 (SD) Average symptom duration placebo group: 37.5 ± 6.7 (SD)	3 mg or 10 mg melatonin	Placebo	Post-concussion Symptom Inventory Score (PCSI)	Change in PCSI score: Melatonin 3 mg to placebo median (95% CI): -2 (-13 to 6) Melatonin 10 mg to placebo Median (95% CI): - 4 (-7 to 14) p = 0.36	Eight events involved a known melatonin side effect, and 5 were potentially related. Ten events were associated with a mild functional impact. The frequency ( × 2 = 1.755; p = 0.42) and severity ( × 6 = 6.619; p = 0.36) of adverse events did not differ between groups.	High	None
Jensen et al. 1990³³	19	Adults admitted to the County Hospital of Aarhus, Denmark Average age manual therapy group: 32.3 (19-48 range) Average age cold pack group: 30.8 (21-45 range) Symptom duration: 9-12 months of persisting symptoms	Manual therapy to cervical spine	Cold packs to neck	Mean pain index (as measured over 28 registrations in mm)	Trial end-point (5 weeks) Mean change in Mean Pain Index (SD): Manual therapy: -305.7 (329.5) Cold pack: 67.4 (362.4) p < 0.05	Not reported on	Low	Risk of bias and imprecision
Walker et al. 2008⁴⁴	26	Average age of participants: 38.6 ± 13.5 (SD) Average symptom duration (months): 12.7 ± 18.5 (SD)	Neurofeedback therapy	None	Global improvement score (GIS) Number of sessions	GSI mean (SD): 72.7 (27.6) Mean number of sessions (SD): 19.1 (9.7)	Not reported on	Very low	Risk of bias, imprecision and publication bias
Leung et al. 2016¹⁷	6	American Veterans Age: 50.2 ± 4.4 (SD) Average PCS duration: 4.4 ± 2.2 (SD)	rTMS	None	Numerical Pain Rating Scale (NPRS)	Pre-treatment mean NPRS (SD): 5.5 (1.4) Post-treatment mean NPRS (SD): 2.67 (1.8)	Not reported on.	Very low	Risk of bias, imprecision and publication bias
Koski et al. 2015²³	15	Adults referred by physiatrist at a TBI clinic in Montreal, Quebec, Canada Average age: 34.3 ± 10.8 (SD) Time since injury (years): 0.5-28	rTMS	none	Post-concussion Symptom Scale (PCSS)	Mean pre-treatment PCSS (SD): 45.8 (17.2) Mean change post-pre-treatment PCSS (SD): -14.6 (16.1)	Left the study after a few rTMS sessions because of worsening symptoms (n = 2) Side effects reported by the 12 participants were headache (n = 3), a single episode of vertigo (n = 1), anxiety about the procedures (n = 1), and increased sleep/disturbance (n = 3).	Very low	Risk of bias and imprecision
Huang et al. 2017²⁵	6	Adults Age range: 27-41 Average symptom duration (months): 48.2 ± 25.2 (SD)	Low-intensity pulses using transcranial electrical stimulation (LIP-tES) with electroencephalography (EEG) monitoring	None	RPQ	Average change in RPQ pre-post- treatment (SD): 52.76 (26.4)	Not reported on	Very low	Risk of bias, imprecision and publication bias
Hugentobler et al. 2015⁴²	6	Adolescents and young adults Age range: 15-19 Average symptom duration range (days): 33-192	Individualized physical therapy treatment plan based on presentation during evaluation	None	Post-concussion Symptom Scale (PCSS)	Change in PCSS evaluation to final session: Patient 1: -6 Patient 2: -2 Patient 3: -38 Patient 4: -23 Patient 5: -21 Patient 6: -54 No other descriptive statics were provided.	Not reported on	Very low	Risk of bias, imprecision, and publication bias
Leddy et al. 2010²⁸	12	Adults being seen at University at Buffalo Concussion Clinic Average age: 27.9 ± 14.3 (SD) Average symptom duration (weeks): 19 (6-40 range)	Subsymptom aerobic exercise once per day for 5-6 days per week	None	Exercise tolerance test and post-concussion symptoms via the Graded Symptom Checklist (GSC)	Mean exercise tolerance via exercise time pre-treatment (SD): 9.75 (6.38) Mean exercise tolerance via exercise time post-treatment (SD): 18.67 (2.53) Mean GSC pre-treatment (SD): 9.67 (5.87) Mean GSC post-treatment (SD); 5.42 (4.54) p = 0.002	No adverse events reported	Very low	Risk of bias and imprecision
Gagnon et al. 2009⁴¹	16	Children and adolescents seen at the Montreal Children's Hospital. Age range: 10-17	The Montreal Children's Hospital Rehabilitation After Concussion program (supervised physical conditioning, education, visualization, and coordinated exercises)	None	Post-concussion Scale-Revised Score (PCSRS)	Initial mean PCSRS (SD): 30.0 (20.8) Discharge mean PCSRS (SD): 6.7 (5.7)	Not reported on	Very low	Risk of bias, imprecision, and publication bias
Hammerle et al. 2019³⁴	48	Adult patients seen at Brooke Army Medical Center Brain Injury Rehabilitation Service physical therapy clinic Average age: 33.48 ± 8.88 (SD) Time since injury (years): < 1 year = 60.4% > 1 year = 39.6%	Cervical Spine Proprioceptive Retraining (CSPR)	Standard care (vestibular rehabilitation therapy VRT)	Patient self-reported dizziness symptoms (improved or not improved)	Post-treatment dizziness not improved (n = 22): VRT = 18 CSPR = 4 Post-treatment dizziness improved (n = 26) VRT = 4 CSPR = 22 Adjusted Odds ratio (95% CI): 30.12 (4.44-204.26) p < 0.001	Not reported on	Low	Risk of bias
Kennedy et al. 2017³⁵	21	Adults referred to physiotherapist in private practice. Average age: 26.5 ± 12.0 (SD) Time since injury (months): < 1 = 15% 1-3 = 63% 4-6 = 7% > 6 = 9% Unclear = 7%	Physical therapy to cervical spine	None	Patient specific functional scale increase (minimum clinical important difference = 2.2) Numerical pain rating scale decrease (NPRS) (minimum clinical important difference = 2.2)	Mean patient specific functional scale increase (SD) = 3.8 (1.9) Mean Numerical pain rating scale decrease (SD) = 4.6 (1.8)	Not reported on.	Very low	Risk of bias and imprecision
Marshall et al. 2015³⁶	5	Adults being seen at Canadian multi-disciplinary concussion clinic Age of participants (range): 19-59 Time since injury (range in weeks): 5-124	Manual therapy / rehabilitation to cervical spine	None	Case series reporting various outcomes including post-concussion symptoms and Balke physical exertion protocol	Patient 1: Fully completed Balke test with no symptoms and return to sport Patient 2: At 3 months patient was symptom free Patient 3: 80% reduction in post-concussion symptom severity score Patient 4: 47% reduction in symptom severity score with complete resolution by eighth visit. Patient 5: 3 treatments over 6 weeks resulted in complete resolution of symptom score.	Not reported on	Very low	Risk of bias, imprecision, and publication bias
Miller et al. 2020²⁴	7	Adults recruited from Danish Concussion Patient Union Average age: 34 Average time since injury (months): 11	Transcranial low-frequency pulsating electromagnetic fields (T-PEMF)	None	RPQ	Percentage change of RPQ from baseline to 5 weeks: Patient 1: -48% Patient 2: +86% Patient 3: -22% Patient 4: -2% Patient 5: -39% Patient 6: -61% Patient 7: +13% Percentage change of RPQ from trial end-point (5 weeks) to 3 months: Patient 1: -12% Patient 2: -5% Patient 3: +14% Patient 4: -37% Patient 5: -11% Patient 6: 0% Patient 7: +25%	Worsening headaches, worsening of symptoms overall, brain fog, redness of eyes, dizziness, fatigue, photosensitivity, nausea.	Very low	Risk of bias, imprecision, and publication bias
Usmani et al. 2021⁴⁶	49	Adults recruited from New York University Concussion Center Average age: 40.1 ± 14.6 (SD) Average time since injury (months): 6.7 ± 2.5 (SD)	Smartphone delivered progressive muscle relaxation (PMR)	None	Patient reported headache symptoms	Baseline average headache frequency in days per month (SD): 17.7 (9.3) Month 1 follow-up average headache frequency in days per month (SD): 10.1 (8.6) Month 2 follow-up average headache frequency in days per month (SD): 10.4 (8.6) Month 3 follow-up average headache frequency in days per month (SD): 8.9 (8.9)	Not reported on	Very low	Risk of bias and imprecision
Kennedy et al. 2021³⁷	11	Adults presenting to a multi-disciplinary concussion clinic Average age: 37 ± 16 (SD) Average time since injury (weeks): 7.5 ± 8.9 (SD)	Manual therapy / rehabilitation of cervical spine	None	RPQ; NDI; DHI	Baseline RPQ (SD): 35.4 (7.2) Baseline NDI (SD): 32.9 (7.3) Baseline DHI (SD): 51.8 (15.4) Post-treatment RPQ (SD): 14.1 (8.7) Post-treatment NDI (SD): 12.5 (7.5) Post-treatment DHI (SD): 38.0 (13.4) 6-month follow-up RPQ (SD): 14.1 (10.7) 6-month follow-up NDI (SD): 22.5 (17.0) 6-month follow-up DHI (SD): 16.0 (not reported) 12-month follow-up RPQ (SD): 10.5 (8.4) 12-month follow-up NDI (SD): 18.0 (6.9) 12-month follow-up DHI (SD): 26.0 (17.1)	No adverse effects from treatment were reported.	Very low	Risk of bias, imprecision, and publication bias
Wetzler et al. 2017³⁸	11	Adult ex-professional football players who presented for treatment at Upledger Institute in West Palm Beach, Florida, United States Age range: 24-73 Time since injury (years): 2-33	CranioSacral Therapy (CST), Visceral Manipulation (VM) and Neural Manipulation modalities (NM)	None	Pain rating score	Baseline Symptom Severity Score from 0-10: Patient 1: 4 Patient 2: 7 Patient 3: 9 Patient 4: 8 Patient 5: 8 Patient 6: 5 Patient 7: 5 Patient 8: 7 Patient 9: 4 Patient 10: 8 Patient 11: 7 Follow-up (3 months) Symptom Severity Score from 0-10: Patient 1: 1 Patient 2: 4 Patient 3: 5 Patient 4: 4 Patient 5: 5 Patient 6: 2 Patient 7: 4 Patient 8: 4 Patient 9: 3 Patient 10: 4 Patient 11: 3 P = 0.0448	Not reported on	Very low	Risk of bias, imprecision, and publication bias
Elbogen et al. 2021⁴⁵	41	Adult veterans recruited through veteran organization and medical centers in southeastern United States Average age: 38.5 ± 10 (SD) Time since injury (years): 13.11 ± 9.8 (SD)	Neurofeedback delivered with a portable electroencephalograph (EEG)	None	Average pain intensity via NRS	Baseline NRS (SD): 6.4 (1.24) Follow-up NRS: 5.39 (1.70) p < 0.001 Cohen's d = 0.61	No adverse events reported.	Very low	Risk of bias, imprecision, and inconsistency
Belanger et al. 2022⁴⁷	479	Veterans Average age in Concussion Coach group (SD): 42.1 (12.2) Average age in Treatment as usual group (SD): 41 (10.9) Time since injury in Concussion Coach group (SD): 9 (9) Time since injury in Treatment as usual group (SD): 8 (7)	Concussion Coach (CC)	Treatment as usual (AU)	Neurobehavioral Symptom Inventory (NSI)	3-month follow-up: Probability of reduced NSI score (95% CI): CC: 0.35 (0.32-0.37) TAU: 0.29 (0.27-0.32) Odds ratio: 1.29	Not reported on	Moderate	Risk of bias and inconsistency
McGeary et al. 2022³¹	193	9/11 combat veterans Age: 39.7 ± 8.4 (SD) Average headache symptoms duration: 2 years	Cognitive Behavioral Therapy (CBTH) or Cognitive Processing Therapy (CPT)	Usual care	6-item Headache Impact Test (HIT-6)	HIT-6 Post-treatment CBTH (95% CI): -3.4 (-5.4 to -1.4) p < 0.01 Post-treatment CPT (95% CI): -6.5 (-12.7 to -0.3) p = 0.21	Adverse events identified were not study related	Moderate	Risk of bias
Langevin et al. 2022⁴³	60	Adults 18-65 years- Average age (SD): Experimental group: 38.9 (14.56) 39.3 (15.08) days since injury Control group: 39.07 (12.63) 38.77 (14.06) days since injury	6-week combined cervicovestibular rehabilitation program with symptom-limited aerobic exercise program (EXP)	6-week symptom-limited aerobic exercise program (CTRL)	Postconcussion Symptom Scale (PCSS)	Mean PSCC Scores (SD) Baseline: EXP: 62.83 (23.69) CTRL: 61.77 (22.59) Week 6: EXP: 17.21 (13.45) CTRL: 23.85 (19.98) Week 12: EXP: 16.46 (16.12) CTRL: 16.18 (14.85) Week 26: EXP: 13.96 (12.63) CTRL: 14.67 (16.85)	Not reported on	Moderate	Risk of bias

p < 0.05; ^p < 0.0004.

rTMS, repetitive transcranial magnetic stimulation; RPQ, Rivermead Postconcussion Questionnaire; PCS, Post-concussion syndrome; FU1, follow-up 1 month; FU2, follow-up 2 months; post-tx, post-treatment; CBT, cognitive behavioral therapy; SD, standard deviation; NDI, Neck Disability Index; DHI, Dizziness Handicap Inventory.

Summary of evidence

rTMS

All trials investigated the effectiveness of rTMS on an adult population. Evidence suggests that this form of rehabilitation, as measured by the RPQ, NRS, or self-reported headache intensity/frequency, shows statistically significant improvements (Table 3). Across all RCTs, the use of rTMS was consistently compared with a sham TMS treatment. Evidence from the five non-controlled trials also suggests that rTMS reduces persistent post-concussion symptoms, specifically headache intensity and frequency as measured by the NPRS, PCSS, RPQ, and participant-reported headache intensity/frequency (Table 3). The quality of the evidence for rTMS, as rated by the GRADE criteria, ranged from very low to moderate. The reasons for downgrading were risk of bias, imprecision, and publication bias secondary to specific issues with sample sizes, study design, and a lack of reporting on confidence intervals.

Aerobic exercise

Leddy and coworkers identified that after 12 weeks of sub-symptom aerobic training, adult participants reported a significant decrease in self-reported symptoms.²⁶ Sub-symptom aerobic training required participants to exercise aerobically for 20 minutes six times per week, at an intensity of 80% of the heart rate attained when their symptoms were aggravated on a baseline treadmill exertion test. The comparison group were advised to perform a 12-week low-impact breathing and stretching program. Kurowski and coworkers also reported 6 weeks of sub-symptom aerobic exercise training lead to significant reductions in post-concussion symptom inventory when compared with a full-body stretching program; however, they studied an adolescent population.²⁷ Again, participants were instructed to perform sub-symptom exercise aerobically 5–6 days per week. The comparison group was instructed to perform upper and lower body stretching for 6 weeks. Lastly, Leddy and coworkers showed via a non-controlled trial that adult participants who engaged in 5–6 days of sub-symptom aerobic exercise had significant change in post-concussion symptoms via a graded symptom check list (Table 3).²⁸ The quality of the evidence, as rated by the GRADE criteria, for aerobic exercise ranged from very low to moderate. The reasons for downgrading were risk of bias and imprecision secondary to issues with study design and lack of reporting on confidence intervals.

Physical therapy directed to the cervical spine

One RCT³² demonstrated that physical therapy produced a statistically significant improvement in physically dominant persistent concussion symptoms. Both the intervention and control groups received weekly session with a physiotherapist for 8 weeks. The intervention group received additional cervical spine and vestibular rehabilitation.

Of the six non-controlled trials, four of them were descriptive case series. All trials investigated changes in post-concussion symptoms as measured by various patient-generated outcome measures after manual therapy directed to the cervical spine. The articles tentatively suggest that manual therapy directed to the cervical spine is effective at reducing physical persistent concussion symptoms in adults (Table 3). The quality of the evidence for physical therapy directed to the cervical spine, as rated by the GRADE criteria, ranged from very low to moderate. The reasons for downgrading were risk of bias, imprecision, and publication bias secondary to specific issues with sample sizes, study design, and a lack of reporting on confidence intervals.

Neurofeedback

Two non-controlled clinical trials evaluated the effectiveness of neurofeedback on physical post-concussion symptoms in adults. Walker and coworkers afforded neurofeedback therapy for 26 adult patients with persistent symptoms.⁴⁴ Participants were given five training sessions designed to instruct them on how to normalize abnormal quantitative electroencephalography (qEEG). Overall, 88% of participants, after a mean of 19 sessions, had a change in global improvement score of ≥50%.⁴⁴ Elbogen and coworkers examined the effectiveness of neurofeedback on physical post-concussion symptoms with a portable EEG in 41 veterans with chronic pain following an mTBI.⁴⁵ They reported that after a mean of 33 neurofeedback sessions over a course of 3 months, participants had a statically significant change in self-reported pain intensity. The quality of the evidence for neurofeedback, as rated by the GRADE criteria, ranged from very low to moderate. The reasons for downgrading were risk of bias, imprecision, inconsistency, and publication bias secondary to issues with either sample sizes, study design, study procedures, or a lack of reporting on confidence intervals.

Progressive muscle relaxation (PMR)

One non-controlled clinical trial examined the feasibility and impact of PMR on persistent headaches following a concussion in adults. Usmani and coworkers reported that despite the lack of compliance, participants on average reported a decrease in monthly headaches from 15 at baseline to 9.7 by the 3rd month follow-up.⁴⁶ The quality of the evidence for PMR was very low, as rated by the GRADE criteria. The reasons for downgrading were risk of bias and imprecision secondary to issues with study design and a lack of reporting on confidence intervals.

CBT

Two of the three RCTs^29,30 examined the impact of CBT on persistent concussion symptoms as compared with waitlist controls in adults. Both articles failed to show any significant change post-treatment for physical dominant symptoms as measured by the RPQ or a headache diary. McGeary and coworkers compared CBT to usual care in veterans with post-traumatic headaches attributed to mTBI. They found a significant improvement in participants' headaches as compared with usual care and cognitive processing therapy.³¹ The quality of the evidence for CBT ranged from moderate to low, as rated by the GRADE criteria. The reasons for downgrading were risk of bias, and imprecision secondary to issues with study design (lack of concealment in blinding), and a lack of reporting on confidence intervals.

Multimodal rehabilitation

The majority of trials utilizing multimodal rehabilitative care were randomized (3/5). Thastum and coworkers investigated the effects of an 8-week program inclusive of CBT in adults, as applied by a neuropsychologist, occupational therapist, and physiotherapist, with gradual return to activities.⁴⁰ The authors found a significant change in RPQ score when compared with enhanced usual care (EUC). Rytter and coworkers compared a 22-week program that included neuropsychological treatments, group therapy, and individualized exercise training (S-REHAB), with usual care in adults.³⁹ They showed that the S-REHAB significantly reduced symptoms as measured by the RPQ when compared with usual care.³⁹ Finally, Langevin and coworkers examined the effects of a symptom-limited aerobic exercise program combined with cervicovestibular rehabilitation as compared with symptom- limited aerobic exercise training alone.⁴³ They found that in adults with persistent symptoms, there was no additional benefit noted for the group randomized to receive the additional cervicovestibular rehabilitation. However, they reported that both groups demonstrated significant improvements from baseline on the Postconcussion Symptom Scale following the 6-week programs, which remained significant at the 26-week follow-up assessment.⁴³ The quality of the evidence for multimodal rehabilitation ranged from very low to moderate, as rated by the GRADE criteria. The reasons for downgrading were risk of bias, imprecision, and publication bias secondary to specific issues with sample sizes, study design (observational trials), and a lack of reporting on confidence intervals.

Smartphone application

One RCT examined the effectiveness of an interactive smartphone application, Concussion Coach, on veterans' persistent concussion symptoms.⁴⁷ Belanger and coworkers compared the Concussion Coach application to treatment as usual over a 3-month period. They reported a significantly greater probability (odds ratio of 1.29) of reduced persistent concussion symptoms for those randomized to the smartphone application as compared with treatment as usual.⁴⁷ The quality of the evidence for smartphone application was moderate, as rated by the GRADE criteria. The reasons for downgrading were risk of bias and inconsistency

Melatonin supplementation

One RCT examined the effectiveness of melatonin supplementation on post-concussive symptoms in a pediatric population. The trial compared 3 mg or 10 mg supplementations with a placebo. The changes observed were not statistically significant between the placebo and either treatment group.⁴⁸ Additionally, there were no group differences on actigraphy sleep parameters after treatment, as measured by a wrist-worn accelerometer. The quality of the evidence for melatonin supplementation was high, as rated by the GRADE criteria.

Weight of evidence

The quality of the evidence ranged from “very low” to “high.” Excluding the single trial investigating the effects of melatonin supplementation, multimodal rehabilitation consistently demonstrated the highest level of evidence. Three of the five trials were found to be among the strongest methodological trials in this literature search, illustrating preliminary evidence for improved physical persistent concussion symptoms.^39,40,43 The quality of the evidence of the three trials was “moderate,” and downgrading for these trials was the result of risk of bias or imprecision.

Adverse events

Eleven of the 32 acceptable articles reported on adverse events (Table 3).^{22
–24,27,28,30,31,37,40,45,48} No serious adverse events were reported.

Discussion

There have been well-executed reviews on the topic of rehabilitation in those with persistent concussion symptoms; however, the focus of discussion in prior reviews has been on the effects of cognitive therapies for alleviating cognitive, mood, or overall symptoms.

It is well documented that the potential source for the cluster of concussion symptoms includes the brain, neck, vestibular system, oculo-motor system, and affective disorders.^49,50 To date, clinical trials have directed management of persistent concussion symptoms toward addressing cognitive impairments. By broadening the potential origin of symptoms, clinicians can be better directed to determine most probable sources of symptoms and therefore can more accurately direct treatment.

Other systematic reviews

The results of this review agree with those of previous reviews.^4,5,51,52 In their review, Rytter and coworkers reported that there is weak evidence for the benefits of graded physical exercise, vestibular rehabilitation, manual therapy to the neck, psychological treatment, and interdisciplinary rehabilitation.⁴ Makdissi and coworkers reported that there is preliminary evidence for the benefits of symptom-limited exercise, targeted physical therapy, and CBT.⁵ Bergersen and coworkers noted several methodological limitations resulting in avoidance to make strong treatment recommendations after reviewing the literature for effective psychotherapies for persistent concussion symptoms.⁵¹ Finally, most recently, Möller and coworkers reported that problem-solving therapy and CBT have been shown to improve depressive symptoms and psychological function, and that interdisciplinary rehabilitation reduces residual symptoms.⁵² Considerable methodological limitations prevented strong recommendations from the reviews for any therapy studied.

Clinical implications

This review has shown that there is preliminary evidence for the effective treatment of physically dominant symptoms in those with persistent concussion symptoms. It is unfortunate that currently, the majority of trials investigating the efficacy and effectiveness of non-pharmacological treatments for persistent physical concussion symptoms have been non-controlled clinical trials with systematic errors in methodology, or case studies leading to caution when interpreting the results.

Strength and limitations

This review has several strengths. We defined explicit inclusion and exclusion criteria to identify relevant articles from the searched literature. In addition, we had two reviewers independently screen and critically appraise the trials to minimize error and bias. Lastly, the two reviewers used well-accepted and valid sets of criteria for critically appraising controlled trials and multi-arm cohort trials with the SIGN criteria⁹ and used the Quality Assessment Tool for all other non-controlled trials.¹⁰

This review also has a number of limitations. First, scientific judgment varies among reviewers, which may affect the process of study appraisal. Training on the use of standardized critical appraisal tools and the use of a consensus process to reach decisions were done to help avoid bias. Second, the fact that a considerable number of the primary and secondary outcome measures utilized by the various trials, although standardized, were derived from the participants' own self-report, is an additional limitation. This subjective nature of the data may have influenced the results. However, given that the therapeutic strategies were chosen on the basis of the subjective complaints of the participants, this was considered a strength to the internal validity of the study. Third, for feasibility reasons, this review did not attempt to search grey literature.

Recommendation

The majority of trials suffer from significant methodological flaws limiting the generalizability and interpretation of the results. More specifically, the majority of articles reported here consistently failed to include a treatment control group, did not recruit sufficient participants to properly power the trial, or failed to use reliable and validated clinical outcome measures.

Future research should be focused on conducting RCTs comparing multimodal forms of care with current standards of care utilizing validated and reliable clinical outcome measures. Choice of outcome measures is important, especially with regard to interventional trials. They should ideally be responsive to the intervention and have enough resolution to answer the research question posed. It is also essential to apply validated and reliable clinical outcome measures, such as the Rivermead Post-Concussion Questionnaire or the Post-Concussion Symptom Scale, when assessing for change post-care. Unfortunately, a significant minority of trials are electing to use self-reported pain (numerical pain rating scale) or functional neuroimaging in isolation, which both come with drawbacks. Further, to assist with patient demographic diversity and generalizability, the study population should be recruited from large concussion assessment centers, similar to several found in this review.^{18,20,22
–24,26

–33,39
–41,43,44,46
–48} We strongly encourage future trials to adopt the most recent presented definition of post-concussion syndrome by Tator and coworkers, which requires a participant to report any three or more symptoms lasting at least 1 month following the diagnosis of a concussion.⁷ This can assist with creating homogeneity across study groups with respect to biological, psychological, and social factors. Lastly, for interventional trials, authors are encouraged to “isolate” the intended effects of the studied intervention(s). This can improve the specificity of the study and consequently improve the visibility.

Conclusion

The best available evidence on rehabilitative treatments of persistent concussion symptoms suggests that various forms of non-pharmacological therapy may provide a reduction in physical symptoms. However, significant methodological limitations of the current trials preclude us from making strong recommendations. Instead, given the encouraging preliminary findings from the multimodal therapy trials, we suggest further research in this area.

Transparency, Rigor, and Reproducibility Summary

The study was registered with PROSPERO on June 12, 2022 (CRD42022341100). A search of the literature was conducted in PubMed, MEDLINE, the Cochrane library, CINAHL, and Embase using terms related to physical post-concussion symptoms. After careful review of relevant full-text articles, 32 studies were included for analysis. Because of study methodology heterogeneity, we could not perform a meta-analysis.

Footnotes

Authors' Contributions

N.M. was responsible for conceptualization, methodology, formal analysis, investigation, writing – original draft, writing – review and editing, and visualization. S.G. was responsible for investigation, and writing – review and editing. M.R.P. was responsible for conceptualization, writing – review and editing, and supervision. S.K. was responsible for conceptualization, writing – review and editing, and supervision. All authors read and approved the final manuscript.

Funding Information

No funding was received for the support of the research, authorship, and/or publication of this article.

Author Disclosure Statement

Milos R. Popovic is co-founder, director, and shareholder in the companies MyndTec Inc. and ERNE Inc. He is also a consultant for the company Fourier Intelligence. The other authors have no competing interests to disclose.

Supplementary Material

Supplementary Appendix I

Supplementary Appendix II

References

Hunt

, Zanetti

, Kirkham

, et al. Identification of hidden health utilization services and costs in adults awaiting tertiary care following mild traumatic brain injury in Toronto, Ontario, Canada. Concussion, 2016; 1(4):CNC21; doi: 10.2217/cnc-2016-0009

McCrory

, Meeuwisse

, Dvořák

, et al. Consensus statement on concussion in sport-the 5^th international conference on concussion in sport held in Berlin, October 2016. Br J Sports Med, 2017; 51(11):838–847; doi: 10.1136/bjsports-2017-097699

McMahon

, Hricik

, Yue

, et al. Symptomatology and functional outcome in mild traumatic brain injury: results from the prospective TRACK-TBI study. J Neurotrauma, 2014; 31(1):26–33; doi: 10.1089/neu.2013.2984

Rytter

, Graff

, Henriksen

, et al. Nonpharmacological treatment of persistent postconcussion symptoms in adults: a systematic review and meta-analysis and guideline recommendation. JAMA Netw Open, 2021; 4(11):e2132221; doi: 10.1001/jamanetworkopen.2021.32221

Makdissi

, Schneider

, Feddermann-Demont

, et al. Approach to investigation and treatment of persistent symptoms following sport-related concussion: a systematic review. Br J Sports Med, 2017; 51(12):958–968; doi: 10.1136/bjsports-2016-097470

Lucas

. Posttraumatic headache: clinical characterization and management. Curr Pain Headache Rep, 2015; 19(10):48; doi: 10.1007/s11916-015-0520-1

Tator

, Davis

, Dufort

, et al. Postconcussion syndrome: demographics and predictors in 221 patients. J Neurosurg, 2016; 125(5):1206–1216; doi: 10.3171/2015.6.JNS15664

World Health Organization. Rehabilitation. 2021 [cited 2022 November 14] Available from: https://www.who.int/news-room/fact-sheets/detail/rehabilitation

Harbour

, Miller

. A new system for grading recommendations in evidence based guidelines. BMJ, 2001; 323(7308):334–336; doi: 10.1136/bmj.323.7308.334

10.

Thomas

Quality Assessment Tool for Quantitative Studies. Effective Public Health Practice Project McMaster University: Toronto; 2003.

11.

Campbell

, McKenzie

, Sowden

, et al. Synthesis without meta-analysis (SWiM) in systematic reviews: reporting guideline. BMJ, 2002; 368:l6890; doi: 10.1136/bmj.l6890

12.

Santesso

, Glenton

, Dahm

, et al. GRADE guidelines 26: informative statements to communicate the findings of systematic reviews of interventions. J Clin Epidemiol, 2020; 119:126–135; doi: 10.1016/j.jclinepi.2019.10.014

13.

Viera

, Garrett

. Understanding interobserver agreement: the kappa statistic. Fam Med, 2005; 37(5):360–363.

14.

Cohen

. A coefficient of agreement for nominal scales. Educ Psychol Meas, 1960; 20:37–46; doi: 10.1177/001316446002000104

15.

Abrams

, Gillies

, Lambert

. Meta-analysis of heterogeneously reported trials assessing change from baseline. Stat Med, 2005; 24(24):3823–3844; doi: 10.1002/sim.2423

16.

Moher

, Liberati

, TetzlaffJ, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ. 2009; 339:b2535; doi: 10.1136/bmj.b2535

17.

Leung

, Fallah

, Shukla

, et al. rTMS in alleviating mild TBI related headaches––a case series. Pain Physician, 2016; 19(2):E347–354; doi: 10.36076/ppj/2016.19.E347

18.

Choi

, Kwak

, Lee

, Chang

. Effect of high-frequency repetitive transcranial magnetic stimulation on chronic central pain after mild traumatic brain injury: a pilot study. J Rehabil Med, 2018; 50(3):246–252; doi: 10.2340/16501977-2321

19.

Moussavi

, Suleiman

, Rutherford

, et al. A pilot randomised double-blind study of the tolerability and efficacy of repetitive transcranial magnetic stimulation on persistent post-concussion syndrome. Sci Rep. 2019; 9(1):5498; doi: 10.1038/s41598-019-41923-6

20.

Leung

, Shukla

, Fallah

, et al. Repetitive transcranial magnetic stimulation in managing mild traumatic brain injury-related headaches. Neuromodulation, 2016; 19(2):133–141; doi: 10.1111/ner.12364

21.

Leung

, Metzger-Smith

, He

, et al. Left dorsolateral prefrontal cortex rTMS in alleviating MTBI related headaches and depressive symptoms. Neuromodulation, 2018; 21(4):390–401; doi: 10.1111/ner.12615

22.

Stilling

, Paxman

, Mercier

, et al. Treatment of persistent post-traumatic headache and post-concussion symptoms using repetitive transcranial magnetic stimulation: a pilot, double-blind, randomized controlled trial. J Neurotrauma, 2020; 37(2):312–323; doi: 10.1089/neu.2019.6692

23.

Koski

, Kolivakis

, Yu

, et al. Noninvasive brain stimulation for persistent postconcussion symptoms in mild traumatic brain injury. J Neurotrauma, 2015; 32(1):38–44; doi: 10.1089/neu.2014.3449

24.

Miller

, Prener

, Dissing

, et al. Transcranial low-frequency pulsating electromagnetic fields (T-PEMF) as post-concussion syndrome treatment. Acta Neurol Scand, 2020; 142(6):597–604; doi: 10.1111/ane.13300

25.

Huang

, Swan

, Quinto

, et al. A pilot treatment study for mild traumatic brain injury: neuroimaging changes detected by MEG after low-intensity pulse-based transcranial electrical stimulation. Brain Inj, 2017; 31(13–14):1951–1963; doi: 10.1080/02699052.2017.1363409

26.

Leddy

, Cox

, Baker

, et al. Exercise treatment for postconcussion syndrome: a pilot study of changes in functional magnetic resonance imaging activation, physiology, and symptoms. J Head Trauma Rehabil, 2013; 28(4):241–249; doi: 10.1097/HTR.0b013e31826da964

27.

Kurowski

, Hugentobler

, Quatman-Yates

, et al. Aerobic exercise for adolescents with prolonged symptoms after mild traumatic brain injury: an exploratory randomized clinical trial. J Head Trauma Rehabil, 2017; 32(2):79–89; doi: 10.1097/HTR.0000000000000238

28.

Leddy

, Kozlowski

, Donnelly

, et al. A preliminary study of subsymptom threshold exercise training for refractory post-concussion syndrome. Clin J Sport Med, 2010; 20(1):21–27; doi: 10.1097/JSM.0b013e3181c6c22c

29.

Potter

, Brown

, Fleminger

. Randomised, waiting list controlled trial of cognitive-behavioural therapy for persistent postconcussional symptoms after predominantly mild-moderate traumatic brain injury. J Neurol Neurosurg Psychiatry, 2016; 87(10):1075–1083; doi: 10.1136/jnnp-2015-312838

30.

Kjeldgaard

, Forchhammer

, Teasdale

, et al. Cognitive behavioural treatment for the chronic post-traumatic headache patient: a randomized controlled trial. J Headache Pain, 2014; 15(1):81; doi: 10.1186/1129-2377-15-81

31.

McGeary

, Resick

, Penzien

, et al. Cognitive behavioral therapy for veterans with comorbid posttraumatic headache and posttraumatic stress disorder symptoms: a randomized clinical trial. JAMA Neurol, 2022; 79(8):746–757; doi: 10.1001/jamaneurol.2022.1567

32.

Schneider

, Meeuwisse

, Nettel-Aguirre

, et al. Cervicovestibular rehabilitation in sport-related concussion: a randomised controlled trial. Br J Sports Med, 2014; 48(17):1294–1298; doi: 10.1136/bjsports-2013-093267

33.

Jensen

, Nielsen

, Vosmar

. An open study comparing manual therapy with the use of cold packs in the treatment of post-traumatic headache. Cephalalgia, 1990; 10(5):241–250; doi: 10.1046/j.1468-2982.1990.1005241.x

34.

Hammerle

, Swan

, Nelson

, et al. Retrospective review: effectiveness of cervical proprioception retraining for dizziness after mild traumatic brain injury in a military population with abnormal cervical proprioception. J Manipulative Physiol Ther, 2019; 42(6):399–406; doi: 10.1016/j.jmpt.2018.12.002

35.

Kennedy

, Quinn

, Tumilty

, et al. Clinical characteristics and outcomes of treatment of the cervical spine in patients with persistent post-concussion symptoms: a retrospective analysis. Musculoskelet Sci Pract, 2017; 29:91–98; doi: 10.1016/j.msksp.2017.03.002

36.

Marshall

, Vernon

, Leddy

, et al. The role of the cervical spine in post-concussion syndrome. Phys Sportsmed, 2015; 43(3):274–284; doi: 10.1080/00913847.2015.1064301

37.

Kennedy

, Chapple

, Quinn

, et al. Can the neck contribute to persistent symptoms post concussion? Long-term follow up from a prospective descriptive case series. J Man Manip Ther, 2021; 29(5):318–331; doi: 10.1080/10669817.2021.1920276

38.

Wetzler

, Roland

, Fryer-Dietz

, et al. CranioSacral therapy and visceral manipulation: a new treatment intervention for concussion recovery. Med Acupunct, 2017; 29(4):239–248; doi: 10.1089/acu.2017.1222

39.

Rytter

, Westenbaek

, Henriksen

, et al. Specialized interdisciplinary rehabilitation reduces persistent post-concussive symptoms: a randomized clinical trial. Brain Inj, 2019; 33(3):266–281; doi: 10.1080/02699052.2018.1552022

40.

Thastum

, Rask

, Næss-Schmidt

, et al. Novel interdisciplinary intervention, GAIN, vs. enhanced usual care to reduce high levels of post-concussion symptoms in adolescents and young adults 2–6 months post-injury: a randomised trial. EClinicalMedicine, 2019; 17:100214; doi: 10.1016/j.eclinm.2019.11.007

41.

Gagnon

, Galli

, Friedman

, et al. Active rehabilitation for children who are slow to recover following sport-related concussion. Brain Inj, 2009; 23(12):956–964; doi: 10.3109/02699050903373477

42.

Hugentobler

, Vegh

, Janiszewski

, et al. Physical therapy intervention strategies for patients with prolonged mild traumatic brain injury symptoms: a case series. Int J Sports Phys Ther, 2015; 10(5):676–689.

43.

Langevin

, Frémont

, Fait

, et al. Cervicovestibular rehabilitation in adults with mild traumatic brain injury: a randomized clinical trial. J Neurotrauma, 2022; 39(7–8):487–496; doi: 10.1089/neu.2021.0508

44.

Walker

, Norman

, Weber

. Impact of qEEG-guided coherence training for patients with a mild closed head injury. J Neurother, 2008; 6(2):31–43; doi: 10.1300/J184v06n02_05

45.

Elbogen

, Alsobrooks

, Battles

, et al. Mobile neurofeedback for pain management in veterans with TBI and PTSD. Pain Med, 2021; 22(2):329–337; doi: 10.1093/pm/pnz269

46.

Usmani

, Balcer

, Galetta

, et al. Feasibility of smartphone-delivered progressive muscle relaxation in persistent post-traumatic headache patients. J Neurotrauma, 2021; 38(1):94–101; doi: 10.1089/neu.2019.6601

47.

Belanger

, Toyinbo

, Barrett

, et al. Concussion coach for postconcussive symptoms: a randomized, controlled trial of a smartphone application with Afghanistan and Iraq war veterans. Clin Neuropsychol, 2022; 36(8):2093–2119; doi: 10.1080/13854046.2021.1936188

48.

Barlow

, Brooks

, Esser

, et al. Efficacy of melatonin in children with postconcussive symptoms: a randomized clinical trial. Pediatrics, 2020; 145(4):e20192812; doi: 10.1542/peds.2019-2812

49.

Wallis

, Lord

, Bogduk

. Pain and psychological symptoms of Australian patients with whiplash. Spine (Phila Pa 1976). 1996; 21(7):804–810; doi: 10.1097/00007632-199604010-00006

50.

Bogduk

, Govind

. Cervicogenic headache: an assessment of the evidence on clinical diagnosis, invasive tests, and treatment. Lancet Neurol, 2009; 8(10):959–968; doi: 10.1016/S1474-4422(09)70209-1

51.

Bergersen

, Halvorsen

JØ

, Tryti

, et al. A systematic literature review of psychotherapeutic treatment of prolonged symptoms after mild traumatic brain injury. Brain Inj, 2017; 31(3):279–289; doi: 10.1080/02699052.2016.1255779

52.

Möller

, Lexell

, Wilbe Ramsay

. Effectiveness of specialized rehabilitation after mild traumatic brain injury: a systematic review and meta-analysis. J Rehabil Med, 2021; 53(2):jrm00149; doi: 10.2340/16501977-2791

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.07 MB

0.09 MB