Active Rehabilitation Following Acute Mild Traumatic Brain Injury: A Systematic Review

Abstract

Purpose:

Following mild traumatic brain injury (mTBI), patients are often advised to restrict physical activity until full symptom resolution followed by gradual return to activity. However, extended rest periods may prolong recovery and contribute to persistent symptoms. Emerging evidence suggests early active rehabilitation that increases heart rate without exacerbating symptoms may improve mTBI patient recovery. This review aims to: (1) appraise evidence on active rehabilitation intervention for mTBI recovery within one-month of injury (i.e., exercise type, duration, intensity, etc.); and (2) recommend evidence-based rehabilitation protocols.

Method:

Pubmed, CINAHL, PsychARTICLES, SportDISCUS, and AMED databases were searched using key terms “mild Traumatic Brain Injury”, “Rehabilitation”, “Acute”, and their synonyms. Evidence was appraised using Cochrane RoB-2 and ROBINS-I.

Results:

434 citations were initially identified with seven papers systematically reviewed. Within the reviewed articles, only three were randomized controlled trials with low risk of bias, and four were non-randomized trials with low to moderate risk of bias. Findings highlighted that a range of active rehabilitation protocols were used with different exercise modalities (primarily treadmills and static cycling), durations (9–20 min, or until symptomatic, for 30–50 days or symptoms resolved), and intensities (low, moderate or high). Active rehabilitation did not cause any serious adverse events (i.e., death, hospitalisation etc.), and six studies reported that it did not exacerbate mTBI symptoms in any participants (with one participant having symptom worsening in one study). Overall, the majority of reviewed studies (n = 4) showed that active rehabilitation decreased time to mTBI symptom resolution compared to controls (strict rest/stretching exercises).

Conclusions:

Individualized active rehabilitation prescribed within one-month post-mTBI appears to be safe and effective at decreasing recovery time to symptom resolution in mTBI. However, there is a lack of consensus regarding specific intervention protocols that needs to be addressed before adoption within clinical practice.

Mild traumatic brain injury (mTBI) (or concussion) can be caused by indirect or direct impact to the head, neck, or body and is followed by onset of neurological impairments.¹ There is no ‘gold-standard’ for mTBI diagnosis and current diagnostic approaches have difficulty detecting subtle symptoms,¹ making mTBI difficult to diagnose. In addition, treatment protocols are highly variable, making mTBI difficult to manage.^2,3 Clinical symptoms include headache, dizziness, nausea, balance deficits, vision deficits, and memory loss.² It has been suggested that most people with mTBI recover within 2–4 weeks;⁴ however, approximately 20% of individuals will experience lasting chronic symptoms (>12 weeks), defined as post-concussion syndrome (PCS).⁵ Despite improvements in understanding in the structural changes associated with moderate and severe TBI which validated neurofilament light as a fluid biomarker of axonal damage after moderate-to-severe TBI, imaging and biomarkers of recovery in mTBI are still non-specific and unproven.⁶, ⁷, ⁸ Physical, cognitive, and emotional impairment, such as anxiety, decreased concentration and decreased quality of life, are associated with PCS.⁹ Therefore, research has begun to investigate evidence-based ways to mitigate PCS and accelerate recovery at an early stage following an mTBI.¹⁰ Typical management of mTBI is predominantly concerned with reducing symptom severity, preventing PCS, and decreasing time to symptom resolution.¹¹ Previous guidelines have recommended strict rest (SR) until recovery,⁴ but recent shift in mTBI guidelines² suggest that active rehabilitation (AR) (i.e., cardio-vascular or strengthening exercise that aims to increase heart rate) may be safe and effective for decreasing time to recovery and preventing occurrence of PCS.^1,7,8,12,13 AR is suggested to improve recovery by increasing exertional tolerance in those following mTBI, which may relate to improved physiological and autonomic nervous system (ANS) function.¹ However, there are inconsistencies regarding best practice (i.e., ideal time and type of intervention).^3,11

The benefits of AR, such as improved sleep, mental health, cognitive health, cardiorespiratory fitness, and bone and muscular strength, reduced risk of disease (i.e., hypertension, stroke, diabetes, cancers, etc.), reduced risk of falls, and maintenance of healthy body weight, are well-established in healthy individuals, and across numerous health-related conditions.¹⁴ Guidelines for mTBI treatment do not currently include AR,^15,16 but there is recent research investigating whether the benefits of AR extend to individuals with mTBI.^17,18 Most recently, the 5th Concussion in Sport Group consensus recommended an initial rest period (24–48 hours) followed by gradual initiation of AR.² This is largely based on evidence by Thomas and colleagues¹⁹ in which mTBI patients were randomized into a SR or usual care group, which found that SR worsened symptoms and led to slower symptom resolution. Furthermore, in a large multi-centre cohort study of 3,036 adolescents with mTBI,²⁰ individuals who participated in AR within seven days post-injury were significantly less likely to experience mTBI symptoms (27.8%) compared to SR participants (40.1%).²⁰ It is thought that activity that increases heart rate (HR) without exacerbating mTBI symptoms may have a positive effect on recovery, likely through neurophysiological mechanisms, such as increased proliferation of neural stem cells²¹ and neuroplasticity.²²

It is now widely accepted that SR is not beneficial following an mTBI,² however the most effective mode, intensity, frequency, and duration of AR interventions remain unknown.²³ Despite this, there has been development of standardized rehabilitation protocols such as the Buffalo Concussion Treadmill Test (BCTT) and Buffalo Concussion Bicycle Test (BCBT).^11,18 Furthermore, a recent review by Baker and colleagues¹¹ investigated the literature regarding various modes and intensities used in AR interventions, however this review was limited to only examining rehabilitation deployed within 7 days post-mTBI and did not provide comprehensive details on the methodologies or results (rehabilitation duration, intensity, type, effect, feasibility, etc.), therefore this previous work did not provide comprehensive analysis of AR interventions within the standard acute period of 2–4 weeks.² Here we provide a comprehensive review of AR interventions within acute mTBI.

This study aimed to review the effectiveness of AR intervention in acute mTBI and its specific components, including intensity, frequency, duration, and exercise type. Therefore, the objectives of this systematic review are to: (1) appraise evidence involving AR interventions used to manage acute mTBI, including mode, intensity, frequency, and duration; and (2) recommend evidence-based approaches of mTBI rehabilitation protocols based on current empirical evidence and provide future research directions.

Methods

Study design

This systematic review was written according to the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) checklist²⁴ (Appendix 1).

Eligibility criteria

Inclusion criteria were: (1) original research including randomized controlled trials (RCT), cohort studies, case-control studies, case series, and quasi-experimental studies; (2) written in English language; (3) mTBI diagnoses; (4) AR intervention with specified intensity, duration, frequency, and type of exercise conducted within one month; (5) evaluation of the symptomatic response to AR intervention. Exclusion criteria were participants who were diagnosed with PCS, animal studies, and other non-AR interventions, including oculomotor, and vestibular rehabilitation.

Search criteria and study selection

A literature search was conducted in five databases (PubMed, SportDISCUS, CINAHL, PsychARTICLES, and AMED) up to August 2022. Key terms for the search were “mild Traumatic Brain Injury”, “Rehabilitation” and “Acute”, and their synonyms (Table 1). Articles related to animal models, non-AR, or persistent (chronic, i.e., >12 weeks post-injury) mTBI symptoms were excluded using separate key terms. Duplicates were removed. An initial title and abstract screen was conducted by two independent reviewers (BC, DP), and full texts were reviewed if additional information about the studies' eligibility criteria was required. A secondary search of similar systematic reviews’ reference lists was used to identify potential additional relevant articles.

Table 1.

Key Search Term

Search terms used	Limiters	Search engine	Total citations
Mild traumatic brain injury: “mild traumatic*” OR “concussion” OR “mTBI” TITLE-ABS-KEY AND Rehabilitation: “rehabilitation*” OR “exercise*” OR “exercise therapy” OR “subthreshold” OR “intensity” OR “aerobic” OR “resistance” OR “active” TITLE-ABS-KEY AND Acute: “acute*” NOT (“animal” OR “vestib*” OR “ocul*” OR “post-concussion*” OR “PCS”) TITLE-ABS-KEY	English language; original research (RCT; clinical trials; case series; observational studies; cohort studies)	AMED CINAHL SportDISCUSS PsychARTICLES PUBMED	20 103 64 50 190

mTBI = mild traumatic brain injury; RCT = randomized controlled trial; PCS = Post-concussion syndrome.

Data management and extraction

Data were extracted by two reviewers (BC, DP). Study characteristics included study design, location of study, aims, participant characteristics, outcomes measured, and results. Intervention data included type, frequency, duration, intensity, compliance, equipment, and supervision, in line with the Template for Intervention Description and Replication (TIDieR) checklists²⁵ (Appendix 2).

Quality appraisal

The Oxford Center for Evidence-Based Medicine Levels of Evidence (2011) determined level of evidence (LoE) (Table 2).²⁶ TIDieR checklists were used to appraise intervention quality.²⁵ Two risk of bias (RoB) appraisal tools: The Cochrane Risk of Bias version 2 (RoB-2) for RCTs, and the Cochrane Risk of Bias in Non-Randomized Studies – of Interventions (ROBINS-I) for non-randomized study designs, were used. Overall RoB was rated as “high”, “moderate”, or “low” in RoB-2 or “critical”, “serious”, “moderate”, or “low” in ROBINS-I RoB-2 and ROBINS-I appraisals were reviewed by two reviewers (BC, DP) for quality assurance.

Table 2.

Oxford Centre for Evidence-Based Medicine (2011) Levels of Evidence (LoE)

LoE hierarchy	Criteria for LoE
I	Systematic reviews of RCTs
II	RCTs, observational studies with significant effects
III	Non-randomized control studies, cohort studies, control arms of RCTs, follow-up studies
IV	Case series, case-control studies, historical control studies
V	Mechanism-based reasoning

LoE = levels of evidence.

Data synthesis

To synthesize the collected data, the York Center for Reviews and Dissemination (CRD) framework (2009)²⁷ was used. A meta-analysis was not conducted due to heterogeneous intervention and various study outcomes.

Results

Search results

A total of 434 articles were identified (Table 1 and Figure 1).²⁸ Two reviewers (BC, DP) completed a title/abstract screen and reviewed 21 full texts (BC, DP). Finally, a total of seven articles were selected for inclusion. Most excluded articles did not include AR or participants with acute mTBI (full details for exclusion of full-text reviewed articles is shown in Appendix 3).

Figure 1. PRISMA flow diagram including citations identified, removed, excluded with reasons, and included.²².

Participant characteristics

A total of 584 participants were included in the reviewed studies (see Table 3). Female and male participants were recruited to every study except for one study of males only.²⁹ Participants were included in each study if they were diagnosed with a symptomatic acute mTBI. Two studies did not define the acute period,^30,31 but other studies provided timelines, for example, Hinds and colleagues³⁰ used 5 days, and Maerlenderm and colleagues³¹ involved participants immediately after initial diagnosis.

Table 3.

Summary of Study Characteristics

Study, design, and setting	Primary aim(s)	Population (number, sex, age, days since injury, days)	Intervention	Symptom outcomes and recovery definition	Outcomes	Key results
Hinds and colleagues RCT pilot study.30 Concussion clinic; USA	To determine if concussed individuals demonstrate cardiovascular dysregulation during exercise, and if ANS recovery correlates with resolution of symptoms	70 participants, 40 adolescent athletes with SRC diagnosed by physician, and 30 healthy adolescent athletes without SRC in past 6 months EG: 40 concussed adolescent athletes (M: 23, F: 17), Age: 15.5 ± 0.9, Days since injury: 5 ± 1.1 CG (N = 30, M:18, F:12), Age: 15.9 ± .05 Days since concussion: not applicable	BCTT	Recovery: resolution of symptoms, no symptom exacerbation with treadmill exertion, normal physical exam. Symptoms self-reported.	HR; RPE – 6–20 on Borg Scale	• Participants had lower HR at onset of exercise when concussed compared to recovered (72 ± 6.3 vs. 74 ± 6.7 BPM, p > 0.05) • Greater RPE was reported at every intensity level, despite exercising at lower workloads
Leddy and colleagues36 RCT University and community sports medicine centers; USA	To evaluate the systematic assessment of exercise tolerance in adolescents shortly after SRC and the prognostic utility of such assessment	54 adolescent athletes presented to clinic within 10 days of SRC, and diagnosed by physician EG (N = 27, M: 18, F: 9) Age: 15.2 ± 1.5 Days since injury: median = 5, IRQ = 3 RG (N = 27, M: 19, F: 8) Age: 15.6 ± 1.4 Days since injury: median = 5, IRQ = 3	EG: BCTT on Visit #1 and visit #2 (14 days apart); daily symptom reporting RG: Standard care, no BCTT on visit #1, BCTT visit #2 (14 days after visit #1); daily symptom reporting	Recovery: symptoms resolution to normal, confirmed by normal independent physical exam. Symptom resolution defined as PCSS score of <5 points for males and <9 points for females, exercise tolerance on BCTT and normal cognitive performance on ImPACT.	Days from injury to recovery	• Faster recovery time for EG (8.29 ± 3.85 days) compared with RG (23.93 ± 41.73 days) • At day 14, exercise group had significantly fewer symptomatic subjects than the rest group. No subjects in exercise group had delayed recovery, whereas 4 RG subjects did (113.25 ± 73.6 days)
Leddy and colleagues32 Parallel Multi-Center RCT 4 outpatient concussion management clinics; USA and Canada	To assess the effectiveness of sub-symptom threshold aerobic exercise (AE) vs. a placebo-like stretching group (SG) prescribed to adolescents in the acute phase of recovery from SRC	103 concussed adolescents, presented to clinic within 10 days of SRC and diagnosed by physician AE: (N = 52, M: 28, F: 24) Age: 15.3 ± 1.5 Days since injury: 4.9 ± 2.2 SG: (N = 51, M: 27, F: 24) Age: 15.4 ± 1.7 Days since injury: 4.8 ± 2.4	AE: Daily aerobic exercise at prescribed target HR SG: Prescribed stretching programmes	Recovery: symptoms resolution to normal, confirmed by normal physical exam. Ability to exercise to exhaustion without symptom exacerbation. Symptom resolution defined as PCSS score of 7 points or fewer, for 3 consecutive days	Days from injury to recovery; BCTT result	• AE subjects recovered in a median of 13 days, compared with stretching participants in 17 days • Incidence of subjects with delayed recovery was longer in stretching group (N = 7) compared with exercise group (N = 2) • Total symptom score appeared to decrease more rapidly in exercise group
Leddy and colleagues29 Quasi-experimental trial University sports medicine centers; USA and Canada	To study the effect of early prescribed aerobic exercise group (EG) vs. rest group (RG) on rate of recovery in male adolescents acutely after SRC	54 concussed male adolescent athletes Presented to clinic within 9 days of SRC and diagnosed by physician EG (N = 24): Age: 15.1 ± 1.4. Days since injury: 4.8 ± 2.5 RG (N = 30) Age: 15.3 ± 1.4 years Days since injury: 4.5 ± 2.1	EG: Daily progressive exercise programme RG: Relative rest.	Recovery: no symptoms or return to baseline of level of symptoms Symptom resolution defined as PCSS score of 7 points or fewer, for 3 consecutive days	Days from initial visit to recovery; symptom scores (PCSS)	• Recovery time from initial visit was significantly shorter in EG (8.29 ± .9 days vs. 23.93 ± 41.7 days, p = 0.048) • Mixed effects linear models showed that all symptom clusters decreased with time and that there was no significant interaction between treatment group and time • No EG participants experienced delayed recovery (>30 days) whereas 13% (4/30) of RG participants experienced delayed recovery
Maerlender and colleagues31 Controlled study Concussion clinic; USA	To determine if exertion has an effect on recovery from exercise, and to refute the hypothesis that exertion negatively impacts recovery	28 collegiate athletes diagnosed with SRC, presented median of 2 days post SRC and diagnosed by physician EG (N = 13, M: 5, F: 8). (demographics not presented). RG (N = 15, M: 3, F: 12). (demographics not presented)	Exertion: Daily, moderate intensity physical exertion Standard: Standard care recommendations	Recovery: Symptom resolution, no symptom exacerbation with exertion. Symptoms rated via self-reported VAS (0–10 points).	Days from initial visit to recovery; Symptom change following exercise; Neuro-physiological variables	• 1.8 symptom increase per ride in the light ride category • 0.55 symptom increase per ride in the moderate ride category • 7.0 symptom increase per ride in the strenuous ride category
Popovich and colleagueas34 Retrospective cohort study University SRC clinic at a tertiary care center; USA	To assess the safety of supervised exercise (SE) in acute sport-related concussion (SRC) and its influence on recovery	126 adolescent athletes presented to clinic within 30 days of sustaining a SRC. NR who diagnosed SE (N = 26, M: 20, F: 6) Age: 15.3 (10.2–19.4). Days since injury 5.4 (1–15) CG (N = 98, M: 68, F: 30), Age: 15.4 (8.7–19.8) Days since injury: 10.5 (1–30)	SE: Supervised exercise (SE) programme CG: No SE programme, or supervised exercise after 16 days of injury	Recovery: early clearance for return to sport by clinician. Symptoms self-reported via SCAT3.	Days from concussion to RTP; Number of days symptomatic from injury	• Days to recovery and typical vs. prolonged recovery were not significantly different between groups (SE: 27.5 ± 36.6, CG: 23.5 ± 39.4, p = 0.7060) • Symptom severity scores decreased in both groups over 14 days (p < 0.0001), were similar (p = 0.2984), and did not significantly increase the day after the BCTT (p = 0.1960) • Lower HR on visit day 1 was strongly associated with prolonged recovery time (p = 0.0032)
Willer and colleagues33 Quasi-experimental trial. University concussion management clinics; USA and Canada	To compare adolescents with SRC who were prescribed rest with two arms of a RCT comparing aerobic exercise with placebo-like stretching group (PG) and rest group (RG).	151 concussed adolescent athletes. Presented to clinic within 10 days of SRC, diagnosed by physician EG (N = 52, M: 28 F: 24) Age: 15.3 ± 2 Days since injury: 4.9 ± 2.2 PG (N = 51, M: 27, F: 24) Age: 15.4 ± 2 Days since injury: 4.8 ± 2.4 RG (N = 48, M: 36, F: 12) Age: 15.4 ± 1 Days since injury: 4.3 ± 2.0	EG: Daily aerobic exercise (treadmill or cycling) at prescribed target HR PG: Prescribed standardized stretching programme 20 min daily; Rest advice RG: Relative rest	Recovery: symptoms resolution to normal, confirmed by normal physical exam. Ability to exercise to exhaustion without symptom exacerbation Symptom resolution defined as PCS scale score of 7 points or fewer, for 3 consecutive days	Days from injury to recovery	• The PG recovered in 16 days, which was significantly delayed compared with EG in 13 days • The PG recovered in 17 days • Four percent of the EG, 14% of the PG, and 13% of the RG had delayed recovery

AE = aerobic exercise; ANS = autonomic nervous system; AR = active rehabilitation; BCTT = Buffalo Concussion Treadmill Test; CG = control group; EG = exercise group; HR = heart rate; PCSS = post-concussion symptom scale; RG = rest group; RPE = rate of perceived exertion; RTP = return to play; SCAT3 = sports concussion assessment tool 3rd edition; SRC = sport related concussion; SE = supervised exercise; VAS = visual analogue scale.

Active rehabilitation interventions

Duration, timing, intensity, and delivery of interventions

In five studies, the BCTT was the initial intervention to determine sub-symptom threshold.^29,30,32,33 Following initial BCTT symptom exacerbation test, Table 4 shows that these studies allowed participants to use a treadmill, stationary bicycle, walk, or jog at their sub-symptom threshold. Two studies used the BCTT as the main intervention throughout,^29,30 and others used a modified BCTT programme with an additional graded, individualized exercise programme based on exercise tolerance.³⁴ One study, Maerlender and colleagues,³¹ employed a stationary bicycle protocol. Sub-symptom threshold (80% of maximum HR), a valid and reliable test for symptom exacerbation,³⁵ was used to measure intervention outcomes in two studies.^32,33 A visual analogue scale (VAS) was used to determine individualized symptom exacerbation, with exacerbation defined as a ≥2 point³⁶ or ≥3 point³⁴ change, with initial symptom score taken before AR intervention began. Maerlender and colleagues³¹ used a VAS for symptom reporting, however they did not calculate a sub-symptom threshold and merely told participants to stop if they felt uncomfortable. Rate of Perceived Exertion (RPE) (6–20 on Borg Scale) and average HR taken every minute was another systematic method used to determine symptom exacerbation.^30,34,36

Table 4.

Summary of Interventions

Study	Treatment group (N)	Duration of treatment period	Timing (duration/frequency of each episode)	Procedure	Equipment	Supervision	Compliance
Hinds and colleagues30	40	Until asymptomatic and recovered based on evaluation of blinded physician	9 min at approximately same time each day	BCTT – exercise intensity increased each min until termination, RPE indicated after each min (6–20 on BORG Scale). HR recorded at rest, at exercise time zero, and average HR each min until test completion	Treadmill; Polar Electro Heartrate Monitor, Model FT1	NR	NR
Leddy and colleagues36	27	Twice 14 days apart	Until symptom exacerbation (≥3 point change from pre-treadmill resting symptom score on 1–10 VAS) or fatigue (RPE ≥ 17)	BCTT – 3.2–3.6 mph at 0% incline, increase incline after 1 min 1 degree, repeat each min, maintaining same speed until patient reaches symptom exacerbation or fatigue. HR, RPE (6–20, Borg Scale) and symptom score (1–10, VAS) assessed at each minute	Treadmill; Polar HR monitor Model #FIT N2965	2 CPR-trained research personnel	No significant difference between groups
Leddy and colleagues32	52	Until asymptomatic or 30 days	20 min a day or until symptom exacerbation every day	Stationary bike, treadmill, walk or jog to symptom exacerbation (HR – 80% of exacerbation from BCTT)	Treadmill or stationary bicycle; Polar H7 Bluetooth HR sensor and Fitness Tracker	NR	No significant difference between groups
Leddy and colleagues29	24	Until asymptomatic	20 min every day or until symptom exacerbation or voluntary exhaustion	Treadmill or stationary bicycle (HR – 80% of exacerbation from BCTT). 5 min warm up and 5–10 cool down.	Treadmill or stationary bicycle; Polar HR Monitor Model #FIT N2965	NR	Symptom scores missing: 6.8% N = 2 participants excluded for missing 3+ days
Maerlender and colleagues31	13	Asymptomatic or 50 days	20 min every day unless uncomfortable	Stationary bike at perceived exertion level of mild to moderate (0–6 on RPE scale)	Schwinn Airdyne Stationary Bicycle; Actical actigraphs to measure the amount of physical activity	Athletic trainer	N = 3 participants lost to lack of recovery > 50 days.
Popovich and colleagues34	26	Individualized based on patient exercise tolerance	At least 10 min every day; if symptom exacerbation of 3+ points of 1–10 VAS, decrease intensity. Stop if symptoms worsen.	Initial modified BCBT symptom exacerbation threshold test, followed by individualized Stationary bike, elliptical machine, dynamic rotational exercises using medicine balls, agility drills, sport specific exercises	Exercise space; Treadmill; Stationary bicycle; Elliptical, Medicine balls	Athletic trainer or physician	N = 2 patients lost to follow-up
Willer and colleagues33	52	Until asymptomatic or 30 days	20 min a day or until symptom exacerbation, every day	Stationary bicycle, treadmill, walk or jog to symptom exacerbation (HR – 80% of exacerbation from BCTT)	Treadmill or stationary bicycle; Polar HR monitor	NR	No significant difference in compliance

BCTT = Buffalo concussion treadmill test; HR = heart rate; NR = not reported; RPE = rate of perceived exertion; VAS = Visual Analogue Scale.

Overall, duration of exercise across the studies was variable with nine minutes,³⁰ 20 minutes,^29,31, ³², ³³ or individualized duration until symptom exacerbation^34,36 being used. Most studies required participants to exercise daily and finished the intervention when participants were asymptomatic,²⁹, ³⁰, ³¹, ³², ³³, ³⁴ however two of these had an arbitrary intervention cut-off of 30 days if not recovered, as mTBI was no longer considered acute.^32,33 Maerlender and colleagues³¹ ended the intervention at 50 days, whereas Leddy and colleagues'³⁶ intervention lasted 14 days.

Equipment and supervision

The majority (n = 6) of reviewed studies used a treadmill for AR, particularly for the initial BCTT test^29,30,32, ³³, ^34,36 (Table 4). Two studies required a treadmill for further visits.^30,36 Maerlender and colleagues³¹ required a Schwinn Airdyne Stationary bike for each visit, and the option of stationary bike, treadmill, elliptical, walking, or jogging was permitted in three studies.^29,32,33 HR monitors were utilized in five studies to monitor sub-symptom threshold levels.^{29,30,32,33,36} Each study except two required dedicated sport environments, such as a gym or clinic, however two studies allowed participants to walk or jog outside if they did not have access to sport facilities.^29,33 In the 3 studies that reported the personnel involved, participants were supervised by CPR trained supervisors, athletic trainers, and physicians.²⁹, ³⁰, ³¹, ³², ³³, ^34,36

Compliance and safety

There were no significant differences with compliance reported in any of the studies (Table 4). Participants were lost due to lack of follow-up,³⁴ lack of symptom reporting,²⁹ lack of recovery,³¹ and missed study days.³² Compliance was not reported in one study.³⁰ No serious adverse events occurred in any of the reviewed studies, such as death, events that required either inpatient hospitalization or the prolongation of hospitalization, are life-threatening, result in a persistent or significant disability/incapacity or result in a congenital anomaly/birth defect. One study had a non-serious adverse event where a participant had a severe increase in symptoms and withdrew from the study.³²

Main outcomes

Days to recovery and symptom severity

Table 3 shows that recovery and symptom severity were defined and measured in different ways within the reviewed studies, with some examining when subjective symptom reports (i.e., VAS), balance or cognitive tests returned to baseline recordings and others using time until the person returned to activities. However, the primary outcome measures for studies were generally similar, that is, time to recovery and symptom resolution (i.e., self-reported, objectively examined by physical examination or no symptom exacerbation with exertion). Overall, most reviewed studies reported reduction in recovery time with AR compared to control rehabilitation (i.e., stretching, SR, etc.). Specifically, three studies reported significantly fewer days to recovery from initial injury or clinic visit with AR compared to control,^32,33,36 with one other study also reporting non-significant reduction in recovery time²⁹ (Table 3). However, two studies reported no difference between AR and control groups for days to recovery,^31,36 with Maerlender and colleagues³¹ reporting that the AR treatment group had non-significant prolonged recovery.

Heart rate

Measurement of HR allows for non-invasive determination of autonomic nervous system (ANS) function after mTBI.³⁷ Five of the reviewed studies recorded participant HR during the intervention in order to record the threshold of symptom onset and ensure sub-maximal exercise (Table 4). Of particular relevance, Hinds and colleagues³⁰ found participants had lower HR at onset of exercise when concussed, similar to previous research.³⁸ Additionally, Leddy and colleagues³⁶ reported that this lower HR was associated with prolonged recovery time, which may indicate ANS dysfunction following mTBI.³⁸

Quality appraisal

Overall, LoE was 2 (N = 3) or 3 (N = 4) (Table 5).²⁶ RoB-2 and ROBINS-I quality appraisal was completed by two independent reviewers (BC, DP) with agreement made on scoring to minimize personal bias. RCT RoB was rated as low overall^31,32,36 (Table 6). There were some concerns regarding missing outcome data in two studies.^31,32 However, RCTs were generally methodically sound, as they included information regarding randomization, allocation concealment, valid and reliable outcome measurement, and appropriate selection of the reported results. Overall, RoB for the non-randomized studies was rated as low³⁴ or moderate^29,30,33 (Table 7). Due to the nature of including quasi-experimental and cohort studies, the main RoB was the lack of randomisation and risk of confounding within these studies. In each study, blinding was not possible. Additionally, the results could not be found as causal due to lack of control groups or randomisation.³⁹ Given the range in RoB of the reviewed studies (n = 7) caution must be taken when interpreting the results and applying to practice.

Table 5.

Level of Evidence Results

Study	LoE criteria	LoE score
Hinds and colleagues30	RCT	II
Leddy and colleagues36	Retrospective comparative trial	III
Leddy and colleagues32	RCT	II
Leddy and colleagues29	Cohort comparison arm of RCT	II
Maerlender and colleagues31	RCT	II
Popovich and colleagues34	Retrospective cohort study	III
Willer and colleagues33	Cohort comparison arm of RCT	III

LoE = level of evidence.

Table 6.

Assessing Risk of Bias in Randomized Control Trials using COCHRANE RoB

	Study
Domain	Leddy and colleagues32	Leddy and colleagues36	Maerlender and colleagues31
Risk of bias arising from the randomization process	Low	Low	Low
Risk of bias due to deviations from the intended interventions	Low	Low	Low
Missing outcome data	Some concerns	Low	Some concerns
Risk of bias in measurement of the outcome	Low	Low	Low
Risk of bias in selection of the reported result	Low	Low	Low
Overall risk of bias	Low	Low	Low

Table 7.

Assessing Risk of Bias in Non-Randomized included Studies using ROBINS-I

	Study
Domain	Leddy and colleagues29	Willer and colleagues33	Popovich and colleagues34	Hinds and colleagues30
Bias due to confounding	Low	Moderate	Moderate	Moderate
Bias in selection of participants into the study	Moderate	Moderate	Low	Low
Bias in classification of interventions	Low	Low	Low	Low
Bias due to deviations from intended interventions	Low	Low	Low	Low
Bias due to missing data	Moderate	Moderate	Low	No information
Bias in measurement of outcomes	Low	Low	Low	Low
Bias in selection of reported result	Low	Low	Low	Low
Overall risk of bias	Moderate	Moderate	Low	Moderate

Discussion

This review examined AR for acute mTBI and highlighted that AR is being increasingly studied, with variance in the parameters of AR being used. The aim of this review was to critically appraise evidence regarding AR interventions for acute mTBI rehabilitation with seven studies meeting the eligibility criteria. The included studies found that AR for adolescents with mTBI is safe and may have positive recovery outcomes with consideration of the following themes.

Intensity of active rehabilitation

Intensity has been cited as the most important aspect of an exercise programme.⁴⁰ All seven studies found that light (i.e., walking, swimming, stationary cycling) to moderate (i.e., sport-specific or non-contact drills) AR did not exacerbate symptoms and was well-tolerated in the acute mTBI period. Generally, included studies used a sub-symptom threshold as an individualized, moderate intensity, which is consistent with intensity protocols within the PCS and youth/adolescent (<18 years old) concussion literature.^11,41,42 This supports the findings of a large, prospective study in children/adolescents that determined that light AR within seven days of mTBI injury was associated with lower incidence of PCS compared to no AR participation.²⁰ However, due to the quality and observational design of reviewed studies, causation cannot be inferred due to potential confounding factors from self-reported symptoms. Additionally, the definition of “light” intensity between studies is unclear, and generally relies on self-reported, non-objective measures. For example, Willer and colleagues³³ argued that a light AR may involve stretching exercises, which has a similar recovery time to a SR group (17 days and 16 days to recovery, respectively), but significantly longer recovery time to a moderate AR group (i.e., incremented treadmill walking/running) (13 days to recovery), concluding that moderate AR may have favourable outcomes compared to both SR and light intensity AR. Further clarification and standardisation of terminology for AR intensity are required before clinical adoption of findings.

While it is currently contested whether light intensity AR is effective, it is also unclear whether high intensity AR facilitates positive recovery outcomes. Maerlender and colleagues³¹ found that high intensity AR significantly slowed time to recovery compared to the low and moderate intensity AR groups. Although high intensity AR appeared to negate any benefits to recovery, there was no report of worsening of symptoms or serious adverse effects by any intensity of AR group.³¹ However, the study generalizability was limited due to the small sample size (N = 28) of the study. Conversely, a review across age ranges (child to adult) by Baker and colleagues¹¹ suggested that high intensity AR may not induce serious adverse effects if it is incrementally introduced (i.e., start low and increase intensity within symptom tolerance). Therefore, this review highlights that larger, high-quality RCTs are required in adults (>18 years old) to clarify intensity of AR required to improve mTBI recovery in acute stages.

Lack of high-quality evidence using high or low intensity AR interventions limits the ability to determine whether AR truly decreases time to recovery, with more evidence needed. However, the current evidence from the reviewed studies and similar systematic reviews¹¹ shows that moderate, sub-symptom threshold intensity may be effective.

Type of active rehabilitation

Generally, a graded exercise protocol using sub-symptom threshold intensity may allow clinicians to determine an individualized and safe AR intervention. The most common mode of exercise in the reviewed studies was standardized, unimodal protocols, including treadmill and stationary bicycle exercise. This is largely due to the recognized reliability and validity of the BCTT and BCBT, and the accessibility of treadmill and stationary bicycle equipment, which are typically found at multiple types of health facilities, such as physiotherapist clinics, gyms, and physician offices.³⁶ Additionally, much of the early research of AR interventions comparing SR and AR for PCS participants employed similar unimodal protocols,^35,43,44 which acted as groundwork evidence for acute mTBI AR research.^44,45

Alternatively, multi-modal interventions included choice of treadmill, bicycle, walking, or jogging,^32,33 and a graded, progressive AR intervention utilizing an elliptical machine, dynamic rotational exercises using medicine balls, agility drills, and sport-specific exercises.³⁴ Other multi-modal interventions, such as non-contact sport-specific activities and self-reported activity^20,46 were excluded in the study selection process due to lack of specified and complete AR interventions. Many multi-modal interventions rely on retrospective study designs and self-reported levels of PA.^41,46 Generally, lack of intervention standardisation can increase RoB score, meaning results may not be directly related to outcomes. While multi-modal interventions may not be as effective as unimodal,⁴⁷ they still appear to have a positive effect on recovery outcomes.

Duration and frequency of interventions

Reviewed studies prescribed various durations and frequencies of aerobic exercise, between 9 and 20 minutes, or until symptom exacerbation. Frequency of each episode ranged from twice total in a 14 day period,³⁶ every day for 30 days,^32,33 50 days,³¹ or daily until symptom-free.^29,30,34 Due to heterogeneity between studies, it is difficult to determine optimal duration and frequency. Recent findings by Howell and colleagues^41,42 demonstrate that compliance to AR prescription in children with mTBI may be more critical to reducing time to recovery than the mode or intensity itself. This has implications for intervention frequency, suggesting that lack of compliance may negatively impact recovery outcomes. Overall, higher frequency of exercise during the acute period may be associated with lower symptom severity although it is possible that this is related to the fact that participants who experienced less severe concussion may exercise more, instead of AR reducing the severity of reported symptoms. However, this was not assessed in the current study, and supporting evidence is still required.

Time to recovery and delayed recovery

Three studies reported AR interventions significantly reduced time to recovery compared to stretching or control groups. Additionally, AR did not significantly increase symptom severity scores or prolong recovery. This supports a recent systematic review and meta-analysis which found that several modes and intensities of AR in the acute phase yield safe and positive outcomes.¹¹ Therefore, the current review adds to the findings, and extends the acute phase from 14 days to one-month post-mTBI.

Maerlender and colleagues³¹ found that while AR appears safe, it may not decrease time to recovery or reduce symptom scores. Alternatively, Leddy and colleagues³⁶ reported 4 out of 30 participants in the rest group experienced delayed recovery, while no participants in the AR group did. The same authors also reported that 4% of participants in their AR group experienced delayed recovery, whereas 14% of participants in a stretching group had delayed recovery (highlighting the benefits of AR).³² The quasi-experimental design by Willer and colleagues,³³ as an extension of Leddy and colleagues³² additionally found that the participants in the control group also experienced higher incidence of delayed recovery (13%) compared to the AR group (4%). These findings highlight the controversy of AR following mTBI, with some conflicting results that may stem from a lack of standardized approach in some research studies. Generally, this review found that the current literature supports an active approach to mTBI management, as there are no adverse outcomes, and it may decrease time to symptom resolution.

Conclusion

This review suggests that sub-symptom, progressive, and individualized AR intervention, may be well-tolerated and safe for acute mTBI. Overall, this review found that light, moderate, and progressive AR may improve time to symptom resolution and reduce delayed recovery and prevent PCS. This supports current evidence for sub-symptom threshold exercise for acute mTBI in adolescents. Additional research in the timing, duration, frequency, and mode of exercise, as well as intervention compliance and long-term follow-up is needed in the form of high quality RCTs.

Funding

This work was supported by Northumbria University as part of a doctoral research development fund programme (D Powell) and by the Private Physiotherapy Education Foundation (PPEF – #368; PI: Stuart). Dr Stuart is supported by grants from the Parkinson’s Foundation (PDF-FBS-1898, PDF-CRA-2073). Julia Das is supported by a Northumbria University PhD studentship in collaboration with Senaptec Ltd. Rachel Mason is supported by a Northumbria University PhD studentship in collaboration with DANU Sports. Lisa Graham is supported by a Northumbria University PhD studentship in collaboration with Head Diagnostics.

Key Messages

What is already known on this subject

Typical treatment following a mild traumatic brain injury involves advising patients to restrict their physical activity and rest, with potential to return to usual activities gradually as symptoms subside. Published guidance to support this clinical advise and protocol is limited, which leads to a lack of consistency within clinical practice.

What this study adds

To our knowledge, this study represents a comprehensive examination of active rehabilitation used within the acute stages of mild traumatic brain injury. The findings show that active rehabilitation that is undertaken within one month of mild traumatic brain injury is safe and decreased the time taken for symptoms to resolve. However, there is a lack of consensus on specific intervention protocols used and this needs further research before these findings can be used regularly in clinical practice.

Funding Statement

This work was supported by Northumbria University as part of a doctoral research development fund programme (D Powell) and by the Private Physiotherapy Education Foundation (PPEF – #368; PI: Stuart). Dr Stuart is supported by grants from the Parkinson’s Foundation (PDF-FBS-1898, PDF-CRA-2073). Julia Das is supported by a Northumbria University PhD studentship in collaboration with Senaptec Ltd. Rachel Mason is supported by a Northumbria University PhD studentship in collaboration with DANU Sports. Lisa Graham is supported by a Northumbria University PhD studentship in collaboration with Head Diagnostics.