Exercise Intolerance After Mild Traumatic Brain Injury Occurs in All Subtypes in the Adult Population

Abstract

Thematically grouped symptom clusters are present during the acute timeline of post-mild traumatic brain injuries (mTBI), representing clinical profiles called subtypes. Exercise intolerance has not been evaluated within the subtype classifications and, because guidelines support early submaximal aerobic exercise, further knowledge is required in regard to the exercise capabilities among the concussion subtypes. This cross-sectional study (n = 78) aimed to characterize the presence of exercise intolerance within the clinical subtypes and to explore performance on the Buffalo Concussion Treadmill Test (BCTT) in the adult subacute (2–12 weeks post-injury) mTBI population. All participants were evaluated using the BCTT to determine exercise tolerance. We first used the Neurobehavioral Symptom Inventory (NSI) questionnaire to assign each participant a primary subtype(s). To further explore all five subtypes (headache, cognitive, vestibular, ocular motor, and mood), participants were assessed using a multitude of thematically grouped assessments including self-reported questionnaires, clinical tests of vestibular and ocular motor function, balance function, and computerized cognitive testing. Thirty-seven (47%) subjects were exercise tolerant and 41 (53%) were exercise intolerant. There was no difference in the distribution of primary subtypes between the exercise tolerant and exercise intolerant groups. In addition, no significant differences were found between the exercise tolerant and exercise intolerant groups on other thematically grouped subtype assessments. The exercise intolerant group had a significantly higher resting heart rate (HR), lower percentage of age-predicted maximum HR achieved, lower Borg Rate of Perceived Exertion (RPE), and could walk on the treadmill for less time (lower duration) compared with the exercise tolerant group. The current findings suggest that exercise intolerance is common and pervasive across all five mTBI subtypes. A comprehensive mTBI assessment should include evaluation for exercise intolerance regardless of the primary clustering of symptoms and across patient populations. Therefore, early referral to physical therapists, athletic trainers, or medical clinics that can perform the BCTT may be helpful to initiate appropriate exercise prescriptions for patients with mTBI.

Introduction

Concussions, referred to as mild traumatic brain injuries (mTBI), represent a heterogeneous collection of symptoms with varying degrees of clinical presentations. An mTBI may be caused by direct traumatic injury to the head or elsewhere to the body with forces transmitted to the brain.¹ The result is a multi-faceted functional neuropathological disturbance thought to reflect changes in global microstructure, cellular injury, inflammatory cascades, and metabolic responses causing a diverse range of symptoms.^2–4

Recent research has suggested the presence of thematically grouped symptom clusters during the acute timeline post-injury, representing organized clinical profiles called subtypes.^5–9 Clinical utilization and analysis of the Rivermead Postconcussion Symptoms Questionnaire and other objective testing can assist in delineating concussion subtypes.^5,10 To date, five predominant subtypes have been suggested and include: (1) cognitive, (2) ocular-motor, (3) headache/migraine, (4) vestibular, (5) anxiety/mood, as well as two associated conditions: (1) sleep disturbance and (2) cervical strain.^5,10–12 The goal of this new micro-classification of mTBI was to encourage early initiation of targeted therapies (i.e., cognitive behavioral therapy for cognitive subtype classifications or physical therapy for vestibular subtype classifications) to improve the speed of recovery and decrease prolonged symptom resolution back to baseline.^11,12

Although not yet identified as a subtype, exercise intolerance has been extensively studied as a marker of prolonged recovery as well as a primary marker of physiological readiness in return to sport protocols.^13–15 Exercise intolerance is currently defined as the inability to exercise at or near age-appropriate maximum heart rate because of concussion symptom exacerbation.^2,16,17 It is well known that exercise intolerance after mTBI exists and is thought to be because of autonomic nervous system dysfunction and attenuated cerebrovascular physiology.^2,17–21 The presence of exercise intolerance after mTBI is assessed with a provocative graded aerobic treadmill test known as the Buffalo Concussion Treadmill Test (BCTT).^2,18,19,22 Patients are asked to perform a progressive exertion protocol to gradually increase cardiovascular demand to identify the threshold at which symptoms worsen. As a potential physiological biomarker of mTBI, the BCTT has been shown to be safe for assessing exercise tolerance after mTBI, and results are informative for the development of individualized periodized treatment plans.^22–24 Through BCTT evaluation, performing submaximal mild symptom exacerbation aerobic exercise has been reported to speed recovery in the acute phase (<16 days) of an mTBI.^3,13,18,21 Similarly, the early application of targeted aerobic exercise treatment (within 2–10 days of mTBI) utilizing the BCTT has shown a reduction in the incidence of persistent post-mTBI symptoms in adolescents beyond the 30-day timeframe.²⁵ Evidence also supports the reduction of persistent (>30 days post-injury) mTBI symptoms with the intervention of aerobic exercise prescribed by the BCTT.^16,26–28

Currently, exercise intolerance has been primarily studied on sport-related, adolescent mTBI. Many mTBIs, however, are not from sports injuries and include falls, motor vehicle accidents, and violence.^29,30 In addition, a recent study examining point of entry for patients with an mTBI at a large academic center found the mean age of patients to be 30.5 years (range 3–89 years).⁸ Patients often do not engage in physical therapy, where exercise tolerance is likely evaluated, for several months after injury.³¹ As a result, there is a large gap on assessing exercise tolerance for the diverse population of patients with mTBI seen in concussion clinics with a wide range of ages, various mechanisms of injury, and potentially more sedentary lifestyles than sport-related injuries.

Previous work has not evaluated exercise intolerance within the subtype classifications and, because guidelines support early submaximal aerobic exercise, further knowledge is required in regard to the exercise capabilities among varying subtype populations. Most subtype classification to date has been based solely on subjective symptom questionnaires rather than objective measures. With an emerging focus on using objective measures to classify deficit after mTBI, it may be important to consider objective measures of thematically grouped assessment tools for subtype classification. The purpose of this study was to (1) investigate exercise intolerance across subtypes using both subjective questionnaires and objective, thematically grouped assessment tools for classifying subtypes, and (2) explore performance on BCTT in the adult subacute mTBI population.

Methods

Participants

A convenience sample of 78 people with subacute mTBI (2–12 weeks post-injury) participated in this study (Table 1). Participants were part of a larger study investigating the effect of early physical therapy in people with subacute mTBI (ClinicalTrials.gov identifier: NCT03479541).³²

Table 1.

Participant Demographics for the Exercise Tolerant and the Exercise Intolerant Mild Traumatic Brain Injury Groups

	Exercise tolerant (n = 37)	Exercise intolerant (n = 41)	Group difference p value
Age (years)	37.5 (11)	38.62 (10.92)	0.07
Height (m)	1.68 (0.09)	1.67 (0.09)	0.61
Mass (kg)	71.60 (17.21)	69.33 (14.24)	0.52
Sex (M/F/Other)	7 M / 30 F	6 M / 33 F / 2 Other
BMI	25.2 (5.6)	24.8 (11.2)	0.74
NSI Total Score (out of 88)	31.7 (12.9)	38.5 (16.4)	0.04
Days since injury	75.5 (33.3)	73.9 (29.8)	0.67
Overall recovery (0-100)	65.14 (20.67)	59.76 (23.07)	0.29
Exercise recovery (0-100)	51.62 (26.56)	41.24 (29.72)	0.12
Injury mechanism (N and Percent)
Bike	2 (5.41%)	3 (7.32%)
Fall	7 (18.92%)	9 (21.95%)
Motor Vehicle Accident	14 (37.84%)	12 (29.27%)
Sport	8 (21.62%)	7 (17.07%)
Other	6 (16.22%)	10 (24.39%)

BMI, body mass index; NSI, Neurobehavioral Symptom Inventory.

Means and standard deviations are presented unless noted otherwise for injury mechanism.

Independent t test was used for testing group differences.

Participants were recruited from the Portland metropolitan and nearby areas, encompassing various locations such as hospitals, clinics, universities, community recreation centers, gymnasiums, sporting facilities, cafes, and public noticeboards. In addition, participants were also recruited from the Oregon Health & Science University concussion clinic, as well as affiliated and supporting medical clinics.

Inclusion and exclusion criteria and mTBI definitions have been described elsewhere.³² Briefly, participants were included if they: (1) had a diagnosis of mTBI (based on Veterans Affairs/Department of Defense criteria) made by a physician and were 2–12 weeks from their injury at enrollment,³³ (2) were between 18 and 60 years old, (3) had a graded symptom checklist total symptom severity score >15, and endorsed any symptoms on either headache, nausea, dizziness, blurred vision, or balance problems from the Sport Concussion Assessment Tool version 3, and (4) no more than minimal cognitive impairment (≤8 on Short Blessed Test).³⁴

The exclusion criteria consisted of participants (1) having other musculoskeletal, neurological, or sensory deficits that could explain dysfunction; (2) being in severe pain during the evaluation (≥7/10 subjective rating); and (3) being pregnant.

This study was approved by the Joint Institutional Review Board Committee of Oregon Health & Science University and Veterans Administration Portland Health Care System.

Protocol

Participants attended data collection sessions where demographics, self-reported questionnaires, clinical tests of vestibular and ocular motor function, exercise tolerance and balance function tests were collected.

All participants were evaluated using the BCTT to determine exercise tolerance.³⁵ The BCTT was modified to accommodate a wide range of age and baseline physical activity. Specifically, the initial treadmill speed was based on participant tolerance. Participants were classified as exercise tolerant if they were able to achieve 85% of their age predicted heart rate (HR) max (220-age) or rated perceived exertion ≥17/20 on the Borg Rate of Perceived Exertion (RPE) without a significant increase in symptoms.³⁶ Exercise intolerance was defined as an increase in symptoms (≥ 3/10) and an inability to achieve 85% HRmax or RPE ≥17/20.³⁶

Participants were asked to complete the Neurobehavioral Symptom Inventory (NSI) questionnaire, a 22-item test with each item rated from 0 (none) to 4 (very severe). The NSI comprises a total score and subscale scores, including vestibular, somatosensory, affective, and cognitive, which can be calculated from the 22 items.³⁷ The assessment has good internal consistency and stability.^38,39

Subtype assignment for exercise tolerance

We used the NSI questionnaire to assign each participant a primary subtype(s). The presence of a primary subtype was identified if the average of all items in that category was ≥2 (indicating a moderate to very severe effect on daily function).

If an average score was <2 in any given category, there was no subtype associated, and participants were classified as none. Of note, participants could have more than one subtype if those categories were ≥2 and were classified as having mixed subtypes. This categorization was consistent with procedures from other studies.^10,40–43 We used the following items from the NSI per subtype.

Headache subtype

We used the four items from the NSI—‘‘headaches,’’ ‘‘nausea,’’ ‘‘sensitivity to light,’’ and ‘‘sensitivity to noise’’—for the headache profile.

Cognitive subtype

We used the four items from the NSI—‘‘poor concentration, clumsy,’’ ‘‘forgetfulness, can't remember things,’’ ‘‘slowed thinking, difficulty getting organized, can't finish things,’’ and ‘‘fatigue, loss of energy, getting tired easily’’—for the cognitive profile.

Vestibular subtype

We used the two items from the NSI—‘‘feeling dizzy’’ and ‘‘loss of balance’’—for the vestibular profile.

Ocular motor subtype

We used the one item from the NSI—‘‘vision problems, blurring, trouble seeing’’—for the ocular motor profile.

Mood subtype

We used the four items from the NSI—‘‘feeling anxious or tense,’’ ‘‘feeling depressed or sad,’’ ‘‘irritability, easily annoyed,’’ and ‘‘poor frustration tolerance, feeling easily overwhelmed by things’’—for the mood profile.

Other outcome measures representing subtypes

Items from the NSI and the subscale scores were used in combination with other self-reported, clinical and instrumented assessments to represent the clinical subtypes presented below.

Headache subtype measurements

Headache Impact Test-6 (HIT-6)

The HIT-6 includes six self-rated items to measure the effect of headaches on daily life and overall function.⁴⁴ Each item is rated from never (6 points) to always (12 points), for a total score range of 36-78. Scores ≤49 represent little or no impact; 50–55 represent some impact; 56–59 represent substantial impact, and ≥60 indicate severe impact. The test has good reliability and construct validity.⁴⁴

Headache (NSI)

The NSI headache item is rated 0–4 (where 0 represents “none,” and 4 indicates “very severe”) and was used from the NSI to measure the impact of headache on daily life functioning. The single-item NSI headache measure is substantively similar to that provided by the HIT-6.⁴⁵

Cognitive subtype measurements

Automated Neuropsychological Assessment Metrics (ANAM)

ANAM is a computerized battery of neurocognitive tests examining attention, concentration, reaction time, memory, processing speed, and decision-making and provides a composite score. Lower scores indicate worse performance.⁴⁶ The test has been shown to have excellent test-retest reliability.⁴⁷

NSI Cognitive subscale

This is a subscale of the NSI that is summed from the following individual items: poor concentration, forgetfulness, difficulty making decisions, and slowed thinking. Scores range from 0 (no impact of cognitive symptoms) to 16 (severe impact of cognitive symptoms).³⁷

Vestibular subtype measurements

Vestibular/Ocular Motor Screening (VOMS) total change score

Participants were asked to subjectively report each symptom of headache, dizziness, nausea, and fogginess on a 0–10 scale at baseline (before performing the VOMS tool) and after completing each of the following tasks: horizontal and vertical smooth pursuits, horizontal saccades, vertical saccades, convergence, horizontal vestibulo-ocular reflex (VOR), vertical VOR, and visual motion sensitivity.⁴⁸

To calculate the change score, the sum of symptoms (headache, dizziness, nausea, and fogginess) were calculated after each task and then subtracted from the sum of baseline symptoms.⁴⁹ The total change score consists of the sum of the seven individual task change scores.

NSI Vestibular subscale

This is a subscale of the NSI that is summed from the following individual items: feeling dizzy, loss of balance, and poor coordination/clumsy. Scores range from 0 (no impact of vestibular symptoms) to 12 (severe impact of vestibular symptoms).³⁷

Sway area

Participants wore an inertial measurement unit (128 Hz - Opal V2; APDM Inc., Portland, OR) at the level of the fifth lumbar vertebra to measure postural sway area. Participants stood for 30 sec with their feet together on a foam surface, hands on hips, and eyes closed. Higher postural sway area values indicate worse balance.⁵⁰

Dizziness Handicap Inventory (DHI)

The DHI is a 25-item self-reported assessment to determine the self-perceived handicap of dizziness.⁵¹ Each item is scored 0 (no), 2 (sometimes), or 4 (yes) for a total of 100 points. Higher scores relate to higher levels of perceived handicap. The DHI has been shown to have excellent test-retest reliability and content validity in patients with mTBI.^52,53

VOR average gain (crHIT)

The computerized rotational head impulse test (crHIT) assesses horizontal VOR function.⁵⁴ The crHIT was performed in a computer-controlled rotational chair (Neuro Otologic Test Center [NOTC]; Neurolign USA LLC, Pittsburgh PA) that presented 12 impulsive, whole-body, earth-vertical axis rotations (six clockwise and six counterclockwise; direction randomized; peak velocity 150 degrees/sec) with instructions to fixate gaze on an earth-fixed laser dot in a dark room.

Eye movements were recorded (Neurolign I-Portal 2019) using infrared video-oculography (VOG) and analyzed (Neurolign VEST 2019 software). VOR eye movements were analyzed to compute VOR gain (ratio of slow phase eye velocity to rotation chair velocity) for each rotation and then averaged for each direction and across all directions.

Ocular motor subtype measurements

Saccade horizontal and vertical latency

Tests of the horizontal and vertical saccadic eye movement system were performed in the Neurolign NOTC system with head restrained in a dark room with infrared VOG recordings (Neurolign I-Portal 2019) of eye position. A laser dot visual target was displayed on the wall with participants instructed to maintain their gaze on the target as it jumped to different positions.

Twenty randomized target jumps occurred in the horizontal plane (within ±15 degree range) for the horizontal saccade test and in the vertical plane (within ±9 degree range) for the vertical saccade test with each target position maintained for a 1.7 to 2.3-sec duration. Eye movements were analyzed to determine the average onset latency following target displacement for each direction.

Smooth pursuit horizontal and vertical velocity gain

Tests of the horizontal and vertical smooth pursuit eye movement system were performed in the Neurolign NOTC system with head restrained in a dark room with infrared VOG recordings (I-Portal 2019) of eye position. A laser dot visual target was displayed on the wall with participants instructed to maintain their gaze on the target as it moved back and forth through a ±10 degree range.

Horizontal and vertical smooth pursuit tests were performed at 0.3 Hz (5 cycles). Eye movements were analyzed (Neurolign VEST 2019 software) to determine eye velocity gain (ratio of eye velocity to stimulus velocity after removing any saccadic eye movements).

Mood subtype measurements

ANAM Mood Scale

The ANAM Mood Scale includes individual scores on seven items: anxiety, anger, depression, fatigue, happiness, restlessness, and vigor.⁵⁵ Each dimension includes six items rated on a 7-point Likert scale of mood intensity, with higher values reflecting a greater degree of endorsement of each of component of mood. The ANAM Mood Scale has been shown to be valid and have excellent test-retest reliability.⁵⁵

NSI Affective subscale

This is a subscale of the NSI that is summed from the following individual items: fatigue, sleep, anxious, depressed, irritability, and frustration. Scores range from 0 (no impact of affective symptoms) to 24 (severe impact of affective symptoms).³⁷

Statistical analyses

We evaluated group (exercise tolerant vs. exercise intolerant) demographic differences using independent t tests. A chi-square test was used to assess whether there was a difference in the distribution of the primary subtypes between the exercise intolerant and exercise tolerant groups. To examine group differences in each subtype assessment, we used independent t tests, and significance values were corrected for multiple comparisons using a Bonferroni correction.

To explore whether there was a pattern of group differences among other subtype assessments (i.e., larger differences across assessments for one subtype assessment compared with another), Cohen d effect sizes (Cohen d_s) were calculated.⁵⁶ Cohen d_s were classified using the following scale: none (≤0.2), small (0.2–0.5), moderate (0.5–0.8), and large (> 0.8).⁵⁷

To gain a deeper understanding on how exercise intolerant people performed on the BCTT, we used summary statistics to evaluate performance on the BCTT (i.e., max HR achieved and duration maintained on the BCTT). To explore how baseline symptoms were related to the BCTT performance, Pearson product-moment correlation coefficients were calculated to assess the relationship between the maximum HR achieved and NSI total score for both groups.

Results

Exercise intolerance in people with subacute mTBI

Thirty-seven (47%) subjects were exercise tolerant and 41 (53%) were exercise intolerant. The exercise tolerant group and exercise intolerant group were similar for age, height, mass, sex, body mass index, and days since injury (Table 1). The most common mechanism of injury for both groups was motor vehicle accident: exercise tolerant (37.84%) and exercise intolerant (29.27%) (Table 1). While all participants were >2 weeks after their mTBI, the exercise intolerant group had a significantly higher NSI total score compared with the exercise tolerant group (Table 1).

Primary subtype identification of exercise intolerance

There was no association or difference in the distribution of primary subtypes between the exercise intolerant and exercise tolerant groups (p = 0.43; Table 2).

Table 2.

Primary Subtype Assignment Based on the Neurobehavioral Symptom Inventory for the Exercise Tolerant and the Exercise Intolerant Mild Traumatic Brain Injury Groups

Groups	Subtype
	Cognitive	Headache	Ocular-Motor	Mood	Vestibular	Mixed	None
Exercise intolerant	7 (17.07%)	4 (9.76%)	6 (14.63%)	8 (19.51%)	3 (7.32%)	6 (14.63%)	7 (17.07%)
Exercise tolerant	7 (19.44%)	2 (5.56%)	3 (8.33%)	4 (11.11%)	1 (2.78%)	5 (13.89%)	14 (38.89%)
Total	14	6	9	12	4	11	21

Exploring patterns of exercise intolerance across subtypes using other subtype-related measures

There were no significant differences between the exercise tolerant and exercise intolerant groups on subtype-related assessments after correcting for multiple comparisons (all p's > corrected alpha of 0.002; Table 3). Both groups had impairments in all assessments when compared with normative values from the literature (Table 3).

Table 3.

Assessments for Each Subtype for the Exercise Tolerant and Intolerant Groups

Subtype related-assessments	Exercise tolerant	Exercise intolerant	Group difference p value	Normative value
Headache
Headache Impact Test (HIT-6)	57.38 (8.33)	61.46 (8.65)	0.037	42.42 (5.76)59
Headache (NSI)	2.06 (1.06)	2.56 (1.2)	0.057	0.3 (0.7)60
Cognitive
Cognitive composite score (ANAM)	0.79 (1.03)	0.48 (0.89)	0.157	0.38 (1.06)61
Cognitive subscale score (NSI)	6.83 (3.92)	8.56 (3.35)	0.041	1.01 (1.68)62
Vestibular
Dizziness Handicap Inventory (DHI)	25.68 (18.67)	33.85 (18.90)	0.059	0.66 (2.20)63
Total Change score (VOMS)	9.95 (11.04)	12.9 (16.82)	0.37	2.26 (4.73)64
Vestibular subscale score (NSI)	3.58 (2.38)	4.8 (2.39)	0.028	0.62 (1.31)62
Sway area	0.67 (0.6)	0.61 (0.44)	0.602	0.38 (0.2)50
VOR average gain (crHIT)	1 (0.07)	1 (0.06)	0.983	0.96 (0.03)65
Ocular motor
Saccade horizontal latency	209.88 (27.23)	206.02 (27.51)	0.562	170 (20)66
Saccade vertical latency	221.07 (28.48)	211.58 (28.21)	0.17	180 (20)66
Smooth pursuit horizontal gain	0.90 (0.13)	0.93 (0.05)	0.19	-
Smooth pursuit vertical gain	0.76 (0.14)	0.79 (0.11)	0.271	-
Mood
Anger scale (ANAM)	12.38 (14.08)	15.55 (19.31)	0.416	13.6 (19)67
Anxiety scale (ANAM)	18.38 (16.23)	28.5 (23.32)	0.031	15.2 (16.4)67
Depression scale (ANAM)	17.68 (17.67)	26.8 (23.78)	0.062	15.5 (20.1)67
Fatigue scale (ANAM)	39.49 (21.72)	46.93 (23.31)	0.153	27.5 (21.6)67
Happiness scale (ANAM)	41.05 (19.95)	33.98 (18.39)	0.109	-
Restless scale (ANAM)	17.65 (17.12)	27.6 (23.76)	0.04	17.3 (17.9)67
Vigor scale (ANAM)	27.14 (17.26)	24.43 (16.19)	0.479	50.9 (23.5)67
Affective subscale score (NSI)	10.69 (4.65)	12.15 (5.83)	0.235	2.27 (3.1)62

NSI, Neurobehavioral Symptom Inventory; ANAM, Automated Neuropsychological Assessment Metrics; DHI, Dizziness Handicap Index; VOMS, Vestibular/Ocular Motor Screening; crHIT, computerized rotational head impulse test.

Outcomes presented as means and standard deviations. Normative values were based on data from existing literature (reference indicated by superscript in Normative value column).

Cohen d_s effect sizes ranged from 0.006 to 0.51 representing no effect to a moderate effect in the differences between groups (Fig. 1). There was no pattern, however, to the magnitude of group difference across concussion subtype-related assessments.

FIG. 1.

Effect size calculations for each subtype-related assessment between the exercise tolerant and exercise intolerant group.

Exploring performance measures on the BCTT

The exercise intolerant group had a significantly higher resting HR, lower percentage of age-predicted maximum HR achieved, lower RPE, and could walk on the treadmill for less time (lower duration) compared with the exercise tolerant group (Table 4). There were no adverse events during the BCTT. The maximum HR achieved during the BCTT was not correlated with baseline symptoms subjectively recorded on the NSI for both the exercise intolerant (r = 0.184, p = 0.249) and exercise tolerant groups (r = 0.291, p = 0.085) (Fig. 2).

Table 4.

Buffalo Concussion Treadmill Test Performance Outcomes for the Exercise Tolerant and the Exercise Intolerant Mild Traumatic Brain Injury Groups

	Exercise tolerant (n = 37)	Exercise intolerant (n = 41)	Group difference p value
BCTT outcomes
Resting HR (BPM)	74.75 (17.66)	81.61 (11.37)	0.048
HR Achieved (%)	81.03 (0.08)	67.07 (0.09)	< 0.001
RPE	15.78 (1.70)	13.08 (2.13)	< 0.001
Pre symptoms	3.59 (3.18)	4.90 (3.44)	0.043
Post symptoms	3.73 (3.48)	8.20 (3.57)	< 0.001
Change in symptoms	0.14 (1.60)	3.29 (0.78)	< 0.001
Duration (min)	10.92 (3.19)	6.56 (2.61)	< 0.001

BCTT, Buffalo Concussion Treadmill Test; HR, heart rate; RPE, Borg Rate of Perceived Exertion.

Data presented as means and standard deviations.

t test was used for testing group differences.

FIG. 2.

Correlation between percentage of age predicted maximum heart rate achieved and Neurobehavioral Symptom Inventory (NSI) total score for the exercise intolerant group and the exercise tolerant group.

Discussion

This study was the first to evaluate the presence of exercise intolerance across various mTBI subtypes in a diverse adult mTBI population. Utilizing multiple thematically grouped self-reported assessments, clinical scales, and objective measurements, we were able to evaluate all of the primary mTBI subtypes identified to date, including mood, cognitive, ocular-motor, headache, and vestibular. Our main finding was that our cohort of adults with diverse subtypes showed no difference in presence of exercise intolerance versus exercise tolerance across subtypes. This was observed regardless of whether we classified the subtypes based on a primary designation from a symptom-based tool (NSI) or considered a more comprehensive range of both subjective and objective measures for subtype-themed assessment tools. Therefore, our study supports the recommendation that patients, regardless of subtype classification, may benefit from an evaluation of exercise tolerance as part of their clinical examination. This can provide prognostic value in clinical care because it has been shown to provide insight into the possibility of prolonged recovery.²

This is the first study to assess exercise intolerance in an adult population with a more diverse mechanism of injury. At present, the BCTT is primarily used in the young and sport-related mTBI population.^15,24 Specifically, a recent systematic review summarized that 64% of the studies were performed in athletes, only 18% of the studies included only non-athletes, and 18% had a mixed population.⁵⁸ Similarly, 55% of the studies included only adolescents (approximately 15 years old) compared with 36% of the studies included only young adults (between 20–30 years old on average), and 9% of studies had a mixture of young adults and adolescents.⁵⁸ In our adult cohort, the mean age was approximately 38 years old and was primarily a non-athletic population (19% had a sport-related injury across both exercise tolerant and intolerant groups). We had no adverse events reported in our study that supports the notion that aerobic exercise is safe in adult mTBI patients with diverse injury mechanisms.

Our finding that exercise intolerance can be present across all subtypes has important clinical implications. We recently published a retrospective chart review on points of entry for patients after sustaining an mTBI and found that the primary point of entry is through the emergency department and primary care.⁸ The BCTT is primarily performed later in the context of a specialized sport medicine clinic suggesting that, unless people were referred to a specialized clinic, exercise intolerance could go undiagnosed. Further, a person with a primary subtype of mood or cognition might never get seen by an athletic trainer or physical therapist and therefore would not be assessed for exercise intolerance. An important reason to identify exercise intolerance is that subthreshold aerobic exercise can reduce symptoms after mTBI.^3,13,15 As a result, prescribing an appropriate level of aerobic exercise may facilitate a faster rate of return to activity after mTBI.

Our results showed that no specific subtype consistently exhibited a difference between the groups, even when considering a range of diverse assessments such as subjective, computerized, instrumented, and laboratory-based evaluations for subtype classification. Specifically, we found no clustering or phenotypic collection among the metrics used to identify subtypes. Similarly, there was no difference in the distribution of primary subtypes (based on the NSI questionnaire) between the exercise tolerant and exercise intolerant groups. Both groups had a similar percentage of participants who had more than one subtype (exercise intolerant: 15% and exercise tolerant: 14%). The percentage of participants without a subtype (i.e., no category on the NSI reached a level of moderate symptom reporting), however, was higher in the exercise tolerant group (39%) compared with the exercise intolerant group (17%). One potential reason for this difference could be attributed to how we defined the presence of a primary subtype. Specifically, to reach the level of primary designation, one had to score at the moderate level of symptoms. If individuals had very mild symptoms, they would not qualify for a primary subtype and would therefore be classified as having no subtype. This was the case for 21 individuals. Overall, our findings support the inability of subtype classification to identify those who are exercise tolerant versus intolerant. We demonstrated that both groups have deficits across all subtype-related measures based on existing normative values. Despite these deficits, there was no difference after correcting for multiple statistical tests between exercise tolerant and intolerant groups across the full range of subtype assessment tools we used. This is important because there are no clear subtype assessment tools that help identify exercise intolerance. As a result, the BCTT is a valuable assessment that can identify autonomic dysfunction related to mTBI.²⁷

Total mTBI symptoms (NSI) were unrelated to the maximum HR achieved during the BCTT in our study, which suggests that evaluating symptoms by using a questionnaire alone cannot provide direction on whether an individual is exercise intolerant. In addition, self -reported exercise tolerance during recovery did not differ between groups. Therefore, exercise intolerance cannot be identified based on self-perceived exercise recovery. We found that although the exercise tolerant people could exercise for longer without significant symptom increase and achieve higher HRs, they may also be very symptomatic at other periods during their day. On the contrary, exercise intolerant people could be unable to complete the BCTT because of symptom increase with exercise, yet have low symptom burden when assessed with the NSI. Therefore, our findings indicate that exercise tolerance should be assessed in people with mTBI regardless of their symptom severity.

This study has some limitations because we focused on exercise tolerance identified on the BCTT rather than autonomic dysfunction more broadly. We did not utilize any direct physiological measure in regard to subtype classification or exercise response. Other studies have discussed various methods to evaluate autonomic dysfunction (through evaluation of vagal tone, cerebral vasculature changes, and reactivity),²³ but we did not perform this testing. While we found that the exercise intolerant group had high scores on the anxiety scale (from the ANAM test), we did not use a specific anxiety scale to quantify anxiety. Future research should investigate the potential role of anxiety in exercise intolerance after mTBI. Further, a selection bias might have been present in this study because it was conducted within a research study context and focused on individuals with self-reported balance symptoms. Consequently, there is a possibility of an overrepresentation of participants classified with a vestibular subtype. Contrary to expectations, however, our findings did not support this assumption, because we also identified several deficits unrelated to balance. Finally, because of the complicated recommendations for subtype identification,⁵ we utilized multiple testing metrics to identify each individual subtype. Therefore, using multiple testing metrics could create challenges when trying to identify specific subtypes.

Conclusions

The important findings from our study suggest that exercise intolerance is common and pervasive across all five mTBI subtypes. A comprehensive mTBI assessment should include evaluation for exercise intolerance regardless of the primary clustering of symptoms and across patient populations. Therefore, early referral to physical therapists, athletic trainers or medical clinics that can perform BCTT may be helpful to initiate appropriate exercise prescriptions for patients with mTBI that ideally will improve outcomes across all subtypes. Future research should examine the neurophysiological processes mediating exercise intolerance after mTBI, efficient and accurate diagnostic testing to identify these processes, the impact of current rehabilitation protocols on recovery, and the development of unique new rehabilitation protocols to improve mTBI treatment efficacy.

Authors' Contributions

PA: Conceptualization, Methodology, Formal analysis, Investigation, Writing - Original Draft, Writing - Review & Editing, Visualization

KRC: Methodology, Software, Formal analysis, Investigation, Writing - Review & Editing

JLW: Methodology, Formal analysis, Investigation, Writing - Review & Editing

JDS: Writing - Original Draft, Writing - Review & Editing

JCC: Writing - Review & Editing

LAK: Conceptualization, Methodology, Resources, Writing - Review & Editing, Supervision, Funding acquisition

Funding Information

This work was supported by the Assistant Secretary of Defense for Health Affairs endorsed by the Department of Defense, through the Congressionally Directed Medical Research Program under Award Number W81XWH-17-1-0424. An integrated SQL database at Oregon Health & Science University has housed all the data and is supported by the Oregon Clinical and Translational Research Institute funded by a grant from the National Center for Advancing Translational Sciences (NCATS), National Institutes of Health, through Grant Award Number UL1TR002369.

Author Disclosure Statement

No competing financial interests exist.

Transparency, Rigor, and Reproducibility Summary

The study was pre-registered at ClinicalTrials.gov (NCT03479541). The analysis plan was registered prior to beginning data collection at the Physical Therapy & Rehabilitation Journal (https://academic.oup.com/ptj/article/100/4/687/5707558). Participants were blinded to results of the other assessments throughout the study, even after primary clinical observations were complete. All equipment and software used to perform acquisition and analysis are widely available from companies. The key inclusion criteria (e.g., primary diagnosis or prognostic factor) are established standards in the field. There is no current or planned replication studies ongoing to our knowledge. De-identified data from this study are available in a FITBIR data repository (https://fitbir.nih.gov/) and can be accessed at http://doi.org/ 10.23718/FITBIR/1518823 under a creative commons attribution license (CC-BY 4.0). The authors agree to provide the full content of the manuscript on request by contacting the corresponding author.