Use of the Buffalo Concussion Treadmill Test in Community Adults with Mild Traumatic Brain Injury (mTBI)

Abstract

Introduction

The Buffalo Concussion Treadmill Test (BCTT) is used to establish exercise tolerance for rehabilitation and identify injury subtypes for youth athletes after mild traumatic brain injury (mTBI). Its utility in adult community members is unknown.

Objective

Primary: Describe how adults with and without mTBI tolerate the BCTT. Secondary: Explore relationships between baseline factors, mTBI-related symptoms, and BCTT duration.

Design

Prospective, observational, longitudinal

Setting

Academic medical center

Participants

N=37 adults treated in a level 1 trauma center emergency department with mTBI; N=24 uninjured controls (UC).

Interventions

N/A

Main Measures

Participants completed two visits three weeks apart (1 week and 1 month post-mTBI) including a 15-minute BCTT, the Rivermead Post Concussion Symptoms Questionnaire, and pre-injury International Physical Activity Questionnaire. Analyses characterized BCTT response and associations between baseline factors, RPQ scores, and BCTT duration.

Results

Persons with mTBI discontinued earlier than UC at 1-week post-injury utilizing standard discontinuation criteria for exercise intolerance. The percentage of mTBI participants with signs of possible mTBI-related intolerance was 55.6% at 1 week (36.1% for mTBI-related symptom exacerbation, 19.4% for exertion/fatigue before reaching 85% of one’s age-predicted maximum heart rate [HR]) and 48.0% at 1 month (40.0% mTBI-related symptom exacerbation, 8.0% exertion without reaching the target HR). Thirty percent of UCs completed the BCTT at both assessments. UCs met discontinuation criteria for increased non-specific symptoms (e.g., pain/general discomfort and increased Visual Analogue Scale ratings; 39–61%) and physical exertion (9–26%). Shorter duration was associated with higher BMI (r=−0.42 – −0.45), shorter height (r=0.22 – 0.29), female gender (r=−0.26 – −0.27), and greater RPQ symptoms (r=−0.28 – −0.47).

Conclusion

The BCTT exacerbates mTBI-related symptoms in adult community members. Participant characteristics and non-injury factors influence performance. The findings imply the BCTT could be useful in clinical assessments of adults with mTBI. Interpretation should account for the unique characteristics of non-athletes.

Introduction

There are over 2.8 million TBI-related emergency department (ED) visits, hospitalizations, and deaths in the U.S. annually.¹ Mild traumatic brain injury (mTBI), a subtype of TBI classically defined by admission Glasgow Coma Scale (GCS) scores of 13–15, accounts for roughly 94.5% of all brain injury-related emergency department visits.² However, the phrase “mild” can be misleading because mTBI causes a diverse array of physical, cognitive, and emotional symptoms that impact life function and quality.^3,4 These symptoms can contribute to further functional impairment; in fact, 15–25% more adults with mTBI report persistent symptoms for up to a year when compared to adults with other traumatic injuries.^5,67 Unfortunately, many adult level 1 trauma center patients with mTBI go undiagnosed and receive no follow-up care for their injuries; these are missed opportunities to better manage mTBI in the community and support patients’ return to normal life.

Prescribed exercise may be a promising way to help persons with mTBI recover more quickly. Historically, mTBI patients were treated with bed rest and reassurance. Recommendations to rest may have stemmed from emerging understanding of the physiological consequences of mTBI and observations that persons with recent mTBI display physical and neuropsychological impairments and increased risk of repeat mTBI.⁸ Mild TBI causes a mismatch between energy supply and demand in the brain, with an initial hypermetabolic state paired with relatively low cerebral blood flow (CBF).^9,10 These neurophysiological changes have downstream consequences for other bodily systems, such as the cardiovascular system due to altered autonomic nervous system (ANS) function.¹¹ Persons with recent mTBI often experience symptoms and objective evidence of ANS dysfunction in response to physical exertion and stress,^10,12,13 a problem proposed to stem from disrupted cerebral vascular function and its effect on other systems.⁹

However, recent research has revealed that strict rest or “cocoon therapy” may worsen mTBI-related symptoms and prolong recovery, sparking a paradigm shift in the clinical management of sport-related mTBI.¹⁴ Consequently, clinical management guidelines for sport-related mTBI have been revised to emphasize brief periods of relative rest followed by more active rehabilitation approaches.¹⁵ Although supporting evidence of the physiological mechanisms is needed, it has been proposed that appropriately timed and dosed physical activity can help correct the aforementioned metabolic imbalances and ANS dysfunction that purportedly contribute to intolerance of physical activity in the compromised acute mTBI state.^16–19 There is now strong (level I) evidence in adolescents that aerobic exercise initiated within 10 days of injury improves symptom outcomes from sport-related mTBI, and much active research is ongoing to further develop evidence-based interventions to rehabilitate athletes with mTBI.^16,17,19,20 These findings raise the possibility that exercise may be helpful to manage mTBI in the broader adult community, yet data are currently lacking to support this practice.

Considering the slower and often more incomplete clinical recovery of the general community mTBI population,^7,21,22 there is a need to evaluate the degree to which tools and management strategies being used in sport-related concussion could inform the clinical management of the broader mTBI population. In individuals with sport-related concussion, exercise testing—such as with the Buffalo Concussion Treadmill Test (BCTT)—has been validated as a safe assessment in the acute post-mTBI period. The primary uses of the BCTT are to identify the right dose (intensity/duration) of exercise to prescribe to individuals with mTBI and to identify “physiological mTBI,” a subtype of mTBI characterized by exercise intolerance considered reflective of ANS disruptions.

The primary objective of the present study was to explore the utility of using the BCTT in a more diverse adult mTBI sample. We assessed prospectively recruited emergency department patients with mTBI at 1 week and at 1 month post-injury with a modified BCTT and characterized their clinical (symptom) response. A smaller sample of uninjured community controls (UC) were also given the same BCTT protocol to obtain data on the response of a general adult population to the task. Our primary aim was to describe the degree to which the BCTT elicited symptoms (e.g., mTBI-related symptoms, non-mTBI symptoms, feelings of high physical exertion) necessitating early discontinuation of the test in persons with mTBI. We hypothesized that the BCTT would commonly elicit mTBI-related symptoms, in community adults, similar to what has been observed in youth sport-related mTBI samples. Secondarily, we examined the association between non-injury factors (e.g., demographics, physical activity) and tolerance of the BCTT (BCTT duration) as well as the association between BCTT duration and the magnitude of mTBI symptoms over time. Finding substantial associations, for example, between patient factors and BCTT tolerance may indicate lower validity of the BCTT to detect mTBI-related exercise intolerance in some subgroups or indicate a need to develop separate interpretation guidelines for the BCTT in some subgroups. The findings may inform understanding how the BCTT might be used to manage mTBI in community adults and facilitate the translation of advances in sport-related mTBI treatment to the broader adult mTBI community.

Methods

Participants

This prospective cohort studyenrolled individuals with acute mTBI (N=37) treated at our Level 1 trauma center emergency department and uninjured controls (UC, N=24) from the community. Concurrent peripheral injuries diagnosed at discharge are included in eTable 1. One mTBI participant was dropped from analysis due to withdrawal before treadmill testing. Controls were recruited to be friends of mTBI participants and though community advertisements, with a goal to obtain group-level matching on age and gender. The Institutional Review Board at the principal investigator’s institution approved the study and all participants provided written informed consent.

Inclusion and Exclusion Criteria

Inclusion criteria for all participants were: age 18–65 years old; English-speaking; able to provide informed consent; have no history of concussion or TBI in the 6 months prior to injury; no history of serious cardiovascular, neurologic, autonomic or endocrine disease; no medical contraindications to participate (including no peripheral injuries that would preclude doing treadmill testing); and not currently pregnant. Further, the mTBI group needed to be 7 ± 3 days from injury at the time of enrollment and meet the American Congress of Rehabilitation Medicine definition of mild traumatic brain injury (mTBI) published in 1993, which requires evidence of a disruption in normal physiological brain function after head trauma. Per this definition, all participants demonstrated (via medical records) or reported (via structured clinical interview with research staff) some alteration in mental status after head trauma, but where any loss of consciousness (LOC) was no more than 30 minutes, Glasgow Coma Scale (GCS) scores were 13–15, and posttraumatic amnesia was no more than 24 hours.²³ To restrict the sample to “uncomplicated” mTBI (i.e., concussion), we additionally required that there be no acute intracranial findings if any clinical neuroimaging was performed.

Assessment Protocol

Participants were assessed at 1-week and 1-month post-injury (or twice 3 weeks apart for UCs). Each assessment took about 2 hours and comprised a variety of questionnaire, interview, and performance-based assessments. Ratings of pre-injury physical activity and mTBI-related symptoms were provided before the BCTT was administered; therefore, ratings were not influenced by recent physical exertion.

Modified Buffalo Concussion Treadmill Test (BCTT):

The BCTT was developed to measure mTBI-related physiological dysfunction and exercise intolerance in individuals with mild traumatic brain injury (mTBI), especially adolescents.¹⁶ For example, when used to dose prescribed exercise, the BCTT can be used to establish the duration at which aerobic exercise induces mTBI-related symptom exacerbation or extremely high perceived exertion, providing a maximum exercise duration for initial exercise prescriptions. In the present study, participants walked on a treadmill at 3.2 mph starting with a 0-degree incline, and the incline was increased 1 degree every minute until a discontinue criterion or the test endpoint (15 minutes) was reached. At the end of each minute, participants were asked to report their level of exertion using the Borg Scale (range 6–20, where higher ratings reflect greater perceived exertion). Severity of any mTBI-related symptoms was further monitored using the Visual Analog Scale (VAS). The VAS is a scale used evaluate discomfort or pain levels, ranging from 0–10. UCs were asked to use the VAS to rate any physical symptoms such as pain or headache they experienced during the assessment.

This protocol was adapted from the original protocol in three ways to account for untrained adults: (1) the starting speed was held constant at 3.2 mph (whereas in the original protocol some participants start at a faster speed), (2) the maximum duration of the task was shortened from 20 to 15 minutes, and (3) to accommodate saliva collection not relevant to this paper, the treadmill was paused for 2 minutes after 5 minutes of walking and then restarted. Discontinuation criteria included: Borg rating of 18 (exhaustion) or higher, increase in VAS score of 3 points from the pre-test resting score, any rapid progression of complaints reported to the examiner (e.g., severe focal pain), or report of feeling unable to continue the test safely.

Examiners recorded the reasons for discontinuation. The primary BCTT outcomes were task duration (in minutes) and the behavioral reason for discontinuing the test (mTBI-related symptom exacerbation, other symptom exacerbation, exertion, other discontinuation, or didn’t discontinue). Reasons for discontinuation were recorded in a categorical fashion by examiners and verified when possible by a second rater who reviewed notes on case report forms and video recordings of exams. When the examiner recorded the reason for discontinuation as significant symptom exacerbation, symptoms were classified by the first and senior author (AD, LN) as mTBI-related or non-mTBI-related based on review of examiner notes and video recordings of the exams. Symptoms such as headache and dizziness/imbalance were considered possibly mTBI-related. Conversely, symptoms such as bodily pain were categorized as peripheral in origin. It is important to note that no controls reported the types of symptoms considered to be mTBI-related. Discontinuation from exertion was considered when subjects reported a score ≥18 on the Borg scale.

In addition to reporting the behavioral reasons for discontinuation, we performed secondary analyses that considered heart rate (HR) data monitored at 2-minute intervals during the test using the Tango M2 Stress Test Monitor. The Tango Monitor was designed to overcome noise, motion, and physical difficulties associated with exercise and cardiac stress testing. In secondary analyses, participants who discontinued the BCTT prematurely due to exhaustion (i.e., Borg discontinuation threshold) and who did not achieve at least 85% of age-predicted maximum HR at the time of discontinuation were considered (alongside those who experienced mTBI symptom exacerbation) to have signs of mTBI-related exercise intolerance.²⁴

Rivermead Post Concussion Symptoms Questionnaire (RPQ):

The RPQ is a validated self-report inventory of symptoms commonly experienced after mTBI.²⁵ This questionnaire contains 16 items ranging from 0–4 (0=not experienced at all, 1=no more of a problem [than pre-injury], 2=A mild problem, 3=A moderate problem, 4=A severe problem). Participants were prompted to report the symptoms experienced over the past 24 hours prior to testing. Ratings of 1 are treated as 0 for summation; total scores can range from 0–64.

mTBI Recovery Interview.

Participants were asked to list their mTBI-related symptoms via interview and whether their symptoms were resolved (Y/N).

International Physical Activity Questionnaire (IPAQ):

The IPAQ is a physical activity questionnaire designed to assess physical activity of adolescents and middle-aged adults (15–69 years old).²⁶ In the short form used in the present study, participants reported the duration of activity at four intensity levels (vigorous, moderate, walking, sitting) in the 7 days prior to injury (or past 7 days for UCs) from occupation, exercise, recreation, etc. Activity duration is reported in metabolic equivalent of task (MET) units, and participants’ activity levels were categorized as inactive, minimally active, and highly active per the definitions suggested by Craig et al.²⁷

Statistical Analysis

Statistical analyses were conducted using IBM SPSS Statistics version 27. Sample demographic and injury characteristics were summarized as means (standard deviations) or frequencies (percentages). Group comparisons in sample characteristics were performed using Mann-Whitney U Tests and Fisher Exact Tests.

A linear mixed effects model was used to evaluate the impact of Group, Time, and Group x Time on the duration of the BCTT task; incorporating a random intercept for each participant and Sidak correction to correct follow-up tests for multiple comparisons. BCTT duration was reasonably normally distributed but is displayed using boxplots to facilitate clinical interpretability. Reasons for discontinuation were summarized descriptively with percentages.

To address our secondary aim, we quantified the association of BCTT duration with non-injury variables of age, gender, body mass index (BMI), and ratings of pre-injury physical activity using partial correlations for the full sample, controlling for the effect of group. This was done to maximize sample size for correlational analysis; sensitivity analyses performed within group verified similar results to the pooled-sample correlations reported. For consistency, Pearson (r) correlations were used to quantify the degree of association, at each timepoint separately, between BCTT duration and mTBI-related symptom severity (RPQ total score) within the mTBI group. Sensitivity analyses using nonparametric (Spearman) correlations for RPQ associations found similar results. Finally, a post-hoc multivariable general linear model was used to identify the independent predictors of BCTT duration at 1 week post-injury to follow up on the finding that multiple non-injury variables were associated in a bivariate fashion with BCTT duration at this timepoint.

Results

Participant Characteristics

Table 1 provides the demographic and injury characteristics of the sample. Groups were well matched on gender, BMI, socioeconomic status (SES), and type of health insurance type. Persons in the UC group were an average of 7 years older than mTBI participants (p = 0.014). Moreover, the mTBI group was more often non-white and had a lower mean education (M difference = 2 years). The distribution of participants with mTBI by cause of injury was: Motor Vehicle/Traffic Collision (77.8%), Fall (11.1%), Assault (2.8%), Struck By/Against (2.8%), and Other (5.6%). Using existing activity criteria ranges and IPAQ MET results, mTBI subjects were Inactive (44.0%), Minimally Active (52.0%), and Highly Active (4.0%). Activity ratings were highly variable across participants and were not statistically significantly different between mTBI and UC groups (p=0.344).

Table 1.

Sample Characteristics, M (SD) unless otherwise noted

	mTBI n=36	Uninjured Control n=24	p
n with BCTT at 1-week / 1-month visit	36/25	24/235
Age (years), range	32.24 (11.34), 18.72 – 62.73	39.28 (10.54), 20.64 – 54.65	0.014
Female gender, n (%)	17 (47.2%)	10 (41.7%)	0.874
Race, n (%)			< 0.001
Black	23 (63.9%)	4 (16.70%)
White	11 (30.6%)	15 (62.5%)
Other/not reported	2 (5.6%)	5 (20.8%)
Body mass index, M (SD), range	28.24 (6.22), 13.75–45.72	26.28 (4.73), 19.00–36.80	0.174
Height, in	68.13 (4.57)	68.00 (4.54)	0.898
Socioeconomic status index1	27.00 (13.19)	34.57 (18.26)	0.130
Health insurance type, n (%)			0.486
Commercial	13 (36.1%)	13 (54.2%)
Government	19 (52.8%)	10 (41.7%)
None	2 (5.6%)	0 (0.0%)
Other/unknown	2 (5.6%)	1 (4.2%)
Years of education	12.58 (1.81)	14.67 (1.93)	< 0.001
Pre-injury physical activity METs (IPAQ)2
Total MET	1193.7 (1589.2)	874.3 (1160.3)	0.424
Walking MET	414.5 (593.9)	278.4 (320.5)	0.920
Moderate MET	208.0 (454.8)	292.5 (452.3)	0.250
Vigorous MET	571.2 (1103.1)	303.3 (651.1)	0.329
Pre-injury physical activity category (IPAQ)			0.344
Inactive	11 (44.0%)	15 (62.5%)
Minimally active	13 (52.0%)	7 (29.2%)
Highly active	1 (4.0%)	2 (8.3%)
Cause of injury, n (%)
Motor vehicle/traffic collision	28 (77.8%)	-
Fall	4 (11.1%)	-
Assault	1 (2.8%)	-
Struck by/against	1 (2.8%)	-
Other	2 (5.6%)	-
Acute injury characteristics, n (%)
Admission GCS 153	36 (100%)
Head CT
Not performed	15 (41.7%)
Negative	21 (58.3%)
Loss of consciousness	10 (27.8%)	-
Posttraumatic amnesia	12 (33.4%)	-
Retrograde amnesia	5 (13.9%)	-
Other altered mental status4	36 (100.0%)	-
mTBI-related symptom severity
RPQ total score, 1 week	20.33 (15.6)	-
RPQ total score, 1 month	16.96 (15.2)	-

Note. BCTT, Buffalo Concussion Treadmill Test; IPAQ, International Physical Activity Questionnaire; MET, metabolic equivalent of task; mTBI, mild traumatic brain injury

¹Hollingshead socioeconomic status index (composite of individual education and occupation)

²The first 11 participants in the mTBI cohort did not complete the IPAQ questionnaire due to delayed add on time to the protocol.

³for 13 mTBI participants, admission GCS was not found in the medical record but was presumed to be 15 based on provider notes.

⁴Included self-report of other indicators of altered mental status such as confusion and disorientation

⁵1 UC completed the 2^nd visit by phone due to the COVID-19 pandemic, so BCTT data were not available at this timepoint.

Prevalence of Exercise Intolerance

Figure 1 depicts box plots of BCTT duration by group and timepoint, with mTBI participants stratified by their reported mTBI symptom recovery status. (Inferential statistics combined symptomatic and asymptomatic mTBI participants to maximize statistical power.) At 1 week, median (Med) BCTT duration for the mTBI group as a whole was 7.00 minutes (Inter Quartile Range [IQR]=2.11–11.25, range=0.00–15.00). At 1-month, median BCTT duration for the mTBI group was 8.27 minutes (IQR = 4.18–12.58, range = 0.10–15.00). Controls had a median 1-week duration of 11.86 minutes (IQR = 11.00–15.00, range = 2.24–5.00) and a 1-month Med=9.11 minutes (IQR = 6.36–15.00, range = 1.60–15.00). A linear mixed effects model with effects of Group and Time revealed a significant Group x Time interaction, F(1,46.306) = 10.98, p = 0.002, due to a significant effect of Group (lower duration in mTBIs) at the first (1-week) visit (p = 0.029) that was nonsignificant at the second (1-month) visit (p = 0.965). Adding BMI as a covariate (based on analyses below indicating this was an independent predictor of BCTT duration) did not change these findings, Group x Time F(1,46.157) = 9.72, p = .003; post-hoc group comparisons at 1-week p < .001 and 1-month p = .377.

Figure 1.

Boxplot of Buffalo Concussion Treadmill Test (BCTT) duration at 1-week and 1-month postinjury in symptomatic mild traumatic brain injury (mTBI), asymptomatic mTBI, and uninjured control (UC) Groups. Symptomatic status determined by an interview question about whether mTBI participants felt their mTBI-related symptoms had remitted. Among the two outliers at 1 week in the UC group, one was a 54-year-old female (BCTT duration = 2.2 minutes, height 154.9 cm BMI = 32 kg/m²) who reported inability to continue due to the speed of the treadmill, and the second was a 51-year-old female (BCTT duration 4.93 min, height 170.2 cm BMI = 37 kg/m²), who discontinued due to exertion. These two individuals had similar performance on the task at the second visit (1.6 and 4.89 minutes, respectively).

Figure 2 portrays the frequency of different behavioral reasons for discontinuing the BCTT stratified by group and time. The percentage of participants with mTBI who completed the BCTT was 13.9% at 1-week and 16.0% at 1-month post-injury. Reasons for discontinuing the BCTT prior to completion were variable; discontinuing due to mTBI-related symptom exacerbation was common in the mTBI group (36.1% and 40.0% at 1 week and 1 month, respectively). Commonly, subjects with aggravated mTBI symptoms had complaints of dizziness, unsteady gait, loss of balance, and worsening headache. Peripheral (Other) symptoms reported by participants included indicating pain from peripheral injury, musculoskeletal pain, chest pain, shortness of breath, and exertional fatigue.

Figure 2.

Percentage of participants who discontinued the Buffalo Concussion Treadmill Test (BCTT) before the 15-minute time limit, stratified by the reason for discontinuation.

The percentage of UC participants who completed the BCTT was 30.4% at both timepoints. Most UC participants discontinued due to other symptom exacerbation (39.1% visit 1, 60.9% visit 2), which included peripheral irritation/pain. This was followed by exertion (26.0%, 9.0%) and other (4.3%, 0.0%). Taken together, the reasons UC discontinued were generally due to self-report of non-specific symptoms (e.g., general pain, limb pain, breathing difficulties/chest discomfort) and/or or meeting the pre-specified quantitative discontinuation criteria (Borg ≥18 or VAS increase ≥3).

As a secondary analysis we examined who, among the mTBI participants who discontinued to physical exertion, did so before their HR had exceeded 85% of their age-predicted maximum HR. As discussed earlier, experiencing a high level of subjective exertion paired with a relatively low HR response to exercise could be considered a clinical sign of the ANS consequences of mTBI. At 1 week, only 1 of 8 mTBI participants who discontinued the BCTT due to exertion had achieved the target HR of > 85% their age-predicted maximum HR (vs. 4 of 6 UC). The Fisher’s exact test statistical value from discontinuing from exertion only is 0.1023, which is not significant at p<0.05. Therefore, we summed the 13 mTBI participants who discontinued due to mTBI-related symptom exacerbation with the 7 who discontinued for exertion without achieving their target HR, to yield a total percentage of mTBI participants displaying signs of mTBI-related exercise intolerance of 20/36 (55.6%). At 1 month, 2/4 mTBI participants who discontinued due to exertion failed to meet the HR threshold which, combined with the individuals who discontinued due to mTBI-related symptom exacerbation, yielded a total percentage of mTBI participants with signs of mTBI-related exercise intolerance of 12/25 (48.0%).

Correlates of BCTT Duration

Pre-injury baseline factors.

Table 2 summarizes associations between BCTT duration (in minutes) and other factors. Within the full sample (controlling for group), being female (p=0.041) and being shorter in height was associated with longer BCTT duration at 1 week (p = 0.027), with small-medium effect sizes. A higher BMI was moderately-strongly associated with lower BCTT duration at both time points (p = 0.001). IPAQ MET scores were not associated with BCTT duration. In a multivariable general linear model predicting BCTT duration from gender, height, and BMI, BMI was the only significant predictor (e.g., predicting 1-week BCTT duration, BMI p = 0.009, group p < .001, vs. female and height p ≥ 0.558; see eTable 2).

Table 2.

Association Between Baseline (Pre-Injury) Variables, BCTT Duration, and Clinical Recovery

	r (p) with BCTT Duration (Minutes)
	1 Week	1 Month
Baseline Factors
Age	−0.14 (.296)	−0.03 (.831)
Female	−0.27 (.041)	−0.26 (.080)
BMI	−0.42 (.001)	−0.45 (.001)
Height (inches)	0.29 (.027)	0.22 (.139)
IPAQ Total MET	−0.04 (.779)	−0.18 (.288)
IPAQ Walking MET	−0.03 (.868)	0.25 (.124)
IPAQ Moderate MET	−0.11 (.451)	0.06 (.709)
IPAQ Vigorous MET	−0.02 (.888)	0.09 (.583)
mTBI-Related Symptoms
RPQ Total 1 Week*	−0.34 (.044)	−0.47 (.017)
RPQ Total 1 Month*	−0.28 (.160)	−0.45 (.025)

Note: Correlations (r) with baseline factors reflect partial correlations between baseline factors and BCTT duration, controlling for group (mTBI, control). Correlations between RPQ and BCTT duration reflect Pearson correlations within the mTBI group. Correlations bolded where p < .05. BCTT, Buffalo Concussion Treadmill Test; mTBI, mild traumatic brain injury; IPAQ, International Physical Activity Questionnaire; RPQ, Rivermead Post Concussive Symptoms Questionnaire.

mTBI-related symptoms.

Within the mTBI group, reporting more severe mTBI-related (RPQ) symptoms at 1 week was associated with a shorter-duration BCTT exam at both 1 week and 1 month (r = −.34 and −.47, respectively).

Discussion

In this prospective longitudinal cohort study, we characterized how adult community members with and without mTBI responded to the BCTT, which was developed for and is widely used to assess athletes with sport-related mTBI. In particular, the BCTT is used both for the purpose of establishing exercise prescriptions to facilitate rehabilitation and to identify persons with signs of ANS dysfunction contributing to exercise intolerance. For example, the duration an individual can participate in the BCTT can be used to set initial exercise prescriptions, with typical recommendations to discontinue activity before this time to avoid significant symptom exacerbation and to increase exercise duration periodically as tolerated. Furthermore, persons who display mTBI-related symptom exacerbation during the BCTT or who discontinue due to exertion before their HR reaches 85% of their maximum HR have been proposed to fall into a unique “physiological” subtype of mTBI characterized by autonomic nervous system dysfunction and are hypothesized to especially benefit from exercise-based rehabilitation. In our sample of persons treated at a level 1 trauma center emergency department with mTBI, only 14% (1-week post-injury) and 16% (1-month post-injury) completed a 15-minute version of the BCTT without meeting discontinuation criteria of significant symptom exacerbation, exertion, or observable signs of difficulties such as imbalance.

Reasons for early discontinuation of the exam were variable; consistent with our hypothesis, a large percentage of the mTBI group terminating the exam due to increased mTBI-related symptom exacerbation (36%–40% over time). Another 8–17% discontinued due to exertion before their HR had exceeded 85% of age-predicted maximum (which is considered to be a physiologic sign of mTBI in light of fatigue and reduced autonomic flexibility being associated with mTBI).^9,28 Combining these groups, 55.6% (1 week) and 48.0% (1 month) of mTBI participants displayed evidence of mTBI-related exercise intolerance. These findings imply that exercise intolerance is common and detectable by the BCTT in adult community members who present to an ED with mTBI, which indicates the BCTT might support clinical management in this population. Furthermore, incorporating objective HR data increased the detection of possible signs of mTBI-related exercise intolerance, highlighting the relevance of considering both subjective symptoms and observable signs of exercise intolerance, including objective physiological data when possible.

Of course, most mTBI symptoms are not specific to brain injury, and it is possible other contributing factors contributed to participants’ BCTT outcomes (e.g., peripheral injuries, medications, being inactive before injury). However, the higher rate of early discontinuation on the BCTT seen in the mTBI group supports that the BCTT detected some mTBI- and injury-related sequelae. This assertion is further supported by our finding expectable associations mTBI symptom severity (i.e., RPQ scores) and duration of the BCTT. For example, BCTT duration at 1-month post-injury was associated with higher symptom burden both cross-sectionally but also at the earlier study visit (1-week), implying that high symptom burden predicts a slower return of exercise tolerance in community members with mTBI. This agrees with prior studies that show elevated symptom burden soon after injury correlated with slower clinical recovery.²⁹

We did not anticipate the high rate (70%) at which the UC group, comprised of relatively healthy community adults without acute injury, met discontinuation criteria on the BCTT. The main reason for this was an increase in “other symptom exacerbation,” which included subjects’ report of general pain or discomfort or specific limb pain. Other UCs reported breathing difficulties/chest discomfort as well as finding the standardized treadmill speed too fast. In the mTBI group, these sorts of symptoms were also counted as “Other symptom exacerbation” and were included in estimates of possible mTBI-related exercise intolerance.

Although we did not record specific symptoms reported by each participant, many UCs discontinued due to reporting increased discomfort on the VAS (i.e., ratings that increased by 3 or more points) or high exertion on the Borg scale. Of note, this occurred in individuals who otherwise felt they could continue with examination. The discontinuation criteria were designed to avoid over-exerting acutely brain-injured persons; however, the fact that these criteria were frequently met by non-injured persons implies that one should not interpret BCTT duration as a sign of mTBI due to potential confounders. Rather, we would suggest attending to the reasons one discontinues (e.g., the degree to which they show mTBI-related symptom exacerbation or a flatter HR response than would be expected by their perceived exertion, implying abnormal autonomic response). Our findings in the control group may have been influenced by the fact that the sample was typically physically inactive (62.5%) or minimally active (29.2%), and preinjury health-related factors (especially BMI) were associated with BCTT duration. The data suggest that it may be helpful to tailor the BCTT more to the general community, either by reducing the intensity (speed, duration) of the test or by revising discontinuation criteria to help distinguish true mTBI-related exercise intolerance from general physical limitations or the influence of exertion on the subjective report of physical symptoms in untrained individuals.³⁰ Additional research is required to better understand the uses of ANS dysfunction in community adults with and without mTBI.

Although BMI was the most robust preinjury predictor of BCTT duration in our sample, female gender also showed an inverse relationship with BCTT duration. This deserves replication and further attention given the documented sex/gender differences in mTBI recovery and response to the BCTT.³¹ However, gender differences vary substantially across studies that have different sample characteristics, time since injury, and outcomes, revealing nuances in the topic that need to be clarified.^28,32

This study was limited by sample size, which necessitated predominantly descriptive analyses. Thus, the findings warrant replication and extension in a larger cohort. However, prospective recruitment and longitudinal follow-up of persons with mTBI and community controls is resource intensive. Other issues to consider include concurrent injuries in our sample which may have contributed to additional pain and may have caused symptoms misinterpreted as mTBI exacerbation. Additionally, although we categorized reasons for discontinuation into discrete bins (e.g., mTBI-related symptom exacerbation), it is not always possible to definitively attribution the nonspecific symptoms of mTBI to mTBI versus other factors. Another limitation is that we were only able to collect heart rate at 2-minute intervals, which may have reduced the precision with which we assessed end-of-task heart rate. Lastly, other statistically significant differences between mTBI and UCs were noted, which may have limited our ability to compare findings between groups. However, this concern is tempered by our finding that the primary effects of interest (mTBI versus UC group comparisons) remained robust after controlling for covariates, while other factors (age) were neither associated with group in a direction that would support our hypothesis nor substantially related to BCTT duration in our sample. Other variables that showed group differences (race, education) theoretically are not associated with BCTT performance.

Although key associations were identified, improvements to tailor the BCTT to adult community members with mTBI must be investigated further given that injuries in the general trauma population are diverse and multifactorial. Considering the response of the UCs to the BCTT, a number of modifications might be introduced to isolate symptom exacerbation of concussive symptoms (dizziness, unsteady gait, loss of balance, and worsening headache) from discontinuation due to reports of peripheral symptoms (i.e., peripheral injury, musculoskeletal, chest pain, shortness of breath) and aerobic deconditioning. New protocols with similar efficacy to the BCTT, such as the Buffalo Concussion Bike Test, may prove to be more applicable to testing level I trauma center patients with mTBI due to a lower likelihood of exacerbating peripheral injury symptoms common to this patient population.³³

Intolerance to exercise is proposed to be an indicator of physiological concussion in young athletes.¹⁶ Similarly, our study indicates that the BCTT may detect physiological consequences of mTBI in community adults. Importantly, exercise intolerance is prevalent at least 1-month post-injury. Although preliminary, the findings provide strong early evidence that adopting treadmill exercise tolerance testing in adult community members with mTBI informs understanding of individual differences in response to exercise, and may help to advance evidence-based clinical management practices for the often undertreated adult/non-athlete mTBI population.

Conclusion

The BCTT elicits exercise intolerance and exacerbates mTBI-related symptoms in adult community members, similar to the athletes for which it was designed. The findings suggest that the BCTT could be useful in clinical assessments of adults with mTBI and in prescribing tailored exercise programs to community adults with mTBI in the future. At the same time, the high rate at which non-injured controls met BCTT criteria for early discontinuation indicates that the exam may need to be modified for the general adult community, and that more research is needed to develop and validate exercise testing and active rehabilitation strategies for the general adult community.