A Preliminary Study of the Effect of Early Aerobic Exercise Treatment for Sport-Related Concussion in Males

Abstract

Objective:

To study the effect of early prescribed aerobic exercise vs. relative rest on rate of recovery in male adolescents acutely after sport-related concussion (SRC).

Design:

Quasi-experimental design.

Setting:

University sports medicine centers.

Participants:

Exercise Group (EG, n=24, 15.13±1.4 years, 4.75±2.5 days from injury) and Rest Group (RG, n=30, 15.33±1.4 years, 4.50±2.1 days from injury).

Interventions:

EG performed a progressive program of at least 20 minutes of daily sub-threshold aerobic exercise. RG was prescribed relative rest (no structured exercise). Both groups completed daily online symptom reports (Post Concussion Symptom Scale, PCSS) for 14 days.

Main Outcome Measures:

Days to recovery after treatment prescription. Recovery was defined as return to baseline symptoms, exercise tolerant, and judged recovered by physician examination.

Results:

Recovery time from Initial Visit was significantly shorter in EG (8.29±3.9 days v. 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.

Conclusions:

These preliminary data suggest that early sub-threshold aerobic exercise prescribed to symptomatic adolescent males within one week of SRC hastens recovery and has the potential to prevent delayed recovery.

INTRODUCTION

The 2017 Concussion in Sport Group (CISG) consensus statement says that sport-related concussion (SRC) results in a range of clinical signs and symptoms that typically follow a sequential course of resolution.¹ Furthermore, the clinical signs and symptoms cannot be explained by drug, alcohol, or medication use, other injuries (such as cervical injuries, peripheral vestibular dysfunction, etc.) or other comorbidities (e.g., psychological factors or coexisting medical conditions). Adolescent athletes have a significant rate of SRC² and have been shown to base their perception of recovery primarily on physical symptoms (e.g., headache, nausea, fatigue, etc.).³ It appears that adolescents take longer to recover from concussion than do adults and children,^4,5 and while symptoms are highly variable and not specific to brain injury,^6,7 they are nevertheless the most common metric used to define recovery from concussion.⁸

Rest has long been a mainstay of concussion treatment because of fear of exacerbation of symptoms, delayed recovery, and/or brain re-injury with exertion early after concussion.⁹ Nevertheless, exercise to the maximum capacity of each sport without exacerbation of symptoms is recommended before an athlete returns to play after SRC.¹ Regular aerobic exercise is good for the human brain¹⁰ and emerging evidence suggests that moderate levels of physical activity^11,12 or prescribed sub-threshold aerobic exercise¹³ do not delay recovery in athletes after SRC and may in fact speed recovery in those with persistent post-concussive symptoms (PPCS).^14–17 Regular aerobic exercise has been reported to have beneficial effects on cognition,¹⁸ mood,¹⁹ and sleep.²⁰ Too much exercise can, however, produce post-concussion-like symptoms in concussed patients and delay recovery.²¹ Since the most recent international consensus statement recommends a more active approach to concussion treatment,²² it would be useful to know how sub-threshold exercise affects specific symptoms in adolescents after SRC when compared with adolescents who were instructed to rest. Thus, the purpose of this study was to compare early sub-threshold aerobic exercise with prescribed rest on days to recovery from concussion for adolescent males. Since complete rest is not the current standard of care, we used a historical cohort control group that was prescribed rest and compared it with a cohort of concussed male adolescents who were given an individualized sub-threshold aerobic exercise prescription early after SRC. We consider this study preliminary because participants were not randomized to the respective treatments. We hypothesized that sub-threshold aerobic exercise would reduce the days to recovery after treatment prescription. If the time to recovery was different between the groups we planned to conduct post-hoc analyses of the daily symptom scores to determine if there was differential response to exercise or rest across types of symptoms: physical, cognitive, sleep, and affective.

METHODS

Study Design

The study was approved by the University at Buffalo Institutional Review Board. We compared two cohorts of adolescents who sustained acute SRC. It is important to note that this was not a randomized trial but rather a comparison of two convenience samples that were otherwise similar in age, sex, athletic background, and time since injury but received different recommendations from their treating physicians. We focused on male adolescents for this comparison because they matched well across the groups with respect to age, initial symptoms, and athletic ability. The Rest Group (RG) is from a previously published study that recruited participants between March 2013 and February 2015.²³ The Exercise Group (EG) is from an ongoing randomized controlled trial (clinicaltrials.gov: NCT02710123) that began in September 2016. In both groups, adolescent male athletes came to one of three University Concussion Management Clinic less than 10 days from injury. They were diagnosed with a concussion by experienced sports medicine clinicians according to international guidelines¹ using history (including a concussion symptom checklist and cognitive evaluation) and physical examination. All EG performed the Buffalo Concussion Treadmill Test (BCTT) at the first clinical visit (Initial Visit) to calculate the exercise prescription whereas not all RG performed the BCTT since they were not part of a study of the effects of exercise assessment. Previous research has shown that performing the BCTT within 10 days of injury does not affect recovery.²³ If eligible for the study, a research assistant explained the study and consent was obtained the same day. Parental consent was obtained for minors. RG participants were instructed to rest according to the previous standard of care (i.e., no structured exercise). EG was given an individualized sub-threshold exercise prescription based upon their BCTT performance.¹ Both groups followed up with the physician at 7 and 14 days after the initial visit.

Symptom Reporting

Participants in both groups reported symptoms on a password-protected online data form each day between 7 and 10 PM for two weeks. Participants were instructed to begin reporting symptoms the day after the Initial Visit (called Day 1). To encourage compliance with daily symptom reporting, participants received daily email or text message reminders to access the online data form to record symptoms using the Sport Concussion Assessment Tool 3²⁴ Post Concussion Symptom Scale (PCSS), a validated instrument with normative data for males and females.²⁵ Symptom recovery was defined as return to a baseline level of symptoms, which was defined as a symptom severity score of 7 or less on the PCSS, for three consecutive days.^22,26 For initial symptom scores between 5–7 we used the criterion of no symptoms to define recovery. We categorized PCSS symptoms into the following clusters: physical, cognitive, sleep and affective.^27,28 (Table 1).

Table 1.

Symptom clusters

Physical (8 symptoms, max score 48)	Cognitive (6 symptoms, max score 36)	Sleep (3 symptoms, max score 18)	Affective (4 symptoms, max score 24)
Headache Pressure in head Neck Pain Nausea or vomiting Dizziness Blurred vision Sensitivity to light Sensitivity to noise	Feel slowed down Feeling like “in a fog” “Don’t feel right” Difficulty concentrating Difficulty remembering Confusion	Fatigue or low energy Drowsiness Trouble falling asleep	More emotional Irritability Sadness Nervous or Anxious

Participants

Male adolescents (aged 13–18 years) who sustained SRC within 1–9 days of clinic presentation were evaluated by a study physician who diagnosed the concussion. Potential participants were identified after a standard clinical evaluation that consisted of a thorough history (including standardized concussion symptom questionnaire and cognitive evaluation²⁹) and physical examination by one of several physicians with extensive experience in concussion management. The physical examination was standardized among the study physicians via a 2-hour training session and included instruction on assessment of cervical, oculomotor and vestibular function.³⁰ A sample concussion-specific physical examination form is attached in a supplementary file. No participants were kept out of school for more than 2 days and no subjects were participating in physical, vision or vestibular therapy during the first 4 weeks of the study period. Clinical recovery for the EG cohort was determined by physicians blinded to treatment group when participants reported a baseline level of symptoms, had a normal physical examination, and could exercise to age-appropriate maximum without exacerbation of symptoms.⁸ Physicians could not be blinded for the RG cohort because everyone in that study received the same treatment. For those participants not recovering within 4 weeks, a multi-modal treatment approach was utilized.³¹

Participants were excluded from both groups using the following criteria: (1) evidence of focal neurological deficit; (2) inability to exercise because of orthopedic injury, cervical spine injury, diabetes or known heart disease; (3) increased cardiac risk; (4) current diagnosis of ADHD, learning disorder, depression, or anxiety; (5) history of moderate or severe TBI; (6) more than 3 prior concussions (because these factors are associated with delayed recovery³²); (7) inability to understand English; (8) recovery in 2 days or less from the Initial Visit (because exercise would not have enough time to be effective); (9) sustaining another head injury before recovery; (10) having an initial PCSS score of 5 or below; or (11) not completing at least 75% of daily symptom reports or having missed 3 or more days of reporting symptoms in a row. Only data on males from both groups is presented in this paper.

Sub-threshold Aerobic Exercise Prescription

The sub-threshold aerobic exercise prescription for EG was calculated as 80% of the heart rate (HR) achieved at symptom exacerbation on the BCTT.³³ Previous studies have shown that the BCTT is a reliable method to determine the symptom exacerbation exercise threshold in concussed patients.³⁴ Before the test began, the participants rated their overall symptom state on a Visual Analogue Scale (VAS, 0–10). Then they walked on a level treadmill at 3.2 or 3.6 mph (depending on patient height and comfort level) at 0 degree incline. The incline was increased by 1 degree after each minute for the first 15 minutes and then the speed by 0.4 mph every minute thereafter. Each minute the HR (by Polar HR monitor, Model #FIT N2965, Kempele, Finland), symptom severity (VAS), and Borg Rating of Perceived Exertion (RPE)³⁵ were recorded until symptom exacerbation or voluntary exhaustion, followed by a cool down period to resting HR. Voluntary exhaustion was defined as ≥17 on the RPE scale while symptom exacerbation was defined as an increase of 3 points or more from the pre-exercise VAS value (a point or more for an increase in symptoms and a point for appearance of each new symptom). Participants were instructed to report symptoms, not to push through symptoms, and the examiner observed for visible signs of distress, which would prompt test cessation. EG participants were instructed to exercise at home or in a gym with supervision each day for 20 minutes on a treadmill or stationary bike at the prescribed HR with a 5 minute warm-up and a 5–10 minute cool-down. They were instructed to stop exercise if symptoms were exacerbated or at 20 minutes, whichever came first. Exercise that drastically increases HR, like weight lifting or sprinting, was prohibited. Participants were given a Polar HR monitor to measure their HR at home during exercise. They were explicitly told not to participate in any physical activity which could result in a head injury and were instructed to advance daily cognitive activities according to symptom tolerance. They were instructed to complete a daily record of their symptoms (PCSS) and activity level online.

Relative Rest Prescription

RG participants were assessed by a clinician on their Initial Visit and were told that rest was recommended to give their concussed brain a chance to heal. Rest was described as not participating in any sports or any other forms of exercise and they were excused from gym class. They were also explicitly told not to participate in any physical activity which could result in a head injury. They were instructed to advance daily cognitive activities according to symptom tolerance. They were asked to provide a daily record of their symptoms and activity level the same as EG. Although they reported their daily physical activity level via the on-line form, RG participants were not strictly monitored for adherence to the recommended rest program.

Statistical Analysis

The primary question being addressed in this study was whether male adolescents recover from concussion at a different pace when prescribed sub-threshold exercise compared with adolescents who were advised to rest until asymptomatic. A series of analyses were completed to determine the differences, if any, between the two groups. Welch’s two-sample t-tests with unequal variances at level 0.05 were used to assess group-wise (EG vs RG) differences in (1) age, (2) previous concussions, (3) days from injury to Initial Visit, and (4) symptom scores at first visit. Fisher’s Exact Test (2-sided) at level 0.05 was used to assess group-wise differences in Initial Visit physical examination findings. Since the purpose of these analyses was to determine whether the groups were comparable we elected the conservative approach of not correcting for multiple comparisons.

Welch’s two-sample t-tests with unequal variances at level 0.05 were used to assess differences in recovery time between groups and Fisher’s Exact Test (2-sided) at level 0.05 was used to assess the rates of delayed recovery (> 30 days).²² Post hoc analyses were planned with the proviso that the initial primary analysis demonstrated a significant difference between the two interventions. These analyses were considered preliminary since they focused on a novel issue; namely, consideration of whether certain symptom clusters varied across treatments. Thus a series of t-tests were conducted comparing (1) Day 1 to 14 total symptom score, (2) Day 1 to 14 physical symptom score, (3) Day 1 to 14 cognitive symptom score, (4) Day 1 to 14 sleep symptom score and (5) Day 1 to 14 affective symptom score. Finally, 95% confidence intervals (CI) based on a normality assumption for symptom scores were calculated for each day. Missing values were imputed using the average of the day before and the day after score. If scores for two consecutive days were missing, they were imputed using an average of the score on the latest day before and the score on the soonest day after. Day 14 missing values were imputed with the subject’s score from Day 13. Mean values and 95% CI (based on normality) were calculated for each group for each cluster on each day. Mixed effects linear models were used to analyze the total symptom scores and the symptom cluster scores from Day 1 to Day 14. Fixed effects included group and time (days) with their interaction term in the model. Random effects were the intercepts for each subject. A p-value ≤ 0.05 was considered significant, and because of the preliminary nature of these post hoc analyses we did not adjust the p-value to account for multiple comparisons. All data analyses were performed using the R programming language.³⁶

RESULTS

A total of 67 participants were consented for the study. Three participants withdrew from RG because they did not return to the clinic and 3 participants withdrew from EG because they did not return to the clinic or were not willing to perform the exercise prescription. Three participants from RG and 2 participants from EG were removed because they failed to complete at least 75% of the daily symptom reports or missed 3 or more days in a row. One participant from RG and 2 from EG recovered in 2 days or less from the Initial Visit and were excluded. Hence, RG contained 30 participants and EG contained 24 participants. Symptom scores were missing for 6.8% of the day/participant cells and were imputed as described above.

RG and EG were not statistically different in age, concussion history, time from injury to Initial Visit, Day 1 symptom scores, or Initial Visit physical examination findings (Table 2).

Table 2.

Demographics, Day 1 symptom scores, and Initial Visit physical examination findings

	EG (n=24)	RG (n=30)	p-value*
Age in years	15.13 ± 1.42	15.33 ± 1.40	0.593
Sex	Male	Male	-
Previous Concussion	0.71 ± 0.81	0.33 ± 0.61	0.066
Injury to First Clinic Visit in days	4.75 ± 2.47	4.50 ± 2.13	0.697
Day 1 Total Symptom Score (95% CI) Maximum score = 132	22.79 ± 13.80	28.43 ± 17.95	0.197
Day 1 Physical Symptom Score (95% CI) Maximum score = 48	11.92 ± 6.60	14.80 ± 8.50	0.167
Day 1 Cognitive Symptom Score (95% CI) Maximum score = 36	5.58 ± 5.30	7.70 ± 4.95	0.140
Day 1 Sleep Symptom Score (95% CI) Maximum = 18	2.83 ± 2.93	4.20 ± 3.57	0.128
Day 1 Affective Symptom Score (95% CI) Maximum = 24	2.46 ± 3.45	1.73 ± 2.07	0.370
Initial Visit Oculomotor Deficit1	63% (15/24)	73% (22/30)	0.298**
Initial Visit Vestibular Deficit2	54% (13/24)	50% (15/30)	0.631**
Initial Visit Cervical Deficit3	42% (10/24)	20% (6/30)	0.064**

¹presence of abnormal visual tracking, saccadic eye movements or convergence/accommodation (observed and/or symptom producing).

²presence of abnormal tandem gait or vestibular-ocular reflex (VOR).

³presence of cervical tenderness or abnormal range of motion.

^*Welch’s two-sample t-tests with unequal variances unless specified.

^**Fisher’s Exact Test (2-sided).

Recovery time from Initial Visit was significantly faster for EG than RG (8.29 ± 3.85 days vs 23.93 ± 41.73 days, p = 0.048). The null hypothesis was rejected. Post hoc analysis revealed that recovery time from initial injury was faster for EG but did not quite reach significance (p = 0.052). By the end of daily symptom monitoring (Day 14), EG had significantly fewer participants who remained symptomatic in total (p = 0.028), physical (p = 0.028), cognitive (p = 0.027), and sleep (p = 0.011) clusters. EG had slightly more participants with affective symptoms on Day 14 but it was not significant (p = 0.816). It is important to note that none of the 24 EG participants had delayed recovery whereas 4 RG participants (13%) had delayed recovery. The average recovery time for the 4 RG participants with delayed recovery was 113.25 (± 73.6) days. The sample size of slow recovery participants was too small to conduct a meaningful statistical comparison. Recovery time since Initial Visit, total recovery time, % not recovered for total symptoms and individual symptom clusters on Day 14, and incidence of delayed recovery are presented in Table 3.

Table 3.

Recovery time since Initial Visit, Total Recovery Time, and incidence of delayed recovery

	EG (n=24)	RG (n=30)	p-value
Recovery time since Initial Visit (days)	8.29 ± 3.85	23.93 ± 41.73	0.048*
Total Recovery Time (days)	13.04 ± 4.89	28.43 ± 41.78	0.052*
Total Symptoms (% not recovered by 14 days)	8% (2/24)	33% (10/30)	0.028**
Physical Symptoms (% not recovered by 14 days)	8% (2/24)	33% (10/30)	0.028**
Cognitive Symptoms (% not recovered by 14 days)	4.2% (1/24)	27% (8/30)	0.027**
Sleep Symptoms (% not recovered by 14 days)	0% (0/24)	23% (7/30)	0.011**
Affective Symptoms (% not recovered by 14 days)	8% (2/24)	7% (2/30)	0.816**
Delayed Recovery (recovery > 30 days)	0% (0/24)	13% (4/30)	0.063**

^*Welch’s two-sample t-tests with unequal variances

^**Pearson Chi-square test (2-sided)

Figures 1 and 2a-d show the mean recovery trajectories for total symptoms and symptom clusters. The total (Figure 1, p < 0.001), physical (Figure 2a, p < 0.001), cognitive (Figure 2b, p < 0.001), sleep (Figure 2c, p = 0.002), and affective (Figure 2d, p = 0.004) clusters decreased significantly over time; however, there was no significant difference in the rate of improvement of each cluster (interaction of time and group, p = 0.658, 0.483, 0.686, 0.420, and 0.777, respectively). Total symptom score became significantly lower in EG compared with RG from Day 4 forward (p = 0.020). Physical symptom score became significantly lower in EG compared with RG from Day 4 forward (p = 0.024). Sleep symptom score was significantly lower in EG compared with RG on Day 2 (p = 0.043), was not significantly different on Day 3, yet returned to be significantly lower from Day 4 forward (p < 0.001). Affective symptoms were not significantly different between the two groups from Day 1 to Day 14.

Figure 1.

Mean total symptom score recovery trajectory for 14 days with 95% CI *indicates significant difference (p < 0.05)

Figure 2.

Mean physical (2a), cognitive (2b), sleep (2c), and affective (2d) symptom cluster score recovery trajectory for 14 days with 95% CI *indicates significant difference (p < 0.05)

DISCUSSION

This non-randomized cohort study provides preliminary evidence for the efficacy of controlled and individualized sub-threshold aerobic exercise as an active early (within one week of injury) treatment for SRC in symptomatic male adolescents when compared with the previous standard of care (relative rest). While male adolescents prescribed sub-threshold exercise recovered faster than those who were instructed to rest, the results should be interpreted with caution given that they were not randomly assigned and the groups were drawn from different years. We considered whether the rest group included more severely injured athletes; however, athletes who were recruited for the exercise group were more likely to have had a previous concussion and did not differ on physical examination, suggesting they were at least equal to the earlier sample in terms of concussion severity. There is the possibility that the perception of concussion is affected by the way the concussion is managed in the clinic. The adolescent who is told to go home and rest and avoid all physical activity after concussion may be more likely to adopt the ‘sick role’ and focus attention on his state of health. Conversely, adolescents advised to exercise could feel “permitted” to adopt a more proactive approach to recovery and therefore less likely to attend to and report every symptom as potentially related to the concussion. Nevertheless, EG and RG were well matched on age, concussion mechanism of injury, time to clinic, initial symptom scores, and initial physical exam findings, making them very comparable groups.

The data on participants with delayed recovery is important for several reasons. First, this small group of participants greatly influenced the mean recovery time for the rest group. Second, a possible important advantage for early exercise-based treatment is that it may prevent delayed recovery in some patients. Research with larger samples in randomized trials is required to validate this observation but it is encouraging nonetheless since patients with very long recovery times present a particular dilemma to health care providers and their families.³⁷ The potential for early moderate physical activity to significantly reduce the incidence of delayed recovery has been reported in a large observational study of concussion in the emergency department.³⁸

Prescribed aerobic exercise appears to improve recovery time but the rates of improvement were highly variable regardless of intervention. Thus, this finding needs to be examined with a larger sample size and with females. It appears from these data that most facets of symptoms, except for affective symptoms, improve in a similar manner and rate regardless of whether there is spontaneous improvement (i.e., during rest) or improvement that is aided by exercise. In the statistical comparison using mixed effects linear models, the rate of change of symptom scores was significant with time but not between groups. Such analyses are greatly influenced by the high variability of symptom reports, something observed regularly with adolescents, and do not take into account the fact that a treatment, in this case exercise, is not likely to have an immediate effect on symptoms. After a week, when one might expect to see a physiological effect of exercise, the slope of symptom reduction is modified by the prior week’s slope and therefore it is much less likely that we would find a statistically significant difference. We observed a slight increase in total symptoms on Day 7 (that lasted for one day) in EG but not RG. Further analysis showed that this was due to an increase in the physical symptom of nausea on Day 7 (the day of the second study visit). One possibility is that subjects performed exercise at home after having a physical examination and BCTT in the clinic on Day 7, which effectively doubled the dose of exercise that day and could have exacerbated symptoms. We instructed participants at the start of the study to begin exercise the day after the Initial Visit to avoid symptom exacerbation but it was not in our protocol to inform them to avoid exercise the same day after attending their second clinic visit. These results are informative because clinics that incorporate systematic testing of exercise tolerance shortly after SRC should advise patients not to exercise until the following day.

The results of this study provide cautious support for the 2017 Berlin CISG international consensus statement that complete rest until symptom resolution may not be an effective intervention for SRC and that closely monitored active rehabilitation programs involving sub-symptom threshold and submaximal exercise may facilitate recovery. Future studies should evaluate the optimal timing, mode, duration, intensity, and frequency of treatment during the post-acute time period following SRC.^22,39 The fact that participants in the exercise group recovered twice as fast as those in the rest group may reflect a beneficial effect of exercise on the physiology of concussion, the fact that prolonged rest has been shown to delay symptomatic recovery in adolescents after concussion,¹² or both. The rapidity of the response to exercise means that reduced cardiovascular risk is not the mechanism. Rather, aerobic exercise may enhance neuroplasticity acutely to promote neuron growth and repair after concussion via increased levels of brain derived neurotrophic factor (BDNF).⁴⁰ Voluntary exercise has been shown to increase BDNF in pre-clinical models of simulated concussion in rodents⁴¹ and a recent study showed that exercise initiated within 1–3 days post-concussion increased BDNF levels (and significantly improved motor and cognitive functioning) in concussed rodents, the most robust effect being observed in the pre-trained “rat athletes” when compared with rats restricted from exercise.⁴² This study suggests that the exercise-trained animal has a different, more beneficial response to early exercise after concussion versus the untrained, unfit animal. It is possible that early exercise after concussion stimulates BDNF action since one study in human athletes showed that BDNF levels increased immediately following ramp incremental cycle ergometry to exhaustion.⁴³ Concussion also adversely affects function of the autonomic nervous system (ANS).⁴⁴ A potential mechanism for the rapid effect of exercise may therefore involve improved ANS function. Regular aerobic exercise training increases parasympathetic and decreases sympathetic activity in the human heart at rest, reduces resting heart rate, and decreases submaximal exercise heart rate by reducing sympathetic activity to the heart during exercise.⁴⁵ Cardiac autonomic activity can be acutely modulated during and immediately after graded-work load exercise,⁴⁶ with some evidence that females have a stronger parasympathetic response than males to an acute bout of exercise.⁴⁷

Sleep disturbance is a very common complaint after SRC.⁴⁸ Controlled aerobic exercise would be expected to improve sleep because regular aerobic exercise has been shown to improve sleep quality.^49,50 One physiological effect of concussion appears to be a disruption in regulation of ANS control that affects the balance between parasympathetic and sympathetic output. We have shown that during the acute phase after SRC, concussed patients demonstrated an impaired ability to shift from parasympathetic to sympathetic control over heart rate at the onset of exercise.⁵¹ Sleep involves activation of the parasympathetic branch of the ANS and increased ANS arousal interferes with sleep.⁵² Concussed patients may therefore have difficulty activating the parasympathetic response necessary for good sleep, which regular exercise could ameliorate by improving ANS balance and/or responsivity.^53,54 Hence, it was not surprising that sleep symptoms were the first to become significantly lower in the exercise group. Affective symptoms did not improve faster with aerobic exercise in our study, which was unexpected because regular aerobic exercise has been shown to reduce symptoms of anxiety and depression as soon as within 10 days of starting daily training.⁵⁵

Limitations of the study include that the subjects were recruited at two different times, there was no randomization, and we only assessed male adolescents. Rest Group participants were not strictly monitored for adherence and some of them may have been exercising unbeknownst to the researchers, although this would have reduced rather than enhanced the effect of exercise on recovery that we observed. Subjects were not blinded to treatment; thus, intervention bias is a possible confounder since EG was given an active treatment and may have been more likely to report improvement versus RG. We think this is unlikely since not all symptom clusters improved to the same degree with exercise and we used other measures (physical examination and exercise tolerance) to diagnose concussion or to confirm recovery. The results should not be generalized to females, younger children, adults with risk factors for heart disease, or to concussions incurred by other mechanisms of injury (car accidents, work injuries, etc.). The use of early exercise treatment in athletes with ADHD, a learning disorder, psychological disorders, or a history of 3 or more concussions requires further study. Another potential limitation is that the concussion centers involved in this trial are accustomed to treating patients with exercise after concussion. It will require extra training on the part of others to gain experience and become comfortable with the nuances of exercise testing and treatment of concussed patients with aerobic exercise early after injury.

CONCLUSION

This study provides preliminary evidence for the efficacy of individualized sub-threshold aerobic exercise treatment during the first week after SRC to safely speed recovery in symptomatic male adolescents. It must be emphasized that this is not equivalent to return to sport-specific play; rather, it is an early active intervention intended to speed recovery to the point where it is safe for the athlete to begin the return-to-sport process. Our data provide preliminary evidence that a primary benefit of early exercise-based treatment is a reduction in the number of patients that experience prolonged recovery (>30 days), potentially a very important result. With respect to specific symptom clusters, it appears that most symptoms recover together regardless of intervention or symptom category. These preliminary findings require confirmation in randomized controlled trials using larger samples of both sexes to confirm whether early sub-threshold aerobic exercise promotes recovery after SRC, prevents delayed recovery, and whether exercise or any other form of treatment differentially affects the symptom clusters reported after concussion.

Supplementary Material

Supplemental Data File (.doc, .tif, pdf, etc.)

Supplementary File 1. Brief Buffalo Concussion Physical Exam assessment form.