Secondary Logo

Journal Logo

Original Investigation

Changes in Metabolism and Caloric Intake after Sport Concussion: A Case Series

Daniell, Brooke1; Bernitt, Candace1; Walton, Samuel R.2,3; Malin, Steven K.4; Resch, Jacob E.1

Author Information
Translational Journal of the ACSM: Fall 2020 - Volume 5 - Issue 12 - e000129
doi: 10.1249/TJX.0000000000000129
  • Free



Sport concussion (SC) is defined as a traumatic brain injury (TBI) induced by biomechanical forces to the head, neck, or body (1). During a concussive event, results from animal studies suggest that neuronal strain results in a neurometabolic cascade of events characterized by an ionic flux subsequent to extracellular glutamate binding to the N-methyl-d-aspartate receptors (2–5). In order to correct for this ionic imbalance, an initial hypermetabolic state ensues with increased glucose oxidation (6–8) that, in turn, induces anaerobic glycolysis and an acidic environment (7,8). Concurrently, cerebral blood flow is reduced by approximately 50% for up to 10 d or more (7–9), which limits substrate delivery to the affected areas and further exacerbates the energy crisis. This imbalance between energy “supply” and “demand” leads to a hypometabolic state, thereby hindering the restorative processes for recovery (7,8). Despite our understanding of the pathophysiology of SC in the brain, the whole-body metabolic consequences of SC have yet to be investigated.

Resting metabolic rate (RMR) represents the daily energy requirements to maintain vital, life-sustaining bodily functions including both sleep and conscious but resting states (10). RMR accounts for approximately 60% of the body’s total energy expenditure (TEE) with additional calories expended through physical activity (PA; ≈30%) and the thermic effect of food (≈10%) (10–12). To maintain energy balance (EBal), energy consumption via dietary caloric intake (CI) must equal TEE (13). Clinical studies of moderate and severe TBI in adult humans demonstrate that whole-body RMR is elevated during the first days to weeks after injury and returns to predicted levels throughout the recovery process (14–16). However, there are currently no existing data examining whole-body RMR or CI in humans after an SC. This is clinically relevant because understanding the metabolic and CI needs of patients with SC could elucidate potential avenues for therapeutic dietary and exercise intervention(s) to facilitate recovery from injury. For instance, if there is an increase in whole-body energy demand that is not being met by energy supply (i.e., dietary intake), perhaps increasing energy consumption could reduce recovery time.

Therefore, the purpose of the current study was to examine the effect of SC on energy expenditure (RMR and TEE) and consumption (CI and EBal) throughout clinical recovery in high school and collegiate student-athletes. We hypothesized that RMR would be highest during the acute phase after an SC and then would decline as the student-athletes recovered over a 2-wk period. Similarly, we hypothesized that CI would be increased throughout recovery to mirror the hypothesized whole-body energy demand associated with SC, such that there would be no difference between the total amounts of energy consumed and expended.



A convenience sample of concussed student-athletes were recruited from a large National Collegiate Athletic Association Division I university and local high schools between 2015 and 2017 for this case series. Diagnosis of SC was made by a certified athletic trainer or physician trained in concussion assessment according to the most recent International Consensus Statements on Concussion in Sport (1,17), and participants were subsequently referred to the study team (B. D. or C. B.). This study was approved by our institutional review board. Written informed consent was obtained from each participant 18 yr or older, and both consent (parent/guardian) and assent (participant) were obtained if the participant was younger than 18 yr. Concussed participants were excluded from our study if they had sustained any concurrent injury (e.g., fracture), had received another concussion within the previous 6 months, or had a condition known to affect metabolism (e.g., thyroid dysfunction).


Self-reported symptoms

The Revised Head Injury Scale (HIS-R) was administered at the start of each study session. The HIS-R includes three symptom-related outcomes: the presence of 22 concussion-related symptoms, and the duration and severity of 22 concussion-related symptoms over the prior 24 h (18). Duration and severity are rated on Likert scales for each endorsed symptom, from “briefly” [1] to “always” [6] for duration and “not severe” [0] to “as severe as possible” [6] for severity. All individual duration scores are summed into a total duration sum score, and all severity scores are summed into a total severity sum score. The duration and severity sum scores have shown strong sensitivity (77.5%–97.5%) and specificity (100%) in correctly identifying the presence or absence of a clinically diagnosed concussion (18). The HIS-R was selected specifically because it incorporates a unique duration component. The HIS-R’s concussion-related symptoms are similar to other clinically used inventories, but we felt that duration was also a potentially important piece of the puzzle regarding postconcussion symptom burden.

Energy expenditure

RMR was determined in the supine position by indirect calorimetry (VMax metabolic cart; Carefusion, Yorba Linda, CA). The VMax metabolic cart has demonstrated both good reliability (coefficient of variation = 8.4%) and concurrent validity (R2 = 0.966) when measuring RMR compared with a Deltatrac monitor (19,20). Respiratory exchange ratio (a measure of substrate utilization) was measured by the VMax metabolic cart during indirect calorimetry and was used as an internal assessment of the validity of our measurements (10).

Physical activity

A Fitbit Charge HR (Fitbit, Inc., San Francisco, CA) was used to record daily step count. The Fitbit Charge HR has shown good evidence of reliability (intraclass correlation coefficients (95% confidence intervals), ≥0.70 (0.46–0.86)) and concurrent validity when compared with a handheld step counter (intraclass correlation coefficients (95% confidence interval) as high as 0.74 (0.54–0.87)) in the measurement of walking steps (21).

Dietary recall

Participants were instructed to record all food and beverage intake on a standardized paper form provided by the research team. Self-reported dietary intake, including food and beverage, was recorded daily by participants for each assessment day and the two subsequent days. A 3-d recording window was selected to capture habitual variations in diet and minimize participant burden of recording their intake every day over the course of the study. Specific instructions were provided with regard to recording the quantity, size, and/or volume of each serving, food/beverage brand, and any additional condiments or additives (e.g., ketchup) as previously performed by our team (22). Participant-reported dietary intakes were then entered into the MyFitnessPal (MyFitnessPal, Inc., Baltimore, MD) online portal by a study team member. MyFitnessPal was used to calculate total caloric content of food. MyFitnessPal has been shown to be similar to traditional standardized paper-based dietary recall with regard to overall CI (mean difference <14 kcal) and individual macronutrient CI (mean differences <7 kcal) (23).


Recruited participants reported to the research laboratory for their first assessment within 72 h of their diagnosed SC (T1), 7 d after their first assessment (T2), and again 7 d after their second assessment (T3). All assessments took place between the hours of 0600 and 0900 to account for the effects of circadian rhythm on metabolism. Participants were asked to abstain from consuming anything but water for 6 h before each assessment. After obtaining informed consent/assent at T1, subjects completed a health history questionnaire and height and weight were measured. The HIS-R was completed before assessment with indirect calorimetry at each time point to account for the quantity, severity, and duration of self-reported symptoms.

For indirect calorimetry, participants rested without falling asleep in the supine position for 30 min during which respiratory gasses were collected using a ventilated hood. The VMax metabolic cart was calibrated according to the manufacturer’s guidelines before each assessment. The last 5 min of steady-state data was averaged for statistical analyses. Afterwards participants were provided dietary food journals and Fitbits along with instructions to record their dietary intake and final step count for the day of testing and the two subsequent days. Participants were compensated with a $10 gift card upon receipt of the dietary intake journal after each session.


RMR was measured through indirect calorimetry and is presented in kilocalories expended per day (kcal·d−1). PA level was determined using step count as a surrogate measure. The average daily step count for the day of assessment and the 2 d after assessment were used to represent each assessment time point’s PA level (“sedentary,” <5000 steps per day; “low active,” 5000–7499 steps per day; “somewhat active,” 7500–9999 steps per day; “active,” 10,000–12,499 steps per day; and “highly active,” ≥12,500 steps per day) (24). TEE was calculated by multiplying RMR by the analogous PA correction factor, which is both activity level- and sex-specific (% above RMR; “sedentary,” 15% for men and 15% for women; “lightly active,” 40% for men and 35% for women; “moderately active,” 50% for men and 45% for women; “very active,” 85% for men and 70% for women; and “exceptionally active,” 110% for men and 100% for women) (13). CI was the average number of calories per day consumed by the participant on the day of each assessment and the following 2 d. EBal was calculated as the difference between CI and TEE. A positive value indicated that the participant consumed more energy than they expended, and vice versa.

Statistical Analyses

Separate repeated-measures ANOVA was utilized to assess changes in RMR, TEE, CI, EBal, and PA over time. Post hoc paired-sample t-tests were used to assess which time points were different from each other. Total number of symptoms, total symptom duration sum score, and total symptom severity sum score were then correlated with RMR, TEE, CI, EBal, and PA. Because of the nonnormally distributed nature of self-reported symptoms, Spearman’s ρ correlations were utilized. All analyses were performed with α = 0.05.


A total of 10 concussed participants were included in this study (5 males and 5 females). There were six high school participants (three males (65.4 ± 8.2 kg, 173.1 ± 7.2 cm) and three females (60.0 ± 2.1 kg, 169.0 ± 6.6 cm)) and four collegiate participants (two males (75.7 ± 16.0 kg, 177.5 ± 7.8 cm) and two females (78.0 ± 13.8 kg, 182.5 ± 3.5 cm)). High school participants sustained their injuries while participating in boy’s soccer, girl’s lacrosse, boy’s lacrosse, baseball, or girl’s basketball. Collegiate participants sustained their injuries while participating in softball, men’s lacrosse, women’s volleyball, or men’s track and field. Five participants reported for their initial assessment within 24 h after sustaining an SC, four reported between 24 and 48 h, and one reported between 48 and 72 h after injury.

Energy Expenditure

Measured respiratory exchange ratio remained within a physiological range (0.71–0.92) for each individual at all assessments time points, indicating that our measurements were appropriate (10). Although there was an approximate 177.5 kcal·d−1 decrease in RMR from T1 to T3, no significant differences were observed for RMR across time (F(2) = 1.52, P = 0.24, η2 = 0.15). Similar to RMR, TEE was also not different across time (F(2) = 2.29, P = 0.13, η2 = 0.20; Fig. 1). The SD of measured RMR values increased over time, indicative of increased variability within the group throughout recovery. This in turn prompted post hoc analyses to explore the potential that participant sex was a mediating factor of this effect in order to inform future research questions for the prospective study of energy expenditure after SC. As such, results of these exploratory statistical analyses are not meant to be interpreted with regard to the presence or absence of statistical significance. We examined RMR over time with regard to sex (male vs female) using a separate repeated-measures ANOVA. RMR in males seemed to change across time (F(2) = 5.05, P = 0.04, η2 = 0.56) with an increase in RMR between T1 and T3 (mean difference, 206.5 ± 107.0 kcal·d−1; P = 0.01; Fig. 2). However, females did not seem to change over time (F(2) = 3.07, P = 0.10, η2 = 0.43) despite a tendency to decrease between T1 and T3.

Figure 1
Figure 1:
TEE and CI throughout recovery from SC. Group mean values are presented separately for TEE and CI. Error bars represent 1 SD from the mean in either direction.
Figure 2
Figure 2:
RMR throughout recovery from SC. Measured RMR and reported step counts for each individual at each assessment time point are presented alongside demographics measured at the first assessment (with 72 h after SC). aAge in years. bHeight in centimeters. cWeight in kilograms. dAverage values are presented as mean ± SD.

CI and EBal

A significant decrease in CI was observed over time (F(2) = 4.73, P = 0.02, η2 = 0.35; Fig. 1). Specifically, participants consumed 385.4 kcal·d−1 less at T3 compared with T2 (P = 0.04). In addition, EBal decreased over time (F(2) = 4.35, P = 0.03, η2 = 0.33), with a statistically significant decrease from T1 to T3 of 531.2 kcal·d−1 (P = 0.04). This change indicated that participants were consuming 592.5 ± 604.3 kcal·d−1 more than they were expending at T1 and consuming 61.3 ± 397.5 kcal·d−1 more than they were expending at T3. EBal at T2 showed that participants were overeating by 404.6 ± 598.4 kcal·d−1.

PA Level

Interestingly, no significant differences were observed over time for step count (F(2) = 1.18, P = 0.33 η2 = 0.13). At T1, four concussed participants were classified as being “highly/exceptionally active,” one was “somewhat/moderately active,” two were “low/lightly active,” one was “sedentary,” and one did not report step count for this time point (Fig. 2). At T2 and T3, all participants were either categorized as “somewhat/moderately active” or “highly/exceptionally active.”


At T1, all participants reported that they had experienced concussion-related symptoms within the previous 24 h (median [interquartile range]: number of symptoms, 6.5 [4.5–10.5]; total symptom duration sum score, 19 [15–27.75]; total symptom severity sum score, 14.5 [9.5–20.5]). At T2, only one participant (participant C in Fig. 2) reported experiencing symptoms (drowsiness, feeling slowed down, and difficulty concentrating), with durations and severities at 1 out of 6 for each symptom. No participants self-reported symptoms at T3. Correlation analyses were performed for T1 only, and no statistically significant relationships were identified between total symptom score, duration, or severity and any of the primary outcome variables (all, p ≥ 0.13).


Although we hypothesized that concussed participants would experience a hypermetabolic state after injury similar to previous reports in more severe TBI (up to 200% of predicted values [25]), our present observation does not support this notion. In fact, correlational analyses performed at the initial assessment time point did not reveal any significant relationships between symptom burden and energy expenditure or intake. As the majority of our participants reported no further concussion-related symptoms at the second and third time points, correlational analyses were only performed at T1. These findings indicate that symptom burden in people with SC may not be physiologically linked to whole-body metabolism. It has been proposed that the strongest predictor for a slower recovery from SC is the symptom burden reported by an individual within the first day to few days after a diagnosed concussion (26,27). In our sample, participants experienced relatively mild symptom burden, which may have biased our results toward not detecting discernable changes over time. In either case, our data align more so with the “physiological recovery” that has been demonstrated in animal models of concussion (8,14–16,28). Indeed, concussed participants expended approximately 177.5 kcal·d−1 less at T1 than at T3, which may be a physiologically meaningful, albeit not statistically significant, change over time (11).

There are likely many reasons for why we did not observe a statistically significant change in RMR throughout recovery from SC. First, we observed high intersubject variability at each assessment time point. More specifically, RMR increased in concussed males throughout recovery, whereas a decrease or no effect was observed in females. The causes for the disparate SC recovery trajectories in males and females are unknown, but could be related to different hormonal (29) (e.g., progesterone), autonomic (30) (e.g., heart rate variability), or psychological (31) (e.g., mood state) responses to injury, which were not measured in the current study. In any case, it is worth noting that the sample sizes in the current study are modest, and future work with a large sample and an uninjured control group is required to confirm or refute our findings. Another consideration relates to the relationship between whole-body and brain metabolism. As neurometabolic changes occur within the brain during the acute phase of concussion recovery, our results suggest that SC may not result in systemic metabolic changes. Although changes in whole-body RMR and intracranial substrate utilization have been observed in more severe TBIs as discussed previously (14–16,25), it is reasonable to expect that SC (a subset of mild TBI) may not result in measurable whole-body metabolic consequences. In consideration of this, our preliminary and exploratory findings suggest that further research is warranted to understand the effect of SC on RMR when compared with healthy, matched controls, and further to explore the potential for disparate pathophysiological responses to SC between sexes.

Lastly, PA restriction acutely after SC is recommended in order to avoid exacerbation of symptoms (1,17). As such, an assumption of the current study was that all concussed student-athletes would be sedentary at T1 and potentially T2. The assumption of limited activity directly associates with decreased energy expenditure due to decreased activities of daily living. However, only one concussed participant was classified as “sedentary” at T1 (Fig. 2, participant G), whereas all other participants were more active; 40% of whom were considered to be “highly active” at that same time point (4/10; Fig. 2, participants E, H, I, and J). At T2, no concussed participants were classified as “sedentary.” This in turn may explain why we observed less effect of SC on RMR in our case series. Of note are the observed decreases in step count at T3 compared with T2 in some individuals (Fig. 2, participants A, E, H, and I). Although it is uncertain what caused these reductions, steps were a component of calculating TEE, which was subsequently used to estimate EBal. Despite these decreases in measured PA in a large portion of the sample, EBal was close to isocaloric on the group level at T3. These findings are suggestive of patient-specific responses to SC, and further study is required to better understand EBal and its relationship with clinical recovery.

A primary focus of our study was the influence of concussion and the associated symptom burden on RMR throughout clinical recovery. In alignment with the concussion management protocols at each participant’s institution, participants were progressed through a progressive return to play protocol after reporting symptom-free (1,17). However, we did not have access to participants’ medical records and were unable to actively track the progression of return to sport activities and unrestricted return to play clearance. This is a limitation that should be accounted for in future studies incorporating PA assessments after SC.

We hypothesized that concussed participants would consume a similar number of calories to what they expended throughout recovery, thereby reflecting relative EBal. In contrast, we observed an average energy surplus at T1 and T2, and a relative balance at T3. These findings suggest that although SC promoted an overconsumption of food and higher energy intake relative to energy expenditure initially, the effect was short-lived. Overconsumption could be the result of one or more influences that we did not measure in the current study (e.g., altered appetite, stress eating, more time to eat, etc.), and higher EBal could also have been related to generally reduced PA in the cohort. It is unclear why SC would promote such a change in feeding behavior during the early phase of recovery because PA and RMR were generally unaltered between T1 and T2. Further investigation is required to understand the mechanism by which food intake is regulated after SC and if a diet-based intervention may improve patient care.

Our study is not without limitations. A modest sample of 10 participants were examined during the current study, and thus, our results should be interpreted with caution. Our sample included adolescent participants who may have been in the midst of active growth and development, and we did not assess menstrual cycle phase in our female participants. In addition, body composition was not assessed and influence of lean body mass on RMR cannot be inferred from these data. Subsequently, these factors could independently or collectively have influenced the range of RMR and related outcomes we observed. We elected to use self-reported measures for step count and dietary intake. Although the Fitbit Charge HR automatically calculates step count for participants, the Fitbit is a wrist-worn device, which may also inflate step count through various activities (e.g., clapping hands) and may not be worn for some sport activities (e.g., lacrosse games). Future work should consider use of actigraphy or other more accurate wearable devices to provide a more precise representation of daily PA and intensity. Participants were instructed to manually record their daily dietary intake. Compliance to our dietary recall journal may have been hindered because of participants either not feeling well enough to record their consumed food items (e.g., due to symptom burden) or simply prioritizing other tasks. Nevertheless, a strength of the current study is that individuals with SC were studied immediately after injury and repeated measures were obtained over approximately 2 wk to gain insight toward the recovery process.

In conclusion, SC did not seem to statistically alter RMR. However, our exploratory findings suggest that males and females may experience differing whole-body metabolic consequences after SC. In addition, we observed for the first time that excess food consumption occurs within the first week after SC. The clinical relevance of these metabolic and dietary findings warrants future attention to optimize patient care.

The authors would like to thank all of the athletic trainers who helped in the recruitment of our participants.

There are no conflicts of interest or financial disclosures to report for the current study. The results of this study do not constitute endorsement by the American College of Sports Medicine.


1. McCrory P, Meeuwisse W, Dvorak J, et al. Consensus statement on concussion in sport-the 5th international conference on concussion in sport held in Berlin, October 2016. Br J Sports Med. 2017;51(11):838–47.
2. Osteen CL, Giza CC, Hovda DA. Injury-induced alterations in N-methyl-d-aspartate receptor subunit composition contribute to prolonged 45calcium accumulation following lateral fluid percussion. Neuroscience. 2004;128(2):305–22.
3. Katayama Y, Becker DP, Tamura T, Hovda DA. Massive increases in extracellular potassium and the indiscriminate release of glutamate following concussive brain injury. J Neurosurg. 1990;73(6):889–900.
4. Takahashi H, Manaka S, Sano K. Changes in extracellular potassium concentration in cortex and brain stem during the acute phase of experimental closed head injury. J Neurosurg. 1981;55(5):708–17.
5. Fineman I, Hovda DA, Smith M, Yoshino A, Becker DP. Concussive brain injury is associated with a prolonged accumulation of calcium: a 45Ca autoradiographic study. Brain Res. 1993;624(1–2):94–102.
6. Kawamata T, Katayama Y, Hovda DA, Yoshino A, Becker DP. Administration of excitatory amino acid antagonists via microdialysis attenuates the increase in glucose utilization seen following concussive brain injury. J Cereb Blood Flow Metab. 1992;12(1):12–24.
7. Yoshino A, Hovda DA, Kawamata T, Katayama Y, Becker DP. Dynamic changes in local cerebral glucose utilization following cerebral conclusion in rats: evidence of a hyper- and subsequent hypometabolic state. Brain Res. 1991;561(1):106–19.
8. Giza CC, Hovda DA. The new neurometabolic cascade of concussion. Neurosurgery. 2014;75(Suppl 4):S24–33.
9. Barkhoudarian G, Hovda DA, Giza CC. The molecular pathophysiology of concussive brain injury. Clin Sports Med. 2011;30(1):33–48.
10. Brooks GA, Fahey TD, Baldwin KM. Exercise Physiology: Human Bioenergetics and Its Applications. 4th ed. New York (NY): McGraw-Hill; 2005.
11. Hill JO, Wyatt HR. Role of physical activity in preventing and treating obesity. J Appl Physiol. 2005;99(2):765–70.
12. Ravussin E, Bogardus C. Relationship of genetics, age, and physical fitness to daily energy expenditure and fuel utilization. Am J Clin Nutr. 1989;49(5):968–75.
13. Heyward VH. Advanced fitness assessment and exercise prescription. In: Designing Weight Management and Body Composition Programs. Champaign (IL): Human Kinetics Publishers, 2002; 2004.
14. Chiolero R, Schutz Y, Lemarchand T, et al. Hormonal and metabolic changes following severe head injury or noncranial injury. JPEN J Parenter Enteral Nutr. 1989;13(1):5–12.
15. Bruder N, Lassegue D, Pelissier D, Graziani N, Francois G. Energy expenditure and withdrawal of sedation in severe head-injured patients. Crit Care Med. 1994;22(7):1114–9.
16. Clifton GL, Robertson CS, Choi SC. Assessment of nutritional requirements of head-injured patients. J Neurosurg. 1986;64(6):895–901.
17. McCrory P, Meeuwisse W, Aubry M, et al. Consensus statement on concussion in sport: The 4th international conference on concussion in sport, Zurich, November 2012. Br J Sports Med. 2013;48(4):554–75.
18. Resch JE, Brown CN, Schmidt J, et al. The sensitivity and specificity of clinical measures of sport concussion: three tests are better than one. BMJ Open Sport Exerc Med. 2016;2(1):e000012.
19. Cooper JA, Watras AC, O’Brien MJ, et al. Assessing validity and reliability of resting metabolic rate in six gas analysis systems. J Am Diet Assoc. 2009;109(1):128–32.
20. Schadewaldt P, Nowotny B, Strassburger K, Kotzka J, Roden M. Indirect calorimetry in humans: a postcalorimetric evaluation procedure for correction of metabolic monitor variability. Am J Clin Nutr. 2013;97(4):763–73.
21. Fokkema T, Kooiman TJ, Krijnen WP, Van Der Schans CP. Reliability and validity of ten consumer activity trackers depend on walking speed. Med Sci Sports Exerc. 2017;49(4):793–800.
22. Heiston EM, Eichner NZM, Gilbertson NM, et al. Two weeks of exercise training intensity on appetite regulation in obese adults with prediabetes. J Appl Physiol. 2019;126(3):746–54.
23. Teixeira V, Voci SM, Mendes-Netto RS, da Silva DG. The relative validity of a food record using the smartphone application MyFitnessPal. Nutr Diet. 2018;75(2):219–25.
24. Tudor-Locke C, Bassett DR Jr. How many steps/day are enough? Preliminary pedometer indices for public health. Sports Med. 2004;34(1):1–8.
25. Haider W, Lackner F, Schlick W, et al. Metabolic changes in the course of severe acute brain damage [Internet]. Eur J Intensive Care Med. 1975;1(19–26):19–26.
26. Meehan WP 3rd, Mannix R, Monuteaux MC, Stein CJ, Bachur RG. Early symptom burden predicts recovery after sport-related concussion. Neurology. 2014;83(24):2204–10.
27. Meehan WP 3rd, O’Brien MJ, Geminiani E, Mannix R. Initial symptom burden predicts duration of symptoms after concussion. J Sci Med Sport. 2016;19(9):722–5.
28. Prins ML, Matsumoto J. Metabolic response of pediatric traumatic brain injury. J Child Neurol. 2016;31(1):28–34.
29. Wunderle K, Hoeger KM, Wasserman E, Bazarian JJ. Menstrual phase as predictor of outcome after mild traumatic brain injury in women. J Head Trauma Rehabil. 2014;29(5):E1–8.
30. Hutchison MG, Mainwaring L, Senthinathan A, Churchill N, Thomas S, Richards D. Psychological and physiological markers of stress in concussed athletes across recovery milestones. J Head Trauma Rehabil. 2017;32(3):E38–48.
31. McCauley SR, Wilde EA, Miller ER, et al. Preinjury resilience and mood as predictors of early outcome following mild traumatic brain injury. J Neurotrauma. 2013;30(8):642–52.
Copyright © 2020 by the American College of Sports Medicine