WITH APPROXIMATELY 1.6 TO 3.8 million sport/recreation-related concussions (SRCs) annually in the United States,1 the injury has received unprecedented attention from scientists, clinicians, and media. Large-scale initiatives, such as the Center for Disease Control and Prevention's HEADS UP program, have been created to disseminate information about concussion for athletes, parents, coaches, referees, and medical providers. Despite these efforts, 50% to 70% of concussions go unreported or undiagnosed2–4 due to limited knowledge regarding concussion signs and symptoms,3 delayed symptoms,5–8 and pressures to continue to play.9,10 Athletes report feeling pressured to remain in competition following injury,9–11 but do not understand the associated risks.12
Recently, our research team13 reported that adolescent athletes who continued to play after sustaining a concussion took twice as long to recover (44.4 ± 36.0 days) compared with athletes immediately removed from play (22.0 ± 18.7 days), and were 8.8 times more likely to exhibit protracted recovery (ie, ≥21 days). Continued play was associated with worse symptoms and neurocognitive scores up to 1-month after injury. Similarly, Asken et al14 reported that NCAA Division I athletes who continued to play following SRC took an average of 4.9 additional days to recover and play status predicted total days missed after controlling for other empirically established predictors of prolonged recovery (eg, sex, concussion history, learning disability, or attention-deficit hyperactivity disorder history). Athletes who continued to play were 2.2 times more likely to take more than 7 days to recover.
Although these studies provide evidence for the deleterious effects of continued play postconcussion, they do not address whether the duration athletes continue to play influences these outcomes. Length of postinjury play may progressively exacerbate SRC based on the pathophysiological processes underlying the injury. In the acute stages of concussion, biomechanical forces on the brain result in potassium efflux, sodium and calcium influx, and indiscriminate glutamate release.15,16 Membrane ionic pumps, requiring ATP to operate, work to restore cellular homeostasis, depleting neurons of intracellular energy.15,16 These events occur in the context of decreased cerebral blood flow, resulting in a neurological energy crisis.15,16 Continued aerobic activity may exacerbate this process by further exerting an already depleted system.15,17 Animal models and retrospective human studies provide evidence for impaired recovery with early physical and cognitive activity,18–20 and research has demonstrated that immediate physical activity following a concussion decreases neuroplasticity18 and increases neuroinflammation and cognitive impairment.19,21 Continuing to play following a concussion also exposes the athlete to additional opportunities for head trauma during a period of neurological vulnerability, as research suggests secondary head trauma within 24 hours of initial concussion is associated with exacerbated axonal injury, astrocytic reactivity,15 and poorer neurocognitive outcomes.17 Additional sport exposure following a concussion may result in a dose-response effect on concussion severity and outcomes.
The current study sought to determine whether there is a dose-response relationship between postconcussion sport exposure and recovery outcomes. We hypothesized that athletes who continued to play for a longer period would take progressively longer to recover. We also hypothesized that continuing to play for a longer period would be associated with worse symptoms and neurocognitive performance.
The current study was a secondary analysis of prospective, repeated-measures data13 from 130 athletes (aged 11-19 years) seeking care for an SRC at a concussion specialty clinic between September 1, 2014, and December 1, 2014. Inclusion criteria included (1) patient research registry enrollment with written informed consent; (2) first clinical visit within 7 days of injury; (3) ability to identify the moment during a game or practice that they sustained a head impact resulting in on-field SRC symptoms or changes in mental status (For any athlete experiencing posttraumatic amnesia, confusion/disorientation, or brief loss of consciousness, information regarding removal from play timing was obtained from an ATC, sport medicine physician, or parent. In practice, all of the athletes in the current study with these on-field concussion signs were immediately removed from play per current state concussion policy, making for a straightforward evaluation of this criterion in these athletes.); (4) no diagnosed or reported brain injury during the previous 3 months; and (5) no history of diagnosed learning disability and/or hyperactivity disorder.
Definition of sport-related concussion
Concussions were diagnosed by a certified athletic trainer (ATC) or sports medicine physician using the following criteria: (1) clear mechanism of injury; and (2) one or more on-field signs (eg, loss of consciousness, amnesia, disorientation, confusion, and imbalance) and/or symptoms (eg, headache, dizziness, nausea, and mental fogginess), and impairment on sideline concussion assessment measures (eg, Sport Concussion Assessment Tool-3: SCAT-3).
Determining continued play status groups
Demographics, medical history, and injury information were obtained via clinical interview. Questions from a removal from play clinical intake form described previously13 were used to identify the amount of time athletes continued to participate following SRC. This form included questions regarding the athletes' ability to recall the moment of injury, changes in mental status, presence of symptoms, removal from play status, and duration of continued play (if applicable). Groups were defined as (1) immediately removed from play (Removed), (2) continued to play for 15 minutes or less (Short-Play), and (3) continued to play for more than 15 minutes (Long-Play).
Neurocognitive testing and symptoms
The Immediate Post-Concussion Assessment and Cognitive Testing (ImPACT)22 was used to measure neurocognitive performance. The ImPACT is a computerized neuropsychological battery composed of 6 subtests that yield 4 composite scores including verbal memory, visual memory, visual motor speed, and reaction time. The ImPACT includes the Post-Concussion Symptom Scale (PCSS),23 a self-report measure of symptom severity involving 22 symptoms (eg, headache and dizziness) rated from 0 (none) to 6 (severe) on the Likert scale. The psychometric properties of ImPACT and the PCSS are reported elsewhere in the literature.24–29
Recovery time was defined as the number of days from the date of injury to the date of clearance for full participation by a licensed healthcare provider (ie, physician and neuropsychologist) with expertise in clinical concussion care. Consistent with international consensus,8 athletes were determined to meet criteria for return to play once symptom-free at rest and following physical exertion. In addition, when baseline ImPACT data were available, athletes were required to perform within normal limits of their baseline (ie, 80% confidence intervals utilizing reliable change indices). When baseline data were not available, neurocognitive performance was compared with normative values.
Athletes were enrolled in a university institutional review board-approved concussion research registry. During each of 2 consecutive clinical visits, athletes completed the ImPACT and PCSS, followed by an in-person clinical interview by a neuropsychologist trained in concussion. Neurocognitive and symptom data were gathered from athletes at 1 to 7 and 8 to 30 days following injury. Baseline neurocognitive and symptom data were retrieved from medical records for each patient when available to inform the determination of recovery. On-field signs and symptoms of SRC were assessed retrospectively via the clinical interview during the first clinical visit. Parents accompanying athletes to the clinic provided information in the event the athlete could not recall their on-field signs and/or symptoms, and additional information regarding length of time continued to play was obtained from sports medicine staff (eg, athletic trainer).
Descriptive statistics (eg, means, frequencies, and standard deviations) were computed for the total sample and each group for demographic characteristics and injury information. A univariate analysis of variance (ANOVA) was performed to compare differences among the 3 continued play status groups for recovery time. To investigate the association and risk for protracted recovery as a function of continued play status, recovery time was dichotomized at 21 days (with ≥21 days representing protracted recovery per previous research13,30,31) and a χ2 analysis with odds ratios (ORs) with 95% confidence interval (CI) was conducted. Between-group comparisons of neurocognitive performance (verbal and visual memory, visual motor speed, reaction time) and total symptom severity were conducted utilizing mixed-factorial, repeated-measures ANOVA, with time (1-7, 8-30 days) as the within-subjects factor. Following Bonferroni correction (for 5 univariate analyses), the α level was set at P < .01. Post hoc analyses were conducted to determine group differences in the dose-response pattern at these time points. All analyses were conducted utilizing SPSS version 22.0.32
There were no differences between the groups for demographic information or on-field severity markers for SRC (see Table 1). Across the sample, days from injury to first clinical visit was mean = 5.21, standard deviation (SD) = 4.38, while days from injury to second visit was mean = 14.93, SD = 6.90.
Results of repeated-measures ANOVA revealed a significant between-groups main effect for continued play status on recovery time (F(2,105) = 14.06, P < .001, η2 = 0.26). Post hoc analyses (see Table 2) revealed that participants in the Long-Play (>15 minutes) group took longer to recover than those in the Removed group (P < .001, d = 1.28) and the Short-Play (≤15 minutes) group (P = .01, d = 0.72). There were no differences in recovery time between Removed and Short-Play groups (P = .12, d = 0.71).
The χ2 analysis with ORs with 95% CI revealed a significantly greater likelihood of protracted recovery based on continued play status (χ2 = 26.13, P < .001, V = 0.49). Participants in the Short-Play group were 5.43 times more likely (P < .001, 95% CI = 1.91-15.46; see Table 3) to experience a protracted recovery compared with participants immediately removed from play, and participants in the Long-Play group were 11.76 times more likely (P < .001, 95% CI = 4.00-34.59; see Table 3) to experience protracted recovery relative to participants immediately removed from play. The Short-Play and Long-Play groups did not differ statistically (OR = 2.17, P = .21, 95% CI = 0.64-7.40; see Table 3) regarding the odds for protracted recovery.
Neurocognitive performance and symptoms
Results of a series of 3 (continued play status) × 2 (clinical visit) mixed-factorial ANOVAs revealed a within-subjects main effect for time (F(1,74) = 30.86, P < .001, η2 = 0.29) and a significant between-subjects effect for continued play status (F(2,74) = 8.81, P < .001, η2 = 0.19) on verbal memory. Specifically, post hoc 1-way ANOVA with Bonferroni-corrected comparisons revealed that the Removed group performed better than the Short-Play and Long-Play groups at days 1 to 7, while at days 8 to 30, the Removed group performed better than the Long-Play group (see Table 4). The visual memory composite also yielded a within-subjects effect for stage of recovery (F(1,74) = 7.19, P = .01, η2 = 0.09) and a between-subjects effect for continued play status (F(2,74) = 12.52, P < .001, η2 = 0.25), with post hoc analyses indicating the Removed group performed better than the Short-Play and Long-Play groups at days 1 to 7 and 8 to 30 (see Table 4). Similarly, a within-subjects main effect for stage of recovery (F(1,74) = 56.01, P < .001, η2 = 0.43) and between-subjects effect for continued play status (F(2,74) = 17.35, P < .001, η2 = 0.32), as well as a significant continued play status × stage of recovery interaction (F(2,74) = 6.66, P < .001, η2 = 0.15), were supported for the visual motor speed composite. Post hoc analyses revealed the Removed group performed better than both the Short-Play and Long-Play groups at days 1 to 7 and 8 to 30 (see Table 4). Reaction time also yielded a within-subjects main effects for stage of recovery (F(1,74) = 18.78, P < .001, η2 = 0.20) and between-subjects effect for continued play status (F(2,74) = 13.41, P < .001, η2 = 0.27). Specifically, the Removed group performed better than both the Short-Play and Long-Play groups at days 1 to 7 and 8 to 30, while the Short-Play group performed better than the Long-Play group at days 1 to 7 (see Table 4). PCSS total symptom score yielded a within-subjects effect for stage of recovery (F(1,74) = 105.86, P = .00, η2 = 0.59) and between-subjects effect for continued play status (F(2,74) = 8.50, P < .001, η2 = 0.19). Post hoc analyses revealed that, at days 1 to 7, the Short-Play and Long-Play groups reported significantly more severe symptoms than the Removed group. At days 8 to 30, only the Long-Play group reported more severe symptoms than the Removed group (see Table 4).
The current study is the first to examine a dose-response for continuing to play immediately following SRC on recovery and related outcomes. Results provided partial support for a dose-response effect of continuing to play following SRC, extending recent findings indicating worse recovery outcomes among athletes who continued to play following SRC.13,14 In the current study, participants in the Long-Play group exhibited the longest recovery time, which was over twice as long as those participants immediately removed from competition (19 vs 44 days) and 16 days longer than participants in the Short-Play group (28 vs 44 days). Continued play status was associated with increased odds of protracted recovery, with only 27% of participants in the Removed group demonstrating protracted recovery, whereas 67% and 81% of participants demonstrated protracted recovery in the Short-Play and Long-Play groups, respectively. Continuing to play for a longer duration was also associated with greater neurocognitive impairment and more severe total symptom scores in patients continuing to play relative to those immediately removed at both 1 to 7 and 8 to 30 days postconcussion. The 2 groups composed of participants who continued to play were significantly different on reaction time at 1 to 7 days postconcussion, with the Short-Play group demonstrating faster reaction time than the Long-Play group.
Our results regarding a dose-response for continuing to play after SRC have implications for the clinical management of athletes. The findings emphasize the importance of removing athletes from play immediately after SRC, as continuing to play even for a few minutes may have adverse consequences on recovery time. The potential reduction in morbidity associated with immediate removal from play is substantial. In the current study, 52% of adolescent athletes reported continuing to play for some length of time following their concussion. Given that there are approximately 1.9 million adolescents with concussion each year in the United States,33 this number translates to 988 000 athletes who risk prolonged recovery by continuing to play. Immediate removal from play could reduce the estimated 380 000 athletes each year who experience prolonged recovery.1,33,34 This highly modifiable risk factor following SRC could have a significant impact on reducing recovery time and associated healthcare costs.
Although the present study was the first to examine a dose-response effect for continuing to play following SRC, it is limited by several factors. The sample size was small, and represents a potentially biased group of patients who were seen within 1 to 7 days of injury at a sports concussion clinic. As such, the results may not generalize to all athletes following concussion. In addition, much of the relevant clinical information obtained regarding on-field signs and symptoms, as well as duration of post-injury play, was gathered via retrospective patient and/or observer report, and may not be entirely accurate as a result. Finally, the present study did not obtain data regarding athletes' exposure to specific events during postinjury play (eg, additional head trauma, aerobic, or dynamic exertion). As such, it cannot be determined whether the detrimental effects of continued play are a result of further trauma or continued aerobic and dynamic activity.
Future studies should directly assess what happens to athletes when they continue to play following a concussion to address this ambiguity. Due to limitations associated with retrospective recall, researchers could utilize accelerometers together with video analysis to quantify additional impacts to the head. Similarly, accelerometers and video analysis can be used to estimate aerobic activity (ie, distance traveled) in certain sports. In addition, although current recommendations are clear regarding the importance of immediately removing athletes from play when there is a suspected concussion, the present findings indicate that this often does not occur, as 52% of athletes in the current study reported continuing to play. Researchers should examine which factors contribute to athletes remaining in play including recognition, identification, and reporting of signs and symptoms; appropriate training of coaches, officials, and athletic trainers to identify this injury; and lack of medical coverage and concussion management policies.
Despite consensus statements emphasizing the importance of immediate removal from play postinjury7,8,35 and previous research demonstrating the deleterious effect of continued play on outcome and recovery,13 concussions continue to be underreported and underdiagnosed.2–4 To our knowledge, the present study is the first to provide preliminary evidence for a dose-response effect of continuing to play following SRC on recovery time and related clinical outcomes. The present results extend previous findings in animal models18,19,21 to human subjects. The current findings demonstrate the implications of continuing to play following SRC and highlight the importance of identifying and removing athletes from play. However, the mechanisms for the findings reported in this study are not clear. Moving forward, researchers need to determine whether it is additional aerobic and/or dynamic physical and cognitive activity or impacts to the head that result in the effects reported here. Removal from play represents a modifiable risk factor for prolonged recovery that should be emphasized in education and awareness programs for sport medicine professionals, coaches, officials, parents, and athletes.
1. Langlois JA, Rutland-Brown W, Wald MM. The Epidemiology and impact
of traumatic brain injury: a brief overview. J Head Trauma Rehabil. 2006;21(5):375–378.
2. Llewellyn T, Burdette GT, Joyner AB, Buckley TA. Concussion
reporting rates at the conclusion of an intercollegiate athletic career. Clin J Sport Med. 2014;24:76–79.
3. McCrea M, Hammeke T, Olsen G, Leo P, Guskiewicz K. Unreported concussion
in high school football players: implications for prevention. Clin J Sport Med. 2004;14(1):13–17.
4. Meehan Iii WP, Mannix RC, O'brien MJ, Collins MW. The prevalence of undiagnosed concussions in athletes. Clin J Sport Med. 2013;23(5):339–342.
5. Bruce JM, Echemendia RJ. Delayed-onset deficits in verbal encoding strategies among patients with mild traumatic brain injury. Neuropsychology. 2003;17(4):622–629.
6. Duhaime AC, Beckwith JG, Maerlender AC, et al Spectrum of acute clinical characteristics of diagnosed concussions in college athletes wearing instrumented helmets: clinical article. J Neurosurg. 2012;117(6):1092–1099.
7. Giza CC, Kutcher JS, Ashwal S, et al Summary of evidence-based guideline update: evaluation and management of concussion
in sports: report of the Guideline Development Subcommittee of the American Academy of Neurology. Neurology. 2013;80(24):2250–2257.
8. McCrory P, Meeuwisse W, Aubry M, et al Consensus statement on Concussion
in Sport—The 4th International Conference on Concussion
in Sport held in Zurich, November 2012. J Sci Med Sport. 2013;16:178–189.
9. Kroshus E, Baugh CM, Daneshvar DH, Stamm JM, Laursen RM, Austin SB. Pressure on sports medicine clinicians to prematurely return collegiate athletes to play after concussion
. J Athl Train. 2015;50(9):944–951.
10. Kroshus E, Garnett B, Hawrilenko M, Baugh CM, Calzo JP. Concussion
under-reporting and pressure from coaches, teammates, fans, and parents. Soc Sci Med. 2015;134:66–75.
11. Kurowski BG, Pomerantz WJ, Schaiper C, Ho M, Gittelman MA. Impact
of preseason concussion
education on knowledge, attitudes, and behaviors of high school athletes. J Trauma Acute Care Surg. 2015;79(3):S21–S28.
12. Youth CoS-RCi. Sports-Related Concussions in Youth: Improving the Science, Changing the Culture. Washington, DC: Board on Children, Youth, and Families, Institute of Medicine, National Research Council; 2013.
13. Elbin RJ, Sufrinko A, Schatz P, et al Removal from play after concussion
and recovery time. Pediatrics. 2016;138(3).
14. Asken BM, McCrea MA, Clugston JR, Snyder AR, Houck ZM, Bauer RM. “Playing through it”: delayed reporting and removal from athletic activity after concussion
predicts prolonged recovery. J Athl Train. 2016;51(4):329–335.
15. Giza CC, Hovda DA. The new neurometabolic cascade of concussion
. Neurosurgery. 2014;75(suppl 4):S24–S33.
16. Giza CC, Hovda DA. The neurometabolic cascade of concussion
. J Athl Train. 2001;36(3):228–235.
17. Longhi L, Saatman KE, Fujimoto S, et al Temporal window of vulnerability to repetitive experimental concussive brain injury. Neurosurgery. 2005;56(2):364–374.
18. Griesbach GS, Hovda DA, Molteni R, Wu A, Gomez-Pinilla F. Voluntary exercise following traumatic brain injury: brain-derived neurotrophic factor upregulation and recovery of function. Neuroscience. 2004;125(1):129–139.
19. Griesbach GS, Gomez-Pinilla F, Hovda DA. The upregulation of plasticity-related proteins following TBI is disrupted with acute voluntary exercise. Brain Res. 2004;1016(2):154–162.
20. Majerske CW, Mihalik JP, Ren D, et al Concussion
in sports: postconcussive activity levels, symptoms, and neurocognitive performance. J Athl Train. 2008;43(3):265–274.
21. Piao CS, Stoica BA, Wu J, et al Late exercise reduces neuroinflammation and cognitive dysfunction after traumatic brain injury. Neurobiol Dis. 2013;54:252–263.
22. Immediate Post-Concussion
Assessment Cognitive Testing [computer program]. Version 6.0. Pittsburgh, PA: NeuroHealth Systems; 2006.
23. Kontos AP, Elbin RJ, Schatz P, et al A revised factor structure for the postconcussion symptom scale: baseline and postconcussion factors. Am J Sports Med. 2012;40(10):2375–2384.
24. Elbin RJ, Schatz P, Covassin T. One-year test-retest reliability of the online version of ImPACT
in high school athletes. Am J Sports Med. 2011;39(11):2319–2324.
25. Iverson GL, Lovell MR, Collins MW. Validity of ImPACT
for measuring processing speed following sports-related concussion
. J Clin Exp Neuropsychol. 2005;27(6):683–689.
26. Lovell MR, Iverson GL, Collins MW, et al Measurement of symptoms following sports-related concussion
: reliability and normative data for the postconcussion scale. Appl Neuropsychol. 2006;13(3):166–174.
27. Schatz P. Long-term test-retest reliability of baseline cognitive assessments using ImPACT
. Am J Sports Med. 2010;38(1):47–53.
28. Schatz P, Ferris CS. One-month test-retest reliability of the ImPACT
test battery. Arch Clin Neuropsychol. 2013;28(5):499–504.
29. Schatz P, Sandel N. Sensitivity and specificity of the online version of ImPACT
in high school and collegiate athletes. Am J Sports Med. 2013;41(2):321–326.
30. Lau BC, Kontos AP, Collins MW, Mucha A, Lovell MR. Which on-field signs/symptoms predict protracted recovery
from sport-related concussion
among high school football players? Am J Sports Med. 2011;39(11):2311–2318.
31. Kontos AP, Elbin RJ, Lau B, et al Posttraumatic migraine as a predictor of recovery and cognitive impairment after sport-related concussion
. Am J Sports Med. 2013;41(7):1497–1504.
32. IBM SPSS Statistics for Windows [computer program]. Version 21.0. Armonk, NY: IBM Corporation; 2012.
33. Bryan MA, Rowhani-Rahbar A, Comstock RD, Rivara F, Seattle Sports Concussion
Research Council. Sports- and recreation-related concussions in US youth. Pediatrics. 2016;138(1).
34. Collins M, Lovell MR, Iverson GL, Ide T, Maroon J. Examining concussion
rates and return to play in high school football players wearing newer helmet technology: a three-year prospective cohort study. Neurosurgery. 2006;58(2):275–286.
35. Broglio SP, Cantu RC, Gioia GA, et al National Athletic Trainers' Association position statement: management of sport concussion
. J Athl Train. 2014;49(2):245–265.