Jourruil of Athk'üc Training 2a08;43(3):265-274© by the National Athletic Trainers" Association. Incwww.nata.org/jat original research

Concussion in Sports: Postconcussive ActivityLevels, Symptoms, and Neurocognitive PerformanceCynthia W. Majerske, MD, MS*; Jason P. Mihalik, MS, CAT(C), ATCt;Dianxu Ren, PhD*; Michael W. Collins, PhD*; Cara Camiolo Reddy, MD*;Mark R. Lovell, PhD*; Amy K. Wagner, MD*^University of Pittsburgh, Pittsburgh, PA; tUniversity of North Carolina. Chapei Hili, NC

Context: Evidence suggests that athletes engaging in high-intensity activities atter concussion have more difticulties withcognitive recovery.

Objective: To examine the role postinjury activity level pfaysin postconcussive symptoms and performance on neurocogni-tive tests in a population of student-athletes.

Design: Retrospective cohort study with repeated measuresof neurocognitive pertormance and symptom reporting.

Setting: University-based sports concussion clinic.Patients or Other Participants: Ninety-five student-athletes

(80 males, 15 females: age = 15.88 ± 1.35 years) wereretrospectively assigned to 1 of 5 groups based on a postinjuryactivity intensity scale.

Main Outcome Measure(s): We employed a regressionanalysis for repeated measures to evaluate the relationship ofactivity intensity to symptoms and neurocognitive outcome up to

33 days after concussion. Postconcussion symptom scores andneurocognitive (verhal memory, visual memory, visual motorspeed, and reaction time) scores served as the primaryoutcome measures.

Results: Level of exertion was significantly related to alloutcome variables (P < .02 tor all comparisons). With mul-tivariate analysis, activity intensity remained significant withrespect to visual memory (P - .003) and reaction time [P <.001).

Conclusions: Activity level after concussion affected symp-toms and neurocognitive recovery. Athletes engaging in highlevels of activity after concussion demonstrated worse neuro-cognitive performance. For these tasks, those engaging inmoderate levels of activity demonstrated the best performance.

Key Words: exertion, rehabilitation, mild traumatic braininjuries, student-athletes

Key Points

Symptom status and neurocognitive performance were affected by postconcussive activity levels and age and sex of theathlete.Younger adolescents experienced more pronounced deficits in verbal and visual memory than older teenagers afterconcussion.Moderate levels of exertion were associated with better symptom and neurocognitive performance prognosis, suggestingcontrolled exertion may improve outcome after concussion. More study in this area is needed.

M ore than 1000000 mild traumatic brain injuries(TBIs) occur each year in the United States.' Theaverage incidence of mild TBI from all causes in

the United States is estimated at 503 per 100000, withbicycles and sports accounting for 26.4% in the 5-year-oldto 14-year-old age group.i Conservative estimates indicatethat more than 300 000 sport-related concussions occureach year in the United States^: more than 60 000 cases ofconcussions occur at the high school level, with footballaccounting for the majority of these.-^ Approximately 4%of high school and collegiate football players sustainconcussions during each season.^ ^ With more than1 250000 student-athletes participating at the high schoollevel/' this is an especially important population toexamine. A discussion of sex diHerences in clinicalmeasures of concussion has recently been published^showing that males tended to perform better on visualmemory, while females performed better on verbal memorytasks; females also typically endorsed a higher number ofsymptoms during preseason baseline screening. Epidemio-

logie studies on collegiate football players have alsodemonstrated that concussed players' symptoms typicallyresolved within 7 days and neurocognitive functionreturned to baseline levels within 5 to 7 days.*^ Not muchis known about the age-related differences in postconcus-sive symptoms, and whether symptoms of concussion cansignificantly interfere with the cognitive activities student-athletes require for schooiwork is an area yet to beexplored.

Concussion is a mild TBI that results from a biome-chanical insult to the brain that initiates a destructiveneurometabolic cascade of events.^ The cascade beginswith the release of excitatory neurotransmitters. whichresult in cellular membrane disruption and ionic imbal-ances. Increasing amounts of adenosine triphosphate(ATP) are required in an attempt to correct these ionicimbalances, and an increase in glucose metabolism isobserved within the first 24 hours after concussion. Thisincreased glucose metabolism, combined with an initialdecrease in cerebral blood flow, results in a mismatch

Journal of Athletic Training 265

between the energy required and that available to brainstructures. The increase in glucose metabohsm is followedby a period of reduced glucose uptake and metabolism,which may last for as long as 1 month.^ Exercise alsomodulates glucose uptake in the brain and increasesCortisol in a dose-dependent manner, both of which couldworsen the neuronal energy mismatch after concussion. i"i '

Most of these studies have been performed on animalmodels. Although the generalizability to humans is limited,these models display many pathophyslologic and behav-ioral characteristics noted in the human condition. As such,they may provide us with some insight for clinicalphenomena observed in human brain injuries. Furtherworsening of traumatically induced metabolic mismatchmay occur with exercise or other types of activity.Worsening of postconcussive memory deficits and process-ing speed after cognitively demanding tasks has also beendemonstrated.'- In addition, variations in postconcussivesymptoms with respect to subjective levels of stress havebeen reported.'-^

Given the proposed metabolic abnortnalities associatedwith concussion, which may last longer than once believed,the decision on when and how to return an athlete to playhas i"eceived attention as a national public health issue.'•*Postconcussive activity has been clinically noted to worsensymptoms and cognition after concussion. Most of theexisting guidelines have been specifically developed forreturn lo play rather than return to the functional activitiesof school, work, or daily chores. Therefore, we aimed toexamine the role of postinjury activity level in post-concussive symptoms and performance on neurocognitivetests in a population of student-athletes. Our hypothesiswas that athletes who engaged in high levels of activityafter a concussion would have higher symptom severityscores and slower recoveries than those who engaged inlower activity levels during recovery.

METHODS

We used a retrospective cohort design to assess therelationship of symptom status and neurocognitive func-tion after sport-related concussion in student-athletes. Aretrospective chart review resulted in the inclusion of 95participants in our study (80 males: age = 15.81 ± 1.35years; 15 females: age = 16.31 ± 1.32 years) during the2002-2003 and 2003-2004 academic years. All participantswere evaluated through a university hospital system'ssports medicine concussion program. This ongoing clinicalprogram includes computerized neurocognitive tests toassist team medical staff in making return-to-play decisionsafter sport-related concussion. In order to provide us withthe inf^ormation necessary to answer our research ques-tions, patient records needed to meet 4 criteria: (I)inclusion of the patient's current academic status. (2)information pertaining to postinjury activity level insufficient detail to accurately categorize the patient, (3)the injury must have been sustained during sport partic-ipation, and (4) data from at least 2 clinical follow-up visitswere available. Patients with a history of learningdisability, seizure disorder, or attention deficit disorderwere not included in our study. As some prescribedmedications can adversely affect cognitive function, pa-ticnls who were taking any form of medication at the time

Table 1. Symptoms Rated in the Post-ConcussionSymptom

HeadacheNauseaEmesisBalance problemsDizzinessFatigueTrouble falling asleepSleeping more than usualSleeping less than usualDrowsinessIrritability

Sensitivity to lightSensitivity to noiseIncreased sadnessNervousnessFeeling more emotionalNumbness or tinglingFeeling slowed downSensaiion of being "in a fog'Difficulty with concentrationDifiiculty with memoryVisual problems

^ Each item is graded from 0 (asymptomatic) to 6 (severely symptomatic).

of injury and subsequent clinical evaluations were alsoexcluded. Use of data from patient records was approvedby a university institutional review board.

Using these criteria, our sample was obtained afterscreening 297 patient records. Most of the screened chartswere excluded because either the concussion was not theresult of sport participation or information pertaining tothe patient's current academic status was not provided.Nine athletes were excluded from the analysis of neuro-cognitive scores due to insufficient normative data tocalculate standard scores. Thus. 86 athletes were analyzedwith standardized scores on neurocognitive tests.

Instrumentation

Immediate Postconcussion and Cognitive Test (ImP.ACT).The Immediate Postconcussion and Cognitive Test (Im-PACT; ImPACT Applications. Inc. Pittsburgh. PA) is acomputer-administered neuropsychological test batteryconsisting of 7 individual test modules that measureaspects of cognitive functioning, including attention,memory, reaction time, and information processing speed.A thorough description of the ImPACT test battery andrationale for the development of the individual tests havebeen provided previously.i^ Five forms of ImPACT exist,with the word memory stimuli and design memory stimulidifferent and alternating for each form. Also, each time thetest is given, the stimuli are infmitely randomized(including the stimuli for each alternate form), furthercircumventing the typical practice effects one sees withpaper-and-pencil testing. No practice effects have beendemonstrated for 3 of 4 composite scores, with theinformation processing speed composite score having aminimal 3-point practice effect from time I to time 2."» TheImPACT has been previously shown to be a reliable^'' andvalid'"^ tool in the assessment of concussion. This test hasalso demonstrated good sensitivity and specificity in priorstudies of young athletes."*

Postconcussion Symptom Scale (PCSS). The ImPACTalso yields a Postconcussion Symptom Scale (PCSS) that isnow being utilized throughout both amateur and profes-sional sports. 19 This Likert scale consists of 22 symptomscommonly associated with concussion, which are gradedfrom 0 (asymptomatic) to 6 (severely symptomatic). All thestudent-athletes were required to provide a self-report ofsymptoms based on the PCSS, which included bothcognitive (attention deficits and perceived memory dys-function) and noncognitive (headache, nausea, dizziness.sleep disturbance, emotional changes, and photophobia)

266 Volume 43 • Number 3 • June 2008

symptoms {Table I ). Reliability and normative data for thePCSS have been previously e.stabtished,-"

Main Outcome Measures

Neurocognitive Function and Symptom Status. TheimPACT yields outcome measures for verbal memory,visual memory, visual motor speed, reaction time, andimpulse control composites. A faster reaction time and alower symptom score indicate a better response. Converse-ly, higher scores for verbal and visual memory and visualmotor processing speed indicate better performance. Dueto the psychometric properties of the imPACT. weevaluated each composite score .separately as an individualoutcome measure. In order to do so, each composite scorewas converted to a standard score (Z-score) based on age-and sex-matched normative data available on the ImPACTdevelopers' Web site; normative symptom data werederived from 707 high school students.-" These standardZ-scores have been converted to percentiles for the purposeof presentation because most clinical certified athletictrainers are accustotiied to viewing ImPACT results in thisway. Symptom status was recorded as a total symptomscore and was assessed by PCSS testing performed at eachfollow-up visit. Patients referred to our medical center arefrom the community, referred by other physicians, or sentto us by certified athletic trainers working in the area.Baseline testing results are often not available, so we didnot include baseline as a time interval in our analyses.

Independent Variables

Concussion Grade and Return to Play. Concussion gradewas determined by a single rater using the ColoradoConcussion Scale (CCS),21 with information obtained fromchart review regarding the presence or absence ofconfusion, amnesia, and loss of consciousness immediatelypostconcussion. Return to play was a categoric variablescored as the athlete was returned to play (ye.<i) or notrettn-ned to play {no) in the same event in which theconcussion was sustained.

Activity Intensity Scale. An activity intensity scale (AIS)was developed for the purpose of this study and futurestudies. Level of activity was determined by a single raterfor each time interval in which an evaluation occurred. TheAIS consists of 5 categories designed to be ordinal innature based on information recorded in the chart; noschool or exercise activity {()). school activity only (7),school activity and light activity at home (eg, slow jogging,mowing the lawn) (2), school activity and sports practice(i), and school activity and participation in a sports game{4). Our AIS was not designed as a guide for return-to-playprogression btit rather as a categorization of the student-athlete's activity level {whether it was cognitive or physical)after injury. Data were derived from self-reported activitynoted in the patient records by the clinician performing theclinical evaluation. In our study, the same investigatorrated both the CCS and the AIS.

Reliability Assessment of Independent Variables

The CCS has been used for several years to evaluateconcussion severity. Although the criteria for calculatinginjury severity (presence of confusion, amnesia, or loss of

consciousness) intuitively lend themselves to the possibilityof extracting this information from patient charts, to ourknowledge, no authors lo date have reported the reliabihtyor validity of chart extraction of the information requiredto calculate this measure. Additionally, the informationrequired to determine the AIS at each poslinjtuy assess-ment point (time 1 through time 5) was taken from chartreview. Student-athletes were included in the study ifadequate information was available in the chart to assign ascore. The AIS is a new measure for which reliabilitymeasurements have not been reported. As such, weestablished interrater reliability estimates for each of thesemeasures using a subset of charts from student-athletesevaluated clinically at a university medical center's sportsconcussion program. Twenty-three charts, which included64 separate evaluations, were reviewed by 2 independentraters; athletes included in the review met the sameinclusion criteria as above. After reviewing the criteriaassociated with establishing CCS and AIS grades, eachrater independently obtained information relevant toassessing CCS and AIS from each chart. The 2 raters thena.ssigned a CCS grade for each athlete and an AIS score foreach athlete evaluation. The 2 physicians had the CCS andAIS criteria in front of them as they independentlyevaluated the content of the patient charts for the purposeof the interrater reliability assessment. The principalinvestigator {a physical medicine and rehabilitation physi-cian) continued to assign CCS and AIS grades for theremainder oi' the cases in our study. Two chnicalneuropsychologists with more than 30 years of combinedexperience dealing with athletes who have sustained sport-related concussion conducted the initial and any follow-upclinical evaluations of the patients in our study. A standardclinical interview form was used, including questionspertaining to the type, frequency, duration, and intensityof both physical and cognitive exertion. Two physicalmedicine and rehabilitation physicians experienced intraumatic brain injury evaluation and rehabilitationconducted the initial chart review.

Data Analysis

We considered a number of independent variables fortheir relationship to postconcussive symptoms and neuro-cognitive performance. These variables included age, sex,concussion grade, self-reported history of concussions,return-to-play status, type of sport in which the concussionoccurred, level of postinjury activity as assessed by the AIS,and time of evaluation (in days). Age was categorized as 13to 15 (mean = 14.39 ± 0.66) or 16 to 18 (mean = 16.81 ±0.65) years. Because the outcome measures were obtainedin a clinical setting and at different periods of time after theinjury, outcome data were organized into discrete andclinically relevant time periods (time postconcussion) inwhich the athlete completed testing as follows: day ofinjury to day 3 (time 1), between days 4 and 7 (time 2),between days 8 and 14 {time 3), between days 15 and 21(time 4), and between days 22 and 33 {time 5), For thoseathletes who had 2 testing sessions within a specific timeinterval. ImPACT composite and symptom scores wereaveraged, and the higher of the AIS scores was used. In anygiven postconcussion time interval, no more than 5 athletesreceived 2 evaluations.

Journal of Athletic Training 267

Table 2. Participants' Demographic Information and Injury Data

Variable Number (Percentage)

Sex

MalesFemales

Sport

FootballOther

Concussion grade»Grade 1Grade 2Grade 3

Prior concussions?ti

NoYes

Return to play?Did not return to playReturned to play

80 (84.2)15 (15.8)

56 (59.0)39(41.0)

37 (39.0)44 (46.3)14 (14.7)

54 (58.1)39 (41.9)

58(61.1)37 (39.0)

^ Concussion grade was determined using Ihe Colorado ConcussionScale.

" Data were missing for 2 participants.

Summary statistics, including mean and standard errorof the mean (SEM) for all continuous variables, werecalculated. The effect size was calculated based on thestandardized mean difference between groups (ie. Cohend).-- Frequency distributions were detertnitied for categoricvariables. Because postconcussion symptoms and neuro-cognitive performance were measured at multiple timeintervals for each patient, repeated measurement analyseswere performed using a mixed-effects model with acompound symmetry covariance structure to test theoverall time effect within a group for each outcomemeasure. With this technique., multivariate regressionmodels were created to evaluate AIS and other clinicallyimportant factors affecting postconcussion symptoms andneurocognitive performance across a recovery periodlasting as long as 33 days. All independent variables wereconsidered when constructing each mtiltivariate regressionmodel, and interactions among time periods, sex, return toplay, age, concussion grade, history of concussions, andsport were explored for each multivariate model. Thereference category for the AIS in multivariate regressionanalysis was the category consisting of school activity atidlight activity at home, such as slow Jogging or mowing thelawn (AIS category 2), as preliminary analysis suggestedthat athletes in this category achieved the best outcomes.Means adjusted for individual time-dependent covarianceare reported for AIS in the multivariate table. Meanspresented in the multivariate table were also adjusted forother independent variables included in the models.

In order to assess interrater reliability for chartextraction of the CCS and the AIS, K coefficients andassociated confidence intervals were calculated for each ofthese variables. All analyses were performed using SAS(version 8; SAS Institute Inc, Cary, NC). The level ofsignificance was set a priori at .05.

RESULTS

Descriptive Analysis

Demographic information and injury data regardingconcussion grade, history of concussions, and return toplay status are provided in Table 2. Unadjusted standard-ized percentile means for the verbal memory, visualmemory, visual tnotor processing speed, and reaction timecomposites on the 86 athletes for whom standardizedscores could be calculated are shown in Table 3. Thesedata illustrate that the mean percentiles for each neuro-cognitive test at time 1 were considerably impaired relativeto normative scores but displayed improvement over time.Data at time 5 demonstrated small decrements in allpercentiles. Additional analysis of the athlete populationat each follow-up time point suggests that a largerproportion of time 5 athletes were female, with as manyas 53.3% ofthe female athletes still being followed at time5, whereas only 26.3% ofthe male athletes were still beingfollowed at this time (P = .037). None of the otherindependent variables—concussion grade, number ofprior concussions, return to play, or age—were differentbetween time 5 athletes and athletes evaluated at othertime points. Thus, these independent variables did notappear to contribute to the likelihood of prolonged clinicalfollow-up of athletes in our sample in the same way thatsex did.

Regression Analysis

Multivariate regression analysis revealed that adjustedsymptom scores decreased {P < .001) at each time interval,suggesting an improvement in the self-report of syrnptomsover time. Neurocognitive scores demonstrated improve-ments {P ^ .002) over time. We observed a trend betweentotal symptom score and intensity of activity afterconcussion (P = .08). Table 4 provides the adjusted means,standard errors, effect sizes, and percentile rankingspertaining to our multivariate analyses. Some ofthe largesteffect sizes noted in the analysis were in activity intensityand outcome.

After adjusting for other variables, a main effect wasnoted for intensity of activity (AIS score) on visual memory(P = .003) and reaction time (P < .001) (Figure 1).Evaluation of adjusted means suggests that athletes

Table 3. Unadjusted Mean Standardized ImPACT Composite Z-Scores (Percentile Rank)

Time Postconcussion

Composite score Time 1 Time 2 Time 3 Time 4 Time 5

Verbal memoryVisual memoryVisual motorReaction

-1.77(3.14)-1.34(9.85)-0.91 (17.36)

1.73(3.36)

-0.77 (19.22)-0.64 (22.96)-0.42 (32.64)

0.72 (21.19)

-0.74 (23.58)-0.34 (37.45)-0.06(51.20)

0.41 (35.57)

-0.11 (45.22)-0.06 (50.80)

0.53 (68.08)-0.42(68.79)

-0.90 (33.72)-0.27 (49.60)

0.08 (59.48)-0.32 (67.36)

A higher reaction time composite Z-score indicates slower performance. Increases in composite Z-scores vt/ould result in a lower percentile rank.

268 Volume 43 • Number 3 • June 2008

V)

3O

•ocnuiEoo.E(/)c:o

m lu55

CM 01 COi n CM l oo i CO .,"

S3 11 w î i T- ^ CM

• - CM CO

CO ^ ÖCO CO Tj-

CO T - O )

CM t o cri

CO CM ^

CM

1

a>

o

o

II

i n

oo

o II

in oCM O¡^ d1 I

a ° *

i n CD -=,

E cu TJ -= O) S. S s.

u. O £• S.

Il <U 3 . § .

T- —. T — r -..I Oi in oo (M CO

CO in ^ CM CM CM

.2 O

<

0>

D)O

5

ose

CD>COo, .oc£

iw0)oÜCO

E

*—OG)

3^rac

ssio

u¡2OO

ra

_>.5Q

—

œoE•o

yû i• -

omCD 2 >

CD J :

< o

E *^— CM

(D

OCO

ra"ô5en

nce

a

roSoÜc0)

T3

<DE

ove

>O

impr

at

\eU

<

>

(D. ^COniLU

li

E

wer

e

cTO

CO

"o(0O)

O•Ö-S

i= o

lira 2 - ëü c ^ ^ro 3 — w

S CO X :• t .9Q. 2

Journal of Athletic Training 269

60

50

Verbal memory Visual memory Visual motor speed

Outcome Measure

Reaction time

Figure 1. Effects of exertion on recovery. Athletes participating in school activity and light activity at home (eg, slow jogging, mowingthe lawn) performed better than athletes experiencing other levels of exertion. Percentiles of time-adjusted mean scores were derivedfrom the multivariate analysis. "Level" refers to the activity intensity scale in which 0 indicates no school or exercise activity and 4indicates school activity and full sport participation.

engaging in the highest activity levels (AIS = 4) over timehad the worst visual memory (adjusted mean seore =-2.22). scoring below the 2nd percentile. In fact, adjustedmean comparisons from the multivariate analysis suggestthat athletes with AIS = 3 and AIS = 4 were moreimpaired in visual memory than AIS = 2 (^ ^ .05 allcomparisons). Additionally, reaction times were the worstfor the AIS = 4 group (adjusted mean score ^ 3.51). withperformance below the 1st percentile. Comparison ofadjusted means derived from the muhivariate analysissuggests that athletes with an AIS ^ 4 had slower reactiontimes than those with AIS = 2. Those athletes whoreturned to piay in the same contest performed better overtime on visual memory subtests than those who did notreturn to play {P < .001).

Age was also associated with verbal memory (P = .02)and visual memory (P = .03) over time, with youngerathletes performing more poorly on both composite scores.Concussion grade was not associated with any of theoutcome measures. Females performed worse on visualmotor speed than males {P = .05). Analyses evaluatinginteractions among independent variables demonstrated aninterrelationship among sex. return lo play, and performaneeon visual memory tests (P < .001). For females, those whodid not return to play had lower visual memory scores thanthose who did return to the event in which ihe concussion

occurred (Figure 2). In fact, those females returning to playin the same contest had an adjusted mean performance thatwas above average normative values. In contrast, no suchrelationship was observed in males.

25• Returned !• play

• Did not return to play

Male Female

Sex

Figure 2. Visual memory between the sexes and return-to-playstatus. Percentiles of adjusted mean scores were derived from themultivariate analysis. Only significant interaction terms arereported.

270 Volume 43 • Number 3 • June 2008

Reliability Assessment of Outcome Measures

The K coefficient for obtaining similar CCS scores for 2independent raters through chart review was 0.87 (95%conlidence interval = 0.70. 1.04), and the K eoefficient fordelineating the AIS score with 2 independent raters was0.87 (95% confidence interval = 0.78. 0.97). These restiltsindicate excellent interrater reliabilities ior each of thesemeasures.

DISCUSSION

The results of this study demonstrate an importantrelationship between postconcussion activity and perfor-mance on visual memory and reaction time neurocognitivetests. Although we found no statistically significantrelationship between symptom scores and levels of activityfollowing injury, clinical trends were observed in oursample. The fact that only 2 ImPACT composite scoreswere sensitive to cognitive changes after concussion is notsurprising; different cognitive domains are known to beserved by different brain regions and systems. Also, speedand reaction time can be considered to be more subcorticalfunctions and memory functions more cortical in nature. Inother words, cerebral concussion injuries can affect variousneural networks. Thcret^ore. every test in a concussionbattery is unlikely to be significant in a particular sample.Thus, our belief that clinicians should use a multifacetcdapproach to the evaluation and management of headinjuries, paying careful attention to address the manyaspects of brain function that may be impaired after injury,is strengthened. Younger athletes demonstrated worsepostconcussion neurocognitive performance than the olderathletes in our sample. Sex associations were noted withposlconcussion management and neurocognitive perfor-mance. Additionally, the results suggest that an AIS scorecan be reliably assigned based on chart review byindependent observers and that the visual memorycomposite score appears to be more sensitive than otherneurocognitive scores in identifying poor outcomes.

Exertion

The poor performance in the high activity group (AISlevel 4) over the course of postinjury assessments mightbest be explained by the effect of exertion on neurocogni-tive performance. Exercise is considered an integral aspectof most rehabilitation protocols relating to athletic injuries;however, the scientific understanding of rehabilitation inthe context of concussion remains inconclusive at best. Inanimal studies, re sea re hers--^-4 have demonstrated thatvoluntary exercise in the uninjured brain results in anupregulalion of trophic factors, specifically brain-derivedncurotrophic factor, which contributes to experientialneural plasticity. However, recent studies in animals haveshown that after experimental TBI. rats engaging involuntary exorcise early after injury perfornied morepoorly on tasks of learning acquisition and memory andhad a reduction in plasticity-related proteins comparedwith nonexercising injured rats and healthy rats thatunderwent a sham treatment.25 in contrast, injured ratssubjected to a delayed voluntary exercise paradigm showedincreases in neurotrophins and better performance oncognitive tasks,^? indicating that the timing of voluntary

exercise after injury is important. These animal studies lendsome support to our findings, which indicate poorerperformance on visual memory and reaction time compos-ite tests in athletes who engaged in the highest activity level(AIS = 4) after concussion.

Although physical activity has been the primary focus ofpostconcussion activity restrictions, cognitive activityshould also be con.sidcred. All of the athletes in this studywere student-athletes and. therefore, involved in some sortoí' cognitive activity after concussion. Given that bothmental and physical exertion can change the metabolicactivity of the brain, •<> cognitive activities could alsopotentially worsen the metabolic mismatch after concus-sion. Our resuits suggest that cognitive rehabilitationinterventions and other cognitive activities (eg, school-work) in the population with mild TBI may need to bestudied for their effects on central nervous systemphysiology after TBI and tailored so that they are notdetrimental to recovery. For the high school or collegiateathlete, this might include reducing coursework, shorteningthe school day, rescheduling examinations, and offeringpersonal one-on-one class sessions, in addition to reducingphysical exertional demands. Anecdotally. in these chartreviews, cognitive activities associated with school ap-peared to be an issue for many concussed athletes.Although we did not separate this aspect out with theAIS, future investigators should quantify how cognitiveactivities specifically contribute to recovery and perhapseven study a graded return to school or cognitive activities.

Postconcussion symptoms, visual memory, and reactiontime with activity levels over time were related in thesestudent-athletes. Our effect sizes suggest that athletes whoengaged in the highest level of activity tended todemonstrate the worst neurocognitive scores and slowestreaction times, while on average, those who engaged in theintermediate levels of activity had the best scores andfastest reaction times. We speculate that most athletes inthe highest-intensity activity group probably experienced aless severe initial injury, but by continuing with high levelsof activity, they began to present similarly to those^athletesexperiencing a more severe or symptomatic initial injury.Conversely, athletes who were initially most symptomaticmay have been more inclined to limit their activity afterinjury. Due to the retrospective design employed in thisstudy, we are unable to determine if this is true. Futureprospective studies should better document these findingsin order to more appropriately answer these researchquestions. The AIS was somewhat sensitive to symptomreporting and reached statistical significance with 2composite scores of neurocognitive function. Visual motorprocessing speed and verbal memory were not related tothe AIS. However, the visual motor processing speedcomposite was not sensitive to any independent variables,and the verbal memory composite was sensitive only toage. Lack of associations with performance within each ofthese domains may be due to heterogeneity of the injury,the sensitivity of the AIS scale, the sensitivity of thecomposite scores, or all 3 factors.

Overall, athletes performing the best on neurocognitivetests and reporting the lowest symptom scores engaged inintermediate levels of activity after concussion; hence,further investigation into possible mechanisms to explainthis relationship is needed. One possible way to prospec-

Journal of Athletic Training 271

tively investigate the effects of exertion (both cognitive andphysical) would be to reevaluate athletes 7 days after a fullreturn to play categorized into 2 groups: those returning toplay based on a traditional management protocol (eg,American Academy of Neurology return-to-play guide-lines f') and those cleared for return after a multifacetedprotocol including symptom scoring, computerized neuro-psychological testing, assessment of postural stability, andmodification and supervision of physical exertion, as wellas tapering of academic and cognitive tasks. Outcomemeasures could consist of neurocognitive test scores,postural stability, academic achievement, and on-fieldperformance measures, to name a few.

AgePrevious authors ' - s have reported conflicting results

regarding the relationship of age and recovery from TBI,but most have examined only severe injury and theextremes of age, with both the elderly and infants andtoddlers having the worst outcomes. The literature is sparseregarding the effect of age on recovery after concussion inadolescents, with the exception of one study-'' demonstrat-ing that high school athletes have slower cognitive recoverythan collegiate athletes. Our results are consistent with thiswork, in part, by showing worse performance on verbaland visual memory tests in the younger student-athletes inour sample. Our findings warrant further study into therole age-specific mechanisms of TBI pathophysiology mayhave in mediating this association.3"^ ''

SexAlthough several researchers^^ 39 have suggested ncuro-

protection for females after TBI, other studies'*"-^- suggestpoorer outcomes for females after TBI. Additionally, highschool and eollegiate female athletes have a higherincidence of concussion than their male counterparts- -'*^and may be at greater risk for postconcussive syndromeafter mild TBI,-** Our data indicate that females werefollowed clinically for a longer period of time than malesand were more impaired on visual motor speed testing thantheir male counterparts; no differences were observed fortheir performance in other domains of neurocognitivefunction. Although previous authors have demonstratedbaseline sex difTerences on neurocognitive tests, withfemales generally scoring lower on visuospatial tests thanmales, - - ^ our data account for these baseline differencesby using standardized Z-scores. Although the number offemales in our sample was small, these findings areconsistent with other data suggesting that females mayperform worse after TBI.

As a whole, athletes who returned to play in the samecontest in which the injury was sustained were lessimpaired on visual memory testing, likely indicating lesssevere signs and symptoms of injury on the field and ahigher probability of appropriate decision making bycoaches and athletic trainers regarding return to play. Infact, as a group, athletes returning to play scored aboveaverage on visual memory over the time course studied.Interestingly, postconcussion management (return to play),sex. and neurocognitive performance were related in thatfemales who returned to play in the same contest in whichthe concussion occurred performed significantly better on

visual memory tasks than those who were removed fromthe contest. Conversely, males performed the same,whether or not they returned to play. This may indicatedifferences in early management based on sex, possiblyrelated to sex differences in immediate symptom report-ing47 or other factors not evaluated. However, the lack ofdifferences in neurocognitive performance for males in thiscomposite score and rcturn-to-play status suggests thatperhaps some male athletes require more vigilant monitor-ing to determine the appropriateness of being returned toplay after eoncussion.

Concussion GradeNumerous concussion grading scales have been present-

ed in the literature. Unfortunately, return-to-play guidelinesystems have been predicated on anecdotal experience andtend to rely heavily on loss of consciousness as a measureof severity. Previous investigators^« have demonstratedthat amnesia, not loss of consciousness, predicts cognitivedeficits after concussion. Although the retrospectiveassignment of CCS grade has not been previously reported,the CCS lends itself well to the identification and extractionof injury-related tactors (presence or absence of confusion,amnesia, and loss of consciousness) needed to assign ascore. The chnical charts for this study were provided by 2investigators who routinely provide detailed informationregarding symptoms and events surrounding concussions.The choice of concussion scale, along with the completeand detailed histories consistently provided for athletes,likely contributed to the high interiater reliabilities notedfor the CCS in this study.

LimitationsThis study represents an initial analysis relating activity

levels after concussion in student-athletes and theirlongitudinal neurocognitive performance and symptomreporting. The retrospective nature of the study meansthat it is subject to a number of considerations andlimitations when we interpret the results. Athletes in thisstudy were students seeking treatment for sport-relatedconcussion and. therefore, may not represent all athleteswho sustain concussion. Because they were evaluated inthe clinical setting, they may have had more initialsymptoms, more or less motivation, or more socialsupport, all of which could affect symptom reportingand performance on neurocognitive tests. Data werecollected in a clinical setting, so the time for each follow-up visit varied among athletes. However, we attempted toaddress this issue by categorizing the data into discreteand clinically relevant time intervals. Even when we usedthis approach, a significant number of athletes had fewerthan 5 observations for the data set. Reasons for this arevaried and include time to referral and scheduling ofappointments, as well as limited need for follow-up at thetime intervals defined for analysis. Additionally, baselinetesting was not available tor all athletes, precluding use ofthis factor in analysis. However, the use of Z-scoresderived from a normative data set available on theImPACT developers' Web site, along with publishednormative symptom data,2" helped us differentiatebetween inherent and TBI-related differences in age andsex on performance.

272 Volume 43 • Number 3 • June 2008

The level of activity was self-reported and recordedclinically, which could introduce recall bias into activityreporting. The AIS has not yet been prospcctively validatedto evaluate how well AIS scores correspond with a specificmeasurable level of exertion; however, the interraterreliability data presented in this study suggest that themeasure can be consistently obtained from chart review.Further. AIS levels for each individual at each time pointwere captured through our mixed-effects regression anal-ysis and allowed us to assess specifically how activity levelsover time affected longitudinal outcomes. However,authors of future prospective studies should quantify andmanipulate cognitive and exercise load after concussion fora more accurate measure of exertion levels and use thisinformation comparatively with the AIS. The AIS level Imay have been too broad, and levels 2 to 4 may haveshared commonalities. Within AIS level I, we acknowledgedifierences among representative activities within eachcategory (eg, mowing the lawn, slow jogging). With thistype of retrospective analysis, we could not control for orquantify the speed of slow jogging, duration of running, andwhat was considered school activity. We did not demarcatethe différence between regular school class activity andparticipation in physical education class. This is a generallimitation to self-reporting in retrospective sttidies in the areaof sport-related concussion, a factor that all researchers inthis area acknowledge needs improvement. The AIS is aninitial scale, and we anticipate that further i-efiiiements mayhelp to increase its sensitivity and further strengthen theinterrater reliability ofthe instrument.

This study was limited to student-athletes, andfurther work is required to generalize these results tolarger, more diverse populations over longer time periods.However, the longitudinal nature of this study—up to 33days postconcussion—and the potential importance ofactivity intensity for recovery in all persons with mildTBI strengthen the conclusions and management implica-tions raised. Importantly, our AIS scale was not de-signed as a guide for return-to-play progression butrather as a categorization strategy of the student-athlete'sactivity level (whether it be academics or athletics) afterinjury.

Conclusions and Management Recommendations

This study provides support for an individualized.graded return-to-play protocol but also highlights thenotion that concussion management may need to includerecommendations regarding return to all activities,including school, work, and daily chores, and not justreturn to sport-specific activity. Additionally, our dataindicate differences in recovery based on age and sex.which suggest that different strategies for postconcussionactivities may be needed in males and younger student-athletes. Future prospective studies should quantify bothcognitive and physical exertion after concussion. At firstglance, it appears that a moderate level of exertion (AIS= 2) may be relatively beneficial in recovery fromconcussion. These findings may lead to rehabilitationstrategics, for which we will need to consider the roles ofboth cognitive and physical exertion on outcomes ofstudent-athletes recovering from sport-rclated concus-sion.

ACKNOWLEDGMENTS

Support for this project was contributed in part byN1HK08HD4Ü833. We thank Jamie Pardini. PhD, for herassistance in preparing the human subjects approval for thisstudy. We also thank Kevin M. Guskiewicz. PhD. ATC, FACSM.for his critical review of the manuscript.

FINANCIAL DISCLOSURES

Michael W. Collins. PhD. and Mark R. Lovell. PhD. arcstockholders in ImPACT Applications. Inc. the company thatdevelops and sells ImPACT software. None of the other authorshave any financial or other conflicts of interests to disclose.

REFERENCES1.

9.

11).

11.

12.

13.

14.

15.

16.

17.

18.

Baziirian JJ. McClung J. Shall MN. Cheng YT. Klcsher W. Kraus J.Mild iraumalic brain injury in the United Slates. 1998-2000. BrainInj. 20()5;19(2):85 91.

Thiirman DJ. Branche CM. Sniczek JE. The epidemiology of sports-related tratiniatie brain injuries in the Uniled Slates: recentdevelopments. J Head Trautna Rehahil. 1998:13(2):! -8.Powell JW. Barber-Foss KD. Traumatic brain injury in high schoolathletes. JAMA. Í999:282(10):958 -963.

Maroon JC. Field M. Lovell M. Collins M. Bost J. The cvitluaiion ofathletes with cerebral coneussion. Clin Xcurosurg. 2()02;49:319 .132.McCrea M, Kdly JP. Kiuge J. Ackley B, Randolph C Standardiyxxlassessment of concussion in football playees. Neurology. 1 W7;48(3);586-588.National Federation of State High School As.sociaiions Handbook:1996 High School Athletics Participation Survey. Kansas City. MO:National Federation of Slate High School Associations; 1997.Covassin T. Swanik CB. Sachs M. et al. Sex differences in baselineneuropsychological fiuiction and concussion symptoms of collegiateathletes. Br J Sports Med. 2006:4(K 11 ):923-927.

McCrea M. Giiskiewiez KM. Marshal! SW. et al. Acute clïects andrecovery time following coiicu.ssion in collegiate football players: theNCAA Concussion Study. JAMA. 2Ü03;290(t9):2556-2563.Giza CC. Hovda DA. The neurometabolic cascade of concussion.J Athl Train. 2OOI;36(3):228-235.

Dalsgaard MK. Quistorlï B. Danielsen ER. Selmer C. Vogelsang T.Sécher NH. A reduced cerebral metabolic ratio in exercise reflectsmetabolism and not accumulation of lactate within the human brain.J Physiol. 2(l()4:554(pt 2):571-578.

Radosevich PM. Nash JA. Lacy DB. O'Donovan C, Williams PF,Abumrad NN. Effects of low- and high-intensity exercise on plasmaand cerebrospinal lluid levels of ir-beta-endorphin. ACTH. cortisol,norepinephrine and glucose in the conscious dog. Brain Res.1989:498(0:89-98.Hanna-Pladdy B. Berry ZM. Bennett T. Phillips HL. Gouvier WD.Stress as a diagnostic challenge for postconcussive symptoms:sequelae of mild traumatic brain injury or physiological stressresponse. Clin Neiiropsychol. 2{H1I;15(3):289 304.Gouvier WD, Cubic B, Jones G, Brantlcy P. Cutlip Q. Posiconcus-sion symptoms and daily stress in normal and head-injured collegepopulations. Arch Clin Neiiropsychol. 1992;7(3):193-2t I.Kelly JP. Traumatic brain injury and concussion in sports. JAMAÍ999;282(1Ü):989-99I.

Maroon JC. Loveii MR. Norwig J, Podell K, Powell JW, Hartl R.Cerebral concussion in athletes: evaluation and neuropsychologicaltesting. Neurosurgcry. 20()0:47(3):659-672.tverson GL. Lovell MR, Collins MW. Interpreting change onImPACT following sport concussion. Clin Neurop.syehol. 2003; 17(4)'460 467.Iverson GL, Lovell MR. Collins MW. Validity of ImPACT formeasuring processing speed following sports-related concussion.J Clin Exp Ncuropsyehol. 2005;27(6):683-689.Schatz P. Pardini JE, Lovei! MR, Collins MW, Podell K. Sensitivityand specificity of the tmPACT Tesi Battery tor concussion in athletes.Arch Clin S'europsycliol. 2OO6:2l(l):9l-99.

Journal of AthletJc Training 273

19. Lovell MR, Collins MW. Neuropsychological assessment of thecollege football player. J Head Trauma Rehabii 1998;l3(l2):9-26.

20. Lovell MR, Iverson GL. Collins MW, et al. Measurement ofsymptoms following sports-related concussion: reliability and nor-mative data for the post-concussion scale. .Appl NeiiropsychoL2i)0():13(3):l66 174.

21. Kelly JP. Nichols JS, Filley CM. Lillehei KO, Rubinstein D.Kleinschmidt-DeMasters BK. Concussion in sports: guidelinesfor the prevention of catastrophic outcome. JAMA. 199I;266(20):2867-2869.

22. Cohen J. Siaiisikal Puwci- Analysis for ihe Behiiviiiral Sciviiccs. 2ndcd. Hillsdiik. NJ: Lawrence Erlbauni Assoc: 1988.

23. Gomez-Pi ni lilt F. So V. Kesslak JP. Spatial learning and physicalactivity contribute to the induction of fibroblasi growth factor: neuralsubstrates Tor increased cognition associated with exercise. Nciiro.'ici-encc. 1998;85( 0:53-61.

24. Griesbach GS. Hovda DA. Molteni R. Wu A, Gomez-Pinilla F.Voluntary exercise following traumalic brain injury: brain-derivedneurotrophic factor upregulation and recovery of function. Ncuro-.scic-iicc. 2O1W:125(1):129-I39.

25. Griesbach GS. Gomez-Pinilla F. Hovda DA. The upregulation ofplasticity-related proteins following TBI ¡s disrupted with acutevoluntary exercise. Brain Res. 2()04:lül6(2):154-!62.

26. Practice parameter: the management of concussion in sports {summarystatement). Report o\' the Quality Standards Subcommittee of theAmerican Academy ofNeurology. Neurology. 1997:48(3):581-585.

27. Farin A. Deutsch R. Biegon A. Marshall LF. Sex-related differencesin patients with severe head injury: greater susceptibility to brainswelling in female patients 50 years of age and younger. J Neurosurg.2003;98(l):32-36.

28. Mosenthal AC. Lavery RF, Addis M. et ill. Isolated traumatic braininjury: age is an independent predictor of mortality and earlyoutcome. J Trauma. 2(IO2:52(5):9Ü7-911.

29. Field M. Collins MW. Lovell MR. Maroon J. Does age play a role inrecovery from sports-related concu.ssion? A comparison of highschool and collegiate athletes. J Fediair. 2003:l42(5):546-553.

30. Biagas KV. Grundl PD, Kochanek PM, Schiding JK. Nemoto EM.Posttraumatic hyperemia in immature, mature, and aged rats:autoradiographic determination of cerebral blood flow.J Neurali-auma. 1996:13(4):189 200.

31. Grundl PD, Biagas KV. Kochanek PM. Schiciing JK. Barmada MA,Nemoto CM. Farly cerebrovascular response to head injury inimmature and mature rats. J Neuwirawna. 1994;! 1(2): 135 -148.

32. McDonald JW. Johnston MV. Physiological and pathophysiologicalroles of excitatory amino acids during central nervous systemdevelopment. Brain Res Brain Res Rev. 1990:15(l):41-70.

33. McDonald JW, Silverslein FS. Johnston MV. Neurotoxicity of N-methyl-D-aspartatc is markedly enhanced in developing ral centralnervous system. Hrain Res. 1988:459(l):200-203.

34. Wagner AK. Bayir H. Ren D, Pticcio A. Zafonte RD. Kochanek PM.Relationships between cerebrospinal Huid markers of excitotoxicity.ischemia, and oxidative damage after severe TBI: the impact ofgender, age. and hypothermia. J Neurotrauma. 2004;2!(2):125-136.

35. Behl C. Oestrogen as a neuroprotective hormone. Nat Rev Neurosa.2002;3(6):433 442.

36. Mendelowitsch A, Ritz MF. Ros J, Langemann H, Gratzl O. 17beta-Estradiol reduces cortical lesion size in the glutamate excitotoxicitymodel by enhancing extracellular lactate: a new neuroprotectivepathway. Brain Res. 2001:901(1-2):230-236.

37. Roof RL, Hall ED. Estrogen-related gender difference in survival rateand cortical blood flow after impact-acceleralion head injury in rats../ Neuromnima. 2000; 17( 12): I 155-1169,

38. Roof RL. Hoffman SW, Stein DG, Progesterone protects againstlipid peroxidalion following traumatic brain injury in rals. Mal CiiemNeuropalhol. 1997:31(1 0:1-11.

39. Zliang YQ, Shi J, Rajakumar G. Day AL, Simpkins JW. Effects ofgender and eslradiol ireatment on focal brain ischemia. Brain Res.1998;784(l-2):321-324.

40. Farace E. Alves WM. Do women fare worse? A metaanalysis ofgender differences in traumatic brain injury outcome. J Ncurosurg.2000;93(4):539-545.

41. Morrison WE. Arbelaez JJ. Fackler JC. De Maio A, Paidas CN.Gender and age effects on outcome after pédiatrie traumatic braininjury. Pediair Crii Care Med. 2OO4;5(2):145-I5L

42. Wagner AK. Hammond FM, Sasser HC. Wiercisiewski D. NortonHJ. Use of injury severity variables in determining disability andcommunity integration after traumatic brain injury, J Trauma.2000:49(3):411-419.

43. Covassin T, Swanik CB, Sachs ML, Epidemiological considerationsof concussions among intercollegiate athletes. Áppl Neuropsychol.2003;10( 0:12-22.

44. Bazarian JJ. Wong T. Harris M, Leahey N. Mookerjee S. DombovyM, Epidemiology and predictors of post-concussive syndrome afterminor head injury in an emergency population. Brain Inj.I999;13(3):I73-189,

45. Hiscock M. Israelian M. Inch R, Jacek C, Hiscock-Kalil C. Is there asex difference in human laterality? II. An exhaustive survey of visualiaterality studies from six neuropsychology journals. J Clin ExpNeuropsyelwt. 1995:17|4):590-610.

46. Parsons TD. Larson P. Kratz K. et al. Sex diíTerences in mentalrotation and spatial rotation in a virtual environment. Neuropsyeho-¡ogio. 2()04;42(4):555-562.Broshek DK. Kaushik T. Freeman JR. Erlanger D, Webbc F. Barth47.

48.

JT. Sex differences in outcome following sporls-rclaled concussion.J Neurosurg. 2OO5;IO2(5):856-863.Collins MW. Iverson GL. Lovell MR, McKeag DB. Norwig J.Maroon J. On-tleld predictors of neuropsychological and symptomdeficit following sports-related concussion. Clin J Span Meit.2003:l3(4):222-229,

Cvnthia W. Majerske, MD. MS, contributed to conception and design: acquisition and analysis and interpretation of the data: and drafting.critical revision, and final approval of the article. Jason P. Miiialik. MS. CAT(C). ATC. contributed to conception and design: analysisand interpretation of the data: and drafting, critical revision, and final approval of the article. Dianxu Ren. PhD. contributed to analysisand interpretation of the data and drafting and final approval of the article. Michael W. Collins, PhD. contributed to conception and designand critical revision and final approval of the article. Cara Camiolo Reddy. MD. contributed to acqui.sition and analysis and interpretationof the data and critical revision and final approval of the article. Mark R. Lovell. PhD. contributed to conception and design, analysis andinterpretation of the data, and critical revision and final approval of the article. Amy K. Wagner. MD. contributed to conception and design:acipiisition ami analysis and interpretation of the data: am! drafting, critical revision, and final approval of the article.Address correspondence to Amy K. Wagner. MD. Department of Physical Medicine and Rehabilitation. University of Pitt.slmrgh MedicalCenter. 201 Kaufmann Building, 3471 5th Avenue, Pittsburgh, PA ¡5213. Address e-mail to wagneraki^^upnicedu.

274 Volume 43 • Number 3 • June 2008