Original article The effect of performance demands on lower extremity biomechanics during landing and cutting tasks

Boyi Di ,Willim E.Grrtt ,Mihl T.Gross ,Drin A.Pu ,Roin M.Qun ,Bing Yu ,*

a Division of Kinesiology and Health,University of Wyoming,Laramie,WY 82070,USA

b Department of Orthopaedic Surgery,Duke University,Durham,NC 27710,USA

c Division of Physical Therapy,University of North Carolina at Chapel Hill,Chapel Hill,NC 27514,USA

d Department of Exercise and Sport Science,University of North Carolina at Chapel Hill,Chapel Hill,NC 27514,USA

e Kevin Granata Biomechanics Lab,Department of Biomedical Engineering and Mechanics,Virginia Tech,Blacksburg,VA 24061,USA

Abstract Background:Anterior cruciate ligament(ACL)injuries commonly occur during the early phase of landing and cutting tasks that involve sudden decelerations.The purpose of this study was to investigate the effects of jump height and jump speed on lower extremity biomechanics during a stop-jump task and the effect of cutting speed on lower extremity biomechanics during a side-cutting task.Methods:Thirty-six recreational athletesperformed astop-jump task under 3 conditions:jumping fast,jumping for maximum height,and jumping for 60%of maximum height.Participants also performed a side-cutting task under 2 conditions:cutting at maximum speed and cutting at 60%of maximum speed.Three-dimensional kinematic and kinetic data were collected.Results:The jumping fast condition resulted in increased peak posterior ground reaction force(PPGRF),kneeextension moment at PPGRF,and knee joint stiffness and decreased knee f lexion angle compared with the jumping for maximum height condition.The jumping for 60%of maximum height condition resulted in decreased knee f lexion angle compared with the jumping for maximum height condition.Participants demonstrated greater PPGRF,knee extension moment at PPGRF,knee valgus angle and varus moment at PPGRF,knee joint stiffness,and knee f lexion angle during the cutting at maximum speed condition compared with the cutting at 60%maximum speed condition.Conclusion:Performing jump landing at an increased jump speed resulted in lower extremity movement patterns that have been previously associated with an increase in ACL loading.Cutting speed also affected lower extremity biomechanics.Jump speed and cutting speed need to be considered when designing ACL injury risk screening and injury prevention programs.2095-2546/?2019 Published by Elsevier B.V.on behalf of Shanghai University of Sport.This is an open access article under the CCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords:ACL injury;Injury prevention;Kinematics;Kinetics;Loading mechanism;Risk factor

1.Introduction

Anterior cruciate ligament(ACL)injuries are common sports-related knee injuries.The annual incidence rate of ACL injury isapproximately onein every 3000 citizens.1,2ACL injuries not only bring f inancial burden to the health service3but also cause devastating consequencesto patients'quality of life and can lead to secondary injuriesand disorders.4,5Understanding ACL injury mechanismsiscrucial for developing evidencebased injury prevention strategies.6,7Previously,investigators have shown that certain movement patterns such as decreased knee f lexion angle,increased impact ground reaction force(GRF),and increased internal kneeextension moment areassociated with increased ACL loading.7-11Kneevalgus/varusangle and valgus/varus moment may also load the ACL when an anterior shear force is applied to the proximal tibia at a small knee f lexion angle.9,12In addition,ACL injuriestypically occur during the early phaseof landing and cutting tasksthat involve sudden decelerations.13-17Furthermore,investigators recently quantif ied knee kinematics near the time of ACL injury based on tibiofemoral bone bruises and found that the knee was close to full extension near the time of injury.18Therefore,investigatorshaveassessed lower extremity biomechanicsassociated with ACL loading during jump landing,cutting,and a combination of landing and cutting tasks that simulate the maneuvers that are believed to cause ACL injuries.19-25

During competitive situations,athletic tasks may be performed with different performance demands.For example,a jump task may be performed for maximum jump height or speed for different competitive situations.26For the same reason,a cutting task may be performed with different cutting speeds.Although performancedemand isan important component in the completion of an athletic task,the effect of the performancedemand on thelower extremity biomechanicsthat have been previously associated with ACL injury remains largely unknown.Previously,investigators have focused on the effect of drop height on landing biomechanics and found that impact GRFgenerally increaseas thedrop height increases.27,28Although jump-landing biomechanics have been commonly assessed with maximum jump height as the performance demand,10,19-21,29-31it is unknown whether increasing jump height can alter landing mechanicsin away that would increase the risk of an ACL injury.Impact GRF may increase when individualsjump at an increased speed,32but theeffect of jump speed on knee kinematics and kinetics is unclear.In addition,individuals have been commonly assessed using a controlled speed during cutting tasks without clear understanding of how cutting speed may affect lower extremity biomechanics.23,33,34Screening and training athletes without knowing whether the task demand is associated with ACL loading may result in misinterpretation of screening results and mislead injury prevention programs.

As such,the purpose of the current study was to investigate the effects of jump height and jump speed on lower extremity biomechanics during a stop-jump task and the effect of cutting speed on lower extremity biomechanics during a side-cutting task.We hypothesized that increasing jump speed and increasingjump heightwhenperformingastop-jump task would result in increased peak posterior GRF(PPGRF),internal kneeextension moment,knee joint stiffness,knee valgus angle,knee varus moment,and decreased knee f lexion angle.We further hypothesized that increasing cutting speed when performing a side-cutting task would result in increased PPGRF,internal kneeextension moment,kneejoint stiffness,kneevalgusangle,knee varus moment,and decreased knee f lexion angle.

2.Methods

2.1.Participants

Based on previous studies on the effect of drop height and jump-landingtechniqueonlower extremity biomechanics,32,35-37amedium to largeeffectsizewasexpected for thecurrentstudy.Assuming an effect size of 0.5 for a pairwise comparison,a sample size of 34 was needed for a type I error of 0.05 and a power of 0.8.Eighteen male and 18 female recreational athletes(age:22.3±3.3 years;height:1.74±0.09 m;body mass:70.9±9.8 kg)who had experience in playing sports that involved landing and cutting tasks participated in the current study.Theexclusioncriteriaincluded(1)havingalower extremity injury that prevented participation in physical activity for more than 2 weeks over the previous 6 months;(2)having a history of an ACLinjury or other major lower extremity injuries;(3)possessing any condition that prevented maximal participation effort in sporting activities;or(4)pregnancy.38This study wasapproved by the University of North Carolinaat Chapel Hill Institutional Review Broad.Participants signed informed consent forms prior to participation.

2.2.Protocol

Participants performed a vertical stop-jump task for 3 experimental conditions:(1)jumping for maximum height,(2)jumping fast,and(3)jumpingfor 60%of maximumheight.The vertical stop-jump task consistsof an approach run followed by a 1-footed takeoff,a 2-footed landing on 2 force plates,and a 2-footed takeoff.24,30During the jumping for maximum height condition,participants were instructed to jump as high as possible following the 2-footed takeoff.During the jumping fast condition,participants were instructed to jump as fast as possible during the 2-footed landing while still trying to jump as high as possible following the 2-footed takeoff.During the jumping for 60%of maximum height condition,participants jumped for 60%of maximum height following the 2-footed takeoff.In our pilot study,we observed that 60%of maximum height gave ageneral representation of jumping with decreased jump height while participants still maintained a f luid jumping motion.For thiscondition,participants'maximum jump height wasf irstmeasured using aVertec(Sports Imports,Hilliard,OH,USA),and 60%of maximum height was calculated and corresponded to acertain height of the Vertec.Participantspracticed 60%of maximum height until they felt comfortable that they could consistently jump to thetargeted height.Participantsused a single-hand contact technique with the Vertec during both evaluation trials for maximum jump height and practice trials for 60%of maximum height.The Vertec was then removed to be consistent with other jumping conditions,and participants were instructed to maintain 60%of maximum jump height during thejumping for 60%of maximum height condition.The actual jump heightwasnotmonitored duringdatacollectionbut wascalculated during data processing based on marker coordinate data.

Participants also performed a side-cutting task with the dominant leg(self-reported preferred leg to jump for distance)for 2 experimental conditions:(1)cutting at maximum speed and(2)cutting at 60%of maximum speed.The side-cutting task consisted of an approach run followed by a 1-footed landing onaforceplateand alateral cutat45°fromtherunning direction.24,39During the cutting at maximum speed condition,participantswereinstructed to run asfast aspossibleand cut as fast as possible.During the cutting at 60%of maximum speed condition,participants cut at 60%of maximum running and cutting speed.For this condition,a regular timer was manually started and stopped by theinvestigator to quantify thetimefrom the start position to the end position when participantsran and cut as fast as possible.Participants then practiced to complete the task from the same start and end positions using 167%of the total time that was used during the cutting at maximum speed condition.Assuch,with thesamestart and end positions but 167%of the total time,participants were expected to achieve60%of maximum cutting speed.Participantspracticed 60%of maximum cutting speed until they felt comfortablethat they could consistently run and cut at the targeted speed.The actual speed was not monitored during data collection but was calculated during data processing based on marker coordinate data.

The order of stop-jump and cutting tasks and the order of different experimental conditions for each task were randomized.A minimum of 5 practice trials were performed before 5 off icial trials were collected for each experimental condition.Participants had a 3-min rest between experimental conditions and a 30-s rest between trials to reduce the effect of fatigue.

2.3.Data collection

Participants wore Spandex shorts and shirts as well as their own athletic shoes during data collection.Participants performed overground running and self-selected stretching for 5 min to warm-up.Retroref lectivemarkerswereattached bilaterally on participants'acromioclavicular joints,anterior superior iliac spines(ASIS),posterior superior iliac spines(PSIS),greater trochanters,lateral and medial femoral condyles,tibial tuberosity,lower shank,lateral and medial malleoli,heels,f irst and f ifth metatarsal heads,and f irst toes.38Three-dimensional coordinates of ref lective markers were collected using a data acquisition system with 8 Peak Motus video cameras(Peak Performance Technologies,Centennial,CO,USA)at a sampling rate of 120 Hz.20,22,24GRF data were collected using 2 Bertec 4060A force plates(Bertec Corporation,Columbus,OH,USA)at a sampling rate of 1200 Hz.19,20,25

2.4.Data reduction

The data during the landing phase were examined only for the dominant leg.The coordinate and GRF data were f iltered using a fourth-order zero-phase-shift low-pass Butterworth f ilter at a frequency of 10 Hz and 200 Hz,respectively.The cutoff frequency for thecoordinatedatawasestimated using an established method for best accuracy of calculating 2nd time derivatives.40The cutoff frequency for GRF data was obtained from power spectrum analyses of GRFs that demonstrated that most signals of GRFs are below 200 Hz when sampled at 1200 Hz.The use of these cutoff frequencies was consistent with previous studies.38,41

Methods to calculate jump height,approach speed,takeoff speed,and contact time were described in a previous study.38The center of the pelvis was def ined as the center of the left and right ASIS and the left and right PSIS.Jump height was determined by the difference between the maximum vertical coordinatesof thecenter of thepelvisduring jumping trialsand vertical coordinates of the center of pelvis during static trials.The instantaneous speed of the center of the pelvis at the moments of toe-touch and toe-off was calculated to determine approach and takeoff speed.Contact timewascalculated asthe total time from toe-touch to toe-off.

Procedures to def ine joint centers and segment reference frames and methods to calculate joint angles and resultant moments were consistent with previous studies.20,38The hip joint center wasdef ined asapoint in thepelvisreferenceframe and was located at 19%,30%,and 14%of the inter-ASIS distanceposterior,distal,and medial to theASIS,respectively.42The knee joint center was def ined as the midpoint between the lateral and medial femoral condyles.Theanklejoint center was def ined asthemidpointbetween thelateral and medial malleoli.The pelvis reference frame was def ined using bilateral ASIS and the middle point of bilateral PSIS.The thigh reference frame was def ined using the hip joint center,knee joint center,and lateral femoral condyle.The shank reference frame was def ined using the knee joint center,ankle joint center,and lateral femoral condyle.Cardan angles between thigh and shank reference frames were calculated in an order of f lexionextension,varus-valgus,and internal-external rotation.43Segment masses,center of mass locations,and segment moments of inertia were based on modif ied Clauser methods.44An inverse dynamics approach was used to calculate lower extremity joint resultant forces and resultant moments.45Joint resultant moments were transferred to the distal segment'sreferenceframeand expressed asinternal moments.Joint stiffness wascalculated aschangesinjointresultantmomentsdivided by changesin jointangles.Forceswerenormalized to body weight.Moments were normalized to the product of body weight and body height.Data calculations were performed in an MS3D70 computer program package(MotionSoft,Chapel Hill,NC,USA).

For the stop-jump task,performance variables included jump height and contact time.For the side-cutting task,performance variables included approach speed,takeoff speed,and contact time.PPGRFduring the early landing phase is considered a critical time point for ACL loading.10,20In addition to PPGRF,knee f lexion angle and knee extension moment are important when assessing ACL injury risk.6,8-11,18Knee varus/valgus angle and knee varus/valgus moment may contribute to ACL loading.9,12Therefore,for both stop-jump and side-cutting tasks,kinematic and kinetic variables associated with ACL injury risk included knee f lexion angle at initial contact,PPGRF,knee f lexion angle at PPGRF,knee extension moment at PPGRF,knee varus/valgus angle at PPGRF,knee varus/valgusmomentat PPGRF,peak kneef lexion angle,kneef lexion rangeof motion from initial contactto peak f lexion,and sagittal plane knee joint stiffness from initial contact to peak f lexion.

2.5.Statistical analysis

Performance,kinematic,and kinetic variables were compared among 3 stop-jump conditionsusing analysisof variance with repeated measures.Only signif icant analyses of variance were followed by paired t tests.Performance,kinematic,and kinetic variables were compared between the 2 side-cutting conditionsusing paired t tests.An outlier wasdef ined asavalue that deviated from the mean by more than 3 times the standard deviation and signif icantly affected the signif icance level of a statistical test.A type Ierror ratelessthan or equal to 0.05 was chosen as indication of statistical signif icance.The Holm stepdown procedurewasused to adjust thetype Ierror rate of each paired t test to keep the overall type I error rate no greater than 0.05.46Statistical analyses were performed in SPSS 16.0(SPSSInc.,Chicago,IL,USA).

Table1Performanceoutcomesand kinematic and kinetic variables(mean±SD)duringthe3stop-jump conditions(jumpingfast,jumping for maximum height,and jumping for 60%of maximum jump height).

For thestop-jump task,p valuesof analysisof variancewere lessthan 0.001 for jump height,contacttime,kneef lexion angle at initial contact,PPGRF,knee f lexion angle at PPGRF,knee extension moment at PPGRF,peak knee f lexion angle,knee f lexion range of motion,and knee joint stiffness,but not for knee varus/valgus angle at PPGRF(p=0.294)or knee varus/valgus moment at PPGRF(p=0.470).Paired t tests were performed between each pair of jumping conditions for the 9 variablesthat showed signif icant analysisof variance(Table1).For the side-cutting tasks,the knee joint stiffness data of 1 participant were identif ied as outliers and not included in the analysis.Paired t tests were performed between 2 cutting conditions for all 12 variables(Table 2).A total of 39 paired t tests were performed for both the stop-jump and side-cutting tasks(Tables 1 and 2).The largest p value for a signif icant paired t test was 0.006 after the adjustment for the overall type Ierror rate(Table 1).

For thestop-jump task(Table1),jump height wasthegreatest during the jumping for maximum height condition,the second greatest during thejumping fast condition,and theleast during thejumping for 60%of maximum height condition.The actual jump height during the jumping for 60%of maximum height condition was 65.6%±13.1%(mean±SD)of the jump height during the jumping for maximum height condition.Contact time was signif icantly shorter during the jumping fast condition compared with theother 2 jumping conditions.Knee f lexion angle at initial contact was signif icantly smaller during the jumping for 60%of maximum height condition compared with the other 2 conditions.PPGRF,knee extension moment at PPGRF,and knee joint stiffness were signif icantly greater during the jumping fast condition compared with the other 2 jumping conditions.Knee f lexion angle at PPGRF was signif icantly greater during the jumping for maximum height condi-tion compared with the other 2 conditions.Peak knee f lexion anglewasthegreatest during thejumping for maximum height condition,the second greatest during the jumping for 60%of maximum height condition,and the least during the jumping fast condition.Knee f lexion range of motion was signif icantly smaller during the jumping fast condition compared with the other 2 jumping conditions.

Table2Performance outcomes and kinematic and kinetic variables(mean±SD)during the 2 side-cutting conditions(cutting at maximum speed and cutting at 60%of maximum speed).

3.Results

For theside-cutting task(Table2),participantsdemonstrated signif icantly greater approach and takeoff speeds and shorter contact time during the cutting at maximum speed condition compared with the cutting at 60%of maximum speed condition.Theactual approach and takeoff speedsduring thecutting at 60%of maximum speed condition were 54.9%±8.5%and 55.5%±8.2%of those during the cutting at maximum speed condition,respectively.Participants demonstrated signif icantly greater knee f lexion angle at initial contact,PPGRF,knee f lexion angle at PPGRF,knee extension moment at PPGRF,knee valgus angle at PPGRF,knee varus moment at PPGRF,peak kneef lexion angle,and kneejoint stiffness,and decreased knee f lexion range of motion during the cutting at maximum speed condition compared with thecutting at60%of maximum speed condition.

4.Discussion

The purpose of the current study was to investigate the effects of jump height and jump speed on lower extremity biomechanicsduring a stop-jump task and the effect of cutting speed on lower extremity biomechanics during a side-cutting task.The results of performance outcomes support that participants achieved different performance demands during different jumping and cutting conditions.The performance outcomes during the stop-jump task were consistent with a recent study,26which has shown that jump height and jump speed are2 differenttask demands,and itisunlikely to jump for maximum height and highest speed at the same time.The f indings of kinematic and kinetic variables partially support our hypothesis.

The f indings of this study support the hypothesis that performing stop jump at an increased speed would result in increased PPGRF,internal knee extension moment,and knee joint stiffnessand decreased knee f lexion angle.Different from drop-landing and drop-vertical jump tasks,19,28,35the stop-jump task begins with an approach run and involves sudden decelerations in the anterior-posterior direction during landing.To achieve the goal of jumping as fast as possible,participants landed stiffer,as indicated by the decreased knee f lexion angle at PPGRF,peak knee f lexion angle,and knee f lexion range of motion compared with the jumping for maximum height condition.Thisstiff landing pattern ensured that participantscould absorb the approach momentum in a short time and reduce the total contact time.However,PPGRF,kneeextension moment at PPGRF,and knee joint stiffness increased as compensation for the decreased contact time.The f indings of increased PPGRF are consistent with the study by Walsh et al.,32who found greater impact GRFwhen participantslanded and jumped with a shorter contact time.Meanwhile,investigators have previously shown decreased impact GRFwhen participants utilized a soft landing pattern that is characterized by increased knee f lexion angles and contact time,indicating a decreased jump speed.38,47Thef indingsof previousstudiesand thecurrentstudy suggest that jump speed is a sensitive factor associated with lower extremity biomechanics during jump landing.On the other hand,jump speed did not result in signif icant differences in knee valgus angle and knee varus moment at PPGRF,which could be associated with the predominance of sagittal plane motion during the stop-jump task.Performing jump landing at anincreased speed may imposeatask demand that isassociated with increased sagittal plane loading of the ACL.

Thef indingsof thisstudy do not support thehypothesisthat jumping for a greater height would result in increased PPGRF,internal knee extension moment,and knee joint stiffness and decreased knee f lexion angle.Actually,jumping for maximum heightresulted in increased kneef lexion angleat initial contact,knee f lexion angle at PPGRF,and peak knee f lexion angle but similar PPGRF,knee extension moment at PPGRF,knee valgus angle at PPGRF,knee varus moment at PPGRF,knee f lexion rangeof motion,and kneejoint stiffnesscompared with jumping for 60%of maximum height.Previously investigators have focused on the effect of drop height on lower extremity biomechanics during drop-vertical jump tasks.27,28Jumplanding task have been usually completed with participants jumping for maximum jump height.10,19-21,29,30The results of the current study,however,suggest that jumping for 60%of maximum height may represent a scenario that is associated with increased ACL loading compared with jumping for maximum height,becauseof thedecreased kneef lexion angle.8The average knee f lexion angle at PPGRF during the jumping for 60%of maximum height condition wasbelow 30°,which is considered a critical knee f lexion angle associated with greater ACL loading.48,49A low knee f lexion angle may amplify ACL loading in combination with other loading mechanismssuch as anterior shear force and knee valgus/varus moment.9,12From a mechanical perspective,adecreased jump height indicated that less kinetic energy needed to be generated during the takeoff phase of landing.During the jumping for maximum height condition,it was postulated that participants utilized a selfoptimized joint range of motion for maximizing force production during the takeoff.On the other hand,during the jumping for 60%of maximum height condition,participants utilized a decreased knee f lexion angle strategy during landing,which corresponded to a joint range of motion associated with less forceproduction during thetakeoff.Thef indingsof thecurrent study do not support thebelief that increasing jump height may changelanding mechanicsin away that would increasetherisk of an ACL injury.Assessing jump-landing mechanics with maximum jump height as the performance demand,therefore,may not represent ACL injury scenarios.

The f indings of this study support the hypothesis that performing side-cutting at an increased speed would result in increased PPGRF,knee extension moment at PPGRF,and knee joint stiffness and decreased knee f lexion range of motion.An increased speed also resulted in an increase in the knee valgus angle at PPGRF and the knee varus moment at PPGRF.The averagekneef lexion anglesat PPGRFduring both side-cutting conditions were less than 30°.However,performing sidecutting at an increased speed also resulted in increased knee f lexion angle at initial contact,knee f lexion angle at PPGRF,and peak knee f lexion angle.Previous investigators have studied the effects of performance demands such as reaction time and fatigue on lower extremity biomechanics during cutting.50,51Participants have been commonly tested with cutting speed as a control variable.23,33,34In the current study,similar to the stop-jump task,participants utilized a movement pattern with decreased knee f lexion angle during the cutting with 60%of maximum speed condition.This decrease may be associated with a decreased task demand to produce force and generate kinetic energy.During the cutting with maximum speed condition,participants started with greater knee f lexion at initial contact but reduced knee f lexion range of motion to reduce the total contact time,and the task demand of great force production within a short contact time resulted in greater PPGRF,knee extension moment at PPGRF,and knee joint stiffness.The f indings suggest that cutting speed could signif icantly modify lower extremity biomechanics and support the notion of controlling cutting speed when assessing cutting mechanics.In addition,performing cutting task at different speedsmay poseloadsto the ACL from differencemechanisms.

There were several limitations of the current study.We evaluated only jump height and jump speed as performance demands for the stop-jump tasks,and cutting speed was the only performance demand for the side-cutting task.Other task demands,such astheanticipated vs.unanticipated natureof the task,jumping/cutting directions,and fatigue,were not evaluated and could interact with jump height,jump speed,and cutting speed to alter the movement patterns in the lower extremity.Participants practiced the 60%of maximum jump height and 60%of maximum cutting speed conditions before data were collected.Participants'actual jump height and cutting speed during these 2 conditions were calculated using marker coordinates during data processing but were not monitored during datacollection owing to softwarelimitations.Differences were observed between the targeted and actual jump height and cutting speed during these 2 conditions.These differences could be caused by different measurement methods(Vertec and regular timer vs.markers)between data collection and data reduction.Participants'variation in maintaining targeted jump height and cutting speed may also contribute to these differences.The purpose of the current study was to compare lower extremity biomechanics between conditions with maximum performancedemandsand relatively lower performance demands.In addition,the differences between the targeted(60%)and actual jump height(66%)and cutting speed(55%)were only 5%-6%.As such,this discrepancy in jump height and cutting speed may affect the exact magnitudes of dependentvariablesbut should notaffect thegeneral changesin dependent variables and the conclusion of the current study.In addition,only 1 decreased jump height condition and 1 decreased cutting speed condition were studied.Real-time monitoring of jump height and cutting speed may improve the consistency in achieving the targeted jump height and cutting speed and allow evaluation of the effect of small incremental changes in jump height and cutting speed on lower extremity biomechanics.Weassessed only lower extremity biomechanics that have been previously shown to be associated with ACL injury.Estimated ACL loading becomes inconclusive when loading variablessuch askneef lexion angleand kneeextension moment change in different directions.Future research to directly measure ACL length or strain could provide a better understanding of changes in ACL loading as a function of different performance demands.

5.Conclusion

Performing jump landing atincreased jump speed resulted in lower extremity movement patterns that have been previously associated with an increase in ACL loading.Cutting speed also affected lower extremity biomechanics.Jump speed and cutting speed need to beconsidered when designing injury risk screening and injury prevention programs.More dynamic tasks with decreased contact time could be used in the development of injury prevention programsduring thef inal stagesof training as well asbeing incorporated into thef inal stagesof rehabilitation as athletes are returned to sport after an injury to insure that they are ready to meet the demands of athletic competition.

Authors'contributions

BD and BY contributed to the design of the study,data collection,data reduction,and data analysis and drafted the manuscript;WEG,MTG,DAP,and RMQ contributed to the design of the study and data analysis and helped to draft the manuscript.All authors have read and approved the f inal version of the manuscript,and agree with the order of presentation of the authors.

Competinginterests

The authors declare that they have no competing interests.