Neuromusculoskeletal Models Provide Insights into the Mechanisms ...
7 Pages
English

Neuromusculoskeletal Models Provide Insights into the Mechanisms ...

Downloading requires you to have access to the YouScribe library
Learn all about the services we offer

Description

www.acsm-essr.org ..... Sports 37: 1931Y1938, 2005. 14. Thelen, D.G., E.S. Chumanov, D.M. Hoerth, T.M. Best, S.C. Swanson,. L. Li, M. Young, and B.C. ...

Subjects

Informations

Published by
Reads 43
Language English

ARTICLE

NeuromusculoskeletalModelsProvideInsights
intotheMechanismsandRehabilitationof
HamstringStrains
DarrylG.Thelen,
1
ElizabethS.Chumanov,
1
MarcA.Sherry,
2
andBryanC.Heiderscheit
1
1
UniversityofWisconsin-Madison,Madison,WI;and
2
UniversityofWisconsinHealthSportsMedicineCenter,
Madison,WI

THELEN,D.G.,E.S.CHUMANOV,M.A.SHERRY,andB.C.HEIDERSCHEIT.Neuromusculoskeletalmodelsprovide
insightsintothemechanismsandrehabilitationofhamstringstrains.
Exerc.SportSci.Rev.,
Vol.34,No.3,pp.135
Y
141,2006.
Neuromusculoskeletalmodelsareusedtoinvestigatehamstringmechanicsduringsprinting.Weshowthatpeakhamstringstretchoccurs
duringlateswingphaseandisinvariantwithspeed,butdoesdependontendoncomplianceandtheactionofothermusclesinthe
lumbopelvicregion.Theinsightsgainedarerelevantforimprovingthescientificbasisofhamstringstraininjurypreventionand
rehabilitationprograms.
KeyWords:
muscleinjury,tendoncompliance,stretchshorteningcycle,forwarddynamics,computer
simulation,biomechanics,motionanalysis,sprinting

INTRODUCTION
Acutehamstringstrainsareacommoninjuryinsports
involvingsprinting.Straininjuriesarecharacterizedby
observabledisruptionofthemusculotendonjunction(7),
withpostinjuryremodelinginvolvingbothscartissue
formationandmuscleregeneration(6).Theinjurycan
causeanathletetomissafewdaystoafewweeksofsport.
Moreproblematicisthehighrecurrencerate,withapprox-
imatelyoneofthreeathletesreinjuringwithinayearof
returningtosport(9).Theseobservationshighlightthe
prevalenceofhamstringstraininjuriesandthechallengein
preventingtheinitialinjuryandsubsequentreinjury.
Theresidualeffectsofapriorhamstringstrainmaybe
identifiable,andtheassociatedriskforreinjuryreducibleusing
newapproaches.Forexample,Proske
etal.
(10)identifieda
shiftintheisokinetickneestrengthprofileinpreviously
injuredlimbs,aneffectthatmaybeamendablevialength-
eningcontractiontraining.SherryandBest(11)haveshown
thatarehabilitationprogramfocusedonearlymovementand

Addressforcorrespondence:
DarrylG.Thelen,Ph.D.,DepartmentofMechanical
Engineering,UniversityofWisconsin-Madison,1513UniversityAve.,Madison,WI
53706(E-mail:thelen@engr.wisc.edu).
Acceptedforpublication:
April20,2006.
AssociateEditor:ChrisMcGibbon,Ph.D.
0091-6331/3403/135
Y
141
ExerciseandSportSciencesReviews
Copyright
*
2006bytheAmericanCollegeofSportsMedicine

oCypirhg t @0260b yht emArecinaC loelego fpSrostM dei531

icen .neuromuscularcontroldramaticallyreducedhamstringreinjury
ratescomparedwithatraditionalstretchingandstrengthening
approach.Althoughthesearepromisingobservations,the
underlyingmechanismsarenotcompletelyunderstood.
Wehaveusedaneuromusculoskeletalmodelofsprinting
toanalyzepotentialhamstringinjurymechanisms.The
modeldescribestherelationshipamongmuscleexcitations,
activationdynamics,musculotendoncontractionmecha-
nics,andsegmentalaccelerations(13,15).Thus,themodel
hasallowedustorelatemechanicsatthemusclelevelto
themovementproducedatthewholebodylevel.Inthis
article,wereviewtheuseofaneuromusculoskeletalmodel
toinvestigatetheeffectsofsprintingspeed,musculotendon
properties,andcoordinationonhamstringmechanics
duringsprinting.Theinsightsgainedarerelevantfor
improvingthescientificbasisofinjurypreventionand
rehabilitationprograms.

HAMSTRINGINJURYREHABILITATION

Hamstringstraininjuriesmostcommonlyoccurinthe
bicepsfemorislongheadandexhibitastrongtendencyto
recur.Forexample,imaginganalysisof170recentlyinjured
athletesfoundthatapproximately80%ofhamstringstrain
injuriesinvolvedthebicepsfemoris(7).Furthermore,a
reviewof858hamstringstrainsinAustralianfootballers
showedtherateofrecurrencewas12.6%duringthefirstweek
ofreturntosportand8.1%forthesecondweek.Thecumulative

nUuahtrozider peorudtcoi nfot ih sraitlc esip orihibet.d
riskforreinjuryforthe22-wkseasonwas30.6%(9).Thehigh
reinjuryratemaybeduetotheuseofinappropriatecriteriafor
determiningsuitabilityforreturntosportor,alternatively,that
traditionalrehabilitationmethodsareinsufficientforreducing
riskforreinjury.
Rehabilitationprotocolsforacutemusclestrainshave
traditionallyemphasizedhamstringstretchingandstrength-
eningexercises.SherryandBest(11)prospectivelycom-
paredsuchanapproachwithaprogressiveagilityandtrunk
stabilization(PATS)rehabilitationprogram.PATS
includedexercisesthatemphasizedearlymovementand
coordinationofthepelvisandtrunkmuscles(Fig.1).Inthe
first2wkafterreturningtosport,noneofthe13athletes
participatinginthePATSprogramexperiencedareinjury,
comparedwith6(54.5%)of11athletesthatperformed
isolatedhamstringstretchingandstrengtheningexercises.A
significantreductioninreinjuryratewasstillevidentin
PATSparticipantsevenafter1yrofreturningtosport.
However,thestudywasunabletorelatereinjuryriskto
commonclinicalmeasuressuchasstrength,flexibility,
speed,andverticaljumpheight.
AlthoughSherryandBest(11)haveshownapromising
clinicaloutcome,itremainsunclearwhichneuromuscular
factorsareresponsibleforthereducedreinjuryriskinthe
PATSgroup.Onehypothesisisthatimprovedneuro-
muscularcontrolofthelumbopelvicregionallowsthe
hamstringstofunctionatsafelengthsandloadsduring
athleticmovement,therebyreducinginjuryrisk(11).An
alternativeexplanationisthattheuseofearlysubmaximal
loadinglimitstheresidualadverseeffectsofscartissue
formedearlyintheremodelingprocess.
RecentobservationsbyProske
etal.
(10)suggestresidual
effectsmayindeedbepresentinapreviouslyinjuredmuscle.

j1Theystudiedisokinetic(60
-I
s)kneeflexionexercises
performedbynineathleteswhohadexperiencedmultiple
unilateralhamstringstrainswithinthepast5yrs.Atthe
timeoftesting,allathleteswereinjury-freeforatleast1
monthandparticipatinginsport.Interestingly,therewere
nostrengthdeficitsinthepreviouslyinjuredlimbs.How-
ever,peakisokinetictorquewasgeneratedatanaverage12
-
greaterkneeflexionangleinthepreviouslyinjuredlimb
comparedwiththeuninjuredlimb.Theauthorsattributed
thedifferencetoashorteroptimalmusculotendonlengthfor
activetensioninthepreviouslyinjuredmuscle.Suchashift
canbeatrainingeffect,forexample,repeatedperformance
ofconcentricstrengtheningexercisesinrehabilitation.
Alternatively,theshiftcanreflectthepresenceofresidual
scartissueatthemusculotendonjunction(6).Scartissueis
stifferthanthecontractiletissueitreplacesand,thus,may
alterthemechanicalenvironmentseenbythemusclefibers.
Specifically,adecreaseinseriescompliancewouldshift
peakforcedevelopmenttoshortermusculotendonlengths
asobserved.Proske
etal.
(10)havealsoshown,atleastin
healthycontrolsubjects,thattheperformanceofcontrolled
lengtheningcontractionexercisescanfacilitateashiftin
peakforcedevelopmenttolongermusculotendonlengths.
Theirinitialdatasuggestthattheincorporationofsuch
exercisesintotrainingmayreducehamstringinjuryrates.
Althoughtherearepromisingnewapproachesfor
improvingthepreventionandrehabilitationofhamstring
strains,questionsremainregardingtheunderlyingmecha-
nisms.Fundamentally,astraininjuryistheresultof
exceedingthelocalmechanicallimitsofthemuscletissue.
Thus,itisrelevanttoinvestigatethemechanicalbehavior
ofthehamstringsduringpotentiallyinjurioustasks.Inthis
article,wereviewtheuseofaneuromusculoskeletalmodel

Figure1.
SherryandBest(11)comparedtheeffectivenessoftworehabilitationprogramsinreducingreinjuryratesinathleteswhosustainedanacute
hamstringstrain.ASSgroup(
N
=11)performedstaticstretching,isolatedprogressivehamstringstrengthening,andgraduatedreturntoactivity.APATS
group(
N
=13)performedagilityexercisesbeginningwithmovementsprimarilyinthefrontalandtransverseplanes,thenprogressingtomovementsin
thesagittalplane.Exercisesrequiringmuscleactivitytomaintainthespineandpelvisinadesiredposture(bridges)werealsoperformed.Foreach
rehabilitationprogram,athletesprogressedfromphase1tophase2whentheycouldwalkwithanormalgaitpatternanddoahigh-kneemarchinplace
withoutpain.ComparedwiththeSSgroup,therewasastatisticallysignificantreductionininjuryrecurrenceinthePATSgroupat2wkandat1yrafter
returntosport.PATSindicatesprogressiveagilityandtrunkstabilization;SS,stretchingandstrengthening.

136
ExerciseandSportSciencesReviews

oCypirhg t @0260b yht emArecinaC loelego fpSrostM decini.eU antuohirez derrpdocuitnoo fhtsia trcieli srpwww.acsm-essr.org

hobitide.
toestimatethestretch,loading,andworkdonebymuscles
duringsprinting.Theseanalysesareusedtoaddressthe
followingspecificquestions:Whenduringthesprintinggait
cyclearethehamstringssusceptibletoinjury?Howdoes
speedaffecthamstringmechanicsandpotentiallycontribute
toinjuryrisk?Howcanchangesinmusculotendonproper-
tiesaffectinjurypotential?Canthecoordinationof
individualtrunkmusclesinfluencehamstringmechanics?
HAMSTRINGINJURYMECHANISMS
Despitethefrequencyofhamstringstrainsduringsprint-
ing,itremainsuncertainwheninthegaitcyclethemuscle
isinjured.Ithasbeensuggestedthatinjuriesmayoccur
duringlateswing,whenthehipisflexedandthekneeis
extended.However,othershavespeculatedthatthe
potentiallylargeloadsassociatedwithgroundcontactmay
causeinjury.Animalmodelsclearlyshowthatmuscleinjury
isassociatedwithexcessivefiberstretchduringalength-
eningcontraction(8).Thus,wehaveusedamodelof
sprintingtoquantifythestretchofthebiarticularham-
stringsthroughoutthegaitcycle(14).Thisinformation,
coupledwithEMGmeasurements,wasusedtoassesswhen
thehamstringswereundergoinganactivelengthening
contractionandwhythebicepsfemorislongheadmaybe
moresusceptibletoinjury.
Agenericmodelofthepelvisandlowerextremitywas
firstcreated,whichincludedadescriptionofthemuscu-
lotendonpathsofthesemimembranosus,semitendinosus,
andbicepsfemorislongheadmuscles(Fig.2A).Themodel
incorporatedimportantdifferencesbetweentheindividual
hamstringmomentarmsatthehipandknee(14).Whole-
bodykinematicswerecollectedwhilesubjects(
N
=14)

sprintedonahigh-speedtreadmillatspeedsrangingfrom
80%to100%ofmaximum.Themodelwasscaledtoeach
subjectusingsubject-specificmeasuresofsegmentlengths.
Jointangleswerecomputedtofitthescaledmodeltothe
measuredmarkerkinematics.Thisprocessprovidedquanti-
tativeestimatesofmusculotendonstretchforeachofthe
hamstringmusclesthroughoutthegaitcycle(Fig.2B).
Wefoundthatpeakhamstringstretchoccursduringthe
lateswingphaseofsprintingbeforefootcontact(14).
Electromyographydataindicatethatthehamstringsare
activeatthissamephaseofthegaitcycle(13).Thus,the
hamstringsareundergoinganactivelengtheningcontrac-
tionduringlateswing,creatingthepotentialconditionsfor
astraininjurytooccur.Themagnitudeofpeakstretchwas
significantlygreaterforthebicepsfemorislonghead
(stretched9.5%beyondlengthinanuprightposture)than
thesemimembranosus(7.4%)andsemitendinosus(8.1%)
muscles.Differencesinthekneeflexionmomentarms
betweenthemedialandlateralhamstringsaccountedfor
theintermusclevariationsinpeakmusculotendonstretch.
Specifically,becausethekneeisslightlyflexedduringlate
swing,thesmallerkneeflexionmomentarmofthelateral
hamstrings(bicepsfemoris)resultsinthismuscleunder-
goinggreaterstretchrelativetoupright.Weconcludedthat
intermuscledifferencesinhamstringmusclegeometrymight
beacontributingfactortothegreaterpropensityforbiceps
femorisinjury.
Althoughprovidinginterestinginformation,itisdifficult
todirectlyassesswhenaninjuryoccursbasedonthe
kinematicanalysisofinjury-freerunningtrials.Through
unexpectedcircumstances,werecentlycompletedananal-
ysisofwhole-bodykinematicsobtainedatthetimeofan
acutehamstringstraininjury(5).Amaleathleteinjuredhis
rightbicepsfemoriswhilerunningat5.36m
I
s
j
1
onan

Figure2.
A.Subject-specificmodelswereobtainedbyscalingsegmentlengthsofagenericmusculoskeletalmodelbasedonanatomicalmarker
positionsrecordedofasubjectinastandingposture.Aninversekinematicsroutinewasthenusedtodeterminethesegmentjointanglesthatbestalign
thescaledmodelwiththemeasuredmarkerkinematicsofthesubjectsprintingonatreadmill.B.Shownaretheensemble-averaged(
N
=14subjects)
estimatesofmusculotendonstretch,normalizedtouprightmusculotendonlengths,foreachofthebiarticularhamstringsthroughoutthesprintinggait
cycle.Peakstretchisreachedduringlateswingphasewhenbothfeetwereofftheground.Thelateralhamstrings(bicepsfemoris)arestretched
significantlymorethanthemedialhamstringsbecauseofdifferencesinthekneeflexionmomentarmsbetweenmuscles(14).

Volume34
c
Number3
c
July2006

oCypirhg t @0260b yht emArecinaC loelego fpSrostM decini.eU antuohirez derrpdocuitMechanismsofHamstringStrains
137

noo fhtsia trcieli srphobitide.
inclined(15%)treadmill.Weusedstatisticaltechniquesto
identifywhenindividualmarkertrajectoriesdeviatedfroma
periodicpattern,indicatingtheearliestmechanicalresponse
toinjury.Thisinformation,combinedwithestimatesof
neuromuscularlatenciesandelectromechanicaldelay,was
usedtoidentifya130-msportionofthelateswingphaseof
thegaitcyclewhentheinjurymostlikelyoccurred.
Maximumstretchofthebicepsfemorisandpeakhip
extensionandkneeflexionmomentsalsooccurredwithin
thesuspectedperiodofinjury.Weconcludedthatthe
bicepsfemoriswaslikelyinjuredasaresultofalength-
eningcontractionduringthelateswingphaseofthe
runninggaitcycle.

SPEEDEFFECTSONHAMSTRINGMECHANICS
Hamstringinjuriesarecommonlyassociatedwithhigh-
speedsprinting.Therefore,anassessmentofhowhamstring
mechanicsvarywithspeedisrelevanttounderstanding
injurymechanisms.Wehaveusedaneuromusculoskeletal
modeltoinvestigatetheinfluenceofspeedonthestretch,
loading,andworkdonebythehamstringsduringthelate
swingphaseofsprinting(13).Musculotendonmechanics
wererepresentedbyaHill-typemodelinwhichacontrac-
tileelementisinparallelandserieswithelasticelements.
Theserieselasticelementwasassumedtorepresentthe
complianceofthetendonandaponeurosis.Theinputto
eachmusclewasanidealizedexcitationlevelthatcanvary
between0and1(fullexcitation).Theforcesdevelopedby
themusculotendonactuatorsactedontheskeletonviathe

linked-segmentequationsofmotion.Therefore,theneuro-
musculoskeletalmodelrelatedthemuscleexcitations,
muscleforces,andthesegmentalaccelerationsgenerated
inthesystem.
Acomputedmusclecontrolalgorithm(12)wasusedto
determinetheexcitationpatternsof26lowerextremity
musclesthatdrovethelowerextremitytotrackmeasured
hipandkneemotionofsprintingathletes(Fig.3A).
Validationofthemodelwasassessedbycomparing
estimatedandmeasuredjointanglesandmuscleactivation
patterns.Subject-specificjointanglesweretrackedclosely,
withlessthan2
-
ofaverageerrorinhipandkneeangles
(13).Thetimingofcomputedmuscleexcitationsclosely
matchedmeasuredEMGactivities(Fig.3B).
Thesimulationsshowthatthehamstringsundergoa
stretchshorteningcycleduringthelatterhalfofswing
phase.Bicepsfemorismuscleexcitationsincreaserapidly
between70%and80%ofthegaitcycle,continuingthrough
theendofswingphase.Maximalhamstringloadingis
reachedslightlybeforepeakmusculotendonstretchofthe
musculotendonunit(Fig.3B).Asaresult,thehamstrings
aredoingasubstantialamountofnegativework(integralof
negativemusculotendonpower)between70%and90%of
thesprintinggaitcycle.
Peakhamstringmusculotendonstretchwasfoundtobe
invariantacrosstherangeofspeeds(80
Y
100%ofmax-
imum)considered(Fig.4).However,thenegativemuscu-
lotendonworkdonebythehamstringsincreasedconsiderably
withspeed.Thehamstringslikelyfunctiontoabsorbandredis-
tributethekineticenergyoftheswinglimbbeforefootcontact.
Becausekineticenergyincreasesinproportiontothevelocity

Figure3.
A.Acomputedmusclecontrolalgorithmwasusedtodeterminemuscleexcitationsthat,wheninputtedtotheneuromusculoskeletalmodel,
producedsimulatedkinematicsthatcloselyreplicatedmeasuredkinematics.Ateachtimestepofthesimulation,thecontrollerdeterminedmuscle
excitationsthatwouldproducethedesiredaccelerationswhileminimizingacostfunctiontoresolvemuscleredundancy(12).Withthemuscleexcitations
asinput,thesetofactivation,musculotendon,andmusculoskeletaldynamicequationswereintegratedforwardtosimulatethekinematicsthatthe
muscleactionsproduce.B.TimingofthesimulatedbicepsfemorisexcitationsareconsistentwithmeasuredEMGpatterns(mean
T
1SDoffivesubjects).
Thesimulationprovidesestimatesofforce,stretch,andpowerdevelopmentofthemuscle,tendon,andmusculotendoncomponents(13).

138
ExerciseandSportSciencesReviews

oCypirhg t @0260b yht emArecinaC loelego fpSrostM decini.eU antuohirez derrpdocuitnoo fhtsia trcieli srpwww.acsm-essr.org

hobitide.
squared,thenegativeworkdonebythehamstringsincreasesata
ratethatexceedsthepercentagechangeinspeed.
Theseobservationsdemonstratethatmaximummuscle
stretchmaybereachedatlowersprintingspeedsthanthe
maximumenergyabsorptioncapabilitiesofthemuscle.
Stretchandnegativeworkrequirementsmaycouple
togetherathighspeedstocontributetoinjuryrisk.For
example,thecumulativenegativeworkdoneoverrepeated
maximalstretch-shorteningcontractionsmaypredisposea
muscletoinjury(2).Alternatively,injurycanresultfrom
stride-to-stridevariabilityinthepeakstretchimposed,with
anexcessivestretchinasinglestrideleadingtotheonsetof
injury.Itshouldbenotedthatouranalysesofmuscle
loadingandworkwerelimitedtotheswingphaseof
sprinting,whenthehamstringsareundergoinganactive
lengtheningcontractionandseemsusceptibletoinjury.
However,similaranalysisofstancephaseiswarrantedto
characterizethenetloadingandworkdonebytheham-
stringsduringsprinting.

EFFECTOFTENDONCOMPLIANCEON
HAMSTRINGMECHANICS
Itisimportanttorecognizethatthemotionofthe
musculotendonunitisnotnecessarilyrepresentativeofthe
behaviorofindividualmusclefiberswheretheinjuryoccurs.
Forexample,Griffiths(4)showedthatastretchimposedon
acontractingmusculotendonisoftentakenupbythe
tendon,allowingthemusclefiberstomaintainanisometric
lengthorevencontinuetoshorten.Griffithsproposedthat
tendoncomplianceactsasamechanicalbufferthatreduces
thestretchofmusclefibersandprotectsagainstinjury.
Therefore,changesinmusculotendonpropertiescanaffect
fiberstretchand,hence,contributetoinjuryrisk.For
example,residualpostinjuryscartissueatthemusculoten-
donjunctioncanalterthemechanicalenvironmentofthe
musclefibers.Also,ithasbeenshowninananimalmodel
thatseriescomplianceseemstodecreasewithrepeated
maximalstretch
Y
shorteningcontractions(2).Atime-

Volume34
c
Number3
c
July2006

oCypirhg t @0260b yht emArecinaC loelego fpSrostM decini.e Figure4.
A.Shownisthelowerextremitypostureat
thetimeofpeakhamstringmusculotendonstretch.
Slightlygreaterhipandkneeflexionoccursatamaximal
sprintingspeedcomparedwithasubmaximalspeed.B.
However,theadditionalhamstringstretchduetohip
flexionisoffsetbyshorteningduetokneeflexion,
resultinginapeakstretchthatisinvariantwithspeed.In
contrasttothekinematicquantities,thenegativemuscu-
lotendonworkincreasessubstantiallywithsprintingspeed.

dependentdecreaseintendoncompliancewouldconceiv-
ablyalterthestretchandtheloadingofthefibersduring
repetitiveathleticmovementsuchassprinting.
Weusedasensitivitystudytoinvestigatehowthestretch
andnegativeworkseenatthefiberleveldependsonthe
complianceofthein-seriestendoncomponentduring
sprinting(13).Anominalmuscle-actuatedsimulationof
theswingphaseofsprintingwasfirstcreated(Fig.3A).We
thenvariedthetendoncomplianceofthebicepsfemoris
longheadandreconductedthesimulationwithanaltered
muscleexcitationpatternthatretainedtheoverallmuscu-
lotendonbehavior,albeitwithalteredmusculotendon
mechanics.Adecreaseintendoncompliance,withina
normalphysiologicalrange,necessitatedasubstantial
increaseinmusclefiberstretchandnegativemusculotendon
work(Fig.5).Thisillustratesthatwithareductionin
tendoncompliance,anincreasedriskforinjurycanoccur
unlesstherunningpostureadaptstoallowthemusculoten-
doncomplextooperateatshorterlengths.

INFLUENCEOFMUSCLECOORDINATIONON
HAMSTRINGMECHANICS
Anintriguingaspectofthepromisingrehabilitation
outcomesdemonstratedbySherryandBest(11)involves
thepotentialbenefitoftrunkstabilizationexercises.In
particular,theauthorssuggestthattheabilitytocontrolthe
lumbopelvicregionduringhigher-speedskilledmovements
maypreventhamstringreinjury.Neuromusculoskeletal
modelsprovideatoolbywhichtoquantitativelyassess
howindividualmusclesinfluencehamstringmechanicsand,
hence,injurypotential.
Theapproachtakenwastofirstgenerateasimulationof
thedouble-floatphaseofsprinting,thelastpartofthegait
cyclewhenpeakhamstringstretchoccurs.Thesimulation
generatedincluded58forcetrajectoriesforindividual
musclescrossingthelowertrunk,hip,knee,andankle.We
thenperturbedthemuscleforcetrajectoriesoneatatimeby
asmallamount(1N)andresimulatedthemovementfor
eachperturbation.Thechangeinhamstringmusculotendon

nUuahtrozider peorudtciMechanismsofHamstringStrains
139

noo fhtsia trcieli srphobitide.
stretchbecauseofeachperturbationwasascertained.The
changeinstretchwasthenscaledbytheaveragemuscleforce
overdoublefloat,resultinginanestimateoftheinfluenceof
eachmuscleonhamstringstretch.
Ouranalysissuggeststhatthecontralateralhipflexors(
i.e.
,
iliopsoas)haveaslargeaninfluenceonhamstringstretchas
thehamstringsthemselves(Fig.6).Thisoccursbecausethe
iliopsoascandirectlyinduceanincreaseinanteriorpelvic
tilt,whichinturnnecessitatesgreaterhamstringstretch.
Otherproximalmusclesactingonthepelvissuchasthe
abdominalobliquesanderectorspinaealsosubstantially
influencehamstringstretch.Themoredistalmusclesacting
aboutthekneeand/oranklehadmuchlessinfluenceon
hamstringmechanics.Thisanalysisillustratesamechanism
bywhichneuromuscularcontroloftrunkandpelvismuscles
canaffecthamstringstraininjuryrisk.
FUTUREDIRECTIONSANDRESEARCH
Thedevelopmentanduseofneuromusculoskeletal
modelsinherentlydependsonexperimentaldatabothto
formulateappropriatescientificquestionsandtovalidate
model-basedestimates.Giventhehighrateofinjury
recurrence,thereisaneedforabetterdescriptionofthe
structureandfunctionoftheremodeledmusculotendon

140
ExerciseandSportSciencesReviews

oCypirhg t @0260b yht emArecinaC loelego fpSrostM decini.e Figure5.
A.Force-length
Y
velocitypropertiesofmuscle
andforce-strainpropertiesoftendonwereusedtodescribe
musculotendoncontractiondynamics.Thetendonforce-
straincurvewasparameterizedby
?
0T
,definedasthe
tendonstrainresultingfromtheapplicationofmaximum
isometricforce(F
0M
).Simulationsoftheswingphaseof
sprintingwereperformedwitharangeof
?
0T
toestimate
theeffectoftendoncomplianceonfiberstretchandwork.
B.Adecreaseintendoncompliance,withinphysiologically
reportedranges,substantiallyincreasesthestretchand
negativeworkattributabletothemusclecomponentofthe
bicepsfemoris(13).Thisillustratesthatdecreasedseries
compliancebecauseoftheperformanceofrepeated
stretch-shorteningcontractions(2)orasaresultofpost-
injuryscarring(6)couldcontributetoincreasedinjuryrisk.

complexafterinjury.Inthisregard,magneticresonance
imagingmaybeausefultoolforcharacterizingtheextentof
scartissueinremodeledmuscle.Furthermore,newultrasound
approachesformeasuringcontractionmechanicscanbeused
tocharacterizemechanicalstrainsnearapriorinjurysite,at
leastundercontrolledcircumstancesinwhichthese
invivo
measurementscanbeperformed(3).Thesedatawouldbe
importantforunderstandingremodelingfactorsthatmight
influencereinjuryrisk.Theeventualcouplingofimaging-
basedmeasurementswithcontrolledinterventionstudiesmay
providenewunderstandingoftheeffectsofspecificexercises
onmusclefunctionandmechanics.
Thereisacomplimentaryneedforfurtherrefinementofthe
neuromusculoskeletalmodelstocharacterizelocalized
mechanicsaroundthemusculotendonjunctionwherethe
injuryoccurs.Hill-typemusculotendonmodelsinherently
assumeauniformmechanicalstraindistributionalongthe
musclefiber,whichmaynotbethecasewheninjuryoccurs.
Forexample,ithasbeensuggestedthatstraininjurymaybe
precededbydisruptionofindividualsarcomeres,leadingto
largelocalmechanicalstrains,whichpropagatetothe
musculotendonjunction(10).Thesubstantialchangein
stiffnessbetweenfibersandtendoncanthencontributeto
thefiberstearingatthejunction.Thedevelopmentand
incorporationofmorerefinedsofttissuemodelsareneededto

nUuahtrozider peorFigure6.
A.Anominalsimulationofthedouble-
floatphaseofsprintingwasgenerated.Individual
muscleforcetrajectorieswerethenperturbed(by1N
throughoutdoublefloat)andthesimulationregener-
ated.Perturbation-inducedchangesinhamstring
stretchweredeterminedandthenscaledbythe
associatedaveragemuscleforcetoquantifythe
muscle’sinfluence.B.Shownarethemusclesthat
exhibitedthegreatestinfluenceonbicepsfemoris
musculotendonstretch.Apositiveinfluencemeans
themuscleactstoincreasestretch.Thisanalysisshows
thatthecontralateraliliopsoas,abdominalobliques,and
ipsilaterallowbackmusclesallsubstantiallyinfluence
hamstringstretchviatheiractionsonthepelvisand,
thus,illustratesamechanismbywhichtrunkmuscles
mayaffectinjuryrisk(11).
www.acsm-essr.org

udtcoi nfot ih sraitlc esip orihibet.d
provideamoredetaileddescriptionofthisinjurymechanism
andcanalsoprovideameansforassessinghowremodeling
canalterlocalizedstrainpatterns.Therecentdevelopmentof
finiteelementmodelsofmusclethataccountforimportant
architectureeffectsmayfacilitatesuchanalyses(1).
Quantifyingtheroleofneuromuscularcoordinationin
preventinginjuryorreinjuryremainsachallenging
endeavor.Trunkstabilizationexercisesarewidelycitedas
providingsuchbenefitsforarangeofpathologies,although
thespecificmannerinwhichthebenefitsareachieved
remainstenuous.Inthisarticle,weshowedthatlumbo-
pelvicmusclescouldindeedinfluencelowerextremity
musclemechanics.However,complimentaryexperimental
approachesarestillneededtoinvestigatehowtrainingtruly
facilitatesimprovedcoordinationofthesemusclesduring
athleticmovements.Anunderstandingofthechangesin
coordinationcanthenbecoupledwithsimulationsof
movementtointerpretthemechanicalramifications.
CONCLUSIONS
Hamstringstrainsareacommonandrecurrentinjury
amongsprintingathletes.Theeffectivepreventionand
rehabilitationofsuchinjuriesremainschallenging.Inthis
article,wehaveshownthatneuromusculoskeletalmodels
enhancethefundamentalunderstandingoffactorsthat
affectbothinjuryandrehabilitationmechanisms.The
continueddevelopmentanduseofsuchmodels,together
withcontrolledexperimentalobservations,areimportant
forscientificallyestablishingeffectiveinjurypreventionand
rehabilitationprograms.
Acknowledgements
TheauthorsthankAllisonArnold,ThomasBest,LiLi,SteveSwanson,and
MichaelYoungfortheircontributionstothisresearch.Thisstudywas
supportedbytheAircastFoundationandNationalFootballLeague
Charities,aswellasanNationalScienceFoundationGraduateFellowship
toE.Chumanov.

Volume34
c
Number3
c
July2006

oCypirhg t @0260b yht emArecinaC loelego fpSrostM decini.e References
1.Blemker,S.S.,andS.L.Delp.Three-dimensionalrepresentationof
complexmusclearchitecturesandgeometries.
Ann.Biomed.Eng.
33:
661
Y
673,2005.
2.Butterfield,T.A.,andW.Herzog.Quantificationofmusclefiberstrain
duringin-vivorepetitivestretch-shorteningcycles.
J.Appl.Physiol.
99:593
Y
602,2005.
3.Fukunaga,T.,Y.Kawakami,K.Kubo,andH.Kanehisa.Muscleand
tendoninteractionduringhumanmovements.
Exerc.SportSci.Rev.
30:106
Y
110,2002.
4.Griffiths,R.I.Shorteningofmusclefibresduringstretchoftheactivecat
medialgastrocnemiusmuscle:theroleoftendoncompliance.
J.Physiol.
436:219
Y
236,1991.
5.Heiderscheit,B.C.,D.M.Hoerth,E.S.Chumanov,S.C.Swanson,
B.J.Thelen,andD.G.Thelen.Identifyingthetimeofoccurrenceofa
hamstringstraininjuringduringtreadmillrunning:acasestudy.
Clin.
Biomech.(Bristol,Avon)
20:1072
Y
1078,2005.
6.Kaariainen,M.,T.Jarvinen,M.Jarvinen,J.Rantanen,andH.Kalimo.
Relationbetweenmyofibersandconnectivetissueduringmuscleinjury
repair.
Scand.J.Med.Sci.Sports
10:332
Y
337,2000.
7.Koulouris,G.,andD.Connell.Evaluationofthehamstringmuscle
complexfollowingacuteinjury.
Skeletal.Radiol.
32:582
Y
589,2003.
8.Lieber,R.L.,andJ.Friden.Mechanismsofmuscleinjurygleanedfrom
animalmodels.
Am.J.Phys.Med.Rehabil.
81:S70
Y
S79,2002.
9.Orchard,J.,andT.M.Best.Themanagementofmusclestraininjuries:an
earlyreturnversustheriskofrecurrence.
Clin.J.SportMed.
12:3
Y
5,2002.
10.Proske,U.,D.L.Morgan,L.Brockett,andP.Percival.Identifying
athletesatriskofhamstringstrainsandhowtoprotectthem.
Clin.Exp.
Pharmacol.Physiol.
31:546
Y
550,2004.
11.Sherry,M.,andT.Best.Acomparisonoftworehabilitationprograms
inthetreatmentofacutehamstringstrains.
J.Orthop.SportsPhys.
Ther.
34:116
Y
125,2004.
12.Thelen,D.G.,F.C.Anderson,andS.L.Delp.Generatingdynamic
simulationsofmovementusingcomputedmusclecontrol.
J.Biomech.
13:321
Y
328,2003.
13.Thelen,D.G.,E.S.Chumanov,T.M.Best,S.C.Swanson,and
B.C.Heiderscheit.Simulationofbicepsfemorismusculotendon
mechanicsduringtheswingphaseofsprinting.
Med.Sci.Sports
37:
1931
Y
1938,2005.
14.Thelen,D.G.,E.S.Chumanov,D.M.Hoerth,T.M.Best,S.C.Swanson,
L.Li,M.Young,andB.C.Heiderscheit.Hamstringmusclekinematics
duringtreadmillsprinting.
Med.Sci.Sports
37:108
Y
114,2005.
15.Zajac,F.E.,andM.Winters.Modelingmusculoskeletalmovement
systems:jointandbodysegmentaldynamics,musculoskeletalactuation,
andneuro-muscularcontrol.In:
MultipleMuscleSystems
V
Biomechanics
andMovementOrganization
,editedbyJ.M.WintersandSL.-Y.Woo.
NewYork:Springer-Verlag,1990,pp.121
Y
148.

nUuahtrozider peorudtciMechanismsofHamstringStrains
141

noo fhtsia trcieli srphobitide.