Mtthew Zellmer ,Thoms W.Kernozek ,*,Nghmeh Gheidi ,Jordn Hove ,Michel Torry

a La Crosse Institute for Movement Science,Physical Therapy Program,Department of Health Professions,University of Wisconsin-La Crosse,La Crosse,WI 54601,USA

b Department of Kinesiology and Recreation,Illinois State University,Normal,IL 61761,USA

Abstract Background:Patellar tendinopathy(PT)or“jumper's knee”is generally found in active populations that perform jumping activities.Graded exposure of patellar tendon stress through functional exercise has been demonstrated to be effective for the treatment of PT.However,no studies havecompared how anterior knee displacement variations during thecommonly performed forward step lunge(FSL)affect patellar tendon stress.Methods:Twenty-f ivesubjects(age:22.69±0.74 years;height:169.39±6.44 cm;mass:61.55±9.74 kg)performed 2 variationsof an FSL with the anterior knee motion going in front of the toes(FSL-FT)and the knee remaining behind the toes(FSL-BT).Kinematic and kinetic data were used with an inverse-dynamics based static optimization technique to estimateindividual muscle forcesto determine patellar tendon stress during both lunge techniques.A repeated measures multivariate analysis was used to analyze these data.Results:Thepeak patellar tendon stress,stressimpulse,quadricepsforce,kneemoment,kneef lexion,and ankledorsif lexion angleweresignif icantly greater(p＜0.001)during the FSL-FT as compared to the FSL-BT.The peak patellar tendon stress rate did not differ between the FSL-FT and FSL-BT.Conclusion:The use of an FSL-FT as compared to an FSL-BT increased the load and stress on the patellar tendon.Because a graded exposure of patellar tendon loading with other closed kinetic chain exerciseshasproven to beeffective in treating PT,consideration for the prescription of variations of the FSL and further clinical evaluation of this exercise is warranted in individuals with PT.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:Biomechanics;Kinematics;Kinetics;Knee;Modeling;Rehabilitation

1.Introduction

Patellar tendinopathy(PT)or“jumper's knee”is generally found in active populations that perform jumping activities as demonstrated by theincreased prevalencein volleyball(31.9%)and basketball(44.6%)athletes.1Raisingand loweringexercises like squats,lunges,and heel raising and lowering exercises have been shown to be effective in treating lower extremity tendinopathies,and theseexercisesareoften described aseccentric strength training even though they both use concentric and eccentric muscle activation.2-6However,the reasons for eccentric loading regimen effectiveness are largely unknown.3,7Different neuromuscular changes could be responsible for the eccentricexercisebenef itsinrehabilitation.7Despitetheforward step lunge(FSL)being a commonly used exercise in lower extremity strength and rehabilitation programs8dueto dynamic balancedemands,9manipulation of lower extremity kinematics to increase patellar tendon stress,10and an eccentric lowering phase that involves all 4 quadriceps muscles,11it is rarely included in rehabilitation protocols for patients with PT.12,13Rather few investigationshavefocused on the FSL compared to the squat.Physical therapists often use different kinds of squat techniques during rehabilitation programs.14-17Most of the previousliteratureregarding PTrehabilitationexerciseshasfocused on thesquat on a decline board,14,16-18which ishypothesized to increasepatellar tendon stressby increasing thekneeextension moment on a decline board.Longpréet al.19reported greater quadriceps muscle activation and a higher internal knee extension moments during lunging vs.squatting.It seems that the use of the lunge and a variation in technique may serve to increase patellar tendon stress through functional exercise during rehabilitation.

There are different techniques for lunges,including variationsin step length,10walking or jumping lunges,11or different trunk positions.20Keepingthekneebehind thetoesisacommon cueduringperformingaproperformof squatandlunges.21,22The amountof anterior kneemotion relativeto thefoothasshown to affect quadriceps forces during an FSL,which may allow for increased stressonthepatellar tendon.Escamillaetal.10reported an increase of up to 30%in quadriceps forces,estimated by electromyography,during short FSL ascompared to along FSL and concluded that thepatellofemoral joint force and stressare smaller during long FSL compared to short FSL.However,by manipulating step length in the lunge they effectively altered knee f lexion angle,but not necessarily where the knee was in relationship to the foot.

Itisyetunknown how manipulating FSL by either keeping the knee over the foot and behind the toes vs.keeping the knee over the foot but allowing the knee to move in front of the toesin the sagittal plane with standardized step length may affect patellar tendon stress.Thus,the aim of our study wasto provide a comparison of patellar tendon stress during 2 variations of the FSL exercise:with the knee translated in front of the toes(FSL-FT)and with thekneetranslated behind thetoes(FSL-BT).Changes in shank position and knee-jointf lexion will alter thedirection of the ground reaction force vector and the moment arms.These variables may inf luence knee-joint loading.Therefore,we hypothesized thattherewill bean increased patellar tendon stress with FSL-FTduetotheincreased anterior motionof thekneeand resultant increased knee-f lexion angle and quadriceps force.

2.Materials and methods

2.1.Subjects

Twenty-f ive healthy female subjects(age:22.69±0.74 years;height:169.39±6.44 cm;mass:61.55±9.74 kg)participated in this study.The inclusion criteria included ascore of 4 or greater on the Tegner activity scale23and no report of pain or knee symptoms associated with patellofemoral pain syndrome or PT that had limited their functional or recreational activity in the past 12 months.Participants with a traumatic knee injury in the past 6 months on either knee,a surgery on either lower extremity in thepast 12 months,history of cardiovascular pathology,or currently pregnant were excluded.Informed consent was given by all participants regarding the testing protocol as approved by the Institutional Review Board at the University of Wisconsin-La Crosse.

2.2.Protocol

All participants were f itted with the same type of footwear(Model 629;New Balance,Boston,MA,USA)and completed a5-min activewarm-up by walking at aself-selected paceon a treadmill.Because the criteria for lunge step length varied in the literature,8,10,20,24-26100%of the leg length of the subject(apex of greater trochanter to apex of lateral malleolus on the right leg)wasused to standardize the step length for both FSL techniques.

A metronome,set at 0.5 Hz,was used to standardize a 2-s descending phaseand a 2-s ascending phase.The right leg was thelead leg for all participants.Prior to each trial,ademonstration and verbal explanation was provided by the same researcher for each participant.The order of the 2 FSL techniques was randomly determined through a coin f lip.For both FSL techniques,the subjects were cued to maintain an erect trunk posturewhilemaintaining their armsabducted at 90°with their elbowsfully extended.Wecontrolled trunk position asthis has been shown to inf luence kinematics,kinetics,and muscle activity during forward lunge.19We instructed each participant to“remain upright throughout each trial”in order to better examine the effect of anterior knee translation on patellar tendon stress.Lead-leg footcontact occurred just prior to astep marker that wasplaced at 100%of thestep length.Thelead-leg foot placement and contact position was constrained to a heel strike to foot f lat position while executing the lunge.The knee of the lead leg made contact with a guide cord that was placed at the level of the participants'knee to ensure proper anterior knee movement for the respected FSL technique.Thus,the major difference between the 2 FSL techniques was the placement of the guide cord.For the FSL-FT,the guide cord was placed at 110%of the step length in order to result in knee motion past the participants'toes(Fig.1A).For the FSL-BT,the guide cord was placed directly over the step marker(100%of leg length)such that anterior kneemotion did not go past the participants'toes(Fig.1B).Each subject performed 5 consecutiverepetitionsfor each FSL technique.Thetrialswererepeated if the subject demonstrated an inability to maintain temporal standardization during either the descending or ascending phases,improper anterior knee motion during the descending phase,or an inability perform the lunge with the correct step length.

Fig.1.(A)Forward step lunge with knee in front of toes(FSL-FT)and,(B)Forward step lunge with knee behind toes(FSL-BT).

2.3.Instrumentation

Forty-seven ref lective markers were placed on the body at the head,trunk,pelvis,and bilateral upper and lower extremities.27Head markers were placed on the right,left,top,and front.Trunk markers were placed on C7 and T10 spinous processes,navel,xiphoid process,sternal notch,and on the right scapula.Bilateral upper extremity markerswereplaced at the acromion process;near the deltoid insertion;medial and lateral humeral epicondyles;theforearm;at theulnar and radial styloid processes;and at thesecond metacarpophalangeal joint.Markers def ining the pelvis were placed at bilateral anterior superior and posterior superior iliac spinesalong with 1 marker being placed at the apex of the sacrum.Lower extremity markers were placed bilaterally on the greater trochanter,anterior thigh,lateral femoral epicondyle,anterior tibia,and lateral malleolus.The foot segment consisted of 3 markers placed on the shoe at the heel,the great toe,and the f ifth metacarpophalangeal joint.All markerswereleft in place during data collection.Fifteen cameras(Motion Analysis Corp.,Santa Rosa,CA,USA)were used to collect motion analysis data at 180 Hz.Synchronized analog data were collected at 1800 Hz with the useof 4 forceplatforms(Model 4080;Bertec Corp.,Columbus,OH,USA).

2.4.Data processing

Raw kinematic and analog data were f iltered with a second order Butterworth low-pass f ilter using an 8 Hz cutoff frequency.Joint angles,kinetic data,and muscle forces were processed using Human Body Model(Motek Medical,Amsterdam,the Netherlands).Muscle forces were calculated based ona44-degreeof freedom(DOF)musculoskeletal model with 16 rigid segments.27The head relative to the pelvis was modeled with 3-DOF.The trunk was modeled as 3 segments with 3-DOF,upper arm with 6-DOF,elbow with 2-DOF,and wrist with 2-DOF.Thepelvissegment had 6-DOFand wasable to rotate and translate in all 3 dimensions with respect to the ground.The kneewasmodeled as a1-DOFhingejoint and the ankle joint was modeled with 2-DOF,respectively.The inertial characteristicsof thesegmentsused in themodel werebased on participants'total body mass and segment lengths.28Overall,300 muscle tendon units were represented in the model in which muscle parameters such as muscle insertion and wrapping points were determined as in Delp et al.29

Muscle forces were estimated from the joint moments by minimizing a static cost function where the sum of squared muscleactivationswasrelated to maximum musclestrengthsat each time step of the model.30The static optimization problem was solved using a recurrent neural network.31The muscle forces and joint moments were normalized by weight.

The muscle forces from the Human Body Model were then used to quantify thetotal patellar tendon forceby summing the muscle forces of the rectus femoris,vastus medialis,vastus lateralis,and vastus intermedius throughout each repetition.The patellar tendon stress was calculated by dividing the patellar tendon forceby thecross-sectional area.Thepatellar tendon cross-sectional area was determined from Hansen et al.32and applied to all participants.Stressratewasthen determined from the instantaneous slope of the stress vs.the time curve.Stress impulse was determined based on the integrated stress time curve during each lunge repetition.

2.5.Statistical analysis

Sample size was calculated a priori usingβ=0.20 and α=0.05.A minimum sample size of 6 was determined based on peak patellar-tendon forcefrom Frohm et al.14A multivariate analysis of variance with repeated measures was performed using anα=0.05.Follow-up univariate tests were then performed to detect differencesbetween the2 typesof lungetechniques for each variable.Statistical calculations were completed in SPSS Version 22.0(IBM Corp.,Armonk,NY,USA)software.

3.Results

Multivariate analysis indicated there were differences in kinetic(Wilksλ=0.071,p＜0.001)and kinematic(Wilks λ=0.007,p＜0.001)variables between the FSL-FT and FSL-BT lunge techniques.The peak patellar tendon stress was 11.1%greater during the FSL-FT than the FSL-BT(Table 1).No differencewasfound betweenthe2 FSL techniquesfor peakpatellar tendon stress rate.Patellar tendon stress impulse was 18.8%greater during the FSL-FT as compared to the FSL-BT.

Table1Kinetic and kinematic variables for the forward step lunge behind the toes(FSL-BT)and forward step lung in front of toes(FSL-FT)techniques(mean±SD).

Fig.2.Ensembleaverageand standard deviation timeseriesgraphsof theforward step lunge(FSL)(mean±SD).(A)Patellar tendon stress;(B)kneef lexion angle;(C)knee extension moment;and(D)ankle plantar f lexion moment.FSL-BT=forward step lunge with knee behind toes;FSL-FT=forward step lunge with knee in front of toes.

Fig.2 illustrates the patellar tendon stress,knee f lexion angle,knee extension moment,and ankle plantar f lexion moment during FSL-BT and FSL-FT.Fig.2A shows that there was a similar trend for the mean patellar tendon stress during both FSL techniques,although the peak stress occurs at about 60%during FSL-FT and about 75%during FSL-BT.With respect to patellar tendon stress(Fig.2A)and knee f lexion angle(Fig.2B),during the descending phase(0-50%FSL),there was a progressive increase in patellar tendon stress,and during the ascending phase(50%-100%FSL)there was a progressive decrease in patellar tendon stress.In addition,the mean patellar tendon stress was 25.4%greater during the FSL-FT and 38.5%greater during the FSL-BT at the midpoint of theascending phase(75%FSL)ascompared to thedescending phase(25%FSL).Fig.2C depicts the knee extension moment,which reaches a peak at about 58%during FSL-FT and 70%during FSL-BT.Fig.2D depicts the ankle plantar f lexion moment,which peaksat about 65%during FSL-FT and 78%during FSL-BT.FSL-FT has a larger knee extension and ankle plantar f lexion moment overall.These peak knee extension and ankleplantar f lexion momentsfor both tasksoccurred during the ascending phase but slightly later during the FSLBT.These peak moments occurred earlier with the FSL-FT with similar timing as the peak patellar tendon stress.

The peak quadriceps forces and knee moment had a greater magnitude,12.6%and 25.8%,respectively,for the FSL-FT compared to the FSL-BT(Table 1).Likewise,peak hip f lexion moment(16.5%)and knee f lexion angle(13.2%)were greater during the FSL-FT ascompared to the FSL-BT(Table1).Peak hip f lexion anglewas11.0%greater during the FSL-BT condition.There was no difference found between trunk f lexion angles during the 2 FSL techniques.Peak knee f lexion angle,peak ankle dorsif lexion angle,and peak ankle plantar f lexion moment were 11.6%,58.2%,and 59.6%,respectively,greater during the FSL-FT than the FSL-BT(Table 1).

4.Discussion

Theprimary purposeof thisinvestigationwastodeterminehow alterations in sagittal anterior knee motion during an FSL affect patellar tendon stress.The study f indings support our hypotheses that the performance of a lunge with the knee translating beyond thetoes,asdemonstrated by the FSL-FT,resulted in greater peak patellar tendon stress and patellar tendon stress impulse,while FSL-FT displayed higher knee and ankle f lexion angles with less hip f lexion angle,higher knee extension moment,plantar f lexion moment,and less hip extensor moment compared to FSL-BT.These f indings could play a role in the formulation of future regimesutilized for thetreatment of individualswith PT.

The effects of a higher mechanical load in symptomatic tendons of individuals suffering from lower extremity tendinopathies have been well documented in the literature,with f indings of decreased neovascularization4and tendon hypertrophy,5,15as well as increased blood circulation.33The present data showed the peak patellar tendon stress and stress impulsewere,respectively,11.1%and 18.8%greater during the FSL-FT than the FSL-BT.The increased patellar tendon stress during the FSL-FT appears to provide greater patellar tendon loading and therefore may result in superior patellar tendon structural adaptations and improvements as compared to FSL-BTexercisewith individualswith PT.Themanipulation of performing an FSL-BT may also serve as a progression to the FSL-FT during rehabilitation.

Numerous studies have reported that a linear relationship exists between knee f lexion angle and quadriceps force during closed kinetic chain functional movements.14,34-36In the present study,a similar relationship was identif ied as there was both a greater peak kneef lexion angleand peak quadricepsforceduring the FSL-FTascompared to FSL-BT.Escamillaetal.10performed a comparable study depicting changes in patellofemoral joint stressduring step length variationsin the FSL.They reported an even larger increase in quadriceps force(20%-30%)during lunges with increased knee f lexion angle as compared to our study(12.6%),which appears due to our standardized step length.Stressisforceper area,so theincreaseinquadricepsforce during the FSL-FTappearsto beamain contributing factor in the greater patellar tendon stress.Because the FSL is considered to beaclosed kinetic chainexercise,ankle,and hip f lexionangleare increased and decreased,respectively,during FSL-FT.Not only joint angles are different,but also hip,knee,and ankle joint moments are different.Fig.2C and Fig.2D depicted the knee extension and ankle plantar f lexion moment,which is greater during FSL-FT.This difference is expected due to the more anterior location of the body center mass from the ankle during FSL-FT.This highlights the difference in the location of the ground reaction force vector between movements and how it seemsto inf luencethemoment armsrelativeto each joint.With a higher knee extension,plantar f lexion moment,and lesser hip extensor moment during FSL-FT of the lead leg,higher stress occursin the patellar tendon.

There were differences in peak patellar tendon stress and impulse between the 2 FSL techniques but no difference between peak patellar tendon stressrate(Table 1).Although we controlled time by standardizing the 2-s descending and ascending phase during the motions,FSL-FT demonstrated higher knee and ankle displacement rather than FSL-BT.Fig.2A showed thepatellar tendon stresswherethepeak stress occurs at about 60%during FSL-FT and 75%FSL-BT.The peak kneeextension and plantar f lexion occur later in themovement as well for the FSL-BT.In combination,these may affect the timing differences in peak patellar tendon stress.The peak patellar tendon stress curve slope looks similar until 20%and after 70%of motion(about 50%of total lungemotion).But the peak knee f lexion occurs at about 60%of the squat for both techniques(Fig.2B).These in combination indicated that the FSL-FT used a more rapid knee extension velocity on average as the overall rate of the exercises was controlled.With the similar rate of movement performed,the FSL-FT requires the greater knee and ankle motion,which increases the quadriceps force and therefore increases the patellar tendon stress.Even though the timing was largely controlled by using of the metronome,it is plausible that these increases in joint loading are inf luenced by both different knee positionsand different movement dynamics requiring greater muscle force in the lead leg.

Whilethemechanical loading of the patellar tendon hasbeen shown to be effective for the treatment of PT,3,15,16it is also important to notethatincreased ratesof loading havebeen found to placesoft tissue,such astendons,at risk for injury.37,38Therefore,rehabilitation programsincorporating either the FSL-BTor FSL-FT may consider a graded increase in loading rate as performance of these movements at a faster cadence would likely increase these loading rates.There were differences between stress impulses during the FSL techniques whereas the FSL-FT showed a higher impulse.Fig.2A shows the area under patellar tendon stress curve was higher during FSL-FT than FSL-BT.

Fig.2A represented patellar tendon stress during entire lungemotion,which included descending and ascending phase.Taken together with knee f lexion angle(Fig.2B),the peak patellar tendon stress appears to occur during the ascending phase.J?nhagen et al.,11supported the description of phases when they reported rectus femoris eccentric contraction after 54%of step and jump lunges.Thesef indingsmay havebeen the result of a greater quadriceps force requirement needed to accelerate the center of mass during the ascending phase as opposed to the decelerating of the center of mass during the descending phase.Our musculoskeletal model that uses static optimization accounts for the dynamics of the lunge activity and theforceproduction of musclethat span morethan asingle joint likethequadriceps.Other studiesanalyzing closed kinetic chain exercises with the same resistance applied throughout the entire repetition have supported this notion as they have reported greater quadricepsactivity during the ascending phase rather than the descending phase of this movement.39-43Meanwhile,somestudiesindicatethat theremay beaneuromuscular inhibition associated with maximum quadriceps contraction during the descending phase of these exercises.44,45

In this study,we controlled trunk f lexion because the manipulation of trunk position has been shown to alter patellar tendon stress.20Our f indings show no difference between peak trunk f lexion angles during either FSL technique.Farrokhi et al.20reported a 19.2%increase in knee extension impulse with lungeswherethetrunk wasin moreextension ascompared to when f lexed.These authors hypothesized that this occurred due to an increase in the knee extension moment due to the posterior location of the body center mass from the knee.This same concept has been applied to the prescription of declined squats for PT.Studies have shown that declined squat had greater kneef lexion coupled with amoreerect trunk position as compared to the traditional squat on a horizontal surface.14,18Therefore,thesechangesin thesagittal trunk position during an FSL-FT may allow for the ability to further manipulate the amount of patellar tendon stress based on the tissue irritability and the rehabilitation stage of an individual with PT.

There are some limitations that should be considered.Only femalesubjectsparticipated and thereforefurther investigations using male subjects may be warranted.In addition,static optimization wasutilized to estimatethemuscleforces.Without in vivo measurements of muscle force,it is diff icult to determine thetrueaccuracy of thistechnique.Our musculoskeletal model uses 1 degree of freedom for the knee joint,which may affect muscle force estimation and therefore may overestimate patellar tendon load.Lastly,a reference32was utilized for determining the patellar tendon cross-sectional areas instead of a direct measurement,whichmay haveaffected theresultantpeak patellar tendon stress.Many of these limitations,although important,may not inf luence the overall conclusions of our study becauseweutilized arepeated measuresdesignwhereall model parameters and assumptions were systematically applied to all subjects performing both lunge techniques.

5.Conclusion

Our data demonstrate that the patellar tendon stress,stress impulse,quadriceps force,knee extension moment,ankle plantar f lexion,and knee f lexion angle are higher during FSL-FT than in FSL-BT.Therefore,the patellar tendon undergoes greater loading during the FSL-FT compared to the FSLBT.Because the squat is performed commonly during PT rehabilitation,further research appears warranted wherein decline squats and the FSL-FT could be examined for their effectiveness in providing patellar tendon stress as well as with rehabilitation outcomes of individuals with PT.

Acknowledgment

This work has received funding from a Graduate Student Research,Service,and Educational Leadership Grant at the University of Wisconsin-La Crosse.

Authors'contributions

TWK was involved in the design of the study and project conception,datacollection,dataprocessing and performed statistical analysis;NGwasinvolved in thedesign of thestudy and project conception,data collection,data processing and performed statistical analysis,and data interpretation and drafting the original manuscript;MZ was involved in data collection,dataprocessing and analysis,performed statistical analysis,and datainterpretation and drafting theoriginal manuscript;JH was involved in data collection,data processing and analysis;MT were involved in data interpretation and drafting the original 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.