Patellofemoral Compressive Force and Stress During the Forward and Side Lunge With and Without a Stride

Available online at www.sciencedirect.com

www.elsevier.com/locate/clinbiomech

Clinical Biomechanics 23 (2008) 1026–1037

Patellofemoral compressive force and stress during the forwardand side lunges with and without a stride

Rafael F. Escamilla a,*, Naiquan Zheng b, Toran D. MacLeod a,h, W. Brent Edwards c,Alan Hreljac d, Glenn S. Fleisig e, Kevin E. Wilk f, Claude T. Moorman III g,

Rodney Imamura d

a Department of Physical Therapy, California State University, 6000 J Street Sacramento, CA 95819-6020, USAb The Center for Biomedical Engineering, Department of Mechanical Engineering and Engineering Science, University of North Carolina, Charlotte, NC, USA

c Department of Kinesiology, Iowa State University, Ames, IA, USAd Kinesiology and Health Science Department, California State University, Sacramento, CA 95819-6020, USA

e American Sports Medicine Institute, Birmingham, AL, USAf Champion Sports Medicine, Birmingham, AL, USA

g Duke University Medical Center, Durham, NC, USAh Department of Physical Therapy, Center for Biomedical Engineering Research, University of Delaware, Newark, DE, USA

Received 31 July 2007; accepted 2 May 2008

Abstract

Background. Although weight bearing lunge exercises are frequently employed during patellofemoral rehabilitation, patellofemoralcompressive force and stress are currently unknown for these exercises.

Methods. Eighteen subjects used their 12 repetition maximum weight while performing forward and side lunges with and without astride. EMG, force platform, and kinematic variables were input into a biomechanical model, and patellofemoral compressive force andstress were calculated as a function of knee angle.

Findings. Patellofemoral force and stress progressively increased as knee flexion increased and progressively decreased as knee flexiondecreased. Patellofemoral force and stress were greater in the side lunge compared to the forward lunge between 80� and 90� knee angles,and greater with a stride compared to without a stride between 10� and 50� knee angles. There were no significant interactions betweenlunge variations and stride variations.

Interpretation. A more functional knee flexion range between 0� and 50� may be appropriate during the early phases of patellofemoralrehabilitation due to lower patellofemoral compressive force and stress during this range compared to higher knee angles between 60� and90�. Moreover, when the goal is to minimize patellofemoral compressive force and stress, it may be prudent to employ forward and sidelunges without a stride compared to with a stride, especially at lower knee angles between 0� and 50�. Understanding differences in patel-lofemoral compressive force and stress among lunge variations may help clinicians prescribe safer and more effective exercise interventions.� 2008 Published by Elsevier Ltd.

Keywords: Knee pain; Knee kinetics; Knee biomechanics; Patellofemoral rehabilitation

1. Introduction

Patellofemoral rehabilitation results in millions of dollars(USA) in medical cost each year, and employing exercise

0268-0033/$ - see front matter � 2008 Published by Elsevier Ltd.

doi:10.1016/j.clinbiomech.2008.05.002

* Corresponding author.E-mail address: [email protected] (R.F. Escamilla).

therapy during patellofemoral rehabilitation has demon-strated beneficial, positive changes in cost effectiveness, painseverity, and functional disability (van Linschoten et al.,2006). Moreover, the quality of patellofemoral rehabilita-tion is paramount in determining the health and well beingof the patient, and a faster return to function (vanLinschoten et al., 2006). Consequently, judicious thought

mailto:[email protected]

R.F. Escamilla et al. / Clinical Biomechanics 23 (2008) 1026–1037 1027

should be given in choosing patellofemoral rehabilitationexercises.

Patellofemoral pain syndrome (PFPS), otherwise knownas anterior knee pain, is the most common cause of kneepain in the outpatient setting, and accounts for 25–30%of all knee pathologies treated (Devereaux and Lachmann,1984; Dixit et al., 2007; Fredericson and Yoon, 2006).PFPS primarily affects younger (typically the 18–40 agerange) active individuals, both athletes and non-athletes(Dixit et al., 2007; Fredericson and Yoon, 2006; LaBella,2004) and both males and females (Dehaven and Lintner,1986), although older individuals can also be affected.Because the etiology of PFPS is poorly understood andmulti-faceted, it remains one of the most difficult clinicalchallenges in rehabilitative medicine (Wilk et al., 1998).Although patellofemoral rehabilitation after injury can bea long and arduous process, the use of appropriate exer-cises can improve this process by decreasing rehabilitationtime and improving function (Boling et al., 2006; Heintjeset al., 2003; Natri et al., 1998; Witvrouw et al., 2004; Witv-rouw et al., 2000).

Weight bearing exercises are frequently employed duringpatellofemoral rehabilitation, and are specific to many func-tional activities such as walking, running, and jumping (Bol-ing et al., 2006; Heintjes et al., 2003; Natri et al., 1998;Witvrouw et al., 2004; Witvrouw et al., 2000). The use ofweight bearing exercises have been shown to be effective,both in short and long term outcomes, in decreasing anteriorknee pain and enhancing functional performance (Bolinget al., 2006; Heintjes et al., 2003; Natri et al., 1998; Witv-rouw et al., 2004; Witvrouw et al., 2000). Clinicians use thesetypes of exercises to minimize anterior knee pain and muscleloss, strengthen hip and thigh musculature, enhance balanceand stability, and minimize the risk of future injuries andassociated costs of health care (van Linschoten et al.,2006). Understanding how patellofemoral compressiveforce and stress (force per unit patella contact area) varyamong weight bearing exercises allows physical therapists,physicians, and trainers to prescribe safer and more effectiveknee rehabilitation treatment to patients recovering fromknee injury or surgery. For example, if an exercise generatesgreater patellofemoral compressive force and stress between60� and 90� knee flexion compared to 0–30� knee flexion, the0–30� knee flexion range can be initiated at an earlier stage inpatellofemoral rehabilitation, while the 60–90� knee flexionrange can be initiated at a later stage in the rehabilitationprocess. There may also be differences in patellofemoralcompressive force and stress between different exercises overa specific knee flexion range.

Lunges are common weight bearing exercises used byathletes and other individuals with healthy knees to trainthe hip and thigh musculature. Therapists and trainers alsouse lunges and similar weight bearing exercises duringpatellofemoral rehabilitation for PFPS patients to allowthem to recover faster and return to function earlier (Bol-ing et al., 2006; Heintjes et al., 2003; Natri et al., 1998;Witvrouw et al., 2004; Witvrouw et al., 2000). Lunges

can be done with varying techniques, such as forward orside lunges, as well as lunging with a stride (striding for-ward or sideways with one leg and returning back toupright position with feet together), or lunging without astride (keeping both feet stationary throughout the forwardor side lunge movements). However, patellofemoral com-pressive force and stress generated while performing for-ward and side lunges with and without a stride iscurrently unknown.

Patellofemoral compressive force, which can elevatesubchondral bone stress (Besier et al., 2005), is a plausiblecause of pain in individuals with PFPS. Patellofemoralstress can result in cartilage degeneration and a decreasein the ability of the cartilage to distribute patellofemoralcompressive force (Besier et al., 2005). Therefore, it isimportant for clinicians to understanding the magnitudesof patellofemoral compressive force and stress generatedduring different weight bearing rehabilitation exercises.The purpose of this study was to compare patellofemoralcompressive force and stress between forward and sidelunges with and without a stride. It was hypothesized thatpatellofemoral compressive force and stress would increaseas knee flexion increased, would be greater with a stridecompared to without a stride, and would be greater inthe forward lunge compared to the side lunge.

2. Methods

2.1. Subjects

Eighteen healthy individuals (9 males and 9 females)without a history of patellofemoral pathology participatedwith an average(SD) age, mass, and height of 29(7) y,77(9) kg, and 177(6) cm, respectively, for males, and 25(2)y, 60(4) kg and 164(6) cm, respectively, for females. Inaddition, all subjects were required to be able to performall exercises pain free and with proper form and techniquefor 12 consecutive repetitions using their 12 repetition max-imum (12 RM) weight.

To control the electromyographic (EMG) signal quality,the current study was limited to males and females that hadaverage or below average body fat, which was assessed byBaseline skinfold calipers (Model 68900, Country Technol-ogy, Inc., Gays Mill, WI, USA) and appropriate regressionequations and body fat standards set by the American Col-lege of Sports Medicine. Average (SD) body fat was 12(4)%for males and 18(1)% for females. All subjects providedwritten informed consent in accordance with the Institu-tional Review Board at California State University, Sacra-mento, USA, which approved the research conducted andinformed consent form.

2.2. Exercise descriptions

2.2.1. Forward lungeEach subject performed the forward lunge (Fig. 1a) both

with and without stride. The starting and ending positions

Fig. 1. (a) Forward lunge; (b) side lunge.

1028 R.F. Escamilla et al. / Clinical Biomechanics 23 (2008) 1026–1037

for the forward lunge with a stride were the same, whichinvolved standing upright with both feet together. Fromthe starting position for the forward lunge with a stride,the subject held a dumbbell weight in each hand and lungedforward with the right leg toward a force platform atground level. The dumbbell weight used during the forwardlunge with stride was normalized by each subject’s 12 RMweight, and this same weight was used during the forwardlunge without stride. The mean (SD) mass of both dumb-

bells during the forward lunge was 49(11) kg for malesand 32(8) kg for females. At right foot contact the rightknee slowly flexed until maximum right knee flexion of90–100� was obtained as the left knee made contact withthe ground. From this ending position the subject immedi-ately pushed backward off the force platform and returnedto the starting position. A metronome was used to ensurethat the knee flexed and extended at approximately 45�/s.Each subject was instructed to use as long a stride length


as was comfortable. A tester ensured that the stride waslong enough so that at the lowest position of the lungethe stride knee was maintained over the stride foot withouttranslating forward beyond the toes. The average (SD)stride length (measured from left toe to right heel) duringthe forward lunge of 89(4) cm for males and 79(6) cm forfemales. During the forward lunge, a longer stride lengthis commonly preferred by individuals compared to ashorter stride length to ensure that the stride knee doesnot progress beyond the toes at the lowest position of theforward lunge (Fig. 1a). The forward lunge without stridewas performed the same as the forward lunge with stridewith the exception that both feet remained stationarythroughout each repetition during the forward lunge with-out stride. That is, from the lowest position of the forwardlunge shown in Fig. 1a, the subject simply fully extendedboth knees, and then flexed both knees back to return tothe lowest position of the forward lunge shown in Fig. 1a.

2.2.2. Side lunge

Each subject performed the side lunge (Fig. 1b) bothwith and without a stride. The starting and ending posi-tions for the side lunge with a stride were both the same,which involved standing upright with both feet together.From the starting position for the side lunge with a stride,the subject held a single dumbbell weight down between thelegs and lunged sideways with the left knee remaining fullyextended and the right leg moving toward a force platformat ground level. The dumbbell weight used during the sidelunge with stride was normalized by each subject’s 12 RMweight, and this same weight was used during the side lungewithout stride. On the average (SD), the dumbbell massused during the side lunge was 34(9) kg for males and20(5) kg for females. At right foot contact the right footwas turned out approximately 30–45� relative to the leftfoot (subject’s preference) and the right knee flexed slowlyat approximately 45�/s until the right knee flexed approxi-mately 90–100� (Fig. 1b) as the left knee remained fullyextended. From this ending position the subject thenpushed backward off the force platform and returned tothe starting position. A metronome was used to ensure thatthe knee flexed and extended at approximately 45�/s. Eachsubject was instructed to use as long a stride length as wascomfortable. A tester ensured that the stride was longenough so that at the lowest position of the lunge the strideknee was maintained over the stride foot without translat-ing forward beyond the toes. The average (SD) stridelength (measured from inside of left heel to inside of rightheel) during the side lunge was 94(5) cm for males and83(3) cm for females. Using a longer stride length com-pared to a shorter stride length during the side lunge iscommonly preferred by individuals as it allows the left kneeto remain straight and the right knee to flex approximately90–100� and remain over the foot. The side lunge withoutstride was performed the same as the side lunge with stridewith the exception that both feet remained stationarythroughout each repetition during the side lunge without

stride. That is, from the lowest position of the side lungeshown in Fig. 1b, the subject simply fully extended the rightknee (left knee was already extended), and then returnedback to the lowest position of the side lunge by flexingthe right knee.

2.3. Data collection

Each subject came in for a pre-test one week prior to thetesting session. At that time the experimental protocol wasreviewed and the subject was given the opportunity to askquestions. In addition, each subject’s 12 RM was deter-mined for the forward and side lunges by utilizing the mostweight they could lift for 12 consecutive repetitions. More-over, each subject’s stride length (as previously defined)was measured for the forward and side lunges.

Blue Sensor (Ambu Inc., Linthicum, MD, USA) dispos-able surface electrodes (type M-00-S) were used to collectEMG data. These oval shaped electrodes (22 mm wide and30 mm long) were placed in a bipolar electrode configurationalong the longitudinal axis of muscle, with a center-to-centerdistance of approximately 3 cm between electrodes. Prior topositioning the electrodes over each muscle, the skin was pre-pared by shaving, abrading, and cleaning with isopropylalcohol wipes to reduce skin impedance. As previouslydescribed, (Basmajian and Blumenstein, 1980) electrodepairs were then placed on the subject’s right side for the fol-lowing muscles: (1) rectus femoris; (2) vastus lateralis; (3)vastus medialis; (4) medial hamstrings (semitendinosis andsemitendinosus); (5) lateral hamstrings (biceps femoris);and (6) gastrocnemius.

EMG data were collected at 960 Hz using a NoraxonMyosystem EMG unit (Noraxon USA, Inc., Scottsdale,AZ, USA). The amplifier bandwidth frequency was 10–500 Hz, the input impedance of the amplifier was20,000 kX, and the common-mode rejection ratio was130 Db. EMG signals were processed through an analog todigital (A/D) converter by a 16 bit A/D board.

Spheres (3.8 cm in diameter) covered with 3 MTM reflectivetape were attached to adhesives and positioned over the fol-lowing bony landmarks: medial and lateral malleoli of theright foot; upper edges of the medial and lateral tibial pla-teaus of the right knee; posterosuperior greater trochantersof the left and right femurs; lateral acromion of the rightshoulder; and third metatarsal head of the right foot.

A six camera Peak Performance motion analysis system(Vicon-Peak Performance Technologies, Inc., Englewood,CO, USA) was used to collect 60 Hz video data. Each sub-ject performed each lunge variation with their right foot onan AMTI force platform (Model OR6-6-2000, AdvancedMechanical Technologies, Inc.) and their left foot on theground (Fig. 1a and b), with force platform data collectedat 960 Hz. Once the electrodes were positioned, the subjectwarmed up and practiced the exercises as needed, and datacollection commenced. Video, EMG, and force platformdata were electronically synchronized and collected as eachsubject performed in a randomized manner one set of three


continuous repetitions (trials) during the forward lungewith stride, forward lunge without stride, side lunge withstride, and side lunge without stride.

Subsequent to completing all exercise variations, EMGdata were collected during maximum voluntary isometriccontractions (MVIC) to normalize the EMG data collectedduring each lunge variation. The MVIC for the rectus femo-ris, vastus lateralis, and vastus medialis were collected in aseated position at 90� knee and hip flexion with a maximumeffort knee extension. The MVIC for the lateral and medialhamstrings were collected in a seated position at 90� kneeand hip flexion with a maximum effort knee flexion. MVICfor the gastrocnemius was collected during a maximumeffort standing one leg toe raise with the ankle positionedapproximately halfway between neutral and full plantar flex-ion. Two 5 s trials were randomly collected for each MVIC.

2.4. Data reduction

Video images for each reflective marker were trackedand digitized in three-dimensional space with Peak Perfor-mance software, utilizing the direct linear transformationcalibration method (Shapiro, 1978). Testing of the accu-racy of the calibration system resulted in reflective ballsthat could be located in three-dimensional space with anerror less than 0.7 cm. The raw position data weresmoothed with a double-pass fourth order Butterworthlow-pass filter with a cut-off frequency of 6 Hz (Escamillaet al., 1998). Joint angles, linear and angular velocities,and linear and angular accelerations were calculated in atwo-dimensional sagittal plane of the knee utilizing appro-priate kinematic equations (Escamilla et al., 1998).

Raw EMG signals were full-waved rectified andsmoothed with a 10 ms moving average window (linearenveloped) throughout the knee range of motion for eachrepetition. These EMG data were then normalized for eachmuscle and expressed as a percentage of each subject’shighest corresponding MVIC trial. The MVIC trials werecalculated using the highest EMG signal over a 1 s timeinterval throughout the 5 s MVIC. Normalized EMG datafor the three repetitions (trials) were then averaged at cor-responding knee angles between 0� and 90�, and were usedin the biomechanical model described below.

2.5. Biomechanical model

As previously described (Escamilla et al., 1998; Zhenget al., 1998), a biomechanical model of the knee was usedto continuously calculate patellofemoral forces throughouta 90� knee range of motion during the knee flexing (0–90�)and knee extending (90–0�) phases of the lunge exercises.Resultant force and torque equilibrium equations were cal-culated using inverse dynamics and the biomechanical kneemodel (Escamilla et al., 1998; Zheng et al., 1998). Momentarms of muscle forces and angles of the line of action formuscles were represented as polynomial functions of theknee flexion angle using data from Herzog and Read (1993).

Quadriceps, hamstrings, and gastrocnemius muscleforces were calculated as previously described (Escamillaet al., 1998; Zheng et al., 1998). Because the accuracy ofcalculating muscle forces depends on accurate calculationsof a muscle’s physiological cross-sectional area (PCSA),maximum voluntary contraction force per unit PCSA,and the EMG-force relationship, resultant force and tor-que equilibrium equations may not be satisfied. Therefore,each muscle force Fm(i) was modified by the following equa-tion: Fm(i) = ciklikviAirm(i)[EMGi/MVICi], where Ai wasPCSA of the ith muscle, rm(i) was MVIC force per unitPCSA of the ith muscle, EMGi and MVICi were EMGwindow averages of the ith muscle EMG during exerciseand MVIC trials, ci was a weight factor (explained below)adjusted in a computer optimization program to minimizethe difference between the resultant torque from the inversedynamics (Tres) and the resultant torque calculation fromthe biomechanical model (Tmi), kli represented each mus-cle’s force–length relationship as function of hip and kneeangles (based on muscle length, fiber length, sarcomerelength, pennation angle, and cross-sectional area) (Wick-iewicz et al., 1983), and kvi represented each muscle’sforce–velocity relationship based on a Hill-type model foreccentric and concentric muscle actions using the followingequations from Zajac (1989)) and Epstein and Herzog(1998):

kv ¼ ðb� ða=F 0ÞvÞ=ðbþ vÞ concentric

kv ¼ C � ðC � 1Þðbþ ða=F 0ÞvÞ=ðb� vÞ eccentric

with F0 representing isometric muscle force, v = velocity,a = 0.32 F0, b = 3.2 l0/sec, and C = 1.8. Muscle force fromeccentric contractions was scaled up by 1.8 times the iso-metric muscle force F0. Forces generated by the knee flex-ors and extensors at MVIC were assumed to be linearlyproportional to their physiological cross-sectional area.Muscle force per unit physiological cross-sectional area atMVIC was 35 Ncm�2 for the knee flexors and 40 Ncm�2

for the quadriceps (Cholewicki et al., 1995; Narici et al.,1992; Narici et al., 1988; Wickiewicz et al., 1984).

The objective function used to determine each ith mus-cle’s coefficient ci was as follows:

min f ðci Þ ¼Xnm

i¼1

ð1� ciÞ2 þ kðT res �Xnm

i¼1

T miÞ2;

subject to clow 6 ci6chigh, where clow and chigh were lowerand upper limits for ci, and k was a constant. The weightfactor ‘‘c” was to adjust the final muscle force calculation.The bounds on ‘‘c” were set between 0.5 and 1.5. The tor-ques predicted by the EMG driven model matched well(<2%) with the torques generated from the inverse dynam-ics. The assumptions associated with this model are (1) thetorque from cruciate ligament forces were ignored (2) otherforces and torques out of the sagittal plane were ignored.

Patellofemoral force was a function of patellar tendonforce and quadriceps tendon force. Patellar tendon forcewas calculated by the quadriceps tendon force and the ratio


of the patellar tendon force and the quadriceps tendonforce, as previously described (van Eijden et al., 1986;van Eijden et al., 1987). The angles between the patellartendon, quadriceps tendon, and patellofemoral joint wereexpressed as functions of knee angle (van Eijden et al.,1986; van Eijden et al., 1987).

Patellofemoral stress, which was calculated every 10�between 0� and 90� knee angle, was expressed as the ratioof patellofemoral force (calculated from the biomechanicalmodel described above) and patellar contact area (Escam-illa et al., 1998; Zheng et al., 1998). Patellar contact areaswere determined at 10� intervals between 0� and 90� kneeangle. Contact areas from in vivo MRI data from Salsichet al. (2003) (who used both male and female subjects withhealthy knees and had them perform weight bearing exer-cise using resistance, similar to the current study) were usedat 0� (146 mm2), 20� (184 mm2), 40� (290 mm2), and 60�(347 mm2) knee angles. These four contact area valuesformed a near linear relationship as a function of kneeangle, resulting in a line of best fit equation ofy = 3.55x + 135 (r = 0.98) with y = contact area andx = knee angle. This line of best fit equation was used todetermine contact areas at 10� knee angle (171 mm2), 30�knee angle (242 mm2), and 50� knee angle (313 mm2).The contact areas at 40�, 50�, and 60� knee angles wereused to develop the line of best fit equationy = 2.81x + 176 (r = 0.99), which was used to determinecontact areas at 70� knee angle (373 mm2), 80� knee angle(401 mm2), and 90� knee angle (429 mm2). Like the currentstudy, a near linear relationship between patellar contactarea and knee angles has been reported between 0� and90� knee angles in several studies involving weight bearingexercises (Besier et al., 2005; Cohen et al., 2001; Hinterw-immer et al., 2005; Patel et al., 2003; Salsich et al., 2003).

2.6. Data analysis

To determine significant differences in patellofemoralforces between the two lunge variations (forward lungeand side lunge) and two stride variations (with stride andwithout stride), patellofemoral forces were statistically ana-lyzed every 10� during the 0–90� knee flexing phase and the90–0� knee extending phase using a two factor (lunge vari-ations and stride variations) repeated measure Analysis ofvariance. Bonferroni t-tests were used to assess pairwisecomparisons. The level of significance used was P < 0.01.Because each patellofemoral stress value was derived bydividing each patellofemoral force value by a constant foreach lunge variation and each stride variation, patellofe-moral stress values between lunge and stride variations willhave the same P values as the corresponding patellofemor-al force values in which they were derived.

3. Results

Patellofemoral force and stress progressively increasedas knee flexion increased and progressively decreased as

knee flexion decreased (Figs. 2–7). Table 1 shows patellofe-moral force values as a function of knee angle between for-ward and side lunges and between with a stride andwithout a stride. From Table 1, between 80� and 90� kneeangles of the knee flexing phase and at 90� knee angle of theknee extending phase patellofemoral force and stress weregreater in the side lunge compared to the forward lunge.Between 10� and 50� knee angles of the knee flexing phaseand between 50� and 20� knee angles of the knee extendingphase patellofemoral force and stress were significantlygreater with a stride compared to without a stride. Therewere no significant interactions between lunge variationsand stride variations.

4. Discussion

Both patellofemoral compressive force and stress canresult in the degeneration of patellofemoral cartilage andanterior knee pain from subchondral bone, synovial plicae,infrapatellar fat pad, retinacula, joint capsule, tendons andligaments (Biedert and Sanchis-Alfonso, 2002). Anotherproposed mechanism of anterior knee pain is increasedpatellar cartilage stress from patellofemoral compressiveforce leading to subchondral bone stress (Biedert and San-chis-Alfonso, 2002), and it has been demonstrated that thesubchondral bone plate is rich in pain receptors (Wojtyset al., 1990). Therefore, understanding differences in patel-lofemoral compressive force and stress among forward andside lunge variations is helpful to the clinician when pre-scribing therapeutic exercises to individuals with PFPS.

Contrary to our original hypothesis, patellofemoralcompressive force and stress were greater while performingthe side lunge compared to the forward lunge, but only athigher knee angles between 80� and 90�. From these data itcan be concluded that loading the patellofemoral joint issimilar between forward and side lunges except at higherknee angles, where patellofemoral loading was greater inthe side lunge.

As hypothesized, patellofemoral compressive force andstress were greater with a stride compared to without astride, but only between 0� and 50� knee angles. Performingforward and side lunges without a stride within a smallerknee angle range (eg, 0–50�) may be easier and safer to startwith earlier during patellofemoral rehabilitation when thegoal to minimize patellofemoral compressive force andstress, while the performing forward and side lunges witha stride may be more appropriate later during patellofe-moral rehabilitation due to greater patellofemoral com-pressive force and stress compared to without a stride.

Patellofemoral compressive force and stress curves weresimilar in shape to each other due to proportional increasesin patellofemoral compressive force and patellar contactarea with increased knee flexion. One exception was withknee angles between 80� and 90�, which resulted in adecrease in patellofemoral stress. This occurred becausealthough patellar contact area increased between 80� and90�, patellofemoral compressive force did not increase

Fig. 2. Mean (SD) patellofemoral compressive force between forward and side lunges with a stride.

Fig. 3. Mean (SD) patellofemoral stress between forward and side lunges with a stride.


proportionally, but instead began to plateau. This impliesthat patellofemoral stress decreased during forward andside lunges as maximum knee flexion was approached.These findings are consistent with patellofemoral compres-sive force and stress data during the barbell squat fromEscamilla et al. (1998) and Salem and Powers (2001).Escamilla et al. (1998) reported that patellofemoral com-pressive forces increases until 75–80� knee flexion, and thenbegan to level off and plateau or slightly decrease. Salemand Powers (2001) reported no significant differences inpatellofemoral compressive force or stress at 75�, 100�,and 110� knee flexion. It can be concluded from these squatdata that injury risk to the patellofemoral joint may notincrease with knee angles between 75� and 110� due to sim-ilar magnitudes in patellofemoral stress during these kneeangles, with the benefit of increased quadriceps, ham-

strings, and gastrocnemius activity when training at higherknee angles (75–110�) compared to training at lower kneeangles (0–75�) (Escamilla et al., 1998).

Another consideration during patellofemoral rehabilita-tion is what knee flexion range of motion to employ whileperforming lunge exercises. Because patellofemoral com-pressive force and stress both increased with increased kneeflexion and decreased with decreased knee flexion, a morefunctional knee flexion range between 0� and 50� may beappropriate during the early phases of patellofemoral reha-bilitation due to lower patellofemoral compressive forceand stress that is generated during this range of motion.Higher knee angles between 60� and 90� may be moreappropriate later in the rehabilitation process due to higherpatellofemoral compressive force and stress that are gener-ated during this range of motion. For example, during the

Fig. 4. Mean (SD) patellofemoral compressive force between forward and side lunges without a stride.

Fig. 5. Mean (SD) patellofemoral stress between forward and side lunges without a stride.


knee extending phase of the forward lunge, patellofemoralcompressive force ranged from 66 to 1176 N between 0�and 50� knee angle and from 1577 to 2191 N between 60�and 90� knee angle (Table 1), and patellofemoral stress ran-ged from 0.43 to 3.76 MPa between 0� and 50� knee angleand from approximately 4.54–5.11 MPa between 60� and90� knee angle. This same pattern of increased patellofe-moral compressive force and stress with increased kneeflexion has been reported during the barbell squat and legpress (Escamilla et al., 1998; Escamilla et al., 2001; Salemand Powers, 2001; Steinkamp et al., 1993; Wallace et al.,2002). These authors reported that patellofemoral com-pressive force and stress progressively increased from 0�to approximately 90�, peaked near 90�, and then progres-sively decreasing from approximately 90–0�. Computeroptimization techniques demonstrated similar results dur-ing a simulated squat (Cohen et al., 2003).

Peak patellofemoral force and stress magnitudes fromthe current study are greater compared to weight bearingfunctional exercises such as walking (Heino Brechter andPowers, 2002) and going up and down stairs (Brechterand Powers, 2002), but less than other weight bearingactivities, such as the squat and leg press (Escamillaet al., 1998). Escamilla et al. (1998) reported peak patellofe-moral force magnitudes between 4500 and 4700 N at 90�knee angle during the 12 RM squat and leg press usinghealthy subjects, resulting in patellofemoral stress magni-tudes between 11 and 12 MPa. These peak force and stressmagnitudes during the squat and leg press are approxi-mately 60% greater compared to peak force and stress mag-nitudes in the current study, which also occurred at 90�knee angle. Peak patellofemoral force and stress in healthysubjects during fast walking reportedly are approximately900 N and 3.13 MPa, respectively (Heino Brechter and

Fig. 6. Mean (SD) patellofemoral compressive force between forward lunge with and without a stride.

Fig. 7. Mean (SD) patellofemoral compressive force between side lunge with and without a stride.


Powers, 2002), which are approximately 2–3 times lowerthan the peak force and stress magnitudes in the currentstudy. However, peak patellofemoral force and stress mag-nitudes in healthy subjects going up and down stairsreportedly are approximately 2500 N and 7 MPa, respec-tively (Heino Brechter and Powers, 2002), which are similarto the peak force and stress magnitudes in the currentstudy.

Unlike healthy subjects, patients with patellofemoralpathologies have demonstrated smaller patellar contactareas and greater patellofemoral stress during some weightbearing functional activities.(Brechter and Powers, 2002;Heino Brechter and Powers, 2002). However, for bothhealthy subjects and patients with patellofemoral patholo-gies it is currently unknown what patellofemoral force orstress magnitudes, and over what time duration, can ulti-mately lead to patellofemoral pathology or exacerbate an

existing condition. There are many factors that may con-tribute to patellofemoral pathology, such as: (1) imbalanceor malalignment of the extensor mechanism, which canlead to lateral patellar subluxation or tilt; (2) muscle weak-ness, such as weak quadriceps and hip external rotators; (3)overuse or trauma; (4) muscle tightness, such as tight quad-riceps, hamstrings, or iliotibial band; (5) lower extremitymalalignment, such as patella alta; genu valgum, femoralneck anteversion, excessive Q angle; and excessive rearfootpronation. It can only be surmised that relatively largepatellofemoral force and stress magnitudes over time maylead to or exacerbate patellofemoral pathology, especiallyin individuals that exhibit some of the above factors andare predisposed to patellofemoral problems. Clinicianscan use information regarding patellofemoral force andstress magnitudes among different weight bearing exercises,technique variations, and functional activities to more

Table 1Mean (±SD) patellofemoral force (N) values between lunge variations (forward lunge versus side lunge) and stride variations (with a stride versus withouta stride)

Knee angle for knee flexing phase Exercise variations Stride variations

Forward lunge Side lunge P-value With stride Without stride P-value

0� 69 ± 62 46 ± 46 0.115 74 ± 76 49 ± 39 0.35010� 159 ± 143 127 ± 149 0.483 220 ± 181 80 ± 53 0.009*

20� 207 ± 152 194 ± 183 0.550 301 ± 198 120 ± 63 0.002*

30� 356 ± 190 332 ± 245 0.999 459 ± 247 242 ± 109 0.002*

40� 628 ± 236 573 ± 328 0.390 732 ± 314 481 ± 180 <0.001*

50� 1059 ± 425 926 ± 408 0.127 1150 ± 485 856 ± 292 0.003*

60� 1524 ± 550 1533 ± 579 0.847 1683 ± 629 1378 ± 443 0.02670� 1944 ± 672 2009 ± 619 0.445 2058 ± 730 1896 ± 550 0.26380� 2161 ± 657 2493 ± 702 0.009* 2344 ± 742 2267 ± 645 0.38290� 2185 ± 654 2668 ± 788 <0.001* 2411 ± 748 2419 ± 774 0.669Knee angle for knee extending phase90� 2191 ± 662 2551 ± 783 <0.001* 2309 ± 686 2400 ± 788 0.86280� 2102 ± 739 2239 ± 743 0.181 2146 ± 776 2182 ± 713 0.23970� 1937 ± 786 1779 ± 675 0.482 1886 ± 852 1846 ± 621 0.07260� 1577 ± 706 1363 ± 561 0.271 1579 ± 772 1387 ± 503 0.02550� 1176 ± 525 999 ± 447 0.211 1217 ± 574 983 ± 386 0.008*

40� 780 ± 344 680 ± 339 0.206 833 ± 370 643 ± 293 0.006*

30� 504 ± 240 445 ± 252 0.250 567 ± 234 395 ± 228 0.002*

20� 312 ± 180 268 ± 169 0.170 383 ± 177 230 ± 150 0.009*

10� 168 ± 119 160 ± 111 0.841 228 ± 126 116 ± 75 0.0200� 66 ± 46 66 ± 82 0.999 73 ± 56 63 ± 69 0.743

* Significant difference (P < 0.01) between lunge variations or stride variations.


effectively make informed decisions regarding which exer-cise they choose to employ during patellofemoralrehabilitation.

There are some limitations in the current study. Firstly,MRI knee kinematic data have shown during the weightbearing squat that the femur moves and rotates underneatha relatively stationary patella, and if this femoral rotation isexcessive it may result in an increase in patellofemoralstress on the contralateral patellar facets (Doucette andChild, 1996; Li et al., 2004; Powers, 2003). This implies thatexcessive medial femoral rotation may place more stress onthe lateral patellar facets, while excessive lateral femoralrotation may place more stress on the medial patellar fac-ets. Unfortunately, collecting MRI knee kinematic datawhile performing the lunge is not currently possible dueto technology limitations, so it is unknown how much fem-oral rotation occurs during the lunge, how this rotationvaries among individuals (healthy and pathologic), and iffemoral rotation occurs in the lunge similar to how itoccurs in the squat.

Another limitation is the effect of Q-angle on patellofe-moral compressive force and stress. From cadaveric dataduring a simulated squat it was shown that an increasedQ-angle significantly caused a lateral shift and medial tiltand rotation of the patella, which may increase patellofe-moral stress (Mizuno et al., 2001). It is currently not feasi-ble to effectively measure lateral shift and medial tilt androtation of the patellar while performing the lunge. More-over, increased medial femoral rotation, which alsoincreases Q-angle, is also difficult to measure accuratelywhile performing the lunge.

There are also limitations in the biomechanical model.Firstly, muscle and patellofemoral forces were estimatedfrom modeling techniques and not measured directly,which is currently not possible in vivo. Secondly, patellofe-moral stress magnitudes were measured using patellar con-tact area values from MRI data from the literature andwere not measured directly. However, the contact areasused from the literature were determined during loadedweight bearing exercise in healthy male and female sub-jects, similar to the current study. Moreover, the near linearand direct relationship between contact area and kneeangle has been shown to be similar among studies (Besieret al., 2005; Cohen et al., 2001; Hinterwimmer et al.,2005; Patel et al., 2003; Salsich et al., 2003). Thirdly, thereare limitations regarding the magnitude of patellofemoralcontact areas (and concomitant stress magnitudes), inwhich the literature reports a wide array of values (Besieret al., 2005; Cohen et al., 2001; Hinterwimmer et al.,2005; Patel et al., 2003; Salsich et al., 2003). Differencesin patellar contact area magnitudes and concomitant patel-lofemoral stress magnitudes among weight bearing studiesare due to many factors, such as sex, mass, measuring tech-niques, and loading magnitudes. Nevertheless, althoughpatellofemoral stress magnitudes during weight bearingexercises or functional activities in the current study areapproximations only and not exact values, these varyingmagnitudes may be helpful to clinicians in deciding inter-ventions to employ during patellofemoral rehabilitation.Finally, the current study was limited to healthy subjectswho were able to perform the lunge in the sagittal planeof motion without transverse and frontal plane motions.


Future studies are needed during the lunge and otherweight bearing exercises to investigate the effects of trans-verse plane rotary motions and frontal plane valgus/varusmotions on patellofemoral force and stress magnitudes,which may occur with individuals with patellofemoralpathology.

5. Conclusions

Patellofemoral compressive force and stress magnitudeswere greater at higher knee angles and smaller at lowerknee angles, were greater during the side lunge comparedto the forward lunge between 80� and 90� knee angles,and were greater with a stride compared to without a stridebetween 10� and 50� knee angles. A more functional kneeflexion range between 0� and 50� may be appropriate dur-ing the early phases of patellofemoral rehabilitation due tolower patellofemoral compressive force and stress duringthis range compared to higher knee angles between 60�and 90�. Moreover, when the goal is to minimize patellofe-moral compressive force and stress, it may be prudent toemploy forward and side lunges without a stride comparedto with a stride. Understanding differences in patellofemor-al compressive force and stress among lunge variationshelp clinicians prescribe safer and more effective exerciseinterventions.

Acknowledgements

The authors would like to thank Lisa Bonacci, ToniBurnham, Juliann Busch, Kristen D’Anna, Pete Eliopou-los, & Ryan Mowbray for all their assistance during datacollection and analyses.

References

Basmajian, J.V., Blumenstein, R., 1980. Electrode Placement in EMGBiofeedback. In: Baltimore. Williams and Wilkins.

Besier, T.F., Draper, C.E., Gold, G.E., Beaupre, G.S., Delp, S.L., 2005.Patellofemoral joint contact area increases with knee flexion andweight-bearing. J. Orthop. Res. 23, 345–350.

Biedert, R.M., Sanchis-Alfonso, V., 2002. Sources of anterior knee pain.Clin. Sport. Med. 21, 335–347, vii.

Boling, M.C., Bolgla, L.A., Mattacola, C.G., Uhl, T.L., Hosey, R.G.,2006. Outcomes of a weight-bearing rehabilitation program forpatients diagnosed with patellofemoral pain syndrome. Arch. Phys.Med. Rehab. 87, 1428–1435.

Brechter, J.H., Powers, C.M., 2002. Patellofemoral joint stress during stairascent and descent in persons with and without patellofemoral pain.Gait Posture 16, 115–123.

Cholewicki, J., McGill, S.M., Norman, R.W., 1995. Comparison ofmuscle forces and joint load from an optimization and EMG assistedlumbar spine model: towards development of a hybrid approach. J.Biomech. 28, 321–331.

Cohen, Z.A., Henry, J.H., McCarthy, D.M., Mow, V.C., Ateshian, G.A.,2003. Computer simulations of patellofemoral joint surgery patient-specific models for tuberosity transfer. Am. J. Sport. Med. 31, 87–98.

Cohen, Z.A., Roglic, H., Grelsamer, R.P., Henry, J.H., Levine, W.N.,Mow, V.C., Ateshian, G.A., 2001. Patellofemoral stresses during openand closed kinetic chain exercises. An analysis using computersimulation. Am. J. Sport. Med. 29, 480–487.

Dehaven, K.E., Lintner, D.M., 1986. Athletic injuries: comparison by age,sport, and gender. Am. J. Sport. Med. 14, 218–224.

Devereaux, M.D., Lachmann, S.M., 1984. Patello-femoral arthralgia inathletes attending a sports medicine clinic. Br. J. Sport. Med. 18, 18–21.

Dixit, S., DiFiori, J.P., Burton, M., Mines, B., 2007. Management ofpatellofemoral pain syndrome. Am. Fam Physician 75, 194–202.

Doucette, S.A., Child, D.D., 1996. The effect of open and closed chainexercise and knee joint position on patellar tracking in lateral patellarcompression syndrome. J. Orthop. Sport. Phys. Ther. 23, 104–110.

Epstein, M., Herzog, W., 1998. Theoretical Models of Skeletal Muscle:Biological and Mathematical Considerations. John Wiley & Sons.,New York.

Escamilla, R.F., Fleisig, G.S., Zheng, N., Barrentine, S.W., Wilk, K.E.,Andrews, J.R., 1998. Biomechanics of the knee during closed kineticchain and open kinetic chain exercises. Med. Sci. Sport. Exer. 30, 556–569.

Escamilla, R.F., Fleisig, G.S., Zheng, N., Lander, J.E., Barrentine, S.W.,Andrews, J.R., Bergemann, B.W., Moorman 3rd, C.T., 2001. Effects oftechnique variations on knee biomechanics during the squat and legpress. Med. Sci. Sport. Exer. 33, 1552–1566.

Fredericson, M., Yoon, K., 2006. Physical examination and patellofe-moral pain syndrome. Am. J. Phys. Med. Rehab. 85, 234–243.

Heino Brechter, J., Powers, C.M., 2002. Patellofemoral stress duringwalking in persons with and without patellofemoral pain. Med. Sci.Sport. Exer. 34, 1582–1593.

Heintjes, E., Berger, M.Y., Bierma-Zeinstra, S.M., Bernsen, R.M.,Verhaar, J.A., Koes, B.W., 2003. Exercise therapy for patellofemoralpain syndrome. Cochrane Database Syst. Rev. CD003472.

Herzog, W., Read, L.J., 1993. Lines of action and moment arms of themajor force-carrying structures crossing the human knee joint. J. Anat.182, 213–230.

Hinterwimmer, S., Gotthardt, M., von Eisenhart-Rothe, R., Sauerland, S.,Siebert, M., Vogl, T., Eckstein, F., Graichen, H., 2005. In vivo contactareas of the knee in patients with patellar subluxation. J. Biomech. 38,2095–2101.

LaBella, C., 2004. Patellofemoral pain syndrome: evaluation and treat-ment. Prim Care 31, 977–1003.

Li, G., DeFrate, L.E., Zayontz, S., Park, S.E., Gill, T.J., 2004. The effectof tibiofemoral joint kinematics on patellofemoral contact pressuresunder simulated muscle loads. J. Orthop. Res. 22, 801–806.

Mizuno, Y., Kumagai, M., Mattessich, S.M., Elias, J.J., Ramrattan, N.,Cosgarea, A.J., Chao, E.Y., 2001. Q-angle influences tibiofemoral andpatellofemoral kinematics. J. Orthop. Res. 19, 834–840.

Narici, M.V., Landoni, L., Minetti, A.E., 1992. Assessment of humanknee extensor muscles stress from in vivo physiological cross-sectionalarea and strength measurements. Eur. J. Appl. Physiol. 65, 438–444.

Narici, M.V., Roi, G.S., Landoni, L., 1988. Force of knee extensor andflexor muscles and cross-sectional area determined by nuclear magneticresonance imaging. Eur. J. Appl. Physiol. 57, 39–44.

Natri, A., Kannus, P., Jarvinen, M., 1998. Which factors predict the long-term outcome in chronic patellofemoral pain syndrome? A 7-yrprospective follow-up study. Med. Sci. Sport. Exer. 30, 1572–1577.

Patel, V.V., Hall, K., Ries, M., Lindsey, C., Ozhinsky, E., Lu, Y.,Majumdar, S., 2003. Magnetic resonance imaging of patellofemoralkinematics with weight-bearing. J. Bone Joint Surg. Am., 2419–2424.

Powers, C.M., 2003. The influence of altered lower-extremity kinematicson patellofemoral joint dysfunction: a theoretical perspective. J.Orthop. Sport. Phys Ther. 33, 639–646.

Salem, G.J., Powers, C.M., 2001. Patellofemoral joint kinetics duringsquatting in collegiate women athletes. Clin. Biomech. 16, 424–430.

Salsich, G.B., Ward, S.R., Terk, M.R., Powers, C.M., 2003. In vivoassessment of patellofemoral joint contact area in individuals who arepain free. Clin. Orthop. Relat. Res., 277–284.

Shapiro, R., 1978. Direct linear transformation method for three-dimensional cinematography. Res. Q. 49, 197–205.

Steinkamp, L.A., Dillingham, M.F., Markel, M.D., Hill, J.A., Kaufman,K.R., 1993. Biomechanical considerations in patellofemoral jointrehabilitation. Am. J. Sport. Med. 21, 438–444.


van Eijden, T.M., Kouwenhoven, E., Verburg, J., Weijs, W.A., 1986. Amathematical model of the patellofemoral joint. J. Biomech. 19, 219–229.

van Eijden, T.M., Kouwenhoven, E., Weijs, W.A., 1987. Mechanics of thepatellar articulation Effects of patellar ligament length studied with amathematical model.. Acta Orthop. Scand. 58, 560–566.

van Linschoten, R., van Middelkoop, M., Berger, M.Y., Heintjes, E.M.,Koopmanschap, M.A., Verhaar, J.A., Koes, B.W., Bierma-Zeinstra,S.M., 2006. The PEX study - Exercise therapy for patellofemoral painsyndrome: design of a randomized clinical trial in general practice andsports medicine [ISRCTN83938749]. BMC Musculoskelet Disord. 7,31.

Wallace, D.A., Salem, G.J., Salinas, R., Powers, C.M., 2002. Patellofe-moral joint kinetics while squatting with and without an external load.J. Orthop. Sport. Phys. Ther. 32, 141–148.

Wickiewicz, T.L., Roy, R.R., Powell, P.L., Edgerton, V.R., 1983. Musclearchitecture of the human lower limb. Clin. Orthop., 275–283.

Wickiewicz, T.L., Roy, R.R., Powell, P.L., Perrine, J.J., Edgerton, V.R.,1984. Muscle architecture and force-velocity relationships in humans.J. Appl. Physiol. 57, 435–443.

Wilk, K.E., Davies, G.J., Mangine, R.E., Malone, T.R., 1998. Patellofe-moral disorders: a classification system and clinical guidelines fornonoperative rehabilitation. J. Orthop. Sport. Phys. Ther. 28, 307–322.

Witvrouw, E., Danneels, L., Van Tiggelen, D., Willems, T.M., Cambier,D., 2004. Open versus closed kinetic chain exercises in patellofemoralpain: a 5-year prospective randomized study. Am. J. Sport. Med. 32,1122–1130.

Witvrouw, E., Lysens, R., Bellemans, J., Peers, K., Vanderstraeten, G.,2000. Open versus closed kinetic chain exercises for patellofemoralpain. A prospective, randomized study. Am. J. Sport. Med. 28, 687–694.

Wojtys, E.M., Beaman, D.N., Glover, R.A., Janda, D., 1990. Innervationof the human knee joint by substance-P fibers. Arthroscopy 6, 254–263.

Zajac, F.E., 1989. Muscle and tendon: properties, models, scaling, andapplication to biomechanics and motor control. Crit. Rev. Biomed.Eng. 17, 359–411.

Zheng, N., Fleisig, G.S., Escamilla, R.F., Barrentine, S.W., 1998. Ananalytical model of the knee for estimation of internal forces duringexercise. J. Biomech. 31, 963–967.

Documents

Patellofemoral Compressive Force and Stress During the Forward and Side Lunge With and Without a Stride