Patellofemoral Joint Force and Stress duringthe Wall Squat and One-Leg Squat

RAFAEL F. ESCAMILLA', NAIQUAN ZHENG 2, TORAN D. MACLEOD 3, W. BRENT EDWARDS 4 ,RODNEY IMAMURA5 , ALAN HRELJAC5 , GLENN S. FLEISIG6 , KEVIN E. WILK7 ,CLAUDE T. MOORMAN 1118, and JAMES R. ANDREWS 6.9

1Department of Physical Therapy, California State University, Sacramento, CA; 2Department of Mechanical Engineering andEngineering Science, The Center for Biomedical Engineering, University of North Carolina, Charlotte, NC; 3Department ofPhysical Theraphy, Center for Biomedical Engineering Research, University of Delaware, Newark, DE; Department ofKinesiology, Iowa State University, Ames, IA; 5Kinesiology and Health Science Department, California State University,Sacramento, CA; 6American Sports Medicine Institute, Birmingham, AL; 7Champion Sports Medicine, Birmingham, AL;8Duke Sports Medicine Center, Duke University Medical Center, Durham, NC; and 9Andrews-Paulos Research andEducation Institute, Andrews Institute, Gulf Breeze, FL

ABSTRACT

ESCAMILLA, R. F., N. ZHENG, T. D. MACLEOD, W. BRENT EDWARDS, R. IMAMURA, A. HRELJAC, G. S. FLEISIG, K. E.

WILK, C. T. MOORMAN, and J. R. ANDREWS. Patellofemoral Joint Force and Stress during the Wall Squat and One-Leg Squat.

Med. Sci. Sports Exerc., Vol. 41, No. 4, pp. 879-888, 2009. Purpose: To compare patellofemoral compressive force and stress duringthe one-leg squat and two variations of the wall squat. Methods: Eighteen subjects used their 12 repetition maximum (12 RM) weightwhile performing the wall squat with the feet closer to the wall (wall squat short), the wall squat with the feet farther away from the wall(wall squat long), and the one-leg squat. EMG, force platform, and kinematic variables were input into a biomechanical model to

calculate patellofemoral compressive force and stress as a function of knee angle. To asses differences among exercises, a one-factorrepeated-measure ANOVA (P = 0.0025) was used. Results: During the squat ascent, there were significant differences in patellofemoral

force and stress among the three squat exercises at 900 knee angle (P = 0.002), 800 knee angle (P = 0.002). 70' knee angle (P <0.001), and 600 knee angle (P = 0.001). Patellofemoral force and stress were significantly greater at 90' knee angle in the wall squatshort compared with wall squat long and one-leg squat, significantly greater at 700 and 800 knee angles in the wall squat short and longcompared with the one-leg squat and significantly greater at 600 knee angle in the wall squat long compared with the wall squat shortand one-leg squat. Conclusions: Except at 600 and 900 knee angles, patellofemoral compressive force and stress were similar betweenthe wall squat short and the wall squat long. Between 600 and 900 knee angles, wall squat exercises generally produced greaterpatellofemoral compressive force and stress compared with the one-leg squat. When the goal is to minimize patellofemoral compressiveforce and stress, it may be prudent to use a smaller knee angle range between 00 and 500 compared with a larger knee angle rangebetween 600 and 90'. Key Words: BIOMECHANICS, KINETICS, CLOSED CHAIN EXERCISES, KNEE

In the outpatient setting patellofemoral pain- syndrome.(PFPS) is the most common type of knee pain andaccounts for 25-30% of all knee pathologies treated

(10,11,16). Because the etiology of PFPS is poorly under-stood and multifaceted, it remains one of the most difficultclinical challenges in rehabilitative medicine (39). PFPSprimarily affects younger active individuals approximately18-40 yr old (although older individuals can also beaffected), athletes and nonathletes (11,16,21), and males

Address for correspondence: Rafael F. Escamilla, Ph.D., P.T., C.S.C.S.,F.A.C.S.M., Professor, Department of Physical Therapy, California StateUniversity, 6000 J Street, Sacramento, CA 95819-6020; E-mail: rescamil@csus.edu.Submitted for publication July 2007.Accepted for publication September 2008.

0195-9131/09/4104-0879/0MEDICINE & SCIENCE IN SPORTS & EXERCISE®Copyright © 2009 by the American College of Sports Medicine

DOI: 10.1249/MSS.0b0I3e31818e7ead

and females (9). Although patellofemoral rehabilitation canbe a long and arduous process, the use of appropriateexercises can improve this process by decreasing rehabili-tation time and improving function (4,18,26,40,41).

High patellofemoral joint compressive force (patellofe-moral force) can result in PFPS from numerous soft tissues,such as synovial plicae, infrapatellar fat pad, retinacula,joint capsule, and patellofemoral ligaments (3). Patellofe-moral force can also elevate subchondral bone stress(patellofemoral force per unit patella contact area) in thepatellofemoral joint (2). Because the subchondral bone plateis rich in pain receptors (42), increased subchondral bonestress from high patellofemoral force may also result inPFPS (3). Patellofemoral joint stress can result in a cartilagedegeneration and a decrease in the ability of the cartilage todistribute patellofemoral force (2). Therefore, understandingwhat patellofemoral force and stress magnitudes are gener-ated among patellofemoral rehabilitation exercises may behelpful to clinicians when prescribing therapeutic exercisesto individuals with PFPS.

879

Weight-bearing exercises, such as the squat, are frequentlyused during patellofemoral rehabilitation and are specificto many functional activities such as walking, running,and jumping (4,18,26,40,41). The use of weight-bearingexercises have been shown to be effective, both in short- andlong-term outcomes, in decreasing PFPS and in enhancingfunctional performance (4,18,26,40,41). Therapists usethese types of exercises to minimize PFPS and muscle loss,to strengthen hip and thigh musculature, to enhance balanceand stability, and to minimize the risk of future injuries andassociated costs of health care (35). However, all weight-bearing exercises may not produce similar magnitudes ofpatellofemoral force and stress. Moreover, using varyingtechniques within a weight-bearing exercise may also affectpatellofemoral force and stress.

Wall squats and one-leg squats are common weight-bearing exercises used by athletes and other individualswith healthy knees to train the hip and the thighmusculature. Therapists and trainers also use wall squats,one-leg squats, and other similar weight-bearing exercisesduring patellofemoral rehabilitation for PFPS patients toallow patients to recover faster and return to function earlier(4,18,26,40,41). Wall squats can be performed with varyingtechniques, such as positioning the heels farther from orcloser to the wall. Positioning the heels farther from the walltypically results in the knees being maintained over the feet atthe lowest position of the squat, while positioning the heelscloser to the wall typically results in anterior knee translationbeyond the toes at the lowest posidion of the squat.Performing a one-leg squat also causes the stance knee totranslate forward beyond the toes at the lowest position of thesquat. Clinicians and trainers often believe that anteriortranslation of the lead knee beyond the toes during squattingtype exercises increases patellofemoral force and stress, butthere is currently no evidence to support this belief.

Understanding how patellofemoral force and stress varyamong weight-bearing exercises will allow clinicians andtrainers to prescribe safer and more effective knee rehabil-itation treatment to patients with PFPS or to athletes duringtraining. For example, if performingthe wall squat with theheels closer to the wall (causing greater anterior translation ofthe lead knee over the toes) results in greater patellofemoralforce and stress compared with performing the wall squatwith the heels farther away from the wall (causing the leadknee to be maintained over the foot), a wall squat with theheels closer to the wall may be-discouraged during trainingand rehabilitation. There may also be differences in patello-femoral force and stress over a specific knee flexion range ofmotion between wall squat and one-leg squat exercises.Excess patellofemoral force and stress over time may lead toPFPS in individuals with asymptomatic patellofemoral jointsor may exacerbate PFPS in patients with patellofemoralpathology and retard the rehabilitation process.

Currently, patellofemoral force and stress magnitudesduring the wall squat or one-leg squat are unknown.Therefore, the purpose of this study was to compare

patellofemoral force and stress during the wall squat withthe feet farther away from the wall (wall squat long), thewall squat with the feet closer to the wall (wall squat short),and the one-leg squat. It was hypothesized that patellofe-moral force and stress would be greater in the one-leg squatcompared with the wall squat long, similar between the one-.leg squat and the wall squat short and greater in the wallsquat short compared with the wall squat long.

METHODS

Subjects

Eighteen healthy individuals (nine males and ninefemales) without a history of. patellofemoral pathologyparticipated, with an average age, mass, and height of29 + 7 yr, 77 + 9 kg, and 177 ± 6 cm, respectively, for malesand 25 ± 2 yr, 60 + 4 kg, and 164 + 6 cm, respectively, forfemales. All subjects were required to perform the wall squatand the one-leg squat exercises pain-free and with properform and technique for 12 consecutive repetitions'using their12 repetition maximum (12 RM) weight.

To control the EMG signal quality, the current study waslimited to males and females that had average or belowaverage body fat, which was assessed by Baseline skinfoldcalipers (Model 68900; Country Technology, Inc., GaysMill, WI), and appropriate regression equations and body fatstandards set by the American College of Sports Medicine.Average body fat was 12% ± 4% for males and 18% ± 1% forfemales. All subjects provided written informed consent inaccordance with the Institutional Review Board at CaliforniaState University, Sacramento, which approved the researchconducted and informed consent form.

Exercise Description

Wall squat (Fig. 1A and B). The wall squat beganwith the right foot on an AMTI force platform (Model

FIGURE 1-Wall ýquat with feet farther from wall (wall squat long;A); wall squat with feet closer to wall (wall squat short; B).

880 Official Journal of the American College of Sports Medicine http://www.acsm-msse.org

OR6-6-2000; Advanced Mechanical Technologies, Inc.) andtheir left foot on the ground, both knees fully extended (0'knee angle), the back flat against the wall, and a dumbbellweight held in both hands with the arms straight and at thesubject's side. From this position, the subject slowly flexedboth knees and squatted down until the thighs were ap-proximately parallel with the ground (resulting in approxi-mately 90-1000 of knee flexion in the wall squat long andapproximately 100-1 100 of knee flexion in the wall squatshort), and in a continuous motion, the subject returned backto the starting position. A metronome was used to helpensure that the knees flexed and extended at approximately45's-I. The surface of the wall was smooth, and a towelwas positioned between the wall and the subject to min-imize friction as the subject slid down and up the wall. Thestance width (distance between inside heels) was 32 ± 6 cmfor males and 28 + 7 cm for females, and the foot angle wasapproximately 0' (feet pointing approximately straightahead), and both stance and foot angle were according tosubject preference.

The wall squat was performed with two techniquevariations, wall squat long (Fig. IA) and wall squat short(Fig. I B). The foot position relative to the wall for the wallsquat long was determined using a heel-to-wall distance thatresulted in the legs being approximately vertical at thelowest position of the squat (Fig. IA), with the knees abovethe ankles, which is commonly recommended by cliniciansand trainers. The average heel-to-wall distance for the wallsquat long was 45 ± 3 cm for males and 41 + 3 cm forfemales. The heel-to-wall distance for the wall squat shortwas one half the distance of the heel-to-wall distance for thewall squat long. This distance was chosen because theshorter heel-to-wall distance for the wall squat shortresulted in the anterior surface of the knee translatingbeyond the distal end of the toes at the lowest position ofthe wall squat short (Fig. 1B), which is typically discour-aged by clinicians and trainers.

One-leg squat. The one-leg squat started with thesubject standing on one leg with the right foot on the AMTIforce platform, the right knee fully extended, the left kneebent approximately 90', and a single dumbbell weight heldwith both hands in front of the chest (subject preference).From this position, the subject slowly flexed the right knee andsquatted down until the right thigh was approximately parallelwith the ground (resulting in approximately 100-110' of kneeflexion) with the trunk tilted forward approximately 30-40'(Fig. 2). In a continuous motion, the subject returned back tothe starting position. A metronome was used to help ensurethat the right knee flexed and extended at approximately45's-1. Like the wall squat short, at the lowest position ofthe one-leg squat, the distal surface of the knee translatedbeyond the distal end of the toes (Fig. 2).

Data Collection

Each subject came in for a pretest I wk before the testingsession. The experimental protocol was reviewed, the

FIGURE 2-One-leg squat.

subject was given the opportunity to practice the one-legsquat and wall squat exercises, and each subject's heel-to-wall distances for the wall squat short and long weredetermined. In addition, to normalize intensity between thewall squat and the one-leg squat exercises, each subject's 12RM was determined. To determine the weight used for thewall squat short and long, each subject used their 12 RMweight while performing the wall squat using a heel-to-walldistance that was halfway between the heel-to-wall dis-tances for the wall squat short and wall squat long, and thisweight was used for the wall squat short and wall squat longduring the testing session. The mean total dumbbell massused was 56 + 9 kg for males and 36 + 8 kg for females forthe wall squat short and wall squat long and 15 ± 3 kg formales and 10 + 3 kg for females for the one-leg squat.

Blue Sensor (Ambu Inc., Linthicum, MD) disposablesurface electrodes (type M-00-S) were used to collect EMGdata. These oval-shaped electrodes (22 mm wide and 30mm long) were placed in a bipolar electrode configurationalong the longitudinal axis of each muscle, with a center-to-center distance of approximately 3 cm between electrodes.Before positioning the electrodes over each muscle, the skinwas prepared by shaving, abrading, and cleaning withisopropyl alcohol wipes to reduce skin impedance. Aspreviously described (1), electrode pairs were then placedon the subject's right side for the following muscles: a)rectus femoris; b) vastus lateralis; c) vastus medialis; d)medial hamstrings (semimembranosus and semitendinosus);e) lateral hamstrings (biceps femoris); and f) gastrocnemius(midpoint between medial and lateral heads).

Spheres (3.8 cm in diameter) covered with 3MrM reflec-tive tape were attached to adhesives and positioned overthe following bony landmarks: a) third metatarsal head of theright foot ; b) medial and lateral malleoli of the fight leg; c)upper edges of the medial and lateral tibial plateaus of the

Medicine & Science in Sports & Exercise® 881WALL SQUAT AND ONE-LEG SQUAT

right knee; d) posterosuperior greater trochanters of theleft and right femurs; and e) lateral acromion of the rightshoulder.

Once the electrodes and the spheres were positioned, thesubject warmed up and practiced the exercises as needed, anddata collection commenced. A six-camera Peak Performancemotion analysis system (Vicon-Peak Performance Technol-ogies, Inc., Englewood, CO) was used to collect 60-Hz videodata. Force data were collected at 960 Hz using an AMTIforce platform (Model OR6-6-2000, Advanced MechanicalTechnologies, Inc.). EMG data were collected at 960 Hzusing a Noraxon Myosystem unit (Noraxon USA, Inc.,Scottsdale, AZ). The EMG amplifier bandwidth frequencywas 10-500 Hz, with an input impedance of 20,000 kfl,and the common-mode rejection ratio was 130 dB. Video,EMG, and force data were electronically synchronized andcollected as each subject performed in a randomizedmanner one set of three continuous repetitions (trials)during the wall squat short, wall squat long, and one-legsquat.

Subsequent to completing all exercise trials, EMG datawere collected during maximum voluntary isometric con-tractions (MVIC) to normalize the EMG data collectedduring each exercise (14). The MVIC for the rectus femoris,vastus lateralis, and vastus medialis were collected in aseated position at 900 knee and hip flexion with a maximumeffort knee extension (14). The MVIC for the lateral and themedial hamstrings were collected in a seated position at 90'knee and hip flexion with a maximum effort knee flexion(14). MVIC for the gastrocnemius was collected during amaximum effort standing one-leg toe raise with the anklepositioned approximately halfway between neutral and fullplantar flexion (14). Two 5-s trials were randomly collectedfor each MVIC.

Data Reduction

Video images for each reflective marker were tracked anddigitized in three-dimensional space with Peak Performancesoftware, using the direct linear transformation calibrationmethod (31). Testing of the accuracy of the calibrationsystem resulted in reflective balls that could be located inthree dimensional space with an error less than 4-7 mm.The raw position data were smoothed with a double-passfourth-order Butterworth low-pass filter with a cutofffrequency of 6 Hz (14). Joint angles, linear and angularvelocities, and linear and angular accelerations werecalculated in a two-dimensional sagittal plane of the kneeusing appropriate kinematic equations (14).

Raw EMG signals were full-waved rectified, smoothedwith a 10-ms moving average window, and linear envel-oped 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 data

for the three repetitions (trials) were then averaged atcorresponding. knee angles between 00 and 90' and wereused in the biomechanical model described below.

Biomechanical Model

As previously described (14,44), a biomechanical modelof the knee was used to continuously calculate patellofe-moral forces throughout a 900 knee range of motion duringthe knee flexing (descent) phase (0-90') and knee extend-ing (ascent) phase (90-0') of the wall squat and one-legsquat (Fig. 3). Resultant force and torque equilibriumequations were calculated using inverse dynamics and thebiomechanical knee model (14,44). Moment arms formuscle forces lines of action angles for muscles wererepresented as polynomial functions of the knee flexionangle using data from Herzog and Read (19).

Quadriceps, hamstrings, and gastrocnemius muscle forceswere calculated as previously described (14,44). Becausethe accuracy of calculating muscle forces depends onaccurate calculations of. a muscle's physiological cross-sectional area (PCSA), a maximum voluntary contractionforce per unit PCSA, and the EMG-force relationship,resultant force and torque equilibrium equations may not besatisfied. Therefore, each muscle force F,,(i) was modifiedby the following equation:

F.o(i) = cik1 kviAio-.i[EMGi/MVICi],

where Ai was the PCSA of the ith muscle, om(i) was theMVIC force per unit PCSA of the ith muscle, EMGiand MVICQ were the EMG window averages of the ithmuscle EMG during exercise and MVIC trials, ci was aweight factor (values given below) adjusted in a computer

FIGURE 3-Computer optimization with input from measured kneetorque from inverse dynamics and predicted knee torque from musclemodel, where TK = resultant knee torque, FK resultant knee force,I = moment of inertia about leg center of mass, a = angular accelerationof leg, m = mass of leg, a = linear acceleration of leg, g = gravitationconstant 9.80 m.s-2, F,.t = external force acting on foot, Text = externaltorque acting on foot, FQ = quadriceps force, Fp = patellar tendonforce, FPF = patellofemoral compressive force, FH = hamstrings force,and FG = gastrocnemius force. Note: to simplify the drawing, the equaland opposite forces and torques acting on the distal leg and proximalankle are not shown.

882 Official Journal of the American College of Sports Medicine hfttp://www.acsm-msse.org

optimization program to minimize the difference betweenthe resultant torque from the inverse dynamics (T,,) andthe resultant torque calculation from the biomechanicalmodel (Tmi), k1i represented each muscle's force-lengthrelationship as function of hip and knee angles (based onmuscle length, fiber length, sarcomere length, pennationangle, and cross-sectional area) (37), and k,,i representedeach muscle's force-velocity relationship based on a Hill-type model for eccentric and concentric muscle actionsusing the following equations from Zajac (43) and Epsteinand Herzog (13):

k,, (b- (a/Fo)v)/(b + v) concentrick, = C- (C -1)(b+ (a/Fo)v)/(b - v) eccentric

with F0 representing isometric muscle force, v = velocity,a = 0.32Fo, b = 3.20 s-1, and C = 1.8. Muscle force fromeccentric contractions was scaled up by 1.8 times theisometric muscle force Fo. Forces generated by the kneeflexors and extensors at MVIC were assumed to be linearlyproportional to their PCSA. Muscle force per unit PCSA atMVIC was 35 N.cm- 2 for the knee flexors and 40 N.cm-2

for the quadriceps (6,24,25,38).

The objective function used to determine each ithmuscle's coefficient ci was as follows:

min f(ci) = I(I cj)2 +k (Tre, - ' T,)',

subject to Clow < C _< Chi,h, where clo, and Chigh were lowerand upper limits for ci, and A was -a constant. The weightfactor c was to adjust the final muscle force calculation. Thebounds on c were set between 0.5 and 1.5. The torquespredicted by the EMG-driven model matched well (<2%)with the torques generated from the inverse dynamics. Theassumptions associated with this model are 1) that the torquefrom cruciate ligament forces was ignored and 2) that otherforces and torques out of the sagittal plane were ignored.

Patellofemoral force was a function of patellar tendonforce and quadriceps tendon force. Patellar tendon force wascalculated by the quadriceps tendon force and the ratio of thepatellar tendon force and the quadriceps tendon force, aspreviously described (33,34). The angles between thepatellar tendon, quadriceps tendon, and patellofemoral jointwere expressed as functions of knee angle (33,34).

Patellofemoral stress, which was calculated every 10'between 0' and 900 knee angles, was expressed as the ratio ofpatellofemoral force, calculated from the biomechanicalmodel described above (14,44), and the patellar contactarea. Patellar contact areas were determined at 100 intervalsbetween 00 and 90' knee angles. Contact areas from in vivoMRI data from Salsich et al. (30), who also used both maleand female subjects with healthy knees and had themperform weight-bearing exercise using resistance, were usedat 00 (146 mm2), 200 (184 mm 2), 400 (290 nIM2), and 60'(347 mm?) knee angles. These four contact area values

formed a near linear relationship as a function of knee angle,resulting in a line of best fit equation of y = 3.55x + 135(r = 0.98), with y = contact area and x = knee angle. Thisline of best fit equation was used to determine contact areasat 100 knee angle (171 mm 2), 300 knee angle (242 mm2),and 50' knee angle (313 mm2). The contact areas above at40', 500, and 600 knee angles were used to develop the lineof best fit equation, y = 2.8 lx + 176 (r = 0.99), which wasused to determine contact areas at 70' knee angle(373 mm 2), 800 knee angle (401 mm 2), and 900 knee angle(429 mm2). Like the current study, a near linear relationshipbetween patellar contact area and knee angles has beenreported between 0' and 900 knee angles in several studiesinvolving weight-bearing exercises (2,8,20,27,30).

Data Analysis

To determine significant differences among the wallsquat long, wall squat short, and one-leg squat, patellofe-moral force and stress were analyzed every 100 during the0-900 descent phase and the 90-00 ascent phase using aone-factor repeated-measure ANOVA. Bonferroni t-testswere used to assess pairwise comparisons. To minimize theprobability of type I errors secondary to the use of aseparate ANOVA for each knee angle, a Bonferroniadjustment was performed with the level of significanceestablished at 0.0025 (0.05/20 knee angles). A separate setof analyses was not performed for patellofemoral jointstress values because stress values for each knee angle werederived from dividing force data by a constant, therefore notaffecting statistical results.

RESULTS

Descriptive patellofemoral force and stress data duringthe wall squat and the one-leg squat are shown in Figures 4and 5. Visual observation of the data indicates thatpatellofemoral force and stress progressively increasedduring the squat descent and progressively decreased duringthe squat ascent, except between 900 and 700 during thesquat ascent in which patellofemoral force and stressprogressively increased. During the squat descent, therewere no significant differences in patellofemoral force andstress among the three squat exercises. During the squatascent, there were significant differences in patellofemoralforce and stress among the three squat exercises at 900 kneeangle (P = 0.002), 80' knee angle (P = 0.002), 700 kneeangle (P < 0.001), and 600 knee angle (P = 0.001).Patellofemoral force and stress were significantly greater at900 knee angle in the wall squat short compared with wallsquat long and one-leg squat, significantly greater at 700and 800 knee angles in the wall squat short and longcompared with the one-leg squat, and significantly greaterat 600 knee angle in the wall squat long compared with thewall squat short and one-leg squat. At the lowest position ofthe wall squat short, the knees translated beyond the toes

Medicine & Science in Sports & Exercise® 883WALL SQUAT AND ONE-LEG SQUAT

4000

S3000

2000

S1000

00 20 40 60 80 100 80 60 4(

Descent AscentKnee Flexion Angle (deg)

SWall

Squat Short- Wall Squat Long- One Leg Squat

FIGURE 4-Mean (SD) patellofemoral joint compressive force for one-leg squat and wall squat exercises.

9 + 2 cm, whereas at the lowest position of the one-legsquat the knee translated beyond the toes 10 + 2 cm.

DISCUSSIONAs hypothesized, patellofemoral force and stress were

greater during the wall squat short compared with the wallsquat long, but only at 90' knee angle during the squatascent. At 90' knee angle, the knees translated beyond thetoes in the wall squat short, but the knees remained over thefeet in the wall squat long. Also, as the knees translatedforward beyond the toes in the wall squat short, theorientation of the leg tilted forward (Fig. 1B), changingthe direction of the patellar tendon force, which potentiallymay increase patellofemoral force compared with thevertical leg position in the wall squat long at 90' kneeangle (Fig. IA). Therefore, anterior knee translation andforward tilt of the leg may be related to increasedpatellofemoral force and stress. The results of the current

12

10

8-g" g

S60

'ý 40

S2 T

00 20 40 60 80

Descent

Knee- Wall Squat Short

- Wall Squat Long

- One Leg Squat

study support the belief of many clinicians that anteriorknee translation beyond the toes while performing squattingtype exercises increases patellofemoral force and stresscompared with maintaining the knees over the feet.

The wall squat short and the one-leg squat both resultedin similar amounts of anterior knee translation at maximumknee flexion, but patellofemoral force and stress weresignificantly lower in the one-leg squat compared with thewall squat short between 90' and 70' knee angles duringthe squat ascent (Figs. 4 and 5). The primary cause of thegreater patellofemoral force and stress between 90' and 70'knee angles in the wall squat short compared with the one-leg squat was greater quadriceps force during the wall squatshort. Between 90' and 70' knee angles during the squatascent, the estimated quadriceps forces that were calculatedin the current study using our EMG-driven knee modelwere approximately 30-40% greater in the wall squat shortcompared with the one-leg squat. In contrast, between 900and 70' knee angles during the squat ascent, the estimated

100 80 60 40Ascent

0

Flexion Angle (deg)

FIGURE 5-Mean (SD) patellofemoral joint stress for one-leg squat and wall squat exercises.

884 Official Journal of the American College of Sports Medicine hftp://www.acsm-msse.orq

hamstring forces that were calculated were approximately

60-70% greater in the one-leg squat compared with the wall

squat short. One reason for greater quadriceps force and lesshamstrings force in the wall squat short compared with theone-leg squat is because the trunk is erect in the wall squat

short (Fig. LA) and tilted forward 30-40' in the one-leg

squat (Fig. 2). The erect trunk in the wall squat shortproduced a line of force from the center of the mass of thelifter-dumbbell system (lifter's mass plus dumbbell mass)that resulted in a relatively small hip moment arm and hip

extensor muscle moment and a relatively large kneemoment arm and knee extensor muscle moment (Fig. I B).In contrast, the forward trunk tilt in the one-leg squatproduced a line of force from the center of the mass of thelifter-dumbbell system that resulted in a relatively large hip

moment arm and hip extensor muscle moment and arelatively small knee moment arm and knee extensormuscle moment (Fig. 2).

Although friction was minimized during the wall squatby using a smooth wall and a towel between the subject andthe wall, the normal force that the wall applied to thesubject's back during the wall squat exercises resulted in anincreased friction force on the subject as they slid down andup the wall. This friction force acted opposite the force ofgravity during the squat descent but acted in the same

direction as the force of gravity during the squat ascent.Therefore, the friction force made it easier for the subject tocontrol the rate of sliding down the wall by producing aknee extensor torque but made it more difficult for thesubject to slide up the wall by producing a knee flexortorque. Because the one-leg squat did not have a friction

force compare to the wall squat, this provides one plausibleexplanation why quadriceps force and patellofemoral forceand stress were greater in the ascent phase of the wall squat

exercises compared with the one-leg squat.If acceleration, whose magnitudes were very small, was

discounted during the wall squat and a static analysis wasperformed at a point on the feet where the ground reactionforce acted, gravity acting on the center of mass of the lifter-dumbbell system would produce a torque about the feet thatmust be countered by an equal and opposite torque generatedby the normal force and static friction force that the wallexerts on the lifter's back (Fig. IA and B). Because duringthe wall squat long the heels were twice as far from the wallcompared with .the wall squat short (Fig. IA and B), thenormal force must be greater in the wall squat longcompared with the wall squat short. Because friction forceis directly proportional to the normal force, the downward-acting friction force on the subject while sliding up the wallwas greater in the wall squat long compared with the wall

squat short, making it more difficult for a subject to slide upthe wall during the wall squat long. However, because thenormal force generates a knee extensor torque duringthe wall squat exercises, the greater normal force during

the wall squat long compared with the wall squat short may

have made it easier for the subject to slide up the wall

during the wall squat long. Therefore, compared with thewall squat short, during the wall squat long, the greaterfriction force made it more difficult to slide up the wall butthe greater normal force made it easier to slide up the wall.The varying and apposing actions of the normal force andfriction force during the wall squat long and the wall squat

short may help explain why the quadriceps force andresulting patellofemoral force and stress were generallysimilar between these two exercises, with the only excep-tions at 600 and 90' knee angles during the squat ascent. Itis unclear why patellofemoral force and stress weresignificantly greater in the wall squat long at 600 kneeangle and significantly greater in the wall squat short at90' knee angle.

The wall squat short and long as defined in the currentstudy may represent two extremes in heel-to-wall distancesthat can be used while performing wall squat exercises. It is

unlikely that the heel-to-wall distances used in the currentstudy would ever be greater when performing the wall squatlong or less when performing the wall squat short. It is

also possible that patellofemoral force and stress may be

different than the results of the current study if the wallsquat were performed with heel-to-wall distances some-where between those used for the wall squat short and long,and this should be the focus of additional studies. Theremay be an optimal heel-to-wall distance that minimizespatellofemoral force and stress.

Another consideration during patellofemoral rehabilita-tion is what knee flexion range of motion to use whileperforming squat exercises. Because patellofemoral forceand stress generally increased with greater knee angles anddecreased with smaller knee angles, a more functional kneeflexion range between 00 and 500 may be more appropriatefor patellofemoral patients compared with higher kneeangles between 600 and 900. For example, during the squatascent phase of the wall squat short, patellofemoral forceranged from approximately 75 to 1400 N between 00 and500 knee angle and from approximately 2100 to 3650 Nbetween 60' and 90' knee angles, and patellofemoral stressranged from approximately 0.5 to 4.4 MPa between 0' and500 knee angles and from approximately 5.9 to 8.9 MPa

between 600 and 900 knee angles. This same pattern ofincreased patellofemoral force and stress with larger kneeangles has been reported during the barbell squat and legpress (14,15,29,32,36). These authors reported that patello-

femoral force and stress progressively increased from 00 toapproximately 900, peaking at approximately 90', and thenprogressively decreasing from approximately 90' to 00.Computer optimization techniques demonstrated similarresults during a simulated squat (7).

Peak patellofemoral force and stress magnitudes from thecurrent study are less than some weight-bearing exercises,such as the barbell squat and the leg press (14), but morethan some weight-bearing functional activities, such aswalking (17) and going up and down the stairs (5).Escamilla et al. (14) reported'peak patellofemoral force

Medicine & Science in Sports & Exercisee 885WALL SQUAT AND ONE-LEG SQUAT

magnitudes between 4500 and 4700 N at.900 knee angleduring the 12 RM barbell squat and leg press using healthysubjects, resulting in patellofemoral stress magnitudesbetween 11 and 12 MPa. Having healthy subjects squatwith a barbell using a 35% bodyweight load, Wallace et al.(36) reported peak patellofemoral force magnitudes near2500 N and patellofemoral stress magnitudes near 13MPa, both occurring at 90' knee angle. Peak force and stressmagnitudes during the barbell squat and leg press areapproximately 25-50% greater compared with peak forceand stress magnitudes in the current study, which alsooccurred at 90' knee angle. In contrast, peak patellofemoralforce and stress in healthy subjects during fast walkingreportedly are approximately 900 N and 3.13 MPa, respec-tively (17), which are approximately two to three timeslower than the peak force and stress magnitudes in thecurrent study. However, peak patellofemoral force andstress magnitudes in healthy subjects going up and downthe stairs reportedly are approximately 2500 N and 7 MPa,respectively (17), which are only 15-30% less comparedwith the peak force and stress magnitudes in the currentstudy. Understanding patellofemoral force and stress mag-nitudes among varying resistance exercises and functionalactivities is helpful to clinicians and trainers when decidingwhich interventions to use for patients with PFPS.

Unlike healthy subjects, patients with PFPS exhibitsmaller patellar contact areas and greater patellofemoralstress during some weight-bearing functional activities(5,17). Compared with healthy individuals, patients withPFPS had 40% smaller patellar contact areas and 110%greater peak patellofemoral stress (6.61 MPa) during fastwalking (17). Moreover, Hinterwimmer et al. (20) reportedthat patients with patellar subluxation had 40-55% smallerpatellar contact areas compared with healthy individuals,which implies greater patellofemoral stress in these patientscompared with healthy individuals. These patellofemoralstress data involving patients with PFPS implies that PFPSpatients may be at higher risk of experiencing pain anddiscomfort while performing wall squat and one-leg squatexercises, especially at higher knee angles where patellofe-moral force and stress are greatest,. and this should be thefocus of future studies.

Unfortunately, it is currently unknown what patellofe-moral force or stress magnitudes and over what timeduration can ultimately lead to patellofemoral pathology.There are many factors that may contribute to patellofemoralpathology, such as 1) overuse or trauma; 2) imbalance ormalalignment of the extensor mechanism, which can lead tolateral patellar subluxation or tilt; 3) muscle weakness, suchas weak quadriceps and hip external rotators; 4) muscletightness, such as tight quadriceps, hamstrings, or iliotibialband; and 5) lower extremity malalignment, such as patellaalta, genu valgum, femoral neck anteversion, excessiveQ-angle, and excessive rearfoot pronation. It can only besurmised that relatively large patellofemoral force and stressmagnitudes over time may lead to patellofemoral pathology,

especially in individuals that exhibit some of the abovefactors and thus are predisposed to patellofe'Moral problems.Nevertheless, clinicians can use information regardingpatellofemoral force and stress magnitudes among differentweight-bearing exercises, technique variations, and function-al activities to be able to make more informed decisionsregarding which exercise they choose to use during patello-femoral rehabilitation.

Patellofemoral force and stress curves were similar inshape due to proportional increases in patellofemoral forceand patellar contact area with increased knee angles. Oneexception was at higher knee angles between 70' and 900,in which patellofemoral stress began to plateau or decreaseto a greater extent than patellofemoral force. This occurredbecause although patellar contact area increased between70' and 90', patellofemoral force did not increase propor-tionally but instead began to plateau or decrease. Thesefindings are consistent with patellofemoral force and stressdata during the barbell squat from Escamilla et al. (14) andSalem and Powers (29). Escamilla et al. (14) reported thatpatellofemoral forces increases until 75-80° knee flexionand then began to plateau or slightly decrease. Salem andPowers (29) reported no significant differences in patellofe-moral force or stress at 75', 100', and 1100 knee flexion. Itcan be concluded from these squat data that injury risk to thepatellofemoral joint may not increase with knee anglesbetween 750 and 1100 due to similar magnitudes inpatellofemoral stress during these knee angles, with thebenefit of increased quadriceps, hamstrings, and gastrocne-mius activity when training at higher knee angles (75-1100)compared with training at lower knee angles (0-70') (14).

There are limitations in the current study. Firstly, MRIknee kinematic data have shown during the weight-bearingsquat that the femur moves and rotates underneath arelatively stationary patella, and if femoral rotation isexcessive, it may result in an increase in patellofemoralcontact area, force, and stress on the contralateral patellarfacets (12,22,28). This implies that excessive medialfemoral rotation during the squat ascent may place morestress on the lateral patellar facets, whereas excessive lateralfemoral rotation during the squat descent may place morestress on the medial patellar facets. Unfortunately, collect-ing MRI knee kinematic data while performing the wallsquat or the one-leg squat is not currently possible due tolimitations in equipment design, so it is, unknown howmuch femoral rotation occurs during the wall squat or theone-leg squat, how this rotation varies among healthyversus pathologic individuals while performing theseexercises, and if femoral rotation occurs in the wall squatand the one-leg squat similarly to how it occurs in otherweight-bearing exercises.

Another limitation is the effect of Q-angle on patellofe-moral force and stress. From cadaveric data during asimulated squat, it was shown that an increased Q-anglesignificantly caused a lateral shift and medial tilt androtation of the patella, which may increase patellofemoral

886 Official Journal of the American College of Sports Medicine http://www.acsm-msse.org

force and stress (23). Unfortunately, it is currently difficult

or impossible to effectively measure lateral shift and medial

tilt and rotation of the patellar while performing squat typeexercises. Moreover, increased medial femoral rotation,which also increases Q-angle, is also difficult to measure

accurately while squatting.There are also limitations in the biomechanical model.

First, muscle and patellofemoral forces were estimated frommodeling techniques and not measured directly, which is

currently not possible in vivo. Second, patellofemoral stressmagnitudes were measured using patellar contact area

values from MRI data from the literature and were notmeasured directly. However, the contact areas used fromthe literature were determined during loaded weight-bearing

exercise in healthy male and female subjects, similar to the

current study. Moreover, the near linear and direct relation-ship between contact area and knee angle has been shown to

be similar among studies (2,8,20,27,30). This implies thatthe patellofemoral stress curve patterns shown in Figure 5

using contact areas from the literature will be similar topatellofemoral stress curve patterns if contact areas weremeasured directly using MRI. The patellofemoral stress

patterns are important to clinicians in determining whatknee range of motions stress increases or decreases.

There are limitations regarding the magnitude of patello-

femoral contact areas (and concomitant stress magnitudes),in which the literature reports a wide array. For example,both Patel et al. (27) and Besier et al. (2), who also usedloaded weight-bearing exercise, reported approximately

40-50% higher patellofemoral contact areas compared withcontact areas data from Salsich et al. (30). Using these larger

contact areas from Patel et al. (27) and Besier et al. (2)would have resulted in smaller patellofemoral stress

magnitudes than those reported in the current study.Differences in patellar contact area magnitudes and con-comitant patellofemoral stress magnitudes among weight-

REFERENCES

1. Basmajian JV. Blumenstein R. Electrode Placement in EMGBiofeedback. Baltimore: Williams and Wilkins; 1980. pp. 79-86.

2. Besier TF, Draper CE, Gold GE, Beaupre GS, Delp SL.Patellofemoral joint contact area increases with knee flexion andweight-bearing. J Orthop Res. 2005;23(2):345-50.

3. Biedert RM, Sanchis-Alfonso V. Sources of anterior knee pain.Clin Sports Med. 2002;21(3):335-47, vii.

4. Boling MC, Bolgla LA, Mattacola CG, Uhl TL, Hosey RG.Outcomes of a weight-bearing rehabilitation program for patientsdiagnosed with patellofemoral pain syndrome. Arch Phys MedRehabil. 2006;87(li): 1428-35.

5. Brechter JH, Powers CM. Patellofemoral joint stress during stairascent and descent in persons with and without patellofemoralpain. Gait Posture. 2002;16(2):115-23.

6. Cholewicki J, McGill SM, Norman RW. Comparison of muscleforces and joint load from an optimization and EMG assistedlumbar spine model: towards development of a hybrid approach.J Biomech. 1995;28(3):321-31.

7. Cohen ZA, Henry JH, McCarthy DM, Mow VC, Ateshian GA.Computer simulations of patellofemoral joint surgery. Patient-

bearing studies are due to many factors, such as sex (males

generally have greater contact areas than females), mass

(larger individuals generally have greater contact areas thansmaller individuals), measuring techniques, and loadingmagnitudes (greater loading and quadriceps contractionresults in increased contact areas and less patellofemoralstress (2,30). Nevertheless, although patellofemoral stress

magnitudes during weight-bearing exercises or functionalactivities are approximations only and not exact values,

understanding how magnitudes vary among exercises mayb6 helpful to clinicians in deciding interventions to useduring patellofemoral rehabilitation.

Finally, the current study wvas limited to healthy subjects

who were able to perform the wall squat and the one-leg

squat in the sagittal plane of motion without transverse andfrontal plane motions. Future studies are needed during thewall squat and the one-leg squat to investigate the effects oftransverse plane rotary motions and frontal plane valgus/

varus motions on patellofemoral force and stress magni-tudes, which may occur with individuals with patellofe-moral pathology. Moreover, the results of this study are

specific to performing a one-leg squat holding a singledumbbell at the chest with a forward trunk tilt ofapproximately 30-40' and performing the wall squatexercises with a dumbbell in each hand with the arms atthe side. Using different technique variations during the

one-leg squat and the wall squat exercises may affectpatellofemoral force and stress magnitudes, and this should

also be the focus of future studies.

The efforts of Dr. Bonnie Raingruber and the funding from theNational Institute of Child Health and Human Development'sExtramural Associates Research Development Award programmade this research possible. Also acknowledged are Lisa Bonacci,Toni Burnham, Juliann Busch, Kristen D'Anna, Pete Eliopoulos, andRyan Mowbray for their assistance in data collection and analyses.

specific models for tuberosity transfer. Am J Sports Med. 2003;31(l):87-98.

8. Cohen ZA, Roglic H, Grelsamer RP, et al. Patellofemoral stresses

during open and closed kinetic chain exercises. An analysis usingcomputer simulation. Am J Sports Med. 2001,;29(4):480-7.

9. Dehaven KE, Lintner DM. Athletic injuries: comparison by age,sport, and gender. Am J Sports Med. 1986; 14(3):218-24.

10. Devereaux MD, Lachmann SM. Patello-femoral arthralgia inathletes attending a sports medicine clinic. Br J Sports Med.1984;18(l):18-21.

11. Dixit S, DiFiori JP, Burton M, Mines B. Management ofpatellofemoral pain syndrome. Am Fain Physician. 2007;75(2):194-202.

12. Doucette SA, Child DD. The effect of open and closed chainexercise and knee joint position on patellar tracking in lateralpatellar compression syndrome. J Orthop Sports Phys Ther. 1996;23(2):104-10.

13. Epstein M, Herzog W. Theoretical Models of Skeletal Muscle:Biological and Mathematical Considerations. New York: Joh"Wiley & Sons; 1998. p. 238.

Medicine & Science in Sports & Exercises 887

71

T1

WALL SQUAT AND ONE-LEG SQUAT

14. Escamilla RF, Fleisig GS, Zheng N, Barrentine SW, Wilk KE,Andrews JR. Biomechanics of the knee during closed kineticchain and open kinetic chain exercises. Med Sci Sports Exere.1998;30(4):556-69.

15. Escamilla RF, Fleisig GS, Zheng N, et al. Effects of techniquevariations on knee biomechanics during the squat and leg press.Med Sci Sports Exerc. 2001;33(9):1552-66.

16. Fredericson M, Yoon K. Physical examination and patello-femoral pain syndrome. Am JPhvs*Med Rehabil. 2006;85(3):234-43.

17. Heino Brechter J, Powers CM. Patellofemoral stress duringwalking in persons with and without patellofemoral pain. MedSci Sports Exerc. 2002;34(10):1582-93.

18. Heintjes E, Berger MY, Bierma-Zeinstra SM, Bemsen RM,Verhaar JA, Koes BW. Exercise therapy for patellofemoral painsyndrome. Cochrane Database Syst Rev. 2003;(4):CD003472.

19. Herzog W, Read U. Lines of action and moment arms of themajor force-carrying structures crossing the human knee joint.JAnat. 1993;182(Pt 2):213-30.

20. Hinterwimmer S, Gotthardt M, von Eisenhart-Rothe R, et al. Invivo contact areas of the knee in patients with patellar subluxation.JBiomech. 2005;38(10):2095-101.

21. LaBella C. Patellofemoral pain syndrome: evaluation and treat-ment. Prim Care. 2004;31(4):977-1003.

22. Li G, DeFrate LE, Zayontz S, Park SE, Gill TJ. The effect oftibiofemoral joint kinematics on patellofemoral contact pressuresunder simulated muscle loads. J Orthop Res. 2004;22(4):801-6.

23. Mizuno Y, Kumagai M, Mattessich SM, et al. Q-angle influencestibiofemoral and patellofemoral kinematics. J Orthop Res. 2001;19(5):834-40.

24. Narici MV, Landoni L, Minetti AE. Assessment of human kneeextensor muscles stress from in vivo physiological cross-sectionalarea and strength measurements. Eur J Appl Physiol. 1992;65(5):438-44.

25. Narici MV, Roi GS, Landoni L. Force of knee extensor and flexormuscles and cross-sectional area determined by nuclear magneticresonance imaging. Eur J Appl Physiol. 1988;57(l):39-44.

26. Natri A, Kannus P, Jarvinen M. Which factors predict the long-termoutcome in chronic patellofemoral pain syndrome? A 7-yr prospec-tive follow-up study. Med Sci Sports Exerc. 1998;30(11):1572-7.

27. Patel VV, Hall K, Ries M, et al. Magnetic resonance imaging ofpatellofemoral kinematics with weight-bearing. J Bone Joint SurgAm. 2003;85-A(12):2419-24.

28. Powers CM. The influence of altered lower-extremity kinematicson patellofemoral joint dysfunction: a theoretical perspective.J Orthop Sports Phys Ther. 2003;33(li):639-46.

29. Salem GJ, Powers CM. Patellofemoral joint kinetics duringsquatting in collegiate women athletes. Clin Biomech (Bristol,Avon). 2001;16(5):424-30.

30. Salsich GB, Ward SR, Terk MR, Powers CM. In vivo assessmentof patellofemoral joint contact area in individuals who are painfree. Clin Orthop Relat Res. 2003(417):277-84.

31. Shapiro R. Direct linear transformation method for three-dimensional cinematography. Res Q. 1978;49(2):197-205.

32. Steinkamp LA, Dillingham MF, Markel MD, Hill JA, KauftnanKR. Biomechanical considerations in patellofemoral joint rehabil-itation. Am J Sports Med. 1993;21(3):438-44.

33. van Eijden TM, Kouwenhoven E, Verburg J, Weijs WA. Amathematical model of the patellofemoral joint. J Biomech. 1986;19(3):219-29.

34. yan Eijden TM, Kouwenhoven E, Weijs WA. Mechanics of thepatellar articulation. Effects of patellar ligament length studiedwith a mathematical model. Acta Orthop Scand. 1987;58(5):560-6.

35. van Linschoten R, van Middelkoop M, Berger MY, et al. ThePEX study-exercise therapy for patellofemoral pain syndrome:design of a randomized clinical trial in general practice and sportsmedicine [ISRCTN83938749]. BMC Musculoskelet Disord. 2006;7:31.

36. Wallace DA, Salem GJ, Salinas R, Powers CM. Patellofemoraljoint kinetics while squatting with and without an external load.J Orthop Sports PhIvs Ther. 2002;32(4):141-8.

37. Wickiewicz TL, Roy RR, Powell PL, Edgerton VR. Musclearchitecture of the human lower limb. Clin Orthop. 1983;(179):275-83.

38. Wickiewicz TL, Roy RR, Powell PL, Perrine JJ, Edgerton VR.Muscle architecture and force-velocity relationships in humans.J Appl Physiol. 1984;57(2):435-43.

39. Wilk KE, Davies GJ, Mangine RE, Malone TR. Patellofemoraldisorders: a classification system and clinical guidelines fornonoperative rehabilitation. J Orthop Sports Phys Ther. 1998;28(5):307-22.

40. Witvrouw E, Danneels L, Van Tiggelen D, Willems TM,Cambier D. Open versus closed kinetic chain exercises in.patellofemoral pain: a 5-year prospective randomized study. Am JSports Med. 2004;32(5):1122-30.

41. Witvrouw E, Lysens R, Bellemans J, Peers K, Vanderstraeten G.Open versus closed kinetic chain exercises for patellofemoralpain. A prospective, randomized study. Am J Sports Med. 2000;28(5):687-94.

42. Wojtys EM, Huston LJ, Taylor PD, Bastian SD. Neuromuscularadaptations in isokinetic, isotonic, and agility training programs.Am J Sports Med. 1996;24(2): 187-92.

43. Zajac FE. Muscle and tendon: properties, models, scaling, andapplication to biomechanics and motor control. Crit Rev BiomedEng. 1989;17(4):359-411.

44. Zheng N, Fleisig GS, Escamilla RF, Barrentine SW. An analyticalmodel of the knee for estimation of internal forces during exercise.J Biomech. 1998;31(l0):963-7.

888 Official Journal of the American College of Sports Medicine hfttp://www.acsm-msse.org

COPYRIGHT INFORMATION

TITLE: Patellofemoral Joint Force and Stress during the WallSquat and One-Leg Squat

SOURCE: Med Sci Sports Exercise 41 no4 Ap 2009

The magazine publisher is the copyright holder of this article and itis reproduced with permission. Further reproduction of this article inviolation of the copyright is prohibited.