Journal of Bodywork & Movement Therapies (2011) 15, 235e241

ava i lab le at www.sc ienced i rec t . com

journa l homepage : www.e lsev ier . com/ jbmt

POSTURE PHYSIOLOGY

Misalignment of the knees: Does it affect humanstance stability

OG

Y

Cassio Marinho Siqueira a,*, Gabriel Bueno Lahoz Moya a,Rene Rogieri Caffaro a, Carolina Fu a, Andre Fabio Kohn b,Cesar Ferreira Amorim c, Clarice Tanaka a

UR

EP

HY

SIO

L

a Department of Physical Therapy, Communication Science & Disorders and Occupational Therapy,Faculty of Medicine, University of Sao Paulo, Sao Paulo, Brazilb Department of Biomedical Engineering, Escola Politecnica, University of Sao Paulo, Sao Paulo, Brazilc Department of Mechanical Engineering , Sao Paulo State University e UNESP e FEG,Guaratingueta, Sao Paulo, Brazil

Received 5 June 2009; received in revised form 21 August 2009; accepted 25 August 2009

TA

TIO

Ne

PO

ST

KEYWORDSPosture;Balance;Equilibrium;Knee;Standing stability;Postural alignment

* Correspondence to: Cassio MarinhoTel.: þ55 11 3069 6867; fax: þ55 11 3

E-mail address: fisioterapia.ichc@h

1360-8592/$ - see front matter ª 200doi:10.1016/j.jbmt.2009.08.005

N&

RE

HA

BIL

I

Summary Introduction: Data describing the relationships between postural alignment andstance stability are scarce and controversial.Objective: The aim of this study was to evaluate the effects of sensory disturbances on kneealignment in upright stance and the effects of knee hyperextension on stance stability.Method: Kinetic and kinematic data of 23 healthy adult women were collected while quietlystanding in four sensory conditions. Kinematic data: knee angle (dependent variables) varia-tions were analyzed across sensory conditions. Kinetic data: as subjects with hyperextendedknees showed a clear tendency to flex their knees as balance challenge increased, center ofpressure (COP) parameters (dependent variables) were analyzed in each sensory conditionamong trial sub-groups: Aligned-Trials (knee angle< 180�), Hyperextended-Trials (>180�)and Adjusted-Trials (>180� initially, turned <180� under challenging conditions).Results: Differences were found in mean velocity of COP in two conditions showing that kneealignment can affect stance stability.Conclusion: Knee hyperextension is a transient condition changing under postural challenges.Knee hyperextension affected postural control as mean velocity was the highest in the hyper-extended group in natural standing sensory condition and lowest with sensory disturbance.ª 2009 Elsevier Ltd. All rights reserved.

IO

Siqueira, Av. Dr. Eneas de Carvalho Aguiar, 255, Cerqueira Cesar, 05403-000, Sao Paulo-SP, Brazil.069 7969.cnet.usp.br (C.M. Siqueira).

9 Elsevier Ltd. All rights reserved.

PR

EV

EN

T

236 C.M. Siqueira et al.

PR

EV

EN

TIO

N&

RE

HA

BIL

ITA

TIO

Ne

PO

ST

UR

EP

HY

SIO

LO

GY

Introduction

Postural alignment is thought to affect postural control as itdetermines the effort required to support the body againstgravity (Kantor et al., 2001). In addition, it is known thatpostural alignment affects sensory afferent inputs to thecentral nervous system, the location of the center-of-gravity (Danis et al., 1998) and it is related to the posturaladjustments necessary to control body balance (Kantoret al., 2001; Raymakers et al., 2005; Saha et al., 2007).

However, reports of the relationship between posturalalignment and stability are scarce and remain inconclusive.Postural alignment has been reported not to correlate withstability in able-bodied subjects (Ferreira et al., 2007), orto have low correlation in able-bodied and vestibularlyimpaired patients (Danis et al., 1998). Daniel et al. (1998)also noted that the six most unstable vestibularly impairedpatients revealed the highest correlation between trunkinclination and stability. Yet, higher center of pressure(COP) displacement, and more frequent correlationsbetween stability and posture parameters, were found inscoliotic, compared with non-scoliotic girls (Nault et al.,2002; Allard et al., 2004; Dalleau et al., 2007).

Biomechanical and neurological factors have beenreported to affect postural responses; similar muscleactivation patterns were observed between healthy chil-dren simulating the crouched posture of cerebral palsy,and children with cerebral palsy (Woollacott et al.,1998).

It is possible that, environmental conditions affectpostural alignment and that postural alignment affectsstance stability. A better understanding of the relationshipbetween postural alignment and stability is crucial forclinical practice (see Box 1). Because stability depends onpostural chain mobility (Kantor et al., 2001), misalignmentof a given segment may demand compensatory movementsin other segments in order to maintain stability. On theother hand, postural instabilities may demand positions ofsegments other than the ideal alignment. Thus, posturalalignment should be considered to improve stability andpostural stability may have an important role in posturalalignment.

We evaluated stance stability in subjects with kneeshyperextension, a postural misalignment in which the kneesare extended beyond 180�. In this condition, knee

Box 1. Clinical relevance.

The understanding of the relationship between knee hyperextthe practitioner taking in consideration that:

� Knee hyperextension is a postural condition highly dependan isolated musculoskeletal condition.� As clinical practice guidelines suggest, knee hyperextens

patients are more vulnerable to knee injuries.� Knee hyperextension changes according to postural ch

a diversity of postural conditions when treating knee hyp� Diversity of postural conditions includes visual and propr� Treatment of knee hyperextension needs to include dyn� All these considerations have to be taken seriously into

stabilization becomes more dependent on passive struc-tures, and quadricep muscle activation is no longer neces-sary (Kapandji, 1980; Loudon et al., 1998; Nguyen andShultz, 2007). Knee hyperextension is clinically relevantdue to increased risk of injury, especially to the anteriorcruciate ligament (Cowan et al., 1996; Bonci, 1999; Devanet al., 2004; Ramesh et al., 2005). Additionally, it seems tobe related to increased risk of falls in elderly populations(O’Brien et al., 1997).

The objective of this study was to evaluate the effects ofvisual and proprioceptive disturbance on knee alignmentand on postural control during upright stance.

Methods

Subjects

Twenty-three healthy women (22.7� 3.2 years, BMI20.3� 1.5) were recruited to participate in this study.Subjects were included using a clinical assessment toguarantee participants with aligned and hyperextendedknees. Exclusion criteria were neurological, musculoskel-etal or respiratory pathology, sensory systems diseases,previous surgery in lower limbs or trunk, dizziness, cogni-tive impairments, consumption of alcohol 24 h prior to thetest, complaints of any pain or fatigue at the moment of thetest. All subjects read and signed an informed consent formthat was approved by the Ethics Committee of the ClinicsHospital, School of Medicine, University of Sao Paulo.

Procedures

Information from three 30-s trials was acquired in each offour test conditions when subject stood still:

1) on the force plate with eyes open (NO FOAM/EO),condition with all sensory inputs available;

2) on the force plate with eyes closed (NO FOAM/EC);3) on foam (Airex balance pad�, 6 cm thick, density 57 g/

dm3) placed on top of the force plate with eyes open(FOAM/EO); and

4) on foam placed on top of the force plate with eyesclosed (FOAM/EC), the most challenging condition withvision and proprioceptive inputs disturbed.

ension and postural control will assist the clinical work of

ent on postural control. Thus, it should not be treated as

ion has to be treated to improve knee alignment as such

allenges. Thus, it is important to consider the need oferextension.ioceptive perturbations.amic conditions.account in the final phase of knee injury rehabilitation.

Figure 1 Example of knee angle variation across trials(subject 7). The horizontal dotted line marks the limit betweenthe hyperextension (above the line) and alignment of the knees(below the line). Initially hyperextended, in the last five trialsthe subject behaved as aligned. In the trial 5 knee anglesfluctuated below and above the line being excluded.

Figure 2 Mean knee angle for the aligned andhyperextended subjects across test conditions. *Significantintra-group difference from NO FOAM/EO (p< 0.001) and fromNO FOAM/EC (p< 0.05).

Misalignment of the knees 237

PR

EV

EN

TIO

N&

RE

HA

BIL

ITA

TIO

Ne

PO

ST

UR

EP

HY

SIO

LO

GY

Subjects were asked to stand quietly on the force plate,barefoot, with feet parallel, forward oriented, apart at hipwidth (Prieto et al., 1996) and arms crossed against thechest.

Data acquisition and signal processing

Kinematic data were collected using a Panasonic�PV-GS250 digital camcorder at 60 Hz placed to the right sideof subject. Spherical retro-reflective landmarks wereplaced at the right side of the subjects at the lateral mal-leolus, the head of the fibula, and the greater trochanter.

Signal was digitized and 2D-analyzed using Ariel PostureAnalysis System software with the appropriate calibration.The mean angle (A-knee) and range of motion (ROM-knee)of knee were calculated for each trial.

Kinetic data were sampled at 100 Hz using an AMTI OR6-71000 force plate (AMTI, Newton, MA). Instant COP and thefollowing COP variables were calculated in Matlab (6.5,Mathworks) environment: 1) root mean square (RMS) of theanterioreposterior COP displacement, 2) mean velocity(MV), mean of COP displacement velocities between eachtwo consecutive sample points in the anterioreposteriordirection; and 3) area of the ellipse containing 85% of thetotal COP displacement (Area) (Oliveira et al., 1996).

Data analyses

Sensory disturbance and knee alignment (kinematicanalyses)To analyze the effects of visual and proprioceptive distur-bance on knee alignment during upright stance, two groupsof subjects based on the knee angle in the NO FOAM/EOcondition, in all natural sensory inputs which are inte-grated, were considered: aligned group (knee angle< 180�)and hyperextended group (knee angles> 180�).

Kinematic variables (A-knee and ROM-knee) werecompared across test conditions in each group usingGeneral Linear Models Analysis of Variance (Anova) andTukey post hoc tests with significance level of 0.05.

Sensory disturbance and postural control (kineticanalyses)Visual and proprioceptive disturbance showed an effect onknee alignment across trials, revealing a tendency to flexthe knees when the balance challenge was increased. Kneeangle could oscillate within-trial from >180� to <180�

under challenging conditions. Therefore, to verify theeffects of visual and proprioceptive disturbance on kneealignment and postural control during upright stance, trialsbased on the behavior of knee angles, across test condi-tions, were assessed in three study sub-groups. Example ofknee angle variation across trials in test conditions is shownin Figure 1.

Trials were assigned to Aligned-Trials or Hyperextended-Trials if the knee angles were <180� or >180� throughoutthe entire acquisition time of the trial, respectively.

Trials presenting knee angle >180� previously thatchanged to <180� throughout the entire 30-s of acquisitiontime, were assigned as Adjusted-Trials. Trials in which theknee angle fluctuated around 180� were excluded.

The COP variables were compared among the trialsgroups in each sensory condition using ManneWhitney andKruskalleWallis tests with significance level of 0.05.

Also, Spearman correlation tests were performed toverify correlations between kinetic and kinematicvariables.

Results

All subjects were able to maintain balance during the 30-strials without any retake.

Sensory disturbance and knee alignment(kinematic analyses)

Regarding kinematic analysis, there were significantbetween-groups differences in knee angle (p< 0.001), butnone in knee range of motion, in all test conditions.

Figures 2 and 3 show the variation of knee angle andknee range of motion throughout the test conditions in bothgroups.

Figure 3 Mean range of motion of knee for the aligned andhyperextended subjects across test conditions. *Significantintra-group difference from NO FOAM/EO (p< 0.001).**Significant intra-group difference from NO FOAM/EO(p< 0.001), from NO FOAM/EC (p< 0.001) and from FOAM/EO(p Z 0.004). Note the tendency for knee flexion and increasedknee range of motion when balance is threatened.

238 C.M. Siqueira et al.

PR

EV

EN

TIO

N&

RE

HA

BIL

ITA

TIO

Ne

PO

ST

UR

EP

HY

SIO

LO

GY

Within-groups analysis showed that while knee angletended to decrease with sensory disturbance, knee range ofmotion tended to increase in both groups. Knee angle didnot vary in the Aligned group, whereas the Hyperextendedgroup showed significantly lower knee angles in the FOAM/EC than in NO FOAM/EO (p< 0.001) or NO FOAM/EC(p< 0.05) conditions. There was no significant difference inknee range of motion across test conditions for the Alignedgroup, but the Hyperextended group had higher knee rangeof motion in the FOAM/EC condition, than in the FOAM/EO(p< 0.004), NO FOAM/EC (p< 0.001), and NO FOAM/EO(p< 0.001) conditions. Knee range of motion was alsohigher in the FOAM/EO condition than in the NO FOAM/EO(p< 0.001) condition.

Kinetic analyses

Table 1 shows the distribution of the number of trials,based on the behavior of knee angles during trials, acrosstest conditions.

Table 2 shows the COP variable outcomes in eachsensory condition for the three groups. Between-groupscomparisons in each test condition showed that in the NOFOAM/EO condition, the Hyperextended group hada significantly higher MV than the Aligned group (p< 0.001).

Table 1 Sub-groups distribution and number of trials based on t

Sub-group Knee angle No of trials p

NO FOAM/EO

Aligned-Trials <180� 27Hyperextended-Trials >180� 40Adjusted-Trialsa <180�

Excluded-Trialsb Inconsistent 2Total of trials 69

a Trials of subjects who had knee angle> 180� in previous test condb Trials in which knee angle oscillated within-trial from >180� to <

In the NO FOAM/EC condition, there were no differencesbetween-groups. Only a trend (p< 0.079) was found, withthe Aligned group tending to present a higher mean ellipseArea than the Hyperextended group.

In the FOAM/EO condition, there were no significantdifferences among the three sub-groups. In the FOAM/ECcondition, however, there were significant differences inthe MV-ap of the three sub-groups (KruskaleWallis,p< 0.003). ManneWhitney tests showed that theAligned-Trials sub-group had a higher and marginally lowerVM-ap than the Hyperextended group (p< 0.016) and theAdjusted group (p Z 10), respectively. The Adjusted grouphad a higher VM-ap than the Hyperextended group(p< 0.003).

Spearman tests showed significant negative correlationsbetween the COP variables and knee angle. For Hyper-extended group, the correlation coefficient (r) ranged from�15% (Area) to �18% (MV). For Aligned group, r ranged from�29.8% (RMS) to �37.3% (MV). For the whole sample,r ranged from �17.3% (MV) to �19.4% (Area).

The correlations between COP variables and knee rangeof motion were higher, especially for Hyperextended group(r Z 68% for RMS, 64% for MV, and 67% for Area). For Alignedgroup, r values were 38% for RMS, 44.5% for MV, and 46% forArea. When considering the whole sample, r ranged from55.5% for RMS to 59.6% for Area.

Discussion

This study provides original information about the rela-tionship between knee alignment and stance stability; itassessed changes in knee motion and effects of kneealignment on stability during sensory disturbances. Therewas a clear tendency for knee flexion and knee motion toincrease when challenges to balance increased. Addition-ally, the mean velocity of COP displacement varied amonggroups depending on their knee alignment in two conditionsof different sensory disruptions.

The results suggest that knee flexion with increasedmotion was a motor strategy used to endure threats tobalance under more demanding test conditions. In fact,subjects performing balance-demanding tasks such asskating and surfing, keep their knees semi-flexed. Healthyand vestibularly impaired subjects were reported withgreater knee flexion with feet together and eyes closed,than with feet apart and eyes open (Danis et al., 1998) andthe authors argue that body segments such as the knee or

he behavior of knee angles during trials across test conditions.

er test condition

NO FOAM/EC FOAM/EO FOAM/EC

27 27 2738 31 25

6 104 5 7

69 69 69

itions and turned <180� in challenging test conditions.180�. No case of <180� to >180� was registered.

Table 2 Mean and standard deviations of COP variables in which sensory condition for each group and sub-groups.

Test conditions Groups COP variables

MV-ap (cm/s) RMS-ap (cm) Area (cm2)

NO FOAM/EO Aligned 0.59� 0.16 0.35� 0.28 0.92� 1.21Hyperextended 0.70� 0.12* 0.32� 0.14 0.82� 0.57

NO FOAM/EC Aligned 0.81� 0.27 0.47� 0.37 2.4� 4.79Hyperextended 0.87� 0.19 0.43� 0.17 1.7� 1.43***

Test conditions Sub-groups MV-ap (cm/s) RMS-ap (cm) Area (cm2)FOAM/EO Aligned-Trials 1.05� 0.41 0.57� 0.33 2.99� 2.38

Hyperextended-Trials 1.15� 0.32 0.61� 0.23 3.32� 1.98Adjusted-Trials 1.20� 0.50 0.52� 0.19 2.31� 1.20

FOAM/EC Aligned-Trials 2.47� 0.58 1.41� 1.95 9.29� 4.52Hyperextended-Trials 2.07� 0.62 0.93� 0.28 8.63� 4.41Adjusted-Trials 2.95� 0.78** 1.07� 0.29 11.87� 5.78

MV-ap: mean velocity in anterioreposterior direction; RMS-ap: root mean square in anterioreposterior direction; Area: area of COPdisplacement.*Significant difference between-groups (p Z 0.001).**Significant difference among sub-groups (KruskaleWallis, p Z 0.003).***Trend toward significance between-groups (p Z 0.079).Adjusted-Trials� Hyperextended-Trials (ManneWhitney, p Z 0.003); Adjusted-Trials� Aligned-Trials (ManneWhitney, p Z 0.10);Aligned-Trials� Hyperextended-Trials (ManneWhitney, p Z 0.016).

Misalignment of the knees 239

PR

EV

EN

TIO

N&

RE

HA

BIL

ITA

TIO

Ne

PO

ST

UR

EP

HY

SIO

LO

GY

trunk can move in only one direction when fully extended.However, individuals use other degrees of freedom toadjust their body sway under challenging standingpositions.

The knee flexion of the majority of our subjects (19 outof 23) in the most challenging condition reinforces thisidea. In the NO FOAM/EO condition 57% of trials showedhyperextended knees against 36% in the FOAM/EC condi-tion. It is likely that the subjects who still kept their kneeshyperextended in the most challenging condition did notsense a sufficient threat to change their motor strategy.Given test conditions with higher postural demands, theymight have flexed their knees. However, other studies mustbe done to evaluate the balance parameters underlying thischange of motor strategy.

Subjects initially classified as hyperextended, flexedtheir knees in more challenging conditions. Even in thesame sensory condition, subjects initially hyperextended,sometimes had trials in which the knees were kept hyper-extended and others in which knees turned aligned, leadingto the establishment of an Adjusted-Trials sub-group.Despite similar biomechanics between the Aligned andAdjusted-Trials sub-group (both with flexed knees), theneural control over the knees might not be similar.

Subjects with hyperextended knees, who were led toflex them as a response to a postural challenge, exploreda new motor experience in order to adopt an unusual kneeposition. It is likely that, in this unusual position, thesensory acquisition and integration and motor processesmay be different than for those subjects spontaneouslyaligned. This result was revealed in quite standing condi-tions and might be clinically and functionally more relevantwhen performing dynamics tasks.

This is the idea underlying the decision to consider theAdjusted-Trials sub-group instead of considering onlybiomechanical parameters of knee alignment and include

these trials into the Aligned group when evaluating theeffects of knee alignment on stability.

The mean velocity of the COP has been reported to bethe most discriminating variable between conditions andsubject populations in terms of balance control (Hufsch-midt et al., 1980; Prieto et al., 1993; Prieto et al., 1996;Cornilleau-Peres et al., 2005; Raymakers et al., 2005).

While RMS has been related to the postural stabilityachieved by the postural control system, the mean velocityhas been linked to the amount of regulatory activity asso-ciated with the level of stability (Hufschmidt et al., 1980;Prieto et al., 1993; Prieto et al., 1996; Cornilleau-Pereset al., 2005; Raymakers et al., 2005; Mezzarane and Kohn,2008).

In this study, the Hyperextended group had a higher MVthan the Aligned group in the NO FOAM/EO condition. In thetwo intermediary conditions, no differences were found forMV. However, in the most challenging condition (FOAM/EC)results revealed the lowest MV in the Hyperextended-Trialssub-group and the highest in the Adjusted-Trials sub-group.

Our results showed only a tendency for lower COPamplitude displacement (Area and RMS) in the NO FOAMconditions in the Hyperextended group. Significant differ-ences other than those in the MV were not found, probably,due to the high variability in the Aligned group (coefficientof variance in some cases above 100%). p-values rangedfrom 0.079 to 0.20 in the comparisons of RMS and AREAbetween Aligned and Hyperextended groups.

The higher MV in the Hyperextended group in the NOFOAM/EO condition might possibly be due to the constrictionof knee and ankle motion, by knee hyperextension leading tocompensatory movements in proximal segments (Daniset al., 1998; Kantor et al., 2001). A possible hypothesis toexplain the inversion of the MVresults in themost challengingcondition (FOAM/EC condition) is that, as the foam is nota firm support, it is deformed by the movements of the distal

240 C.M. Siqueira et al.

PR

EV

EN

TIO

N&

RE

HA

BIL

ITA

TIO

Ne

PO

ST

UR

EP

HY

SIO

LO

GY

segments of the lower limb which require faster movementsthat are reflected in the higher MV of the Adjusted-Trials andAligned-Trials sub-groups, while the constriction of the distalsegments in the Hyperextended-Trials sub-group preventsthis from happening.

The Romberg quotient is a classical computation of thevisual contribution to posture and when calculated fromthe mean velocity leads to a more reliable measurement(Romberg quotient Z MV(EC)/MV(EO)) (Cornilleau-Pereset al., 2005). The calculation of the Romberg quotient fromthe means of the groups showed that closing the eyes whilestanding on foam increased MV by 146% in the Adjusted,135% in the Aligned-Trials sub-group and only 80% in theHyperextended-Trials sub-groups.

These results reinforce the idea that knee alignmentaffects stance stability because subjects with knee hyper-extension showed more regulatory activity as demonstratedby their higher MV in the NO FOAM/EO condition. Duringchallenging balance conditions they exhibited a motorstrategy by flexing their knees, and even when showingknee angles similar to the aligned subjects (Adjusted-Trialssub-group) they tend to show higher MV.

Of the rare reports available that address this topic, fewreach the same conclusion. We believe that the discrepancymight arise as much from variations in postural and stabilityparameters as from the level of challenges to balance.

Able-bodied individuals under easy balance tasks mightcompensate for existing misalignment effects withmovements of other segments in order to maintain a safebalance position (Tarantola et al., 1997; Danis et al.,1998; Kantor et al., 2001), as also suggested by higherenergy expenditure without differences in COP displace-ment in flexed trunk positions (Saha et al., 2007). Webelieve that changes in stance stability will be noticeableonly when the compensatory mechanisms reach theirlimits under more challenging stance conditions (Daniset al., 1998) or in severe postural misalignment (Naultet al., 2002; Allard et al., 2004; Dalleau et al., 2007; Sahaet al., 2007). This may also be the reason that MV is themost discriminative COP variable. The central nervoussystem must safely maintain center-of-gravity whenperforming balance tasks; therefore misalignments wouldlikely require more regulatory activity, as expressed bythe mean velocity, to ensure stability (Tarantola et al.,1997; Saha et al., 2007).

We, like other investigators (Danis et al., 1998; Naultet al., 2002; Ferreira et al., 2007), were primarily exam-ining correlations between postural alignment and stability;however we found greater correlations between COP vari-ables and the range of motion of the knee rather than theknee angle itself. Therefore, in future studies the amountof regulatory activity associated with a given posturalmisalignment should be considered. Additionally, it wouldbe interesting to examine how postural misalignmentaffects responses of dynamic posture to internal andexternal threats.

Limitations

A large sample size would be needed to increase thestatistical power of these results. Moreover, an increased

postural challenge, by adding a higher sensory distur-bance, would enhance the differences found in thisstudy.

Conclusion

Our study has revealed that knee hyperextension is a tran-sient condition showing a tendency to flex as posturalchallenge increases. Also, knee hyperextension affectspostural control as mean velocity is the highest in thehyperextended group in natural standing sensory condi-tions, and lowest with sensory disturbance. This studyreinforces that, when treating knee hyperextension, eitherwith the purpose of obtaining alignment, or treating kneesfor a hyperextension injury, postural control has to betaken into consideration.

References

Allard, P., Chavet, P., Barbier, F., Gatto, L., Labelle, H.,Sadeghi, H., 2004. Effect of body morphology on standingbalance in adolescent idiopathic scoliosis. Am. J. Phys. Med.Rehabil. 83 (9), 689e697.

Bonci, C.M., 1999. Assessment and evaluation of predisposingfactors to anterior cruciate ligament injury. J. Athl. Train. 34(2), 155e164.

Cornilleau-Peres, V., Shabana, N., Droulez, J., Goh, J.C.,Lee, G.S., Chew, P.T., 2005. Measurement of the visual contri-bution to postural steadiness from the COP movement: meth-odology and reliability. Gait. Posture 22 (2), 96e106.

Cowan, D.N., Jones, B.H., Frykman, P.N., Polly Jr., D.W.,Harman, E.A., Rosenstein, R.M., Rosenstein, M.T., 1996. Lowerlimb morphology and risk of overuse injury among male infantrytrainees. Med. Sci. Sports Exerc. 28 (8), 945e952.

Dalleau, G., Allard, M.S., Beaulieu, M., Rivard, C.H., Allard, P.,2007. Free moment contribution to quiet standing in able-bodied and scoliotic girls. Eur. Spine J. 16 (10), 1593e1599.

Danis, C.G., Krebs, D.E., Gill-Body, K.M., Sahrmann, S., 1998.Relationship between standing posture and stability. Phys.Ther. 78 (5), 502e517.

Devan, M.R., Pescatello, L.S., Faghri, P., Anderson, J., 2004. Aprospective study of overuse knee injuries among femaleathletes with muscle imbalances and structural abnormalities.J. Athl. Train. 39 (3), 263e267.

Ferreira, E., Duarte, M., Marques, A.P., Maldonado, E.P., 2007.Correlation of postural alignment in sagittal view and posturalcontrol. Santos-SP, Brazil, The International Journal for theMultidisciplinary Study of Voluntary Movements. Prog. MotorControl VI 11, S162.

Hufschmidt, A., Dichgans, J., Mauritz, K.H., Hufschmidt, M., 1980.Some methods and parameters of body sway quantification andtheir neurological applications. Arch. Psychiatr. Nervenkr. 228(2), 135e150.

Kantor, E., Poupard, L., Le Bozec, S., Bouisset, S., 2001. Does bodystability depend on postural chain mobility or stability area?Neurosci. Lett. 308 (2), 128e132.

Kapandji, I.A., 1980. Fisiologia Articular. Manole, Sao Paulo.Loudon, J.K., Goist, H.L., Loudon, K.L., 1998. Genu recurvatum

syndrome. J. Orthop. Sports Phys. Ther. 27 (5), 361e367.Mezzarane, R.A., Kohn, A.F., 2008. Postural control during

kneeling. Exp. Brain Res. 187 (3), 395e405.Nault, M.L., Allard, P., Hinse, S., Le Blanc, R., Caron, O.,

Labelle, H., Sadeghi, H., 2002. Relations between standingstability and body posture parameters in adolescent idiopathicscoliosis. Spine 27 (17), 1911e1917.

Misalignment of the knees 241

Nguyen, A.D., Shultz, S.J., 2007. Sex differences in clinicalmeasures of lower extremity alignment. J. Orthop. Sports Phys.Ther. 37 (7), 389e398.

O’Brien, K., Culham, E., Pickles, B., 1997. Balance and skeletalalignment in a group of elderly female fallers and nonfallers.J. Gerontol. A Biol. Sci. Med. Sci. 52 (4), B221eB226.

Oliveira, L.F., Simpson, D.M., Nadal, J., 1996. Calculation of areaof stabilometric signals using principal component analysis.Physiol. Meas. 17 (4), 305e312.

Prieto, T.E., Myklebust, J.B., Hoffmann, R.G., Lovett, E.G.,Myklebust, B.M., 1996. Measures of postural steadiness:differences between healthy young and elderly adults. IEEETrans. Biomed. Eng. 43 (9), 956e966.

Prieto, T.E., Myklebust, J.B., Myklebust, B.M., 1993. Character-ization and modeling of postural steadiness in the elderly:a review. Trans. Rehabil. Eng. 1 (1), 26e34.

Ramesh, R., Von Arx, O., Azzopardi, T., Schranz, P.J., 2005. Therisk of anterior cruciate ligament rupture with generalised jointlaxity. J. Bone Joint Surg. Br. 87 (6), 800e803.

Raymakers, J.A., Samson, M.M., Verhaar, H.J., 2005. The assess-ment of body sway and the choice of the stability parameter(s).Gait. Posture 21 (1), 48e58.

Saha, D., Gard, S., Fatone, S., Ondra, S., 2007. The effect of trunk-flexed postures on balance and metabolic energy expenditureduring standing. Spine 32 (15), 1605e1611.

Tarantola, J., Nardone, A., Tacchini, E., Schieppati, M., 1997. Humanstance stability improves with the repetition of the task: effect of footposition and visual condition. Neurosci. Lett. 228 (2), 75e78.

Woollacott, M.H., Burtner, P., Jensen, J., Jasiewicz, J.,Roncesvalles, N., Sveistrup, H., 1998. Development of posturalresponses during standing in healthy children and children withspastic diplegia. Neurosci. Biobehav. Rev. 22 (4), 583e589.

PR

EV

EN

TIO

N&

RE

HA

BIL

ITA

TIO

Ne

PO

ST

UR

EP

HY

SIO

LO

GY