The Knee xxx (2009) xxx–xxx

THEKNE-01262; No of Pages 7

Contents lists available at ScienceDirect

The Knee

ARTICLE IN PRESS

Dynamic alignment and its association with knee adduction moment inmedial knee osteoarthritis

Nasim Foroughi a,⁎, Richard M. Smith a, Angela K. Lange a, Michael K. Baker a,Maria A. Fiatarone Singh a,b,c, Benedicte Vanwanseele a

a Exercise, Health and Performance Research Group, Faculty of Health Sciences, University of Sydney, Sydney, NSW, Australiab Faculty of Medicine, University of Sydney, Sydney, NSW, Australiac Hebrew SeniorLife and Jean Mayer USDA Human Nutrition Centre on Aging at Tufts University, Boston, MA, USA

⁎ Corresponding author. Exercise, Health and Performof Health Science, University of Sydney, PO Box 170, LidTel.: +61 433 663 590; fax: +61 2 9351 9204.

E-mail address: n.foroughi@usyd.edu.au (N. Forough

0968-0160/$ – see front matter. Crown Copyright © 20doi:10.1016/j.knee.2009.09.006

Please cite this article as: Foroughi N, eosteoarthritis, The Knee (2009), doi:10.101

a b s t r a c t

a r t i c l e i n f o

Article history:Received 22 June 2009Received in revised form 10 September 2009Accepted 26 September 2009Available online xxxx

Keywords:OsteoarthritisAngular velocityDynamic alignmentKnee adduction moment

Lower limb dynamic alignment represents the limb position during functional loading conditions andobtains valuable information throughout the gait cycle rather than a single instant in time. This study aims todetermine whether dynamic alignment is altered in medial knee osteoarthritis (OA) and how dynamicalignment is related to knee adduction moment (KAM). Community-dwelling women (n=17) with medialOA in at least one knee, according to the American College of Rheumatology criteria and 17 body mass index-matched women without OA were recruited. A three-dimensional motion analysis system was used tocollect the gait data at self-selected habitual and maximal speeds. Clinical evaluation of lower extremities,physical function, pain, habitual level of physical activity, quality of life and physical self-efficacy wereassessed. Shank adduction angle and shank mean angular velocity were significantly greater in the OA groupcompared to the controls from heel strike to 30% stance. KAM was not different between the groups(p=0.542). Dynamic alignment variables were the best predictors of KAM. Health-related quality of life,habitual level of physical activity, lower extremity muscle strength and balance performance were impairedin the OA group compared to the controls. The importance of variables that contribute to dynamic alignmentand the contribution of limb alignment to KAM were highlighted in this study. Detection of postural changessuch as altered dynamic alignment in early stages of OA will lead to the institution of joint-protectivemeasures including changes in footwear, orthotics, gait re-training, use of assistive devices to reduce weight-bearing loads, strengthening and balance enhancing exercises, better analgesia, or cartilage-preservingpharmacotherapy.

Crown Copyright © 2009 Published by Elsevier B.V. All rights reserved.

1. Introduction

Osteoarthritis (OA) leads to an impaired gait pattern and loss offunctional independence and its estimated prevalence is proposed toreach to 7million in 2050 [1]. The knee adduction moment (KAM)indicates the medial compartment loading of the knee joint [2–4] andhas been associated with the development and progression of OA [4,5].Results from the previous studies comparing KAM in patients with kneeOA to a healthy control group show a large variability [4,6–9]. Forexample, some studies have reported a greater KAM in patients withmore sever OA than those with less severity [4,8–10], whereas othersfound a difference on KAM between the OA patients and healthycontrols [6,7]. Mechanicalmodel used tomeasure KAM can influence itsmagnitude andwaveform shape [11] and itsmeasurement needs access

ance Research Group, Facultycombe, NSW, 1825, Australia.

i).

09 Published by Elsevier B.V. All rig

t al, Dynamic alignment an6/j.knee.2009.09.006

to force plates and inverse dynamics calculationswhich restrict its use towell equipped gait laboratories.

Increased degree of static varus alignment measured by means ofradiographs is associatedwith increased OA severity [12], developmentandprogression [13–15], and is correlatedwithKAM[5]. There are somelimitations involved in the assessment of alignment obtained from full-length standing radiographs such as limited accessibility to someclinicians, unnecessary exposure to radiation, and economical costs [16].In addition, foot position [17] and weight-bearing status duringradiography significantly affects frontal plane knee alignment [18].

Hunt et al. [19] suggested that standard quantitative gait analysisprovides a valid measure of dynamic lower limb alignment and may bemore appropriate to reduce the risk of harm to the patients with kneeOA. Unlike static alignment, dynamic alignment reflects the jointloading behaviour during dynamic conditions such as gait and providesuseful information about the limb during the gait cycle rather than asingle instance in time. The dependent variables that describe thedynamic alignment are frontal plane shank (to provide the absoluteorientation of the shank in space), thigh, and knee angles. Together

hts reserved.

d its association with knee adduction moment in medial knee

2 N. Foroughi et al. / The Knee xxx (2009) xxx–xxx

ARTICLE IN PRESS

these variables show the alignment of the limb and its absolute positionin the laboratory frame of reference.

Studies investigating the association between alignment anddynamic knee joint loads have reported that compensatory kinematicgait characteristics such as toe-out angle [8,20–22] and lateral trunklean [23,24] have a potential confounding influence on dynamic kneejoint loads. Capturing these altered characteristics is not possible in aninstance of time; however, measuring dynamic alignment duringdynamic activities such as walking is an advantage. It provides helpfulinformation for tailoring therapeutic interventions based on anindividual's unique dynamic loading [23]. Dynamic alignment inassociationwith dynamic loads during gait in patients withmedial OAhas not been compared to a healthy control group. Comparing thesesvariables between OA-affected and healthy controls during gait canprovide greater insight for clinicians in refining the evaluation andrehabilitation of the OA population.

Several investigations found that patients with OA habitually walkmore slowly, have shorter stride length [6,25,26], increased toe-outangle [20], have reduced knee range of motion [2], have increasedknee extension angle at weight acceptance, greater knee and hipadduction moments, and increased axial loading in all lowerextremity joints [8]. Only one study investigated the gait character-istics at maximal speed; however, this study does not support thatmechanical overload may be associated with knee OA [27].

Therefore, we conducted a cross-sectional analysis of a femalecohort with medial knee OA and a body mass index (BMI)-matchedcontrol group without OA symptoms. We hypothesized that:

1) the shank and thigh adduction angles at peak KAM (~30% stance)and shank mean angular velocity would be significantly larger inthe OA group compared to the controls at both habitual andmaximal walking speeds.

2) patients with medial OA would have a higher KAM.3) gait pattern differences between the OA and control group would

be exaggerated at maximal speed.

2. Materials and methods

2.1. Study design and recruitment

This is a cross-sectional substudy of baseline data from a subset ofparticipants enrolled in a randomized single blind controlled trial ofprogressive resistance training (PRT) for knee OA (REACH study [28],Australian New Zealand Clinical Trials Registry Number12605000116628) and healthy controls. The study was approved bythe Human Research Ethics Committee of the University of Sydney onthe 21st of December 2004 (Reference Number 7849) and all thesubjects granted informed consent. Subjects were recruited from April2005 to December 2006 by advertising in local newspapers, publishingarticles, giving talks to local community and senior citizen centres,sending flyers to local businesses, andword-of-mouth. All themeasuresof the REACH study were collected prior to randomization at baseline.The participants in the REACH study may have had medial, lateral orboth compartments affected in one or both knees. The OA subgroup ofthe current study was selected from this cohort if they had only medialOA and matched with the controls based on BMI.

2.2. OA group

Community-dwellingwomen (n=17) over the age of 40 years withmedial OA in at least one knee according to the American College ofRheumatology criteria [29] were included in the OA group. Magneticresonance images (MRI) were used to grade the OA severity accordingto Outerbridge Classification [30]. Subjects undertook gait analysis andclinical evaluation including the history of OA onset and physicalexamination of lower extremities. The exclusion criteria included

Please cite this article as: Foroughi N, et al, Dynamic alignment anosteoarthritis, The Knee (2009), doi:10.1016/j.knee.2009.09.006

secondary OA due to trauma, surgery and other rare forms of arthritis;joint injury, injection or surgery within the past six months or jointreplacement; disorders of the nervous system disrupting voluntarymovement, severe functional limitation (unable to walk without theassistance of a person or an assistive device); or cognitive impairment.

2.3. Control group

Seventeen sedentary women over the age of 40 years withoutdiagnosis or symptoms suggestive of OA,whomet the inclusion criteria,were matched for BMI (kg/m2) with the patients in the OA group. Theywere excluded if they reported morning stiffness and/or knee painduring the last six months, a history of surgery in the lower extremitiesprescribed medication related to musculoskeletal diseases, or a historyof neurological or musculoskeletal disorders which may have affectedtheir normal gait pattern.

2.4. Gait testing and analysis procedure

Measurements were performed on the most symptomatic knee inthe OA group diagnosed by clinical examinations. The groups werematched for the knee used for analysis. A three-dimensional motionmeasurement system (Motion Analysis Corporation, Santa Rosa, USA)including 10 Eagle video cameras and two sets of eight channel forceplates (Kistler instruments, Switzerland) were used to collect the data.The camera's shutter speed was 1/1000 of a second with the samplingrate of 100 frames per second (Hz) and the force data was collected atthe sampling frequency of 1000 Hz.

Thirty eight (20-mm-diameter) passive markers were attachedbilaterally using a modified Helen Hayes marker set [31] by double-sided tape on standard bony landmarks of foot, shank, thigh, pelvis andtrunk segments. Each segment was defined using three markers (sixdegrees of freedom) and idealized as a rigid bodywith a local coordinatesystem defined to coincide with a set of anatomical axes. The three-dimensional positions of markers were used to calculate the location ofthe joint centres.

Subjects wore a black Lycra body suit and walked barefoot along a10meter walkway at their self-selected habitual and maximal speeds(five trials/condition). Maximal speed trials occurred after habitualspeed trials and subjects rested for 2–3min between the trials and5–10min between eachwalking condition. A static trialwas collected asa reference to determine body mass and positions of joint centres ofrotation. The average values from five trials at both velocities werecalculated for each subject permitting comparison of average values foreach subject. Gait velocity for each trial was calculated as the averagevelocity of the sacrum marker in the walking direction during two fullstrides and was averaged over the five trials. Segment angles relative tothe laboratory and relative joint angles were calculated using jointcoordinate systems [32].

The tibial alignment is correlated (r=0.831, p<0.001) withmechanical axis measured form weight-bearing static radiographs[33]. The tibial alignmentwasmeasured using an inclinometer with thesubjects standing on a foot map with their weight distributed equallyover both feet (feet ~29 cm apart) and their second metatarsal alignedwith the middle of the heel [33]. The tibial tuberosity and the middle ofthe talar head were identified to assess orientation of the tibia and agravity inclinometerwasmounted to a set of calipers and thearmsof thecalipers were positioned on the two landmarks [33]. The angle of thetibia was measured with respect to the vertical. Varus was defined asnegative values and valgus as positive values.

Two variables are required to establish the frontal plane dynamicalignment in the global frame of reference: frontal plane relative theangle between the thigh and the shank and the shank or thigh absoluteangles relative to the global coordinate system. These variables weremeasured dynamically during gait.

d its association with knee adduction moment in medial knee

3N. Foroughi et al. / The Knee xxx (2009) xxx–xxx

ARTICLE IN PRESS

Three-dimensional internal moments were calculated usinginverse dynamics via Kintrak™ version 6.2 (The University of Calgary,Canada) and were normalized to the individual's body weight andheight (%BW×Ht). Internal moments are the moments resulting fromthe muscles. They balance the external moments which are equal butin opposite direction. The KAMwas defined as the maximum externalknee adductionmoment about the frontal axis of the knee. Lower limbsegments (shank, and thigh) were modelled as a rigid body with alocal coordinate system that coincided with anatomically relevantaxes. Inertial properties of each limb segment were approximatedanthropometrically and translations and rotations of each segmentwere reported relative to neutral positions defined during the initialstanding static trial. Inclination of the shank segment in the directionof knee joint adduction (positive) was defined as shank adductionangle (see Fig. 1).

The reproducibility of the marker placement was determined byperforming all aspects of the experimental protocol in two differentdays (one week apart) by one investigator. The coefficient of variance(%) for knee adduction moment was 0.5%. The kinematic data waschecked during data collection to reduce the cross-talk betweenplanes and the excluded trial from further analysis was replaced withthe mean of other four trials.

2.5. Other characteristics

The stretched stature [34] and fasting weight were measured earlyin the morning and BMI was calculated from those measurements.

2.6. Muscle strength

Knee joint peak muscle strength was measured using Digital K400Keiser pneumatic resistance machines (Keiser Sports Health Equip-ment, Inc., Fresno, CA). One repetition maximum (1RM) tests wereperformed unilaterally on knee extension and bilaterally on kneeflexion twice and the better of two trials was considered for analysis[35].

2.7. Balance test

The static balance time and body sway were measured via forceplatform with the Chattecx Dynamic Balance System (ChattanoogaGroup Inc., Hixson, TN) [36] by employing previously reportedmethods[28]. Balance was tested for 30s per six conditions (eyes open and

Fig. 1. Positive inclination of the shank segment (shank adduction angle).

Please cite this article as: Foroughi N, et al, Dynamic alignment anosteoarthritis, The Knee (2009), doi:10.1016/j.knee.2009.09.006

closed) including narrow bilateral stance on the platform slidingbackward and forward, narrow bilateral stance on the platform tiltingup and down, and unilateral stance of the preferred leg on a stillplatform. Unilateral stance duration with eyes closed, summatedmaximum sway in four directions (medial, lateral, anterior andposterior), and number of trials needed to complete the six conditionswere used as individual measures of balance performance. Overallbalance performance was examined using a balance index which wascalculated by summating all anterior–posterior andmedio-lateral swaymeasures and time results respectively. A lower balance index indicatesbetter balance and has been shown to have high test–retest reliability(r=0.759, p=<0.001) and is sensitive to change with an exerciseintervention [37].

2.8. Questionnaires

TheWestern Ontario andMcMaster University questionnaire (LikertWOMAC VA3-series) was used to assess the pain, stiffness and physicalfunction for the patients withOA [38]. The Physical Activity Scale for theElderly (PASE) [39] andHarvardAlumniQuestionnaire [40] assessed thelevels of habitual physical activity. Medical Outcome Survey 36-item(version2, short-form, SF-36) [41] assessed the health-related quality oflife with lower scores indicating poorer health. Eight health outcomemeasures were extracted from the SF-36 including three physicalparameters [physical functioning, limitationsdue tophysical difficulties,and bodily pain]; three mental health parameters [mental health, rolelimitations due to emotional difficulties, and social functioning]; andtwo combined physical and mental health parameters [general healthand vitality]. Normalized SF-36 scores were used. All questionnaireswere interviewer-administered in a private roomusing visual cards andperformed for both groups.

2.9. Statistical analysis

Statistical calculations were carried out using SPSS (Release 15.0 forWindows, 2006, Chicago: SPSS Inc) and all p values less than 0.05 wereconsidered statistically significant. The data were expressed as meanand standard deviation (SD) or median and range after beingstatistically tested for normality (skewness −1≥1). Logarithmictransformations were performed for non-normally distributed data.Comparisons between groups for normally distributed data were madeby applying an analysis of co-variance (ANCOVA) with subjects' ageadded to all models as a potential confounding variable. Covariatesidentified a priori for potential inclusion into statistical models includednumber of chronic diseases, total PASE score, muscle strength, physicalfunction, and balance. Confounders were defined as covariates thatdiffered between groups by a clinically important or statisticallysignificant amount. Linear regression analyses were performed to testthe a priori hypotheses stated above.

3. Results

3.1. Health and functional characteristics

Both groups were matched for BMI (26±3 vs. 25±4, p=0.827), but age wassignificantly different between the groups (66±8 vs. 51±7 yr, p<0.001) and was addedto all statistical models as a covariate. The participants were classified as overweight orobese if their MRI was 25–29 and 30, respectively (Table 1). The OA group suffered fromOA for a median of 6 years (0–12); 65% were overweight or obese; 44% had more severesymptoms in the right knee, and approximately 30% had severe OA (grade 3–4). Theaverage measure of health-related quality of life was significantly worse in the OA groupfor physical function difficulties (p<0.001) and bodily pain (p=0.003). The habitual levelof physical activity (p<0.001), knee extension muscle strength (p=0.023) and staticbalance stance time (p=0.031) were all lower in the OA group compared to the controls.In the OA group, habitual energy expenditure in physical activity was less than 50% of thatin the controls.

In the following paragraphs the gait characteristics of women with knee OA havebeen compared to the controls during the habitual and maximal walking speeds.

d its association with knee adduction moment in medial knee

Table 1Baseline characteristics of the OA and control group.

Variables OA (n=17) Control (n=17) p

Age (years) 66±8 51±7 0.001a

BMI (kg/m2)b 26±3 25±4 0.827Normal (%) 18.5–24.99 35.3 52.9Overweight (%)≥25.00–29.99 47.1 41.2Obese (%)≥30.00 17.6 5.9

OA grade (range 1–4; %)c

Mild 58.8Moderate 11.8Severe 23.5Very severe 5.9

No. of chronic diseases (n) 1.12±0.8 0.18±0.4 0.005a

Tibial alignment −0.2±5.1 −1.2±3.5 0.515WOMAC

Pain (range 0–20) 5±4Stiffness (range 0–8) 4±2Difficulty (range 0–68) 20±12Total (range 0–96) 28±17

SF-36 (0–100)Physical functioningd 80 (55) 95 (55) 0.001a

Role limitations due tophysical difficultiesd

87 (50) 100 (75) 0.215

Bodily pain 59±17 85±19 0.003a

General healthd 70 (60) 70 (30) 0.740Vitality 63±16 58±18 0.766Social functioningd 100 (87) 100 (62) 0.581Role limitation due toemotional difficultiesd

100 (50) 100 (58) 0.864

Mental healthd 85 (50) 85 (55) 0.639PASEd 114±37 225±160 0.001a

Harvard Alumni Questionnaire(kcal/week)d

1222±1293 2960±3917 0.082

Knee extension 1RM (Nm) 39±17 56±21 0.023a

Bilateral knee flexion 1RM (Nm) 91±22 104±26 0.146Balance index 90±22 77±11 0.95Total time balanced 155 (94) 159 (27) 0.031a

No. of balance trials (range 6–18)d 8 (6) 8 (1) 0.062Velocity (m/s) 1.17±0.17 1.25±0.13 0.906Stride length (m) 1.22±0.11 1.31±0.12 0.733Stride time (s) 1.03±0.11 1.04±0.61 0.798Base of support (m) 0.11±0.02 0.10±0.03 0.195

Values are the mean±SD; OA, osteoarthritis; BMI, body mass index; WOMAC, WesternOntario and McMaster Osteoarthritis Index, lower score equals to less symptoms [1]; SF-36, version 2 of a 36-item short-form health survey [2], lower values indicate poorerhealth; PASE, Physical Activity Scale for the Elderly [3], Harvard Alumni Questionnaire(Paffenbarger Score) [4]; Nm, Newtonmeters; 1RM, one repetitionmaximum;m,meters;s, seconds.

a OA group are different from control group, a lower balance index indicates betterbalance.

b Adapted fromWorld Health Organization (WHO) 1995,WHO 2000 andWHO 2004.c MRI correlation of Modified Outerbridge Classification [5].d Data reported as median (range) if not normally distributed.

Fig. 2. Frontal plane shank Add-Abduction motions of the OA and control group athabitual speed. Adduction is positive. The slim lines accompanying the average time-series indicates the ±95% confidence intervals.

4 N. Foroughi et al. / The Knee xxx (2009) xxx–xxx

ARTICLE IN PRESS

3.2. Frontal plane angles and angular velocity

As shown in Table 2, the shank adduction angle was greater in the OA group, with thelargest difference occurring at 30% of the stance (5.1±2.8 vs. 2.9±2.1, p=0.012). Theshank mean angular velocity from heel strike to 30% stance at habitual speed was greater

Table 2Peak frontal plane angles and moments at habitual and maximal walking speeds.

Variables Habitual speed

OA (n=17) Control (n=17

Knee abduction anglea 7.46±5.2 10.0±3.1Knee adduction anglea 4.51±3.2 2.55±2.2Shank adduction anglea (at 30% stance) 5.12±2.8 2.9±2.1Thigh adduction anglea (at 30% stance) 5.42±3.07 5.63±1.97Knee Adduction momentc 2.80±1.12 2.22±0.59

Values are mean±SD; OA, osteoarthritis.a In degrees.b OA group are different from control group.c Moments in (%BW×Ht).

Please cite this article as: Foroughi N, et al, Dynamic alignment anosteoarthritis, The Knee (2009), doi:10.1016/j.knee.2009.09.006

in the OA group compared to the controls (21.70±8.09 vs. 14.81±6.52°/s, p=0.027; seeFig. 2). Shank angle was not different between groups during late stance when the kneewas less loaded. The same pattern for shank adduction angle was observed at maximalspeed (p=0.02). The thighadduction anglewasnot differentbetween the groups ineitherspeeds (p=0.81, p=0.797). All the other frontal plane angleswere not different betweenthe groups in either walking speeds (p>0.05).

3.3. Knee adduction moment

Peak KAM was not significantly greater in the OA group than controls at eitherhabitual or maximal speeds. Correlation analysis showed that the shank adductionangle in the OA group increased in proportion to KAM (r=0.39, p=0.026) explaining15% of the variance (Table 2).

3.4. The effect of speed

Walking velocity was not significantly different between the OA and control groupat both habitual andmaximal speeds (p=0.906). Themaximal walking speed of the OAgroup and the controls was on average 154% and 159% faster than the habitual speed,respectively. This increase in speed was not significantly different between the groups(p=0.774). To minimize the effect of the inter-individual variability of speed, thedifferences between the variables at maximal speed and habitual speed were used;however, no additional differences in the change scores were observed between thegroups (data not shown).

3.5. Potential relationships

Shank adduction angle at 30% of the stance phases was related to disease severity(r=0.351, p=0.042) and physical function (r=−0.461, p=0.006); however, it was notrelated to BMI (r=0.281, p=0.107), balance index (r=0.179, p=0.310), and kneeextension strength (r=−0.324, p=0.062). Knee adduction moment was also associatedwith disease severity (r=0.464, p=0.006), number of chronic diseases (r=0.335,p=0.039), physical function (r=−0.350, p=0.042) and PASE score (r=−0.342,p=0.042). The differences between groups and the strength of the relationships betweenthe studied variables remained statistically significant after adjusting for other potentialconfounders (number of chronic diseases, total PASE score, muscle strength, physicalfunction, and balance).

Maximal speed

) p OA (n=17) Control (n=17) p

0.403 8.1±4.6 10.39±2.8 0.4860.154 4.57±3.4 2.46±2.0 0.1170.012b 4.94±2.60 2.88±2.3 0.02b

0.81 5.33±2.9 5.57±2.5 0.7970.542 3.41±1.30 2.84±0.90 0.581

d its association with knee adduction moment in medial knee

5N. Foroughi et al. / The Knee xxx (2009) xxx–xxx

ARTICLE IN PRESS

4. Discussion

This study investigated whether dynamic alignment is impaired inwomen with medial knee OA and its relation to KAM. As hypothe-sized, the shank adduction angle at peak KAM (~30% stance) andshank mean angular velocity were significantly larger in the OA groupcompared to the controls at both habitual and maximal walkingvelocities. No significant differences were found for the thigh angle. Incontrary to our a priori hypotheses patients with knee OA did not havea higher KAM compared to the controls and gait pattern differencesbetween the OA and control group were not exaggerated at maximalspeed.

The results of the current study show that the shank adductionangle reached its peak around 30% of the stance phase. The 30% ofstance time point corresponds with the first peak of KAM. Dynamicknee and shank frontal plane angles at 30% of stance were the bestpredictors for KAM and explained 61% of the variation in KAM. Anincrease in the shank adduction angle may result in a larger momentarm and consequently would increase the magnitude of KAM.However, this was not evident in our study and may be due to thesmall sample or mild degree of OA.

To our knowledge this is the first study to demonstrate that shankadduction angle and shank mean angular velocity (from heel strike to30% stance) is higher in medial knee OA than the control group. Changet al. reported a greater KAM in patients with a thrust compared withknees without a thrust [42]. Our findings are similar to the varusthrust of Chang et al. and could be associated with an increasedpossibility of OA progression reported previously [42]. Varus thrustmay be caused by inability of patients to control the shank adductionduring the weight acceptance phase of stance. More in-depth analysisis required to establish this relationship.

Knee adduction moment has been of great interest in the recentliterature. In contrast to our hypothesis, KAM was not significantlydifferent between patients with knee OA and the controls at eitherhabitual or maximal speeds. Nonetheless the peak KAM of thepatients with OA, reported in the current study during habitual gait,was similar to those reported previously [12,18,25,26,43]. Munder-mann et al. has reported that patients with more sever OA havegreater KAM compared to the controls [12]. Our cohort had mild OAand that may explain the undetected differences in KAM observedbetween the groups. The magnitude of KAM in patients with OAduring habitual speed was very close to the peak KAM at maximumspeed in the control group. This could be due to using compensatorygait mechanisms such as toeing-out [20–22,44–46] or lateral trunklean [23,24,47].

Knee adduction moment has been used as an indicator for OAprogression; however, an examination of the literature reveals thatthere is a larger variability in the results of studies reporting KAM inpatients with OA compared to the matched controls with someinvestigations using different movements (e.g. single limb stance for3s) [7], different models [11] or different cohorts [48]. In addition,measuring KAM can be problematic in clinical settings as theaccessibility to a gait laboratory with force plates may be limited.Kinematic analysis only requires three-dimensional video or opticalmotion capture system which makes the use more practical. Inaddition, in this study the shank adduction angle was more sensitiveto distinguish between the OA and control group. Our study did notfind any difference in KAM between the study groups. Furtherinvestigation is recommended to investigate the association of shankadduction angle with OA progression or even OA development.

Greater shank adduction angle at 30% of stance suggests that theresulting alignment adds more pressure on the medial compartment.The medial compartment supports higher loads than the lateralcompartment during walking in medial knee OA due to thischaracteristic malalignment [3], owing to the fact that the GRF vectoracts medially relative to the knee joint centre [2]. Patients' posture

Please cite this article as: Foroughi N, et al, Dynamic alignment anosteoarthritis, The Knee (2009), doi:10.1016/j.knee.2009.09.006

during gait may affect the magnitude of KAM by shifting the GRF lineof action further away from the knee joint centre.

Our hypothesis that patients with medial OA would adopt adifferent gait pattern compared to the controls at maximal speed wasnot supported by the results. Walking at maximal speed increased thevariability in the OA group and that may explain the undetectabledifferences observed between thewalking conditions of the OA group.The OA group walked with a similar gait pattern at their maximumspeed suggesting that they maintained their gait pattern regardless ofthe speed at which theymay have chosen to ambulate even though onaverage the fast speed was approximately 154% of the habitual speed.In this study, maximal gait speed and the difference between thehabitual and maximal walking speeds were not associated withdisease severity indicating that walking speed may not be a stressorfor this cohort during gait. Given the small sample size, the withingroup variability of patients with OA increased when they walked attheir maximum speed decreasing the ability to detect a differencebetween the groups. This may explain the undetectable differencesbetween the groups when walking at maximal speeds.

Our findings are in contrast to the findings of previous studies,which reported impaired spatial and temporal (ST) variables inpatients with OA compared to a control group [6,26]. Both habitualand maximal walking velocities of our study groups were comparableto the walking velocity reported previously by other investigators[27]. Both groups walked at similar velocities which may explain whywe could not detect any differences in the ST variables between thegroups. The majority of biomechanical variables are influenced bywalking velocity [49]. Kaufman et al. [26] reported a slower walkingvelocity for the OA group compared to the controls. This could beexplained by a 20 year mean age difference between the OA andcontrol group in that study. Since walking barefoot significantlyreduces the joint loads, stride, cadence, and lower extremity's range ofmotion [50] all trials were performed barefoot in our study. This mayexplain the observed differences in ST variables compared to previousstudies.

The balance index of the OA group in this study was comparable tothat reported previously [37]. Functional vision and proprioceptionsystems are the requirements of maintaining the complex task ofbalance in old age. Walking with OA may contain some elements ofunanticipated perturbations due to the way pain or knee instability isexperienced.

The SF-36 instrument assesses quality of life from a physical andmental aspect and pain accounts for 40% of its variance in patientswith knee OA [51]. In this study, physical functioning and bodily painscores indicated poorer health in the OA group compared to thecontrols. Published physical functioning values for an OA cohortranged from 19.5 to 27.05 and bodily pain values ranged from 27.1 to32.90 in the literature, while these values were higher in our patientgroup [52–54]. This indicates that general health was better in ourpatient group than previous studies mainly due to lower levels of painor mild OA severity.

4.1. Limitations

There are some limitations to our study design and therefore theconclusions that may be drawn from our data. First, the imputation ofcause and effect of OA and velocity on the pattern of knee joint loadingmust proceed with caution due to the cross-sectional nature of thestudy. Second, increased walking speed was apparently not a strongenough stressor and did not enlarge the differences between thegroups. The motor control system may be more sensitive to othertasks than walking, such as sit-to-stand or stair climbing. Third, thegroups were not matched for age; however, age was considered as acovariate in the statistical analysis. Fourth, there were no radiographicimages available and the MRI images were not obtained for thecontrols, or for the OA group in the clinically least symptomatic knee.

d its association with knee adduction moment in medial knee

6 N. Foroughi et al. / The Knee xxx (2009) xxx–xxx

ARTICLE IN PRESS

We were not able to measure the static alignment of the participantsto that effect. In addition, some of the asymptomatic control subjectsmay have had undiagnosed OA in their lower extremity joints. Fifth,the majority of the patients had mild to moderate grades of OA (70%had grade 1–2) which may explain the small differences observedbetween the groups. Sixth, marker placement in three-dimensionalgait analysis method can be a source of error. However, markerplacement was done on standardized bony landmarks by the sameexperienced examiner to minimize errors. Finally, the small samplesize precludes definitive clinical conclusions at this stage, althoughthe findings, particularly regarding dynamic alignment are encour-aging for further investigation in this area. Prospective studies oflarger cohorts should be undertaken to distinguish between thecontributory and compensatory factors in knee OA.

5. Conclusions

This study showed that dynamic alignment could distinguishbetween patients with medial tibiofemoral OA and matched controlsand that mean shank angular velocity was increased in the OA groupduring early stance at both habitual and maximum speeds. Furtherinvestigation is required to identify whether these parameters candetect changes in early stages of OA disease and whether they arefavourably altered subsequent to the institution of joint-protectivemeasures including changes in footwear, orthotics, gait re-training,use of assistive devices to reduce weight-bearing loads, weight loss,strengthening of specific muscle groups, balance enhancing exercises,better analgesia, or cartilage-preserving pharmacotherapy.

6. Conflict of interest statement

No part of the following work has been published anywhere elseand there is no commercial relationship related to this work. Thetypescript has been read and agreed by all authors. This study wasconducted under no conflict of interest.

Acknowledgments

The authors acknowledge the valuable time and effort ofvolunteers, MSRT of IRAN (Shiraz University), and The University ofSydney R & D Grant (S4201 U3301).

References

[1] Access-Economics, Painful realities: the economic impact of arthritis in Australia. 2007,Report for Arthritis Australia.

[2] Baliunas AJ, Hurwitz DE, Ryals AB, Karrar A, Case JP, Block JA, et al. Increased kneejoint loads during walking are present in subjects with knee osteoarthritis. OsteoarCartil 2002;10:573–9.

[3] Hurwitz DE, Ryals AR, Block JA, Sharma L, Schnitzer TJ, Andriacchi TP. Knee painand joint loading in subjects with osteoarthritis of the knee. J Orthop Res2000;18:572–9.

[4] Sharma L, Hurwitz DE, Thonar EJ, Sum JA, Lenz ME, Dunlop DD, et al. Kneeadduction moment, serum hyaluronan level, and disease severity in medialtibiofemoral osteoarthritis. Arth Rheum 1998;41:1233–40.

[5] Miyazaki T, Wada M, Kawahara H, Sato M, Baba H, Shimada S. Dynamic load atbaseline can predict radiographic disease progression in medial compartment kneeosteoarthritis. Ann Rheum Dis 2002;61:617–22.

[6] Gok H, Ergin S, Yavuzer G. Kinetic and kinematic characteristics of gait in patientswith medial knee arthrosis. Acta Orthop Scand 2002;73:647–52.

[7] KimWY,Richards J, Jones RK,HegabA. Anewbiomechanicalmodel for the functionalassessment of knee osteoarthritis. Knee 2004;11:225–31.

[8] Mundermann A, Dyrby CO, Andriacchi TP. Secondary gait changes in patients withmedial compartment knee osteoarthritis: increased load at the ankle, knee, andhip during walking. Arth Rheum 2005;52:2835–44.

[9] Thorp LE, Sumner DR, Block JA, Moisio KC, Shott S, Wimmer MA. Knee joint loadingdiffers in individuals with mild comparedwithmoderate medial knee osteoarthritis.Arth Rheum 2006;54:3842–9.

[10] Wada M, Maezawa Y, Baba H, Shimada S, Sasaki S, Nose Y. Relationships amongbone mineral densities, static alignment and dynamic load in patients with medialcompartment knee osteoarthritis. Rheum 2001;40:499–505.

Please cite this article as: Foroughi N, et al, Dynamic alignment anosteoarthritis, The Knee (2009), doi:10.1016/j.knee.2009.09.006

[11] Newell RS, Hubley-Kozey CL, StanishWD, Deluzio KJ. Detecting differences betweenasymptomatic and osteoarthritic gait is influenced by changing the knee adductionmoment model. Gait Posture 2008;27:485–92.

[12] Mundermann A, Dyrby CO, Hurwitz DE, Sharma L, Andriacchi TP. Potential strategiesto reducemedial compartment loading inpatientswithkneeosteoarthritis of varyingseverity: reduced walking speed. Arth Rheum 2004;50:1172–8.

[13] Brouwer GM, van Tol AW, Bergink AP, Belo JN, Bernsen RMD, Reijman M, et al.Association between valgus and varus alignment and the development andprogression of radiographic osteoarthritis of the knee. Arth Rheum 2007;56:1204–11.

[14] Cicuttini F, Wluka A, Hankin J, Wang Y. Longitudinal study of the relationshipbetween knee angle and tibiofemoral cartilage volume in subjects with kneeosteoarthritis. J Rheumatol 2004;43:321–4.

[15] Sharma L, Song J, Felson DT, Cahue S, Shamiyeh E, Dunlop DD. The role of kneealignment in disease progression and functional decline in knee osteoarthritis.JAMA 2001;286:188–95.

[16] Hinman RS, May RL, Crossley KM. Is there an alternative to the full-leg radiographfor determining knee joint alignment in osteoarthritis? Arthritis Rheum2006;55:306–13.

[17] Hunt MA, Birmingham TB, Giffin JR, Jenkyn TR. Associations among knee adductionmoment, frontal plane ground reaction force, and lever arm during walking inpatients with knee osteoarthritis. J Biomech 2006;39:2213–20.

[18] Specogna AV, Birmingham TB, Hunt MA, Jones IC, Jenkyn TR, Fowler PJ, et al.Radiographicmeasures of knee alignment in patients with varus gonarthrosis: effectof weightbearing status and associations with dynamic joint load. Am J Sports Med2007;35:65–70.

[19] Hunt MA, Birmingham TB, Jenkyn TR, Giffin JR, Jones IC. Measures of frontal planelower limb alignment obtained from static radiographs and dynamic gait analysis.Gait Posture 2008;27:635–40.

[20] Hurwitz DE, Ryals AB, Case JP, Block JA, Andriacchi TP. The knee adduction momentduring gait in subjects with knee osteoarthritis is more closely correlated with staticalignment than radiographic disease severity, toe out angle and pain. J Orthop Res2002;20:101–7.

[21] AndrewsM, Noyes FR, Hewett TE, Andriacchi TP. Lower limb alignment and foot angleare related to stance phase knee adduction in normal subjects: a critical analysis of thereliability of gait analysis data. J Orthop Res 1996;14:289–95.

[22] GuoM, AxeMJ, Manal K. The influence of foot progression angle on the knee adductionmoment during walking and stair climbing in pain free individuals with kneeosteoarthritis. Gait Posture 2007;26:436–41.

[23] HuntMA,BirminghamTB, BryantD, Jones IC,Giffin JR, JenkynTR, et al. Lateral trunk leanexplains variation in dynamic knee joint load in patients with medial compartmentknee osteoarthritis. Osteoarthritis Cartilage 2008;16:291–599.

[24] Mundermann A, Asay JL, Mundermann L, Andriacchi TP. Implications of increasedmedio-lateral trunk sway for ambulatory mechanics. Clin Biomech 2008;41:165–70.

[25] Messier SP, DeVita P, Cowan RE, Seay J, Young HC, Marsh AP. Do older adults withknee osteoarthritis place greater loads on the knee during gait? A preliminarystudy. Arch Phys Med Rehabil 2005;86:703–9.

[26] Kaufman KR, Hughes C, Morrey BF, Morrey M, An KN. Gait characteristics of patientswith knee osteoarthritis. J Biomech 2001;34:907–15.

[27] Landry S, McKean K, Hubley-Kozey C, Stanish W, Deluzio K. Knee biomechanics ofmoderate OA patients measured during gait at a self-selected and fast walking speed.J Biomech 2006;40:1754–61.

[28] Lange AK, Fiatarone Singh MA, Smith RM, Foroughi N, Baker MK, Shnier R, et al.Degenerative meniscus tears and mobility impairment in women with kneeosteoarthritis. Osteoar Cartil 2007;15:701–8.

[29] Altman RD. Criteria for the classification of osteoarthritis of the knee and hip. Scand JRheumatol 1987;65:31–9.

[30] Outerbridge RE. The etiology of chondromalacia patella. J Bone Joint Surg1961;43B:752–7.

[31] Kadaba MP, Ramakrishnan HK, Wootten ME. Measurement of lower extremitykinematics during level walking. J Orthop Res 1990;8:383–92.

[32] Wu G, Cavanagh PR. ISB recomendations for standardization in the reporting ofkinematic data. J Biomech 1995;28:1257–61.

[33] Vanwanseele B, Parker D, Coolican M. Frontal knee alignment: three-dimensionalmarker positions and clinical assessment. Clin Orthop Relat Res 2009;467:504–9.

[34] Marfell-Jones M, Olds T, Stewart A, Carter JEL. International standards for anthro-pometric assessment; 2006.

[35] de VosNJ, SinghNA, Ross DA, Stavrinos TM, Orr R, Fiatarone SinghMA. Optimal load forincreasingmuscle power during explosive resistance training in older adults. J GerontolA Biol Sci Med Sci 2005;60:638–47.

[36] Mattacola CG, Lebsack DA, Perrin DH. Intertester reliability of assessing posturalsway using the Chattecx Balance System. J Athl Train 1995;30:237–9.

[37] Orr R, De Vos NJ, Singh NA, Ross DA, Stavrinos TM, Fiatarone Singh MA. Powertraining improves balance in healthy older adults. J Gerontol A Biol Sci Med Sci2006;61A:78–85.

[38] Bellamy N, Buchanan W, Goldsmith CH, Campbell J, Stitt LW. Validation study ofWOMAC: a health status instrument for measuring clinically important patientrelevant outcomes to antirheumatic drug therapy in patients with osteoarthritis ofthe hip or the knee. J Rheumatol 1988;15:1833–40.

[39] Washburn RA, Smith KW, Jette AM, Janney CA. The Physical Activity Scale for theElderly (PASE): development and evaluation. J Clin Epidemiol 1993;46:153–62.

[40] Paffenbarger RS, Wing AL, Hyde RT. Physical activity as an index of heart attack risk incollege alumni. Am J Epidemiol 1978:108.

[41] Stewart AL, Hays RD, Ware JE Jr. The MOS short-form general health survey.Reliability and validity in a patient population. Med Care 1988;26:724–35.

[42] ChangA, Hayes K, DunlopD, Hurwitz D, Song J, Cahue S, et al. Thrust during ambulationand the progression of knee osteoarthritis. Arth Rheum 2004;50:3897–903.

d its association with knee adduction moment in medial knee

7N. Foroughi et al. / The Knee xxx (2009) xxx–xxx

ARTICLE IN PRESS

[43] Weidenhielm L, Svensson OK, Brostrom LA, Mattsson E. Adduction moment of theknee compared to radiological and clinical parameters in moderate medicalosteoarthrosis of the knee. Ann Chir Gynaecol 1994;83:236–42.

[44] Wang JW, Kuo KN, Andriacchi TP, Galante JO. The influence of walking mechanics andtime on the results of proximal tibial osteotomy. J Bone Jt Surg 1990;72:905–9.

[45] Wada M, Imura S, Nagatani K, Baba H, Shimada S, Sasaki S. Relationship betweengait and clinical results after high tibial osteotomy. Clin Orthop 1998;354:180–8.

[46] Chang A, Hurwitz D, Dunlop D, Song J, Cahue S, Hayes K. The relationship between toe-out angle during gait and progression ofmedial tibiofemoral osteoarthritis. Ann RheumDis 2007;66:1271–5.

[47] Paley D. Principles of deformity correction. New York: Springer; 2002.[48] Lewek MD, Rudolph KS, Snyder-Mackler L. Control of frontal plane knee laxity during

gait in patients with medial compartment knee osteoarthritis. Osteoar Cartil2004;12:745–51.

[49] Bejek Z, Paroczai R, Illyes A, Kiss RM. The influence ofwalking speed on gait parametersin healthy people and in patients with osteoarthritis. Knee Surg Sports TraumatolArthrosc 2006;14:612–22.

Please cite this article as: Foroughi N, et al, Dynamic alignment anosteoarthritis, The Knee (2009), doi:10.1016/j.knee.2009.09.006

[50] Shakoor N, Block JA. Walking barefoot decreases loading on the lower extremityjoints in knee osteoarthritis. Arth Rheum 2006;54:2923–7.

[51] Maly MR, Costigan PA, Olney SJ. Role of knee kinematics and kinetics on performanceand disability in people with medial compartment knee osteoarthritis. Clin Biomech2006;21:1051–9.

[52] OstendorfM, Van Stel HF, Buskens E, Schrijvers JP,Marting LN, Verbout AJ, et al. Patient-reported outcome in total hip replacemet. A comparison of five instruments of healthstatus. J Bone Joint Surg Br 2004;86:801–8.

[53] Bachmeier CJM, March LM, Cross MJ, Lapsley HMT, Tribe KL, Courtenay BG, et al. Acomparison of outcomes in osteoarthritis patients undergoing total hip and kneereplacement surgery. Osteoarthr Cartil 2001;9:137–46.

[54] Lieberman JR, Dorey F, Shekelle P, Schumacher L, Kilgus DJ, Thomas BJ, et al.Outcome after total hip arthroplasty. Comparison of a traditional disease-specificand quality-of life measurement of outcome. J Arthroplast 1997;6:639–45.

d its association with knee adduction moment in medial knee