1. Top of page
  2. Abstract


To evaluate whether increased lateral trunk lean toward the symptomatic lower extremity during gait in people with medial knee osteoarthritis (OA) immediately alters symptoms or medial knee load, as measured by the external knee adduction moment (KAM).


Participants with medial knee OA (n = 22) underwent 3-dimensional gait analysis to measure KAM peaks (early and late stance) and KAM impulse. Following the analysis of natural gait, participants were trained to lean their trunk toward the symptomatic leg during ipsilateral stance over 3 randomly ordered conditions (6°, 9°, and 12° lean). A projection screen displayed real-time trunk angles and target levels. Pain/discomfort in the knees, the hip, and the back were measured across conditions. Load-modifying effects of increasing lean magnitudes were investigated using linear mixed models. Mediating effects of peak lean timing and participant characteristics (pain and malalignment) were evaluated.


Increased trunk lean reduced all KAM measures (P < 0.001), with larger lean angles achieving greater reductions. Efficacy of load reduction improved with later peak lean timing for all measures of the KAM. Participant characteristics did not mediate the effect of trunk lean on the KAM, and symptoms did not change across conditions (P > 0.05).


Increased trunk lean reduced medial knee load in a dose-response manner. Slightly later achievement of peak trunk lean improved the load-modifying effect of this gait strategy. No immediate symptomatic changes were identified. Future research should determine if long-term implementation of this gait strategy is feasible and whether it can modify disease symptoms and OA progression.


  1. Top of page
  2. Abstract

Knee osteoarthritis (OA) is a common chronic joint condition affecting older people (1). The disease causes a large personal and societal burden (2), which is expected to increase with the aging population and obesity crisis. As knee joint structural deterioration occurs in many people (3), and a cure for OA remains to be discovered (4), therapies with potential to slow disease progression are the focus of current research efforts. Increased medial knee compartment loading is a risk factor for medial knee OA progression (5, 6). As excessive joint load is one of few potentially modifiable risk factors (7), treatments that attempt to reduce load in the medial tibiofemoral compartment are warranted (8–10).

Medial knee load is typically evaluated noninvasively via 3-dimensional (3-D) gait analysis. The variable of most interest is the external knee adduction moment (KAM), which acts to rotate the tibia in a varus direction. The external KAM is a reliable and valid indicator of medial knee load (11, 12) and is most often quantified by evaluating the 2 peaks during stance, as well as the impulse (area under the KAM-time curve). The first and typically largest peak occurs during the load acceptance phase of gait (approximately 25% stance), and the second peak occurs during the propulsion phase (approximately 75% stance) (13–15). Clinical implications of a large KAM are significant as the peak KAM is a strong predictor of medial compartment OA radiographic disease severity (16), rate of disease progression (6), and development of chronic knee pain (17). The KAM impulse has also received increasing research focus as it reflects both the mean KAM magnitude and the absolute duration of load (i.e., stance) (5, 18, 19). Due to the strong clinical relevance of elevated KAM parameters, the success of load-modifying interventions is typically evaluated using measures of the KAM (9).

Some conservative gait modifications have been shown to reduce the KAM (9, 20). Of these, increasing lateral trunk lean (or lateral flexion) toward the symptomatic knee during stance appears to give the greatest reduction in the early stance peak KAM (10). This strategy has been evaluated in 2 studies of healthy young individuals, both demonstrating significant load reductions (10, 21). Increasing trunk lean is postulated to modify knee load by changing the frontal plane center of mass location, subsequently reducing the ground reaction force lever arm at the knee and thus the KAM (22). Although research shows promising results, selective performance of this modification and effective load reduction is yet to be demonstrated in the medial knee OA population. Furthermore, the timing (or phasing) of peak lateral trunk lean is a characteristic of the modification with potential to influence the strategy's load-reducing ability and requires further analysis. Participant-related factors known to influence the KAM magnitude, such as pain severity and knee malalignment (23, 24), should also be assessed for mediating effects of this gait modification on medial knee joint load.

The primary aim of this study was to evaluate the immediate effect of increased lateral trunk lean on medial knee load, as measured by the KAM, in people with medial knee OA. Particularly, we wished to determine the effect of varying trunk lean magnitudes on the KAM. The secondary aim was to determine whether specific participant and gait modification characteristics (including pain severity, mechanical knee alignment, and timing of peak lean) influence load-modifying effects of the strategy. Finally, the effect of ipsilateral trunk lean on acute knee, hip, and back symptoms were investigated to determine symptom-modifying potential and possible adverse effects.

Significance & Innovations

  • In people with medial knee osteoarthritis, ipsilateral trunk lean gait modification significantly reduced all measures of the knee adduction moment in a dose-response manner.

  • No immediate decreases in knee symptoms or increases in hip or back symptoms were identified with this gait modification.


  1. Top of page
  2. Abstract


Individuals with medial knee OA were recruited from the community via advertisements. Participants were included if they fulfilled the American College of Rheumatology clinical and radiographic criteria for knee OA (25) and experienced average knee pain on most days of the previous month (level ≥3 on an 11-point numerical rating scale). Using a radiographic atlas (26) to grade OA features in each compartment, only participants with predominantly medial tibiofemoral OA features were included (determined by greater medial osteophyte presence or greater medial joint space narrowing in the case of equal osteophyte grades). Exclusion criteria included history of knee or hip surgery, knee arthroscopy or injection in the previous 6 months, cardiorespiratory instability, neurologic conditions affecting ambulation, gait aid use, presence of other rheumatologic conditions, significant spinal pain with associated lower extremity symptoms, body mass index >35 kg/m2, and anatomic valgus knee malalignment (≥185°) on radiographs (27). For participants with bilaterally eligible knees, the most symptomatic side was considered the study extremity. This study was approved by the Human Research Ethics Committee, The University of Melbourne, Australia. All participants provided written informed consent.

Measurement of kinematics and kinetics.

Participants underwent 3-D gait analysis in their own shoes under 4 gait conditions in a single session. First, the natural (self-selected) gait condition was recorded, followed by the increased trunk lean gait conditions implemented under 3 predetermined magnitudes of lean. A Vicon motion analysis system captured gait kinematics using 8 MX cameras recording at 120 Hz, which was integrated with 3 force platforms (Advanced Mechanical Technology) embedded in the laboratory floor to collect ground reaction force data at 1,200 Hz. The standard Plug-In-Gait lower body marker set (Vicon) was used, with 20 retro-reflective markers adhered to anatomic landmarks (28). Medial knee and ankle markers were included during an initial static standing trial to determine positioning of knee and ankle joint centers. Three additional markers were placed on the thorax (anteriorly on the manubrium, posteriorly on the spinous processes of T2 and T10) to monitor trunk motion, as recommended in the literature (29, 30). This method of trunk lean angle evaluation is considered optimal as it demonstrated the lowest errors while simultaneously excluding the confounding effects of shoulder and pelvic motion (30). The same method has been used in similar research (21, 31).

The Vicon Plug-In-Gait model (version 2) in Vicon Nexus software was used to calculate the external KAM about an orthogonal axis system located in the shank segment. Knee load was quantified by the early and late stance KAM peaks and KAM impulse. Moments were normalized to body weight times height (Nm/%BW × Ht) and expressed as percentage of stance, except for the KAM impulse. The peak lateral trunk lean angle from the vertical toward the study (ipsilateral) extremity was calculated using a custom-written BodyBuilder model (Vicon), where frontal plane angle in relation to the laboratory coordinate system was the first to be calculated in the Cardan angle sequence describing trunk orientation. Other variables evaluated because of their potential to influence the KAM were gait speed, step width, and stride length (9).

Walking with increased trunk lean.

A physiotherapist (MS) trained participants to walk with increased lean. Three trunk lean conditions were implemented, with participants instructed to reach a peak of 6° lean, 9° lean, and 12° lean. Testing order was randomized for each participant using a Latin square matrix. Trunk lean magnitudes were selected based on previous research conducted on healthy young individuals (21), where magnitudes of lean implemented were 4°, 8°, and 12°. There is evidence of naturally larger trunk lean values in the medial knee OA population, with participants adopting an average peak trunk lean of 2.3° in mild OA, 3.1° in moderate OA, and 5.0° in severe OA (31). Consequently, trunk lean condition magnitudes selected for this study commenced at 6° to ensure the participants leaned farther than most would perform naturally.

Participants were instructed to lean their trunk toward the study extremity during ipsilateral stance phase and to reach maximum trunk lean to the desired level as soon as possible after initial contact of the symptomatic extremity. Usual trunk motion during the contralateral stance phase was encouraged. Participants were instructed to imagine bringing their whole upper body over the stance extremity, rather than laterally flexing the upper trunk (i.e., shoulder depression). Several motor learning principles were implemented during training by the physiotherapist to aid skill acquisition, including standardized instructions, demonstration, feedback via verbal and tactile means, and visual feedback using a full-length mirror (32, 33).

A qualitative movement checklist was used by the physiotherapist to determine if appropriate performance of the modification was achieved. Participants were then trained to achieve the desired magnitude of lean using a real-time visual biofeedback system (34). The biofeedback involved representation of the real-time frontal plane trunk angle, displayed as a scrolling trace on the horizontal axis against time in the vertical axis (Figure 1). Trunk marker positions were calculated by the Vicon Nexus (version 1.4) software in real time and streamed to custom Matlab software (version R2009, The Mathworks) via transmission control protocol/internet protocol. The frontal plane trunk angle was calculated and displayed using the Matlab software. This biofeedback system has demonstrated feasibility in the training of healthy individuals to modify their gait with increased trunk lean (21).

thumbnail image

Figure 1. The real-time movement biofeedback system used for training and walking with increased lateral trunk lean. Participants walked toward the projection screen, which displayed the trunk frontal plane angle in real time (purple trace), as well as the desired target to be achieved. The target area was represented by a green band (corresponding to the desired angle ±1°). Participants were instructed to reach the green target area with the purple trace as soon as possible after initial contact with the ground of the study extremity.

Download figure to PowerPoint

Repeated training and practice occurred (typically 10–20 minutes) until participants had mastered the technique with the desired lateral trunk lean magnitude for the condition. Data collection then commenced. Data capture required 5 good trials per condition, ensuring appropriate execution of the gait modification while adhering to a gait speed similar to the natural walking trials (±5% as measured from photoelectric timing gates placed 4 meters apart on either side of the force platforms). In cases where the target trunk lean angle or gait speed was not achieved, the physiotherapist provided additional verbal feedback and prompted participants to continue attempting the target until they required a rest. Because of difficulty in obtaining prescribed trunk lean magnitudes exactly, trials were not excluded based on the peak trunk motion achieved, and the closest 5 trials to the target were included in the analyses.

Descriptive measures.

Standardized semiflexed anteroposterior knee radiographs were obtained and evaluated by a single reviewer (MS). Disease severity was rated using the Kellgren/Lawrence grading scale (35). Anatomic knee alignment was measured and converted to mechanical alignment using a published regression equation (27). Self-reported pain and physical function were assessed using the Western Ontario McMaster Universities OA Index (WOMAC), where higher scores indicate worse pain and poorer function, respectively (36). The WOMAC pain subscale ranges from 0–20, and the physical function subscale ranges from 0–68.

Affect of increased trunk lean on symptoms.

An 11-point numerical rating scale was used to evaluate pain and discomfort experienced during each condition at both knees, the ipsilateral hip, and lower back (range 0–10, with 0 representing no pain/discomfort and 10 representing worst pain/discomfort imaginable).

Statistical analysis.

Analyses were performed using GenStat (13th edition, VSN International) with an alpha level set at 0.05. All data were checked for normality and homogeneity of variance prior to analyses. Descriptive information was obtained via means and 95% confidence intervals, as well as SDs and frequencies where appropriate. Repeated-measures analysis of variance (ANOVA) was used to determine if differences existed between conditions among independent variables (peak lateral trunk lean angles), to evaluate changes in spatiotemporal gait variables (speed, stride length, and width), and changes in symptoms. Where results were significant, an evaluation of least significant differences between conditions was performed to locate the change.

To evaluate the effect of trunk lean on knee load, restricted maximum likelihood linear mixed modeling was conducted (37), with participants considered as the random factor. Outcome variables were the KAM early and late stance peaks and impulse. The primary independent variable was the peak lateral trunk lean angle toward the study (ipsilateral) extremity. To determine if trunk lean characteristics influenced the load-modifying ability of trunk lean, timing of peak lateral trunk lean was included in subsequent models as a fixed factor. Fixed factors were initially examined for interactions with the independent variable. If significance was not reached with either interaction or addition of variables, the factor was deemed insignificant to the outcome and excluded from the final model. Main effects of all significant variables were evaluated as part of the model. Interpretation of interaction terms was conducted using simple slope tests. Interactions between the independent variable with mechanical knee alignment and WOMAC pain were assessed to determine if they influenced efficacy of load reduction.


  1. Top of page
  2. Abstract

Twenty-two people (13 female, 9 male) participated in this study. Three additional participants who were initially eligible were excluded because they were unable to coordinate ipsilateral trunk lean during stance, and thus could not perform the appropriate gait pattern during training. Participant characteristics are shown in Table 1. Generally, the cohort was overweight with knee varus malalignment. Radiographic knee OA severity and disease symptoms were predominantly mild to moderate in this cohort.

Table 1. Participant characteristics (n = 22)*
  • *

    Values are the mean ± SD unless indicated otherwise. BMI = body mass index; WOMAC = Western Ontario and McMaster Universities Osteoarthritis (OA) Index; K/L = Kellgren/Lawrence.

  • Mechanical alignment derived from radiographic anatomic knee alignment.

Age, years68.4 ± 10.2
Height, meters1.67 ± 0.09
Mass, kg78.3 ± 16.1
BMI, kg/m227.9 ± 4.8
Mechanical knee alignment, degrees178.1 ± 3.9
Symptom duration, years9.3 ± 10.3
Nature of symptoms, no. (%) 
 Unilateral7 (32)
 Bilateral15 (68)
Sex, no. (%) 
 Female13 (59)
 Male9 (41)
WOMAC score 
 Pain (range 0–20)5 ± 3
 Physical function (range 0–68)19 ± 11
K/L radiographic OA severity, no. (%) 
 Grade 2 (mild)9 (41)
 Grade 3 (moderate)9 (41)
 Grade 4 (severe)4 (18)

Discrete gait values and self-perceived measures of knee and back pain/discomfort across gait conditions are reported in Table 2. On average, participants achieved the required peak trunk lean magnitudes toward the study extremity. Differences in trunk lean values across conditions were significant (P < 0.001). Peak lateral trunk lean during natural gait occurred at midstance (51.6% stance), while participants achieved peak lateral trunk lean during the early stance phase of gait for all gait modification conditions (37.7–38.5% stance). As timing of peak lean coincided more closely with the KAM early stance peak (24.6–26.6% stance) than late stance (77.1–77.8% stance), trunk lean values at the time of early stance KAM (mean 5.09°, 7.55°, and 9.26° per condition) were greater than trunk lean values during the late stance KAM peak (mean 3.01°, 4.38°, and 5.56° per condition). Speed and stride length remained unchanged across conditions. Stride width increased with gait modification (P < 0.001), with differences identified between the natural gait condition and all gait modification conditions. All symptoms remained unchanged across conditions (P < 0.05).

Table 2. Descriptive data relating to natural gait and lateral trunk lean gait conditions*
 Natural gaitAttempted 6° leanAttempted 9° leanAttempted 12° lean
  • *

    Values are the mean (95% confidence interval). KAM = knee adduction moment; BW = body weight; Ht = height; N/A = not applicable.

Lateral trunk lean, degrees    
 Peak lateral trunk lean angle2.0 (1.4, 2.7)6.1 (5.5, 6.7)8.7 (8.1, 9.4)11.1 (10.5, 11.8)
 Trunk lean at early stance peak KAM0.9 (0.1, 1.6)5.1 (4.4, 5.8)7.6 (6.8, 8.3)9.3 (8.5, 10.0)
 Trunk lean at late stance peak KAM0.8 (0.0, 1.7)3.0 (2.2, 3.9)4.4 (3.5, 5.2)5.6 (4.7, 6.4)
Gait characteristics    
 Speed, meters/second1.24 (1.16, 1.32)1.25 (1.17, 1.33)1.24 (1.16, 1.32)1.23 (1.15, 1.31)
 Stride length, meters1.35 (1.29, 1.41)1.33 (1.27, 1.39)1.34 (1.28, 1.40)1.34 (1.28, 1.40)
 Stride width, meters0.10 (0.08, 0.11)0.11 (0.10, 0.13)0.12 (0.11, 0.14)0.12 (0.11, 0.14)
Knee load (Nm/%BW × Ht)    
 Early stance peak KAM3.75 (3.28, 4.22)3.40 (2.93, 3.87)3.33 (2.86, 3.80)3.19 (2.72, 3.66)
  Change in early stance peak KAM from natural gaitN/A0.35 (0.12, 0.58)0.43 (0.21, 0.65)0.56 (0.29, 0.83)
 Late stance peak KAM2.05 (1.68, 2.42)1.71 (1.34, 2.08)1.69 (1.32, 2.06)1.56 (1.19, 1.93)
  Change in late stance peak KAM from natural gaitN/A0.35 (0.18, 0.51)0.37 (0.19, 0.55)0.49 (0.30, 0.69)
 KAM impulse (Nm.s/%BW × Ht)1.22 (1.00, 1.44)1.05 (0.83, 1.27)1.03 (0.81, 1.25)0.96 (0.74, 1.18)
  Change in KAM impulse from natural gaitN/A0.17 (0.11, 0.24)0.20 (0.13, 0.27)0.26 (0.19, 0.34)
Timing during stance, % stance    
 Peak ipsilateral trunk lean51.6 (46.9, 56.2)37.7 (33.2, 42.2)37.7 (33.2, 42.2)38.5 (34.0, 43.0)
 Early stance peak KAM26.6 (25.0, 28.2)25.2 (23.6, 26.8)25.1 (23.5, 26.7)24.6 (23.0, 26.2)
 Late stance peak KAM77.7 (74.4, 81.1)77.1 (73.8, 80.3)77.8 (74.5, 81.0)77.7 (74.4, 81.0)
Pain/discomfort, range 0–10    
 Affected knee2.2 (1.3, 3.1)2.3 (1.4, 3.6)2.2 (1.3, 3.2)2.1 (1.3, 3.0)
 Contralateral knee1.1 (0.5, 1.8)1.0 (0.5, 1.6)1.1 (0.4, 1.8)1.1 (0.5, 1.8)
 Ipsilateral hip0.2 (0.0, 0.4)0.6 (0.0, 1.2)0.6 (0.0, 1.3)0.6 (0.0, 1.3)
 Lower back0.6 (0.0, 1.2)1.0 (0.2, 1.7)0.8 (0.3, 1.3)0.8 (0.2, 1.3)

A dose-response effect of lateral trunk lean on all KAM outcomes was evident, as illustrated by the ensemble average graphs of trunk lean angles and KAM for each condition (Figure 2). Linear mixed models demonstrating the effect of increased trunk lean show that trunk lean significantly reduced all measures of the KAM (Table 3). Results imply that larger peak trunk lean angles achieved greater reductions in the early stance peak KAM (P < 0.001), the late stance peak KAM (P < 0.001), and the KAM impulse (P < 0.001).

thumbnail image

Figure 2. Ensemble averages for A, lateral trunk lean angles, and B, external knee adduction moment during the stance phase for each of the 4 conditions: natural gait (solid line), small trunk lean (attempted 6°; dotted line), medium trunk lean (attempted 9°; broken line), and large trunk lean (attempted 12°; broken/dotted line). BW = body weight; Ht = height.

Download figure to PowerPoint

Table 3. Effect of peak trunk lean angle on medial knee loading parameters*
Linear mixed modelEstimate95% CIProbability
  • *

    95% CI = 95% confidence interval; KAM = knee adduction moment.

Early stance peak KAM   
 Intercept3.412.93, 3.89 
 Peak trunk lean angle−5.76 × 10−2−6.87 × 10−2, −4.65 × 10−2< 0.001
Late stance peak KAM   
 Intercept1.741.38, 2.10 
 Peak trunk lean angle−4.87 × 10−2−5.73 × 10−2, −4.01 × 10−2< 0.001
KAM impulse   
 Intercept1.060.84, 1.28 
 Peak trunk lean angle−2.59 × 10−2−2.91 × 10−2, −2.27 × 10−2< 0.001

Variables contributing to the efficacy of load reduction are shown in Table 4. Results demonstrated that timing of peak trunk lean influenced all KAM variables, whereas participants' knee malalignment and WOMAC pain scores did not. Main effects of trunk lean timing demonstrate greater load reductions with earlier lean achievement for the early stance peak KAM (P < 0.001) and KAM impulse (P = 0.016). However, interactions identified between timing of trunk lean and peak trunk lean angle resulted in minimal, but statistically significant, improvements in load reduction with later achievement of peak trunk lean, negating the main effects. This interaction was identified for the early stance peak KAM (P = 0.023), late stance peak KAM (P = 0.012), and KAM impulse (P = 0.001).

Table 4. Final linear mixed models for the effects of increased trunk lean on medial knee loading parameters, with consideration of participant and trunk lean modification characteristics*
Linear mixed modelEstimate95% CIProbability
  • *

    95% CI = 95% confidence interval; KAM = knee adduction moment.

Early stance peak KAM   
 Intercept3.402.91, 3.89 
  Interaction terms   
   Peak trunk lean × timing of peak trunk lean−7.08 × 10−4−1.33 × 10−3, −8.80 × 10−50.023
  Main effects   
   Peak trunk lean−5.06 × 10−2−6.20 × 10−2, −3.92 × 10−2< 0.001
   Timing of peak trunk lean3.64 × 10−35.80 × 10−4, 6.70 × 10−3< 0.001
Late stance peak KAM   
 Intercept1.731.37, 2.09 
  Interaction terms   
   Peak trunk lean × timing of peak trunk lean−6.08 × 10−4−1.09 × 10−3, −1.20 × 10−40.012
  Main effects   
   Peak trunk lean−4.80 × 10−2−5.69 × 10−2, −3.91 × 10−2< 0.001
   Timing of peak trunk lean−2.23 × 10−3−4.63 × 10−3, 1.70 × 10−40.379
KAM impulse   
 Intercept1.050.83, 1.27 
  Interaction terms   
   Peak trunk lean × timing of peak trunk lean−2.87 × 10−4−4.66 × 10−4, −1.08 × 10−40.001
  Main effects   
   Peak trunk lean−2.42 × 10−2−2.75 × 10−2, −2.09 × 10−2< 0.001
   Timing of peak trunk lean3.63 × 10−4−5.19 × 10−4, 1.25 × 10−30.016

The final linear mixed models, representing the effect of increased trunk lean on KAM variables, can be written using the following equations, where a = peak lateral trunk lean (degrees) and b = timing of peak trunk lean (% stance):

  • equation image
  • equation image
  • equation image


  1. Top of page
  2. Abstract

Results of this study demonstrate that increased lateral trunk lean toward the symptomatic lower extremity significantly reduces knee load throughout stance in people with medial knee OA. A dose-response relationship was observed, with larger lean angles leading to greater reductions in KAM variables. Timing of peak trunk lean during stance influenced the modification's load-reducing ability, while pain severity and malalignment did not. Increased trunk lean during gait did not immediately affect symptoms at the knees, hip, or back in this medial knee OA cohort.

Increased lateral trunk lean significantly reduced the KAM in our sample of medial knee OA participants, which is consistent with previous research in healthy individuals (10, 21). This is the first study to implement the trunk lean gait modification in a knee OA sample and evaluate load-modifying effects with increasing magnitudes of lean. On average, our cohort of OA participants achieved early stance peak KAM reductions of 9.3% (0.35 Nm/%BW × Ht) with a peak 6.1° lean, 11.5% (0.43 Nm/%BW × Ht) reduction with 8.7° lean, and 14.9% (0.56 Nm/%BW × Ht) reduction with 11.1° lean. When variability due to participants was considered using the linear mixed model, implementation of 10° of trunk lean is predicted to reduce the early stance KAM by 16.4% on average.

Early stance peak KAM reductions achieved in this study were similar to the 19.9% reductions reported with a peak 12.9° lean in healthy individuals (21), but considerably lower than the 65.0% reduction reported with 10° of lean in another study of healthy individuals (10). The reason for such discrepancy in findings is unknown; however, several methodologic differences among the studies may have contributed. The greater effects demonstrated by Mundermann et al (10) may be due to: 1) their implementation of trunk lean bilaterally (compared to ipsilaterally in the present study), 2) the self-selected magnitude of lean, resulting in an average peak trunk lean angle of 10° with large variability (5° SD), and 3) potentially different timing of trunk lean (not reported), which may have altered the gait modification effect size. Furthermore, the trunk motion evaluation method used by Mundermann et al (10) calculated frontal plane trunk angle as the projection of the angle between a line intersecting the midpoints between anterior superior iliac spines and the scapular acromion processes and the global vertical axis. This method of trunk lean motion cannot exclude the confounding motion of the shoulder and pelvis as readily in comparison to the method chosen for this study. The 10° of lean reported may have possibly underestimated the true lean performed by participants as scapular depression or increased pelvic obliquity may alter the result. Finally, due to biomechanical differences between healthy and knee OA individuals (38), some variability between these study findings and Mundermann et al (10) may be explained. Nevertheless, increased trunk lean has consistently proven effective in reducing medial knee load. As the direct dose-response effect was observed for each KAM parameter, perhaps the largest tolerable trunk lean (up to 12°) should be advocated for optimal load-modifying effects in future research. Trunk lean angles higher than those evaluated in this study may possibly create greater load-modifying effects; however, this should be confirmed by future research.

Achievement of peak trunk lean later in stance minimally increased the load-reducing effect of lateral trunk lean in the range of values implemented. It is difficult to ascertain the reason for such an effect; however, it is possible that participants with delayed timing of lean were subsequently able to maintain greater lean angles over the remaining part of stance. The influence of gait modification timing is similar to the effect identified with contralateral cane use, where timing of the peak body weight support through the cane influenced the cane's load-modifying ability (39). Because neither knee pain nor malalignment influenced load-modifying ability of trunk lean, this modification may be uniformly effective for a range of participant symptoms and malalignment. However, the absence of participant-related mediating effects should be confirmed in larger trials as this study may have been inadequately powered to detect such effects.

According to the literature, an increase in KAM of 1 unit Nm/%BW × Ht (20% of the mean group KAM) is associated with a 6-fold increased risk of radiographic knee OA progression (6). The mean early stance peak KAM reduction of 0.56 Nm/%BW × Ht, achieved with 12° trunk lean, may theoretically slow medial knee OA progression over time.

Greater trunk lean range of motion may be anticipated in knee OA participants (40). Naturally occurring ipsilateral trunk lean magnitude has been shown to increase with pain severity (41) and disease severity, where an average lean of 2.3° was reported in mild OA participants and 5.0° observed in participants with severe OA (31). The naturally occurring trunk lean in our sample of participants was an average 2.0° lean, comparable to previous study findings given the predominantly mild to moderate knee OA severity. Results of this study support that altered trunk lean motion in knee OA participants may be a naturally occurring compensatory mechanism for load reduction in the medial compartment, as hypothesized by previous research (31, 42, 43).

Increased trunk lean did not result in immediate changes in the symptoms of the most symptomatic knee. Importantly, this strategy did not immediately increase or aggravate symptoms at the contralateral knee, ipsilateral hip, or lower back. However, immediate symptomatic changes may not be expected with gait modifications, as seen with cane use in a similar cohort (39). Although this is the first study to evaluate symptomatic effects, potentially adverse symptomatic and biomechanical effects on other joints and dynamic balance changes with large magnitudes of lean warrant future investigations.

Biomechanically, participants with bilateral medial knee OA may achieve load reduction in both knees with bilateral trunk lean gait in preference to the ipsilateral lean gait. Conversely, participants with a diseased lateral compartment may not be optimal candidates for this intervention, as a compressive load shift to the lateral compartment may be detrimental. Results should be considered in light of participant willingness and ability to modify gait over time. Inappropriate performance of this strategy, such as contralateral lean during stance, may lead to increases in medial compartment load due to a medial shift in the center of mass. Also, if participants maintained the prescribed lean during contralateral stance phase, the nonstudy extremity may undergo medial knee load increases. This is particularly relevant as 3 participants could not perform the strategy, signifying that clinical application of ipsilateral lean may be difficult for some. Longitudinal evaluation is required to determine feasibility of long-term gait modification and possibility of symptomatic changes.

This study possesses several key strengths. This is the first study (to our knowledge) to evaluate achievability and effects of isolated trunk lean in medial knee OA participants. Second, the strict testing protocol ensured appropriate and consistent gait modification performance. Third, our real-time biofeedback system enabled the successful implementation of specific trunk lean magnitudes (34). Finally, the linear mixed model statistical method is considered more powerful than regression analysis or repeated-measures ANOVA (37, 44). The approach allowed consideration of random variability attributed to participants and the inclusion of all data points. There are also some limitations to this study. Due to the use of 3-D gait analysis and specialized biofeedback equipment, application of this training approach to the clinical setting is currently limited. Additionally, only immediate biomechanical and symptomatic effects were evaluated, and future research should expand knowledge through longitudinal research.

In conclusion, this study shows that increased lateral trunk lean is a feasible gait modification that reduces medial knee load. A dose-response relationship was identified. Results suggest that timing of peak trunk lean should occur later during the early stance phase to increase efficacy of load reduction. Longitudinal evaluations of increased lateral trunk lean are now required to evaluate effects on knee joint structure and disease symptoms.


  1. Top of page
  2. Abstract

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Ms Simic had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Simic, Hunt, Bennell, Hinman, Wrigley.

Acquisition of data. Simic, Wrigley.

Analysis and interpretation of data. Simic, Hunt, Bennell, Hinman, Wrigley.


  1. Top of page
  2. Abstract
  • 1
    Peat G, McCarney R, Croft P. Knee pain and osteoarthritis in older adults: a review of community burden and current use of primary health care. Ann Rheum Dis 2001; 60: 917.
  • 2
    Access Economics. Painful realities: the economic impact of arthritis in Australia 2007. Canberra: Access Economics Pty; 2007.
  • 3
    Cooper C, Snow S, McAlindon TE, Kellingray S, Stuart B, Coggon D, et al. Risk factors for the incidence and progression of radiographic knee osteoarthritis. Arthritis Rheum 2000; 43: 9951000.
  • 4
    Towheed TE, Maxwell L, Judd MG, Catton M, Hochberg MC, Wells G. Acetaminophen for osteoarthritis. Cochrane Database Syst Rev 2006:CD004257.
  • 5
    Bennell KL, Bowles KA, Wang Y, Cicuttini F, Davies-Tuck M, Hinman RS. Higher dynamic medial knee load predicts greater cartilage loss over 12 months in medial knee osteoarthritis. Ann Rheum Dis 2011; 70: 17704.
  • 6
    Miyazaki T, Wada M, Kawahara H, Sato M, Baba H, Shimada S. Dynamic load at baseline can predict radiographic disease progression in medial compartment knee osteoarthritis. Ann Rheum Dis 2002; 61: 61722.
  • 7
    Felson DT, Lawrence RC, Dieppe PA, Hirsch R, Helmick CG, Jordan JM, et al. Osteoarthritis: new insights. Part 1: the disease and its risk factors. Ann Intern Med 2000; 133: 63546.
  • 8
    Zhang W, Moskowitz RW, Nuki G, Abramson S, Altman RD, Arden N, et al. OARSI recommendations for the management of hip and knee osteoarthritis, part I: critical appraisal of existing treatment guidelines and systematic review of current research evidence. Osteoarthritis Cartilage 2007; 15: 9811000.
  • 9
    Simic M, Hinman RS, Wrigley TV, Bennell KL, Hunt MA. Gait modification strategies for altering medial knee joint load: a systematic review. Arthritis Care Res (Hoboken) 2011; 63: 40526.
  • 10
    Mundermann A, Asay JL, Mundermann L, Andriacchi TP. Implications of increased medio-lateral trunk sway for ambulatory mechanics. J Biomech 2008; 41: 16570.
  • 11
    Birmingham TB, Hunt MA, Jones IC, Jenkyn TR, Giffin JR. Test–retest reliability of the peak knee adduction moment during walking in patients with medial compartment knee osteoarthritis. Arthritis Rheum 2007; 57: 10127.
  • 12
    Zhao D, Banks SA, Mitchell KH, D'Lima DD, Colwell CW, Fregly BJ. Correlation between the knee adduction torque and medial contact force for a variety of gait patterns. J Orthop Res 2007; 25: 78997.
  • 13
    Davies-Tuck ML, Wluka AE, Teichtahl AJ, Martel-Pelletier J, Pelletier JP, Jones G, et al. Association between meniscal tears and the peak external knee adduction moment and foot rotation during level walking in postmenopausal women without knee osteoarthritis: a cross-sectional study. Arthritis Res Ther 2008; 3: 10.
  • 14
    Hunt MA, Birmingham TB, Giffin JR, Jenkyn TR. Associations among knee adduction moment, frontal plane ground reaction force, and lever arm during walking in patients with knee osteoarthritis. J Biomech 2006; 39: 221320.
  • 15
    Thorp LE, Sumner DR, Block JA, Moisio KC, Shott S, Wimmer MA. Knee joint loading differs in individuals with mild compared with moderate medial knee osteoarthritis. Arthritis Rheum 2006; 54: 38429.
  • 16
    Sharma L, Hurwitz DE, Thonar EJ, Sum JA, Lenz ME, Dunlop DD, et al. Knee adduction moment, serum hyaluronan level, and disease severity in medial tibiofemoral osteoarthritis. Arthritis Rheum 1998; 41: 123340.
  • 17
    Thorp LE, Sumner DR, Wimmer MA, Block JA. Relationship between pain and medial knee joint loading in mild radiographic knee osteoarthritis. Arthritis Rheum 2007; 57: 125460.
  • 18
    Kito N, Shinkoda K, Yamasaki T, Kanemura N, Anan M, Okanishi N, et al. Contribution of knee adduction moment impulse to pain and disability in Japanese women with medial knee osteoarthritis. Clin Biomech 2010; 25: 9149.
  • 19
    Maly MR. Abnormal and cumulative loading in knee osteoarthritis. Curr Opin Rheumatol 2008; 20: 54752.
  • 20
    Reeves ND, Bowling FL. Conservative biomechanical strategies for knee osteoarthritis. Nat Rev Rheumatol 2011; 7: 11322.
  • 21
    Hunt MA, Simic M, Hinman RS, Bennell KL, Wrigley TV. Feasibility of a gait retraining strategy for reducing knee joint loading: increased trunk lean guided by real-time biofeedback. J Biomech 2011; 44: 9437.
  • 22
    Hunt MA, Birmingham TB, Bryant D, Jones I, Giffin JR, Jenkyn TR, et al. Lateral trunk lean explains variation in dynamic knee joint load in patients with medial compartment knee osteoarthritis. Osteoarthritis Cartilage 2008; 16: 5919.
  • 23
    Dieppe PA, Lohmander LS. Pathogenesis and management of pain in osteoarthritis. Lancet 2005; 365: 96573.
  • 24
    Lim BW, Hinman RS, Wrigley TV, Sharma L, Bennell KL. Does knee malalignment mediate the effects of quadriceps strengthening on knee adduction moment, pain, and function in medial knee osteoarthritis? A randomized controlled trial. Arthritis Rheum 2008; 59: 94351.
  • 25
    Altman RD, Asch E, Bloch D, Bole G, Borenstein D, Brandt K, et al. Development of criteria for the classification and reporting of osteoarthritis: classification of osteoarthritis of the knee. Arthritis Rheum 1986; 29: 103949.
  • 26
    Altman RD, Hochberg M, Murphy WA, Wolfe F, Lequesne M. Atlas of individual radiographic features in osteoarthritis. Osteoarthritis Cartilage 1995; 3: 370.
  • 27
    Kraus VB, Vail TP, Worrell T, McDaniel G. A comparative assessment of alignment angle of the knee by radiographic and physical examination methods. Arthritis Rheum 2005; 52: 17305.
  • 28
    Hinman RS, Payne C, Metcalf BR, Wrigley TV, Bennell KL. Lateral wedges in knee osteoarthritis: what are their immediate clinical and biomechanical effects and can these predict a three-month clinical outcome? Arthritis Rheum 2008; 59: 40815.
  • 29
    Baker R. Symposium. Modelling concepts in motion analysis. Joint ESMAC and GCMAS meeting. Gait and Clinical Movement Analysis Society 2006; Amsterdam, The Netherlands: 2006.
  • 30
    Armand A, Sangeux M, Hoffmeyer P, Baker R. Optimal markers' placement on the thorax for clinical gait analysis: a preliminary study. Gait Posture 2009; 30: S54.
  • 31
    Hunt MA, Wrigley TV, Hinman RS, Bennell KL. Individuals with severe knee osteoarthritis (OA) exhibit altered proximal walking mechanics compared with individuals with less severe OA and those without knee pain. Arthritis Care Res (Hoboken) 2010; 62: 142632.
  • 32
    Gentile AM. Skill acquisition: action, movement and neuromotor processes. In: Carr JH, Shephard RD, editors. Movement sciences: foundation for physical therapy in rehabilitation. 2nd ed. Maryland: Aspen Press; 2000. p. 11187.
  • 33
    Hodges NJ, Williams AM, Hayes SJ, Breslin G. What is modelled during observational learning? J Sports Sci 2007; 25: 53145.
  • 34
    Wrigley TV, Simic M, Hunt MA, Hinman RS, Bennell KL. Real-time movement biofeedback for walking gait modification in knee osteoarthritis. In: Cobb S, editor. Virtual Rehabilitation 2009 International Conference; June 29– July 2; Haifa, Israel. Piscataway (New Jersey): IEEE; 2009. p. 1325.
  • 35
    Kellgren JH, Jeffrey MR, Ball J. The epidemiology of chronic rheumatism: atlas of standard radiographs of arthritis. Oxford (UK): Blackwell Scientific; 1963.
  • 36
    Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation-study of WOMAC: a health-status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug-therapy in patients with osteoarthritis of the hip or knee. J Rheumatol 1988; 15: 183340.
  • 37
    Verbeke G, Molenbergh G. Linear mixed models for longitudinal data. New York: Springer-Verlang; 2009.
  • 38
    Astephen JL, Deluzio KJ, Caldwell GE, Dunbar MJ. Biomechanical changes at the hip, knee, and ankle joints during gait are associated with knee osteoarthritis severity. J Orthop Res 2008; 26: 33241.
  • 39
    Simic M, Bennell KL, Hunt MA, Wrigley TV, Hinman RS. Contralateral cane use and knee joint load in people with medial knee osteoarthritis: the effect of varying body weight support. Osteoarthritis Cartilage 2011; 19: 13307.
  • 40
    Linley HS, Sled EA, Culham EG, Deluzio KJ. A biomechanical analysis of trunk and pelvis motion during gait in subjects with knee osteoarthritis compared to control subjects. Clin Biomech 2010; 25: 100310.
  • 41
    Van der Esch M, Steultjens M, Harlaar J, van den Noort J, Knol D, Dekker J. Lateral trunk motion and knee pain in osteoarthritis of the knee: a cross-sectional study. BMC Musculoskelet Disord 2011; 12: 141.
  • 42
    Briem K, Snyder-Mackler L. Proximal gait adaptations in medial knee OA. J Orthop Res 2009; 27: 7883.
  • 43
    Mundermann A, Dyrby CO, Andriacchi TP. Secondary gait changes in patients with medial compartment knee osteoarthritis: increased load at the ankle, knee, and hip during walking. Arthritis Rheum 2005; 52: 283544.
  • 44
    Robinson GK. That BLUP is a good thing: the estimation of random effects. Stat Sci 1991; 6: 1532.