To examine the association of varus malalignment with impairments and functional limitations in people with medial knee osteoarthritis (OA).
To examine the association of varus malalignment with impairments and functional limitations in people with medial knee osteoarthritis (OA).
Anatomic radiographic knee alignment was assessed in 107 community volunteers with medial tibiofemoral knee OA. Impairments assessed included pain (Western Ontario and McMaster Universities Osteoarthritis Index [WOMAC]), quadriceps and hamstring isometric strength, and knee varus-valgus laxity. WOMAC, walking speed, step test, and stair climb test were used to assess functional limitations. Participants were categorized into tertiles according to knee alignment (least, moderate, and most varus). Impairments and functional limitations between groups were compared using analyses of variance with and without adjustment for age, sex, and disease severity. Regression analyses were also performed in the entire cohort to further determine the relationship of varus malalignment to impairments and functional limitations.
The most varus group (mean varus 7.7 degrees) did not demonstrate greater impairments or worse functional limitations compared with the moderate varus (4.2 degrees) and least varus (5.0 degrees) groups. In fact, the most varus participants performed significantly better on the step test compared with moderate (P = 0.006) and least varus (P = 0.004) participants. Knee alignment accounted for a significant but small proportion of the variance in step test performance (7%) and quadriceps strength (4%), but did not contribute significantly to the variance in any other parameter measured.
Greater varus malalignment was associated with increased quadriceps strength and improved step test performance, but did not influence the severity of other measured impairments and functional limitations.
Knee osteoarthritis (OA) is the leading cause of chronic disability in the elderly (1). OA involvement at the knee is most common in the medial tibiofemoral compartment (2) and is often associated with varus malalignment (3). Patients with knee OA typically demonstrate impairments such as knee pain (4), muscle weakness (5), and joint laxity (6), which often result in functional limitations such as reduced walking speed (7), balance deficits (8), and difficulty performing activities of daily living (9).
Subgroups within the knee OA population are believed to have differing rates of disease progression and possible variation in treatment response (10, 11). Those with knee varus-valgus malalignment compose a subgroup that has been given increasing attention (11–14). Knee malalignment is a local biomechanical factor that influences the magnitude of forces imposed across the joint. Specifically, varus malalignment increases loading medially (15) and is a major contributing factor to development (12) and progression (16–18) of OA in the medial tibiofemoral compartment. Varus malalignment has been associated with greater disease severity, worsening pain, and functional decline over 18 months (18). Furthermore, knee malalignment mediates the effects of several risk factors for knee OA, including quadriceps weakness and obesity (11, 19). Higher body mass index (BMI) and greater absolute quadriceps strength at baseline increase the risk of disease progression in patients with knee malalignment but not in those with neutrally aligned knees (11, 14). Thus, mounting evidence suggests that the subgroup of knee OA patients with knee malalignment may manifest different characteristics and clinical outcomes compared with their more neutrally aligned counterparts.
Despite evidence suggesting that malalignment is an important mechanical factor in knee OA, no study to our knowledge has examined the association between malalignment and the impairments and functional limitations associated with knee OA. It is likely that knee malalignment stresses and thus accelerates degenerative changes in pain-causing knee structures such as the meniscus, ligaments, and subchondral bone (4, 20, 21), and therefore may increase the severity of knee pain and impairment in surrounding local structures and global body functions. The aim of this study was to examine the association of varus malalignment with impairments and functional limitations in people with medial tibiofemoral OA. We hypothesized that participants with more severe varus malalignment would have increased impairments and greater functional limitations compared with participants with less varus malalignment.
We recruited 107 participants from the community in Melbourne, Australia through advertisements in newspapers and local community clubs for a randomized controlled trial investigating quadriceps strengthening. All participants had tibiofemoral joint OA in at least 1 knee and fulfilled the American College of Rheumatology classification criteria (22): age >50 years, knee pain most days of the past month, and osteophytes apparent on knee radiograph. To ensure medial tibiofemoral joint OA, the following criteria were set: self-reported pain on the medial aspect of the knee, osteophytes in the medial tibiofemoral compartment, and medial joint space narrowing greater than lateral joint space narrowing (23). Participants were excluded if they had a history of lower limb joint replacement; knee surgery, intraarticular steroid, or hylan G-F 20 injection within the previous 6 months; systemic arthritic condition; more than 5 degrees of valgus malalignment on radiograph; were seeking or currently receiving physiotherapy for knee OA; were intending to start or currently participating in a lower limb strengthening program; or had a severe medical condition that precluded safe participation in an exercise program.
Ethical approval was obtained from The University of Melbourne Human Research Ethics Committee, and from the Department of Human Services Radiation Advisory Committee. Written informed consent was provided by participants at enrollment. Figure 1 summarizes the recruitment process. Participants were initially screened over the telephone and those eligible underwent a standardized anteroposterior (AP) weight-bearing radiograph to ascertain knee alignment and OA severity. Participants fulfilling radiographic eligibility criteria were enrolled into the study.
An AP extended weight-bearing radiograph of the most painful knee was used to assess knee alignment and OA severity. When both knees were equally painful, the dominant knee was deemed the study knee. Disease severity was assessed using the Kellgren/Lawrence (K/L) scale (24), in which higher grades indicate greater severity. Anatomic knee alignment was determined using the methods of Moreland et al (25). The anatomic axis of the femur was attained by drawing a line from the center of the tibial spines to a point located 10 cm proximally and bisecting the medial to lateral width of the femur. Similarly, the tibial anatomic axis was obtained by drawing a line from the center of the tibial spines to a point located 10 cm distally and midway between the medial and lateral tibial surfaces. Anatomic knee alignment was indicated by the angle subtended at the point where the 2 lines meet in the center of the tibial spines. Knee alignment was evaluated by 1 investigator (B-WL), with excellent intrarater reliability (intraclass correlation coefficient [ICC] 0.97 based on 10 randomly selected radiographs measured 1 week apart). Mechanical knee alignment was extrapolated using the regression equation from Hinman et al (26): mechanical alignment = 0.915 (anatomic alignment) + 13.895. In this study, alignment is reported as the deviation from neutral (0 degrees) in the varus direction.
The following demographic and anthropometric information was obtained: age, sex, height, body mass, BMI, unilateral or bilateral symptoms, and duration of symptoms. Physical activity levels were measured using the Physical Activity Scale for the Elderly (27), in which scores range from 0 to >400, with higher scores indicating greater physical activity.
The Likert version of the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), transformed to a 100-point scale, was used to assess knee pain (28). Higher scores indicate worse pain.
Quadriceps and hamstring strength were assessed isometrically at 60 degrees knee flexion while sitting using a Kin-Com 125-AP dynamometer (Chattecx Corporation, Chattanooga, TN). A submaximal warm-up was followed by 3 maximal 5-second contractions with a 15-second rest interval in between trials (29). The highest peak force of the 3 trials was multiplied by the lever length (in meters) to obtain the maximum torque, which was then normalized by body mass (Nm/kg).
Knee varus-valgus laxity was assessed using the Kin-Com 125-AP dynamometer with customized modifications (30). Participants were seated with the knee relaxed and flexed at 20 degrees (6, 31) and the foot resting on the horizontal lever arm with its axis of rotation below the tibiofemoral joint. In this gravity-neutral position, the leg was moved passively 10 times from varus to valgus at 5 degrees per second. Varus and valgus angles were determined at the points where 8 Nm of passive resistance was reached. This resistance was similar to that used by other studies (31, 32) and could be achieved by most participants (91.6%) without pain or apprehension. Although muscle activity was not recorded, excessive muscle activity (e.g., due to apprehension) was identified from irregular traces in the real-time varus and valgus angular displacement graphs; the test was repeated when these irregular traces were observed.
In contrast to other methods that do not record force continuously (6, 31, 32), a neutral lever arm angle could be identified at zero force, and on this basis, varus and valgus ranges could be separated from the total lever arm angle data. Intrarater reliability of varus, valgus, and total laxity was excellent when measured a week apart in 8 healthy young adults (ICC 0.94, 0.95, and 0.96, respectively). Post hoc analysis revealed a significant correlation between varus-valgus laxity and both body mass (r = −0.517, P < 0.001) and height (r = −0.640, P < 0.001), but not with BMI (r = −0.153, P = 0.131). Therefore, values of laxity were normalized by height using the following formula:
where x was the slope of the regression graph (log-transformed height plotted against log-transformed laxity). The values of x were −4.65 for varus-valgus laxity, −4.62 for varus laxity, and −4.70 for valgus laxity. Absolute values of knee laxity were used for descriptive purposes, and normalized values were used for all statistical analyses.
The step test was used to assess dynamic standing balance (33). Participants were instructed to place the foot up on a 15-cm step and return it to the floor as many times as possible in 15 seconds while maintaining balance on the symptomatic leg.
For the stair climb test, participants were instructed to ascend and descend a set of stairs with 6 steps (each 17.5 cm in height) at their own pace, and the total time taken was recorded (34). Participants were permitted to use a handrail on 1 side if necessary.
Self-selected natural walking speed was measured using photoelectric beams (Jaycar Electronics, Melbourne, Victoria, Australia) attached to a stopwatch and placed 4 meters apart in the middle of an 8-meter walkway to reduce acceleration and deceleration effects. Average walking speed over 5 trials was recorded.
The Statistical Package for the Social Sciences, version 15 (SPSS Inc., Chicago, IL) was used for analyses. Data were analyzed 2 ways. First, participants were classified into groups based on tertiles of knee alignment rounded to the nearest degree. Participants with knee alignment ≤2 degrees varus were categorized as least varus, those with 3–5 degrees varus were categorized as moderate varus, and those ≥6 degrees varus were categorized as most varus. One-way analysis of variance (ANOVA) was used to compare continuous data across the groups and Tukey's honest significant difference test post hoc analyses were performed where results were significant. Chi-square tests were used to compare categorical data. Analyses were repeated using analyses of covariance with adjustment for age, sex, and disease severity.
In the second method, the relationship of knee malalignment to impairments and functional limitations was determined using separate regression analyses adjusted for age, sex, and disease severity in the entire cohort. The dependent variables were WOMAC pain, WOMAC function, walking speed, step test, stair climb test, quadriceps strength, hamstring strength, varus-valgus laxity, varus laxity, and valgus laxity. For each model, age, sex, and disease severity were initially forced into the model as a block. Knee alignment was then entered to determine whether any additional variance in the dependent variables could be explained by knee alignment. A significance level of P < 0.05 was used for all tests.
Participant characteristics are presented in Table 1 and the distribution of knee alignment is presented in Figure 2. The mean ± SD alignment of the entire cohort was 4.0 ± 3.3 degrees varus, with a range from −4.1 degrees (valgus) to 13.1 degrees (varus). The mean alignment of the least varus, moderate varus, and most varus groups were 0.5 degrees, 4.2 degrees, and 7.7 degrees varus, respectively (P < 0.001). Participants in the most varus group were significantly older than those in both the moderate varus (P = 0.03) and least varus (P = 0.003) groups. Not surprisingly, there was a significant relationship between varus alignment and radiographic disease severity (P < 0.001). Post hoc analysis of standardized residuals revealed that there were significantly more participants with K/L grade 2 in the least varus group (62%), and more participants with K/L grade 4 in the most varus group (77%).
|Entire cohort (n = 107)||Least varus (n = 37)||Moderate varus (n = 36)||Most varus (n = 34)||P|
|Age, years||64.6 ± 8.4||62.0 ± 8.9||63.5 ± 7.8||68.5 ± 7.5||0.003†|
|Body mass, kg||79.4 ± 15.3||76.0 ± 16.2||81.3 ± 14.8||81.2 ± 14.8||0.252|
|Height, meters||1.66 ± 0.10||1.63 ± 0.11||1.67 ± 0.08||1.67 ± 0.10||0.149|
|BMI, kg/m2||29.0 ± 4.8||28.6 ± 5.3||29.2 ± 5.1||29.2 ± 4.2||0.840|
|Varus malalignment, degrees||4.0 ± 3.3||0.5 ± 1.6||4.2 ± 1.1||7.7 ± 1.7||< 0.001†|
|Symptom duration, years||6.7 ± 5.3||5.2 ± 4.6||6.7 ± 5.6||7.2 ± 5.8||0.743|
|Physical activity level‡||169.8 ± 81.5||169.5 ± 78.2||171.0 ± 82.5||168.8 ± 86.3||0.994|
|Female sex, no. (%)||59 (55)||24 (65)||21 (58)||14 (41)||0.120|
|Bilateral symptom, no. (%)||41 (38)||17 (46)||12 (33)||12 (35.3)||0.491|
|Disease severity, no. (%)|
|K/L grade 2||34 (32)||23 (62)||9 (25)||2 (5)|
|K/L grade 3||29 (27)||8 (22)||15 (42)||6 (18)||< 0.001†|
|K/L grade 4||44 (41)||6 (16)||12 (33)||26 (77)|
|WOMAC pain score||35.7 ± 15.1||34.9 ± 16.3||35.7 ± 14.6||36.5 ± 14.6||0.906|
|WOMAC function score||34.8 ± 15.8||32.9 ± 16.6||37.3 ± 14.8||34.2 ± 16.1||0.475|
|Walking speed, meters/second||1.20 ± 0.20||1.20 ± 0.23||1.23 ± 0.17||1.19 ± 0.21||0.723|
|Step test, no.||13 ± 3||13 ± 3||13 ± 4||14 ± 3||0.063|
|Stair climb test, seconds||12.2 ± 4.5||12.8 ± 5.9||12.2 ± 3.9||11.6 ± 3.3||0.544|
|Quadriceps strength, Nm/kg||1.32 ± 0.54||1.22 ± 0.53||1.30 ± 0.58||1.45 ± 0.49||0.205|
|Hamstring strength, Nm/kg||0.71 ± 0.29||0.67 ± 0.31||0.68 ± 0.29||0.78 ± 0.24||0.203|
|Knee laxity, degrees (n = 98)§|
|Varus-valgus||10.8 ± 4.3||12.0 ± 4.9||10.0 ± 3.7||10.3 ± 4.0||0.251|
|Varus||5.5 ± 2.3||6.5 ± 2.5||5.1 ± 2.0||4.9 ± 2.0||0.024†|
|Valgus||5.3 ± 2.3||5.6 ± 2.6||4.9 ± 2.0||5.4 ± 2.2||0.474|
Laxity data were obtained in only 98 participants, as 9 participants could not reach the minimum force threshold of 8 Nm for the test. Post hoc Mann-Whitney U tests revealed that these participants (4 in the least varus, 2 in the moderate varus, and 3 in the most varus group) were significantly shorter and lighter than those who completed the laxity test. Although the most varus group demonstrated significantly lower varus laxity compared with the least varus group (4.9 degrees versus 6.5 degrees, P = 0.024), the difference did not remain significant after adjustment for age, sex, and disease severity (Table 2).
|Least varus (n = 37)||Moderate varus (n = 36)||Most varus (n = 34)||P|
|WOMAC pain score||38.0 ± 2.7||35.7 ± 2.4||33.1 ± 2.8||0.528|
|WOMAC function score||35.2 ± 2.7||37.2 ± 2.5||31.8 ± 3.0||0.383|
|Walking speed, meters/second||1.16 ± 0.04||1.22 ± 0.03||1.23 ± 0.04||0.395|
|Step test, no. in 15 seconds||12 ± 1||13 ± 1||15 ± 1||0.009†|
|Stair climb test, seconds||13.0 ± 0.8||12.3 ± 0.7||11.2 ± 0.8||0.342|
|Quadriceps strength, Nm/kg||1.21 ± 0.09||1.30 ± 0.08||1.45 ± 0.09||0.248|
|Hamstring strength, Nm/kg||0.68 ± 0.05||0.69 ± 0.04||0.77 ± 0.05||0.403|
|Knee laxity, degrees‡|
|Varus-valgus||11.5 ± 0.6||9.7 ± 0.6||11.1 ± 0.7||0.219|
|Varus||6.2 ± 0.3||5.0 ± 0.3||5.3 ± 0.4||0.062|
|Valgus||5.3 ± 0.4||4.7 ± 0.3||5.8 ± 0.4||0.338|
We found a significant difference in step test performance across the alignment groups (P = 0.009) after adjusting for age, sex, and disease severity. Tukey's post hoc analyses revealed that participants in the most varus group had a significantly higher step test score (adjusted mean 15) compared with those in the moderate varus group (adjusted mean 12; P = 0.006) and with those in the least varus group (adjusted mean 12; P = 0.004). Unadjusted and adjusted ANOVA results revealed no significant differences in other measured parameters across the groups (Tables 1 and 2).
Results of the regression analyses in the cohort as a whole demonstrated that malalignment was only a significant predictor of step test score and quadriceps strength (Table 3). Age, sex, and disease severity accounted for 9% of the variance in step test score, and addition of knee alignment to the regression model explained an additional 6% variance (P = 0.003). This model accounted for 15% of the total variance in step test score (adjusted R2 = 0.15, F = 5.8, P < 0.001). Age, sex, and disease severity accounted for 20% of the variance in quadriceps strength, and addition of knee alignment to the regression model explained an additional 4% of variance (P = 0.025). This model accounted for 24% of the total variance in quadriceps strength (adjusted R2 = 0.24, F = 9.1, P < 0.001). Furthermore, each degree of increase in varus malalignment was associated with a 0.04 Nm/kg (3.0%) increase in quadriceps strength and a 0.34-point (2.6%) increase in step test score (or approximately every 3° increase in varus malalignment was associated with a 1-step improvement in step test performance). The relationships between step test score and quadriceps strength and varus malalignment are presented in Figure 3.
|WOMAC pain score||−0.71 (0.52)||0.176|
|WOMAC function score||−0.63 (0.55)||0.255|
|Walking speed, meters/second||0.01 (0.01)||0.508|
|Step test, no. in 15 seconds||0.34 (0.11)||0.003†|
|Stair climb test, seconds||−0.11 (0.15)||0.462|
|Quadriceps strength, Nm/kg||0.04 (0.02)||0.025†|
|Hamstring strength, Nm/kg||0.01 (0.01)||0.158|
|Knee laxity, degrees (n = 98)‡|
The aim of this study was to examine the association of varus malalignment with impairments and functional limitations in people with medial tibiofemoral joint OA. We hypothesized that participants with more severe varus malalignment would have greater impairments and more severe functional limitations than those with less malalignment. Our results do not support our hypotheses. We found that participants with the most varus malalignment did not demonstrate greater impairments or worse functional limitations than participants with moderate or the least varus malalignment. In fact, our data demonstrated that participants with more varus malalignment performed significantly better on the step test (a measure of dynamic standing balance) compared with those with less varus malalignment, after controlling for age, sex, and disease severity. Regression analyses demonstrated that knee alignment accounted for a small but significant proportion of the variance in both step test performance (7%) and quadriceps strength (4%) after accounting for the effects of covariates. Nevertheless, the impact of knee alignment on step test performance and quadriceps strength may be substantial; each degree increase in varus malalignment was associated with improvements in step test performance and quadriceps strength by ∼3%.
To our knowledge, the association between knee malalignment and impairments and functional limitations has not been examined previously in knee OA. It is intriguing that greater varus malalignment was associated with better step test performance and quadriceps strength, especially given that increasing knee malalignment is typically associated with more severe disease (35, 36) and greater risk of disease progression (14, 17, 18). Moreover, the influence of malalignment on the risk of disease progression appears to be greater in knees with more severe OA (37). It is not clear from this cross-sectional study why participants with greater varus malalignment performed better on the step test and demonstrated stronger quadriceps compared with those with less malalignment.
It is possible that stronger quadriceps muscles in more malaligned participants may be a compensatory response to counteract the higher knee adduction moment that is associated with varus malalignment during gait (35, 38). Several authors have suggested that the quadriceps is important in resisting or balancing the external knee adduction moment (39, 40). Despite its relatively small moment arm at the knee, the quadriceps can exert a relatively large knee abduction moment because it is able to develop considerable force as a result of its large cross-sectional area (40). In knees with greater varus malalignment, medialization of the patella tendon is likely to shorten this moment arm, potentially making it more difficult for the quadriceps to resist the adduction moment. As a result of an ongoing attempt to counter the deleterious effects of the adduction moment, it is possible that people with greater varus malalignment may develop stronger quadriceps over time.
There is some indirect evidence to support this theory. The quadriceps demonstrate a higher activation of fast twitch motor units in varus knees compared with neutrally aligned knees, suggesting that the muscle may function less effectively in the presence of malalignment (41). Stronger hip adductors observed in people with knee OA have also been suggested as a muscular adaptation to resist the knee adduction moment (42). On the other hand, we cannot rule out the theory that varus malalignment could also be the effect of increased quadriceps strength. As shown by Sharma et al (11), stronger quadriceps may promote structural deterioration in people with malaligned knees, and hence may cause a higher knee adduction moment and subsequent increases in malalignment. However, causal effects cannot be established from this study, and future longitudinal studies will be required to establish causal relationships between malalignment and quadriceps strength.
The step test was included as a test of dynamic standing balance (33). Previous studies in older adults with chronic knee pain (43) and osteoporosis (44) have demonstrated that quadriceps strength is correlated with performance on both static and dynamic balance tests. Post hoc analysis of our data also showed a significant correlation between quadriceps strength and step test performance (r = 0.56, P < 0.001). Therefore, the improved step test performance in people with greater varus malalignment is likely to be explained by their greater quadriceps strength relative to participants with less malalignment.
Contrary to our hypothesis, we found that participants in the most varus group did not have greater pain than those in the less malaligned groups. In fact, the participants with the most varus malalignment reported pain scores (adjusted data) that were ∼13% lower than those of participants in the least varus group, although the difference was not statistically significant. Post hoc analyses demonstrated that we had 100% power to detect these differences in pain across the groups as significant, so it is highly unlikely that a larger sample would have generated different results. In a longitudinal study of 237 community volunteers with knee OA, the relationship between changes in mechanical alignment and pain over 18 months was examined (18). The authors of that study found that pain increased with increasing amounts of varus or valgus malalignment over time, and that each 5° change of malalignment increased pain by 10 mm on a 100-mm visual analog scale, even after adjusting for age, sex, and BMI. It is not clear why our findings conflict with these.
There are several possible reasons why greater knee malalignment was not associated with more severe knee pain in our study. First, although our data are speculative and only hypothesis-generating, it is possible that a relationship between pain severity and malalignment does not exist. Second, it is possible that participants with varus malalignment may develop a higher threshold of pain tolerance compared with more neutral participants, especially given that greater malalignment is typically associated with greater disease severity (18). Third, pain in OA is multifactorial, and psychosocial factors such as self-efficacy and psychological well-being, which are known to affect pain perception (45, 46), were not assessed. It is possible that factors such as these are more important determinants of pain severity than malalignment. Finally, we did not assess the quantity of pain medication taken by participants; participants with more varus malalignment may have been taking more pain-relieving medications, which could have diluted the differences in pain scores across the groups.
Our results suggest that greater varus malalignment is not associated with more severe functional limitations. This finding was strengthened by being consistent across the various measures of functional limitation used in this study, including self-reported (WOMAC function subscale) and observed measures (walking speed, step test, and stair climb test). Greater quadriceps strength has been associated with less severe knee pain and disability (5), and hence it is likely that the influence of malalignment on function in our study may have been attenuated by the greater quadriceps strength demonstrated in more malaligned participants.
Contrary to our hypothesis, the most varus group demonstrated lower varus laxity compared with the least varus group when data were unadjusted. This could be attributed to osteophytic growth, joint space narrowing, and bony opposition in the medial compartment limiting tibial movement in the varus direction (47). Another plausible explanation is that people with varus knees may have a neutral force position that is biased towards varus due to stretching of lateral soft tissues. However, the difference between groups was no longer significant after adjustment for age, sex, and disease severity, suggesting that one of these latter factors is a more important determinant of laxity than malalignment per se. Regression analyses with inclusion of these covariates in the model did not reveal a significant relationship between varus malalignment and knee laxity. These findings conflict with the study by van der Esch et al (48), who found that increased malalignment was associated with increased varus-valgus laxity in 35 outpatients with knee OA. However, their results were not adjusted for potentially confounding factors, and knee alignment was measured using a goniometer. Differences in laxity measurement techniques and force thresholds used across the studies could also account for the discrepant results. Our knee laxity measurement technique is novel in that it allows us to separate absolute varus laxity from absolute valgus laxity; therefore, our data pertaining to directional laxity cannot be compared with other studies.
A limitation of our study is its cross-sectional design, which prevents the establishment of cause and effect relationships between varus malalignment and the measured parameters. Convenience sampling from the community was used and there may have been a selection bias for healthier and fitter participants, particularly among those with malalignment, which may have resulted in our participants demonstrating less impairment and functional limitation than experienced by typical OA patients. However, the WOMAC pain scores of our participants ranged from 5 to 70 and the WOMAC function scores ranged from 1 to 68 (both out of 100), suggesting that there was a wide variation in pain and functional limitation levels in our participants. In keeping with other authors (12, 49), this study used anatomic axis to define knee alignment. Although the mechanical axis using a full-limb radiograph is generally accepted as the gold standard in measuring knee alignment, several studies have found a good to excellent (r = 0.65 to 0.88) correlation between the anatomic axis and the mechanical axis (26, 36). Therefore, it is unlikely that our results would change with the use of mechanical axis to define knee alignment.
In conclusion, we found that varus knee alignment accounted for a small but significant proportion of the variance in step test performance and quadriceps strength in people with medial tibiofemoral joint OA. More severe varus malalignment did not appear to increase impairments or to worsen functional limitations. On the contrary, each degree increase in varus malalignment was associated with an approximate 3% improvement in step test performance and quadriceps strength. Future longitudinal studies are required to establish causal relationships between varus malalignment and impairments and functional limitations, and to examine the natural history of malalignment.
Dr. Bennell 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 design. Lim, Hinman, Wrigley, Bennell.
Acquisition of data. Lim.
Analysis and interpretation of data. Lim, Hinman, Wrigley, Bennell.
Manuscript preparation. Lim, Hinman, Wrigley, Bennell.
Statistical analysis. Lim.