To evaluate whether increased laxity of the knee during daily physical activities such as stair climbing is associated with progression of knee joint osteoarthritis (OA).
To evaluate whether increased laxity of the knee during daily physical activities such as stair climbing is associated with progression of knee joint osteoarthritis (OA).
During the years 2001–2003, 136 patients with bilateral primary medial compartment knee joint OA were enrolled in this prospective study. Baseline data collected were body mass index (BMI), muscle power, radiographic joint space width, mechanical axis on standing radiography, and anteroposterior (AP) knee laxity before and after physical exercise. After 8 years of followup, 84 patients were reexamined to assess radiographic changes. Radiographic disease progression was defined as progression of >1 grade on the Kellgren/Lawrence scale.
AP knee laxity increased significantly after stair climbing. Patients with OA progression and those without progression did not differ significantly in age, sex, baseline quadriceps muscle strength, mechanical axis, joint space width, and AP knee laxity before exercise. The 2 groups of patients did, however, differ significantly in baseline BMI and change in AP knee laxity due to exercise. The risk of progression of knee OA increased 4.15-fold with each millimeter of increase in the change in AP knee laxity due to exercise and 1.24-fold with each point increase in the BMI.
Our results indicate that patients with OA progression have significantly greater changes in knee joint laxity during physical activities and a higher BMI than patients without OA progression. These findings suggest that larger changes in knee laxity during repetitive physical activities and a higher BMI play significant roles in the progression of knee OA.
The knee is the principal large joint targeted by osteoarthritis (OA). OA of the knee is one of the major causes of joint dysfunction and physically disabling conditions in the elderly (1). It results in disabling knee symptoms in an estimated 10% of the UK population older than age 55 years, a quarter of whom are severely disabled (2). Muraki et al (3) reported that the prevalence of bilateral radiographic knee OA in Japan is 49.5%. The prevalence of knee OA was reported to be 42.6% and 62.4%, respectively among Japanese men and women older than age 40 years (4), and 33.0% in the Framingham study (5).
The cause of knee OA is thought to be multifactorial. Broadly speaking, the risk/prognostic factors can be grouped into systemic and local factors (5–16). It is now recognized that risk factors for the development of OA may differ from those that relate to its progression (17). Reported risk factors for the development of knee OA are heredity, aging, female sex, obesity, history of trauma, knee laxity, occupation, participation in sports, and quadriceps muscle weakness; risk factors for progression are obesity, low bone density, varus–valgus malalignment, and instability of the knee (18). Joint laxity is considered to be one of the major factors involved in both the progression and the development of knee OA (19–23). Sharma et al (24) suggested that varus–valgus laxity is a local risk factor for knee OA and potentially contributes to the progression of OA. Our group reported significantly greater anteroposterior (AP) laxity in patients with mild OA compared with normal control subjects and an increased tendency for varus–valgus laxity in patients with severe knee OA (25). However, these previous studies were performed primarily to assess knee joint laxity under static conditions and not during or after exercise.
Several studies have investigated exercise-induced laxity of the ligamentous structures. Skinner et al (26) studied the change in knee joint laxity in healthy volunteers before and after exercise and observed that exercise led to a significant increase in knee laxity. Weisman et al (27) studied the in vivo effects of cyclic loading in athletes and observed a mean increase in compliance of 16.8% in the medial collateral ligament. Stoller et al (28) determined the effects of exercise on the torsional laxity of healthy knee joints. Their findings indicated a significant increase in laxity, with the peak occurring 10–20 minutes after exercise.
Recently, we observed increased knee joint laxity in patients with knee OA after they participated in an exercise involving stair climbing and descending (29). This increased knee joint laxity after exercise might influence the development or progression of knee OA. To our knowledge, no published prospective study has examined changes in knee joint laxity after physical exercise such as stair climbing in patients with knee OA. Therefore, the purpose of this study was to evaluate whether the change in knee joint laxity that occurs after exercise (stair climbing) is associated with progression of knee OA. We chose stair climbing as the type of exercise because it is frequently performed during activities of daily living and does not require familiarization of the patients to a new locomotor task.
Participants in this prospective study were recruited at the University of Fukui Hospital. During 2001–2003, 136 patients with symptomatic bilateral primary medial compartment knee OA managed at the hospital's orthopedic unit were enrolled. All patients were older than age 60 years and had knee pain during some daily activities. At the start of the study, all of the patients were retired from their jobs. Patients were excluded from the study if they had symptomatic musculoskeletal disorders (e.g., osteonecrosis, osteochondritis dissecans) other than those affecting the knee joints, a history of major trauma or a sports injury of the knee, rheumatoid arthritis, gout, pseudogout, autoimmune disease, or other major systemic diseases. We also excluded patients who had undergone surgery, had leg trauma or a knee joint flexion contracture of ≥20°, as well as those who were unable to perform the stair climbing/descending exercise. All patients had a narrower joint space in the medial compartment than in the lateral compartment, as visualized on standing AP radiographs of the knee. All of the patients reported medial knee joint pain as well as pain in the corresponding patellofemoral joint. All measurements were performed on both knees of each patient; however, to eliminate the confounding variable of bilateral involvement, the analyses were based on the more symptomatic (index) knee. The study was approved by the institutional review board, and written informed consent was obtained from all participants.
Of the 136 patients recruited, 52 underwent total knee arthroplasty (TKA), died, or were lost to followup. The remaining 84 patients completed an 8-year followup observation period. Thus, the study population comprised 6 male patients and 78 female patients with bilateral primary medial tibiofemoral compartment OA who received care at the University of Fukui Hospital.
At the first visit, each patient underwent the following assessments: body mass index (BMI), quadriceps muscle strength in each leg or knee, mechanical axis on standing AP radiography, OA grade, joint space width, AP knee laxity before and after a stair climbing exercise, and change in AP knee laxity due to the exercise. After 8 years, standing AP radiographs of the knee were obtained and were used to evaluate the OA grade and the joint space width of the knee, in the same manner in which these evaluations were performed at study entry (see below for details).
The BMI was calculated as weight in kilograms divided by height in square meters. Radiographs of both knees were obtained for all patients, and the severity of knee OA was graded according to the Kellgren/Lawrence (K/L) scoring system (30). Based on the radiographic OA classification, disease severity was classified as K/L grade 1 (32 knees), grade 2 (56 knees), grade 3 (48 knees), or grade 4 (32 knees).
The quadriceps muscle strength test was performed in all patients, using a Musculator GT-30 (OG Giken) before and after the stair climbing/descending exercise. For this test, the patient sat on a measurement chair with the knee flexed at 90°, while the ankle was positioned under a force sensor. Quadriceps muscle strength was measured isometrically to assess the maximal muscle strength in kilogram weight (kgw). The reproducibility of measurement of quadriceps muscle strength was tested in 5 patients with OA and 5 subjects without OA, each of whom was tested on 3 different days. The mean intraclass correlation coefficient (ICC) was 0.93.
For the baseline measurement, each patient was asked to go up and down a staircase. The staircase was composed of 8 steps; each step was 15 cm deep and 30 cm wide (total height 60 cm and total length 240 cm). The patients were instructed to engage in this exercise for 10 minutes. The patients performed the exercise at 40–60% of their age-predicted maximum heart rate (for women, this was determined using the formula 226 − age), which is a light-to-moderate level applicable to “individuals who are quite unfit” (31).
The AP laxity of both knees was measured before and after the stair climbing exercise. AP laxity was measured with a KT2000 knee arthrometer (MEDmetric). The procedure for assessing translational laxity was comparable with a manual laxity test (drawer test) of the knee. The examiner applied a 13.6-kg (30 lb; 134N) force anteriorly to the tibia at 20° knee flexion, followed by a 9.1-kg (20 lb; 89N) force posteriorly at the same flexion angle, while measuring displacement of the tibia relative to the femur. The sum of maximal anterior and posterior displacement was defined as AP knee joint laxity. The reproducibility of measurements of AP knee joint laxity was tested in 5 patients with OA and 5 subjects without OA, each of whom was tested on 3 different days. The mean ICC was 0.91, and the minimum detectable change at 90% confidence (MDC90, calculated as 1.64 × ✓2 × SD × ✓[1 − ICC]) was 1.8 mm. The change in AP knee joint laxity due to exercise was determined as the difference between knee joint laxity before and after exercise. The knee tests and exercise protocols at baseline are shown in Figure 1.
Standing radiographs of the knee in AP views and standing AP radiographs of the whole leg were obtained for all patients. The full-length weight-bearing AP radiographs of the leg were used to express the varus–valgus alignment of the leg using the mechanical axis, which was determined as the angle between the line connecting the center of the femoral head and the center of the tibial plateau and the line connecting the center of the tibial plateau and the center of the ankle joint. All radiographs were evaluated by a board-certified orthopedic specialist (MS or SW), who was excluded from the process of measuring and assessing other data.
In this study, the severity of tibiofemoral OA was classified using the Kellgren and Lawrence atlas of standard radiographs (32). K/L grade 2 indicates definite OA. However, at the start of the study, patients with K/L grade 1 were included, because all of these patients had knee pain at least sometimes during activities of daily living. The minimum joint space width on the knee radiograph was measured in the medial tibiofemoral compartment at the narrowest point between the distal part of the femur and the proximal part of the tibia. The femoral point (f) was chosen as the lowest point of the convex line of the distal femoral condyle. Then, a line perpendicular to the ground was drawn from point f. The intersection between this line and the dense line of the tibial plateau was marked as point t. The distance between them (f − t; minimal joint space width) was measured with a caliper and corrected for magnification. The minimum joint space width measurements obtained on 2 different days in 10 patients with OA were compared by analysis of variance with repeated measures and using an ICC. The reliability of measurement of the medial compartment was high (ICC 0.94), and the MDC90 was 1.0 mm. In 4 of 84 patients, the joint space width at followup was greater than that at entry (negative radiographic change). Because this change was considered to be attributable to a variation in the radiographic assessment process, we regarded it as “no change” in this study. Radiographic disease progression was defined as an increase of at least 1 grade in the K/L score or a decrease of >0.2 mm in knee joint cartilage thickness annually.
For each patient, AP laxity was measured 3 times in each knee, and the average value was used for analysis. Quadriceps muscle strength was also measured 3 times, and the average value was used for analysis. To eliminate the confounding of bilateral involvement, results from only the more symptomatic side (index knee) of each patient were used in the analyses. The paired t-test was used to compare knee laxity before and after exercise. The unpaired t-test was used for comparison of continuous variables between the nonprogressor group and the progressor group. The chi-square test was used to compare nominal variables. A receiver operating characteristic (ROC) curve analysis was used to identify optimal cutoff values for each continuous variable at baseline, in order to predict which patients would have progression of OA (Figure 2). Optimal cutoff values were determined as the point on the ROC curve closest to the (0,100) point, which yields values for sensitivity and specificity that minimize the value for (1 − sensitivity)2 + (1 − specificity)2. To investigate the sensitivity of the results to loss to followup in the study, we also calculated sensitivity assuming that the 15 patients who underwent TKR could be considered as having disease progression. A multiple logistic regression analysis was performed to evaluate each continuous variable as a predictor of progression of OA and to determine the odds ratios (ORs), adjusted for all variables. Data are expressed as the mean ± SD unless stated otherwise. P values less than or equal to 0.05 were considered significant.
Table 1 summarizes the baseline characteristics of the 136 patients recruited for the study. The 52 patients who did not complete the study tended to be older and had less varus alignment than the 84 patients who did complete the study. Otherwise, there were no clear differences between the 2 groups (Table 1).
|Characteristic||All patients (n = 136)||Patients who completed the study (n = 84)||Patients who underwent TKA, died, or were lost to followup (n = 52)|
|Age, years||72.5 ± 2.1||72.3 ± 3.1||74.0 ± 5.4|
|No. men/no. women||14/122||6/78||8/44|
|Body mass index, kg/m2||24.7 ± 2.2||25.0 ± 2.9||24.7 ± 3.6|
|Quadriceps muscle strength, kgw||14.7 ± 5.3||14.6 ± 4.6||14.9 ± 5.3|
|Mechanical axis, degrees||4.7 ± 0.5||7.5 ± 0.2||2.9 ± 0.4|
|Joint space width, mm||3.1 ± 1.8||2.9 ± 1.8||3.3 ± 1.9|
|AP knee laxity, mm|
|Before exercise||7.8 ± 2.7||8.0 ± 2.6||7.5 ± 2.7|
|After exercise||10.0 ± 2.9||10.3 ± 3.1||9.7 ± 2.7|
|Change||2.2 ± 1.5||2.3 ± 1.4||2.2 ± 1.6|
|Kellgren/Lawrence grade, no. of patients|
The patients who completed the study were categorized into 2 groups based on the radiographic outcome after 8 years of followup; 46 patients had radiographic disease progression, and 38 patients had no progression (Table 2). The proportion of men and women and the distribution of patients according to the radiographic scales (K/L grade and joint space narrowing grade) were similar in the 2 groups (Table 2). In both groups, knee laxity increased significantly after exercise (P < 0.01 and P < 0.001 by paired t-test, in the groups without and with progression, respectively), and the average increase in both groups was larger than the MDC90 value of 1.8 mm. Age, quadriceps muscle strength, mechanical axis, joint space width, and AP knee laxity before exercise at baseline did not differ significantly between the 2 groups (Table 2). However, in the group with radiographic progression, the BMI, AP laxity after exercise, and total change in AP knee laxity due to exercise were significantly greater (P = 0.04, P = 0.05, and P = 0.05, respectively).
|Characteristic||Patients without disease progression (n = 38)||Patients with disease progression (n = 46)||P†|
|Age, years||73.7 ± 3.2||71.5 ± 2.8||0.07|
|No. men/no. women||2/36||4/42||0.12|
|BMI, kg/m2||23.7 ± 1.7||25.7 ± 3.2||0.04|
|Quadriceps muscle strength, kgw||14.6 ± 4.2||14.6 ± 4.8||1|
|Mechanical axis, degrees||7.2 ± 0.2||7.8 ± 0.2||0.63|
|Joint space width, mm||3.9 ± 1.9||3.5 ± 1.7||0.43|
|AP knee laxity, mm|
|Before exercise||8.0 ± 2.2||7.9 ± 2.9||0.59|
|After exercise||10.0 ± 3.2||10.5 ± 2.8||0.05|
|Change||2.0 ± 1.2||2.6 ± 1.3||0.05|
|K/L grade at baseline, no. of patients||0.74|
|K/L grade after 8 years, no. of patients||0.94|
The mean ± SD narrowing of joint space width over 8 years was 1.8 ± 1.2 mm (range 0.6–2.8) in the radiographic progression group and 0.5 ± 0.3 mm (range 0.2–0.7) in the group without progression; this difference was statistically significant. The average joint space narrowing in the progression group was higher and in the nonprogression group was lower than the MDC90 of 1.0 mm. The joint space width after 8 years of followup was significantly narrower in the progression group than in the nonprogression group (P = 0.002).
To compare the predictive value of each baseline variable for radiographic progression, the area under the ROC curve and the optimal cutoff value for each baseline variable were determined using ROC curve analysis (Figure 2 and Table 3). According to this analysis, the optimal cutoff values of AP knee laxity (before exercise), total change in AP knee laxity (before/after exercise), BMI, quadriceps muscle strength, mechanical axis, and joint space width were 9 mm, 3 mm, 26 kg/m2, 16 kgw, 7°, and 3 mm, respectively (Table 3). The cutoff value of 3 mm for change in laxity was larger than the MDC90 value for joint laxity of 1.8 mm.
|Patients with disease progression (n = 46)||Patients without disease progression (n = 38)||Total (n = 84)||AUC||Sens.||Spec.||PPV||PLR|
|AP knee laxity before exercise, mm|
|Change in AP knee laxity due to exercise, mm|
|Quadriceps muscle strength, kgw|
|Mechanical axis, degrees|
|Joint space width, mm|
The sensitivity, specificity, positive predictive value, and positive likelihood ratio for each baseline variable were determined (Table 3). The values for sensitivity changed only slightly when the 15 patients who underwent TKR were included as progressors (results not shown). A change in AP laxity of ≥3 mm due to exercise and a varus–valgus angle of ≥7° were the most sensitive indicators of progression, whereas a BMI of at least 26 kg/m2 and a change in AP laxity of ≥3 mm due to exercise were most specific and had the highest positive predictive values and the highest positive likelihood ratios (Table 3).
A multiple logistic regression analysis was performed, with radiographic disease progression as the dependent variable. Eight independent variables were entered into the analysis (age, sex, BMI, quadriceps muscle strength, mechanical axis, joint space width, AP knee laxity, total changes in AP knee laxity). Of these, the variables shown to be significant were BMI (P = 0.018) and total change in AP knee laxity due to exercise (P = 0.046). After adjustment for all other variables, the odds of progression of knee OA increased 4.15-fold with a 1-mm increase in change in AP knee laxity due to exercise and 1.24-fold with a 1-point increase in the BMI (Table 4).
|Body mass index, kg/m2||0.018||1.24||1.04–1.45|
|Quadriceps muscle strength, kgw||0.076||1.07||0.42–2.36|
|Mechanical axis, degree||0.492||1.20||0.48–2.74|
|Joint space width, mm||0.695||0.96||0.39–2.35|
|AP knee laxity before exercise||0.456||1.29||0.54–3.08|
|Change in AP knee laxity due to exercise||0.046||4.15||1.12–15.37|
The aim of the present study was to characterize the association between the increase in knee joint laxity observed after stair climbing and progression of knee OA. To do so, we evaluated the following variables at the baseline visit: age, sex, BMI, quadriceps muscle strength, mechanical axis, joint space width, AP knee laxity before exercise, and change in knee AP knee laxity due to exercise. By comparing standing AP radiographs obtained at baseline and after 8 years of followup, we identified patients who had radiographic progression of OA.
The findings of our study were as follows: 1) patients with OA progression and those without OA progression did not differ significantly in baseline age, sex, quadriceps muscle strength, mechanical axis, joint space width, and AP knee laxity before exercise; 2) baseline AP knee laxity increased significantly after staircase climbing in both groups of patients; 3) baseline BMI and change in AP knee laxity due to exercise differed significantly between patients with and those without OA progression. After adjustment for all measured variables, the risk of knee OA progression increased 4.15-fold with each millimeter increase in change in AP knee laxity due to exercise and 1.24-fold with each point increase in the BMI. Patients with a change in laxity of >3 mm were twice as likely to have OA progression, and those with a BMI of >26 kg/m2 were 6-fold more likely to have OA progression. Taken together, the results suggest that larger changes in AP laxity due to exercise and a higher BMI are associated with radiographic progression of knee OA.
Risk factors for knee OA can be categorized as systemic factors and local biomechanical factors. Identified systemic risk factors for OA include age, sex, ethnicity, hormonal status, bone density, and genetics, whereas identified local risk factors include obesity, joint injury, joint deformity, sports participation, occupational factors, quadriceps muscle weakness, mechanical environment, and ligamentous laxity (18). It is now recognized that risk factors for the development of OA may differ from those related to its progression (17). The current study had no non-OA control group, which makes it impossible to discuss the role of changes in joint laxity due to exercise on the development of knee OA.
Obesity is one of the 2 risk factors for progression of knee OA identified in this study. Recent studies have proved that being overweight antedates the development of knee OA. Furthermore, in patients with OA, being overweight increases the risk of radiographic progression (33–36). Felson et al reported that a higher baseline BMI increased the risk of incident knee OA in women (OR 1.8 per 5-unit increase) (34). In our study, the OR for the association of BMI with progression of OA over the same period of time was 2.9 per 5-unit increase. The larger value in our study might be explained by the fact that our patients were all Japanese and may have had lower body weights (mean ± SD 56.1 ± 8.2 kg) than those of women in the Framingham study.
In our study, knee laxity before exercise did not differ significantly between the group with OA progression and the group that did not experience disease progression. This finding is consistent with those of an earlier study, in which no association between AP laxity and the incidence of knee OA was observed (21). However, the change in AP knee laxity after the stair climbing exercise was significantly associated with OA progression, with patients experiencing progression showing a greater increase in laxity. Specifically, a 3-mm threshold change in AP laxity was identified, which is larger than the MDC90 value of 1.8 mm; this change could separate progressors from nonprogressors. This larger increase was already apparent after only 10 minutes of stair climbing and descending.
Stair climbing is one of the most demanding daily activities for the quadriceps muscles, because this activity produces tibiofemoral compression loads up to 6-fold greater than body weight (37). In the present study, AP laxity increased significantly after stair climbing in both groups of patients. Exercise-related knee joint laxity has been discussed in previous reports (26–28, 38–41). Skinner et al reported that exercise-related fatigue tends to modify the biomechanical properties of the anterior cruciate ligament and decrease the tone of muscles of the lower extremities through changes in knee joint laxity (26). Weisman et al suggested that changes in compliance of the medial cruciate ligament are influenced by the creep phenomena and changes in properties of the ligament after exercise (27). Stoller et al reported that exercise could result in torsional laxity of the knee (28). Although our study did not investigate directly whether the observed changes in laxity were purely the result of ligamentous changes, the results are consistent with those of the above-mentioned studies and suggest that patients with OA are vulnerable to increases in knee laxity after light-to-moderate physical exercise.
In previous studies, the association between knee OA and joint loading has been investigated (42–44). Jobs in which workers perform repetitious tasks, thus overworking the joints and fatiguing muscles that protect the joints, increase the risk of OA in those joints (18). Other studies have shown that workers whose jobs involve physical labor (e.g., farmers) have high rates of knee OA (43, 44). When specific job tasks were examined, jobs requiring kneeling or squatting along with heavy lifting were associated with especially high rates of both knee OA and hip OA. Forces across the knee increase in the crouching or squatting position; heavy lifting from such a position further increases loading. Other occupational activities, including climbing stairs, walking on uneven ground, standing, and sitting, have been inconsistently linked to a risk of OA (44). In this study, because all subjects were retired from their jobs and thus were not participating in occupational activities, influences from job-related tasks are unlikely.
In previous studies, the knee joint laxity of patients with OA was evaluated in static conditions, i.e., without interposition of exercise or knee loading (20, 25, 45). In daily life, however, the knee joint continues to move (e.g., squatting, walking, standing, climbing stairs), and this movement provides the environment in which knee OA progresses. When evaluating the influence of knee properties on OA progression, it is therefore important to assess knee joint kinematics, state, and stability throughout daily activities. A goal of this study was to evaluate the laxity of OA knee joints during daily activities such as stair climbing, but this was difficult to achieve. We therefore measured knee laxity just after stair climbing as a surrogate for evaluating the knee condition during daily activities.
It is important to recognize that the current study had several limitations. First, although in this study the knee was measured shortly after exercise, this is not the same condition as that during the activity. A reduction in laxity may have occurred between the time of finishing the exercise and the time at which laxity measurements were obtained, which means that our results may underestimate the true changes in laxity due to exercise. Second, the number of subjects in this study was small, partly because more than one-third of the patients were not available for the second part of the study due to TKR, death, or loss to followup. The relatively small number of participants makes it impossible to assess any differences among subgroups of patients with different K/L grades. Third, stair climbing as performed in this study represents a light-to-moderate exercise level, on par with that recommended for “individuals who are quite unfit” (31). Nevertheless, these exercise levels may exceed the activities of normal daily living. The change in laxity measured in this study may therefore overestimate that occurring in daily life. Finally, the measured changes in knee laxity were not standardized by the amount of exercise, because the climbing speed was based on the age-predicted maximum heart rate. Therefore, each patient performed a different amount of exercise. However, an amount of exercise based on the age-predicted maximum heart rate will be more representative of a patient's true daily activities.
In conclusion, our results show that patients with radiographic progression of knee OA had significantly increased knee joint laxity due to physical activities and a significantly higher BMI compared with patients without OA progression. These findings suggest that the development of knee laxity during repetitive physical activities as well as the patient's BMI might play significant roles in the etiology and progression of knee OA. Clarifying the exact role of knee laxity in OA etiology and progression would require a longitudinal study that records the amount and type of physical activity and changes in laxity in a group that includes subjects with and without OA.
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 published. Dr. Uchida 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. Miyazaki, Uchida, Sato, Watanabe, Yoshida, Wada, Shimada, Kuiper, Baba.
Acquisition of data. Miyazaki, Uchida, Sato, Watanabe, Yoshida, Wada, Shimada, Kuiper, Baba.
Analysis and interpretation of data. Miyazaki, Uchida, Sato, Watanabe, Yoshida, Wada, Shimada, Kuiper, Baba.
We thank Keiichiro Kamitani, MD, PhD and Yoshinori Kano, RPT for providing assistance with data collection.