Osteoarthritis (OA), the clinical condition of joint pain and dysfunction caused by the degeneration of articular cartilage and subchondral bone, affects more people than any other joint disease (8). The knee is the most frequently affected weight bearing joint (11). Knee OA is characterized by pain, functional disability, and reduced knee extensor and flexor strength (7,14,16,21). Weakness of the muscles surrounding the knee may reduce their capacity to protect the knee, predisposing it to greater physical stress, structural damage, and joint degeneration (20). Therefore, the therapeutic role of resistance exercise for increasing muscle strength in patients with knee OA is recommended in several international guidelines (2,22,32).
Although substantial deficits in muscular strength have been reported and several recommendations for exercise exist, there is limited evidence to suggest that resistance training can provide more than modest improvements in strength for patients with knee OA (13,19,31). Although substantial gains appear to be possible with higher-intensity resistance exercise programs (3,15,17), most studies have not controlled intensity or have used intensities far below that recommended by the American College of Sports Medicine for optimizing strength gains (26). The lack of information regarding the appropriate intensity of resistance training in patients with knee OA also has been highlighted in a recent systematic review (9). Additionally, although resistance exercise aimed at increasing quadriceps strength has received considerable attention, this has not been the case for the hamstring muscles. Although there is evidence to suggest that stronger quadriceps muscles play an important protective role in the progression of knee OA (20-22,33), hamstring strength has been recently suggested to be more important in the context of self-efficacy and performance (29).
It is important to note that the potential detrimental effects of increased strength on knee joint load, joint compression, and progression of degenerative changes have been recently highlighted (31,35). In patients with knee OA and malalignment of the lower limb, the effect of increased muscular strength is particularly controversial (6). Malalignment is a potent risk factor for structural deterioration of the knee joint (36) because it increases focal loading. It is possible that in the presence of excessive varus alignment, improvements in muscular strength may result in painful and damaging medial knee joint loads. This may be particularly important for younger patients with knee OA who have the capacity for large gains in muscular strength but might paradoxically compromise knee function.
Evaluating the efficacy of resistance exercise for patients with knee OA and varus malalignment is therefore complex and should balance the need to achieve a sufficiently high intensity to evoke strength changes with the potential for exacerbating pain and compromising adherence. Additionally, the potential for increased muscular strength to increase knee joint loads and decrease functional performance must also be considered. Despite the uncertainty surrounding the effects of increased knee extensor and flexor strength for patients with knee OA and varus malalignment, we are unaware of any studies evaluating the efficacy and safety of high-intensity resistance training for these patients.
Our primary objective was to evaluate the short-term efficacy, safety, and adherence of a high-intensity resistance training protocol for patients with advanced knee OA and varus malalignment. We hypothesized that strength would significantly increase without increases in pain and that adherence would be excellent. Our secondary objective was to generate pilot data evaluating the effect of training on dynamic knee joint load, function, and self-efficacy.
Baseline demographic and clinical characteristics of participants are provided in Table 1. Fourteen participants (two females) were recruited from patients on the waiting list for high tibial osteotomy (HTO) surgery at a tertiary care center specializing in orthopedics. All participants were referred by their primary care physician to this center due to unresolved knee pain localized primarily to the medial compartment and were subsequently diagnosed with medial compartment knee OA and varus alignment of the lower limb. A diagnosis of OA was based on the criteria described by Altman et al. (1). Kellgren and Lawrence (24) grades of OA severity and alignment (mechanical axis angle) (37) were determined from double-limb standing hip-to-ankle anteroposterior radiographs that were taken before the first test session (38). Only patients with no history of prior surgery on the limb scheduled for HTO were included. The Physical Activity Readiness Questionnaire (10), a general health-screening questionnaire, was administered to confirm suitability for resistance training. Participants provided written informed consent to participate in this study. The study was approved by the institution's Research Ethics Board for Health Sciences Research Involving Human Subjects.
Participants completed isokinetic resistance training under the supervision of study investigators three times per week for 12 wk, with a minimum of 1 d of rest between sessions. Participants began each session with a 5-min warm-up on a stationary cycle ergometer at low rate (50rpm) and low load (1 kP). After the warm-up, participants completed knee extensor and flexor stretches under the direction of the trainers. In total, each session took approximately 45 min to complete.
We used the Biodex Multi-Joint System 3 dynamometer (Biodex Medical, Shirley, NY) and accompanying software for strength training and testing. During each session, the participant was seated with his or her back against arigid backrest oriented 85° above the horizontal. The participant's pelvis and thigh were secured to the dynamometer using a seatbelt oriented diagonally across the anterior superior iliac spines and over the distal half ofthequadriceps, respectively. The axis of rotation of the dynamometer lever arm was positioned coaxial with the lateral femoral epicondyle. The resistance pad was secured over the distal anterior one third of the lower leg, above the malleoli.
Participants performed three sets of 10 repetitions of reciprocal concentric isokinetic knee extension and flexion at angular velocities of 60, 90, and 120°·s−1. On the basis of the results of baseline strength testing, we provide the participants with a visual target of 60% of their baseline extensor and flexor strength. They were instructed to aim for or exceed the target for each repetition. Participants then performed three sets of fifteen repetitions of reciprocal concentric isokinetic knee extension and flexion at 180°·s−1 with maximum effort.
To monitor progress and to set new strengthening goals, one training session during the third, the sixth, and the ninth week of program was replaced by the test protocols completed at baseline. On the basis of these strength tests, we generated new targets of 60% of peak torques.
Primary Outcomes Measures
Before testing, participants were given a 5-min warm-up on a stationary cycle ergometer at a low rate (50 rpm) and low workload (1 kP). Before each test velocity, participants performed two submaximal (50-65%) repetitions to allow for familiarization with the test speed. Participants performed five reciprocal concentric isokinetic contractions of knee extension and knee flexion with maximum effort at three sequential velocities (60, 90, and 120°·s−1). The participants then completed 30 reciprocal concentric isokinetic contractions of knee extension and knee flexion at 180°·s−1 at maximum effort. Participants were given 2 min of rest between test velocities. Testing was completed on both limbs with the limb scheduled for HTO defined as the index limb. Peak knee extensor and flexor torque (N·m) were calculated by averaging peak torque values from the three highest repetitions out of the five maximum-effort reciprocal isokinetic contractions collected at each of 60, 90, and 120°·s−1 angular velocities. Total knee extensor and flexor work (J) was also calculated for the 30 reciprocal isokinetic contractions at maximum effort at an angular velocity of 180°·s−1. Isokinetic knee extensor and flexor strength in patients with knee osteoarthritis has been previously demonstrated to be reliable and valid (12,28).
At the start of each training session, participants were asked to rate the pain they had experienced since the previous session (operationally defined as activities of daily living (ADL) pain). At the completion of the training session, participants were asked to rate the pain experienced during exercise (operationally defined as training pain). Pain was rated by the participant in the following specific anatomic areas on the index and the opposite limb using a 10-point numerical rating scale (0 was no pain; 10 was extreme pain) corresponding to illustrations depicting: under the knee cap (patellofemoral), inside the knee joint (tibiofemoral), quadriceps, and hamstrings muscles. Participants were also asked at each session to report any adverse event that may have affected the pain in their knee or their ability to exercise or any changes in pain medication relative to baseline.
Attendance and training intensity were monitored to ascertain adherence to this high-intensity program. The total number of sessions attended out of a possible 36 sessions was calculated. Perceived exertion of the patient during each exercise session was recorded using the Borg CR10 scale. The CR10 is a category ratio scale, which is both reliable and valid (5).
Secondary Outcomes Measures
Dynamic knee joint load.
The external adduction moment about the knee during walking was used as a surrogate measure of dynamic knee joint load (4). Participants underwent gait analysis using an eight-camera motion capture system (Motion Analysis Corporation, Santa Rosa, CA) synchronized with a single, floor-mounted force platform (Advanced Mechanical Technology Inc, Watertown, MA). Twenty-two passive-reflective markers were placed on the participant using a modified Helen Hayes marker set (23). During an initial static standing trial on the force platform, four extra markers were placed bilaterally over the medial knee joint lines and the medial malleoli. The static trial was completed to determine marker orientation, position of joint centers of rotation for the knee, and ankle as well as body mass. During the gait analysis, participants were instructed to walk barefoot across the laboratory at a self-selected velocity, whereas kinetic (1200 Hz) and kinematic data (60 Hz) were collected during the middle of several strides. Walking trials were repeated until a total of five trials with clean force platform strikes were obtained from the index limb.
Kinematic and kinetic data from each trial were combined and used to calculate external moments about the knee using inverse dynamic principles. Each lower limb segment (thigh, shank, and foot) was modeled as a rigid body with a local coordinate system that coincided with anatomic axes. Inertial properties of each limb segment were approximated, and translations and rotations of each segment were calculated relative to the marker orientations observed during the initial standing static trial. Walking speed was calculated as the average walking speed between successive foot contacts of the index limb. All postprocessing of gait data was done using commercially available software (Orthotrak; Motion Analysis Corporation, and Excel; Microsoft Corporation, Redmond, WA). The peak knee adduction moment for each of the five walking trials was determined, and an average peak knee adduction moment was calculated. Reliability of the peak knee adduction moment measurement has been previously reported (4).
Performance-based and self-report measures of function.
Participants were asked to walk at a self-selected pace for 6 min around the perimeter of a 24.4-m track, and the total distance walked was measured. This test has been shown to be a reliable measure of physical function for individuals with osteoarthritis (25). The Knee Injury and Osteoarthritis Outcome Score (KOOS), a 42-item patient-administered knee-specific measure, was used to assess the participants' opinions about their knee and general health. The KOOS consists of five subscales: pain, other symptoms, function in daily living (ADL), function in sport and recreation (Sport/Rec), and knee-related quality of life. A normalized score (100 indicating no symptoms and 0 indicating extreme symptoms) was calculated for each subscale. Appropriate reliability and responsiveness of the KOOS have been previously reported (18).
The Arthritis Self-Efficacy Scale (ASES) was used to assess self-efficacy. The score for each of three subscales (function, pain, and other symptoms) is the mean of the items scored 0-10 (with 10 indicating greater self-efficacy) for that subscale. Appropriate reliability, construct, and concurrent validity of the ASES have been previously reported (27).
All statistical analyses were performed using STATISTICA (Statsoft, Tulsa, OK). Each of the strength outcome measures was assessed using a two-factor, limb (index, opposite) by time (0, 3, 6, 9, 12 wk) repeated-measures ANOVA. Because there was very little pain experienced by any participant in the quadriceps and the hamstring muscles, only pain ratings for the patellofemoral and tibiofemoral joint were analyzed. Pain in each of these locations was assessed using a two-factor, activity (ADL, training) by time (1-12 wk) repeated-measures ANOVA. After significant effects, Scheffé posthoc tests were completed. Overall changes in strength, pain, and all secondary outcomes were also reported as mean differences before and after training with 95% confidence intervals (CI).
Knee extension and flexion strength from baseline to week 12 are presented in Figures 1 and 2, respectively. For knee extension peak torque at 60°·s−1, there was a significant main effect for limb (P = 0.002), a significant main effect for time (P < 0.001), and a significant limb × time interaction (P = 0.015). For knee extension peak torque at 90°·s−1, there was a significant main effect for limb (P = 0.01) and a significant main effect for time (P < 0.001). For knee extension peak torque at 120°·s−1, there was a significant main effect for limb (P = 0.02) and a significant main effect for time (P < 0.001). For knee extension total work at 180°·s−1, there was a significant main effect for time (P < 0.001).
At 60°·s−1, post hoc testing revealed significant differences between the index and the opposite limbs at all test times. The index limb had a significant increase in strength from baseline to week 6; however, there were no further significant gains between weeks 6 and 9 or 12. There were no significant differences in the opposite limb from baseline to week 12. When examining the main effect of time, post hoc testing revealed that for 90,120, and 180°·s−1, there were significant gains in extensor strength from baseline to week 3. There were no further significant gains from weeks 3 to 12.
For knee flexion, there was a significant main effect for time (P < 0.001). Post hoc testing revealed similar findings as knee extension, with significant gains between baseline and week 3, but no further significant gains were found from weeks 3 to 12. Overall changes in strength after training are reported in Table 2.
Patellofemoral and tibiofemoral pain data are presented in Figure 3. There were no significant main effects for activity or time. No patient reported any change in medication use to control pain over the course of the study. Overall changes in pain after training are reported in Table 2.
The mean total number of attended exercise sessions was 31.8 ± 2.6 out of a possible 36. Attendance levels were consistent over the course of the study. The mean subjective RPE during training sessions over the course ofthe program was 5.58 ± 1.86 on the Borg C10 scale where a rating of "5" corresponds to "heavy" exertion. During strength testing sessions, the mean RPE was 6.51 ± 2.34, where a rating of "7" corresponds to "very heavy." RPE was consistent over the course of training and testing.
Mean changes in all secondary outcomes after training are reported in Table 3. With the exception of the ASES function subscale, there were no significant changes in knee adduction moment, performance-based or self-report measures of function, and self-efficacy after the intervention.
The present high-intensity resistance training program produced large gains in strength on the index knee for peak knee extension and flexion torque at 60°·s−1 and for total knee extensor and flexor work at 180°·s−1. Specifically, effect sizes ranged from 0.73 to 1.25 during these tests, where strength increased from 28% to 46%, relative to baseline values. The mean knee extensor strength of the index knee was significantly lower at baseline than the opposite knee. The index knee also achieved greater gains than the opposite knee, helping to reduce the side-to-side strength deficits. At baseline, extensor strength in the index knee was on average 43% less than strength in the opposite knee at an angular velocity of 60°·s−1, whereas at week 12, the deficit was reduced to only 22%. For flexor strength, the strength deficit in the index knee was reduced from 14% to 4% after training. Additionally, extensor and flexor strength of the index knee posttraining significantly exceeded the baseline strength values of the opposite limb, and no difference existed for total extensor and flexor work in both the index and the opposite limb at 180°·s−1 between limbs after 12 wk of training.
In addition to the high intensity used and large increases in strength observed, our study has expanded on previously published work by focusing on a specific subgroup of patients with medial compartment knee OA with varus malalignment and advanced disease to warrant surgical intervention. Despite the increased potential to exacerbate symptoms in this group, increases in pain and subsequent modifications to the program were minimal. One patient experiencing knee joint pain took one full session off for rest and consulted with an orthopedic specialist before agreeing to continue training. Two participants had one training session modified to include training only on the index limb due to slight increases medial joint pain in the opposite knee. One participant, with very high baseline pain levels, had his training sessions altered to allow for more warm-up time.
Adherence was generally very good throughout this study. Others have reported that participants tend to drop out when they do not feel that the exercise is improving their knee OA symptoms (15). The participants in this study also maintained consistent levels of intensity throughout the training program that, when combined with attendance, was critical for generating the overall large strength gains. The observed ability for patients with advanced disease to experience sizable increases in strength without increases in pain and decreases in adherence suggests that high-intensity resistance exercise may be appropriate for a broader spectrum of patients than previously thought.
Although there was no change in mean peak knee adduction moment after 12 wk of training, individual results varied widely for this measure. Only one individual experienced an increase in peak knee adduction moment beyond the minimum detectable change at the 95% confidence level (4). This change was not accompanied by an increase in speed, exacerbation of pain, or decreases in function. Thorstensson et al. (39) also did not observe a significant reduction in mean peak knee adduction during gait after an 8-wk general conditioning exercise program. They did find a reduction during one-leg rise after 8 wk, which they suggested was amore sensitive measure than peak adduction moment during walking.
Although there was no change in the mean pain values, individual results also varied widely. One participant incurred a negative change in pain on the KOOS of greater than eight points, which has been suggested to represent the minimal perceptible clinical improvement or decline of each KOOS subscale at the individual level (34). Four participants demonstrated a clinical improvement in pain score (>8 points) relative to baseline.
Similarly, mean levels of function did not change in the current study. Three participants, however, experienced a clinically significant decline in the KOOS symptom (function) subscale relative to baseline, and four demonstrated a clinically significant increase. Although other resistance training studies have shown some modest improvements in function, our results were similar to Thorstensson et al. (40). In these relatively younger samples, having advanced knee OA may have been perceived to be more of a disability than that in the elderly populations with knee OA. Others have noted that self-efficacy explained much of the variance in performance in people with knee OA (30), implying that we should aim to improve an individual's confidence in performing physical tasks in addition to mechanical strategies such as strengthening. Notably, in the current study, despite the absence of change in KOOS and 6 min of walk, a significant increase in self-efficacy was observed in the function subscale of the ASES.
The small number of subjects and the absence of a control group in this pilot study must be acknowledged. The limitations in generalizability of the present results to the greater population of patients with knee OA should also be considered. Although the strength profiles and pain and disability scores at baseline were similar to previous studies, this sample had varus malalignment, was younger than most samples, and consisted largely of males. Although we demonstrated short-term efficacy and safety of a high-intensity resistance training program, future research should examine the long-term effects of high-intensity resistance exercises in patients with knee OA and malalignment, including in radiographic evaluation of disease progression. Furthermore, as a result of this study's small number of subjects, the statistical power of the motion analysis findings is low. A larger study with a control group for comparison and survival analysis to a specific end point such as an osteotomy would offer additional compelling evidence for the validity of this strength training program.
Overall, the present findings suggest that high-intensity resistance training can produce substantial increases in knee extensor and flexor strength in middle-aged patients with advanced knee OA and varus malalignment, without concomitant increases in pain, adverse events, or decreases in adherence. The results of this efficacy study support future clinical trials investigating the effectiveness of high-intensity resistance training for improving various indicators of disease and function in this important subgroup of patients.
This study was supported by the Canada Research Chairs Program (TBB), the Canadian Institutes of Health Research, and the Arthrex Inc.
The results of the present study do not constitute endorsement by ACSM.
1. Altman R, Asch E, Bloch D, et al. Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee. Diagnostic and Therapeutic Criteria Committee of the American Rheumatism Association. Arthritis Rheum
2. American College of Rheumatology Subcommittee on Osteoarthritis. Recommendations for the medical management of osteoarthritis of the hip and knee: 2000 update. American College of Rheumatology Subcommittee on Osteoarthritis Guidelines. Arthritis Rheum
3. Baker KR, Nelson ME, Felson DT, Layne JE, Sarno R, Roubenoff R. The efficacy of home based progressive strength training in older adults with knee osteoarthritis: a randomized controlled trial. J Rheumatol
4. 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
5. Borg G. Borg's Perceived Exertion and Pain Scales
. Champaign (IL): Human Kinetics; 1998.
6. Brandt KD. Is a strong quadriceps muscle bad for a patient with knee osteoarthritis? Ann Intern Med
7. Brandt KD, Heilman DK, Slemenda C, et al. A comparison of lower extremity muscle strength, obesity, and depression scores in elderly subjects with knee pain with and without radiographic evidence of knee osteoarthritis. J Rheumatol
8. Brooks PM. Impact of osteoarthritis on individuals and society: how much disability? Social consequences and health economic implications. Curr Opin Rheumatol
9. Brosseau L, MacLeay L, Robinson V, Wells G, Tugwell P. Intensity of exercise for the treatment of osteoarthritis. Cochrane Database Syst Rev
10. Canadian Society for Exercise Physiology. Physical activity readiness questionnaire [Internet]. Ottawa, ON: Canadian Society for Exercise Physiology [cited 2006 July]. Available at:http://www.csep.ca/communities/c574/files/hidden/pdfs/par-q.pdf
11. Doherty M, Dougados M. Evidence-based management of osteoarthritis: practical issues relating to the data. Best Pract Res Clin Rheumatol
12. Drouin JM, Valovich-McLeod TC, Shultz SJ, Gansneder BM, Perrin DH. Reliability and validity of the Biodex System 3 PRO isokinetic dynamometer velocity, torque and position measurements. Eur J Appl Physiol
13. Ettinger WH, Jr, Burns R, Messier SP, et al. A randomized trial comparing aerobic exercise and resistance exercise with a health education program in older adults with knee osteoarthritis. The Fitness Arthritis and Seniors Trial (FAST). JAMA
14. Felson DT. Clinical practice. Osteoarthritis of the knee. N Engl J Med
15. Fisher NM, Kame VD, Jr, Rouse L, Pendergast DR. Quantitative evaluation of a home exercise program on muscle and functional capacity of patients with osteoarthritis. Am J Phys Med Rehabil
16. Fisher NM, Pendergast DR. Reduced muscle function in patients with osteoarthritis. Scand J Rehabil Med
17. Fisher NM, White SC, Yack HJ, Smolinski RJ, Pendergast DR. Muscle function and gait in patients with knee osteoarthritis before and after muscle rehabilitation. Disabil Rehabil
18. Garratt AM, Brealey S, Gillespie WJ. Patient-assessed health instruments for the knee: a structured review. Rheumatology (Oxford)
19. Hughes SL, Seymour RB, Campbell R, Pollak N, Huber G, Sharma L. Impact of the fit and strong intervention on older adults with osteoarthritis. Gerontologist
20. Hurley MV. The role of muscle weakness in the pathogenesis of osteoarthritis. Rheum Dis Clin North Am
21. Hurley MV. The effects of joint damage on muscle function, proprioception and rehabilitation. Man Ther
22. Jordan KM, Arden NK, Doherty M, et al. EULAR Recommendations 2003: an evidence based approach to the management of knee osteoarthritis: report of a Task Force of the Standing Committee for International Clinical Studies Including Therapeutic Trials (ESCISIT). Ann Rheum Dis
23. Kadaba MP, Ramakrishnan HK, Wootten ME. Measurement of lower extremity kinematics during level walking. J Orthop Res
24. Kellgren JH, Lawrence JS. Radiological assessment of osteo-arthrosis. Ann Rheum Dis
25. Kennedy DM, Stratford PW, Wessel J, Gollish JD, Penney D. Assessing stability and change of four performance measures: a longitudinal study evaluating outcome following total hip and knee arthroplasty. BMC Musculoskelet Disord
26. Kraemer WJ, Adams K, Cafarelli E, et al. American College of Sports Medicine position stand. Progression models in resistance training for healthy adults. Med Sci Sports Exerc
27. Lorig K, Chastain RL, Ung E, Shoor S, Holman HR. Development and evaluation of a scale to measure perceived self-efficacy in people with arthritis. Arthritis Rheum
28. Madsen OR, Bliddal H, Egsmose C, Sylvest J. Isometric and isokinetic quadriceps strength in gonarthrosis; inter-relations between quadriceps strength, walking ability, radiology, subchondral bone density and pain. Clin Rheumatol
29. Maly MR, Costigan PA, Olney SJ. Determinants of self efficacy for physical tasks in people with knee osteoarthritis. Arthritis Rheum
30. Maly MR, Costigan PA, Olney SJ. Contribution of psychosocial and mechanical variables to physical performance measures in knee osteoarthritis. Phys Ther
31. Mikesky AE, Mazzuca SA, Brandt KD, Perkins SM, Damush T, Lane KA. Effects of strength training on the incidence and progression of knee osteoarthritis. Arthritis Rheum
32. Ottawa Panel. Ottawa panel evidence-based clinical practice guidelines for therapeutic exercises and manual therapy in the management of osteoarthritis. Physical Therapy
33. Rogind H, Bibow-Nielsen B, Jensen B, Moller HC, Frimodt-Moller H, Bliddal H. The effects of a physical training program on patients with osteoarthritis of the knees. Arch Phys Med Rehabil
34. Roos EM, Lohmander LS. The Knee Injury and Osteoarthritis Outcome Score (KOOS): from joint injury to osteoarthritis. Health Qual Life Outcomes
35. Sharma L, Dunlop DD, Cahue S, Song J, Hayes KW. Quadriceps strength and osteoarthritis progression in malaligned and lax knees. Ann Intern Med
36. Sharma L, Song J, Felson DT, Cahue S, Shamiyeh E, Dunlop DD. The role of knee alignment in disease progression and functional decline in knee osteoarthritis. JAMA
37. Specogna AV, Birmingham TB, DaSilva JJ, et al. Reliability of lower limb frontal plane alignment measurements using plain radiographs and digitized images. J Knee Surg
38. Specogna AV, Birmingham TB, Hunt MA, et al. Radiographic measures of knee alignment in patients with varus gonarthrosis: effect of weight bearing status and associations with dynamic joint load. Am J Sports Med
39. Thorstensson CA, Henriksson M, von Porat A, Sjodahl C, Roos EM. The effect of eight weeks of exercise on knee adduction moment in early knee osteoarthritis-a pilot study. Osteoarthritis Cartilage
40. Thorstensson CA, Roos EM, Petersson IF, Ekdahl C. Six-week high-intensity exercise program for middle-aged patients with knee osteoarthritis: a randomized controlled trial. BMC Musculoskelet Disord