SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. SUBJECTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Objective

Subjects with unilateral end-stage hip osteoarthritis (OA) who undergo total hip replacement (THR) preferentially require subsequent replacement of the contralateral knee compared with the ipsilateral knee. We investigated whether this nonrandom, preferential evolution of lower extremity OA from the hip to the contralateral knee joint may be related to asymmetries in dynamic joint loading at the knees, particularly the peak external knee adduction moment, which has been associated with the progression of knee OA.

Methods

Gait analysis was performed on 50 subjects who were preoperative for unilateral THR. Twenty-two of these subjects were reevaluated postoperatively 10–23 months after undergoing successful THR. At each analysis, dynamic joint loads in the contralateral knee were compared with those in the ipsilateral knee.

Results

Prior to THR, the peak external knee adduction moment and peak medial compartment load were significantly higher in the contralateral knee. This asymmetry persisted after THR.

Conclusion

Subjects with unilateral end-stage hip OA preferentially require subsequent replacement of the contralateral knee, as compared with the ipsilateral knee. Among patients with unilateral end-stage hip OA, the contralateral knee is subjected to higher dynamic joint loads than is the ipsilateral knee, and this asymmetric loading persists long after subjects have undergone successful THR. Biomechanical factors appear to be involved in the multiarticular evolution of OA of the lower extremities.

Recent investigations into the pattern of progression of lower extremity osteoarthritis (OA) have demonstrated that the evolution of end-stage OA among the large joints of the lower extremities is nonrandom. In particular, individuals with unilateral end-stage hip OA requiring total hip replacement (THR) whose OA then progresses in the knee (the noncognate joint) to require total knee replacement (TKR) have a significantly greater probability that the end-stage knee will be the contralateral, rather than the ipsilateral, joint (1). Similarly, individuals undergoing unilateral TKR for end-stage knee OA more frequently progress to a subsequent contralateral THR (noncognate joint) than an ipsilateral THR (1). This preferential evolution of advanced OA to the contralateral noncognate joint is not replicated in the pattern of progression of the systemic inflammatory arthropathy rheumatoid arthritis (1). Better understanding of the mechanism of this nonrandom evolution may yield significant insights into the pathogenesis of OA.

Biomechanical factors, such as abnormal dynamic joint loading, have been proposed to be important in the development and progression of lower extremity OA (2, 3). Dynamic loading refers to loading during physiologic activity, as opposed to static loading, which occurs when standing. The force across the knee during normal ambulation, the most common form of dynamic lower extremity loading, is approximately 3 times higher than the force during static standing (4). Thus, the “wear-and-tear” process in OA may occur during normal ambulation. An important, noninvasive technique by which to measure dynamic loading is gait analysis, which uses surface skin markers, a camera system, and a force plate on the floor to calculate biomechanical parameters such as the torque, or moment, imposed on the weight-bearing joints (5).

A number of studies have evaluated the relationship between dynamic knee loading and OA in the knees (6–11). The peak external knee adduction moment during gait is a measure of varus torque and, hence, of loading on the medial compartment of the knee. A high external knee adduction moment has been correlated with the radiographic severity of OA in a cross-sectional study of subjects with established OA (6) and imparts a poor surgical prognosis in subjects with a high tibial osteotomy (11). Moreover, there is now evidence that a high adduction moment may predict the subsequent radiographic deterioration of symptomatic knee OA (9, 10).

Normal ambulation and other physiologic activities that occur daily require intimate coordination among the joints of the lower extremities. Although it is likely that deformity and alterations of loading at one lower extremity joint have a significant effect on loading of the other joints of the lower extremities, there have been few studies to evaluate such effects systematically in OA-predisposed joints. Based on the fact that subjects with end-stage hip OA are significantly more likely to undergo contralateral TKR than ipsilateral TKR (1), the current study tested the hypotheses that during ambulation, subjects with unilateral end-stage hip OA place relatively greater loads on their contralateral knee than on their ipsilateral knee and that asymmetric loading is not alleviated by unilateral THR. Support for the hypotheses would suggest that asymmetric dynamic loading contributes to the nonrandom evolution of multiarticular OA of the lower extremities.

SUBJECTS AND METHODS

  1. Top of page
  2. Abstract
  3. SUBJECTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Subjects.

The Gait Laboratory at Rush Medical College maintains an extensive database of subjects who have undergone gait analysis for research purposes. This database was searched for subjects with end-stage primary hip OA who were tested prior to undergoing unilateral THR. Fifty-five subjects who had undergone gait analysis and subsequently underwent unilateral THR were identified.

The main objective of studying these subjects was to assess the multiarticular effects of primary, unilateral end-stage hip OA on lower extremity joint loading. Therefore, exclusion criteria were developed to eliminate any confounding effects on joint loading from pain or primary pathology at the other lower extremity joints. Subjects who had the following features were excluded: 1) a history of an inflammatory arthropathy, 2) significant pain or clinical OA in the contralateral hip (12), 3) documented clinical OA of the knee (13) or significant knee or foot pain, or 4) previous total joint replacement surgery of the lower extremities. Ipsilateral and contralateral hip pain and knee pain were assessed with the Harris Hip Score (14) in all study subjects.

Five subjects were excluded based on these criteria, resulting in 50 preoperative subjects remaining for analysis. Gait analyses were repeated in 24 of the 50 subjects 1–2 years after undergoing successful unilateral THR for end-stage OA. After applying the exclusion criteria, 2 subjects who, in the interim, had developed significant pain at an additional large joint of the lower extremity (as assessed by the Harris Hip Score) were excluded from postoperative analyses. Thus, data on 22 postoperative subjects were available for analysis.

Gait analysis.

Motion was measured with a multicamera optoelectronic system (Computerized Functional Testing Corporation, Chicago, IL) and force with a multicomponent force plate (Bertec, Columbus, OH) (15). External reflective markers were placed on the lower extremities, including the lateral-most aspect of the iliac crest, greater trochanter, lateral joint line of the knee, lateral malleolus, calcaneus, and base of the fifth metatarsal, and joint centers were estimated based on anatomic measurements of each subject. Subjects were instructed to walk at a range of self-selected speeds from slow to fast (16), and data from 6 stride lengths on each side were collected. Representative walking trials at a speed of ∼1 meter/second were chosen for analysis. The average (±SD) speeds of the trials analyzed were 0.98 ± 0.22 meter/second and 0.98 ± 0.24 meter/second on the contralateral and ipsilateral sides of the preoperative group, respectively (P = 0.844) and 1.05 ± 0.17 meter/second and 1.04 ± 0.14 meter/second, respectively, for the postoperative group (P = 0.608).

These position and force data were then utilized to calculate 3-dimensional external moments using inverse dynamics. The external moments (or external torques) that act on a joint during gait (17, 18) are, according to Newton's second law, equal and opposite to the net internal moments produced primarily by the muscles, soft tissues, and joint contact forces (19). The external moments (Nm) are normalized by dividing the subject's body weight (BW) and height and are expressed as a percentage by multiplying by 100. The new units, %BW × height, allow for comparisons between subjects (20).

Whereas the adduction moment describes the distribution of load between the medial and lateral knee compartments, analytical models further estimate the muscle forces and resulting tibial–femoral contact force in the medial and lateral compartments of the knee. A previously published statically determinate muscle model was used to predict the joint reaction force in the medial compartment of the knee during gait (21). In this model, the peak component of the external moments, the knee flexion angle, and the tibial–femoral intersegmental force are input. The output includes the forces on the medial and lateral knee compartments. The latter forces, after normalization, are expressed as body weight. The contact forces on the medial compartment in the current analysis are those occurring without antagonistic muscle activity and, thus, represent a conservative estimate of the joint contact forces.

Statistical analysis.

Two-tailed paired samples t-test was used to identify differences in the peak moments and peak medial compartment load at the knee ipsilateral and contralateral to the preoperative hip. A significance level of α < 0.05 was established a priori.

RESULTS

  1. Top of page
  2. Abstract
  3. SUBJECTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Preoperative subjects.

The baseline characteristics of the 50 preoperative subjects are outlined in Table 1. Subjects were between the ages of 42 and 74 years (mean 62 years). There were 29 men and 21 women, with a mean body mass of 86 kg and a mean height of 1.7 meters. The gait analyses were performed an average (±SD) of 23 ± 23 days (range 0–101 days) before the THR surgery.

Table 1. Baseline characteristics of the 50 preoperative total hip replacement subjects
VariableResult
Age, years 
 Mean ± SD62 ± 8
 Range42–74
Sex, no. of males/no. of females29/21
Body mass, kg 
 Mean ± SD86 ± 14
 Range63–124
Height, meters 
 Mean ± SD1.7 ± 0.1
 Range1.4–1.9

In the 50 subjects who were preoperative for unilateral THR, peak external knee moments were generally higher at the contralateral knee relative to the ipsilateral knee (Table 2). The knee adduction moment, in particular, was significantly higher in the contralateral knee (2.7 ± 0.9 %BW × height) compared with the ipsilateral knee (2.3 ± 1.0 %BW × height) (P = 0.003). In addition, the peak flexion moment (2.3 ± 1.5 versus 1.5 ± 0.9 %BW × height; P < 0.001), the peak extension moment (3.1 ± 1.0 versus 2.1 ± 0.7 %BW × height; P < 0.001), and the peak internal rotation moment (0.9 ± 0.4 versus 0.7 ± 0.4 %BW × height; P = 0.001) were significantly higher in the contralateral knee. The estimated peak medial compartment knee load was also found by mathematical modeling to be significantly higher in the contralateral knee (2.4 ± 0.5 BW) relative to the ipsilateral knee (1.9 ± 0.5 BW) (P < 0.001).

Table 2. Results of peak knee moments of, and medial compartment load on, the contralateral and ipsilateral knees of the 50 preoperative total hip replacement subjects*
Peak external knee moments and loadContralateral kneeIpsilateral knee
  • *

    Values are the mean ± SD %body weight × height, except for the medial compartment load, which is the mean ± SD body weight.

  • P < 0.05 versus the ipsilateral knee.

Flexion moment2.3 ± 1.51.5 ± 0.9
Extension moment3.1 ± 1.02.1 ± 0.7
Abduction moment0.6 ± 0.40.5 ± 0.4
Adduction moment2.7 ± 0.92.3 ± 1.0
Internal rotation moment0.9 ± 0.40.7 ± 0.4
External rotation moment0.1 ± 0.10.1 ± 0.1
Medial compartment load2.4 ± 0.51.9 ± 0.5

Postoperative subjects.

Twenty-two of the 50 subjects had a second gait test performed an average (±SD) of 15 ± 3 months (range 10–23 months) after unilateral THR for end-stage OA. Baseline demographic data on these subjects are given in Table 3 and, overall, are comparable to the data on the preoperative group (P > 0.05). There were 16 men and 6 women who ranged in age from 54 to 76 years (mean 65 years). Their average body mass was 86 kg and average height 1.7 meters.

Table 3. Baseline characteristics of the 22 postoperative total hip replacement subjects
VariableResult
Age, years 
 Mean ± SD65 ± 7
 Range54–76
Sex, no. of males/no. of females16/6
Body mass, kg 
 Mean ± SD86 ± 15
 Range63–126
Height, meters 
 Mean ± SD1.7 ± 0.1
 Range1.5–1.9

Postoperative gait data from these subjects were analyzed in a manner similar to that of the preoperative gait data. These subjects were essentially pain-free at the site of the replaced hip, according to the Harris Hip Score (mean ± SD score 96 ± 4). Table 4 shows their gait analysis data.

Table 4. Results of peak knee moments of, and medial compartment load on, the contralateral and ipsilateral knees of the 22 postoperative total hip replacement subjects*
Peak external knee moments and loadContralateral kneeIpsilateral knee
  • *

    Values are the mean ± SD %body weight × height, except for the medial compartment load, which is the mean ± SD body weight.

  • P < 0.05 versus the ipsilateral knee.

Flexion moment1.7 ± 1.11.7 ± 1.1
Extension moment3.4 ± 1.12.9 ± 1.0
Abduction moment0.4 ± 0.30.5 ± 0.3
Adduction moment2.9 ± 0.82.4 ± 0.9
Internal rotation moment0.9 ± 0.30.9 ± 0.3
External rotation moment0.2 ± 0.10.1 ± 0.1
Medial compartment load2.3 ± 0.42.0 ± 0.4

The peak external knee adduction moment (2.9 ± 0.8 versus 2.4 ± 0.9 %BW × height; P = 0.015) and the peak medial compartment load (2.3 ± 0.4 BW versus 2.0 ± 0.4 BW; P = 0.004), as well as the peak knee extension moment (3.4 ± 1.1 versus 2.9 ± 1.0 %BW × height; P = 0.042) in the postoperative group were, again, significantly higher in the contralateral knee relative to the ipsilateral knee. However, there were no longer significant differences between the contralateral and ipsilateral knee in the peak flexion moment or the peak internal rotation moment (P = 0.989 and P = 0.614, respectively).

DISCUSSION

  1. Top of page
  2. Abstract
  3. SUBJECTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

This is the first study to identify asymmetric dynamic loading of the knee joints in subjects with unilateral end-stage hip OA. These asymmetries in knee loading persist even after the hip pain is resolved by a THR, suggesting that subjects with advanced hip OA may undergo prolonged gait adaptations that could affect disease progression in the other joints of the lower extremities. Specifically, the external knee adduction moment (6–11) and the peak force on the medial compartment were significantly higher in the contralateral knee relative to the ipsilateral knee of subjects preoperative for a unilateral THR, and these values remained significantly elevated up to 23 months after successful THR that rendered the subjects pain-free at the hip. These results can account for the previously observed nonrandom evolution of knee OA seen in a similar population of patients with end-stage hip OA, in which the contralateral knee was almost 2.5 times more likely to develop end-stage OA than was the ipsilateral knee (1).

Results of small studies have indicated that a disruption in lower extremity coordination may lead to subsequent pathologic changes in the joints of the lower extremities. For example, a group of athletes with unilateral lower leg amputation was shown to have a higher incidence of radiographic OA of the “normal” (nonamputated) knee compared with age-matched healthy controls (22). Furthermore, results of gait evaluation in patients who had undergone lower leg amputation suggested that the patients place relatively higher loads on the nonamputated extremity compared with a normal control population (23). This was assumed to be a result of increased magnitude and duration of functional loading of the nonamputated extremity during day-to-day activities. Similarly, Loizeau et al (24) examined 4 postoperative patients who had undergone unilateral THR for end-stage OA and noted the presence of “mechanical dysfunction” based on altered energy absorbed or generated in the large joints of the nonoperated limb (24).

Although these previous investigations hinted at the importance of the biomechanical interrelationship between the joints of the lower extremities, this is the first systematic study to quantify asymmetries in knee joint loading in subjects with hip OA. The specific mechanisms that initiate and maintain differential limb loading are not completely clear. Asymmetric knee loading may be a consequence of gait alterations in response to chronic hip pain and structural pathology. Thus, subjects may unload the symptomatic extremity and relatively “overload” the contralateral limb. Gait analysis has previously shown that the mean peak external knee adduction moment in healthy adults is ∼3.0 %BW × height (8) and approaches 4.0 %BW × height in subjects with radiographic evidence of medial compartment knee OA (8, 25), in those who undergo radiographic deterioration after longitudinal followup (9), and in those who have a poor prognosis after high tibial osteotomy (11).

In the current investigation, although the knee contralateral to the limb with end-stage hip OA had relatively higher loads than did the knee ipsilateral to the end-stage hip, subjects had external knee adduction moments below 3.0 %BW × height in both the contralateral and ipsilateral knees and, therefore, appeared to be “underloading” both knees compared with a normal population or with a knee OA population. Although these subjects appear to be relatively more disposed to developing OA in the knee contralateral, rather than ipsilateral, to their end-stage hip, it is not clear that they are at higher risk for any knee OA compared with the general population (i.e., without hip OA). Perhaps the proportion of subjects who eventually develop end-stage knee OA has biochemical abnormalities in the cartilage that make it more susceptible to wear and tear from nominally “normal” loads (26, 27).

If preoperative pain and pathology in the diseased hip are responsible for differential knee loading in subjects with end-stage hip OA, why does that differential loading persist postoperatively after THR? Perhaps an initial gait adaptation occurs in response to hip pain and, possibly, structural disease that is subsequently maintained through more permanent neuromechanical changes that function in ambulation and are unaffected by THR (28–30). Alternatively, since the peak external knee adduction moment has been correlated with the presence of radiographic knee OA, the relatively higher external knee adduction moment in the contralateral knee may be a reflection of asymptomatic radiographic knee OA that has already developed, is present at the time of THR, and will, in some patients, eventually be manifested by clinical symptoms. Unfortunately, knee radiographs were not available for the subjects of the present study. Regardless of the underlying cause, a better understanding of the specific neuromechanical factors involved in the asymmetric loading in this population may help target future pre- and postoperative physical therapy aimed at the prevention of knee OA (31–33). Similarly, if it is the case that subjects already have established knee OA at the time of THR, then perhaps recognizing this as a potential consequence of end-stage hip OA would encourage radiographic evaluation of the knees and, perhaps, earlier consideration of THR in this population.

Several of the knee moments actually did become symmetric after THR and were no longer significantly different between the knees. The reason for this is unclear. However, it is important that differences in both the peak external knee adduction moment and the peak medial compartment load in the two knees persisted after successful THR. These knee loading parameters are regarded as physiologically the most important parameters in the development and progression of medial compartment knee OA (6–11, 25).

In conclusion, in light of what is known about the effects of loading on knee OA (6–11) and the previous observation of a nonrandom pattern in the evolution of lower extremity OA that favors progression in the contralateral knee (1), the present finding of relatively increased dynamic loading in the contralateral knee of subjects with end-stage hip OA provides further, compelling evidence that dynamic loading is an important risk factor in knee OA and may explain the nonrandom evolution of OA in multiple joints of the lower extremities. Furthermore, the persistence of asymmetric knee loading after THR indicates that fundamental changes in the neuromuscular control of gait or irreversible asymptomatic radiographic knee OA may already be present at the time of development of end-stage hip OA. Future investigations should focus on delineating the specific neuromuscular, biomechanical, and structural factors that may be involved.

REFERENCES

  1. Top of page
  2. Abstract
  3. SUBJECTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  • 1
    Shakoor N, Block JA, Shott S, Case JP. Nonrandom evolution of end-stage osteoarthritis of the lower limbs. Arthritis Rheum 2002; 12: 31859.
  • 2
    Frost HM. Perspectives: a biomechanical model of the pathogenesis of arthroses. Anat Rec 1994; 240: 1931.
  • 3
    Radin EL, Burr DB, Caterson B, Fyhrie D, Brown TD, Boyd RD. Mechanical determinants of osteoarthrosis. Semin Arthritis Rheum 1991; 21 Suppl 2: 1221.
  • 4
    Morrison JB. Bioengineering analysis of force actions transmitted by the knee joint. Biomed Eng 1968; 3: 16470.
  • 5
    Stauffer RN, Chao EY, Gyory AN. Biomechanical gait analysis of the diseased knee joint. Clin Orthop 1977; 126: 24655.
  • 6
    Sharma L, Hurwitz DE, Thonar EJ-MA, 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.
  • 7
    Hurwitz DE, Ryals AR, Block JA, Sharma L, Schnitzer TJ, Andriacchi TP. Knee pain and joint loading in subjects with knee osteoarthritis. J Orthop Res 2000; 18: 5729.
  • 8
    Baliunas AJ, Hurwitz DE, Ryals AB, Karrar A, Case JP, Block JA, et al. Increased knee joint loads during walking are present in subjects with knee osteoarthritis. Osteoarthritis Cartilage 2002; 10: 5739.
  • 9
    Hong R, Hurwitz DE, Block JA, Case JP. High dynamic loading: a risk factor in the progression of knee osteoarthritis (OA) [abstract]. Arthritis Rheum 2001; 44: 2950.
  • 10
    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.
  • 11
    Prodromos CC, Andriacchi TP, Galante JO. A relationship between gait and clinical changes following high tibial osteotomy. J Bone Joint Surg Am 1985; 67: 118894.
  • 12
    Altman R, Alarcón G, Appelrouth D, Bloch D, Borenstein D, Brandt K, et al. The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the hip. Arthritis Rheum 1991; 34: 50514.
  • 13
    Altman R, 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.
  • 14
    Soderman P, Malchau H. Is the Harris hip score system useful to study the outcome of total hip replacement? Clin Orthop 2001; 384: 18997.
  • 15
    Andriacchi TP, Natarajan RN, Hurwitz DE. Musculoskeletal dynamics, locomotion, and clinical applications. In: MowVC, WilsonCH, editors. Basic orthopaedic biomechanics. Philadelphia: Lippincott-Raven; 1997. p. 3768.
  • 16
    Andriacchi TP, Strickland AB. Gait analysis as a tool to assess joint kinetics. In: BermeN, EnginAE, Correia da SilvaKM, editors. Biomechanics of normal and pathological human articulating joints. Dordrecht: Nijhoff; 1985. p. 83.
  • 17
    Johnson F, Leitl S, Waugh W. The distribution of load across the knee: a comparison of static and dynamic measurements. J Bone Joint Surg Br 1980; 62: 3469.
  • 18
    Andriacchi TP. Dynamics of knee malalignment. Orthop Clin North Am 1994; 25: 395403.
  • 19
    Harrington IJ. Static and dynamic loading patterns in knee joints with deformities. J Bone Joint Surg Am 1983; 65: 24759.
  • 20
    Moisio KC, Sumner D, Shott S, Hurwitz DE. Normalization of joint moments during gait: a comparison of two techniques. J Biomech 2003; 36: 599603.
  • 21
    Schipplein OD, Andriacchi TP. Interaction between active and passive knee stabilizers during level walking. J Orthop Res 1991; 9: 1139.
  • 22
    Melzer I, Yekutiel M, Sukenik S. Comparative study of osteoarthritis of the contralateral knee joint of male amputees who do and do not play volleyball. J Rheumatol 2001; 28: 16972.
  • 23
    Burke MJ, Roman V, Wright V. Bone and joint changes in lower limb amputees. Ann Rheum Dis 1978; 37: 2524.
  • 24
    Loizeau J, Allard P, Duhaime M, Landjerit B. Bilateral gait patterns in subjects fitted with a total hip prosthesis. Arch Phys Med Rehab 1995; 76: 5527.
  • 25
    Hurwitz DE, Ryals AR, Case JP, Block JA, Andriacchi TP. The knee adduction moment during gait in subjects with knee osteoarthritis is more closely correlated with static alignment than radiographic disease severity, toe out angle and pain. J Orthop Res 2002; 20: 1017.
  • 26
    Trickey TR, Lee M, Guilak T. Viscoelastic properties of chondrocytes from normal and osteoarthritic human cartilage. J Orthop Res 2000; 18: 8918.
  • 27
    Barker MK, Seedhom BB. The relationship of the compressive modulus of articular cartilage with its deformation response to cyclic loading: does cartilage optimize its modulus so as to minimize the strains arising in it due to the prevalent loading regime? Rheumatology (Oxford) 2001; 40: 27484.
  • 28
    Hurley MV. The role of muscle weakness in the pathogenesis of osteoarthritis. Rheum Dis Clin North Am 1999; 25: 28398.
  • 29
    O'Connor BL, Visco DM, Rogers PI, Mamlin LA, Brandt KD. Serial force plate analyses of dogs with unilateral knee instability, with or without interruption of the sensory input from the ipsilateral limb. Osteoarthritis Cartilage 1999; 7: 56773.
  • 30
    O'Connor BL, Brandt KD. Neurogenic factors in the etiopathogenesis of osteoarthritis. Rheum Dis Clin North Am 1993; 19: 581605.
  • 31
    Messier SP, Loeser RF, Mitchell MN, Valle G, Morgan TP, Rejeski WJ, et al. Exercise and weight loss in obese older adults with knee osteoarthritis: a preliminary study. J Am Geriatr Soc 2000; 48: 106272.
  • 32
    Deyle GD, Henderson NE, Matekel RL, Ryder MG, Garber MB, Allison SC. Effectiveness of manual physical therapy and exercise in osteoarthritis of the knee: a randomized, controlled trial. Ann Intern Med 2000; 132: 17381.
  • 33
    Hurley MV, Scott DL. Improvements in quadriceps sensorimotor function and disability of patients with knee osteoarthritis following a clinically practicable exercise regime. Br J Rheumatol 1998; 37: 11817.