Osteoarthritis (OA) is a degenerative chronic disease of multifactorial etiology, with interrelationship between systemic and local factors 1. Systemic factors include ethnicity, older age, gender, hormonal status, genetic factors, bone density and nutritional factors Local factors result in abnormal mechanical stress on the affected joints, and include obesity, changes in the articular mechanics (ligamentous laxity, misalignment joint and muscle weakness), proprioceptive deficits, injury history of the joint, occupational factors and effects of physical and sporting activities 1,2,3,4.
In the initial stage of OA occurs fibrillation of the articular cartilage surface with deep fissures and calcification of the cartilage layer and subchondral bone sclerosis and osteophyte formation develop with disease progression 1. These structural changes of the articular cartilage and subchondral bone entail an inflammation of the synovial membrane. The result of this pathological process of OA is the reduction of the articular cartilage ability to dissipate loads and maintain joint congruity during weight-bearing activities 5. Studies in developed 6,7 and emerging 8 countries show that the prevalence of symptomatic knee OA taken from radiography in the elderly, ranged from 5.6% to 8.7% in men and 11.4 to 15%, 4% in women.
The clinical criteria for the diagnosis of OA include pain, age over 50 years, morning stiffness (lasting less than 30 minutes) or the crackling active movement 9. The criteria on the degree of severity of knee OA proposed by Kellgren and Lawrence are extensively used these days and include: grade 0 = no osteophytes (no radiographic changes), 1 = possible osteophytes (uncertainty in radiographic change), 2 = definite osteophytes and possible reduction of the articular space (mild OA), 3 = multiple bone spurs and joint space narrowing, sclerosis and possible deformity of the bone contour (moderate OA), and 4 = large osteophytes, severe sclerosis, defined deformity the bone contour and significant reduction in joint space (severe OA) 10.
In the knee joint, the medial tibiofemoral compartment is the place most affected by OA 11. This susceptibility may be explained by the greater body load that occurs, between 60 and 80% of the gait cycle, in the medial compartment 12,13.
The increase in joint mechanical loads on the medial compartment of the knee joint is considered an important factor in the development and subsequent progression of the medial knee OA 14,15. Studies quantifying articular loads in subjects with knee OA have received attention in recent years 13,15,16,17. The overload joint can be estimated by calculating the external adductor moment of the knee, obtained through the assessment of gait kinematics and kinetics 11,18. The moment of force is a result of the combination of kinematic variables, ground reaction forces (GRF) and inertial properties of the body segments 19 and is calculated using the inverse dynamics approach. Applying the third Newton's law, the external moment of force is balanced by an internal moment of equal magnitude but opposite in direction and generated by the muscles, bones and soft tissues 20. Zhao et al. 21 showed a high correlation between the external knee adductor moment and the joint contact forces in the medial compartment of the knee during walking (R = 0.77). Miyazaki et al. 14 found that increased external adductor moment of the knee is a risk factor in the progression of the medial knee OA and an increase of 1% in the kinetic parameter increases 6.5 times the risk of disease progression.
Recent studies 17,18,22 reported that individuals with medial knee OA have greater external adductor knee moment throughout the stance phase of gait compared with asymptomatic individuals. Müdermann et al. 23 analyzed the external adductor moment of the knee during gait of 44 patients with OA of the medial knee compartment of varying degrees of severity. The results showed that individuals with mild to moderate OA had lower external knee adductor moment than individuals at an advanced stage of the disease. Thus, the external adductor moment of the knee during walking increases with disease progression. Added to this, some individuals in this study were able to alter the mechanics of the gait in order to reduce load on the medial compartment of the knee in the early stages of the disease.
Therefore, it is important to study the strategies that reduce the external adductor moment of the knee during gait in individuals with medial knee OA to guide the rehabilitation process, in order to decrease pain and slow the progression of this disease 11,24,25,26. Previous studies had indicated that some spatial-temporal, kinematic and kinetic gait parameters might reduce the external adductor moment of the knee 11,16,24,25. Thus, this review aims to identify these biomechanical strategies and discuss how clinicians could work with their patients to reduce the pain and consequences of the medial knee OA.
A non-systematic review of the literature was conducted using databases MEDLINE, PUBMED and PEDro. Articles analyzed were published between 2000 and 2011. The keywords used in the search were: osteoarthritis (OA), knee, gait, biomechanics and kinetics.
The literature search was focused on cross-sectional observational studies with following criteria: compared the kinematic, kinetic, spatial and temporal variables during gait in individuals with medial knee OA of varying degrees severity with asymptomatic. It was also included studies that identified the correlation between biomechanical variables and reduced load on the medial knee compartment and on the external adductor at the knee, longitudinal studies examining the effect of strategies that reduced the knee external adductor moment, clinical trials that assessed the efficacy of interventions focused on strategies that reduced load on the medial knee improving pain and function of the individuals.
Articles written in English or Portuguese were selected for the review. However, articles with methodological deficiency, for example, lack of homogeneity among groups of individuals studied, lack of a control group that difficult the integrity of the results and selection of individuals with OA on the tibiofemoral side of the knee, or patellofemoral, were excluded from this literature revision.
The titles and abstracts of the articles were analyzed by the authors. Studies that met the inclusion criteria were obtained in full. Each article was evaluated separately according to the methodological quality, and subsequently, the authors discussed the relevance of the articles for the study. If there was a disagreement between the authors, the articles were analyzed until they reached a consensus. Thus, the articles were included in the review of the literature when both authors agree to do so. Articles reviewed should contain information relevant to the topic that could be compared to other studies in other to be included in the review.
Research in the databases resulted in 42 studies relevant to the topic discussed, all in English. However, atfter careful assessment of titles and abstracts by the authors of this study, on 23 articles were included for the preparation of this review. The main grounds for exclusion of some items were the lack of clarification of the knee joint compartment affected by OA and non-allocation of a control group. The lack of a control group undertakes the causal conclusions of a study.
Of the articles investigate, there was one systematic review 27, one longitudinal study 11, two clinical trials 28,29, four correlational studies 23,24,25,30 and the remaining 15 studies were cross-sectional 16,17,18,22,26,28,29,31,32,34,35,36,37,38. The 23 selected articles are summarized on Table 1,1a.
Spatial and temporal gait parameters
Studies have shown significant differences in gait of individuals with and without knee OA. The spatial and temporal changes included: reduced gait speed 22,28,29,31,41, step length and cadence 20,31,42,43 and longer time spent in double support phase 22,32,43.
These changes, however, are dependent on the severity of OA. Recently, Thorp et al. 31 showed a positive correlation (r = 0.302, p < 0.01) between the degree of severity of OA and the duration of the stance phase and a negative correlation with gait speed (r = - 0.341, p < 0.01). Similarly, Astephen et al. 17 demonstrated that individuals with medial knee OA tend to slow the gait with the advance of the disease. The authors attributed these changes to compensatory responses during gait in an attempt to stabilize the knee joint and therefore to decrease pain.
Changes in gait speed have been associated with the biomechanics of the lower extremities in patients with knee OA and asymptomatic 44,45. Zeni and Higginson 33 demonstrated the importance of gait speed in analyzing parameters that are dependent on the magnitude of the GRF and acceleration of body segments in individuals with medial knee OA and asymptomatic. Asymptomatic individuals with moderate to severe knee OA were assessed on self-selected speeds, controlled (1m/s) and the fastest speed tolerated. The fastest speed was determined as the speed that the individuals felt comfortable and still able to maintain the double support phase. The results showed that the magnitude of the moments of the ankle and knee, the angular displacement of the hip in the sagittal plane and frontal of the knee and the vertical anteroposterior peak of the GRF were dependent on speed. Asymptomatic individuals with moderate OA showed a significant increase in all the variables studied at the fastest speed compared to self-selected speed. However, in the severe OA group this difference was not observed. When the speed was controlled (1m/s) no difference between groups in the variables studied were found. The authors concluded that gait speed is a contributing factor to differences in gait parameters between subjects with and without knee OA.
The GRF is the product of mass times the acceleration of body motion, thus increasing acceleration of the body's center of mass results in a higher magnitude of the GRF, and consequently greater moment of forces. Mündermann et al. 23 reported a linear correlation between the peak external adductor moment of the knee and walking speed, suggesting that individuals with mild to moderate OA of the medial compartment are able to decrease this moment by reducing gait speed. Similar results were reported by Thorp et al. 31 and Deluzio and Astephen 46.
Robbins and Maly 34 found that by reducing the speed of gait of healthy individuals, there was a decrease in peak load on the knee but there was an increase in impulse, i.e., the knee received less load but for a longer time.
Angular displacement of the foot and trunk
Increased foot external rotation angle and lateral inclination of the trunk towards the stance limb are kinematic parameters that reduce the external knee adductor moment consequently decrease pain and slow progression of OA medial knee 16,24.
During gait, the vector of the GRF has its point of application on the center of pressure on the sole of foot and travels towards the body center of mass 25. The length of the lever arm of the GRF with respect to the center of rotation of the knee joint in the frontal plane is directly proportional to the magnitude of the external adductor moment of the knee 30. Changes in pressure center positions in the foot or the body's center of mass during walking alter the length of the GRF lever arm. Therefore, an increase in the external rotation angle of the foot shifts laterally the centre of pressure and reduces the length of the GRF lever arm, or the distance of the GRF to the center of rotation of the knee joint, thus reducing the external adductor moment of the knee 24,26.
In the study by Chang et al. 24 a negative correlation between the external rotation angle of the foot and the external adductor moment of the knee in individuals with medial knee OA during the terminal phase of gait support was found. In addition, it was demonstrated that the increase in the external rotation angle of the foot increases the likelihood of progression of the medial knee OA over a period of 18 months. Similarly, Lynn and Costigan 35 observed that increasing the external rotation angle of the foot reduced the electromyography activity of the medial hamstring muscles and increased the activity of the femoral biceps muscle. These changes in neuromuscular control generate an internal abductor, which minimizes the high loads in the medial compartment of individuals with medial knee OA 20,29,32,36.
Hunt et al. 25 in a study with 114 patients with medial knee OA demonstrated a strong correlation between the increased foot external rotation angle and the reduction of the external adductor moment of the knee during the terminal stance, however, lateral inclination of the trunk towards the stance limb was the variable that best explained the variation of the external adductor moment of the knee. In addition, in the systematic review presented by Simic et al. 27 it was reported that increased trunk lateral inclination was the gait change with best scientific evidence in the reduction of the external adductor moment of the knee at initial stance.
The lateral inclination of the trunk towards the stance limb reduces external adductor moment of the knee by moving the body's center of mass laterally, reducing the lever arm of the GRF with respect to the center of knee rotation 25. Mündermann et al. 37 analyzed the relationship between external adductor moment of the knee during gait with the increased and normal trunk inclination in the stance phase of 19 healthy individuals. The results showed that 10 degrees of trunk inclination reduced on average 65% external adductor moment of the knee. In the study by Tanaka et al. 16 there was no significant difference in trunk inclination between asymptomatic individuals with medial knee OA. This result could be explained by the limitations of the study, as cited by the authors, for example, the small number of individuals studied (OA group 12 individuals and 5 in the control group). On the other hand, Linley et al. 22 compared the lateral inclination of the trunk and pelvis during gait among 40 individuals with medial knee OA and 40 asymptomatic pared by age and gender. Using principal component analysis, the results indicated that individuals with medial knee OA had greater lateral inclination of the trunk and pelvis towards the stance limb during initial and mid stance when compared to asymptomatic individuals.
The literature indicates that increased trunk inclination is a modification adopted by individuals with medial knee OA 16,22,41. However, it is unknown if this gait modification would be beneficial in the long run or an effective intervention for those individuals. More studies are needed to understand the role of lateral trunk inclination as an effective intervention for patients with medial knee OA.
Another strategy used to prevent the progression of knee OA is increased hip internal abductor 11,18. The hip abductor muscles during the stance phase of gait are responsible for preventing the drop of the opposite pelvis, thus ensuring a smooth ride of the center of gravity 47. Weakness or decreased force of the hip abductor muscles during stance may lead to a drop of the contralateral pelvis. This would result in a displacement of the body's center of gravity toward the member that is in the swing phase, resulting in an increase of the GRF lever arm with respect to the center of rotation of the knee joint, and therefore increasing the external adductor moment of the knee in stance.
Chang et al. 11 demonstrated the importance of the internal abductor moment of the hip in a longitudinal study conducted with 57 individuals with medial knee OA. The joint space narrowing at the knee joint was the variable used to determine the progression of OA 18 months after the first data collection. The results showed that individuals who had higher peak hip internal abductor moment were less likely to disease progression (odds ratio = 0.48; range of confidence = 0.16 to 0.81). The authors concluded that increasing the forces at the hip abductors while walking reduces the load on the articular medial compartment of the knee, resulting in protecting against the progression of OA.
Astephen et al. 17 reported significant difference in the internal abductor moment of the hip during the loading response and mid stance of gait between individuals with moderate and severe Knee OA. The asymptomatic individuals had greater internal hip abductor moment than the individuals with knee OA. Similar results have been reported by Briem and Snyder-Mackler 38. The authors attributed this finding to the increased lateral trunk inclination to the affected OA side during the stance phase. The lateral shift places the weight of the body closest to the hip joint, reducing the length of the GRF lever arm in relation to the center of rotation of the hip joint and consequently the external force decreases requiring less from the abductors muscles of the hip 38. Sled et al. 39 reported that individuals with medial knee OA have weakness of the abductor muscles of the hip compared to asymptomatic individuals.
Although the strategy of trunk inclination reduces the magnitude of the knee load, it also reduces the demand over the hip abductor muscles during the stance phase of gait leading to weakness of the abductor muscles 38. Therefore, the most efficient way to stop this vicious cycle would be by strengthening the abductor muscles of the hip to prevent the development of compensation in individuals with medial knee OA.
In Thorp's et al. 31 the pilot study a program to strengthen the quadriceps muscles, hamstrings and emphasizing the hip abductors in open and closed kinetic chain resulted in significant reduction of pain and in the external adductor moment of the knee in OA individuals during gait. However, the study had a lower statistical power, a small sample size and no control group, therefore a cause and effect relationship conclusion is impossible.
Moreover, Sled et al. 39 investigated the effect of a home-based program of strengthening the hip abductor muscles on function, pain, on the concentric isokinetic strength and on the external adductor moment during gait in a group of 40 individuals with medial knee OA. The control groups consisted of 40 individuals without diagnostic of OA pared by age, gender with the OA group. The strengthening program had an 8-week period. After the program the medial knee OA group showed significant improvement in pain and function assessed by the test sit-to-stand 5 times (5-STS compared to the control group. The concentric isokinetic strength of the hip abductor muscles increased significantly after the strengthening program in the OA group. However, there was no significant reduction in the external adductor moment of the knee in the OA group. The authors explain this finding by the theory that increasing the strength of the abductor hip muscles could have increased the stability of the trunk, decreasing the lateral bending of the trunk to the stance limb, hence there was an increase in length of the GRF lever arm relative to the knee center of rotation, canceling the reduction in the external adductor moment after knee strengthening program.
Although the interventions that reduce the external adductor knee moment in individuals with medial knee OA include the use of bracing 48, insoles 49 and strengthening of the quadriceps muscle 50, this literature review demonstrates the importance of strengthening the hip abductor muscles for the prevention of compensations, in reducing pain and improving function and in the delay of progression of medial knee OA.
The increase load on the medial compartment of the medial knee during gait is a mechanical risk factor for progression of the disease. The evidence from the literature indicates that the decrease in walking speed, rising up the external rotation angle, the lateral inclination of the trunk towards the limb that is in the stance phase and the increase in the abductor internal moment of the hip are variables that reduce the adductor knee external force during gait in individuals with medial knee OA. Recent studies indicate that the strengthening of the abductor hip muscles reduces pain, improves function and prevents compensation in individuals with medial knee OA. However, there is a need for more randomized clinical trials, since most studies are observational in nature, to clarify the role of the moment of the abductor internal force in hip reduction of external force moment adductor knee during gait.