Effects of specialized footwear on joint loads in osteoarthritis of the knee




Elevated dynamic joint loads have been associated with the severity and progression of osteoarthritis (OA) of the knee. This study compared the effects of a specialized shoe (the mobility shoe) designed to lower dynamic loads at the knee with self-chosen conventional walking shoes and with a commercially available walking shoe as a control.


Subjects with knee OA were evaluated in 2 groups. Group A (n = 28) underwent gait analyses with both their self-chosen walking shoes and the mobility shoes. Group B (n = 20) underwent gait analyses with a control shoe and the mobility shoe. Frontal plane knee loads were compared between the different footwear conditions.


Group A demonstrated an 8% reduction in the peak external knee adduction moment with the mobility shoe compared with self-chosen walking shoes (mean ± SD 49 ± 0.80 versus 2.71 ± 0.84 %BW × H; P < 0.05). Group B demonstrated a 12% reduction in the peak external knee adduction moment with the mobility shoe compared with the control shoe (mean ± SD 2.66 ± 0.69 versus 3.07 ± 0.75 %BW × H; P < 0.05).


Specialized footwear can effectively reduce joint loads in subjects with knee OA, compared with self-chosen shoes and control walking shoes. Footwear may represent a therapeutic target for the treatment of knee OA. The types of shoes worn by subjects with knee OA should be evaluated more closely in terms of their effects on the disease.


Footwear has traditionally been engineered to provide maximum foot support and comfort, and medical alterations to footwear have been largely directed at remediating disorders involving the feet, such as diabetic neuropathy. However, the lower extremity is known to be an interrelated functional and mechanical unit, and alterations at one aspect of the lower extremity (e.g., the foot) can have significant impact on distant areas of the extremity such as the knee (1–7). Therefore, footwear design alone may substantially affect the loading patterns of the entire lower body, and these biomechanic effects may have important implications for conditions in which mechanical factors are important to the pathogenesis and progression of the disease, such as osteoarthritis (OA) of the knee. Nonetheless, the potential impact of footwear on knee OA has not been widely recognized or critically evaluated.

OA is the most common arthritic condition, and symptomatic OA of the knee is a significant source of disability and impaired quality of life (8–10). Despite this, OA treatment remains largely palliative and generally focuses on oral analgesics rather than on the aberrant biomechanic loading that underlies much of its progression. OA of the lower extremity is mediated, at least in part, through aberrant dynamic loads transmitted across the joints. In OA of the medial compartment of the knee, radiographic severity (11), disease progression (12), and pain (13, 14) have all been associated with elevated medial joint loads, whereas reducing loading of the medial compartment may result in symptomatic benefits (15–17).

Recent evidence suggests that modern shoes have a substantial influence on joint loading in subjects with knee OA (2, 6). This is particularly relevant because shoes are worn during most of the day in Western societies, especially during ambulation, the most common daily activity, when the load on the lower extremity joints is significantly greater than it is at rest (18). These observations suggest that rational shoe design may provide a novel approach to reducing biomechanic loading of the knee during gait, and thereby may provide potential palliation for knee OA.

We recently reported that the use of modern, comfortable walking shoes results in an ∼14% increase in dynamic loading of the knees during ambulation compared with barefoot walking among individuals with knee OA (2). These data suggest a biomechanic advantage to the enhanced mobility of a bare foot compared with the constraints and stability provided by modern footwear. Taking into account the loading dynamics of the foot, we designed a “mobility” shoe, intended to affect the biomechanic milieu of the entire lower extremity and thereby to reduce joint loads at the knee during ambulation. Here, we report the biomechanic effects of this mobility shoe, relative to self-chosen conventional walking shoes and to a control shoe, in individuals with knee OA in order to test the hypothesis that the loads borne by osteoarthritic knees can be substantially reduced through innovative footwear design.



Subjects were recruited through 2 separate studies. Both studies were approved through the institution's review board for studies involving human subjects, and informed consent was obtained from all subjects. Both studies evaluated a specialized, flexible shoe engineered in our laboratory to reduce lower extremity joint loads during walking, referred to as the mobility shoe in these studies. In study A, the mobility shoe was compared with conventional walking shoes; in study B, it was compared with a control shoe, a common, commercially available stability shoe (specific details are provided below). Inclusion and exclusion criteria for the studies were identical. Inclusion criteria included the presence of symptomatic OA of the knee, which was defined by the American College of Rheumatology (formerly the American Rheumatism Association) clinical criteria for the classification and reporting of OA of the knee (19) and by the presence of at least 20 mm of pain (on a 100-mm scale) while walking (corresponding to question 1 of the visual analog format of the knee-directed Western Ontario and McMaster Universities Osteoarthritis Index) (20). Although all subjects had bilateral knee OA, the most symptomatic knee on the day of the initial study visit was considered the index knee. OA of the index knee was documented by weight-bearing, full-extension anteroposterior knee radiographs of grade 2 or 3 as defined by the modified Kellgren/Lawrence (K/L) scale (21). The contralateral knee also had radiographic OA of K/L grade 1–3 in severity. Subjects had medial compartment OA defined as medial joint space narrowing (JSN) of ≥1, as well as medial JSN greater than lateral JSN by ≥1 grade (according to the Altman et al atlas) (22). Major exclusion criteria were 1) flexion contracture of >15 degrees at either knee 2) clinical OA of either ankle or either hip 4) significant intrinsic foot disease per a podiatric examination, and 5) body mass index (BMI) >35.


Subjects recruited into study A had gait analyses performed while wearing the mobility shoe, their self-chosen conventional walking shoes, and while barefoot; those in study B underwent gait analyses while wearing the mobility shoe, a commonly prescribed stability shoe, and while barefoot.

Mobility shoes.

The mobility shoe is a flexible, lightweight shoe specifically engineered by our laboratory to mimic essential aspects of barefoot walking (Figure 1). The outsole (the portion of the shoe that contacts the ground) of the mobility shoe consists of flexible polycarbon, with specialized grooves strategically placed at major flexion points of the foot to allow for natural, barefoot-like movement of the foot. The mobility design omits a multilayered, multidense midsole (the portion between the outsole and the subject's foot), using a single, thinner layer of uniform-density foam to avoid manipulating or altering the effects of the external ground reactive force at heel strike or midstance. The heel height varied between 1 and 1.5 cm depending on the shoe size. The shoe consisted of a mesh upper (the portion that wraps around the foot) with laces.

Figure 1.

The mobility shoe. Engineered to mimic essential aspects of barefoot walking.

Conventional shoes.

Subjects' conventional walking shoes varied upon personal preference, but ranged from traditional sneakers to loafers. Heel heights varied.

Stability shoes.

The control shoe in study B was a stability shoe, a brand-name walking shoe (Brooks Addiction Walker; Brooks Sports, Bothell, WA) (Figure 2) readily available and commonly prescribed for foot comfort and stability during walking. Like most stability shoes on the market, it incorporates cuts into the outsole to manipulate the motion across the ball of the foot during propulsion, while leaving the rear foot without significant cuts to promote stability in the latter area. The outsole is made of nonslip carbon rubber. The heel height for the women's shoes is 4 cm and for the men's is 5 cm. The midsole consists of a gel bladder for cushioning and an extended diagonal rollbar to limit pronation. The upper is made of supple leather. It is also reinforced to stabilize the heel and cushioned to improve fit.

Figure 2.

The stability shoe (the control) is commonly prescribed for foot comfort and stability during walking.

All subjects walked at their normal walking speed during gait analyses, and trials during the various footwear conditions were further matched for speed for comparison and analyses. Subjects were allowed to acclimate to the various footwear conditions between testing trials.

Gait analyses.

All subjects underwent gait analysis. Gait analyses were performed while wearing the study shoes in random order for each subject. Gait assessment included collection of 3-dimensional (3-D) kinematics and ground reaction forces (GRFs) using 4 Qualisys optoelectronic cameras (Innovision Systems, Columbiaville, MI) with passive markers, and a multicomponent force plate with a sampling frequency of 120 Hz (Bertec, Columbus, OH). Passive markers were placed at the lateral-most aspect of the superior iliac crest, the superior aspect of the greater trochanter, the lateral knee joint line, the lateral malleolus, the lateral calcaneus, and the head of the fifth metatarsal. For moment calculations, the joint centers of the hip, knee, and ankle were approximated following previously published methods (23, 24). The joint center of the ankle was determined to be the midpoint of the distance from the medial to lateral malleolus. The joint center of the knee was determined to be the midpoint of the distance between the medial and lateral joint lines of the tibiofemoral joint. The joint center of the hip was determined to be 2.5 cm distal to the midpoint of the distance between the anterior superior iliac spine and the pubic tubercle.

Subjects were instructed to walk at a self-selected, normal speed on a 2-inch thick wooden pressboard covered with linoleum. Kinetic components, calculated using processing software developed by the CFTC (Computerized Functional Testing Corporation, Chicago, IL), included frontal plane external joint moments at the knee (25).

These position and force data were then utilized to assess range of motion (ROM) at the joints and to calculate 3-D external moments using inverse dynamics. The external moments that act on a joint during gait 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. The external moments are normalized to the subject's body weight (BW) multiplied by height (H) times 100 (%BW × H) (26).

Primary gait endpoints.

The primary endpoints for the study were gait parameters that reflected the extent of medial compartment knee loading and included the peak external knee adduction moment and the adduction angular impulse. The magnitude of the peak external knee adduction moment has been shown to directly relate to the radiographic and clinical severity of medial knee OA (11), radiographic progression of knee OA (12), and to surgical outcomes (24, 27). The peak external knee adduction moment was defined as the external adduction moment of greatest magnitude during the stance phase of the gait cycle. However, peak joint moments do not include information on the duration of load. In contrast, the adduction angular impulse, defined as the integral of the knee adduction moment over time, is a marker of loading in knee OA that has recently been shown to provide additional information, and it may be more sensitive than the peak external knee adduction moment in predicting the severity of medial compartment radiographic OA grade (28). Adduction angular impulse was calculated by taking the time integral of all the normalized frontal plane joint moments for the whole stance phase of the gait cycle.

Secondary gait endpoints.

Other gait parameters were also measured and compared as secondary endpoints in the study. The GRF is a measure of the reaction to the force that the body exerts on the ground. The GRF was normalized to BW and expressed as %BW. The GRF is one of several components that determines moments, such as the peak external knee adduction moment, at the joints. Stride is the length of steps during walking, and cadence represents the number of steps per minute. ROM at the lower extremity joints was evaluated as well. These measures were included to gain better insight into how overall gait was changing during different footwear conditions.

Statistical analysis.

Statistical analysis was performed using SPSS software (SPSS, Chicago, IL). Descriptive data for all groups and variables were expressed as mean ± SDs for continuous measures. Differences in gait parameters (e.g., speed, cadence, etc.) and knee loads during the different footwear conditions and barefoot walking were measured using repeated-measures analysis of variance. If an overall effect was detected, group differences were further evaluated for significance with post hoc pairwise comparisons using the Newman-Keuls method (29). Relationships between gait parameters and joint moments during the various footwear conditions were evaluated using linear regression. P values less than 0.05 were considered significant.

A power analysis was conducted using the estimated effect sizes based on the peak external knee adduction moment for the 2 samples (30). The effect sizes were large (d = 1.01 for group A and d = 2.23 for group B) and yielded the following power estimates: a sample size of 28 subjects for group A and 20 subjects for group B resulted in at least 98% power to detect a significant difference at the 0.05 level for both groups.


Group A.

Group A consisted of 28 subjects (24 women and 4 men) with a mean ± SD age of 59 ± 9 years who had knee OA and underwent gait analyses (Table 1).

Table 1. Demographic data of the study participants*
 Group A (n = 28)Group B (n = 20)
  • *

    Values are the mean ± SD unless otherwise indicated.

Age, years59 ± 957 ± 9
Sex, female/male, n24/416/4
Weight, kg80 ± 1783 ± 16
Height, meters1.7 ± 0.11.7 ± 0.1
Body mass index, kg/m228.7 ± 5.129.6 ± 4.7

Results for group A are summarized in Table 2. There were no significant differences in speed during the 3 footwear conditions. In this group, both conventional shoes and the mobility shoe resulted in significantly higher peak external knee adduction moments compared with barefoot walking. On further comparison, the peak external knee adduction moment with the mobility shoes was 8% less than that with conventional shoes (P < 0.05). The adduction angular impulse was also 6% less (although nonsignificant) with the mobility shoe, and 13% less (P < 0.05) while barefoot, compared with the subjects' conventional shoes. The difference in the adduction angular impulse between the mobility shoes and barefoot walking was not significant.

Table 2. Results for group A (conventional shoes versus mobility shoes versus barefoot)*
 Conventional shoesMobility shoesBarefoot
  • *

    Values are the mean ± SD. H = height; ROM = range of motion; BW = body weight; GRF = ground reaction force.

  • Significant (P < 0.05) conventional shoes compared with barefoot walking and with mobility shoes.

  • Significant (P < 0.05) mobility shoes compared with barefoot.

  • §

    Significant (P < 0.05) conventional shoes compared with mobility shoes.

Speed, meters/second1.17 ± 0.241.16 ± 0.251.14 ± 0.25
Stride, meters/H0.77 ± 0.080.75 ± 0.080.72 ± 0.07
Cadence, steps/minute108 ± 9§110 ± 11113 ± 12
Knee ROM, degrees62 ± 5§61 ± 558 ± 5
Hip ROM, degrees30 ± 629 ± 628 ± 6
Ankle ROM, degrees30 ± 3§30 ± 426 ± 4
Peak external knee adduction moment, %BW × H2.71 ± 0.842.49 ± 0.802.39 ± 0.82
Adduction angular impulse, %BW × H0.93 ± 0.46§0.87 ± 0.450.81 ± 0.46
GRF, %BW109 ± 10106 ± 46105 ± 13

The GRF was found to be significantly lower both with the mobility shoes (3% lower) and during barefoot walking (4% lower) compared with conventional footwear (P < 0.05).

Assessment of other gait parameters revealed that stride (length of steps) increased and cadence (steps per minute) decreased with both conventional shoes and the mobility shoes compared with barefoot walking. Similarly, ROM during the gait cycle at the knee and ankle joints was significantly greater with both sets of shoes compared with barefoot walking.

Linear regression analysis was used to evaluate the influence of stride, cadence, and joint ROM on the knee loading parameters (peak external knee adduction moment and adduction angular impulse) during the different footwear conditions. Some significant correlations were observed but did not appear to be consistent between the footwear runs. Small associations were observed between knee ROM and peak external knee adduction moment while wearing conventional shoes (r = 0.313, P = 0.043), mobility shoes (r = 0.288, P = 0.058), and while walking barefoot (r = 0.269, P = 0.083). Similarly, knee ROM had a small association with adduction angular impulse while walking with conventional shoes (r = 0.289, P = 0.058), mobility shoes (r = 0.339, P = 0.031), and while barefoot (r = 0.267, P = 0.085). The only other significant correlations were seen between cadence and adduction angular impulse during all the footwear conditions (conventional shoes, r = −0.322; mobility shoes, r = −0.389; barefoot, r = −0.434; P < 0.05 for all correlations). Because cadence represents steps per minute and the adduction angular impulse is an integral of the peak external knee adduction moment over time, it was anticipated that these 2 parameters would be inversely correlated.

Group B.

Group B consisted of 20 subjects (16 women and 4 men) with a mean ± SD age of 57 ± 9 years who underwent gait analyses with the mobility shoe and the stability shoe (Table 1). There were no substantial demographic differences between group A and group B.

Results for group B are summarized in Table 3. Similar to group A, there were no significant differences in speed during the walking conditions. The peak external knee adduction moment was 13% lower while wearing the mobility shoes and 12% lower while walking barefoot compared with that while wearing the stability shoes (P < 0.05 for both comparisons). Similarly, the adduction angular impulse was 10% and 12% lower, respectively, with the mobility shoes and barefoot walking compared with the stability shoes (P < 0.05 for both comparisons). There were no significant differences in the knee loads observed between mobility shoes and barefoot walking.

Table 3. Results for group B (stability shoes versus mobility shoes versus barefoot)*
 Stability shoesMobility shoesBarefoot
  • *

    Values are the mean ± SD. See Table 2 for definitions.

  • Significant (P < 0.05) stability shoes compared with barefoot.

  • Significant (P < 0.05) mobility shoes compared with barefoot.

  • §

    Significant (P < 0.05) stability shoes compared with barefoot and with mobility shoes.

  • Significant (P < 0.05) stability shoes compared with mobility shoes.

Speed, meters/second1.11 ± 0.141.12 ± 0.161.08 ± 0.16
Stride, meters/H0.77 ± 0.060.75 ± 0.060.71 ± 0.07
Cadence, steps/minute104 ± 9107 ± 9110 ± 11
Knee ROM, degrees63 ± 663 ± 657 ± 6
Hip ROM, degrees28 ± 627 ± 526 ± 6
Ankle ROM, degrees30 ± 430 ± 626 ± 5
Peak external knee adduction moment, %BW × H3.07 ± 0.75§2.66 ± 0.692.71 ± 0.67
Adduction angular impulse, %BW × H1.07 ± 0.42§0.96 ± 0.420.94 ± 0.40
GRF, %BW106 ± 9104 ± 9103 ± 11

The GRF was found to be lower both while wearing the mobility shoes (2% lower, P < 0.05) and during barefoot walking (3% lower, P = 0.052) compared with wearing the stability shoes.

Similar differences as those seen in group A were noted between other gait parameters and the different footwear conditions. Again, stride increased and cadence decreased with both sets of shoes compared with barefoot walking. Knee and ankle ROM were also significantly greater with both sets of shoes compared with barefoot walking.

Linear regression again revealed anticipated associations between cadence and adduction angular impulse (stability shoes r = −0.467; mobility shoes r = −0.550; barefoot r = −0.439; P < 0.05 for all correlations). The only other significant association was noted between adduction angular impulse and ROM at the ankle during barefoot walking (r = 0.426, P = 0.031). This association was not present during the other footwear conditions or with ROM at the knee, as had been observed in group A.

The subjects in this group also underwent gait analyses while wearing a modified version of the mobility shoe without the grooves through the sole. This was done in order to determine whether it was the flat nature of the shoe (lack of heel) that was solely responsible for the load reduction or whether the grooves/flexibility offered further load reduction. The mobility shoe was associated with a 5% reduction in the peak external knee adduction moment compared with similar shoes without the sole grooves (2.66 %BW × H versus 2.80 %BW × H; P = 0.011)


This study demonstrates that footwear designed to simulate the essential features of barefoot walking can effectively reduce both the peak knee loads (peak external knee adduction moment) and overall knee loading (adduction angular impulse) during normal ambulation among patients with knee OA, compared both with subjects' self-chosen, comfortable walking shoes and with commonly prescribed stability walking shoes. Because dynamic loading across the knee is pathophysiologically important to the onset and progression of knee OA, these findings suggest that altering the characteristics of footwear might have a substantial impact on the overall disease process of knee OA.

It should be noted that the difference in adduction angular impulse in group A (conventional shoes versus the mobility shoe) approached, but did not reach, statistical significance. However, this is probably a result of less power to detect this difference in this group (see Subjects and Methods). Although group A had a larger sample size, the effect sizes in this group for peak external knee adduction moment and adduction angular impulse were lower, presumably because of the greater variation in data with subjects wearing various conventional footwear versus the standardized control footwear worn in group B.

The mobility shoe in this study was designed to incorporate the potential biomechanic advantages of barefoot walking. For example, conventional walking shoes typically contain partial lifts at the heel, which previous studies have shown actually increase knee loads in healthy women (4); in contrast, the absence of such a heel during barefoot walking may effectively reduce peak torques at the knee. Another factor may be the stiffness imposed by shoe soles. The GRF varied between the types of footwear and, as mentioned previously, the GRF is a determinant of peak moments at the lower extremity joints. The GRF may be affected by the impact and position with which the foot contacts the ground. Thus, the flexible movement of a bare foot may be biomechanically advantageous by allowing for better force application of the foot with the ground. Some advantage may also be related to enhanced sensory input from skin contact with the ground, relative to an insulated foot contacting the ground. The increased sensory input would initiate protective neuromuscular reflexes to help minimize proximal joint impact and load (31–33).

Shoe inserts or lateral wedge orthotics have been examined as potential devices to reduce medial compartment joint loads in knee OA with varying success. In comparison with an 8–13% reduction in peak joint loads with the mobility shoes, most studies of lateral wedge orthotics have demonstrated a 5–8% reduction in load with the inserts compared with the same shoes without inserts. Kuroyanagi et al found larger reduction of 8–13% with lateral wedge insoles, without and with subtalar strapping, respectively (5). However, their results were compared with barefoot walking, and the practical, long-term use of such interventions in traditional footwear is not clear.

The concept of engineering shoes specifically to promote joint load reduction in arthritis is relatively novel. Recently, presumably based on previous lateral wedge orthotic data, Fisher et al reported on a shoe that incorporates a lateral wedge into its design (6); in healthy subjects without OA, reductions of up to 16% were observed in peak knee loads with specific shoe modifications (lateral wedge and lateral stiffness variations) compared with the subjects' conventional walking shoes. To our knowledge, this is the only other report of a shoe designed specifically to lower knee loads, although the mechanisms underlying load reduction and the corresponding design of this shoe are quite different from the mobility shoe described in the present study, and the advantages and long-term effects of each will need to be evaluated in controlled clinical trials.

Interestingly, in addition to significant variations in knee loads among the various footwear conditions, there were also notable differences observed in other gait parameters (stride, cadence, and joint ROM) while wearing shoes compared with walking barefoot. This was also observed in our previous study, in which conventional shoes were compared with barefoot walking in a larger group of subjects with OA of the knee (2). When evaluating whether variations in loads could be attributable to these loads, significant but small associations were seen between knee ROM and both peak external knee adduction moment and adduction angular impulse. Interestingly, these associations were only observed in group A, and not in group B. This may be due to insufficient power to detect these relationships with the smaller sample size in group B. However, in our previous study with >60 subjects with OA, this association was not observed. Furthermore, although variations in ROM were observed, they do not appear to be out of the range of what would be considered normal ROM at each of the lower extremity joints (34).

The other significant association observed in both groups was that between cadence and adduction angular impulse. As mentioned previously, this relationship is expected considering that adduction angular impulse is the integral of the adduction moment during the stance phase of the gait cycle. Because increased cadence would imply a shorter duration of stance phase, the adduction angular impulse would be expected to decrease accordingly. Nevertheless, the peak external knee adduction moment is not related to cadence, and reflects that loads were indeed decreasing while wearing the mobility shoe and walking barefoot compared with walking with conventional shoes and stability shoes.

Several limitations of the current study deserve consideration. First, this study evaluates the short-term variations in joint loads with different types of footwear; it is not yet clear that the observed loading advantages will be maintained after prolonged use of the footwear. Second, although load reduction presumably has clinical benefits in OA, the magnitude of any long-term symptomatic palliation or delay in structural disease progression provided by these shoes will need to be determined in a prospective study, and the long-term safety and comfort of the shoes will need to be established. Finally, this was a relatively small study and may be insufficiently powered to detect subtle relationships and differences in secondary outcomes. Therefore, it should be emphasized that this study was not intended to be a clinical trial or to establish the safety and utility of footwear in clinical practice. The main purpose of the study was to suggest the importance of footwear in terms of lower extremity joint loading, and to examine properties of footwear that may be responsible for knee loading variations. In certain populations, flat, flexible footwear such as the mobility shoe may be an inappropriate choice. Future decisions about footwear should weigh individual patient characteristics, and more detailed information about the clinical effects of the mobility shoe should be examined in a prospective clinical trial.

In summary, we report innovative footwear, designed to simulate barefoot walking, that induces significant reductions in dynamic knee loads during ambulation compared with conventional walking shoes and control shoes. In light of the pathophysiological role of mechanical loading in the progression of knee OA, these findings suggest that footwear may represent a novel therapeutic target for the treatment of knee OA. Moreover, the types of shoes worn by subjects with knee OA should be evaluated more closely in terms of their contribution to the disease, and long-term intervention trials to evaluate the clinical effects of shoe design on pain and disease progression in OA should be considered.


Dr. Shakoor had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Shakoor, Lidtke, Sengupta.

Acquisition of data. Sengupta.

Analysis and interpretation of data. Shakoor, Block.

Manuscript preparation. Shakoor, Block.

Statistical analysis. Shakoor, Fogg.