Skip to main content
NCBI home page
VA Author Manuscripts logoLink to VA Author Manuscripts
. Author manuscript; available in PMC: 2025 Sep 15.
Published in final edited form as: J Orthop Res. 2020 Jun 22;39(3):619–627. doi: 10.1002/jor.24770

Changes in knee adduction moment wearing a variable-stiffness shoe correlate with changes in pain and mechanically stimulated cartilage oligomeric matrix levels

Jennifer C Erhart-Hledik 1,2, Gordhan B Mahtani 1,2, Jessica L Asay 2,3, Eleonora Migliore 1,2, Michelle M Nguyen 2,4, Thomas P Andriacchi 2,3, Constance R Chu 1,2
PMCID: PMC12433613  NIHMSID: NIHMS2099710  PMID: 32497304

Abstract

This study aimed to determine if changes in knee adduction moment (KAM) after 6 months of variable-stiffness shoe wear are associated with changes in symptoms or serum levels of cartilage oligomeric matrix protein (COMP) following a mechanical stimulus in subjects with medial knee osteoarthritis (OA). Twenty-five subjects were enrolled in the study and assigned a variable-stiffness shoe, and 19 subjects completed the 6-month follow-up. At baseline and follow-up subjects underwent gait analysis in control and variable-stiffness shoes, completed Western Ontario and McMaster Universities (WOMAC) questionnaires, and serum COMP concentrations were measured immediately before, 3.5 and 5.5 hours after a 30-minute walking activity. Relationships between changes in KAM (first peak and impulse) and changes in (a) COMP levels in response to the 30-minute walking activity and (b) WOMAC scores from baseline to 6-month follow-up were assessed by Pearson correlation coefficients. Changes in first peak KAM were associated with changes in COMP levels 5.5 hours postactivity from baseline to follow-up (R = .564, P = .045). Subjects with greater reductions in KAM had larger decreases in COMP (expressed as a percent of preactivity levels) at follow-up. Subjects with greater reductions in KAM impulse had significantly greater improvements in WOMAC Pain (R = −.56, P = .015) and Function (R = −.52, P = .028) scores at follow-up. The study results demonstrated the magnitude of reduction in the KAM wearing a variable-stiffness shoe is associated with decreases in mechanically stimulated COMP levels and pain/function. This work suggests that interactions between COMP and joint loading during walking should be further investigated in future studies of treatment outcomes in OA.

Keywords: cartilage oligomeric matrix protein, knee adduction moment, osteoarthritis, pain, variable-stiffness shoe

1 |. INTRODUCTION

Osteoarthritis (OA) is a debilitating disease and a leading source of pain and disability.1 The knee is one of the most commonly affected joints, with knee OA affecting an estimated 14 million individuals in the United States, of which half are under the age of 65.2 Despite the substantial prevalence of knee OA, disease-modifying treatments to prevent or delay progression of the disease are lacking. As such, the rate of total knee replacements has surged, with an estimated 673% increase in demand for primary total knee replacement by 2030.3

A critical barrier to the development of new treatments for knee OA is the detection of treatment effect. The structural standard for disease assessment is radiographs, which have relatively poor sensitivity for monitoring disease progression.4 Thus, assessment of treatment effects via radiographs result in trials requiring large sample sizes with long trial lengths that are costly and of uncertain benefit.5 Biomarkers may have the potential to provide a more rapid assessment of treatment response.6 In particular because OA is characterized by changes in cartilage properties and extracellular matrix components,7 biomarkers of cartilage metabolism may be useful in treatment assessment. There has been limited sensitivity of these markers in isolation in part due to large individual variations in the nominal resting levels of serum-derived biomarkers, as well as the inability of systemically measured markers to reflect local joint conditions.6 However, the mechanosensitivity of certain biomarkers8 may increase their responsiveness to treatment effects by introducing a mechanical stimulus. For example, serum levels of cartilage oligomeric matrix protein (COMP), a component of the articular cartilage extracellular matrix synthesized by chondrocytes,9 have been shown to change following 30 minutes of walking (a “stimulus”).10,11 Furthermore, changes in serum COMP concentrations following a walking stimulus were predictive of disease progression in an OA cohort, while resting COMP levels were not.12 Thus, applying a mechanical stimulus-response method13 may be advantageous in using biomarkers to monitor treatment responses in OA. However, there is a paucity of information on the associations between specific mechanical metrics and biomarker and pain responses.

Medial compartment knee OA offers an opportunity to explore the relationship between joint loading during gait and biomarker responses. Specifically, the medial compartment of the knee is involved in OA approximately 10 times more frequently than the lateral compartment,14 likely due to greater loading on the medial compartment articular cartilage, which is related to an increase in the external knee adduction moment (KAM).15,16 A high peak KAM during walking has been associated with the presence17 and rate of progression of medial compartment knee OA18,19 and increases with OA severity,20 while associations between KAM and symptoms appear to be dependent on radiographic disease severity.21 Thus, reducing the loading on the affected medial compartment of the knee by reducing the peak KAM offers an attractive target for interventions to slow the rate of progression and reduce painful symptoms of the disease. Variable-stiffness shoes have been shown to reduce the KAM and improve pain in patients with medial compartment knee OA.22,23 However, it remains unknown if sustained changes in joint loading affect changes in biomarkers implicated in the OA process.

Thus, the purpose of this study was to investigate if changes in joint loading (KAM) after 6 months of variable-stiffness shoe wear are associated with changes in symptoms or in serum levels of COMP following a mechanical stimulus in subjects with symptomatic medial knee OA. Specifically, it was hypothesized that changes in loading (KAM first peak and impulse) while wearing a variable-stiffness shoe over a 6-month period will be associated with (a) changes in COMP responses to a mechanical stimulus and (b) changes in pain and function.

2 |. METHODS

2.1 |. Study design and level of evidence

Prospective Cohort Study, level II.

2.2 |. Subjects

Twenty-five subjects with symptomatic medial compartment knee OA were enrolled in this 6-month study after giving written consent in accordance with the Institutional Review Board. The study was registered under ClinicalTrials.gov Identifier: NCT02593864. Potential subjects were recruited from sports medicine and orthopedic surgery clinics as well as self-referrals from the community after responding to advertisements. Inclusion criteria included (a) age between 18 and 80 years; (b) symptomatic medial compartment knee OA; (c) full weight-bearing status; (d) able to walk for 30 minutes or longer; (e) able to undergo magnetic resonance imaging (MRI) scan; and (f) agreement and ability to use provided shoe as primary walking shoe (4 or more hours a day) during the 6-month study period. A modified Kellgren-Lawrence24 (KL) grading system was used, such that individuals with a KL 0 were included if they were symptomatic and had evidence of medial knee OA consisting of cartilage loss and/or osteophytes by MRI evaluation. Baseline MRIs were acquired with sagittally oriented three-dimensional fast spin echo (CUBE) images and fat-saturated dual echo steady state images on a General Electric 3 Tesla scanner and evaluated by a musculoskeletal radiologist using a modified MRI Osteoarthritis Knee Score grading system (to assess the presence of medial cartilage loss and osteophytes). Exclusion criteria included (a) inflammatory arthritis, gout or recurrent pseudogout; (b) patellofemoral or lateral compartment disease that is equal to or more extensive than medial disease; (c) symptomatic OA of other lower extremity joints (ankle or hip); (d) BMI >35 kg/m2; (e) surgery of the knee performed less than 6 months prior; (f) use of shoe insert or hinged knee brace; and (g) pes planus (flat foot deformity where most of the sole of the foot touches the ground while standing), and/or unusual foot size or shape.

2.3 |. Intervention

All participants were assigned a variable-stiffness shoe (ABEO SMARTsystem; The Walking Company) to use as their main walking shoe for 6 months. Participants were asked to use the walking shoe for a minimum wear time of 4 hours per day. Shoes were worn bilaterally by all subjects. The variable-stiffness shoe sole is stiffer on the lateral side of the shoe as compared to the medial side. This design has previously been shown to reduce the KAM.22,23,25 For 1 week of each month, all subjects were asked to record and return, via a packet sent in the mail, the number of hours they wore their study shoes per day. The average number of hours of shoe wear per day over the 6-month study period was determined.

2.4 |. Patient-reported outcomes

At baseline and 6-month follow-up, all subjects completed a written Knee injury and Osteoarthritis Outcome Score (KOOS) questionnaire, for calculation of Western Ontario and McMaster Universities (WOMAC) pain, stiffness, and function scales. WOMAC scores were transformed to a 0 to 100 scale, with a score of 100 indicating no pain or dysfunction. All subjects completed the questionnaire about their more affected knee, determined by self-reported pain.

2.5 |. Gait analysis

Subjects performed three walking trials at a self-selected normal walking speed wearing two shoes, the variable-stiffness study shoe and a control walking shoe without the variable-stiffness sole (New Balance 658) at both the baseline and 6-month tests. Subjects began all trials at the same location and made the same number of footsteps until they hit the force plate to reduce variations in walking speed between trials. Gait analysis tests were performed after the serum collection protocol (stimulus-response protocol) or on a separate day from the serum collection protocol. Kinematic and kinetic data were collected using a previously published approach.26 Briefly, motion data were collected using a 10-camera marker-based motion capture system (Qualisys, Sweden) and ground reaction forces were measured with a floor-embedded force plate (Bertec, Columbus, OH). All data were collected at 120 Hz. The foot, shank, and thigh were idealized as rigid bodies, and inertial properties were taken from the literature.27 The software application BioMove (Stanford University) was used to calculate external moments for the subject’s more affected knee for each trial using an inverse dynamics approach from marker, force plate, and inertial segment data.28 Moments were expressed as external moments relative to the tibial anatomical frame based upon the position of anatomical landmarks identified by palpation.29 The KAM was analyzed using two variables that have previously been related to medial knee OA18,19,30: the first peak KAM (KAM1), defined as the maximum moment during the first 50% of stance, and the KAM impulse, defined as the time integral of the KAM curve during the entire stance phase. Average moment values for each shoe for the three trials were calculated, and moments were normalized to bodyweight and height (%Bw × Ht). Walking speed, for comparison between testing conditions as it may affect joint moments, was calculated as the average speed of the pelvis along the posterior-anterior axis of the walkway.

2.6 |. Serum collection protocol

At baseline and 6-month follow-up, serum was collected before and after a 30-minute walk. Subjects were asked to limit their physical activity for 24 hours prior to testing, and from the time they woke up until the beginning of testing, which began between 7:30 and 9:30 AM. Following a 30-minute period where subjects stayed seated while completing the informed consent, a 10 mL blood sample was collected. Subjects then walked at a self-selected pace for 30 minutes on a level treadmill. Following the walking activity, subjects remained in a seated position with minimal physical activity. Blood samples (10 mL) were collected 3.5 and 5.5 hours after the 30-minute walking exercise. The 3.5- and 5.5-hour time points were selected based on a previous literature12 that demonstrated COMP changes at these time points are associated with cartilage thinning in OA. Blood draws were performed using the BD Nexiva closed peripheral IV catheter system (Becton Dickinson, Franklin Lakes, NJ). To maintain patency, the catheter was flushed before and after each draw with 10 mL of 0.9% sodium chloride solution (Covidien Monoject prefill intravenous 12-mL flush syringe; Cardinal Health, Dublin, OH). Each blood draw was collected using the BD Vacutainer SST Venous Blood Collection Serum Tube (Becton Dickinson) and allowed to clot at room temperature. The blood was then centrifuged for 15 minutes at 2000 rpm at 4°C, and the supernatant serum was removed, aliquoted, and stored at −80°C until analysis.

2.7 |. Enzyme-linked immunosorbent assay

Serum concentrations of COMP were measured in duplicate using enzyme-linked immunosorbent assay (ELISA) kits from the same lot number according to manufacturer’s directions (R&D System Quantikine ELISA Human COMP). The mean intra-assay coefficient of variation was 5.6% and the mean minimum detectable dose was 0.010 ng/mL. Serum samples were diluted 1:100, and all samples for any single subject were tested on the same plate. Serum concentrations at the 3.5- and 5.5-hour time points following the 30-minute walk were expressed as a percentage of the preactivity value.

2.8 |. Statistical analysis

The Shapiro-Wilks test was used to test for the normality of the data. Paired two-tailed Student’s t tests or Wilcoxon signed-rank tests (for nonnormally distributed data) were used to evaluate changes in KAM1 and KAM impulse (baseline control shoe vs 6-month variable-stiffness shoe) and in WOMAC scores. Changes in COMP in response to the mechanical stimulus (30-minute walk) at 3.5 and 5.5 hours after the mechanical stimulus were analyzed using Friedman tests with Wilcoxon signed-rank tests utilizing a Bonferroni corrected P value (.016) for post hoc comparisons. Changes in COMP (resting values and 3.5- and 5.5-hour responses) from baseline to follow-up were analyzed with paired two-tailed Student’s t tests. Relationships between changes in KAM and changes in COMP levels in response to the mechanical stimulus and WOMAC scores from baseline to 6-month follow-up were assessed by the calculation of Pearson correlation coefficients, as change data were normally distributed. To investigate effects of potential covariates on outcome variables of changes in COMP levels and WOMAC scores from baseline to 6-month follow-up, exploratory univariate analyses were performed by the calculation of Pearson or Spearman (for nonnormally distributed data) correlation coefficients for covariates of KL grade, age, BMI, shoe wear, and change in walking speed during the serum collection protocol (30-minute walk) and by two-tailed Student’s t tests for covariates of sex and prior surgery. A P value of <.05 was considered significant for all analyses. All analyses were completed with SPSS version 25 (SPSS, Inc., Chicago, IL) using a per-protocol analysis.

3 |. RESULTS

3.1 |. Participant characteristics

Forty-three participants were consented into the study and assessed for eligibility and 25 participants (Table 1) completed baseline testing and were assigned the variable-stiffness shoe (Figure 1). Eighteen of the 25 enrolled participants presented with bilateral knee OA. Twelve individuals did not meet the inclusion/exclusion criteria, and six individuals decided not to participate. Nineteen subjects completed the 6-month study visit; three participants were lost to follow-up and three subjects withdrew from the study. The dropped subjects did not differ from the remaining subjects in any of the demographic or initial pain characteristics. Gait data for one subject at follow-up were unable to be used due to capture error, and thus knee loading change data are presented for 18 subjects. Serum COMP data at baseline and follow-up was available for 15 subjects for resting and 3.5-hour postactivity time points, and for 14 subjects for the 5.5-hour postactivity time point.

TABLE 1.

Demographic data of the study participants

Sex 17 Males/8 females
Age, y 58.6 ± 10.9
BMI, kg/m2 28.2 ± 3.9
Prior Surgery 13 (52%)
Kellgren-Lawrence grade 1.7 ± 1.1
 0a 2
 1 12
 2 4
 3 6
 4 1
Open in a new tab

Note: Data presented as mean ± standard deviation or count.

Abbreviation: MOAKS, MRI Osteoarthritis Knee Score.

a

Modified MOAKS grading (from MRI) of the two participants with KL grade of 0 demonstrated medial femoral and/or tibial cartilage loss as well as the presence of medial osteophytes.

FIGURE 1.

FIGURE 1

Open in a new tab

Subject flow diagram

3.1.1 |. Shoe wear

Average monthly shoe wear reports ranged from 8.0 to 9.0 hours per day, with an average over the 6 months of 8.1 ± 2.2 hours per day, in good compliance with the suggested minimum wear time of 4 hours per day.

3.2 |. Knee loading

KAM1 and KAM impulse were significantly reduced at 6 months when walking in the variable-stiffness shoe as compared to the baseline control (Table 2). The average reductions were −7.2% (P < .001) for KAM1 and −6.9% (P = .009) for KAM impulse over all subjects (Figure 2). KAM1 was reduced in all subjects at 6 months wearing the variable-stiffness shoe as compared to baseline wearing the control shoe, and KAM impulse was reduced in 13 of the 18 subjects (72.2%). No change in walking speed (P = .30; Table 2) was observed between baseline control to 6-month follow-up with the variable-stiffness shoe.

TABLE 2.

Gait, WOMAC, and COMP data at baseline and 6-month follow-up

Variable Baseline Follow-up P
KAM1 (%Bw × Ht) 2.67 ± 0.76 2.47 ± 0.75 <.001
KAM impulse (%Bw × Ht × s) 1.06 ± 0.36 0.99 ± 0.35 .009
Walking speed (m/s) 1.20 ± 0.14 1.22 ± 0.16 .30
WOMAC Pain 75.5 ± 15.1 75.3 ± 21.4 .95
WOMAC Stiffness 64.5 ± 17.8 65.1 ± 22.7 .95
WOMAC Function 79.6 ± 18.0 81.9 ± 18.1 .78
Resting COMP (ng/mL) 2.70 ± 1.98 2.61 ± 1.73 .41
3.5-hour COMP (% of resting) 84.4 ± 11.1 82.5 ± 9.1 .52
5.5-hour COMP (% of resting) 79.4 ± 13.3 84.5 ± 13.2 .10
Open in a new tab

Note: P values presented are from respective paired Student’s t tests or Wilcoxon signed-rank tests (for nonnormally distributed data). Data presented as mean ± standard deviation. KAM1, KAM impulse, and walking speed presented as control shoe at baseline and variable-stiffness shoe at follow-up.

Abbreviations: COMP, cartilage oligomeric matrix protein; KAM, knee adduction moment; KAM1, first peak KAM; WOMAC, Western Ontario and McMaster Universities.

FIGURE 2.

FIGURE 2

Open in a new tab

Ensemble KAM curves (mean ± 95% CIs) for all subjects at baseline (black: control shoe) and 6-month follow-up (red: variable-stiffness shoe). Significant improvements in first peak KAM and KAM impulse were seen on average overall subjects from baseline to 6-month follow-up. CI, confidence interval; KAM, knee adduction moment

3.3 |. Knee loading vs COMP

During the 30-minute walking exercise, subjects walked at a speed of 1.0 ± 0.4 m/s at baseline and 1.1 ± 0.3 m/s at 6-month follow-up. Friedman tests revealed that the 30-minute walk had a significant effect on COMP levels at both baseline (P = .001) and follow-up (P < .001). Specifically, at both baseline and 6-month follow-up (Figure 3), serum COMP levels decreased 3.5-hours (baseline: −15.6%, P = .001; 6 months: −17.5%, P = .001) and 5.5 hours (baseline: −20.6%, P = .001; 6 month: −15.5%, P = .003) after the 30-minute walking activity as compared to preactivity levels (Table 2). No changes were observed in resting COMP values from baseline to 6 months (P = .41; Table 2). Furthermore, there were no differences in mean postactivity changes in COMP values (expressed as a percent of preactivity levels) from baseline to 6 months (3.5 hours: P = .52; 5.5 hours: P = .10).

FIGURE 3.

FIGURE 3

Open in a new tab

Average serum COMP concentrations (mean ± 95% CIs) after a 30-minute walking activity in percent of baseline resting levels. *Significant difference vs preactivity level at baseline; +Significant difference vs preactivity level at 6 months. CI, confidence interval; COMP, cartilage oligomeric matrix protein

Changes in KAM1 from baseline to follow-up were significantly associated with changes in COMP levels 5.5 hours postactivity from baseline to follow-up (R = .564, P = .045; Figure 4). Subjects with a greater reduction in KAM1 over the 6-month follow-up period had larger decreases in COMP at the 5.5-hour postactivity time point at follow-up as compared to baseline. Changes in preactivity (resting) and 3.5-hour postactivity COMP levels were not associated with changes in KAM1 (resting: P = .72; 3.5 hours: P = .75) or KAM impulse (resting: P = .27; 3.5 hours: P = .35) over the 6-month follow-up period. Exploratory analyses of potential covariates of age, BMI, sex, KL grade, prior surgery, shoe wear, and change in walking speed during the serum collection protocol (30-minute walk) demonstrated that no potential covariate was significantly related to the outcome variable of change in COMP at either 3.5 or 5.5 hours (P > .05 for all).

FIGURE 4.

FIGURE 4

Open in a new tab

Association between change in KAM1 from baseline (control shoe) to 6-month follow-up (variable-stiffness shoe) and change in COMP concentration (5.5 hours postactivity expressed as percentage of resting level) from baseline to 6-month follow-up (P = .045). COMP, cartilage oligomeric matrix protein; KAM1, first peak knee adduction moment

3.4 |. Knee loading vs WOMAC

There was variability in change in WOMAC scores from baseline to follow-up, with approximately half of subjects improving over the 6-month follow-up period and half of subjects worsening in pain and symptoms over the follow-up period. As such, on average, over all 19 subjects there were no significant changes (Table 2) in WOMAC Pain (P = .95), Stiffness (P = .95), or Function (P = .78) scores from baseline to 6-month follow-up. However, the variability in changes in WOMAC scores were significantly associated with changes in KAM impulse over the follow-up period. Subjects with greater reductions in KAM impulse had significantly greater improvements in WOMAC Pain (R = −.56, P = .015) and WOMAC Function (R = −.52, P = .028) scores (Figure 5). No associations were observed between changes in KAM impulse and WOMAC Stiffness (P = .15) or between changes in KAM1 and WOMAC scores (Pain: P = .67; Stiffness: P = .44; Function: P = .16). Exploratory analyses of potential covariates of age, BMI, sex, KL grade, prior surgery, and shoe wear demonstrated one significant association between KL grade and change in WOMAC Pain score from baseline to 6-month follow-up (Spearman’s ρ = 0.56, P = .014). No other significant effects of potential covariates on changes in WOMAC scores were observed.

FIGURE 5.

FIGURE 5

Open in a new tab

Subjects with greater improvement (reduction) in KAM impulse with the variable-stiffness shoe from baseline (control) to 6 months had greater improvement (increase) in WOMAC Pain score (P = .015). A similar association was seen for WOMAC Function scores. KAM, knee adduction moment; WOMAC, Western Ontario and McMaster Universities

4 |. DISCUSSION

The results of this study identified an association between changes in a mechanical metric (KAM) and a biomarker (COMP) that may influence cartilage health and degeneration. This finding demonstrated that a mechanical stimulus (a change in joint loading) was significantly associated with a biological response. These results tie together several previous studies separately investigating load-modifying shoes and COMP in OA.12,22,23,25 For example, earlier work22,23,25 demonstrated significant reductions in the first peak KAM with the variable-stiffness shoe, and this work builds upon the prior research and demonstrates a reduction in KAM impulse with 6 months of variable-stiffness shoe wear.

The KAM has been shown to have clinical relevance in medial compartment knee OA, being associated with the presence,17 severity,31 and rate of progression of the disease.18,19,30 The importance of reducing the KAM was supported in this study by the association between KAM1 changes and COMP changes over a 6-month follow-up. Prior work12 has shown that patients with OA with higher relative COMP levels 5.5 hours after a 30-minute walking exercise have greater cartilage loss over 5 years. Thus, reduced KAM over the 6-month period, indicative of reduced medial compartment joint loading, appears to suggest a better COMP response at the 5.5-hour postactivity time point at follow-up. In this study, all subjects successfully reduced the first peak KAM over the 6-month study period, and subjects who decreased their KAM1 to a greater extent (Figure 4) from baseline to follow-up (negative value on x-axis) showed larger decreases in COMP levels at 5.5 hours postactivity at the 6-month follow-up (negative value on y-axis). The associations between pain reduction and reduced KAM impulse suggest that those subjects with improvements in loading also demonstrated better symptomatic responses, suggesting a link between changes in joint loading and changes in pain and function. Prior work23 has shown that a greater peak KAM prior to variable-stiffness shoe use was related to a greater absolute reduction in KAM while using the variable-stiffness shoe. Thus, future work could investigate selecting participants based on their personal KAM.

COMP is a logical consideration for a biomarker of OA as it is a noncollagenous glycoprotein, is expressed in articular cartilage,32 and serum levels of COMP have been shown to be significantly elevated in patients with knee OA.33,34 COMP is mechanosensitive11 and believed to play a role in supporting extracellular matrix interactions in cartilage needed for its load-bearing function.35 COMP is released in response to damage in the cartilage matrix, and it has been suggested that OA may result in more active cartilage turnover, resulting in higher serum COMP levels.36 Thus, COMP may be a promising marker to monitor load-modifying treatment responses in OA. The change in COMP response at follow-up is likely due to altered metabolic activity rather than diffusion due to walking because of the 5.5-hour time lag between the walking activity and the associated serum COMP changes.11 While COMP was chosen as the biomarker of interest for this study based on the literature showing its potential as a marker for OA,12,34,36 future work could investigate other biomarkers, such as markers of cartilage degradation and synthesis, in response to treatment effect.

Of note, changes to baseline resting values of COMP were not associated with loading changes over time, suggesting that the mechanical stimulus-response model,37 wherein the moderate 30-minute walk was used to elicit a response in COMP, may enhance the utility of COMP as a biological treatment indicator. Baseline resting values of COMP in this study were similar to those reported in the literature in knee OA utilizing the same ELISA Kit.38,39 Further this study indicated the KAM during gait is a potential important metric for stimulating the COMP response. Taken together with the finding that the first peak of the KAM18,19 and the KAM impulse30 have been related to OA disease progression, the association found in this study provides a possible mechanistic explanation for the benefit of load reduction.

It is important to note that while all subjects in this study experienced an improvement in first peak KAM over time, only the KAM impulse was significantly correlated to change in WOMAC score. The smoothing effect of the integration over stance phase relative to the variability of a single peak may provide a stronger correlation of the impulse to the pain scores vs the first peak KAM, suggesting future studies of load-modifying interventions should also investigate changes in KAM during the entire stance phase.

The substantial variation in the WOMAC score changes, with approximately half of subjects having improvements in symptoms over the follow-up period, differed from previous studies.22,25 However, the broad range in WOMAC score changes provided a good basis for identifying associations between WOMAC scores and changes in KAM. As described above, the key finding were the associations that demonstrated subjects with improvements in KAM impulse also demonstrated better symptomatic responses, suggesting a link between changes in joint loading and changes in pain and function with the variable-stiffness shoe intervention. Exploratory analyses of potential covariates demonstrated an association between disease severity (KL grade) and change in WOMAC Pain score over the 6-month follow-up period, and therefore future larger studies could also consider the effect of disease severity on study outcomes.

While the results of this study showing associations between changes in joint loading, pain, and biomarker response to a walking stimulus in OA are promising, this study has limitations that should be discussed. The sample size of this initial study was small and future work should confirm the current study findings in a larger population. As COMP levels are known to vary by sex40 and disease severity,34 future work should also investigate the effects of sex and disease severity on associations between changes in joint loading and changes in COMP in a larger cohort. While COMP is found predominantly in the extracellular matrix of cartilage,32 it is also synthesized in tendons, ligaments, and meniscus.32,41 However, subjects were excluded if they had symptomatic OA of other lower extremity joints (ankle or hip), and the concentration of COMP in nonarticular tissues is low when compared to cartilage and meniscus.32 The variable-stiffness shoe has a known effect on the KAM22,23,25; however, prior work has also demonstrated changes in joint moments at the hip and ankle with the variable-stiffness shoe.23 Thus, as COMP is a systemic marker, it is possible that COMP changes may also reflect alterations in hip and/or ankle loading, or changes in contralateral knee loading, as variable-stiffness shoes were worn bilaterally. Furthermore, as the number of steps taken with the shoes each day during the 6-month study period was not collected, it remains unknown if loading frequency with the variable-stiffness shoes affects study outcomes. The KAM was chosen as the gait variable of interest as it has been related to OA disease progression18,19,30 and earlier work22,23,25 demonstrated significant reductions in the KAM with the variable-stiffness shoe. Future work could investigate other biomechanical variables, such as the knee flexion moment or knee flexion angle, which have been related to OA disease progression,19,42 as well as other load-modifying interventions, such as gait retraining. A control group not treated with the variable-stiffness shoe was not included in this study, as the aim of this work was to investigate associations between within-person changes in joint loading with changes in symptoms and COMP levels in response to a mechanical stimulus. Thus, it cannot be determined if the presented data at 6 months for joint loading, symptoms, and COMP differs from a control OA cohort not treated with the intervention shoe. However, prior work22,25 using a randomized study design investigated the variable-stiffness shoe in comparison to a control walking shoe and showed significant effects on joint loading and symptoms. Change in medication usage was not assessed in the present study; however, prior work with the variable-stiffness shoe showed no significant changes in pain medication usage over a 6-month period, or differences in pain medication usage between subjects assigned to a variable-stiffness shoe or a control shoe.25

In conclusion, this study demonstrated that the magnitude of reduction in the first peak KAM wearing a variable-stiffness shoe is associated with decreases in mechanically stimulated COMP levels. This work suggests that interactions between COMP and KAM during walking may provide a potential clinical tool for the detection of treatment outcomes in OA. Future work should investigate the utility of mechanically stimulated COMP levels in association with changes in joint loading, incorporating imaging metrics for assessment of disease progression.

ACKNOWLEDGMENTS

The authors would like to thank Christy Dairaghi for assistance in data collection. This study was supported by VA GrantRX-002045–01 and Department of Defense Grant W81XWH1810590 (to Constance R. Chu). Variable-stiffness shoes were provided by The Walking Company.

REFERENCES

  • 1.Murray CJ, Atkinson C, Bhalla K, et al. The state of US health, 1990–2010: burden of diseases, injuries, and risk factors. JAMA. 2013;310:591–608. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Deshpande BR, Katz JN, Solomon DH, et al. Number of persons with symptomatic knee osteoarthritis in the US: impact of race and ethnicity, age, sex, and obesity. Arthritis Care Res. 2016;68:1743–1750. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am. 2007;89:780–785. [DOI] [PubMed] [Google Scholar]
  • 4.Wick MC, Kastlunger M, Weiss RJ. Clinical imaging assessments of knee osteoarthritis in the elderly: a mini-review. Gerontology. 2014;60:386–394. [DOI] [PubMed] [Google Scholar]
  • 5.Karsdal MA, Henriksen K, Leeming DJ, et al. Biochemical markers and the FDA critical path: how biomarkers may contribute to the understanding of pathophysiology and provide unique and necessary tools for drug development. Biomarkers. 2009;14:181–202. [DOI] [PubMed] [Google Scholar]
  • 6.Kraus VB, Burnett B, Coindreau J, et al. Application of biomarkers in the development of drugs intended for the treatment of osteoarthritis. Osteoarthritis Cartilage. 2011;19:515–542. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Martel-Pelletier J, Boileau C, Pelletier JP, Roughley PJ. Cartilage in normal and osteoarthritis conditions. Best Pract Res Clin Rheumatol. 2008;22:351–384. [DOI] [PubMed] [Google Scholar]
  • 8.Piscoya JL, Fermor B, Kraus VB, Stabler TV, Guilak F. The influence of mechanical compression on the induction of osteoarthritis-related biomarkers in articular cartilage explants. Osteoarthritis Cartilage. 2005;13:1092–1099. [DOI] [PubMed] [Google Scholar]
  • 9.Hedbom E, Antonsson P, Hjerpe A, et al. Cartilage matrix proteins. An acidic oligomeric protein (COMP) detected only in cartilage. J Biol Chem. 1992;267:6132–6136. [PubMed] [Google Scholar]
  • 10.Mündermann A, King KB, Smith RL, Andriacchi TP. Change in serum COMP concentration due to ambulatory load is not related to knee OA status. J Orthop Res. 2009;27:1408–1413. [DOI] [PubMed] [Google Scholar]
  • 11.Mündermann A, Dyrby CO, Andriacchi TP, King KB. Serum concentration of cartilage oligomeric matrix protein (COMP) is sensitive to physiological cyclic loading in healthy adults. Osteoarthritis Cartilage. 2005;13:34–38. [DOI] [PubMed] [Google Scholar]
  • 12.Erhart-Hledik JC, Favre J, Asay JL, et al. A relationship between mechanically-induced changes in serum cartilage oligomeric matrix protein (COMP) and changes in cartilage thickness after 5 years. Osteoarthritis Cartilage. 2012;20:1309–1315. [DOI] [PubMed] [Google Scholar]
  • 13.Andriacchi TP. Osteoarthritis: probing knee OA as a system responding to a stimulus. Nat Rev Rheumatol. 2012;8:371–372. [DOI] [PubMed] [Google Scholar]
  • 14.Ahlback S Osteoarthrosis of the knee. A radiographic investigation. Acta Radiol Diagn. 1968;277:7–72. [PubMed] [Google Scholar]
  • 15.Zhao D, Banks SA, D’Lima DD, Colwell CW Jr, Fregly BJ. In vivo medial and lateral tibial loads during dynamic and high flexion activities. J Orthop Res. 2007;25:593–602. [DOI] [PubMed] [Google Scholar]
  • 16.Noyes FR, Schipplein OD, Andriacchi TP, Saddemi SR, Weise M. The anterior cruciate ligament-deficient knee with varus alignment: an analysis of gait adaptations and dynamic joint loadings. Am J Sports Med. 1992;20:707–716. [DOI] [PubMed] [Google Scholar]
  • 17.Baliunas AJ, Hurwitz DE, Ryals AB, et al. Increased knee joint loads during walking are present in subjects with knee osteoarthritis. Osteoarthritis Cartilage. 2002;10:573–579. [DOI] [PubMed] [Google Scholar]
  • 18.Miyazaki T Dynamic load at baseline can predict radiographic disease progression in medial compartment knee osteoarthritis. Ann Rheum Dis. 2002;61:617–622. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Chehab EF, Favre J, Erhart-Hledik JC, Andriacchi TP. Baseline knee adduction and flexion moments during walking are both associated with 5 year cartilage changes in patients with medial knee osteoarthritis. Osteoarthritis Cartilage. 2014;22:1833–1839. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Foroughi N, Smith R, Vanwanseele B. The association of external knee adduction moment with biomechanical variables in osteoarthritis: a systematic review. Knee. 2009;16:303–309. [DOI] [PubMed] [Google Scholar]
  • 21.Hall M, Bennell KL, Wrigley TV, et al. The knee adduction moment and knee osteoarthritis symptoms: relationships according to radiographic disease severity. Osteoarthritis Cartilage. 2017;25:34–41. [DOI] [PubMed] [Google Scholar]
  • 22.Erhart-Hledik JC, Elspas B, Giori NJ, Andriacchi TP. Effect of variable-stiffness walking shoes on knee adduction moment, pain, and function in subjects with medial compartment knee osteoarthritis after 1 year. J Orthop Res. 2012;30:514–521. [DOI] [PubMed] [Google Scholar]
  • 23.Erhart JC, Mündermann A, Elspas B, Giori NJ, Andriacchi TP. A variable-stiffness shoe lowers the knee adduction moment in subjects with symptoms of medial compartment knee osteoarthritis. J Biomech. 2008;41:2720–2725. [DOI] [PubMed] [Google Scholar]
  • 24.Kellgren JH, Lawrence JS. Radiological assessment of osteo-arthrosis. Ann Rheum Dis. 1957;16:494–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Erhart JC, Mündermann A, Elspas B, Giori NJ, Andriacchi TP. Changes in knee adduction moment, pain, and functionality with a variable-stiffness walking shoe after 6 months. J Orthop Res. 2010;28:873–879. [DOI] [PubMed] [Google Scholar]
  • 26.Chehab EF, Andriacchi TP, Favre J. Speed, age, sex, and body mass index provide a rigorous basis for comparing the kinematic and kinetic profiles of the lower extremity during walking. J Biomech. 2017;58:11–20. [DOI] [PubMed] [Google Scholar]
  • 27.Dempster WT, Gaughran GRL. Properties of body segments based on size and weight. Am J Anat. 1967;120:33–54. [Google Scholar]
  • 28.Andriacchi TP, J T, Hurwitz DE, Nataraja RN. Musculoskeletal dynamics, locomotion, and clinical applications. In: Mow VC, H R, eds. Basic Orthopaedic Biomechanics and Mechano-Biology. 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2005:91–122. [Google Scholar]
  • 29.Dyrby CO, Andriacchi TP. Secondary motions of the knee during weight bearing and non-weight bearing activities. J Orthop Res. 2004;22:794–800. [DOI] [PubMed] [Google Scholar]
  • 30.Chang AH, Moisio KC, Chmiel JS, et al. External knee adduction and flexion moments during gait and medial tibiofemoral disease progression in knee osteoarthritis. Osteoarthritis Cartilage. 2015;23:1099–1106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Mündermann A, Dyrby CO, Hurwitz DE, Sharma L, Andriacchi TP. Potential strategies to reduce medial compartment loading in patients with knee osteoarthritis of varying severity: reduced walking speed. Arthritis Rheum. 2004;50:1172–1178. [DOI] [PubMed] [Google Scholar]
  • 32.Neidhart M, Hauser N, Paulsson M, DiCesare PE, Michel BA, Häuselmann HJ. Small fragments of cartilage oligomeric matrix protein in synovial fluid and serum as markers for cartilage degradation. Br J Rheumatol. 1997;36:1151–1160. [DOI] [PubMed] [Google Scholar]
  • 33.Bi X Correlation of serum cartilage oligomeric matrix protein with knee osteoarthritis diagnosis: a meta-analysis. J Orthop Surg Res. 2018;13:262. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Clark AG, Jordan JM, Vilim V, et al. Serum cartilage oligomeric matrix protein reflects osteoarthritis presence and severity: the Johnston County Osteoarthritis Project. Arthritis Rheum. 1999;42:2356–2364. [DOI] [PubMed] [Google Scholar]
  • 35.Chen FH, Herndon ME, Patel N, Hecht JT, Tuan RS, Lawler J. Interaction of cartilage oligomeric matrix protein/thrombospondin 5 with aggrecan. J Biol Chem. 2007;282:24591–24598. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Saxne T, Heinegard D. Cartilage oligomeric matrix protein: a novel marker of cartilage turnover detectable in synovial fluid and blood. Br J Rheumatol. 1992;31:583–591. [DOI] [PubMed] [Google Scholar]
  • 37.Andriacchi TP, Favre J, Erhart-Hledik JC, Chu CR. A systems view of risk factors for knee osteoarthritis reveals insights into the pathogenesis of the disease. Ann Biomed Eng. 2015;43:376–387. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Tomonaga A, Takahashi T, Tanaka YT, Tsuboi M, Ito K, Nagaoka I. Evaluation of the effect of salmon nasal proteoglycan on biomarkers for cartilage metabolism in individuals with knee joint discomfort: a randomized double-blind placebo-controlled clinical study. Exp Ther Med. 2017;14:115–126. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Das Gupta E, Ng WR, Wong SF, Bhurhanudeen AK, Yeap SS. Correction: correlation of serum cartilage oligometric matrix protein (COMP) and interleukin-16 (IL-16) levels with disease severity in primary knee osteoarthritis: a pilot study in a Malaysian population. PLoS One. 2017;12:e0190542. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Jordan JM, Luta G, Stabler T, et al. Ethnic and sex differences in serum levels of cartilage oligomeric matrix protein: the Johnston County Osteoarthritis Project. Arthritis Rheum. 2003;48:675–681. [DOI] [PubMed] [Google Scholar]
  • 41.Muller G, Michel A, Altenburg E. COMP (cartilage oligomeric matrix protein) is synthesized in ligament, tendon, meniscus, and articular cartilage. Connect Tissue Res. 1998;39:233–244. [DOI] [PubMed] [Google Scholar]
  • 42.Favre J, Erhart-Hledik JC, Chehab EF, Andriacchi TP. Baseline ambulatory knee kinematics are associated with changes in cartilage thickness in osteoarthritic patients over 5 years. J Biomech. 2016;49:1859–1864. [DOI] [PubMed] [Google Scholar]

RESOURCES