Background/Purpose: Biomechanical treatments for medial
knee osteoarthritis (OA) often target the knee adduction moment (KAM).  Foot-mediated
biomechanical interventions for knee OA may operate through altering foot kinematics
and foot center of pressure (COP).  Flexible footwear has recently been shown
to effectively reduce the KAM, but specific mechanisms underlying this
reduction are not clear.  We hypothesize that the use of flexible footwear allows
increased foot pronation and medial shift in foot COP that are associated with
reduction in the KAM.

Methods: Participants with symptomatic medial
compartment knee OA (KL 2 and 3) were evaluated at baseline using simultaneous
barefoot COP analyses and 3-D gait analyses including detailed foot kinematic
modeling.  All participants were provided flexible shoes (Dr. Comfort Flex-OA,
Mequon, WI) and instructed to wear them at least 6 hours/day, 6 days/week. 
Gait and COP analyses were repeated after 12 weeks of wearing the shoes.  Gait
testing consisted of five trials each while barefoot, in subjects “own shoe”,
and in the flexible shoe.  During barefoot trials, plantar pressure
distribution was acquired simultaneously by mounting a pressure platform onto a
force plate and leveling the stacked assembly with the walkway.  COP was
quantified by determining the Medial to Lateral Pressure Index (MLPI) and
normalizing to foot width, with smaller values representing more medialized COP. 
The primary outcomes were the peak KAM and COP during the first half of stance
and foot kinematic variables.   Paired samples t-tests were used to evaluate
changes in kinematics and foot COP.  Pearson correlations were used to evaluate
the association between foot kinematics, shift in COP and percent reduction in KAM.

Results:   28 participants were evaluated (mean
age 59±7 years, 21 women).  At 12 weeks, the KAM in the flexible shoe was
reduced compared with their own shoes at baseline (2.17± 0.82 vs 2.32±0.81%BW*Ht,
p=0.030).  The foot COP medialized compared with baseline (4.61±4.28 vs 5.88±4.39%FW,
p=0.034).  No significant correlation was found between medialization of foot
COP and reduction in KAM (p=0.430).  No significant changes occurred in
measures of foot kinematics including rearfoot eversion (p=0.975), foot
supination (p=0.372) and medial arch (p=0.841).  However, lower maximal
rearfoot eversion at 12 weeks correlated with both the medialization of foot
COP (r=0.438, p=0.02) as well as the percent reduction in KAM with the mobility
shoe at 12 weeks compared to own shoes at baseline (r=0.438, p=0.02).

Conclusion: Walking with flexible shoes for 12
weeks significantly reduces loading at the medial tibiofemoral joint in
participants with knee OA and shifts the foot COP medially during walking. Interestingly,
though, there was no association between the medialization in foot COP and
reduction in knee loading, suggesting that both phenomena could be explained by
an unidentified mechanism.  Also, there were no significant differences found
in foot kinematics over 12 weeks, yet less rearfoot eversion during walking at
12 weeks was associated with the extent of COP medialization and reduction in
medial knee loading.