Background
Knee osteoarthritis (OA) is a leading cause of pain and disability among older adults, with more than 200 million people affected worldwide [
1]. As no cure exists for OA, current treatment is restricted to managing symptoms and improving function in an effort to maximise quality of life. In particular, self-management is fundamental, and clinical guidelines advocate appropriate footwear as important [
2,
3].
Knee joint loading during walking is higher in the medial compared to the lateral compartment, which likely explains why knee OA occurs most commonly in the medial tibiofemoral joint. Increased medial knee joint loading, most commonly inferred using the external knee adduction moment (KAM), has been associated with greater pain and disability in people with knee OA [
4], and worsening structural changes over time [
5]. Accordingly, biomechanical strategies designed to reduce knee loading, such as knee braces, lateral wedges and unloading footwear, have received support in international guidelines for the management of people with knee OA [
2,
3].
Unloading footwear with midsoles that are stiffer laterally than medially have been shown to reduce the KAM in those with knee OA, primarily by reducing the knee ground reaction force (GRF) lever arm (mainly via a lateral shift in the centre of pressure (CoP) under the foot) and the magnitude of the frontal plane GRF [
6]. These findings suggest that reductions in the KAM with unloading shoes are associated with changes evident at the foot. However, research has only measured loads at the shoe-ground interface using force plates; no study to date has evaluated the effect of unloading shoes on in-shoe regional plantar pressures. Research on laterally wedged insoles (that are inserted into the patient’s own footwear) suggests they increase lateral pressures [
7‐
10] and it is reasonable to hypothesise that unloading shoes may have a similar effect.
The actual effect of unloading shoes on changes in the KAM is quite variable across individuals [
11,
12]. Identification of subgroups of “biomechanical responders” is an important OA research priority [
3]. As such, measurement of regional plantar forces permits an increased understanding of how unloading footwear changes regional plantar loading, and may yield important insights into individual responses to this treatment. People with knee OA have a more pronated static foot posture [
13] compared to those without knee OA, and concurrent foot pain is common in people with knee OA [
14]. A pronated foot [
15] and greater rear foot mobility [
16] have been shown to be associated with greater reductions in the KAM with lateral wedge insoles. There is also some evidence that foot posture is associated with plantar pressures during walking [
17]. Thus, it is possible that people with a more pronated foot posture, or greater foot mobility, experience different changes in plantar forces with unloading shoes. This information may help to expand our understanding of how unloading shoes influence foot and ankle biomechanics, and may also help explain why some people experience foot/ankle pain when wearing these types of shoes.
The aims of this exploratory study were to evaluate i) the effects of unloading shoes on in-shoe regional plantar forces, relative to conventional shoes without unloading features and; ii) whether measures of foot posture and/or foot mobility moderate these effects, in people with symptomatic medial knee OA.
Discussion
This study evaluated the effects of unloading shoes on regional plantar forces in people with symptomatic medial knee OA, and whether alterations in these were moderated by measures of foot posture and/or mobility. The results of this exploratory study showed that unloading shoes increased lateral heel and forefoot force, with concurrent reductions in medial sub regional forces. In addition, unloading shoes shifted the CoP anteriorly and laterally, but did not significantly change the arch index. Individual increases in lateral heel force with unloading shoes were variable in magnitude, suggesting individual participant characteristics may influence biomechanical responses to unloading footwear. However, our results showed that foot posture, FMM and navicular drop did not significantly moderate the effect of unloading shoes on any outcome.
Although no previous study has examined changes in regional plantar forces when wearing unloading shoes in people with knee OA, other research has evaluated the effects of similar foot-based biomechanical interventions. Our findings are generally consistent with the effects of lateral wedge insoles that are inserted into the patient’s own footwear. Most research on inserted lateral wedges suggests they increase lateral plantar pressures. For example, Leitch et al. [
7] assessed the effect of 4 and 8 degree lateral heel wedges in participants with and without knee OA, and found that lateral heel wedges increased lateral heel pressure, as well as lateral and anterior CoP displacement. An increase in lateral pressure and/or a lateral shift of the CoP with lateral wedge insoles has also been observed in other studies involving people with OA and healthy individuals [
8‐
10]. However, Erhart et al. [
12] found increased medial-to-lateral pressure ratios in healthy young adults walking in shoes with 4 and 8 degree laterally wedged shoe midsoles (i.e. not inserted wedges), in comparison to unwedged shoes.
Additionally, there is some research evaluating the effects of unloading shoes on CoP measured using force platforms. Kean et al. [
6] observed a lateral shift in CoP in people with knee OA and in overweight asymptomatic individuals, which is consistent with our current findings using a foot-based measurement system to evaluate in-shoe plantar forces. In contrast, Jenkyn et al. [
29] observed a medial shift of the CoP, measured using a force platform, in people with medial knee OA when walking in unloading shoes with a dual, more laterally dense midsole. Reasons for this difference in findings are uncertain. In summary, most findings suggest that current foot-based biomechanical ‘unloading’ interventions for knee OA redistribute plantar forces from medial to lateral, which likely contributes to a laterally shifted CoP; however even when other patterns are observed, group reductions in knee loading are generally found with all these interventions [
6,
11,
29‐
33].
Despite some individual variability in how unloading shoes influenced regional plantar forces, we did not find that measures of foot posture or mobility moderated the effects of unloading shoes on regional plantar forces in our study. There may be several reasons for this. To increase the external validity of our findings and ensure relevance to clinical practice, we used simple static foot measures. However static foot posture is not a good indicator of dynamic foot motion [
34], and although foot mobility and navicular drop are more dynamic measures, other variables such as peak or total rearfoot eversion, or rearfoot moments [
15], during walking may have yielded different results. Our study involved a relatively small sample size, which yielded a limited spread of foot postures, and may mean we were underpowered to detect an interaction between footwear condition and foot posture, foot mobility or navicular drop on plantar forces. We found borderline significant interaction effects for navicular drop on the effect of unloading shoes for peak medial forefoot force and the medial-lateral heel peak force ratio. We interpret this as being a chance finding, because of the number of statistical tests used, and the lack of a consistent pattern across all outcome variables. Alternatively, it is possible that those measures genuinely do not moderate the effects of unloading shoes on regional plantar forces, and other individual characteristics may explain the variability in biomechanical effects of unloading shoes.
Our findings may have clinical implications. Adverse effects at the foot/ankle with unloading shoes have been reported by some people with knee OA. In our recent clinical trial, 20% of participants reported foot/ankle pain, with 4% discontinuing treatment because of these adverse effects [
18]. Importantly, the trial excluded people who reported ankle/foot pain in the previous 6 months, thus it is possible that this study may have underestimated the potential for foot-related adverse effects if people with existing foot and/or ankle problems were to wear such shoes. We observed small increases in lateral forefoot and lateral heel forces with unloading shoes, which may warrant caution with their use in people with concurrent foot/ankle pathology and/or symptoms. As concurrent foot pain is common in people with knee OA [
14], unloading shoes could exacerbate foot problems in this subset of patients.
There are some limitations to this study. First, we only assessed the immediate effect of unloading shoes on plantar forces. Second, although we averaged data over multiple footstrikes, only data from a single walking trial was analysed. Third, our relatively small sample size means that our findings are only generalizable to people with knee OA who have characteristics similar to our sample. Further, although foot posture and mobility did not moderate effects of the shoes, other participant characteristics (such as KL-grade and knee alignment) may play a role. Last, this study only examined the effect of unloading shoes on regional plantar forces, in isolation from other lower limb biomechanical parameters. Future studies should combine regional plantar pressure measurements with gait analysis to determine whether changes in plantar forces are associated with altered knee biomechanics relevant to knee OA, such as the KAM.
Conclusions
In conclusion, this exploratory study evaluated the immediate effects of unloading shoes on plantar force measurements, and whether measures of foot posture and/or mobility moderated these effects, in patients with medial knee OA. Findings demonstrated greater lateral forces with unloading shoes in rear- and forefoot sub regions, with a concurrent decrease in medial forces. Although effects of unloading shoes on plantar forces were not moderated by foot posture, FMM or navicular drop, variability in the individual increases in lateral heel force suggest other individual participant characteristics apart from foot posture may play a role.