Background
Small joint inflammation in the hands and feet are the hallmark of early rheumatoid arthritis (RA). Clinical studies suggest foot pain may be problematic in about one-third of patients in early disease with more frequent involvement of the metatarsophalangeal (MTP) joints (34%) in comparison with the midtarsal (4%) and ankle (20%) regions [
1]. Radiographically, the joints of the feet show damage more often, and earlier than the joints of the hand [
2]. Furthermore, in patients with disease duration of <12 months, edema, synovitis and erosion can be detected by magnetic resonance imaging (MRI) at the MTP joints in patients whose hands are normal [
3]. Ultrasonography has also revealed MTP joint effusion, flexor tenosynovitis and plantar bursa in early disease [
4]. By contrast, tarsal/ankle joint pathology occurs later in the disease with joint space narrowing and erosion in the ankle found in less than 2% of patients within the first 3 years [
5,
6].
Whilst imaging enables detection of joint and soft-tissue pathology, clinical signs of foot disease are often more subtle in early RA and data are lacking. In the forefoot, abnormal MTP joint alignment, subluxation and stiffness were found in 25% of patients within 3 years of disease onset [
7]. In the hindfoot, <10% of cases had moderate to severe deformity by 5 years, although gait analysis can enable detection of collapsing pes valgus within 3 years [
7,
8]. Normal foot joint motion is necessary to allow the body to progress over the supporting foot during stance. This complex structure must facilitate weight acceptance and transfer, contribute shock absorption and stability and distribute pressure evenly on the plantar surface. The consequences of persistent synovitis in terms of disrupting these functions are well documented for established RA [
8‐
13] but the association between joint damage in the feet, impairment and function has not been evaluated in early disease. Nevertheless, strong recommendations have been made for careful attention to foot problems in early disease with an emphasis towards correcting underlying biomechanical faults, chiefly through the use of orthotic devices [
14,
15].
We require a better understanding of the impact of RA in the foot and we have recently developed metrics that accurately and reproducibly measure impairment and disability [
16]. Moreover, three-dimensional (3D) gait analysis permits high-definition measurement of foot function and we have successfully studied this in patients with well-established disease [
8,
12,
13]. Therefore, the aim of this study is to compare clinical disease activity, impairment, disability, and foot function in normal and early rheumatoid arthritis (RA) feet using standardised clinical measures and 3D gait analysis.
Discussion
There has been an increasing use of gait analysis as an objective measure of foot function in RA [
8‐
13]. This prompted us to investigate the relationship between impairment, clinical disease activity and function in a group of RA patients with early disease. Our study showed moderate-to-high levels of disease impact in the foot associated with varying levels of clinical disease activity, pain, deformity and altered function. The LFIS is a new validated measure of disease impact in the foot developed by our own group [
16]. Moderate to high levels of impairment and foot related disability were observed in these early patients. Indeed, the median scores were similar to those currently observed amongst our own clinic patients with established disease of >5 years. Therefore, LFIS like HAQ (a disability metric which predicts poor outcome when high at disease onset) may be an important predictive measure of future localised disease impact [
16,
18].
Despite DMARD treatment in 10/12 patients, all of the early patients were suffering foot pain which may have been related to inflammation or impairment or both. Inflammatory processes may be the predominant determinant of disability in early RA, with structural abnormalities leading to functional impairment in well established disease. In the foot, for example, synovitis is thought to reflect underlying systemic disease activity [
19] whilst pain and disability correlate strongly with patients self perception of foot pain but not disease duration, joint damage or systemic drug treatment [
20]. Furthermore, impairment of gait and mobility are associated with well recognised patterns of pain and foot deformity [
8‐
13]. These problems reported in cohorts of patients with well-established RA are not very different from those reported in this study. We were able to detect functional limitation in gait in patients with <2 years disease duration characterised by slow walking speed and increased double-support and this may suggest early adaptation to underlying systemic and local disease activity and impairments [
9‐
13].
Gait analysis has improved out understanding of foot function in established RA but has not been reported in early disease. Typical forefoot deformities such as hallux valgus, MTP joint subluxation and hammer toe deformities (component scores of the SI) were detected in 75% of the early RA patients and 50% of the control subjects. Although not staged by severity, deformity was worse in the RA cases resulting in a reduced weightbearing contact area for the toes and elevated peak pressures at the plantar MTP joints. Furthermore, plantar pressures are higher at MTP joint which are eroded and of three patients with erosive changes, two had peak values outside normal limits (>2 SD above normal mean limit) [
21]. If untreated, these stresses may be associated with persistent or worsening symptoms and the development of secondary pressure lesions such as callus, bursa and ulceration [
22,
23]. However, some patients can off-load these painful sites by avoiding weight transfer to the forefoot. O'Connell
et al (1998) reported disruption to the rocker function of the foot characterised by; delay of transfer of centre-of-pressure to the forefoot; delay of heel-rise and reduced peak vertical force; and net ankle plantar flexor muscle moment in terminal stance [
11]. These compensations were obvious and marked in some cases but not others and overall no differences were detected other than for a delayed heel-rise in the early cases. Further work is required to determine more fully which of these metrics are the most sensitive to detect early changes. A second off-loading strategy for painful forefoot is to move the foot into a varus posture, especially if the medial MTP joints are involved, and this was detected in 2 of our cases. Fixed varus deformities of the rearfoot are reportedly present in about 2% of RA patients [
24]. In our experience these are difficult to manage clinically. The long term functional consequences of this impairment are unclear.
Inflammatory synovitis and dysfunction of the peritalar joints and the tibialis posterior muscle-tendon unit are postulated mechanisms leading to instability of the subtalar and mid-tarsal joints [
25]. This is clinically recognisable as pes planovalgus which has a reported prevalence of between 46–64% in RA [
8,
12,
26,
27]. The natural history of this progressive deformity is poorly understood but data from the present study suggests it may occur early and progress rapidly. In this study, clinical examination was used to successfully identify synovitis in and around the peri-talar joints. The consequence for these patients may be the progressive development of pes planovalgus. Indeed, 9/12 patients on standing had an exaggerated valgus heel posture above the normal values. Furthermore, we have previously described collapse of the medial longitudinal arch accompanied by an increase in the maximum rearfoot eversion reached during stance when walking [
8,
12,
13]. These two motion deficits were clearly identified in this patient cohort, albeit two cases with moderately severe varus heel deformity appreciably skew the data. Finally, collapse of the medial longitudinal arch creates a larger midfoot weightbearing surface and in the early RA cases this was 21% greater than healthy controls.
In the planovalgus foot, the gastrocnemius-solues complex shows evidence of increased activity in an attempt to minimise the valgus deformity [
28]. The third rocker function of the foot occurs about the forefoot as the heel rises in late stance and this is controlled by concentric contraction of the ankle plantarflexors. However, under manual muscle strength tests, the gastocnemius-soleus muscle group is usually weak in RA patients. The associated functional consequence may be reduced ankle joint power during terminal stance and this was detected for the RA patients in this study [
11,
28]. Since the net muscular moment was normal, the reduced ankle power must be caused by a reduction in the angular velocity at the joint. Factors contributing to reduce ankle joint angular velocity such as walking speed and reduced range of motion were detected amongst the patients but other gait-limiting factors and adaptations to impairments such as pain and deformity could not be fully accounted for. Finally, the blunting of the first peak of the vertical ground reaction force, suggests that some patients may have been cautious in loading the foot during initial foot contact phase, perhaps in an effort to lessen painful symptoms in the rearfoot. These functional changes in the rearfoot in early disease have important implications since we were able to successfully control motion and lessen symptoms by treating patients with customised foot orthoses [
29,
30]. However, the patients in the study by Woodburn
et al. 2002 had median disease duration of 3 years suggesting we should attempt to identify rearfoot dysfunction even earlier and treat accordingly.
One of the limitations of this study is subject selection bias as these 12 cases were recruited following podiatry referral and may therefore represent only a subset of early RA patients with severe foot involvement. The introduction of Early Arthritis Clinics should permit future access to a more generalisable population. The sample size was also too small to permit formal hypothesis testing or statistical analysis of association between impairment and biomechanical data. Finally, due to the current limitations of gait analysis, our biomechanical foot models are insensitive to detection of functional changes at small but important single joints such as the subtalar, talonavicular, and lesser toe MTP joints which are involved in the disease process.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
DET conceived of the study, participated in the design of the study and carried out the gait analyses.
PSH participated in the design of the study and the coordination of the gait analyses.
PE participated in the design of the study and its coordination.
JW conceived of the study, participated in the design of the study, and conducted the gait data analysis and preparation.
All authors read and approved the final manuscript.