Transfer of plantar pressure from the medial to the central forefoot in patients with hallux valgus

While foot pain has always been a highly common symptom among the elderly [1, 2], especially in women [3], the incidence is increasing today, even among younger people [4]. A major cause for these symptoms is hallux valgus deformity, which often leads to severe functional constraints and to transfer metatarsalgia-related forefoot pain. Since the first metatarsophalangeal joint in the physiological situation acts as a pivot for the transfer of body weight during the late stance phase [5], it makes sense that progressive pathological changes, including subluxation in that joint, could interfere with efficient toe-off [6] and thus lead to altered plantar pressure distribution. The way in which these pressures change remains, to date, a matter of debate: Whereas several studies did not detect any significant changes in plantar pressure distribution (reviewed by [6]), two high-quality studies reported increased pressures under the hallux [7, 8] and under the first and second metatarsal heads [7, 9], and one study reported reduced first metatarsal loading [10]. Increased pressures under the small digits (3–5) have also been described [7].

One possible reason for this heterogeneous picture could be the different measurement systems applied: Many pedobarographic or gait analysis studies use single pedobarographic plates [11, 12] or in-shoe pressure measuring systems [13]. Although the plate system offers high spatial resolution, its main disadvantage is the relatively small measuring area, which is especially difficult for fast movement. Longer measurement areas that allow physiological gait can be created with compound measuring plates or carpet walkways; they are, however, costly and remain limited regarding the registered walking distance. In-shoe pressure measuring systems can be worn for a longer time and allow analysis of worn insoles, but their main disadvantage is the relatively low spatial resolution because of the small number of sensors [14]. We therefore used an instrumented treadmill to measure several gait cycles with the patients walking at constant speed. This allows a fluent gait without having to aim for a measuring plate. At the same time, this system offers a reasonable spatial resolution of plantar pressures. Such a set-up has already been used to examine healthy subjects [15] and patients with hallux valgus postoperatively for overall ground-reactive forces [16]. Yet, to our knowledge, no study has examined detailed plantar pressure distribution in patients with hallux valgus by using such an instrumented treadmill.

To better understand changes in plantar pressure distribution could be a valuable key to further improve conservative or surgical therapeutic strategies for patients with hallux valgus. Since the data in the literature with respect to these changes in pressure distribution are still quite heterogeneous, the aim of the present study was to evaluate the effect of hallux valgus deformity on plantar pressure distribution compared with that in healthy feet. We hypothesised that it is possible to differentiate between affected and healthy feet on the basis of these pedobarographic data. In particular, we expected reduced pressures under the medial ray and increased pressures under the second and third metatarsal heads.

The control group, who had no signs of lower extremity pathology, consisted of 19 women and 11 men. They had a median age of 25 years (range 21–61 years) and were significantly younger (p < .001) than the study group. Sex was not equally distributed between these two groups (p = .005). All control subjects were able to walk freely and barefoot on the treadmill and had no pain at rest or strain whatsoever. The mean AOFAS foot score in healthy subjects was 99 (SD 2). The range of motion in the first metatarsophalangeal joint was 67° (SD 17°) in the hallux valgus group and 97° (SD 21°) in the control group (p < .001) (see also Table 1).

The aim of the present study was to evaluate the effect of hallux valgus deformity on gait and plantar pressure distribution compared with that in healthy feet. For our analyses, we used a treadmill equipped with a sensor mat underneath the belt that allowed for continuous measurements of the walking subjects at a high spatial and temporal resolution.

As described in previous studies, no difference was observed in the spatio-temporal gait parameters such as step width, step length, and single-limb support [6, 9, 21, 22]. The only difference, as observed in one study, was reduced walking speed and overall shorter step length [21]. This did not, however, distinguish between the pathological and the healthy foot of each affected individual - a trait common to many of the previously published studies. When analysing the pathological and healthy side separately in our study, still no difference could be detected in the bilateral parameters of step length and single-limb support. If a difference in plantar pressure distribution exists between hallux valgus and healthy feet, it thus seems to not affect the overall gait pattern. Such a possible difference would thus be mostly compensated within the foot and ankle region itself, without affecting sagittal motion in, for example, the knee or hip region. Indeed, when considering the hindfoot, the midfoot, and the forefoot regions separately, no difference in plantar pressures could be observed between the pathological side and the contralateral healthy foot.

We had hypothesised, nevertheless, that it is possible to differentiate between affected and healthy feet on the basis of these pedobarographic data. We had especially expected reduced pressures under the first ray and increased pressures under the second and third metatarsal heads. To further investigate this possibility, we divided the plantar pressure profile into eight masks of interest, as suggested previously by Hutton et al. [18], and developed a special programme to analyse peak pressures in each of these regions during the gait cycle. Using this set-up, we could identify characteristic differences between hallux valgus feet and healthy control feet: pathological feet showed significantly increased pressures under the second and third metatarsal heads. This was accompanied by reduced pressures under the first ray on the same side, although the measured values failed to reach statistical significance. However, the fact that significantly higher values were again measured under the fourth and fifth toes underlines the concept that, by avoiding the first painful ray, more pressure is transferred laterally, the increased pressures under the fourth and fifth toes indicating a walking profile that is terminally more supine. Even though there is also literature saying that no increased pressure can be measured under the second and third metatarsal head [8, 23] our results are consistent with the common clinical finding of transfer metatarsalgia and hyperkeratosis under the second and third metatarsal heads in patients with hallux valgus. Higher pressure values under the third to fifth metatarsal heads in patients were also described by Hutton et al. [18], who compared the plantar foot pressures of 65 patients with hallux valgus and of 64 healthy subjects by using a single force transducer plate. Mickle et al. and Bryant et al. [7, 9] also found significantly higher peak pressures under the second metatarsal head in patients with hallux valgus. They also showed, however, higher metatarsal head I and hallux peak pressures in patients than in healthy subjects, which contrasts with our findings. In contrast, Hutton et al. [18], described significantly lower peak pressures under the hallux, and lower pressures under the first ray were described by Kadono et al. [10]. After thorough examination of all the studies mentioned, the reasons for this discrepancy in observations remain speculative. It is conceivable that different plantar loading patterns are present at different stages of hallux valgus, as soft tissues adapt to forefoot deformity with bunion formation medially, hyperkeratosis under the second and third metatarsal heads, and different connective tissue tension in the metatarsal head region, with the first metatarsal bone going progressively into increased valgus. Greater hallux valgus severity is correlated with lower pressures under the hallux itself [24]. Yet one would expect that such differences in hallux valgus severity would extrapolate to larger samples. Another relevant factor seems to be the pain level, as individuals affected by pain in the first metatarsophalangeal joint may adopt strategies to offload the painful area during gait [25].

The described differences taken together, as previously suggested by Nix et al. [6], can be attributed to differences in study design; patient-related factors such as age, sex, weight, degree of deviation and pain; and different measuring set-ups. Of interest, however, are the results recently reported by Galica et al. in the Framingham foot study [26]: analysis of plantar pressures of over 3000 participants showed significantly lower maximum forces on hallucal loading and higher forces at the lesser toes in patients with hallux valgus. Moreover, the medial forefoot showed lower maximum forces and peak pressures. The latter differences also failed to be statistically significant. Notably, however, in that model, the authors differentiated only the medial from the lateral forefoot, thus including pressures from the second metatarsal head in the medial ray analysis.

Since it can thus be considered established, that pressures indeed increase under the second and third metatarsal head in patients with symptomatic hallux valgus, an apparent goal of surgery should be to reestablish load bearing on the first ray. If certain procedures do this better than others and if patients also indeed benefit from such a strategy will still need to be further investigated. First studies indicate, however, that patients do clinically benefit from surgical procedures leading to increased plantarflexion with higher postoperative plantar pressures under the first ray [27, 28].

Although the treadmill set-up offers many advantages, walking in such an artificial environment leads to a slightly altered gait and plantar pressure profile when compared to normal overground walking. A phenomenon observed in several studies is significantly altered kinematics, with the most prominent example being a reduced hip extension and increased hip flexion when walking on the treadmill. The observed differences are, however, usually within a margin of 2–3°, which can be considered negligible [29‐31]. Stance time also appears to be slightly reduced [30]. With respect to ground reaction forces, most treadmill parameters seem to be significantly reduced in comparison with overground walking. The observed differences are, however, less than 5%. It has been pointed out, that the magnitude of these differences is comparable to the variability in normal gait parameters and within the range of repeatability and that it should thus not affect basic outcomes [29, 31]. Using a treadmill also leads to a lower spatial resolution in comparison to platform based systems. We also focus on peak pressures in our analysis, force-time integrals were not calculated. Even though measurements were performed video-controlled, the experimental setup did not optoelectronically register additional motion parameters which could have further described the subjects’ gait profile. To be able to measure with the same walking speed in all subjects, we chose a walking speed of 1 m/s, which is a the lower margin of customary speeds for treadmill analyses. While this offers a good comparability with overground walking [32], pathologic conditions might not as easily detected as with higher walking speeds.

Several parameters were significantly different from those of the control group, but were not significantly different from the contralateral side. We cannot evaluate how much of this phenomenon is due to the difference in age between the two groups and how much is due to the fact that hallux valgus is rarely a purely unilateral phenomenon. In many patients, the contralateral side also showed a milder form of hallux valgus, although there was no indication for surgery. For this reason, it is essential to compare the results obtained with those from our healthy control group.

Using an instrumented treadmill, we measured step width, step length, single-limb support, and plantar pressures divided into eight different masks from 36 patients with hallux valgus, and we compared these values to those from a healthy control group. Although no difference was observed in the spatiotemporal gait parameters, hallux valgus feet showed significantly higher peak pressures under the second and third metatarsal heads and under the fourth and fifth toes. We also observed reduced pressures under the first ray, although they were not statistically significant. The literature in this respect remains divided; nonetheless, this present combination of measured changes is consistent with the development of transfer metatarsalgia in patients with hallux valgus.