Influence of footwear midsole material hardness on dynamic balance control during unexpected gait termination

1. Introduction 

A study done through interviews with older adults who had fallen while they were hospitalized revealed that in 51% of the falls poorly fitting shoes played a role [1]. Additionally, Sherrington and Menz [2] reported that 71 of 95 subjects, who experience a fall-related hip fracture, were wearing footwear with at least one “theoretically sub-optimal feature” when they fell. Only a small set of specific aspects of footwear and their influence on balance have been investigated. They include midsole material hardness, height of heel collar and outer sole slip resistance (frictional properties). The largest collection of work has been done on midsole material and control of balance by Robbins et al. [3], [4], [5], [6]. His investigations have involved both walking on a narrow beam and detecting ankle angle position in footwear that varied in hardness levels from soft (Shore A Scale 15) to hard (Shore A Scale 50). The outcome from narrow beam walking demonstrated that wearing softer soled footwear resulted in more falls from the beam during similar lengths of travel. During detection of ankle angle position testing the softer footwear resulted in less accurate matching of ankle angle position and the barefoot condition resulted in the best accuracy. The results of this work imply that midsole hardness influences the sensory function of the foot and potentially the control of balance during walking on a narrow beam. However, transition from this work to actual control of balance during gait is limited.

Lord et al. [7] also tested soft versus hard soled shoes during dynamic balance tests and found no differences in performance. Again these results are limited because the dynamic balance testing were voluntary tests rather then compensatory reactionary tests that are more representative of ‘real-life’ balance requirements [8]. The complexity that is inherent to these large prospective studies illustrates the need for controlled research studies into the effects of footwear design characteristics on dynamic control of balance. This will reduce or allow better control of confounding factors such as activity level, mobility and impairment.

The use of a controlled perturbation to gait such as unexpected gait termination would permit investigation of the influence of these footwear characteristics on dynamic balance control. Gait termination has been previously utilized to identify the role of plantar-surface cutaneous sensation and vision during dynamic balance control [9].

The purpose of this study was to determine the influence of different midsole hardnesses on dynamic balance control during unexpected gait termination. We hypothesized that a softer midsole material would result in a larger excursion of the body's COM motion relative to the lateral borders of the BOS and there would be a decreased ability to control the movement of the COM in the anterior direction as indicated by a larger center of pressure−center of mass difference. Also associated with these kinematic changes we also hypothesized that the force loading rates would increase to compensate for lack of mechanical support from the midsole material.

2. Methods 

Twelve healthy young female adults (20–23 years old, height range 164–180cm, and mass range 49–80kg) volunteered to participate in this study. The protocol was approved by the Institutional Ethics Review Board. Participants were asked to walk along an 8-m walkway, looking straight ahead. Randomly during 25% (3 of 12) of the trials, they were signaled (audio buzzer) to terminate gait within the next two steps. The audio cue was triggered by a foot contact force of 10N as detected by force plate one. The first step, where the signal to terminate gait was triggered, and the next two steps, where termination took place, all made contact with individual force platforms (Fig. 1A, [9]). Horizontal and vertical forces and moments under each foot were recorded using securely mounted force platforms and these data were used to calculate the center of pressure (COP). The force signals were sampled at 200Hz and low-pass filtered (10Hz) with a Butterworth dual-pass filter.

View Large Image Download to PowerPoint

Fig. 1. (A) Overhead view of 8-m walkway used for gait termination trials. (B) Simulated foot beds worn by the participants (from left to right: hard, standard and soft).

The whole body center of mass (COM) was calculated using a 13-segment model [10] with data collected from an OptoTrak Motion Analysis System (Northern Digital Inc., Waterloo, Ont., Canada). The three dimensional motion of the feet, legs, arms, trunk, head and the body as a whole were recorded at a sampling rate of 100Hz. The available base of support (BOS) was identified using a combination of force platform data (time of contact), foot marker locations (spatial location) and foot anthropometrics (marker position relative to borders of the foot).

The four experimental conditions were (midsole hardness as measured on a Shore A scale): (1) soft (A15); (2) standard (A33); (3) hard (A50); (4) barefoot; testing order was randomized. Each simulated pair of foot beds were made up of a formed insole of the three hardnesses and a 1cm thick standard (A33) hardness platform (Fig. 1B). These simulated foot beds were fixed to the sole of the foot with double-sided tape and then a snuggly fitting thin sock was pulled over to prevent unwanted movement during walking trials.

The primary outcome measures were maximum, minimum and range of the transverse plane projection of the center of mass location relative to the lateral base of support (COM−BOS), maximum and minimum of the center of mass−center of pressure (COM−COP) difference in the anterior posterior direction and the average vertical force loading rates during the first 100ms of each foot contact. The COM−BOS and COM−COP measures were calculated during the duration of the first single stance phase (on force Plate 1) and final (or second) single stance phase (on force Plate 2, see Fig. 1A). Analysis of the COM−BOS and COM−COP was done only in the transverse plane.

3. Results 

As midsole hardness increased there was an increase in the range of the medial–lateral COM movement measured relative to the lateral border of the BOS during the first single support phase (the phase that immediately followed the signal to terminate gait). The medial–lateral COM−BOS range for the soft (0.14m; max 0.19m to min 0.05m), standard (0.15m; 0.20–0.05m), hard (0.16m; 0.21–0.05m) and barefoot (0.22m; 0.25–0.03m) conditions increased as insole hardness increased (p=0.017, Fig. 2). Significant differences were found between barefoot and the three midsole conditions as well as the soft midsole and the hard midsole conditions.

View Large Image Download to PowerPoint

Fig. 2. Maximum, minimum and range of the medial–lateral COM−BOS differences during the first single support phase after unexpected signal to terminate gait (*significant differences).

The soft and standard midsole hardness conditions decreased the maximum anterior–posterior (AP) COM−COP difference during the final single stance phase of termination when compared to the barefoot condition (soft: 0.227m and standard: 0.229m versus barefoot: 0.25m; p=0.031, Fig. 3). A second interesting finding was that the vertical loading rate was higher in the soft and hard insole as compared to the barefoot condition (soft: 16.1kN/s and hard: 15.7kN/s versus barefoot: 13.1kN/s; p=0.003) during the final stance phase of termination (Fig. 4).

View Large Image Download to PowerPoint

Fig. 3. Effects of midsole hardness on anterior–posterior center of mass−centre of pressure (COM−COP) differences during the final single stance phase before termination of gait (*significant differences).

View Large Image Download to PowerPoint

Fig. 4. Vertical loading rate results for the final single stance phase before termination of gait (*significant differences).

4. Discussion 

The primary outcomes of this study have illustrated the influence of midsole hardness as an impediment to dynamic balance control during responses to gait termination. The decrease in the medial–lateral range of center of mass (COM) movement with respect to the base of support (BOS) is demonstrating a possible constraint being imposed upon the balance control system. Plus the reduction in the maximum center of mass−center of pressure difference is also indicating the restrictive nature of softer midsole material.

As these results indicate, the soft, standard and hard midsole material constrains the range of COM over the BOS. This may be required due to the softer interface that exists between the foot and the environment. With a softer interface the less of a mechanical support is offered to create reactive forces needed to counteract a range of COM movement observed in the barefoot condition. Thus the body appears to adapt and reduces the movement (or range) of COM movement. This could prove to be very detrimental because if a balance perturbation is experienced then the individual wearing soft midsoled footwear will have an impaired ability to effectively respond. Therefore any perturbation encountered, while wearing soft midsoled footwear, will require a greater mechanical response to maintain balance and hence reduce the upper limit of perturbation level that can be effectively dealt with. Examples of mechanical responses may include increases in muscle torque at the hip or knee and maybe even a compensatory step or grasp. Additionally, the soft midsole material may also be insulating sensory information (plantar-surface mechanoreceptors) needed to effectively respond to balance disturbances [9], [11].

The influence of soft midsole material on the center of mass−center of pressure (COM−COP) and the vertical loading rate is theorized in Fig. 5. This figure indicates a reduction in the leverage available (stable surface) when compared to the barefoot condition. Theoretically the changes resulting from the soft midsole material, of a similarly positioned COM vector relative to the ankle joint would be a decrease in COM−COP and an increase in vertical loading. This increase in loading would indicate that more muscular activity is required to stabilize the body when exposed to the same threat. Interestingly the present findings were a reduction in COM−COP and an increase in vertical loading during the final stance phase of termination. Even though no falls occurred in these young adults, the consequences to an older population with impaired sensory and musculoskeletal systems may be more severe.

View Large Image Download to PowerPoint

Fig. 5. Demonstration of the potential effects of soft midsole material on that produced a reduction in centre of mass to centre of pressure (COM−COP) difference and an increase in vertical reaction force (as indicated by vertical loading rate).

These results indicate that detailed analysis of the body's centre of mass relative to the changing base of support and measurement of reactive forces during controlled alterations of footwear design characteristics can provide valuable insight into their effects on control of dynamic balance. This in combination with a retrospective study by Frey and Kubasak [12], who found that 42% of community dwelling older adults reported that they were wearing soft-soled athletic footwear during falls in the past year, indicate the need to determine the effects of midsole hardness on control of balance in older adults. Menz and Lord [13] have presented a review of the multiple footwear characteristics that could theoretically affect postural stability, mobility and falls. The only concrete conclusion from this review of literature is that older adults should avoid wearing high-heeled footwear and that the ideal stable shoe is still somewhat nebulous. They recommended that randomized and controlled investigations been done into what footwear characteristics are required to provide stability for the older adults.

The conclusion from the present study is that variations in midsole material and even the presence of it, impairs the dynamic balance control system. The results opposed our initial hypotheses that the softer material would increase the COM−BOS excursion and the COM−COP differences. However, we did see the hypothesized effect of increased force loading rates for the softer material. Even though no loss of balance occurred in the young participants, these results indicate that further study with older adults is warranted to evaluate the risk that footwear cushioning presents to this population.