Ground reaction force and 3D biomechanical characteristics

of walking in short-leg walkers

Songning Zhang a,c,*, Kurt G. Clowers b, Douglas Powell a

a Biomechanics/Sports Medicine Lab, The University of Tennessee,

1914 Andy Holt Avenue, Knoxville, TN 37996-2700, USAb Anthropometry and Biomechanics Facility, NASA Johnson Space Center, Houston, USA

c Shanghai University of Sport, Shanghai, China

Received 23 September 2005; received in revised form 1 December 2005; accepted 11 December 2005

Abstract

Short-leg walking boots offer several advantages over traditional casts. However, their effects on ground reaction forces (GRF) and three-

dimensional (3D) biomechanics are not fully understood. The purpose of the study was to examine 3D lower extremity kinematics and joint

dynamics during walking in two different short-leg walking boots. Eleven (five females and six males) healthy subjects performed five level

walking trials in each of three conditions: two testing boot conditions, Gait Walker (DeRoyal Industries, Inc.) and Equalizer (Royce Medical

Co.), and one pair of laboratory shoes (Noveto, Adidas). A force platform and a 6-camera Vicon motion analysis system were used to collect

GRFs and 3D kinematic data during the testing session. A one-way repeated measures analysis of variance (ANOVA) was used to evaluate

selected kinematic, GRF, and joint kinetic variables ( p < 0.05). The results revealed that both short-leg walking boots were effective in

minimizing ankle eversion and hip adduction. Neither walker increased the bimodal vertical GRF peaks typically observed in normal walking.

However, they did impose a small initial peak (<1 BW) earlier in the stance phase. The Gait Walker also exhibited a slightly increased vertical

GRF during midstance. These characteristics may be related to the sole materials/design, the restriction of ankle movements, and/or the

elevated heel heights of the tested walkers. Both walkers appeared to increase the demand on the knee extensors while they decreased the

demand of the knee and hip abductors based on the joint kinetic results.

# 2005 Elsevier B.V. All rights reserved.

Keywords: Short-leg walker; Walking boot; Gait; 3D biomechanics; Walking

www.elsevier.com/locate/gaitpost

Gait & Posture 24 (2006) 487–492

1. Introduction

Short-leg rigid immobilization devices are commonly used

in treatment of acute and chronic injuries, and post surgical

interventions [1–10]. Fiberglass short-leg casts have been

traditionally used for these situations. Improvements in

prefabricated short-leg boots have provided an alternative to

traditional cast immobilization [4,11]. Walking boots offer

several advantages over traditional casts: ease of removal for

purpose of exercises, edema treatment, examination and

cleaning, less expensive, and less adverse effects on kinematic

and kinetic gait patterns than a synthetic walking cast [11].

* Corresponding author. Tel.: +1 865 974 4716; fax: +1 865 974 8981.

E-mail address: szhang@utk.edu (S. Zhang).

0966-6362/$ – see front matter # 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.gaitpost.2005.12.003

Indications for use of short-leg walking boots include ankle

and foot fractures, severe ankle sprains, chronic tendinopathy,

post surgical stabilization, and prevention and treatment of

ulceration due to sensory deficit in diabetic patients [11,12].

Several studies [13–16] have examined plantar pressure

distributions wearing different walkers but limited informa-

tion is available about the three-dimensional (3D) lower

extremity kinematics and kinetics of gait while wearing

walking boots [11]. Pollo et al. [11] examined 3D kinematics

and joint moments of walking in several walkers, a cast and

shoes. They concluded that short-leg walking boots elicit

less adverse effects of kinematics and kinetics in gait

compared to the synthetic walking cast. To the knowledge of

the authors, this is the only 3D biomechanical study on

walkers in gait published in the literature. Furthermore, the

S. Zhang et al. / Gait & Posture 24 (2006) 487–492488

information on ground reaction forces (GRF) of gait in

walking boots is not available in the literature. Further

examinations of GRF and related aspects of gait patterns in

short-leg walkers may provide useful information to

clinicians and patients with regard to long-term effects

since walkers may often be worn for a lengthy period of time

(up to six months). Therefore, the objective of this study was

to examine characteristics of lower extremity 3D kine-

matics, ground reaction forces, and joint dynamics during

walking in two different types of short-leg walking boots.

2. Methods

2.1. Subjects

Eleven healthy subjects (age: 27.4 � 7.8 years, body

mass: 72.0 � 13.4 kg, height: 1.76 � 0.08 m) with no

history of major injuries to their lower extremity participated

in the study. Among the subjects, five were female (age:

24.6 � 3.4 years, body mass: 61.3 � 9.0 kg, height:

1.69 � 0.03 m) and six were male (age: 29.7 � 9.9 years,

body mass: 80.9 � 9.3 kg, height: 1.82 � 0.05 m) partici-

pants. Each subject signed an informed consent form

approved by the Institutional Review Board at The

University of Tennessee prior to the actual data collection.

2.2. Experimental protocol and instrumentation

Each subject performed five level walking trials in each

of three conditions: two testing boots and one pair of

laboratory shoes. Prior to testing, each subject walked in one

randomly selected walking boot until he/she felt comfor-

table. The average walking speed of each subject was

determined from three walking trials at a preferred pace in

the walking boot using a pair of photocells (63501 IR,

Lafayette Instrument Inc., IN, USA) placed at shoulder

height. The walking speed was monitored to ensure that it

fell within the range of 10% of the average walking speed.

The boot conditions were arranged with one of the walkers

always being randomly tested first (to obtain the preferred

walking speed) and the shoe condition tested last.

A force platform (600 Hz, American Mechanical

Technology Inc., MA, USA) was used to measure the

ground reaction forces during testing. Three-dimensional

kinematic data of right lower extremity were simultaneously

collected using a 6-camera motion analysis system (120 Hz,

Vicon Motion Systems Ltd., Oxford, UK). Reflective

tracking markers (Fig. 1) were placed through a thin

thermoplastic shell (attached to a Velcro sensitive elastic

wrap) on the thigh and leg, and directly on the foot (the

walker conditions) or shoe (the shoe condition) of the right

side of the body. Tracking markers were placed on both sides

of the pelvis via a Velcro sensitive elastic strap. Anatomical

reflective markers were also placed on the anterior/posterior

iliac spines and iliac crest of both sides of the pelvis, on the

lateral side of the greater trochanter, on the lateral and

medial femoral epicondyles and the malleoli, and on the

head of first and fifth metatarsal, to determine the respective

joint/segment centers at the beginning of the data collection

session. Due to the usage of the walkers, the medial and

lateral malleolar markers in the walker conditions were

placed on the respective sites on the medial and lateral

plastic side arms of the walkers. The widths at the ankle with

and without the walker were measured with a caliper and

two virtual makers were set up to locate the true location

of the ankle and to estimate the ankle joint center. The hip

joint center was estimated from the pelvic and hip

anatomical markers using a modified method by Seidel

et al. [17].

2.3. Short-leg walking boots

Subjects wore two different walking boots, Gait Walker

(DeRoyal Industries, Inc., Powell, TN) and Equalizer

(Royce Medical Co., Camarillo, CA), on the right side

and a laboratory shoe on the left side in the boot conditions

during the test session. The linen wrap inside the walkers

was cut at the heel and lateral part of the mid and anterior

walkers to expose the skin for the attachment of the

anatomical/tracking markers on the foot (Fig. 1). Medial and

lateral plastic leg supports and Velcro straps of the walkers

were not altered; therefore, the integrity of the walkers was

maintained. They also wore a pair of the laboratory running

shoes (Noveto, Adidas) in the shoe condition.

2.4. Data and statistical analysis

Kinematic and GRF data were smoothed at 6 and 20 Hz,

respectively, using a fourth-order Butterworth low-pass

filter. The 3D kinematic and joint kinetic variables were

computed using Visual3D software suite (C-Motion, Inc.,

MD, USA), and the critical events and additional variables

were further determined by a customized computer program.

Internal moments of the lower extremity joints were

computed in Visual3D. The ground reaction forces and

joint moments were normalized to the participant’s body

mass, yielding units of N/kg and N m/kg, respectively. The

inversion/eversion of the ankle joint was computed as the

subtalar joint movement in the coronal plane. A one-way

repeated measures of analysis of variance (ANOVA) was

used to evaluate selected kinematic, GRF, and joint kinetic

variables (SPSS, 12.0). Post hoc comparisons were

conducted with an alpha level ( p < 0.05) adjusted for

multiple comparisons through a Bonferroni procedure.

3. Results

The participants in the study performed the walking trials

at a mean velocity of 1.24 � 0.18 m/s. The female and male

participants walked at similar speeds, 1.22 and 1.26 m/s,

S. Zhang et al. / Gait & Posture 24 (2006) 487–492 489

Fig. 1. Reflective markers placements on the lower extremity and pelvis.

respectively. The ANOVA results showed a significantly

greater maximum knee flexion angle for Gait Walker

compared to the no walker condition (Table 1). No other

variables related to the peaks and ranges of motion (ROM) of

the three lower extremity joints were found significant

between the testing conditions. The peak ankle eversion

angle was found significantly smaller than the no walker

Table 1

Average peak and ROM (8) of lower extremity joint angles in sagittal plane: me

Condition Ankle (8) Knee (8

Dorsiflexion ROM Flexion

No walker 11.9 � 3.4 5.7 � 4.6 15.5 � 8

Gait Walker 11.1 � 4.3 7.3 � 4.3 22.7 � 4

Equalizer 10.4 � 3.8 5.4 � 2.8 19.5 � 6

a Significantly different from Gait Walker.

Table 2

Average peak and ROM (8) of lower extremity joint angles in frontal plane: me

Condition Ankle (8) Knee

Eversion ROM Adduc

No walker �4.5 � 2.3 8.7 � 3.3 4.6 �Gait Walker �2.8 � 4.6 1.8 � 4.9a 3.9 �Equalizer �0.8 � 2.8a 6.6 � 4.6 1.5 �

a Significantly different from Gait Walker.

trials (Table 2). The eversion ROM was greater for the Gait

Walker compared to the no walker condition. In addition, the

hip abduction ROM for the Gait Walker and Equalizer

walkers were significantly smaller than those for the shoes.

In addition to the two vertical GRF peaks associated with

the loading response (Max 2) and terminal stance (Max 3)

commonly observed in walking in shoes, an apparent peak

(Max 1) occurs earlier than the peak of loading response for

the two walker conditions (Fig. 2). The statistical results

indicated no significant differences for the GRF related

variables between the test conditions (Table 3).

For the joint kinetics, the peak plantarflexor moment that

occurs later in the stance phase for the two walker conditions

was greater than the no walker trials (Fig. 3a). In both walker

conditions, the peak knee extensor moments were greater

than the no walker trials. On the other hand, the statistical

results showed a significantly smaller peak dorsiflexor

moment for Gait Walker compared to the no walker and

Equalizer walker conditions during earlier stance (Fig. 3b).

The maximum ankle inversion moment for the Gait Walker

condition was significantly greater than the no walker

(Fig. 4). The peak knee abduction moments for the two

walkers were smaller than the shoe condition; the same

moment variable for Equalizer was also smaller than Gait

Walker. Finally, the peak hip abduction moment for

Equalizer was significantly smaller than the shoe condition.

4. Discussions

The kinematic data from this study showed no major

changes in the peaks and ROMs of lower extremity joint

kinematics in the sagittal plane with the exception of a slight

increase in max knee flexion during the Gait Walker walking

condition. However, the range of motion from heel strike did

an � standard deviation

) Hip (8)

ROM Extension ROM

.7 8.5 � 5.4 0.6 � 10.9 37.1 � 5.4

.9a 9.5 � 4.7 2.4 � 9.5 37.2 � 4.4

.1 10.4 � 3.7 4.0 � 11.4 36.4 � 5.2

an � standard deviation

(8) Hip (8)

tion ROM Adduction ROM

2.9 3.3 � 1.8 5.8 � 2.8 8.3 � 2.4

2.2 2.4 � 2.5 5.1 � 3.6 6.1 � 2.0a

3.4 2.1 � 2.2 6.1 � 3.2 6.0 � 2.3a

S. Zhang et al. / Gait & Posture 24 (2006) 487–492490

Fig. 2. Representative curves of vertical ground reaction force for: no

walker (A), Gait Walker (B), and Equalizer (C).

Fig. 3. Average peak extensor moments (a) and flexor moments (b) lower

extremity joints (N m/kg) in sagittal plane; (1) significantly different from

no walker and (2) significantly different from Gait Walker.

not statistically differ from the Equalizer and the shoe

walking trials. Major differences of joint kinematics were

seen in the frontal plane. The walking trials in the Gait

Walker showed reduced range of motion for subtalar joint

eversion and hip adduction compared to the walking trials in

shoes. The Equalizer walker also exhibited reduced

Table 3

Average peak vertical ground reaction force (N/kg): mean � standard

deviation

Condition Max 1 Max 2 Max 3

No walker – 10.77 � 0.59 10.68 � 0.41

Gait Walker 8.91 � 1.49 10.27 � 0.72 10.47 � 0.59

Equalizer 7.37 � 2.74 10.72 � 0.61 10.43 � 0.44

(–) No apparent peak observed.

maximum eversion angle and hip adduction range of

motion. These data suggest that both walkers restrict

motions of the subtalar and hip joints in the frontal plane.

The reduced hip adduction may be related the restriction

provided by the walkers at the ankle/foot complex. The

previous study [11] demonstrated no significant differences

of hip and knee kinematics in sagittal (flexion/extension)

Fig. 4. Average peak joint moments (N m/kg) of lower extremity joints in

frontal plane; (1) significantly different from no walker and (2) significantly

different from Gait Walker.

S. Zhang et al. / Gait & Posture 24 (2006) 487–492 491

and frontal (adduction/abduction) planes between the walker

conditions. Our data basically agreed with the finding.

The data for both walkers from this study showed the

early peak in the vertical GRF right after the heel strike and

prior to the loading response (Fig. 2). Both walkers have a

Polyurethane outsole, a Polypropylene midsole, and a hard

form as the insole. Their heel height is greater than the

laboratory shoes. A closer examination of the walkers and

shoes used in this study showed that the heel thickness taken

at the mid heel region were 2.4, 3.2, and 3.6 cm on average

for the shoes, Equalizer, and Gait Walker, respectively. The

raised heel height on the walker side artificially increases the

limb length discrepancy. The sole materials/construction,

the restriction of ankle movements, and the heel height may

contribute to the observed initial GRF impact that is absent

from the shoe walking.

The vertical ground reaction force profile for the two

walkers also demonstrated a marked difference that occurs

between the loading response and terminal stance (Fig. 2).

The GRF curve for normal walking in shoes shows a typical

smooth valley between the two peaks. The GRF curve for the

Equalizer walker showed a more similar pattern to the shoe

walking compared to the Gait Walker. The GRF curve for

the Gait Walker trial demonstrated an elevated portion

between the peaks. The elevated portion of the GRF curve is

associated with stance phase when the walker is rolled from

the heel strike to the midstance. An examination of the

outsole of the two walkers showed that both walkers have a

curve at the heel region, which should facilitate the

progression of the body from heel strike to midstance.

The Equalizer walker has a smooth and slightly greater

curve throughout the outsole from the heel to the toe region.

The Gait Walker shows a flatter curve in the outsole,

especially at the region between the heel and the mid-foot,

which is almost entirely flat. The measurements on the shoes

and walkers indicated that the forefoot thickness (taken at

the region of the third metatarsal head) was 1.9, 2.2, and

3.1 cm on average for the shoes, Equalizer and Gait Walker,

respectively. For the testing shoes, the thickness difference

between the forefoot and heel regions is 0.5 cm, whereas the

differences were 1.1 and 0.5 cm for Equalizer and Intuition,

respectively. The differences further verified our initial

observed differences in the sole designs in these walkers.

The greater heel thickness with respect to its forefoot region

was observed in the Equalizer walkers and this may facilitate

transition from the heel strike to the toe-off as the center of

mass progresses forward. This may be especially necessary

while wearing a walker due to the restricted ankle dorsi-/

plantarflexion. On the other hand, the smaller difference of

the heel–forefoot thickness seen in the shoes and Gait

Walker compared to the Equalizer does not affect normal

walking in regular shoes since the ankle joints can dorsi-/

plantarflex freely to accommodate the rolling action needed

to facilitate the forward progression of the center of mass

during the stance. However, this transition process may be

somewhat more restricted wearing the Gait Walker due to

the smaller thickness difference between the heel and

forefoot regions and the flatter outsole curve of the walker,

leading to the elevated vertical ground reaction force around

the midstance. This may be also related to the increased

maximum knee flexion in the earlier stance associated with

the walker to accommodate the need for the forward body

progression. On the other hand, the elevated GRF suggests

that the Gait Walker was able to maintain a low but rather

‘‘constant’’ load and avoid abrupt changes in the ground

reaction force during midstance. This unique characteristic

may benefit patients by decreasing loading rates between the

two GRF peaks and promoting healing by maintaining a

relatively constant load. So far, the authors have not been

able to find published documents on GRF characteristics of

walkers in gait. Pollo et al. [11] report the ground reaction

forces in their study.

The peak knee abduction moment for the Gait Walker and

Equalizer walkers were both found to be significantly

smaller than the shoe condition. The Equalizer walker also

demonstrated a reduction in the peak hip abduction moment.

These reductions occur in early stance phase and may be

related to diminished needs of knee and hip abductors to

restrain adductions of the joints in the early stance due to the

application of the walkers. Pollo et al. also found decreased

knee and hip abduction moments for some of the tested

walkers in early stance phase [11]. The knee moment is

considered to be important in maintaining appropriate

loading to the lateral and medial compartments of the knee

[11] and therefore the mediolateral stability. The increased

inversion ankle moment seen in the Gait Walker is related to

the diminished eversion ROM, suggesting the better

performance of the walker in restricting subtalar joint

motion during the earlier support phase.

The greater peak knee extensor moment for both walkers

compared to the shoe walking trials may be related to the

constraint provided by the arm supports and straps of the

walkers to the ankle joint movements and the increased mass

(walker) attached to the leg and foot, which in turn require

the knee extensors to exert a greater torque to extend the

knee to facilitate the rolling from the heel strike to the toe-off

during the stance phase. The greater heel thickness observed

in the walkers may also increase the length of the walker

side’s limb and thus place it at a slight disadvantage. This

‘‘increased’’ leg length requires the knee extensors to exert

greater amount of torque to raise the center of mass to the

required height for a smooth transition across the midstance.

It was reported that the Equalizer walker along with another

walker (Cam walker) also had greater knee extensor

moments [11]. The authors suggested this increased

moments may lead to increased loading applied to the

tibiofemoral and patellofemoral joints. The joint moment

data from this study also indicated a smaller peak dorsiflexor

moment for the Gait Walker. This reduction occurring in the

earlier stance phase showed a decreased involvement of

dorsiflexors in the walker conditions. This suggests that the

restriction from a walker may reduce the need for the

S. Zhang et al. / Gait & Posture 24 (2006) 487–492492

dorsiflexors to actively oppose the plantarflexion seen in the

earlier stance phase. However, both walkers showed an

increased plantarflexor moment in late stance phase,

suggesting an elevated effort from the plantarflexors during

push-off. It is unclear why this occurred.

5. Conclusion

This study showed both short-leg walking boots,

DeRoyal’s Gait Walker and Royce’s Equalizer, were

effective in minimizing motion of ankle eversion and hip

adduction in frontal plane. Both walkers did not increase the

two peak ground reaction forces observed in normal walking

in shoes. However, they did impose a small initial peak

(<1 BW) in early stance phase. Due to the difference in sole

design, the Gait Walker exhibited a slightly elevated vertical

ground reaction force around midstance. Both walkers

increased the demand on knee extensors while they

decreased the effort of the knee and hip abductors. Both

walkers have different sole materials/construction, restricted

ankle movements, greater weight, and greater heel heights

compared to the shoes used in the study and some of the

observed biomechanics differences including the observed

initial peak may be related to these differences. Although

untested, it is logical to hypothesize that placing an

orthotic insert in the shoe of the unaffected limb may

relieve the initial GRF peak associated with the heel height

difference. Since the walkers may be worn for a long period

of time, the observed initial vertical GRF peak may impose

some adverse effect on the affected limb. The effects of

the observed biomechanical changes of the affected side on

the movements of the unaffected limb are almost entirely

unknown. If the compensatory changes on the unaffected

side do occur, it may cause undesirable outcomes such as

pain in sacroiliac joint and low back due to prolonged usage

of a short-leg walker. Finally, some of the significant

differences are small and their clinical impacts have yet to be

investigated. Therefore, further studies on these aspects of

short-leg walkers are warranted.

Acknowledgments

This study was funded by a grant from DeRoyal

Industries, Inc., a grant from Charlie and Mai Coffey

Endowment, and a grant from the Scholarly Activity and

Research Incentive Fund at The University of Tennessee.

References

[1] Wapner KL, Chao W. Nonoperative treatment of posterior tibial

tendon dysfunction. Clin Orthop 1999;39–45.

[2] Crincoli MG, Trepman E. Immobilization with removable walking

brace for treatment of chronic foot and ankle pain. Foot Ankle Int

2001;22:725–30.

[3] Kader D, Saxena A, Movin T, Maffulli N. Achilles tendinopathy: some

aspects of basic science and clinical management. Br J Sports Med

2002;36:239–49.

[4] Kadel NJ, Segal A, Orendurff M, Shofer J, Sangeorzan B. The efficacy

of two methods of ankle immobilization in reducing gastrocnemius,

soleus, and peroneal muscle activity during stance phase of gait. Foot

Ankle Int 2004;25:406–9.

[5] Maffulli N, Kader D. Tendinopathy of tendo achillis. J Bone Joint Surg

Br 2002;84:1–8.

[6] King DM. Experience with the below-knee total-contact cast in the

management of tibial fractures. Aust N Z J Surg 1975;45:54–6.

[7] Preto A. Patellar tendon bearing and cast braces: total contact orthoses

in the weight-bearing treatment of tibial and femoral shaft fractures.

Ona J 1977;4:10–1.

[8] Crates JM, Richardson EG. Treatment of stage I posterior tibial tendon

dysfunction with medial soft tissue procedures. Clin Orthop 1999;46–9.

[9] Cole BJ, Freedman KB, Taksali S, Hingtgen B, DiMasi M, Bach Jr

BR, et al. Use of a lateral offset short-leg walking cast before high

tibial osteotomy. Clin Orthop 2003;209–17.

[10] Matricali GA, Deroo K, Dereymaeker G. Outcome and recurrence rate

of diabetic foot ulcers treated by a total contact cast: short-term follow-

up. Foot Ankle Int 2003;24:680–4.

[11] Pollo FE, Gowling TL, Jackson RW. Walking boot design: a gait

analysis study. Orthopedics 1999;22:503–7.

[12] Randolph AL, Nelson M, deAraujo MP, Perez-Millan R, Wynn TT.

Use of computerized insole sensor system to evaluate the efficacy of a

modified ankle–foot orthosis for redistributing heel pressures. Arch

Phys Med Rehabil 1999;80:801–4.

[13] Baumhauer JF, Wervey R, McWilliams J, Harris GF, Shereff MJ. A

comparison study of plantar foot pressure in a standardized shoe, total

contact cast, and prefabricated pneumatic walking brace. Foot Ankle

Int 1997;18:26–33.

[14] Crenshaw SJ, Pollo FE, Brodsky JW. The effect of ankle position on

plantar pressure in a short leg walking boot. Foot Ankle Int

2004;25:69–72.

[15] Nawoczenski DA, Birke JA, Coleman WC. Effect of rocker sole

design on plantar forefoot pressures. J Am Podiatr Med Assoc

1988;78:455–60.

[16] Pollo FE, Brodsky JW, Crenshaw SJ, Kirksey C. Plantar pressures in

fiberglass total contact casts vs. a new diabetic walking boot. Foot

Ankle Int 2003;24:45–9.

[17] Seidel GK, Marchinda DM, Dijkers M, Soutas-Little RW. Hip joint

center location from palpable bony landmarks—a cadaver study. J

Biomech 1995;28:995–8.