Comparison of gait properties during level walking and stair ascent and descent with varying loads

doi:10.4236/health.2010.212203

Paper Menu >>

Journal Menu >>

Vol.2, No.12, 1372-1376 (2010) Health

doi:10.4236/health.2010.212203

Comparison of gait properties during level walking and

stair ascent and des cen t wi th v ary ing loa ds

Tomohiro Demura1*, Shin-ich Demura1, Sohee Shin2

1Graduate School of Natural Science & Technology, Kanazawa University, Kanazawa, Ishikawa, Japan;

*Corresponding Author: tomodemu@ed.kanazawa - u.ac.jp;

2Kanazawa University Center for Innovation Venture Business Laboratory Kanazawa, Ishikawa, Japan.

Received 1 September 2010; revised 9 September 2010; accepted 16 September 2010

ABSTRACT

This study aimed to compare gait properties

during level walking and during stair ascent and

descent with varying loads. Fifteen healthy

young men (mean age: 22.1 ± 1.6 years) walked

while holding four different loads relative to

each subject’s body mass (0, 20, 40 and 60% of

body mass: BM) on their backs. Stance time,

swing time, and double support times were se-

lected as gait parameters. All parameters

showed a maximal value during stair ascen t a nd

a minimum value during level walking. Stance

and double support times increased significan-

tly with each load during level walking and during

stair ascent and descent. In conclusion, st-

air ascent and descent creates more unstable

movement than level walking regardless of the

weight of the load. The effect of loads on gait

increases with the weight of the load and be-

comes obvious once the load exceeds 60% of

BM.

Keywords: Gait; Stair Ascent; Stair Descent; Load

1. INTRODUCTION

Climbing up and down stairs is an important activity

in daily life and has been studied mainly from kinetic

and kinematic standpoints. I t was reported that angle and

moment of knee flexion are larger during stair ascent

than during level walking [1,2]. According to Andriacchi

et al. [1], knee flexion moment was roughly three times

larger during stair ascent and hip flexion moment was

about one and a half times larger during stair descent

than during level walking.

The movement of ascending and descending differs

considerably from level walking from a mechanical stand-

point and requires great exertion of leg strength [3,4].

Because the center of gravity move up and down. Fur-

thermore, while ascending, we must raise one foot higher

than each step. Hence, it is assumed that because a single

support time is longer during stair ascent and descent than

during level walking, a large burden is imposed on the

support leg and posture instability is increased.

In daily life, it is very popular to walk with a shopping

bag in one or both hands, which can be a heavy burden.

Until now, the issue of walking with loads has been

mainly studied from the standpoint of walking speed [5]

or energy cost [6]. According to Ghori and Luckwill [7 ],

leg muscle activity increases and the ongoing electro-

myographic activity prolongs during level walking with

loads. In add ition, by holding loads, the body’s center of

gravity is raised and walking becomes more unstable [8,

9]. As a result, to maintain gait stability, gait changes

such as a decrease in swing time [7] or an increase of

double support time [10] occur.

From the above, it is assumed that stair ascent and

descent with loads produces larger gait changes than

level walking and that more instability results. The pur-

pose of this study was to compare gait properties during

level walking and during stair ascent and descent with

varying loads.

2. MATERIALS AND METHODS

2.1. Subjects

Fifteen healthy young men without extremity disorders

participated in this study (age : 22.1 +/– 1.6 years, height:

172.5 +/– 4.9 cm, body mass: 67.6 +/– 5.0 kg). Before

the measurements were taken, the purpose and procedure

of this study were explained in detail and informed con-

sent was obtained from all subjects.

2.2. Material

Gait properties were measured by a gait analysis sys-

tem (Walk Way MG-1000, Anima, Japan) in r eference to

a previous study [11]. The MG-1000 with plate sensors

can determine time, dimensions and the distance of the

T. Demura et al. / Health 2 (2010) 1372-1376

1373

foot or feet when the foot touches its surface and can

measure grounding/non-grounding on the bearing sur-

face as an on/off signal. Data were recorded into a per-

sonal computer at 100 Hz.

The stairs used for the present experiment consisted of

four steps (step width: 90 cm, step height: 20 cm, tread

length: 29 cm) and was within the Japanese building

code [12] step height: under 23 cm, tread length: over 15

cm) (see Figure 1).

2.3. Protocol for Measurement

In this study, we imposed varying weight loads on the

subjects in reference to a previous study by Demura and

Demura [11] (see Figure 2). In short, the subjects walked

with four loads relative to each subject’s body mass (0,

20, 40 and 60% BM) on their backs. The trial order of

each load condition was randomized. The movement

during measurement was explained to the subjects be-

fore the measurements were taken. For stair ascent and

descent, subjects went up the stairs (four steps), and

walked straight about two meters on the stock, made a

right about-face turn, and after standing still for a few

seconds, went down the stairs (see Figure 1). To elimi-

nate the influence of fatigue, the subjects performed

each load condition twice with an enough rest. In addi-

tion, for stair ascent and descent, we used data of four

steps from the first step foot on the stair for analysis.

During level walking, subjects were instructed to walk

straight for eleven meters as normal. To eliminate a fati-

gue effect, the subjects walked under each load condition

twice with a one minute rest. In addition, we used only

the middle five meters of data excluding the first and

final three meters for analysis.

2.4. Parameters

Figure 3 shows the gait property parameters selected

in reference to a previous study Murray et al. [13]. A

mean of two trials was used for analysis. Stance time

equals the duration that the body is supported by one

foot or both feet, that is, the phase in which one foot or

both feet contact the floor. A swing time equals the dura-

tion that one foot is swinging, that is, one foot is off the

floor. This corresponds with single-legged support time.

Double support time equals the duration in which both

feet are in contact with the floor.

2.5. Data Anal ysis

Mean differences among load and movement condi-

tions for parameters were tested by repeated two-way

analysis of variance (ANOVA: load condition × move-

ment). When an interaction was significant, Bonferroni’s

method was selected for a multiple comparison. The

statistical SPSS package ver. 11.0 (SPSS, America) was

used for data analysis. The probability level of 0.05 was

indicative of statistical significance.

3. RESULTS

Table 1 shows results of two-way ANOVA for gait

parameters. A significant interaction is found in all pa-

rameters. In the results of multiple comparisons, stance

time is significantly longer in the ord er of level walking,

stair descent and ascent in all load conditions. In level

walking, it is significantly longer in the order of 0% or

20% BM, 40% BM and 60% BM. It was significantly

longer in 6 0% BM than 0% and 20% BM in stair ascent

and descent and in 60% BM than 40% BM in stair as-

cent.

Swing time is significantly longer in the order of level

walking, stair descent and stair ascent at 0% and 20%

BM, in stair ascent than stair descent at 60% BM, and in

stair ascent and descent than level walking at 40% and

60% BM. In level walking, it is significantly longer at

40% or 60% BM than 0% BM and at 40% BM than 20%

BM.

The double support time in all load conditions was

significantly longer in the order of level walking, stair

20cm

Step height

Ste p wid th

Tread length

Figure 1. The stairs used for the present experiment consisted of four steps (step width: 90 cm, step height: 20 cm, tread length: 29

cm).

T. Demura et al. / Health 2 (2010) 1372-1376

1374

Table 1. Results of 2way-ANOVA during stair ascent and descent, and level walking with loads.

yg, g

Mean SDMean SDMean SDMean SDfactor

ascent 0.780.070.820.070.89 0.110.96 0.12F1134.18*

descent 0.700.050.740.060.80 0.100.85 0.13F218.25*

level 0.630.030.640.040.66 0.040.68 0.05F35.52*

ascent 0.470.040.470.030.48 0.050.49 0.05F155.61*

descent 0.440.040.430.040.46 0.060.47 0.06F21.17

level 0.410.010.400.010.40 0.010.40 0.02F33.57*

ascent 0.150.020.170.020.20 0.040.24 0.05F168.12*

descent 0.140.020.160.020.18 0.030.19 0.04F234.64*

level 0.110.010.120.010.14 0.010.15 0.02F36.74*

F1: movement, F2: loads, F3: interaction, * : p<0.05,

ascent: stair ascent, descent: stair descent, level: level walkin

Double

Support

Ti me

(sec)

a ll loads: level < descent < ascent

ascent: 0%<20, 40%<60%

d e scent: 0% <20, 40, 60% 20% <60%

level : 0%<20%<40% < 60%

Post-hoc

BonferroniF-value

Stance

Ti me

(sec)

a ll loads: level　< descent < ascent

ascent: 0% < 40% < 60% 20% < 60%

descent: 0% < 40, 60% 20% < 60%

level: 0, 20% < 40% < 60%

Swing

Ti me

(sec)

0, 20%: level < descent < ascen t

60%: descent < ascent

40, 60%: level < ascent, descent

0%BM 20%BM 40%BM60%BMANOVA

Figure 2. The va rying wei ght loads on

the subjects.

descent and ascent. In stair ascent, it is significantly

longer in 60% BM than 20% and 40% BM, and in 20%

and 40% BM than 0% BM. In stair descent, it is longer

in 20%, 40% and 60% BM than 0% BM, and in 60%

BM than 20% BM.

4. DISCUSSION

4.1. Effect of Loads

Subjects walked the level plane and stairs with four

loads relative to subject’s body mass (0, 20, 40 and 60%

BM). In level walking and stair ascent and descent,

stance time and double support time increased with loads.

It has been reported that energy consumption [14,15]

and the activity of erector spinae and gastrocnemial

muscles increase [10,16] while walking with loads.

Hence, walking with loads is considered to impose a

large burden on the lower limbs.

Ghori and Lu ckwill [7] rep orted that th e body’s center

of gravity is raised by holding heavy loads and that

walking becomes unstable. Consequently, a decrease in

swing time [7] and an increase in double support time [9,

10] occur to maintain a stable posture. In this study, the

swing time was unchanged in spite of an increase in

loads. However, the percentage of swing time to stride

time decreased. This change is believed to be one of the

strategies to keep a stable posture during walking.

4.2. Comparison of Level Walking, and Stair

Ascent and Descent

Meanwhile, all parameters tended to be longer in the

order of level walking, stair descent and ascent. When

walking the same distance, stair ascent and descent have

more phases which move the body’s center of gravity

more vertically than during level walking. Thus, it can

be simply considered that gait time parameters became

longer during stair ascent and descent than during level

walking due to this extra vertical movement.

Ascending requires greater time and leg strength to

raise the body on a step after the lifted leg has landed. It

is assumed that because of this, ascending the stairs

takes longer than descending them.

In addition, it was reported that an increase in double

support time is to maintain stability of walking [17-20].

Also in this study, the double support time was longer

during stair ascent and descent than during level walking.

From the above, because the burden imposed on the

body is larger during stair ascent and descent than during

level walking, ascending and descending the stairs is

considered to destabilize movement.

T. Demura et al. / Health 2 (2010) 1372-1376

1375

Left

foot

Right

foot

Rig htheel offLeft heel

contact Left heel

off Right heel offRight heel

contact

Double

support

phase

Double

support

phase

Stance

time

Stance

time

Stance

time

Swing

time

Swing

time

Swing

time

Stride

Time (sec)

S ing le suppo rt ph as eS ing le suppo rt ph as e

Figure 3. The gait property parameters regarding time.

Furthermore, gait parameters regarding time increased

in all movements with loads. Th e effect of varying loads

on stair ascent and descent was similar to that on level

walking. In addition, the present results suggested that

the effect of weight loads on gait properties increases

with loads and becomes obvious at 60% BM during lev-

el walking and during stair ascent and descent. This re-

sult agrees with that in the previous study [7,21,22] that

the change in gait became clear at over 50% BM.

5. CONCLUSIONS

In conclusion, stance time, swing time and double

support time are longer in the order of level walking,

stair descent and ascent. In addition, stance time and

double support time increased with loads during level

walking and during stair ascent and descent. Ascending

and descending the stairs is more unstable compared to

level walking regardless of load weight. In addition, the

effect of loads on gait becomes obvious in loads over

60% BM during level walking and while ascending and

descending stairs.

REFERENCES

[1] Andriacchi, T.P., Andersson, G.B., Fermier, R.W., Stem,

D. and Galante, J.O. (1980) A study of lower-limb me-

chanics during stair-climbing. The Journal of Bone &

Joint Surgery, 62, 749-757.

[2] Jevsevar, D.S., Riley, P.O., Hodge, W.A. and Krebs, D.E.

(1993) Knee kinematics and kinetics during locomotor

activities of daily living in subjects with knee arthroplas-

ty and in healthy control subjects. Physical Therapy, 73,

229-238.

[3] Joeseph, J. and Watson, R. (1967) Telemetering electro-

myography of muscles used in walking up and down

stairs. Journal of Bone Joint Surgery, 49-B, 774-780.

[4] James, B. and Parker, A.W. (1989) Electromyography of

stair locomotion in elderly men and women. Electro-

myography and Clinical Neurophysiology, 29, 161-168.

[5] Zarrugh, M.Y. and Radcliffe, C.W. (1978) Predicting

metabolic cost of level walking. European Journal of

Applied Physiology, 38, 215-223.

[6] Hughes, A.L. and Goldman, R.F. (1970) Energy cost of

“hard work”. Journal of Applied Physiology, 29, 570-

572.

[7] Ghori, G.M.U. and Luckwill, R.G. (1985) Responses of

the lower limb to load carrying in walking man. Euro-

pean Journal Applied Physiology & Occupational Physi-

ology, 54, 145-150.

[8] Goh, J.H., Thambyah, A. and Bose, K. (1998) Effects of

varying backpack loads on peak forces in the lumbosa-

cral spine during walking. Clinical Biomechanics, 13,

26-31.

[9] Singh, T. and Koh, M. (2009) Effects of backpack load

position on spatiotemporal parameters and trunk forward

lean. Gait & Posture, 29, 49-53.

[10] Harman, E., Han, K.H., Frykman, P., Johnson, M., Rus-

sell, F. and Rosenstein, M. (1992) The effects on gait

timing, kinetics and muscle activity of various loads car-

ried on the back. Medicine & Science in Sports & Exer-

T. Demura et al. / Health 2 (2010) 1372-1376

1376

cercise, 24, S129.

[11] Demura, T. and Demura, S. (2010) Relationship among

gait parameters while walking with varying loads. Jour-

nal of Physiological Anthropology, 29, 29-34.

[12] Information center for building administration, Order for

enforcement of the building standards act, 2008.

[13] Murray, M.P., Drought, A.B. and Kory, R.C. (1964)

Walking patterns of normal men. Journal of Bone and

Joint Surgery, 46, 335-360.

[14] Borghols, E.A.M. (1978) Influence of heavy weight car-

rying on the cardiorespiratory system during exercise.

European Journal of Applied Physiology, 38, 161-169.

[15] Taylor, C.R., Heglund, N.C., McMahon, T.A. and Looney,

T.R. (1980) Energetic cost of generating muscular force

during running. A comparison of large and small animals.

Journal of Experimental Biology, 86, 9-18.

[16] Norman, R.W. (1979) The utility of combining EMG and

mechanical work rate data in load carriage studies. Pro-

ceedings of the International Society of Electrophysio-

logical Kinesiology, 4, 148-149.（Abstract）

[17] Imms, F.J. and Edholm, O. G. (1981) Study of gait and

mobility in the elderly. Age and Aging, 10, 147-156.

[18] Menz, H.B., Load, S.R. and Fitzpatrick, R.C. (2003)

Age-related differences in walking stability. Age and

Ageing, 32, 137-142.

[19] Oberg, T., Karsznia, A. and Oberg, K. (1993) Basic gait

parameters: Reference data for normal subjects, 10-79

years of age. Journal of Rehabilitation Research and

Development, 30, 210-223.

[20] Lord, S.R., Lloyd, D.G. and Li, S.K. (1996) Senso-

ri-motor function, gait patterns and falls in communi-

ty-dwelling women. Age and Ageing, 25, 292-299.

[21] Kinoshita, H. (1985) Effect of different loads and carry-

ing systems on selected biomechanical parameters de-

scribing walking gait. Ergonomics, 28, 1347-1362.

[22] Martin, J. and Nelson, R.C. (1986) The effect of carried

loads on the walking patterns of men and women. Ergo-

nomics, 29, 1191-1202.