angular movements of the trunk and pelvis when stepping over obstacles of different heights

Angular Movements of the Trunk and Pelvis 219

219

Research in Sports Medicine 11: 219–233, 2003Copyright © Taylor & Francis Inc.ISSN:1543-8627 print / 1543-8635 onlineDOI:10.1080/15438620390254611

Angular Movements of the Trunk and Pelvis WhenStepping Over Obstacles of Different Heights

YUN WANG AND KAZUHIKO WATANABEHiroshima University

This study was performed to investigate the effect of obstacle height on theangular movements of the trunk and pelvis. Seven healthy subjects (all men)participated in the study (age: 22.0 ± 1.2 years; height: 1.73 ± 0.05 m; mass:69.1 ± 7.9 kg). Subjects were instructed to perform unobstructed level walk-ing and to step over obstacles corresponding to 5%, 10%, and 15% of eachsubject’s height. The video-based WINanalyze 3D motion measurement sys-tem was used to measure the three-dimensional (3D) angular movements ofthe trunk and pelvis. In the frontal and horizontal planes, peak-to-peak ampli-tudes of trunk rotation developed as the obstacle height increased. However,peak-to-peak amplitudes of pelvic rotation were not significantly different whenstepping over obstacles of different heights. The coefficients, which revealedthe degree of synchrony of the shoulder and pelvis, showed a trend of decreas-ing as the obstacle height increased. These results suggest that stepping overobstacles poses a significant challenge to the coordination of the trunk andpelvis. Pelvic rotation is important in adjusting the crossing stride and clear-ing an obstacle. In coordination with the pelvic motion, the shoulder plays arole in maintaining dynamic equilibrium when stepping over obstacles.

Keywords gait, obstacle, angular movement, trunk, pelvis

IntroductionOur path is usually obstructed and uneven in real-life situations. People are some-times confronted with a course consisting of undulations or obstacles. Stepping overobstacles is a more challenging task than unobstructed level walking and poses greatrisks (Chou, Kaufman, Robert, et al. 2001; Patla and Prentice 1995; Pryde, Roy, andPatla 1997). It was reported that tripping over obstacles and imbalance during gaitwere two of the most common causes of falls in the elderly (Campbell, Borrie, Spears,et al. 1990; Campbell, Reinken, Allan, et al. 1981; Tinetti and Speechley 1989).

Received 15 June 2003; accepted 28 August 2003.Address correspondence to Yun Wang, c/o Prof. Kazuhiko Watanabe, Lab. Physiology

and Sport Biomechanics, Graduate School of Education, Hiroshima University, Kagamiyama1-1-1, Higashi-Hiroshima, 739-8524, Japan. E-mail: [email protected]

220 Y. Wang and K. Watanabe

So far, the kinematics of the lower extremities when negotiating obstacles havebeen extensively investigated. It has been found that obstacle crossing may causeinappropriate movement of the lower extremities or misplacement of the swingfoot, which could consequently lead to a foot–obstacle contact and result in a fall(Chou, Kaufman, Robert, et al. 2001). However, little information is available re-garding the kinematics of the upper portion of the body. In fact, the mechanics ofthe trunk are important for understanding the motor strategy of human locomotion.Just as Cappozzo, Figura, Leo, et al. (1978) pointed out, the trunk and the pelvismove in a coordinated fashion so that their total mechanical energy variation dur-ing the walking cycle has a lower magnitude than it would have if the trunk movedrigidly with the pelvis.

Saunders, Inman, and Ebernardt (1953), having identified six major determi-nants of gait, reported that three of these are related to the motion of the pelvis. It isconsidered that one of the contributions of the pelvis is a forward rotation on theside of the swinging leg and an opposite rotation near the end of the stance phase.However, the pelvic angular momentum must be counterbalanced, either directlyby counterrotating the thorax or indirectly by swinging an arm. That is, trunk stabi-lization requires movement of the trunk on the pelvis for the compensation of pel-vic movement.

Unlike unobstructed level walking, minimal and precise modification ofmovement is required for safety and efficiency when stepping over an obstacle(Austin, Garrett, and Bohannon 1999). The control strategy of the trunk is animportant factor in achieving smooth locomotion and maintaining body equilib-rium. It is considered that trunk stabilization is a dynamic process whereby stableequilibrium of the trunk in space is maintained (MacKinnon and Winter 1993;Winter 1995). When perturbed, the trunk stabilization process enables the trunk toreturn to its state of rest or to normalize motion and maintain trunk stability inspace.

More recently, the characteristics of the lower extremities when negotiatingobstacles have been analyzed thoroughly. However, the movements of the trunkand the pelvis in the process of obstacle adaptation have not been focused on ex-clusively. Therefore, the purpose of this study was to investigate the 3D movementpattern of the trunk and pelvis during gait and the influence of obstacle height onthese parameters. It is thought that the results might be useful in understanding themaintenance of dynamic stability in obstacle crossing, and might give some in-sights into the nature of falls.

MethodSubjects

Seven healthy subjects (all men) participated in the study (age: 22.0 ± 1.2 years;height: 1.73 ± 0.05 m; mass: 69.1 ± 7.9 kg). Prior to testing, the purpose and pro-cedure of the experiment were explained to the subjects and their written informed


consent was obtained. Subjects were free of a history of skeletal and/or muscularinjury. All of them were right foot dominant.

Instrumentation

The video-based WINanalyze 3D motion measurement system (Mikromak, Ber-lin, Germany) was used to measure the three-dimensional (3D) angular movementsof the trunk of each subject. Three high-quality video cameras (TK-1270, Victor,Tokyo, Japan) with a standard sampling frequency of 60 Hz were used to trackshoulder and pelvis movements in the frontal and horizontal planes, and trunkmovements in the frontal and saggittal planes during gait. Retro-reflective mark-ers, 20 mm in diameter, were attached to the left and right acromion process, theleft and right anterior superior iliac spine (ASIS), the sternum, and the greatertrochanter of the left hip. Additional markers were placed on the feet over thecalcaneal tuberosities to identify footfall events and define complete strides.

The video images were stored on videotapes (SVHS, Sony, Japan) via threevideo camera recorders (NV-CX1 SVHS, Panasonic, Osaka, Japan), which wereinterfaced with three video monitors (TM-100S color TV, Victor, Tokyo, Japan)for the instant qualitative evaluation of the video recordings. The video data weretransformed to digital format and digitized via the WINanalyze automatic motionanalysis system.

Protocol

Subjects walked barefoot on a walkway measuring about 10 m in length and 1 m inwidth. The subjects were allowed several practice trials to become accustomed tothe instrumentation and environment. The experimental protocol included walkingon level ground with no obstructions and stepping over obstacles corresponding to5%, 10%, and 15% of each subject’s height.

The subjects were instructed to walk along walkway, step over the obstaclewith their right foot first and left foot second, and then continue walking to the endof the walkway. They were asked, irrespective of the height of the obstacle, toaccommodate to the obstacles in a natural manner without abrupt adjustments tothe gait pattern or any other unusual adaptations. Also, the subjects were asked tofix their gaze at eye level beyond the walkway.

The obstacle heights were established based on pilot work and previous litera-ture (Chou, Kaufman, Robert, et al. 2001). All obstacles, constructed from wood,were 450 mm in width and 20 mm in depth. They were placed in the center of thewalkway. The heights of the obstacles were randomly selected for each trial. Threetrials were performed for each task.

Measurement Coordinate System

The global reference frame was defined with respect to the laboratory. The X-axiscorresponded to the line of progression of the subject across the walkway and was


positive in that direction. The Y-axis was positive to the subject’s left and the posi-tive Z-axis was upward vertical (Figure 1).

In the frontal plane, angular rotation of the shoulder and pelvic segments wasmeasured with respect to the horizontal line passing through the right acromionand right trochanter, respectively. Rotations of the segments with the left acromionand left trochanter below the horizontal line were defined as positive. The sternumand the point midway between the two ASIS (anterior superior iliac spine) wereused to create the trunk segment. Trunk angular displacement was defined relativeto the vertical line passing through the point midway of the two ASIS. Rotations ofthe trunk segment with the sternum towards the right of this line were defined aspositive. In the horizontal plane, the shoulder and pelvic angular displacementswere measured relative to the Y-axis. Forward rotations of these segments aboutthe right shoulder and right ASIS were defined as positive. In the sagittal plane,trunk angular rotations were measured with respect to the vertical and rotationsanterior to the vertical line were defined as positive (Figure 2).

Figure 1. Coordinate frame and reflective marker set-up used in this study. Pitch, roll,and yaw rotations are defined as the rotations for Y-axis, Z-axis and X-axis, respectively.


Data Analysis

The raw data of the marker positions were filtered with a fourth-order low passButterworth digital filter at a cut-off frequency of 8 Hz.

Figure 2. Segment and joint angle definitions are illustrated for the frontal plane (A),horizontal plane (B), and sagittal plane (C). Figure adapted from Bartonek, Saraste,Eriksson, et al. (2002). Reflective markers were placed on the left and right acromionprocess, the left and right anterior superior iliac spine (ASIS), the sternum, and thegreater trochanter of the left hip. For every segment and joint angle, the arrow directionindicates positive angles from neutral.


The gait cycle was determined based on the vertical translation of the heelmarker. The crossing stride was defined as beginning with the heel contact of theleft foot before stepping over the obstacle and ending with the heel contact of theleft foot after crossing over the obstacle. Gait speed was calculated for each trial bydividing the absolute stride length (the horizontal distance from heel contact toheel contact) by the respective stride time. Stride cycle durations were standard-ized as 100% in order to compare the patterns of different subjects, for differentobstacle heights.

Descriptive statistics were calculated for all anthropometric and kinematicparameters for the entire sample and for the different height obstacles. Peak-to-peak amplitudes (a range from maximum peak to minimum peak during the gaitcycle) of angular displacement in three orthogonal directions were made. One-wayrepeated-measure ANOVAs were carried out for the effect of obstacle height on thetemporal distance parameters and peak-to-peak amplitude of the angular move-ments. If a significant difference was detected, the polynomial test was performedat an α = 0.05 level of significance to determine the nature of the trend (linear,quadratic, or cubic). Pearson’s linear correlation coefficients were used to analyzethe relationships between the shoulder and the pelvic angular rotation in the sameplane. p values of less than 0.05 (two-tailed) were considered statistically signifi-cant. SPSS Version 10.0 was used for all statistical analyses (SPSS, Chicago).

ResultsBasic Characteristics of Gait During the Crossing Stride

The mean values of gait velocity and the associated variables of stride time andstride length/body height in the crossing stride are presented in Table 1. These dataare consistent with previous research documenting the gait of a healthy male (Chouand Draganich 1998). They suggest that the subjects of this study can be consid-ered typical of their age group in the characteristics of their gait.

During the crossing stride, gait velocities of all the subjects decreased linearly(p = 0.009) as the obstacle height increased (Table 1). Stride times during the crossingstride increased linearly (p < 0.003) as the obstacle height increased. The obstacleheight significantly affected the stride length (normalized with respect to the sub-ject height; p = 0.032). There were no significant differences between the meanstride lengths with the obstacles. However, the crossing stride length was signifi-cantly greater than the stride length in the case of unobstructed walking (p < 0.05).

Sagittal Plane Kinematics

In the sagittal plane, the trunk segment showed a movement pattern that oscillatedtwice within the stride cycle. This movement pattern was similar both for unob-structed level walking and in negotiating obstacles. The amplitudes of the trunk


Tabl

e 1

Gai

t Tem

pora

l-Dis

tanc

e M

easu

rem

ents

Dur

ing

the

Cro

ssin

g St

ride

Obs

tacl

e hei

ght

Non

e5%

10%

15%

p va

lues

Gai

t vel

ocity

(m/s

)1.

522

± 0.

100

1.46

3 ±

0.13

01.

427

± 0.

147

1.34

6 ±

0.15

9p

= 0.

009

Strid

e tim

e (s

)0.

984

± 0.

083

1.09

6 ±

0.06

11.

129

± 0.

086

1.19

4 ±

0.11

3p

< 0.

001

Strid

e len

gth/

body

hei

ght

0.87

8 ±

0.02

70.

925

± 0.

056*

0.92

9 ±

0.04

8*0.

922

± 0.

046*

p =

0.03

2

Not

es: p

val

ue w

as fo

r the

AN

OVA

with

repe

ated

mea

sure

s. *

Sign

ifica

ntly

diff

eren

t fro

m u

nobs

truct

ed w

alki

ng (B

onfe

rron

i, p

< 0.

05)

225


motion increased linearly (p = 0.017) with the increases in obstacle height. Thepeak-to-peak amplitudes of the trunk segment are shown in Figure 3.

Figure 3. Trunk pitch rotations in the sagittal plane during the crossing stride. A. Peak-to-peak amplitudes increased linearly (p = 0.017) as the obstacle heights increased. (SDbetween brackets.) B. Typical angular displacements for the trunk in the sagittal planeduring unobstructed level walking and when stepping over obstacles of different heights.

A.

B.


Frontal Plane Kinematics

The shoulder, pelvic, and trunk motions in the frontal plane demonstrated consis-tent patterns within and between participants while the magnitudes of the rotationsdiffered when stepping over obstacles of different heights. The shoulder, pelvic,and trunk movement amplitudes are shown in Figure 4.

Figure 4. The shoulder, pelvis, and trunk yaw rotations in the frontal plane during thecrossing stride as indicated. (A) Peak-to-peak amplitudes of the shoulder increasedlinearly (p = 0.036) as the obstacle height increased. (B) Peak-to-peak amplitudes of thepelvis showed no significant difference (p = 0.222) during “unobstructed level walking”and when stepping over obstacles. (C) Peak-to-peak amplitudes of the trunk increasedlinearly (p = 0.015) when negotiating higher obstacles. (SD between brackets.)


There was a linear (p = 0.036) trend toward increased angular movements ofthe shoulder with the progressive increases in obstacle height. In contrast to this,the amplitudes of pelvic yaw rotation were not significantly different when step-ping over obstacles (p = 0.222).

The peak-to-peak amplitudes of the trunk increased linearly (p = 0.015) as theheight of the obstacle increased. The greatest displacements of the trunk were foundwhen stepping over obstacles measuring 15% of body height.

Horizontal Plane Kinematics

Peak-to-peak amplitudes of the shoulder and pelvic angular movements in the hori-zontal plane can be seen in Figure 5. The amplitude of the shoulder and pelvicrotations were greater than those of the frontal plane. During unobstructed levelwalking, the average amplitude of the shoulder’s roll rotations was 8.4°. For step-ping over obstacles, there was a linear (p = 0.013) trend of increased angular dis-placements with the increase in the obstacle height. It reached 15.0° when the ob-stacle height was 15% of body height. For the pelvic segment, peak-to-peak dis-placements had a linear (p < 0.001) trend of increasing as the obstacle height in-creased. There were no significant differences between the amplitudes of the pel-vic segment when stepping over obstacles of different heights. However, the am-plitudes of the pelvic segment were significantly different (p < 0.05) between the“unobstructed walking” and “obstacle crossing walking.” In addition, the peak-to-peak amplitudes of the pelvic segment showed a significant correlation (r = 0.633,p < 0.01) with the stride length.

Coordinative Characteristics of Trunk and Pelvis

The temporal and directional features of rotations in the frontal and horizontalplanes were similar for all the subjects. The data were averaged and pooled to-gether. Table 2 shows the coordinated characteristics of the shoulder and the trunkduring “unobstructed level walking” and “obstacle crossing walking” in the frontaland horizontal planes. A positive correlation coefficient indicates that the shoulderand pelvic segments move in the same direction with synchrony. On the other hand,a negative correlation coefficient indicates that the movements, although synchro-nous, go in opposite directions.

In the frontal plane, shoulder and pelvic movements went in opposite direc-tions during most of the gait cycle, as shown by the negative values of the correla-tion coefficients. Furthermore, the correlation coefficients revealed the degree ofsynchrony in the two segments. The coefficients had a decreasing trend as the ob-stacles increased in height.

In the horizontal plane, the shoulder movements were negatively correlatedwith the pelvis. Greater correlation coefficients were obtained in the unobstructedlevel walking. It also was found that the correlation coefficients decreased as theheight of the obstacle increased.


Figure 5. The shoulder and pelvic roll rotations in the horizontal plane during thecrossing stride as indicated. (A) Peak-to-peak amplitudes of the shoulder increasedlinearly (p = 0.013) as the obstacle height increased. (B) Peak-to-peak amplitudes of thepelvis had a linear (p < 0.001) increase trend as the obstacle height increased. There wereno significant differences between the amplitudes of the pelvic segment when steppingover obstacles of different heights. However, the amplitudes of the pelvic segment weresignificantly different (p < 0.05) between “unobstructed walking” and “obstacle crossingwalking.” (SD between brackets.)

DiscussionSeveral studies have analyzed the effect of obstacles of different heights duringgait. Most of these studies have focused on the kinematic characteristics of thelower extremities when stepping over obstacles. However, maintaining the dynamicstability in obstructed gait requires not only appropriate modification of the lowerextremities but also adequate intersegmental coordination of the upper body. There-


fore, it is important to understand the effect of obstacle height on the motions of thetrunk and pelvis while negotiating obstacles.

Trunk Adaptations to Obstacle Changes

The results showed a significant effect of obstacle height on peak-to-peak ampli-tudes of the trunk in the frontal plane. The peak-to-peak amplitudes of the trunkincreased significantly with the increases in obstacle height. This finding indicatesthat the higher the obstacle, the more the lateral trunk oscillates; therefore, it is inagreement with the findings of the study on the body’s center of mass (Chou,Kaufman, Robert, et al. 2001). This lateral inclination of the trunk can be explainedby the motor control theory. The basic tasks of motor control are the maintenanceof equilibrium and the adaptation to the environment. To orient the body with re-spect to the external world, the trunk flexing toward the weight-bearing side due tothe swinging limb perturbs the stability of the whole body. As the obstacle heightincreased, an excessive yaw rotation of the trunk counterbalanced the swinginglimbs.

During gait, the body’s center of mass is located outside of the base of supportduring most of each step cycle. However, the forward leaning of the body is care-fully controlled to keep the center of mass within the projected base of support ofthe feet (Winter 1995). When stepping over obstacles, this control becomes moreimportant, as the possibility of trips and falls is high. The results of this studyshowed that, in the sagittal plane, the amplitudes of the trunk motion increasedsignificantly as the obstacle height increased. This finding suggested that the upperbody’s forward leaning improves the stability of the body and may also be a com-pensatory adjustment for negotiating higher obstacles.

Because the greatest body mass is contained in the trunk, trunk stabilization isconsidered a major controlling variable in postural orientation and equilibrium.The results of this study indicate that, when stepping over obstacles, the stability ofthe trunk can be achieved by increasing forward and lateral trunk movement, andthat this adjustment seems to help the body to minimize energy expenditure andincrease the efficiency of walking.

Table 2Correlation Coefficients Between Shoulder and Pelvis

in the Frontal and Horizontal Planes

0% 5% 10% 15%

Frontal plane –0.810 –0.790 –0.761 –0.751Horizontal plane –0.919 –0.812 –0.742 –0.734


Pelvic Adaptations to Obstacle Changes

Recent experimental studies on gait have demonstrated that obstructed gait re-quired adaptations of the lower extremities (Austin, Garrett, and Bohannon 1999;Chou and Draganich 1998). Results from the present study suggest that the modi-fications of pelvic orientation on the frontal plane are also necessary for the adap-tation process. The key mechanism for negotiating obstacles is to elevate the swingfoot by simultaneously increasing the hip and knee flexion in the same limb(McFadyen and Winter 1991). For greater elevation of the swing limb, the pelvicyaw (yawing toward the swing side) was constrained. For this reason, the ampli-tudes of pelvic yaw rotation showed no significant difference with changes in theobstacle height. These gait adaptation processes may guarantee a successful con-trol strategy to maintain the balance of the pelvis. It is interesting to note that dy-namic stability is achieved by adequate muscle strength, appropriate neuromuscu-lar timing, and free passive joint mobility. For elderly people, a decline in muscularstrength occurs. It also is likely that the elderly experience a limited range of mo-tion and joint impairment (Lamoureux, Sparrow, Murphy, et al. 2002). Dynamicwalking balance, such as stepping over obstacles, may challenge the elderly andtherefore be a factor in causing falls.

An important observation from this study is that amplitudes of pelvic roll rota-tion were not significantly affected by obstacle height. It was found that at veloci-ties above 0.75 m/s, pelvic rotation starts to contribute to a lengthening of the stride(Crosbie, Vachalathiti, and Smith 1997). Pelvic rotation is more directly associatedwith the motion of the lower extremities. As the crossing stride length remainedinvariant with changes in obstacle height, the amplitudes of pelvic roll rotation didnot change.

Coordinative Characteristics of the Trunk and Pelvis

One of the major motor functions of human gait is to modify the upper body. Upperbody stability can be accomplished through passive transmission of forces throughthe joints and facilitated by muscles that act across the joints to produce rotation(Thomas, Karen, and Steven 1997). Hinrichs (1990) has demonstrated that thearms swung in opposition to the lower extremities and created movement at theshoulders that opposed pelvic motion during gait. In the present study, shoulderand pelvic movements went in opposite directions during most of the gait cyclewhen stepping over obstacles. The coefficients, which revealed the degree of syn-chrony between the shoulder and the pelvis, showed a decreasing trend with in-creases in the obstacle height. Higher obstacles place more demands on the coordi-nation of both segments.

It is interesting to note the different transitional patterns of the shoulder andthe pelvis with respect to the peak-to-peak amplitude of rotation, as the height ofthe obstacle increases (Figure 6). The peak-to-peak amplitude of shoulder move-ments increased linearly with the height of the obstacle, whereas the pelvic move-


ment did not change with the height of the obstacle. This increased amplitude ofshoulder movement suggests that people tend to move their upper limbs to a greaterextent when negotiating higher obstacles. As the hip and knee flexion increasedwith the increase in obstacle height, it seems that the increased movement of theupper limbs is a reaction to a greater movement of the contralateral leg, which ismeant to maintain dynamic stability. As for the characteristics of pelvic movement,it seems that the invariability of the pelvis is the key motor adaptation for success-fully stepping over obstacles. This kind of adaptation may be related to the safetyand efficiency of the body locomotor system.

ConclusionThe results of the present study indicate that stepping over obstacles poses a sig-nificant challenge to upper body stabilization. Pelvic rotation is important in modi-fying the crossing stride and clearing an obstacle. The shoulder plays a role, incoordination with the pelvis, in maintaining dynamic equilibrium when steppingover obstacles. Dynamic walking balance is achieved by integrating sensory inputfrom the visual, vestibular, and proprioceptive systems (Prince, Corriveau, Hebert,et al. 1997) with adequate muscle strength, appropriate neuromuscular timing, andfree passive joint mobility. Further study of the effects of vision on postural stabil-ity might be required.

Figure 6. Different transitional pattern of the shoulder and pelvis on peak-to-peakamplitudes of rotation with increasing obstacle height.


ReferencesAustin GP, Garrett GE, Bohannon RW (1999) Kinematic analysis of obstacle clearance

during locomotion. Gait and Posture 10:109–120.Bartonek A, Saraste H, Eriksson M, Knutson L, Cresswell AG (2002) Upper body move-

ment during walking in children with lumbo-sacral myelomeningocele. Gait and Pos-ture 15:120–129.

Campbell AJ, Borrie MJ, Spears GF, Jackson SL, Brown JS, Fitzgerald JL (1990) Circum-stances and consequences of falls experienced by a community population 70 yearsand over during a prospective study. Age and Ageing 19:136–141.

Campbell AJ, Reinken J, Allan BC, Martinez GS (1981) Falls in old age: A study of fre-quency and related clinical factors. Age and Aging 10:264–270.

Cappozzo A, Figura F, Leo T, Marchetti M (1978) Biomechanics. Baltimore: UniversityPark Press. p 272.

Chou LS, Draganich LF (1998) Placing the trailing foot closer to an obstacle reduces flex-ion of the hip, knee, and ankle to increase the risk of tripping. Journal of Biomechan-ics 31:685–691.

Chou LS, Kaufman KR, Robert HB, Draganich LF (2001) Motion of the whole body’scenter of mass when stepping over obstacles of different heights. Gait and Posture13:17–26.

Crosbie J, Vachalathiti R, Smith R (1997) Patterns of spinal motion during walking. Gaitand Posture 5:6–12.

Hinrichs RN (1990) Multiple Muscle Systems. New York: John Wiley. p 694–705.Lamoureux EL, Sparrow WA, Murphy A, Newton RU (2002) The relationship between

lower body strength and obstructed gait in community-dwelling older adults. Journalof American Geriatrics Society 50(3):468–473.

MacKinnon CD, Winter DA (1993) Control of whole body balance in the frontal planeduring human walking. Journal of Biomechanics 26:633–644.

McFadyen BJ, Winter DA (1991) Anticipatory locomotor adjustments during obstructedhuman walking. Neuroscience Research Community 9:37–44.

Patla AE, Prentice S (1995) The role of active forces and intersegmental dynamics in thecontrol of limb trajectory over obstacles during locomotion in humans. ExperimentalBrain Research 106:499–504.

Prince F, Corriveau H, Hebert R, Winter DA (1997) Gait in the elderly. Gait and Posture5:128–135.

Pryde KM, Roy EA, Patla AE (1997) Age-related trends in locomotor ability and obstacleavoidance. Human Movement Science 16:507–516.

Saunders JBM, Inman VT, Ebernardt HD (1953) The major determinants in normal andpathological gait. Journal of Bone and Joint Surgery 35(A):543–558.

Thomas MK, Karen LS, Steven JS (1997) Relative contributions of the lower extremityjoint moments to forward progression and support during gait. Gait and Posture6:1–8.

Tinetti ME, Speechley M (1989) Prevention of falls among the elderly. New England Jour-nal of Medicine 320:1055–1059.

Winter DA (1995) Human balance and posture control during standing and walking. Gaitand Posture 3:193–224.

angular movements of the trunk and pelvis when stepping over obstacles of different heights

Documents