Mechanical energy of walking of stroke patients

92

Mechanical

Energy of Walking of Stroke Patients

Sandra J’. Olney, PhD, Trilok N. Monga, MD, Patrick A. Costigan, Faulty

qf Merlic~ir~e.

Quetw’s

Uttiversi~.

k’in,yston.

Onttrrirt.

Cmcttlrr

MSc

K7L

3Nh

ABSTRACT. Olney SJ, Monga TN, Costigan PA: Mechanical energy of walking of stroke patients. Arch Phys Med Rehabil 67:92-98, 1986. l The mechanical energy costs of walking have been studied in ten stroke patients with hemiplegia. A two-dimensional sagittal plane cinematographic analysis of two strides of the subjects’ normal walking was undertaken, yielding continuous information about the mechanical energy costs of the whole body and each of its parts, about the energy types involved, and the amounts of energy conservation. The large head, arms. and trunk (HAT) were found to dominate the total pattern. Three major disturbances were seen. In contrast to normal subjects who show energyconserving negatively correlated potential and kinetic energy curves for the HAT, the subjects who demonstrated the first disturbance showed gross irregularity of the curves, with almost no opportunity for exchange between energy types. In a second disturbance the curves of the HAT showed some energy-conserving portions, but levels of kinetic energy curves were low, resulting in little energy exchange. In the third disturbance, some exchange was evident, but the pattern was dominated by potential energy changes in the form of a single large rise and fall, coinciding with swing phase of the affected leg. Each of these disturbances would require a different approach to treatment. Although mechanical energy analyses do not reflect certain energy costs, such as the effort required to hold the body up against the pull of gravity and that used in contracting antagonist muscles. they could be of considerable assistance in pinpointing costly variations in energy patterns during walking and in determining appropriate treatment procedures.

For many

stroke

patient.\

ing is an important duration

of daily

ficiencia

cannot

using

incidents

during

energy

custs.

appropriate merits.

that the me&mica1

to pinpoint the gait This methods

of

that could

by them.

inef-

atridc

of

Thcsc

technique

Modellcd

anthropmustric

lht

and by

findings.

be used

could

fhr

Ill$h,

+

to dctc‘rmine as linked

+

segments. cycle

energy

segthe ini.

‘~,l,bJ(2

Icvels

tn, is the ITLISS of the scgmcnt.

proportion

of

the total

constant:

h, ix the

reference

point:

height

shanks.

two

head.

energy

The total at each

kinetic

or patcntial

&crease

has occurred.

arms.

cost

conservation

adjacent exert

forcc~

ot’ one

it is aswrncd ccmscrvinp

rychangrs.

ccgnit‘nts

it\

segment

while

is also exchanged

ways. I” First.

neighboring

each

segment

and reaching

Arch Phys Med Rehabil Vol 67, February 1966

to as

between

body and

xcgment ;f

in fig

in citha

are ret’crrrd

eschangc~

subject\.

in l’ig 2. tlere

;I

thcrt:

is

over

the paid

ths

full

ot

bctwcen cycle:

that occur

I and

m

the kinctic

the

during

tt’et.

segnlent

cncr-gy

total

t I J;kg.m’J

for

The

attributed

X0% oi‘ the total cnurgy

curve

below ot’the

energy

peaking

levels

rellccting

the total two

of body

normal

values

I’m-

.h The nori\ ;I cloublc

made

hy c;lch

to the legs encrpy

cost

of the HAT

top ot‘ fig 2. The potential

at the

its lowest

total

history

contributions

changes

to be about ”

ot’ 66.97%

stride

two

comprised

(HAT)-the

an average

Energy

di\playcd

energy.

to he

the large

a rnirrormg

the potential

segmcntx-two

together

body

Icg\.

been estimatcc!

components

i\

ot‘ each of the

cost over

and fall\

scvcn

as

reported

I ). reflecting

of masx:

ruchangc

in time

reported

of the total

(fis

of’ mass:

there

at each instant

mcrgy

The only_ study- that has included

hump

with

is an increase:

that an energy

Enqy

in two on

there

the amount

to the energy

burn of the tliffcrences

trunk

mal shupc

shown

the cater

walking.

of the body

energy

in

tendon-scginunt

is equal

mcl one’ combined

has been

energy

ahovc

of the center

level

total

ttw

body

instant

and

subjects.

in normal

of the segment.

of ;ime

rner~y

Thrsr

within-segment GIII

velocit\’

in the other.

velocitv

of

is modelled thighs.

walking

have

SOIW

xum

transItit rcsults

the gait cycle.

gravitational

01. IWM

of incr-tia about

a small period

it to be ;1 certain

9 is the

r>i its center

momrnt

and w, i> the ungular

assuming

weight”;

V, is the linear

I, ix the rotational

Ii’ over

body

at the

sped

‘I’hc total cncrgy

cncrgy

~LISL~IC activity

energy

constant

is the absolute

of the swinging where

af

cost ih the SLIIII ol‘ ail the rim

total

the

or part of’ the body.

’ ?111,\.,2

body ;1 gait

IS a m~chmlcai

Second.

by all the muscles

is the algebraic

If the body

high

is’-‘: E, =

in time

and specific

information.

E. 01‘ one qmnt

time

gencratetl

are responsible

treatment.

knoun

energy,

at any given

cycle

and during abhorbed

ot‘ cm

energy

absorption

interface’”

highcr

parts of the body

information

and using

stantaneous

certain

and

ot‘ energy

the specific

the general

sqrncnts.

generation

but global met-

thm

II~~~V~‘INXI.

djscont

have shown

by Winter’

aiiaIyscs

energy

sub_jocts supported

also suggest

bc exploited

to pinpnnt

betwcrn

each has ad-

oi‘ studies

of studies

translatitmal

been a~se~d

arc considerably

bc ud Kcsulta

mechanical

hcmiplegic

studies

have

01‘ walk-

the type and

techniques:

A number

costs of walking

in the gait.

l,owcry(’

cods

and nxxhanical

cmt.

limits

subjects than for normal subjects’,’

abolic ~neasures

five

Energy

disadvantages.

that metabolic hemiplepic

or enqy

and frequently

activities.

using both metabolic vantages and

the rf‘titrt.

concern

HAT

components

at midstance

at each heel strike. the xpecd of the body.

is

and

kinetic

time

histoq

is obviom. of each

foot

In contruht. peak> near

MECHANICAL r erGy

co”‘!

! butcons crc!

from

ENERGY,

93

Olney

tars.:

legs

co( $

Klnetlc

(?r

-0tal PoteMal nslc~tionai)

w 39 energy energy energy

+ * C-

-I

iJpt_J( Tfme,

rise’:

pig 2---ln&ntaneous total energ) ol’ thlc tl ,A I’ segment in normal \ralking shown with its potential and translational kinetic components. Keprinted with permission from ‘% inter, Quanhury and Reimer. 1076

Arch Phys Med Rehabil Vol67. February 1986

MECHANICAL ENERGY, Olney

94

Table 1: Gait Cycle, Stride Length and Velocity of Two Strides for Each Subject Subject code QCl2

Age. yr h6

Seu 1:

Side of paresis right

Time since stroke 2yr

I 11111)

Gait cycle, s

Stride length, m

3.x

0.3X

: .h QO4

7.1

M

kit

14,. hlnr)

.1Y

I h

Y)J

I h

YY .25

QOS

hl

F

ryht

-!?I- t OIll~l

I .(I I .1

.2(1

QOh

55

M

rqht

t yr 3mo

37

.-IX

QO7

77

M

Iright

?>I

QOX

h-l

bl

right

“1 I

yrw

hi

M

right

5) r

QI’

43

bl

right

JIllO

YIIIO

7n1(1

2.6

.37

I .1

-17

I 4

.57

2.4

.5l

1.X

37

I Y

.hO

I 4

.60

I .1

.XX Y\

I .h 01.5

Qt7

51 3

M M

right left

using a fourth order, zero lag. low-pass Butterworth filter cutting off at 5Hz, according to the methods validated by Pezzack’s group.” The body was modelled as seven segments: two feet. two shanks, two thighs and an HAT segment. For each segment the necessary anthropometric constants were obtained from Dempster’s work, and based on the subject’s height and mass.i Using computer programs described by Winter,‘Jm’” the data were analyzed to yield kinematic and mechanical energy information about each segment and the total body for each instant in time corresponding to tine frames.

Fig 3-Positions of reflective bodg markers and protractor used in identifying landmarks for use in seven segment link segment model. Arch Phys

Med Rehabil Vol67, February 1986

h) r Imo 7yr

IOmo

I.3

I .IM

I.3

I .I)‘1

I .J

I ,ClU

I.5

I 06

RESULTS The stride characteristics for each subject appear in table 1. Subjects QO2. QOS. Q06. and QOS were the slowest walkers, with speeds not exceeding 0.3m/>. Subjects Q07 and QO9 demonstrated moderate speeds, between 0.3 1 and 0.4 I m/s. The fastest walkers. subjects QOJ. Q17. QIS and QI 7 obtained speeds in the vicinity of O.&O.&n/s. Mechanical energy costs. expressed in joules/kilogram of body massimetcr. are summarized in table 7 with associated kinematic information. Total costs ranged from 0.X.Vkg.m for subject Q07 to 3.931 kg.m for subject QO.5. The percentage of total energy conservation was low for all subjects. with a median of 48% and a range from 21% to 67%. Within-segment exchanges ranged from II% to 40% (median 29% ). Between-segment conservation showed values from 10% to 25% (median 15%‘). For all subjects the shape of the total body curve resembled closely that of the HAT segment. This marked contrast with normai subjects wah particularly obviou:, for slow-walking subjects such as subject QOS (fig 4 top) but was also apparent in the energy histories of one of the best performers, subject QI? (fig 4 bottom). The energy levels for the leg segments were low in magnitude but their patterns followed the expected shapes. For subjects Q04, QO9. Q12, Ql5. and 417 this was the major abnormality. though other subtler differences from normal could also be seen. Examination of the HAT curves with energy types shown separately (figs 5 to 7) revealed three major disturbances in the hemiplegic subjects. The first disturbance was characterized by the absence of mirroring in the shapes of the kinetic energy and potential energy curves. The characteristic of this disturbance was gross irregularity of the kinetic and potential energy curves (fig 5). With the exception of one brief period at midcycle and another between 95% and 100% of the cycle, no within-segment exchanges occurred. A second disturbance (fig 6) showed that there was at least some mirroring of kinetic and potential energy patterns but the

MECHANICAL ENERGY, Olney Table 2: Summary of Energy Costs and Kinematic Data for Two Strides of Each Subject*

SUhjtd rode _ _

\I wk. distance, Jlkg./m

stride.

.Jk.

~~.._.~

DISC’USSION tIcc;lu\c mechanical

cncrgy analyst3

t*reat deal 111rehabilitation. cc Ilmltationx

ot the nethod

and planning

3warcnt‘ss is important

have not hwn

used 2

of the strengths in interpreting

md

results

treatment procedures. This ttxhnique permltz of precise sources of high energy co\ts. locatc5 the pcriotl v. hen they OCCUI in the gait cyc.1~. indicates the \c‘grncntt \ 1 I/W arc rc~p~msibie and the txpc‘ 01’ cncrg! in-

idcntit‘icatioil

95

MECHANICAL

96

ENERGY,

Olney

No findings

of the present study contlict

of hemiplegic

gait.

of the previously example.

low

although

obvious

of calf activity

430

g

410

on the attainment

;

400

cause within-scgmcnt

5

60

One ol‘ the most important cmtratrd

F

LOWER RIGHT -___------

t

I

I

0

Q07.

I

I

I

40

20

I

60

PERCENT

OF GAIT

I

I

80

I 100

CYCLE

: J

comi’ortablt:

F

2 650 7) 1 640 &

index

termed an cncrgy

of

tl

W

90

I

I

I

I

I

LOWER

L

I

I

I

I

speed.

walking with

” Using an

coct‘t’icient to ;I\SCSS the

potential

and kinetic

and coworkers7

demonstrated

segments

of one subject.

though the muits

cncrgy.

this trend t’or- the torso fi)r the total bodl

plotting the index V;I~LIC against the sell’of zevcral subjects

pathologies.

none of whom

conservation

varied directly

relationship

when

and to increase

in walking

correlation

exchange

Metabolic costs

to be lowest

I .?mis.

about

or a decrease

Alxo.

wac subject conservation

bctwcen total ntct-

Mansour

The

oz 620

z

bpeed.

an incrrusc

energy costs.

630

bc-

in \ery IOU total energy cmts.

selected speeds of walking

H. A. T.

Also.

was usually

An cxccption

hi_ch between-segment

tither

were equivocal.

765

energy con\crc.ation velozitich.

costs and the speed of walking:

within-qgnent

775

levels.

in normal subjects have been shown at one’s

785

in thl\ study ih the den-

is ;I well documented rclationshlp

abolic cnrrgy

of 3

the ir;tluence

cxchangcs.

which resulted

There

effects

exchanges were air\ :LV~ ccmsidcrably largel

who had unusuallv

values. I

tindingx

of ndcquatc wfalhing

seen in total conservation

LEFT

have

generating muscle group are

dependcncc of within-scgmcnt

than between-segment

LIMBS

Fol-

patients

the present analysi?.

2 -3 ,

420

in stroke

and the serious

reportsi.’

in this most important

trorn

atudica

the implicuticm

identit‘icd feature5 have beconic clear.

Icvcls

been noted in many deficiency

with ibrrncr

in many cam

reached nor&l

with m~~vemcnt speed Icvcls,

the

with speed for the torso segments.

between speeds and the total mcchmical

however,

on the mechanical energy

is not well

docurnentcd.

expenditure

In ;I report”

of one hemiplogic

sub-

LIMBS

80 70 60

0

40

20 PERCENT

60 OF GAIT

80

100

CYCLE

Fig S-Illustration of close resemblance of shape of total body energy curve to that of HAT segment found in all hemiplegic subjects. Curves of best performing subject (Ql2) are shown bottom figure and of poorer walking (QO8) in top figure.

i

Lowery’?

E

reported average of 55’%. The range ot‘ the present

study (X%

to 67%)

to he between 49%

was also greater than the former. and 58%.

This

might

the greater number of subjects and trials

370

F

POTENTIAL

lfLryz

reported

bc expected from in the present

included

study. With cry’:,

regard to curve shapes,

study

QOC,. Ql?.

would QIS.

body curves similar but without in this study.

the subject reported in Low-

have been categorized and Ql7

of the present

with study.

in shape to the dominating

some of the more serious

The patterns were very similar

February

1966

Q0-I.

having total

HAT

abnormalities

QO’I.

Arch Phys Med Rehabil Vol67,

subjects

segment. identified

to those of subject

0

20 PERCENT

40

60 OF GAIT

80

100

CYCLE

Fig 5-Instantaneous energy curves of the HAT segment with its potential and translational kinetic energy components for subject Q06, showing absence of normal mirroring and irregularity of translational kinetic and potential components.

MECHANICAL

440

z

r.

97

TRANSLATIONAL

-’ ;. ‘5

Obey

TOTAL

>!‘r

LLI

ENERGY,

k I

0 i%1-i

i’ri

PERCENT

4 c!

60

OF GAIT

SC

II?!?

CYCLE

E‘ig f+--Instantaneous energ:? c’urvcs of the HAT segment for suh,jecl Q02. showing some mirroring of potential and kinetic energ! hetwecn 7”r and 30“+ of the cycle and again between 90% and 100~ . I .evels of kinetic energ! are ION.

PERCENT

OF GA/‘-

ZYCLE

Fig 7-Instantaneous energy curves of the H 41‘ secment for suhjcct QM. Potential energy changes d(nnrn;cte and show a hiphiking pattern.

Arch

Phys

Med

Rehabil

Vol 67, February

1966

MECHANICAL

98

conserving exchange between kinetic and potential energy of the body segment and trunk segment. 3. Three abnormal patterns of energy costs could be identified: The first showed no mirroring of the potential and kinetic energy types; the second showed limited mirroring but with low kinetic energies so that energy-conserving exchanges were minimal: the third disturbance. the hip-hiking pattern. was characterized by low kinetic energies and a marked increase in HAT potential energy during the swing phase of the affected leg. Each of the identified disturbances would require a different approach to treatment to restore the conserving exchanges between potential and kinetic energiex. This research indicates that mechanical energy analyses of hemiplegic patients could be valuable in pinpointing specific causes of high energy costs of walking and to assist in determining treatment approaches that specifically address the deficiencies.

ENERGY,

Olney

Mansour

7 f

JM.

Lcsh

MD.

Gond multi-scgniental ological

x.

Nowah

analyk

locomotion.

9. Pezzad

Winter

Arch

DA.

AO.

and power tlou ment Studio

7X3.

Stein&

‘1‘. Gra-

analy\ih of gait: hcmlplegic 57:12

DA:

1425.

1Y76

Awzsment

Winter

DA.

Rcimcr

GD:

in body wgmc‘nts during

1:5%h7.

ot‘dcrivatiw J Biomechanics

Instantaneous

walking.

power

J Hum

Mow-

I Y75

Int Z Angcw

relation anil optimal Physiol

kinschl

xpccd Jurinf

Arbcitsphysiol

Icvcl

17:277-

195x

Robertwn sorption

DG,

Winter

and trand’cr

mechanics 13.

I’s82

AO.

used for motion analysis.

HJ: Energy-qwt_d

walking.

Jimcn-

I977

Quanhury

Ralston

Three

15:s I-50.

Mcd Rehabil Winter

SK:

of normal and path-

Quanbury

tcmpord

KW.

tcchniqucs

10:177-3X’.

11.

Phys

JC. Norman

determining

11.

Simon

human gait. J Biomcchanics

Peat M. Duho HIC.

hame R: Electromyofraphic

IO.

MD.

of cnergctics

Williams

DA:

13:X35451. KR.

Mcchanlcal

amongst

cncrgy gentx~tion.

ssgmenta

during

waking.

ah-

J Bio-

IYXO

Cavanagh PR:

Model

ical power during distance running.

t;v calculation

J Biomcchanics

ot’ mcchan16: 115-12X.

10x3 11.

Winter

DA:

Biomechanic~

of Human

Movement.

NI’,

Wilq.

I’)79 IS.

Winter

DA:

!znergy asscssmcnt

spy Can 3O:IXi-IYI. 16.

G: l3ncrgy

Ing. Arch

Phys

Corcoran

cxpcnditurc of hcmiplqic abbots Mcd Rehabii

PJ. Jchsen RH.

of plastic

and metal

hemiparctlc Dempster

WT:

lef

Spaw

TR-55-159.

Wright

lnman

Ralston

VT.

44:3hX~370.

Brcngchnann

ambulation.

hraccs

Patterson HJ.

Todd

Air F:

Williams

6i Wilkins.

19x1

Knutsson

E. Richards

C: Dit’fcrcnt

trol

in gait of homiparetic

Loway monitoring

LL:

Mechanical and diagnosi\.

Biomechanics

London.

during wall\-

17.

1’163

Energy natural.

CL.

on sped

Simon>

of seated operator. Human

Ohlo

Walking.

types of disturbed

patients.

Brain

energy

analysis

Prwceding~

Cunada. October

Clin

of ankle and knee Orthop

Winter

DA.

Quanbury

AO.

Reimel-

~ “ait.

J Biomechanics

CD:

Andysib

of inhtantn-

9253%25

Locam 5 I: Redlahc Corporution.

b. Sony

Bdtimorc. motor con-

Motion

7. I976

gait: for

Andyscr:

Sony

Campbell.

CA

of Canada. Toronto.

Ontario.

Can-

ada Vanguard

Projector

Mcllwillc.

NL

GTCO

lY7Y

hcmiplcgic pp 5657

3.

Datatizcr:

9X.45 Dcsktup 4341:

Kockvillc,

Computer:

sauga. Ontario. IBM

Head:

IBM

Vanguard

In\trumcnt

Corporation.

MD

Hewlett-Packard

~C’anada) Ltd.,

Missis-

Canada Canada Ltd..

Markham.

Ontario.

Canada

Become active in the goals of the CongressJoin the American Congress of Rehabilitation Medicine

Arch Phys Med Rehabil Vol67, February 1986

175: l47-

19x3

Suppliers

IY7tl WADC-

of Canadian Society IYXO.

and absorption

and 41ou ca~ltmx~.

coat ot

lY55

102:105430. of

gait. PhyGother-

UC: Ettcct\

and cncrgy

Force Bat’.

gcncratmn

ncoux energy of normal

Arch Phyh Med Rehabil 51:6Y-77. requirsments

DA: tst.

153.

References Bxd

Winter during

11, pathological

197X