Development of Gait By Electromyography

Development of Gait by Electromyography Application to Gait Analysis and Evaluation Tsutomu Okamoto, Ph.D. Kayoko Okamoto, Ph.D.

Walking Development Group Osaka, Japan

Copyright © 2007 by Okamoto & Okamoto

Published by

Walking Development Group ~qTOO3£;PJf~pJT

llJ WALKING

G-S04 Tenno 2-6, Ibaraki-shi, Osaka 567-0S76, JAPAN All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher.

ISBN978-4-902473-05-6 Printed in Japan

Preface The gait of a human being continues to change over the course of a lifetime. The first stage is that of neonatal reflex stepping, which is thought to be the origin of bipedal upright walking in human beings. This then develops into young infant stepping at the age of one to two months, followed by inactive stepping and then by voluntary infant supported walking at the age of six to twelve months. Infants then acquire independent walking at around the age of one and begin to acquire mature adult walking at around the age of three. We have analyzed the detailed changes in the development of human gait employing electromyography (EMG) which has enabled us to carry out motion analysis impossible with conventional methods. At present very little longitudinally analyzed post natal gait development data is available anywhere in the world because of the difficulty of carrying out the necessary experiments. It is even more difficult to record electromyographically the neonatal reflex stepping of newborn babies or the moment when babies begin upright, independent walking. Even today the papers that I wrote on this subject in the 1970s and 1980s continue to be cited. We have continued up to the present to carry out additional crosssectional and longitudinal experiments concerning gait development from the newborn baby stage to that of infant independent walking and have in the process accumulated much electromyographical data. The results of our analysis of normal gait development suggest that it can not only contribute to the explanation and clarification of human bipedal upright walking, but also be applied to various areas of research such as the diagnosis of and therapy for various walking disorders and the evaluation of the level of gait function restoration and improvement. We have gathered together in this book the results of our study and analysis of gait carried out over the last 40 years, in the hope that this rare elctromyographical data concerning gait development will contribute to the further development of this field. Part I contains our analysis, based on movement and muscle activity, of the development and changes in gait from birth until the age of eight, that is from the stage of neonatal reflex stepping, thought to be the origin of bipedal upright walking in human beings, through that of the iii

acquisition and mastering of infant independent walking to that of the acquisition of mature adult walking. Part II introduces our application of this to the analysis and evaluation of gait. We have created "An Index of Gait Instability" based on the results of our analysis of the gait development of infant independent walking, which we apply to research into the nature of human stepping and the evaluation of the level of restoration of walking functions in the elderly. We hope that this book will prove useful to those engages in gait studies, not only as a basic reference material analyzing the development of gait, but also as a basis for research, analysis and application in various fields that will help to generate new ideas about human gait. Tsutomu Okamato

iv

Contents Preface Contents Part I

iii v

Development of Gait -Birth to Age Eight-

1 . Newborn Stepping in Neonates and Young Infants Early neonatal period (1 - 2 weeks) Late neonatal period (3 - 4 weeks) Onset of infant period (1- 2 months) Initial infant period (3 - 4 months) Discussion

2 . Independent Walking in Infants 1st day of learning to walk 2 weeks after learning to walk At around 1 month after learning to walk From 2 to 3 months after learning to walk Subsequent development Standing posture on the 1st day of walking Discussion

1

3

8 12 16 18 20 25

28 30 32 34 36 38 40

3. From Newborn Stepping to Mature Walking - Developmental Changes in One Individual-

45

Neonatal stepping Young infant stepping Infant supported walking Infant walking Immature child walking: unsettled muscle activity Mature walking: toward a mature pattern Developmental period of gait Discussion

48 50

52 54

56 58

59 61

v

Part II

Application to Gait Analysis and Evaluation -An Index of Gait Instability-

67

4. An Index of Gait Instability

-Based on the Development of Independent Walking-

69

EMG findings during the development of gait EMG activity in unstable walking Criteria for Instability An Index of Gait Instability

73 79 84 86

5. Application of an Index of Gait Instability (1) -Supported Walking in Normal Neonates and Infants-

89

Until the 1st month of age 92 From 1 to 4 months of age 94 From 6 to 12 months of age 96 Developmental changes in EMG patterns 98 Application of an index of gait instability to supported walking in babies 99 Discussion 101

6. Application of an Index of Gait Instability (2)

-Recovery of Walking in an Elderly Man after Stroke- 107 1 month after the stroke 7 months after the stroke 1 year 7 months after the stroke EMG evaluation of walking stability Discussion

References Appendix Acknowledgements About the Authors

vi

110 112 114 116 117 121 125 131 133

Development of Gait by Electromyography

EMG experiment of infant walking

The electromyographic (EMG) recordings were done with a pen-writing mUltipurpose electroencephalograph, using surface electrodes 5 mm in diameter. The skin at each electrode locus was scratched lightly with a needle, reducing the resistance between pairs of electrodes to less than 5000

n

(Okamoto et aI., 1987).

Neonatal stepping at 3 weeks afte birth

The purpose of this study was to examine the developmental changes in the functional mechanisms of leg muscles in newborn stepping over the first 4 months in ten normal neonates. Neonatal stepping in the first month showed excessive co-activation, that is, co-contraction patterns of mutual antagonists appeared especially during stance phase. The discharge patterns of co-contraction in neonatal stepping began to change to reciprocal patterns in young infant stepping (after the first month), but excessive muscular activities associated with a slightly squatted posture and forward lean still remained . Strong muscle activities of leg extensors due to a parachute reaction of the legs before floor contact, not seen in the neonatal period, began to appear in the young infant period from 1 month of age to 3 months. We suggest that these gradual changes of leg muscular activity in newborn stepping are evoked by development of balance, postural control, and strength, thereby modulating the neonatal stepping reflex.

When a newborn infant is held under the arms in an upright position, well-coordinated walking movements (stepping reflex) appear to be elicited by tactile stimuli on the soles of the feet as they are placed on the floor (Fig. 1-1). McGraw (1940) and Zelazo et al. (1972) have discussed the significance of early stepping movements for development of adult gait. Newborn stepping has been an object of study for a long time. Only a few attempts so far, however, have been made to study characteristics of newborn stepping by electromyography (EMG). Forssberg (1985) noted that the lateral gastrocnemius showed strong activity just before the foot reached the floor (Fig. 1-2). Because this was like a digitigrade pattern, he concluded that man is born with a quadrupedal locomotor program. Thelen (1982, 1987), however, did not find any strong activity in the gastrocnemius before floor contact (Fig. 1-2) . To further study this problem of the EMG pattern in the gastrocnemius before foot contact in stepping, it would be instructive to record EMG data during stepping not only in the neonatal period (up to 1 month of age), but also in the young infant period (after 1 month of age). We have thus closely examined the characteristics of newborn stepping in ten babies during both neonate and young infant periods in terms of the functional mechanisms of leg muscles. Four male and six female neonates were observed from 1 to 4 weeks after birth. Criteria used for selecting the subjects were that they be full-term with birth weight between 2500 g and 4200 g. They were screened by pediatricians to rule out abnormalities and illnesses. Motor development of each subject was within normal limits.

Fig. 1-1.

Newborn stepping at 2 days after birth.

4 Development of Gait

EMGs of all subjects were recorded from the neonatal period (up to the 1st month of age) to the young infant period (from 1 to 4 months after birth) at intervals of 1 to 4 weeks. To induce newborn stepping, the examiner held the neonate under the arms with the soles of the feet touching a horizontal flat surface. Well-coordinated walking movements were observed from around 1 week after birth to around 3 months. We could not induce stepping simply at will, but tended to be successful when the infants were lively, crying, hungry, or slightly excited (Figs. 1-1 and 1-2). For analysis we selected well-coordinated walking movements consisting of three or more steps. The EMGs were recorded from six muscles in the right leg (Fig.1-2): tibialis anterior (fA), lateral gastrocnemius (LG) , vastus medialis (VM), rectus femoris (RF), long head of biceps femoris (BF) , and gluteus maximus (GM), and from two to six muscles in the left leg, usually the TA, LG, RF, and BE

RF: rectus femoris GM : gluteus maximus

(Knee extensor, Hip flexor)

(Hip extensor)

Mutual antagonist: SF

SF : biceps femoris (Knee flexor, Hip extensor) Mutual antagonist: RF

LG : lateral gastrocnemius

VM: vastus medialis (Knee extensor)

TA : tibialis anterior

(Ankle plantar flexor)

(Ankle dorsiflexor)

Mutual antagonist: TA

Mutual antagonist : LG

Fig. 1-2. Muscles chosen for recording EMG.

Newborn Stepping in Neonates and Young Infants 5

Surface electrodes 5 mm in diameter were used. To attenuate artifacts in the surface electrode recordings, skin impedance was lowered by scratching loci of the electrodes lightly with a needle before the electrodes were applied (Okamoto et al., 1987). The EMG recordings were done with an 18-channel pen-writing electroencephalograph (60 mm/sec) with the gain set at 12 mm/0.5mV. An analog pulse signal from the video recording camera (60 frames/sec) was recorded simultaneously with the EMGs. The walking cycle was divided into swing phase (SW) and stance phase (ST) by the video recordings. Movement and EMG recordings obtained during newborn stepping showed some variations both within and among subjects. Variations in stepping form and EMG patterns appeared to depend to some extent on how the infant was supported. We thus selected as representative data those movements and EMG patterns of stepping that were seen relatively frequently in the neonatal or young infant period being observed. For purpose of analysis, longitudinal observations were divided into early neonatal period (from 1 to 2 weeks after birth), late neonatal period (from 3 to 4 weeks), onset of infant period (from 1 to 2 months), and initial infant period (from 3 to 4 months).

Mature adult walking pattern We need to examine normal stable adult walking to compare with gait in terms of developmental processes. Figure 1-3 shows a typical EMG of adult walking (the subject is a female 29 years of age). From the basogram, stance and swing phases can be demarcated. The discharge patterns of the TA and LG, which participate in movement of the ankle joint, showed an almost reciprocal relationship. The TA (an ankle dorsiflexor) discharged through most of swing phase and at the beginning of stance phase, whereas the LG (an ankle plantarflexor), which participates in push off motion, discharged in a strong burst in the latter part of stance phase. The hip and knee muscles, VM, RF, BF, and GM, acted for shock absorption during the transition from swing phase to stance.

6 Development of Gait

Typical EMG pattern of normal adult walking

0.5

k~j~AA:j~A 326

mv]

333

339

346

356

365

372

378

388

Tibialis anterior (TA)

Lateral gastrocnemius (LG) -+---M~"""~I M'H~---1
Vastus medialis (VM) Rectus femoris (RF)

Biceps femoris (BF)

Gluteus maxim us (GM) Ankle

Plantar flexion Dorsiflexion

t

Knee

Extension

t

Hip

Extension Flexion

Flexion

*

Swing

Stance Phase

Pho.e (SW)

(ST)

Swing

Stance

Basogram Foot contact (FC) VTR signal

HC FF

HO

TO

1 sec

.J._"'-"'''''''''''''''Nt'''",,,,'''',_.J,--*-JIJtr-,,_,,f.,.,,,JtUlt.,,--,~ 300

350

400

l~( ~~ ~ HC (Heel Contact)

FF

HO

TO

(Foot Flat)

(Heel Off)

(Toe Off)

Fig. 1-3. Typical adult EMG pattern in leg muscles during walking. Swing phase (SW: short phase), Stance phase (ST: long phase), Basogram: Foot contact (He, FF, HO, TO).

Newborn Stepping in Neonates and Young Infants

7

Early neonatal period (1 - 2 weeks)

ST

sw

ST

~· '·I·'\'·fJ·tW~~t;~i~

(R)

TA

",""

LG ,

VM RF

BF

i

GM

,--

I

(l)

TA

LG VM ----~~~--------~------+_----------~~~~~------+_

BF

-------------------+------+-------------------~------~. STANCE (ST)

SWING (SW)

--'I

___ 1_ s e_c __

0.5 mv

1 week (Y.T.)

NEONATAL STEPPING

Fig. 1-4. EMGs of stepping at 1 week after birth (Y.T.). SW: swing phase, ST: stance phase, (R): right leg, (L): left leg, TA: tibialis anterior, LG: lateral gastrocnemius, VM: vastus medialis, RF: rectus femoris , SF: biceps femoris, GM: gluteus maximus.

8 Development of Gait

Stepping in this period was characterized by quick hip flexion in which the thigh became approximately horizontal in the first part of swing phase. The foot was raised forward and dorsiflexed strongly, as shown in Figures 1-4, 1-5, and 1-6. In the middle part of swing phase, the leg was often held in the flexed position. Then the foot began to reach the floor slowly, the knee extending passively along with the hip. The foot usually contacted the floor with the heel or sole first (Fig. 1-7), but in a few instances toe contact was seen. A fairly squatted posture was often observed during stance phase.

1 week after birth (Fig. 1-4): In the beginning of stance phase, no notable activity was seen in the leg muscles examined. In single stance phase, continuous discharges were frequently observed in the TA, VM, RF, and GM, activities not usually seen in adult gait. The discharge patterns of the VM and GM were highly consistent, but activities at the ankle (the TA and LG) and of two-joint muscles crossing the knee and hip (the RF and BF) showed slight variations. That is, at the ankle, a reversed reciprocal (TA+, LG-) pattern was observed in many cases, but co-contraction (TA+, LG+) and reciprocal (TA-, LG+) patterns were seen in some of the subjects. Across the knee and hip, a reversed reciprocal (RF+, BF-) pattern was observed in many cases, but co-contraction (RF+, BF+) and reciprocal (RF-, BF+) patterns were observed in some of subjects. In the first part of swing phase, continuous discharge in the TA was seen in most cases. Sometimes slight activity of the RF was observed in the same phase. In the latter part of swing phase, when the leg was extending, activities were hardly seen in the LG, VM, RF, BF, and GM.

Fig. 1-5. Foot contact of newborn stepping at 2 days after birth.

Newborn Stepping in Neonates and Young Infants 9

ST

(R)

SW

V, , , , / " " 1,.1/ " I

TA

'i~

LG--________

ST

" " , ,

W. , / , . , , W; " " I I " I WI I I ;

.. , I

......r+~

,

r-__~--~--------------------+_--+_-----

VM RF BF GM ( L)

TA-""""\ LG RF

L

BF

I SW

ST

1 sec

2 weeks (A. I.) NEONATAL STEPPING

Fig. 1-6. EMGs of stepping at 2 weeks after birth (A.I.).

10 Development of Gait

0.5 mv

2 weeks after birth (Fig. 1-6): In the beginning of stance phase, continuous discharges in many leg muscles were observed more often than at 1 week after birth, especially in single stance phase. Continuous discharge patterns of the VM and GM were again consistent, but EMG patterns of the TA and LG and of the RF and BF showed some variations. That is, reciprocal, reversed reciprocal, and co-contraction patterns were seen in those muscles. In the first part of swing phase, continuous discharge patterns of the TA were similar to those of 1 week after birth. In contrast to the first week, continuous discharge in the RF was observed during swing phase in many cases. In the latter part of swing phase, activity in the BF frequently began to appear before foot contact.

Foot contact (Fe) Heel contact with slow leg extension

Foot flat with slow leg extension EMGs of leg extensors before floor contact Early neonatal period (1-2 weeks after birth)

Fig. 1-7. Foot contact of stepping in neonatal period (1-2 weeks after birth). (-): no activity.

Newborn Stepping in Neonates and Young Infants

11

Late neonatal period (3 - 4 weeks)

rt~flflft

sw ST '-H'"""1'-1V.--r--r--t-.'-I"~'-I--''-h'-l--i-W....,~t-H-1'''''''''''i,....J'""'h''''~I'''''''~~h-h'' (R)

TA_ ..Wi

.,'

LG VM RF , BF-!\

,II~

...,.

...

Jill/ILl

'n'

~

''''r-

GM

(LiA"'M~~~~~I'J' LGW~~~ , RF.~""•• "

...

~~~

.\

.';':;'1.;'.;, 1r""'"'Wt""'~ '.Oi.~i~~'jI~""'\""~

....

w..~t~

BF~~ SW

ST

1 sec

3 weeks (H. YJ NEONATAL STEPPING

Fig. 1..8. EMGs of stepping at 3 weeks after birth (H.Y.).

12 Development of Gait

I

0.5 mv

--

SW

hS-rT-r-r--rr-h h----r-r-r--r-,,-...,-,

(R)~ TA

II"

1ft.

""

I

LG Ib", "'I'

RF SF

II.

In

'-\iNo'~

"""rIP

r-"

'W

VM ~~7i 1

~.,~'f

".'

.'\'"',..

~,

.J.

\1"'" /-h-

,I

I\,

GM (L)

,I.

",1."1".1,,,1,,,11, "r'lf

'''''I'

TA /JftOoMiI\--~--"aWJ"'M--..-j>,~~

.., ........~,''I'--

LG~~~----r-----~--~~~~~-

VM~~#~~I~\\\-\fi\'oJ--.'tol\--+III~~1'~~~tr qj.-'~~~4~~~~ftJt!\,jr\""-­

IIIWIlfIM~.r------+---1'-~,~fr""\"\- GMMM~vN(i~/""""'---+--~~~~'4\t'+.'/i'UIf,~,*,-ST

SW

1 sec

I 0.5 mv

4 weeks (T. YJ

NEONATAL STEPPING

Fig, 1-9, EMGs of stepping at 4 weeks after birth (T.Y,),

Newborn Stepping in Neonates and Young Infants

13

As in the early neonatal period, leg flexion was very active in the first part of swing phase in this period (Figs. 1-8 and 1-9). The thigh was outwardly rotated as it was raised diagonally in a forward and lateral direction and the foot dorsiflexed strongly. Then the foot began to approach the floor slowly, the knee extending passively along with the hip. The foot usually contacted the floor with the lateral border first (Figs. 1-10, 1-11, and 1-12), but sometimes the heel, sole, or forefoot made initial contact. The fairly deep squatting posture of the early neonatal period began to become less pronounced during stance phase. 3 and 4 weeks after birth (Figs. 1-8 and 1-9): Throughout stance phase, continuous discharges of leg muscles were observed in many cases. EMG patterns of the VM and GM were consistent as in the early neonatal period. The reversed reciprocal ankle pattern during stance, seen in neonatal former period, was hardly evident, whereas the reciprocal and co-contraction patterns became more frequent. Discharge patterns of the two-joint knee and hip muscles showed reversed reciprocal, reciprocal, and co-contraction patterns as in the early neonatal period. In the first part of swing phase, continuous activity was seen in the TA as in the early neonatal period, but weak bursts of the RF and BF were seen often at the beginning of swing phase. In the latter part of swing phase, activities began to be seen in the LG, VM, RF, BF, and GM in some of the neonates.

VM LG (-), partly (+) f----+--r~

H , partly (+)

Vastus medialis (Knee extensor)

Lateral gastrocnemius (Ankle plantar flexor)

Lateral border with slow leg extension

EMGs of leg extensors before floor contact Late neonatal period (3-4 weeks after birth)

Fig. 1-10. Foot contact of stepping in neonatal period (3-4 weeks after birth). (-): no activity, (+): noticeable activity.

14 Development of Gait

Fig. 1-11. Neonatal stepping at 26 days after birth.

Fig. 1-12. Foot contact of neonatal stepping at 22 days after birth.

Newborn Stepping in Neonates and Young Infants

15

Onset of infant period (1 - 2 months)

SW

ST

"-"-.-v · · · " · , Wr,,....,,,....,,r-r-.,.-,,.-,.,. . "W·

,

j

,

,

,

,

,

U ' , ,, ,••

( R)

TA

BF GM~--...

(Ll

TA'IIA(/~""",~-""'t"''''r'TM~~.~~i''J,rIII..~~.....~IJ1fIt4>~Ji'.

ST

SW 1 sec

0.5 mv

1.5 months (T. YJ YOUNG INFANT STEPPING

Fig. 1-13. EMGs of stepping at 1.5 months after birth (T.Y., same subject as in Fig. 1-9).

16 Development of Gait

After 1 month, as shown in Figure 1-13, leg flexion was performed strongly in the first part of swing phase as in the neonatal period, but the degree of hip flexion tended to decrease slightly. We found mostly plantarflexion of the foot before floor contact rather than dorsiflexion, which had been more prevalent in the neonatal period. The foot usually contacted the floor with the lateral border of the forefoot first (Fig. 1-14). Knee extension began to be performed more actively than in the neonatal period. A half-squatting posture during stance phase tended to increase. 1.5 months after birth (Fig. 1-13): During stance phase, continuous discharges of the VM and GM were seen as in the neonatal period. The ankle muscles likewise exhibited reciprocal and co-contraction patterns as in the late neonatal period. The two-joint knee and hip muscles showed reversed reciprocal, reciprocal, and co-contraction patterns, similar to the neonatal period. In the first part of swing phase, continuous activity of the TA was observed in many instances, as in the neonatal period. In the beginning of swing phase, weak activities of the RF and BF were seen often, but not always. In the latter part of swing phase, activities of the LG, VM, RF, BF, and GM appeared often before foot contact.

LG (-), (+) 1---+--+-.1 Lateral gastrocnemius (Ankle plantar flexor)

•

Foot contact (Fe) Lateral border of forefoot with fast leg extension

EMGs of leg extensors before floor contact Onset of young infant period (1-2 months after birth)

Fig. 1-14. Foot contact of stepping in young infant period (1-2 months after birth). (-), (+): instances of no activity and of noticeable activity intermingled.

Newborn Stepping in Neonates and Young Infants

17

Initial infant period (3 - 4 months)

h,J

SW

ST

W t'

' ihoIh'th ' f ...... ; i .. Wrlr'Ir'Ih'.h

I' "

i' "

;

..

Wit' , .

t'

of

~

\-T

( R)

TA---4~~~------~~~~--~--------~~~~

~ TA~:;~~

(0

LG~ ST

"'~';~

•

SW

t sec

3 months

0.5 mv

CA. I.) YOUNG INFANT STEPPING

Fig. 1-15. EMGs of stepping at 3 months after birth (A.I.. same subject as in Fig. 1-6).

18 Development of Gait

In this period (Fig. 1-15), the lower limb flexed strongly in the first part of swing phase as in the neonatal period, but the total degree of hip flexion tended to decrease slightly. The foot usually approached the floor with a more rapid and vigorous extension of the lower limb, with the toes initially contacting the floor (Fig. 1-16). Knee extension and ankle plantarflexion were visibly active in many cases. A halfsquatting posture during stance phase became more frequent. 3 months after birth (Fig. 1-15): During stance phase, continuous discharges in the VM and GM were observed until onset of the infant period, as mentioned above, but continuous discharges in the anteriorly situated TA and RF tended to decrease or disappear, leading to reciprocal patterns (TA- LG+ and RF- BF+) in most cases. In the first part of swing phase, strong bursts in the TA, RF, and BF were frequently observed. In the latter part of swing phase, strong activities of the LG, VM, BF, and GM appeared often. Although strong activity in the LG and VM were observed shortly before foot contact, activities in the BF and GM were not generally seen. In the course of this period, co-contraction patterns of the ankle (TA+, LG+) and the knee and hip (RF+, BF+), seen fairly often in stance phase in the late neonatal period and onset of the infant period, gave way to reciprocal patterns (TALG+ and RF- BF+) . Strong activities of the LG and VM in the latter part of swing phase, hardly observed during the neonatal period, became remarkably more frequent.

VM (+), partly (-) LG (+), partly (-) f-+-+-~ Lateral gastrocnemius (Ankle plantar flexor)

Vastus medialis (Knee extensor)

Forefoot with fast leg extension EMGs of leg extensors before floor contact Initial young infant period (3-4 months after birth)

Fig. 1-16. Foot contact of stepping in young infant period (3-4 months after birth). (+): noticeable activity, (-): no activity.

Newborn Stepping in Neonates and Young Infants

19

Discussion Although Thelen et al. (1982) reported that when held upright, newborn infants show well-coordinated walking movement that normally cannot be elicited after about 2 months of age, we could induce infant stepping until around 3 months of life in a number of cases. Forssberg (1985) and Thelen et al. (1987) pointed out from movement patterns and EMGs, that the locomotor pattern of the newborn differs markedly from that of an adult. From our results, newborn stepping was characterized by active leg flexion with the thigh becoming horizontal, a somewhat squatted posture, and variable forms of foot contact with the surface (Figs. 1-17 and 1-18). Leg muscle activities in newborn stepping are usually irregular and involve more co-activation than in adult walking, especially in stance phase. For example, in single stance continuous discharge patterns were seen in the knee and hip extensors (VM and GM) in neonatal and infant stepping, associated with a progressively decreasing but ever present squatted posture. These activities in the leg extensors appear to be attributable to the squatted posture itself and would thus not be seen in adult gait. On the other hand, we did observe some similarities in leg muscle activity between newborn stepping and adult gait. As swing phase was beginning, for example, bursts were usually observed in the TA during newborn stepping. Muscle activation seen in flexors of the lower limb at the onset of the stepping cycle becomes incorporated into supported walking seen prior to independent walking, thence into early independent walking, and so on to adult gait. These results suggest that mature walking may evolve from the newborn movement pattern. We could see a developmental trend across the neonatal and young infant periods in stance phase and at the end of swing phase. In stance phase, contractile activity between mutual antagonists varied among co-contraction (TA+ LG+ and RF+ BF+), reciprocal (TALG+ and RF- BF+) , and reversed reciprocal (TA+ LG- and RF+ BF-) patterns. The reciprocal pattern tended to appear more often if the baby happened to be inclined forward and the reversed reciprocal pattern when the baby was inclined backward. Co-contraction might be viewed as an intermediate situation between these two tendencies.

20 Development of Gait

Developmental changes in the pattern of newborn stepping 1week (Y. T.)

Neonatal period (Early. 1- 2 weeks)

A

~~~~j 2 weeks (A.I.)

Neonatal period (Early. 1-2 weeks)

B

jjlttf1 3 weeks (H.Y.)

Neonatal period (Late. 3-4 weeks)

C

fttfttttt 1.5

Infant period (Onset. 1-2 months)

D

months (H.Y.)

frftfflf 3

Infant period (Initial. 3-4 months)

E

iiw~~ 3.5

Infant period (Initial. 3-4 months)

F

months (A.I.)

months (H.Y.)

rl~~~(f

Fig. 1-17. Developmental changes in the pattern of newborn stepping. B: same subject as E. C: same subject as D and F.

Newborn Stepping in Neonates and Young Infants 21

The reversed reciprocal pattern was seen relatively often in the early neonatal period (first 2 weeks), but the other two patterns became more frequent in the late neonatal period (3rd and 4th weeks) and as the infant period began (2nd month). In the initial infant period (3rd and 4th months) the reciprocal pattern became more dominant than the other patterns, although all three patterns could still be observed. Interestingly, this trend anticipates the changes in pattern between mutual antagonists seen as a baby first begins to walk independently and becomes more stable in the ensuing months. At the end of swing phase in the neonatal period, the LG and VM exhibited no activity until the foot actually touched the floor (Fig. 1-18). The foot reached the floor in a relatively passive action of the lower limb, contacting the floor variously with the heel, entire sole, or lateral border. Thelen et al. (1982, 1987) did not find any strong activity in the gastrocnemius before floor contact in the neonatal period. In the second month, at the onset of the infant period, the LG and VM began to become active before actual contact of the foot with the floor, with the lateral part of the forefoot generally touching the floor first. Thelen et al. (1987) and Forssberg (1985) reported that the gastrocnemius showed strong activity just before the foot reached the floor in the young infant period. The activities of the LG and VM subsequently became more pronounced shortly before and during floor contact in the 3rd and 4th months, to the point that one might associate such activity with the parachute reaction. Milani-Comparetti et al. (1967) observed from movement analysis that the parachute reaction of the lower limbs begins to appear at about 4 months after birth. Our observations, if they are of the same phenomenon, suggest that the beginnings of the parachute reaction can be found by EMG much earlier than by visual observation of behavior. These changes in muscle activity during the stance and swing phases of newborn stepping represent what might be considered as the first developmental changes in human bipedal locomotion. Further research would be necessary to elucidate the extents to which these changes can be attributed to maturation of balance, postural control, and strength, as well as to emergence and disappearance of the neonatal stepping reflex itself.

22 Development of Gait

Neonatal and young infant period

Foot contact with leg extension

EMGs of VM and LG before floor contact

Early neonatal VM(-) LG(-)

1-2 weeks after birth Heel contact or foot flat with slow leg extension

Late neonatal VMH, partly (+) LG(-), partly (+)

3-4 weeks after birth Lateral border with slow leg extension

Young infant (onset) VMH, (+) LGH. (+)

1-2 months after birth Lateral border of forefoot with fast leg extension

Young infant (initial) VM( +), partly (-) LG(+), partly (-)

3-4 months after birth Forefoot with fast leg extension

Fig. 1-18. Developmental changes of foot contact in newborn stepping. VM : vastus medialis, LG: lateral gastrocnemius, (-): no activity, (+): noticeable activity, (-), (+): instances of no activity and of noticeable activity intermingled.

Newborn Stepping in Neonates and Young Infants 23

Conclusion In ten neonates first seen at 1 to 4 weeks after birth, EMGs of stepping were recorded at 1 to 4 week intervals until around 4 months of age. During stance phase in neonatal stepping, many leg muscles showed excessive continuous discharges compared with the adult walking pattern. Continuous activity was seen in the vastus medialis and gluteus maximus to maintain a partially squatted posture. Mutual antagonists in the lower limbs variously showed reciprocal and cocontraction patterns during the neonatal period, but the EMG patterns began to shift toward predominantly reciprocal patterns in the young infant period, associated with leaning forward. In the first part of swing phase, activity in the tibialis anterior was observed in most cases. During neonatal stepping, in the latter part of swing phase, muscular activity was not seen in the lateral gastrocnemius or vastus medialis, but during young infant stepping EMG activity in these two muscles became marked before the foot reached the floor, suggesting that muscular activities participating in active ankle plantarflexion and knee extension began to act as a precursor to the parachute response of the lower limb. In summary, these muscular activities of the lower limb characterize the EMG features of newborn stepping. Changes in EMG patterns during newborn stepping, detectable well before corresponding changes can be visually observed in movement analysis, may be the first signs of development in human locomotion.

24 Development of Gait

In order to elucidate electromyographic (EM G) characteristics of infant walking at the onset of independent gait, we longitudinally recorded EMGs from muscles of both legs during the learning process of walking in an infant, from 10 months after birth until about 3 years of age. We found EMG characteristics of infant gait up to around 1 month after learning to walk that are not usually seen in adult gait. In stance phase from foot contact until push off, the role of the vastus medialis for maintaining stability became clear as a slightly squatted position was used to lower the center of gravity. Orderly reciprocal or co-contraction patterns of activity in the rectus femoris and biceps femoris or in the tibialis anterior and gastrocnemius were found to be related to returning the body's center of mass toward its initial position. In the latter half of swing phase, the vastus medialis and gastrocnemius showed strong activities with the knee extending and ankle plantarllexing for active leg extension to prevent falling. These characteristically excessive muscle activities in infant walking are considered to express weak muscle strength and an immature balancing system. As months and years pass, the muscles become stronger and balance matures, obviating the need for so much myoelectric activity.

Normal human infants begin to walk independently when they are about 1 year of age. Thelen et al. (1989) noted that independent walking emerges when a threshold has been reached for muscle strength and ability to balance, but the baby who has just become able to walk independently exhibits a pattern notably different from adult gait. Although a great deal of investigation has been done on development of gait, there are few EMG studies in the area. Crosssectional kinesiological EMG studies on the development of independent gait in babies have been performed by Sutherland et al. (1980), Forssberg (1985), and Thelen et al. (1987), but we have not seen much longitudinal EMG study on the acquisition of gait outside of that by Okamoto et al. (1972, 1983, 1985, 2001, 2003). By means of both longitudinal and cross-sectional EMG and cinematographic findings, we have reported that specific changes can be observed at certain times in the course of that development. That is, during the early stage of independent walking, a baby squats slightly while leaning forward and takes steps with strong active extension of the legs, exhibiting considerable instability. After this early stage of independent walking, the baby exhibits increased stability with the body tilted only slightly forward (childhood walking pattern), and by 3 years of age the body is upright as in adult walking (adult walking pattern). What seem to be most lacking, however, are EMG studies during the very early stage of independent walking in the infant. The purpose here is to explore a little further the onset of independent walking in the infant and to determine EMG characteristics of infant walking by longitudinal observations. The subject was one baby who first began to walk independently at 306 days after birth. We made longitudinal observations on this child from the time she first began to walk independently at 10 months after birth until a stable adult-like walking pattern was achieved at around 3 years of age. Figure 2-1 shows a representative form of infant walking at the onset of independent walking, when the infant succeeded to walk 5 to 10 steps without support. Slight knee flexion was often observed in the supporting leg, the foot base in the double support period was very wide, and the body's center of gravity was lowered during stance. The arms were spread apart and elevated.

26 Development of Gait

The gait in this baby first learning to walk was characterized by quick hip and knee flexion in which the thigh became almost horizontal in the first part of swing phase. The foot was raised forward and slightly outward, then the foot began to reach the floor quickly, the knee extending actively along with the hip. The foot usually contacted the floor with the foot flat and forefoot first, but in a few instances the heel made initial contact. A squatting posture with the body inclined forward was often observed during stance phase. We noticed several other characteristics that differ from adult gait, such as a wide base at the feet and a "high guard" position of abducted arms (Figs. 2-3, 2-5, 2-7, and 2-9). Figures 2-2, 2-4, 2-6, 2-8, and 2-10 show longitudinal developmental changes of EMG activity in the learning process of walking. Compared with corresponding muscular activities of the adult walking pattern, excessive muscular activities and variations appeared during the learning process of infant walking from the 1st day of learning to walk until 2 or 3 months after learning to walk. In the description that follows, we focus attention on EMG activity patterns seen in the infant that deviate from normal adult walking and examine developmental changes in muscle activity related to infant independent walking.

Fig. 2-1. Gait pattern at the onset of independent walking.

Independent Waiking in infants 27

1st day of learning to walk

fffffr (R)

TA LG VM RF BF GM (U TA LG VM RF BF GM sw

ST

( R)

KNEE~ (L)

KNEE

v

~

EXT. ~ FLEX.

~

-----------

V t sec

0.5 mv

1st day of learning to walk

Fig. 2-2. EMGs on the 1st day of independent walking (at 10 months of age). ST: stance phase, SW: swing phase, (R): right leg, (L): left leg, TA: tibialis anterior, LG: lateral gastrocnemius, VM: vastus medialis, RF: rectus femoris, BF: biceps femoris , GM: gluteus maximus, KNEE EXT: extension, KNEE FLEX: flexion.

28 Development of Gait

Figure 2-2 shows a representative excerpt of the EMG patterns of the infant's independent walking on the day when she succeeded in walking 5 to 10 steps for the first time, at 10 months after birth. In stance phase, at the ankle, a pattern of two or three alternating bursts between the TA and LG was most prevalent, but co-contraction of both muscles was also seen frequently. At the knee, the VM was continuously active from foot contact until push off. At the hip and knee, three types of discharge pattern were seen in the biarticular RF and BF muscles. One was a reciprocal (RF-, BF+) pattern in which discharge of the RF tended to decrease or disappear while that of the BF increased. A second was a reversed reciprocal (RF+, BF-) pattern in which discharge of the BF tended to decrease or disappear while that of the RF increased. The third was a co-contraction (RF+, BF+) pattern of the two muscles. When the infant became able to walk continuously, we generally found a reciprocal or co-contraction pattern, although we occasionally observed a reversed reciprocal pattern. At the hip, the GM was continuously active. In swing phase, the LG and VM often showed strong activity in the latter half of the phase.

1st day of learning to walk

Fig. 2-3. Foot prints on the 1st day of independently walking (at 1 year 1 month).

Independent Waiking in infants 29

2 weeks after learning to walk

sw

ST

10.5 months

( R)

TA LG VM RF SF GM (L)

TA LG VM RF SF GM ST

( R)

KNEE ( L)

KNEE

sw

*

EXT . FLEX.

1 sec

2 weeks after learning to walk

Fig. 2-4. EMGs at 2 weeks after learning to walk (at 10.5 months).

30 Development of Gait

0.5 mv

Figure 2-4 shows representative EMG patterns of infant walking at about 2 weeks after learning to walk (at 10.5 months after birth), when the infant was able to take more than 20 steps. In stance phase, at the ankle, the earlier pattern of two or three alternating bursts between the TA and LG changed to one or two alternating bursts, but co-contraction of the two muscles was also seen frequently. At the knee and hip (the VM, RF, BF, and GM), EMG patterns in this period did not differ from those on the 1st day of learning to walk (Fig. 2-2). In swing phase, the LG and VM frequently showed strong activities in the latter half of that phase, as on the 1st day of learning to walk.

Fig. 2-5. Unstable infant independent walking.

Independent Waiking in infants 31

At around 1 month after learning to walk

11 months ( R)

TA~~~~~~~~~~~~mM~~~~~~-+~WM~

GM+-~~~~~~~~~~--~--~~~-r~~~~+-~~~ (L)

TA~~~~NH~~~~~~~~~~~~~~~~~

( R)

KNEE EXT. ..

( Ll FLEX.• KNEE ---~

/---1 sec

1 month after learning to walk

Fig. 2-6. EMGs at 1 month after learning to walk (at 11 months).

32 Development of Gait

I

0.5 mv

Figure 2-6 shows representative EMG patterns of infant walking at around 1 month after learning to walk (at 11 months after birth). At this point, the infant began to walk by herself for long periods. In stance phase, at the ankle, the previous pattern of one or two alternating bursts between the TA and LG disappeared, but cocontraction of both muscles was also seen frequently. The reciprocal (fA-, LG+) pattern tended to increase, and co-contraction (TA+, LG+) of both muscles tended to be seen at about the same frequency as at 2 weeks after learning to walk (Fig. 2-4), but the reverse reciprocal (TA+, LG-) pattern began to decrease or disappear. At the knee, VM activity tended to decrease or disappear. At the hip and knee, although the reciprocal (RF-, BF+) pattern increased, the reverse reciprocal (RF+, BF-) and co-contraction (RF+, BF+) patterns tended to occur much less frequently than at the onset of independent walking. At the hip, activity of the GM in this period did not change from the pattern at the onset of independent walking (Figs. 2-2 and 2-4). In swing phase, discharges of the VM began to decrease in intensity or even disappear in the latter half of that phase, in contrast to the situation at the onset of independent walking (Figs. 2-2 and 2-4). Discharges of the LG, on the other hand, still remained strong in the latter half of swing phase.

~

~

21st day after learning to walk

43rd day after learning to walk

Fig. 2-7. Foot prints of initial infant walking on the 21 st and 43rd days after learning to walk (at 1 year 1 month).

Independent Waiking in infants 33

From 2 to 3 months after learning to walk

( R)

!il"~

~11

1'1'

11\'

ld'llLtk

LG ,~.

VM RF

GM

.I.

l' \

LG 1.1

I.

VM

111

3'I.lj~h'.

j

hUll,

1'1'

,

'If"

.L IdlL~.!

"I fU""llr

T

~II"

!'I

I.

.1,lhlLJ

"11.1.

"1'''

I'

JL

.~.

I

"

.~ J~u "d. ,1J.Jlil, I "'.Jj. 'Ir, I""",!,' '11r'

'JI.J...iIIo, ,1I n"r'

~~h:' llllll

I,~I.

~" .

I''''' ['or

UUi,

:r'll"

1

I~"

,t...

II II~jJll 'r'l'II"I"

~rJl

1.1 'ILl,

~L~ I 'lI'm "

1,\ """'1..".•,

"111 1

ST

~T.

,

'"I:'~U

.oiJ

'r~

~,UJ III.

'I'

~..

dj~J.

l"'lr

~ r

~LlLl.

11"'\lf 11/ It

""r'l"

",i"

RF

,1

..J.'ll...

kL.L 1\

'w

c., .. Ji!b '\ "r\' 11"

r .".

Ilr!\1~ql

iJI.l 1'1"1

TI' I"

,I.

l....i1,dl

r

11'

"1'

II

"'11" ~,~I .Il,lil",

(L)

~I jJo..

I'

I.o!.

.~ ••1..11, 'II " II'" '11' ".Ii""L

SF

I~ IJI

IU., !1~ Ill' !'r

'r

'f'

~

111''1I\~J.ilal'

l~h.J, 1\111

'1'''''

-"

TA 11/

Il.Iil.

·frr".r'"'

11'

GM

w

1!111 I~.

TA

SF

12 months

SW ST ~""""'\J

'IIITI'H ,r.I~I~·1

ItJ,

r

.L

rm"

'1"

...

.1

I It.

Ilfll'

'./.~j ,JIJld.

II 'I'T''''

..L [,IIJljA,

~(l

,~~r'

,lIl.

'I

.lJ,.

,1"""

~r'"

SW

(R)

KNEE

, sec

2 months after learning to walk

Fig, 2·8, EMGs at 2 months after learning to walk (at 12 months),

34 Development of Gait

lJ,,,•

,I.

~~, '~f l'll i~u ~ Jl: 1 ,.

~~

'1'

j~lI..lI..1c

'It '''I

Figure 2-8 shows representative EMG patterns of infant walking at about 2 months after learning to walk (at 12 months after birth). The infant had acquired comparatively stable walking. In stance phase, at the ankle, the co-contraction (TA+, LG+) pattern began to decrease in frequency, whereas the reciprocal (TA-, LG+) and reversed reciprocal (TA+, LG-) patterns were the same as at 1 month after learning to walk (Fig. 2-6). At the knee and hip, there were no obvious changes in EMG patterns of the VM, RF, BF, and GM. In swing phase, although strong discharges of the VM decreased or even disappeared in the latter half of that phase, discharges of the LG still remained the same in the latter half of swing phase as at 1 month after learning to walk (Fig. 2-6).

TO

sw

Fe

1 year

ST

TO

SW 1 sec

FC

1 year 3 months

ST

0.5 mv

1 week after learning to walk

1 sec

I

0.5 mv

3 months after learning to walk

Fig. 2-9. EMGs in mutual antagonists (TA versus LG) of infant independent walking. TO: toe off, Fe: foot contact, SW: swing phase, ST: stance phase,

Left: at 1 week after

learning to walk (at 1 year), Right: at 3 months after learning to walk (at 1 year 3 months). Muscle activity progressed from excessive co-contraction of mutual antagonists to reciprocal patterns.

Independent Waiking in infants 35

Subsequent development

TO

He

TO

HC

. 11.,

TA

.1 "" •

!,...~

I I~'II'J\'

~/~

LG

'JliLJ~jd ~

..Il ,U~

"11' '11"

~ ~.

VM

,

.I

RF

.l/k.

BF

'\1'''

. 1,

rr

"1'"

GM SW

ST

1 sec

SW

10.5mv

-------' 1 year 9 month s IMMATURE CHILD WALKING PATTERN

ST

__,_,e_c_->I mv 0.5

3 years 2 months MATURE ADULT WALKING PATTERN

Fig. 2·10. EMGs of the learning process of walking. TO: toe off, He: heel contact, SW: swing phase, ST: stance phase, Left: at 1 year 9 months (immature child walking pattern), Right: at 3 years 2 months (mature adult walking pattern).

36 Development of Gait

Figure 2-10 (left panel) shows representative EMG patterns of immature childhood walking at 1 year 9 months of age. The infant acquired comparatively stable walking with the body inclined forward (Fig. 2-11) . In stance phase, at the ankle, a reciprocal (fA-, LG+) pattern was observed most often. Reciprocal (RF-, BF+) patterns were also seen at the hip and knee, and continuous activities of antigravity muscles (LG, BF, and GM) were found. In swing phase, discharges of the LG seen at around 2 or 3 months after learning to walk decreased or disappeared in the latter half of swing phase and more greatly resembled the usual adult walking pattern. Discharge patterns of the leg muscles did not appreciably change from 3 months after learning to walk until approaching the third year of age. Figure 2-10 (right panel) shows representative EMG patterns at 3 years 2 months of age, resembling mature adult walking. At this point, the infant appeared to have acquired the adult walking pattern using a strong push-off of the foot with the body erect (Fig. 2-11). In stance phase, at the ankle, reciprocal (TA-, LG+) patterns previously found in the first half of stance phase decreased or disappeared and strong bursts were observed instead in the latter part of stance phase, as in adult walking. At the knee and hip, reciprocal (RF-, BF+) patterns decreased or disappeared. Strong continuous discharges of the LG, BF, and GM, that had been seen until about the end of 2 years of age, began to decrease or disappear. EMG activity patterns that decreased or disappeared at around 3 years of age were closely approximating adult forms.

IMMATURE INFANT WALKING PATTERN

..

IMMATURE CHILD WALKING PATTERN

up to 3 months after learning to walk

3 months - 2 years after learning to walk

1 year - 1.3 years

1.3 years - 3 years

..

MATURE ADULT WALKING PATTERN after 2 years of learning to walk 3 years -

Fig. 2-11. Development of gait pattern from infant walking to mature walking.

Independent Waiking in infants 37

Standing posture on the 1st day of walking

GM~~~n~~~--~~~~~~~~MM~~~~~~~~~ ( L)

TA

::~~~ GM..

... 11,

~~.

••

",., ..

~'" I",."""~¥,,,~ .Jr¥~!..."l~,

"'-1111-'....._ _ __

( R)

KNEE-------." ( L)

EXT . • FLEX.•

K N E E - - - - -...... ( R)

FC ( L)

FC FF

HC

TC

1 sec

0 .5 mv

Standing posture on the 1st day of walking

Fig. 2-12. EMGs of standing posture with a slight squat on the 1st day of independently walking (at 10 months). (R): right leg, (L): left leg, FF: foot flat with the body erect, HC: heel contact with the body inclined backward, TC: toe contact with the body inclined forward.

38 Development of Gait

Figure 2-12 shows EMGs of standing with a slight squat on the 1st day of independent walking. These discharge patterns were similar to those during stance phase on the same day (Fig. 2-2). During maintenance of standing posture at the ankle, alternative bursts between the TA and LG generally showed a reciprocal (TA-, LG+) pattern at toe contact (fC) with the body inclined forward, and a reversed reciprocal (fA+, LG-) pattern at heel contact (HC) with the body inclined backward. Occasionally a co-contraction (TA+, LG+) pattern was seen at toe contact (TC) with the body inclined forward. At the knee, the VM showed continuous strong activity during maintenance of slight knee flexion. At the hip and knee, the three discharge patterns (reciprocal, reversed reciprocal, and co-contraction) between biarticular muscles (RF and BF) could be seen. At the hip, the GM generally showed continuous activity during the maintenance of standing.

Fig. 2-13. Standing posture just before independent walking at 1 year of age.

Independent Waiking in infants 39

Discussion When a baby is just beginning to walk, characteristic EMG patterns can be seen that are excessive when compared to the corresponding patterns in adults. We consider here certain EMG patterns that gradually changed from the time of first learning to walk, principally those in stance phase and in the latter part of swing phase. In stance phase, we have found that excessive muscular activity and patterns peculiar to gait in an infant who has just begun to independently walk, strongly resemble lower limb activity during maintenance of an upright standing posture in the same period of development (Figs. 2-2, 2-4, and 2-12), suggesting that a common mechanism operates both in standing and in the initiation of gait. From a mechanical point of view, at this very early stage, both activities require a low center of gravity and a wide base of support to assure maximum stability. Generally these tasks can be accomplished, even though strength and balance are yet undeveloped, by spreading the legs apart to widen the base of support and by maintaining the knees in slight flexion to lower the center of gravity. During knee flexion in stance phase, continuous discharges of the VM are generally seen until around 1 month after learning to walk (Figs. 2-2 and 2-4). In stationary standing, the VM is continuously active as the baby stands fairly squatted on the 1st day of independently walking (Fig. 2-12). The VM activity seen at the onset of independent gait thus appears to contribute to holding a posture with slight knee flexion, permitting the body's center of gravity to be lowered so that balance is easier to maintain. Mer the first month of walking, such continuous discharges of the VM tend to decrease or disappear (Figs. 2-6, 2-8, and 2-10). This agrees with observations by Okamoto et al. (1985, 2001, 2003) that the load at the knees decreases as strength and balance develop. Another important factor to consider is keeping the vertical projection of the body's center of gravity well within the bounds of the base of support. In our study, the baby who had just begun to walk independently exhibited control over inclination of the trunk during walking or standing, thus keeping the center of gravity within the base of support, by orderly patterns of activity in the leg muscles (Figs. 2-2, 2-4, and 2-12). As mentioned above, three types of discharge patterns were seen in the biarticular RF and BF muscles. First, the reciprocal (RF-, BF+) pattern is considered to be necessary for gait

40 Development of Gait

with an anteriorly inclined trunk. Before strength and balance have matured to the point that push off can be effectively used with the trunk upright, as in adult gait, this pattern tends to increase after 1 month of learning to walk. This pattern is similar to a child's walking pattern (Fig. 2-10, left panel). Second, the reversed reciprocal (RF+, BF-) pattern is considered to help control displacement of the body's center of mass by participating in maintenance of posterior inclination ofthe trunk. Third, a co-contraction (RF+, BF+) pattern is considered to keep balance control with the body erect. The reversed reciprocal and co-contraction patterns are normally seen during the very unstable period of the first month after beginning to walk (Figs. 2-2 and 2-4), but not thereafter. These patterns are not seen in the child or adult walking pattern. These EMG patterns thus suggest that excessive muscular activity is a characteristic feature of balance control when the baby takes steps for the very first time. While these two muscles (RF and BF) act at the hip and knee, Nashner et al. (1985) have pointed out that ankle strategy is the most efficient for returning the body's center of mass to its initial position. Indeed, in our study the TA and LG exhibited alternating reciprocal patterns of activity, thus affording anteroposterior control over the center of gravity to help maintain upright stability. That is, activity of the TA is considered to participate in maintenance of posterior inclination of the trunk, while activity of the LG is considered to be necessary for gait with an anteriorly inclined trunk. We also found variations in the alternating reciprocal patterns of the ankle muscles at about 2 weeks after learning to walk (Fig. 2-14). From the viewpoint of the developmental process, it clear that two or three alternating bursts of these muscles (TA and LG), seen in the very unstable period at the onset of independent walking and stationary standing, disappear at around 1 month after learning to walk (Figs. 2-2 and 2-6). The fact that this alternating burst pattern becomes attenuated with experience of walking further suggests that it is a characteristic EMG feature of balance control when the baby takes steps for the very first time. The TA and LG have previously been reported to co-contract in many instances at the onset of independent walking, and McGraw (1940) pointed out that co-contraction of these mutual antagonists is indicative of maintaining balance by strongly stabilizing the ankle. However, basograms recording using foot contact switches during stationary standing (Fig. 2-12), when the trunk was markedly inclined forward, suggest that the TA in synchrony with the LG acts for inversion to actively prevent falling. In addition to the Independent Waiking in infants 41

pattern of two or three alternating bursts of the TA and LG, as mentioned above, co-contraction of ankle muscles can be considered the expression of an immature balancing system. It would be very difficult for an infant to maintain a prolonged single stance phase at the onset of independent walking. In the adult walking pattern, strong myoelectric discharges during single leg support are hardly seen from foot contact until push off. In contrast, excessive discharges at the onset of independent walking in infant are often observed during single leg support. During single leg support, as shown in Table 2-1, up to around 1 month of learning to walk, the anteriorly located muscles of the lower limb (TA, VM, and RF) are just as active as the posteriorly located muscles (LG, BF, and GM). But after a full month of walking, activity of the anterior muscles tend to disappear. On the other hand, reciprocal EMG (TA-, LG+ and RF-, BF+) patterns seen in childhood gait become more prevalent. Reversed reciprocal EMG (TA+, LG- and RF+, BF-) patterns disappear and are not seen in child and adult gait patterns. This suggests that excessive activity of the anterior muscles indicate marked instability, whereas excessively activity of the posterior muscles should be associated with a lesser degree of instability. In swing phase, up to the first month of walking, the VM (a knee extensor) is generally active from the middle of swing phase until the subsequent foot contact (Figs. 2-2 and 2-4). The LG (an ankle plantarflexor) is likewise active in this part of swing phase during about the first three months of independent gait (Figs. 2-2, 2-4, 2-6, and 2-8). Compared to the situation of standing on both feet, these patterns occur when only the contralateral leg is providing a very small base of support, and the airborne foot is being actively plantarflexed while the knee is being actively extended, suggestive of operation of the protective parachute reflex to prevent falling. It thus becomes clear that when a baby first begins to walk, muscle activity plays a relatively great role in providing stability to maintain posture and to keep the body's center of gravity low and within the base of support. From the early stages of walking, the muscles become stronger and balance matures as months and years pass, obviating the need for so much myoelectric activity. Thus some patterns of EMG activity can be identified that are present in infant walking but are subsequently no longer present in child or adult gait. As the baby matures, these excesses gradually become refined until, at about three years of age, they very much resemble muscle activities of adults.

42 Development of Gait

Table2-1. Developmental changes of EMG pattern during single leg support Joint

Ankle

EMG pattern

1st day

2 wks

1 mon

2-3 mons

Reciprocal (TA-, LG+)

(+)

(+)

(++)

(++)

Reversed Reciprocal (TA+, LG-)

(++)

(+)

(-)

(-)

Co-contraction (TA+, LG+)

(+)

(+)

(+)

(±)

Continuous (VM+)

(++)

(++)

(±)

(±)

Reciprocal (RF-, BF+ )

(+)

(+)

(++)

(++)

Reversed Reciprocal (RF+, BF-)

(±)

(±)

(-)

(-)

Co-contraction (RF+, BF+)

(+)

(+)

(±)

(±)

Continuous (GM + )

(++)

(++)

(++)

(++)

Knee

Knee & Hip

Hip

Frequency of occurrence, (++): very much, (+ ): much, (±): a little, (- ): little.

sw

ST

sw

ST

sw

ST

1 sec

I

0.5 mv

2 weeks after learning to walk

Fig. 2-1 4. Variations in EMG pattern of ankle joint muscles at 2 weeks after learning to walk (at 10.5 months of age). ST: stance phase, SW: swing phase, TA: tibial is anterior, LG: lateral gastrocnemius, Left (ST-l): two or three alternating bursts between the TA and LG, Center (ST-2): one or two alternating bursts between the TA and LG , Right (ST-3): one continuous discharge pattern.

Independent Waiking in infants 43

Conclusion To determine EMG characteristics of infant walking, we longitudinally recorded EMGs using surface electrodes from twelve muscles of both legs in an infant from 306 days after birth. Up to around 1 month after learning to walk, in stance phase the VM showed activity associated with holding a slightly flexed knee joint. Alternating reciprocal patterns between the RF and BF muscles came into playas the body inclined backward and forward, whereas a co-contraction pattern of both muscles appeared when the body was erect. Alternating reciprocal patterns between the TA and LG helped to maintain balance and to prevent falling backward or forward. Cocontraction patterns of these two muscles were seen to stabilize the ankle joint to maintain body balance, preventing strong forward falling. In the latter half of swing phase, the VM and LG showed strong activities with the knee extending and the ankle plantarflexing to prevent falling. These characteristically excessive discharge patterns of infant gait were not seen in subsequent childhood gait or in adult gait, and they began to decrease or disappear after about 1 month of learning to walk. It is in this sense that these leg muscle activities are considered EMG characteristics of infant walking at the onset of independent walking.

44 Development of Gait

Electromyographic (EMG) recordings of the lower limbs were made from a girl from 3 weeks after birth until 8 years of age to determine EMG changes in the development of human bipedal locomotion. Recordings were taken from the tibialis anterior (TA), lateral gastrocnemius (LG), vastus medialis (VM), rectus femoris (RF), biceps femoris (BF), and gluteus maximus (GM) muscles. In each of three developmental stages of gait, primitive walking, supported walking, and independent walking, muscle activity progressed from excessive co-contraction of mutual antagonists to reciprocal patterns. For the stance limb, the predominant reciprocal pattern to emerge was continuous activity of the posteriorly located LG and BF as opposed to the anteriorly located TA and RF In independent walking this preponderance of maintained activity by the LG and BF in stance phase gradually waned over the first 2 years of walking to focused bursts of activity. The developmental changes observed in this girl appear to have been attributable to changes in posture reflecting increased strength and to improvements in control of balance reflecting neuromaturation.

During the first three years of life, human bipedal locomotion develops gradually toward mature walking throughout a series of phases: newborn stepping, infant supported walking, infant independent walking, and child walking (Fig. 3-1). In the 20th century, some studies have provided detailed technical descriptions (kinematics, kinetics, temporal events, and electromyography) of the developmental process of infant locomotion, although to study gait in babies using adult techniques is very difficult. McGraw (1940) analyzed seven selected phases in the development of erect locomotion from newborn stepping to mature erect walking, using film analysis, and pointed out the relations between several reflexes and the development of motor behavior. Touwen (1976) clarified the interactions between reflexes and the development of motor behavior, emphasizing the longitudinal study of motor development. Using EMG can provide information about the maturation of gait that is both significant and otherwise unavailable in conventional motion analysis. Although the study of human locomotion in infants using EMG is difficult, some cross-sectional and longitudinal EMG studies on the development of gait have been done. Forssberg (1985), Thelen et al. (1987), and Okamoto et al. (1972, 1985,2001,2003), have studied the developmental process from newborn stepping until infant supported walking prior to independent walking, and Sutherland et al. (1980) and Okamoto et al. (1972, 1985, 2001, 2003) have researched the learning process from early infant independent walking to mature walking. These studies have generally described developmental changes of various leg muscular activities in both supported and unsupported walking. We are unaware, however, of any studies that have described EMG developmental changes from newborn stepping all the way to mature walking longitudinally in the same individual. The purpose of this study was to study longitudinal developmental changes of human locomotion in terms of leg muscle activity. EMGs of the same subject were recorded over a period of 8 years, from 3 weeks after birth to 8 years of age, so that the entire span of gait development could be examined in one individual.

46 Development of Gait

We made longitudinal observations on a female infant from 3 weeks after birth until 8 years of age. At the beginning, to induce newborn stepping, the examiner held the infant under the arms with the soles of the feet touching a horizontal flat surface. Well-coordinated walking movements were observed fairly consistently from shortly after birth to around 3 or 4 months. Although newborn stepping could not simply be arbitrarily elicited at the will of the examiner, we were able to induce selected well-coordinated walking movements of three or more steps during this period. From 3 weeks after birth to 3 years of age, EMGs were recorded 38 times, at intervals ranging from 2 weeks to 2 months. After that, from 3 to 8 years of age, EMGs were recorded 10 times, about every 6 months. Based on the longitudinal EMG findings of the present investigation, as well as those from previous studies (Okamoto et al.,1972, 1985, 2001), we divided the early development of gait into the following four phases: neonatal stepping, onset of young infant stepping, initial young infant stepping, and infant supported walking. Subsequent maturation of gait was also divided into four phases: onset of infant walking, initial infant walking, immature child walking, and mature walking. The data in Figures 3-2, 3-4, 3-5, 3-6, 3-7, and 3-8 show representative EMG patterns and forms from our longitudinal observations of the same subject (Figs. 3-1 and 3-3).

t1tft~f frffff

Itlrftfi

llUli ~ i I f~€e It~l~~ Jilit/fA j 1111l\~lk Fig. 3-1. Developmental changes of gait in one individual (birth to age eight). Top : neonatal and infant stepping, Middle: infant supported and independent walking, Bottom: child walking (same subject).

From Newborn Stepping to Mature Walking 47

Neonatal stepping (up to 4 weeks after birth)

TO

FC ••11 .

TA

..".

\." ..

LG VM

RF

BF

j,

..1

,J.

,.I..

~~

." ....,"

GM SWING (SW)

STANCE (ST) 1 sec

I 0.5 mv

3 weeks

Fig. 3-2. EMGs of neonatal stepping (at 3 weeks after birth). TO: toe off, Fe: foot contact, SW: swing phase (short phase), ST: stance phase (long phase), TA: tibialis anterior, LG: lateral gastrocnemius, VM: vastus medialis, RF : rectus femoris, BF: biceps femoris, GM: gluteus maximus.

Fig. 3-2 shows EMG patterns of leg muscles at 3 weeks after birth. The stepping in this period was characterized by quick hip and knee flexion in which the thigh became horizontal in the middle part of swing phase. The foot dorsiflexed strongly as it was brought forward. The foot then approached the floor more slowly, the knee extending relatively passively as the hip extended. The foot usually contacted the floor with the lateral border first, but sometimes the heel, sole, or forefoot made initial contact instead. The supporting leg was relatively flexed during stance phase. The TA, RF, and BF exhibited notable myoelectric activity as the ipsilateral foot was leaving the floor to begin swing phase. The TA continued to be active throughout much of swing phase, whereas the RF showed no more than sporadic weak activity during that period,

48 Development of Gait

and the BF was relatively silent until stance phase was being approached. The LG, VM, and GM did not show any remarkable activity during swing phase. During stance phase, the LG, BF, and GM showed relatively continuous activity as antigravity muscles. The VM and RF tended to be active when knee flexion was not very pronounced, that is, during the double-stance phases. In single-stance phase, activities of mutually antagonistic muscles (TA versus LG and RF versus BF) showed reciprocal (TA- and LG+, RF- and BF+) , cocontraction (fA+ and LG+, RF+ and BF+), and reversed reciprocal (fA+ and LG-, RF+ and BF-) patterns, but activities of certain mutually antagonistic muscles were variable and inconsistent (Fig. 3-12).

Fig. 3-3. Neonatal stepping at 3 weeks after birth.

From Newborn Stepping to Mature Walking 49

Young infant stepping (from 1 to 5 months of age)

ffffrJf Ir~~I~ TO

TA

Fe

TO

Fe

lilii! TO

Fe

U'"'

.1,

LG

VM II'

,j;..

RF BF

.,1, F

II

GM

.."

sw

1 sec

I 0.5 mv

1.5 months

SW

ST

SW

ST

1 sec

3.5 months

I 0.5 mv

ST

t sec

I 0.5 mv

5 months

Fig. 3-4. EMGs of young infant stepping (Left: at 1.5 months after birth, Center: at 3.5 months after birth, Right: at 5 months after birth).

1 ) Onset of young infant stepping (from 1 to 2 months of age)

Fig. 3-4 (left) shows EMGs at 1.5 months after birth. Step frequency was more regular than during neonatal stepping. Hip flexion was pronounced in the first part of swing phase as in neonatal stepping, but leg extension began to be more vigorous in the latter part of swing phase. The lateral border of the forefoot initially contacted the floor for the most part. In swing phase, continuous activity of the TA tended to terminate sooner in the latter part of swing phase than during the neonatal period, but bursts of the RF and BF during the phase showed the same tendencies as before. The LG and VM, which had been quiet during swing in neonatal stepping, began to exhibit activity shortly before foot contact, and sometimes activity in the GM was also observed before foot contact. During stance phase, continuous bursts or discharges of the VM were seen often, as in

50 Development of Gait

neonatal stepping. The posteriorly located LG, BF, and GM showed continuous discharges that were comparatively variable in intensity. In single stance, the reciprocal pattern was observed as in neonatal stepping. Incidence of the co-contraction pattern decreased from that at the time of neonatal stepping, and the reversed reciprocal pattern was hardly seen (Fig. 3-12). 2) Initial young infant stepping (from 3 to 5 months of age)

Fig. 3-4 (center) shows EMGs at 3.5 months after birth. Stepping in this period was performed actively as foot contact was audible. Hip flexion was again conspicuous in the first part of swing phase as it was in the neonatal period and in the onset of young infant stepping. Leg extension in the latter part of swing phase was performed in a more active manner than in the onset of young infant stepping, often quite visibly so. Initial contact of the floor was, for the most part, by the forefoot. The supporting leg, especially the knee, tended to be relatively extended in this period. In swing phase, the TA, RF, and BF became less active at toe off than they did in the onset period of young infant stepping. This was especially true for the TA, but that muscle would always exhibit at least a weak burst in the first part of swing phase. In the latter part of swing phase, discharge patterns of the LG and VM tended to increase frequently. During stance phase, continuous bursts or discharges of the VM were seen often, and the LG, BF, and GM showed continuous discharges of variable intensity, as they did in the onset of young infant stepping. Mutually antagonistic muscles in single stance exhibited the reciprocal pattern more readily than previously. The co-contraction pattern was hardly seen, and the reversed reciprocal pattern was not seen at all in this period. Fig. 3-4 (right) shows EMGs at 5 months after birth. Stepping was actually difficult to elicit at that age, but we did manage to induce it and the resulting EMG pattern was basically similar to the initial period of young infant stepping at around 3.5 months after birth.

From Newborn Stepping to Mature Walking 51

Infant supported walking (from 6 to 12 months of age)

TO

TO

Fe

ft~jtrtt ~~liil Fe

TO

Fe

"", 'I

TA+-~--------~~

.lIi,/"Il.

LG+-~~~~~~~

, d,~

'1"'1"'1"

'II'

,',!",'

VM +-~--------~~ RF+-~--------~~ .•1...., "11 "1

BF+-~~~~--~~

I~

I

GM+-~--------~~

sw

sw

ST

1 sec

\

0.5 mv

6 months

sw

ST

1 sec

9 months

.J

'\1'

11 I 0.5 mv

ST 1 sec

I

O.S my

11.5 months

Fig. 3-5. EMGs of infant supported walking (Left: at 6 months after birth, Center: at 9 months after birth, Right: at 11,5 months after birth),

Fig. 3-5 shows EMGs at 6, 9, and 11.5 months after birth. After around 6 months after birth, the infant began to crawl after toys and tended to show comparatively stable stepping when supported upright. At around 11 months, 1 month before walking independently, the infant became able to stand by herself and to walk with one-handed support. The relatively pronounced flexion of the hip seen in the first part of swing phase of the previous period was slightly reduced. Active leg extension tended to disappear in the latter part of swing phase. The heel usually contacted the floor first. The supporting leg was extended. In the first part of swing phase, activity of the RF was sporadic and bursts of the BF disappeared. In the latter part of swing phase, the LG and VM showed minimal activity at 6 and 9 months, but shortly before independent walking (at 11.5 months after birth), strong myoelectric discharges of the LG and VM were sometimes seen. During stance

52 Development of Gait

phase, the continuous activity of the VM seen in earlier stages of development no longer appeared. The LG and BF, antigravity muscles, were markedly active in stance phase throughout this period of supported walking (Fig. 3-6). Mutually antagonistic muscles during this period seldom exhibited co-contraction patterns, but rather worked in reciprocal patterns as a rule.

621

385 days

0.5 mv

TA

LG

VM RF 8F

GM KNEE

Ext. Flex.

FC~ 600

700

1st day of independent walking

Fig. 3-6. EMGs on the 1st day of independent walking (at 1 year of age). SUPPORT: supported walking, INDEPENDENT: independent walking.

From Newborn Stepping to Mature Walking 53

Infant walking (from 1 week to 2 months after learning to walk)

j~R'~~ i~A~~' TO

Fe

jl9Ajil TO

Fe

I

"I.

TA

,II. '·",1, IIIh~LL. [1'1'.1,

LG

1'1

VM RF

"" ~Ii••.

BF~~\~~~~~-+~' GM~~~~~~~--~~

sw

ill.

"~:~',r sw

ST 1 sec

0.5 mv

1 year

HI

111I~I'r

n

~

UI.._ TI~"

.11,

'I,'

sw

ST

1 sec

I

0.5 mv

1.1 years

ST

_-,-,1s"", ee'----.J

0.5 mv

1.3 years

Fig. 3-7. EMGs of learning process of infant independent walking (Left: 1 week after learning to walk at 1 year, Center: at 1 year 1 month, Right: at 1 year 3 months).

1) Onset of infant walking (up to 4 weeks after learning to walk)

Fig. 3-7 (left) shows EMGs of 1 week after learning to walk at around 1 year of age. The infant was able to walk more than 10 steps without external support. The walking in this period was characterized by quick hip flexion in which the thigh was raised forward diagonally in the first part of swing phase. Then the foot reached the floor quickly, the knee extending actively along with the hip. The foot usually contacted the floor with the forefoot or mid-sole first, but occasionally the heel made initial contact. Slight knee flexion was often observed in the supporting leg, the foot base in the double support period was very wide, and the body's center of gravity was lowered during stance. The arms were spread apart and elevated (Fig. 3-6).

54 Development of Gait

At the beginning of swing phase, the TA and RF were strongly active, often accompanied by moderate activity of the BF that had continued from the previous stance phase. Later on in swing phase, the LG, VM, BF and GM all generally became active and, just before foot contact the TA might exhibit a burst of activity. Stance phase was characterized by large amounts of activity in the muscles investigated, variously intermittent or continuous for a given muscle, so that mutual antagonists displayed not only reciprocal patterns in single stance, but also co-contraction and reversed reciprocal patterns (Fig. 3-12). 2) Initial infant walking (from 1 to 2 months after learning to walk)

Fig. 3-7 (center) shows EMGs of 1 month after learning to walk, at 1 year 1 month of age. The infant was becoming accustomed to walking and began to walk by herself for extended periods of time. The femur was not lifted so high in the middle part of swing phase as at the onset of independent walking. Leg extension in the latter part of swing phase was active. The forefoot initially contacted the floor for the most part, but occasionally the entire sole made initial contact. There was a diminution in base width and a gradual decrease in the slight flexion at the knee of the supporting leg, and the body's center of gravity was slightly higher during stance. The upper extremities, which had been extended and abducted earlier, now began to approach the body. At toe off, just as swing phase was beginning, most muscles exhibited the same patterns of activity as at the onset of independent walking 1 month before. During stance phase, activities of the muscles become less sporadic. Continuous discharge of the VM began to appear less and less (Fig. 3-9). Strong maintained activities of the TA and RF seen at onset of infant walking became markedly moderated, whereas in the LG and BF such activities remained unabated though more consistent in nature during stance phase (Fig. 3-9). In singlestance, the discharge patterns of mutual antagonists described in Figure 3-12 showed the reciprocal pattern more often than in the initial period of infant independent walking. The co-contraction pattern decreased comparatively, but was sometimes still seen, while the reversed reciprocal pattern was hardly seen at all in this period.

From Newborn Stepping to Mature Walking 55

Immature child walking Unsettled muscle activity (from 3 months to 2 years after beginning to walk)

AJAlA II ~ll l ~~U~ TO

TA

He

TO

He

TO

He

,.1

IH.

Ill'

IJI~

LG

1,/

' '''"

AL.

'lr'

'WI

W-

VM RF

BF

, '\ I"""

,,\,/,,, I'll

.....1

GM SW

sw

ST

1 sec

0.5 mv

2 years

ST 1 sec

3 years

SW

!

0.5 mv

ST

1 sec

I

0.5 mv

7 years

Fig. 3-8. EMGs of learning process of child walking (Left: at 2 years, Center: at 3 years, Right: at 7 years).

Fig. 3-7 (right) shows EMGs of the child 3 months after learning to walk, at 1 year 3 months of age. The infant had begun to shift to a comparatively stable walking pattern of her own. The thigh was no longer strongly lifted up in the first part of swing phase. Subsequent knee extension, previously prominent in the latter part of swing phase, now began to become more passive. The heel and toe began to touch almost simultaneously. The heights of the hip joints were higher than in infant walking, and the child began to exhibit more force to propel her body forward. The foot base was narrowed to the width of the shoulders. The upper extremities were still held away from the body, although they were only slightly elevated now. At the beginning of swing phase, the TA and RF worked in the same way at 3 months of independent walking as at 1 month. The VM no longer exhibited activity in the middle of swing phase after 3 months

56 Development of Gait

of walking, and activity in the LG was essentially absent throughout swing phase until the very end. During stance phase, activities in the leg muscles changed little between 1 and 3 months after beginning to walk, except for the GM, which exhibited less activity especially in the latter part of this phase. Figure 3-8 (left) shows EMGs at 2 years of age. The child had acquired a comparatively stable walking pattern at that time and was gaining control of movements related to running and fast walking. The thigh was no longer lifted up prominently in the first part of swing phase, nor did the lower leg extend rapidly in the latter part of swing phase. The heel and toe contacted the floor almost simultaneously. Slight knee flexion in the supporting leg was prolonged and the child propelled herself forward with a pumping action of the thighs, the trunk leaning forward slightly and the foot base narrowed. The forearms were slightly elevated during gait at ordinary speed. The TA was notably active as swing phase was about to begin, but neither the RF nor the BF exhibited activity as the foot left the ground. Just before the foot returned to the ground, the TA showed very little activity. During stance phase, continuous discharge patterns of the antigravity LG, BF, and GM were similar to those seen at 1 to 3 months after learning to walk. VM+

LG+

Terminal SW

Terminal SW

VM+

Ak ~rt 0-

Developmental period of gait lyear

H M+

Knee extension Plantarflexion

Onset of infant walking

-

(

~~:~ ~ :~~s_.~~r~~~n~n~ ~o_~.~)_____ Initial infant walking

Immature child walking

LG+, BF+

~F+

TA+

BFl

LO+

Squat

Backward

Forward

++)

(++ )

(++)

(++ )

(+ )

(_)

(+)

(_)

(_)

(++)

(-)

(-)

(-)

(-)

(++)

U -------n-------D.-------.[}-------. ----

(1-2 months after learning to walk)

(3 months - 2 years after learning to walk)

TA+, RF+

Throughout ST Throughout ST Throughout ST

~

Fig. 3-9. Developmental changes of EMG patterns in leg muscles from onset of infant walking to immature child walking. Frequency of occurrence, (++): very much, (+): much, (-): little.

From Newborn Stepping to Mature Walking 57

Mature walking Toward a mature pattern (after 2 years of learning to walk)

Fig. 3-8 (center and right) shows EMGs at 3 and 7 years. The child had begun to acquire stable walking resembling that of an adult in this period. The thigh showed minimal flexion in the first part of swing phase. The foot usually contacted the floor with the heel first and the toes lifted like an adult. Walking with the body inclined forward was seen until around the end of 2 years of age, when the body began to become more erect. The child exhibited strong pushing-off motions of the foot, and the upper extremities were no longer held in any degree of elevation. In swing phase, the TA was consistently active as the foot was leaving the ground, sometimes accompanied by slight activity of the RF. The TA also began to show marked activity just before the foot touched the floor in many instances (Fig. 3-10) . During stance phase, continuous activity of the LG previously found in the first half of stance phase decreased or disappeared and strong bursts were observed instead in the latter part of stance phase. Strong continuous discharges of the BF and GM previously seen in stance phase began to decrease or disappear, thus assuming activity patterns similar to those of adults. TA+ Just before ST

Developmental period of gait

X~TA+ Dorsiflexion

1.3 years Immature child walking (3 months - 2 years after lea rning to walk)

3 years Mature walking (after 2 years of learning t o walk)

(-)

•

(++)

LG+, BF+ Throughout ST

al

LG+Forward (++)

D-(-)

Fig, 3-10, Developmental changes of EMG patterns in leg muscles from immature child walking to mature walking. Frequency of occurrence, (++): very much, (-): little.

58 Development of Gait

Developmental period of gait

After birth

Development of gait

Developmental period of gait

Birth Neonatal reflex stepping (up to 4 weeks after birth) Neonatal stepping

1 month Onset of young infant stepping (1-2 months after birth)

2 months Young infant stepping (Inactive stepping)

6 months

~

,

Initial young infant stepping (3-5 months after birth)

Infant supported walking (6-12 months after birth)

Infant supported walking

1 year

1.1 years

Infant walking

Onset of infant walking (up to 4 weeks after learning to walk) Initial infant walking (1-2 months after learning to walk)

1.3 years

It

Immature child walking (3 months - 2 years after learning to walk)

Immature child walking

3 years

A

Mature walking (after 2 years of learning to walk)

Mature walking

Fig. 3-11 . Developmental period of gait during newborn stepping, infant supported walking, and independent walking.

From Newborn Stepping to Mature Walking 59

Neonatal st epping

Young infant stepping

Infant supported walking

Infant walking

Immature child Mature walking walking

:: ~ ~:-

Reciprocal pattern

:: ~~:

Co-contraction pattern

:::~:::

Reversed reciprocal pattern

Fig. 3-12. Developmental changes of EMG patterns in mutual antagonists (TA versus LG and RF versus BF) during ipsilateral single stance. TA: tibialis anterior, LG: lateral gastrocnemius, RF: rectus femoris, BF: biceps femoris,

+: noticeable activity, - : no activity.

Reciprocal pattern: posterior muscle is active while anterior muscle is inactive, associated with forward inclination of the trunk. At the hip, BF is active and RF is inactive. At the ankle, LG is active and TA is inactive.

Reversed reciprocal pattern: anterior muscle is active while posterior muscle is inactive, associated with backward inclination of the trunk. At the hip, RF is active and BF is inactive. At the ankle, TA is active and LG is inactive.

Co-contraction pattern: muscles on both sides of a joint are simultaneously active.

60 Development of Gait

Discussion During the first 3 years of life (Fig. 3-11), movements related to walking appear to begin with gross patterns of muscle activation, frequently including co-activation of mutual antagonists. Not only in supported walking and then in subsequent independent walking, but even in neonatal primitive walking one can see over time progression from excessive gross activation to more efficient and economical production of muscle activities in the lower limbs (Figs. 3-12 and 3-13). Such findings are evident in both stance and swing phases of the walking patterns. Interestingly, even some specific changes noted in newborn stepping over the first couple of months recur as the baby later masters voluntary supported walking and then independent gait. As the foot leaves the floor, activities of the TA, and to some extent the RF, remain relatively consistent across primitive, supported, and independent modes of walking. The latter two modes might thus be characterized as containing a "primitive" component at toe-off. Activity of the BF, on the other hand, at this same point in the gait cycle, varies from mode to mode as well as within a given mode. In both neonatal primitive walking and independent walking the BF is active along with the TA and RF when the child first performs these modes of locomotion, but the BF subsequently works in a reciprocal pattern with the other two muscles as the child gains experience. Just before the foot returns to the floor, the LG and VM are active at around 3 or 4 months after birth and again during the first few months of independent walking. The appearance of this activity coincides with gradual emergence of the parachute reaction, which Milani-Comparetti (1967) describes as appearing at about 4 months after birth, so we need to consider the possibility that activity of the LG and VM as the foot approaches the floor may be closely related to the parachute reaction at 3 or 4 months. As supported walking becomes more voluntary in subsequent months, this activity of the LG and VM is no longer manifest, nor is it seen 1 to 3 months after first learning to walk independently. The absence of such muscle activity appears to reflect development of balance and postural control. These changes in activity of the LG and VM around floor contact might be interpreted as development from simple reflexes and subcortical motor responses through cortical inhibition of these reflexes to a growing influence of voluntary or cortical motor control. From Newborn Stepping to Mature Walking 61

The manner in which the foot contacts the floor undergoes a similar progression in both supported walking and independent gait. At first the forefoot initially contacts the floor, but as development progresses, the sole of the foot makes initial contact with the floor and subsequent to that the heel makes initial contact. A burst of activity from the TA just before touchdown becomes more distinct as this sequence proceeds, so it might be interpreted as an indicator of stability in gait. In this connection, other behaviors change during this progression that likewise reflect incremental achievement of stability in gait. For example, width of foot placement gradually decreases as the walking pattern becomes more stable, and a "high guard" position of abducted arms becomes "medium guard" and eventually an adult-like "low guard" in the process. After the foot contacts the floor, the muscles in that lower limb can be subjected to greater loads than is possible in swing phase. This is particularly apparent in the VM. In the primitive mode of gait, the VM exhibits considerable activity as the infant pushes the foot against the floor in the extension phase of the primitive pattern. From 6 months, however, the baby is in a supported mode of gait wherein the VM no longer impulsively pushes against the floor and the baby simply relies on the supporting person to bear weight during gait. The VM is not very active even at 11 months, when the baby is close to graduating from the supported mode to independent walking, apparently because she has learned to passively bear weight through the knee on the stance side when the center of gravity of the superincumbent body segments has been brought anterior to the knee joint. At the beginning of the independent mode of gait, however, the situation dramatically changes as the baby suddenly finds herself solely responsible for both maintaining balance and bearing weight. When the foot contacts the floor and the lower limb on that side accepts body weight, the knee remains slightly flexed, presumably to keep the center of gravity of the superincumbent segments low, and thus make the task of balancing easier. Only after gait in the independent mode has progressed to the point when the baby can skillfully bring her weight over and just anterior to the knee on the stance side, can the VM display a brief focused burst during weight acceptance and otherwise be silent or minimally active in stance phase. As the baby moves forward from double support to single support in stance phase, she encounters a more sophisticated task of dynamically maintaining balance. The interplay between the LG and

62 Development of Gait

TA, mutual antagonists at the ankle, as well as between the RF and BF, mutual antagonists at both the hip and the knee, becomes very important during single support. When a baby first begins to walk without external support, co-contraction between each pair of muscles provides gross stability to make this difficult task feasible for the uninitiated. Because this co-contraction pattern first arises from trial and error, the other possibilities of reciprocal and reversed reciprocal patterns also appear (Fig. 3-12). Interestingly, all three of these patterns appear in primitive walking as well, and we have observed that the reciprocal pattern at this very early stage tends to appear when the trunk is inclined forward and the reversed reciprocal pattern when the trunk is leaning back, suggesting that muscles naturally respond to mechanical loading. Since walking can be characterized as a succession of incomplete forward falls, the reciprocal pattern eventually becomes more dominant than the co-contraction pattern as the baby attains dynamic stability in walking forward. By 3 years of age, the refinements of activity between these mutual antagonists are fully in place and the child propels herself forward with precisely measured doses of muscular activity and full dynamic control of balance. One finding to emerge from longitudinal observations was that developmental changes and refinements of excessive muscular activity during newborn stepping and supported walking appear again during the learning process of independent walking. As strength and balance improve in a normal infant, unnecessary muscle activation disappears leading to a series of developmental stages of bipedal locomotion in both supported and unsupported walking (Fig. 3-13). We suggest that the refinement of excessive co-activation, which can serve as a barometer to indicate increasing level of skill in human locomotion, comes from changing posture by improvement of strength and balance control reflecting neuromaturation.

From Newborn Stepping to Mature Walking 63

After birth

Birth

RF+ Initial SW

Dorsiflexion

I".

Hip flexion

Neonatal reflex stepping (up to 4 weeks after birth)

(++)

(++)

Onset of young infant stepping (1-2 months after birth)

(++)

(++)

(++)

(++)

(++)

(+),(-)

Onset of infant walking (up to 4 weeks after learning to walk)

(++)

(++)

Initial infant walking (1-2 months after learning to walk)

(++)

(++)

Immature child walking (3 months - 2 years after learning to walk)

(++)

(+),(-)

(++)

(+),(-)

Development of gait

Developmental period of gait

1

,

TA+ Initial SW

~~.

Neonatal stepping 1 month

2 months

................................ _.....

Initial young infant stepping (3-5 months after birth)

Young infant stepping (Inactive stepping) 6 months

i

Infant supported walking (6-12 months after birth)

Infant supported walking 1 year

1.1 years

I

Infant walking 1.3 years

I

......................... _.......

Immature child walking

3 years

A

Mature walking

(after 2 years of learning to walk)

Mature walking

Fig. 3-13. Developmental changes of EMG patterns in leg muscles during newborn stepping, infant supported walking, and independent walking.

64 Development of Gait

BF+

VM+

LG+

Initial SW

Terminal SW

Terminal SW

"'~

If L.,

Knee flexion Knee extension Plantarflexion

TA+

VM+

TA+, RF+

LG+, BF+

Just before ST Throughout ST Throughout ST Throughout ST

~,~ Dorsiflexion

-};"

~F+

TA+

BFl LG+

Squat

Backward

Forward

(++)

(-)

(-)

(+),(-)

(++)

(+)

(+)

(++)

(+)

(+)

(-)

(++)

(-)

(+)

(+)

(++)

(++)

(-)

(++)

(-)

(++)

(-)

(-)

(-)

(-)

(- )

(-)

(++)

(+),(-)

(++)

(++)

(+),(-)

(++)

(++)

(+)

(+),(-)

(-)

(+)

(+),(-)

(-)

(-)

(++)

(-)

(-)

(-)

(-)

(-)

(-)

(++)

(-)

(-)

(-)

(++)

(- )

(-)

(-)

TA: tibialis anterior, RF: rectus femoris, BF: biceps femoris, VM: vastus medialis, LG: lateral gastrocnemius,

SW: swing phase, ST: stance phase. Frequency of occurrence, (++): very

much, (+): much, (-): little, (+),(-): instances of noticeable activity and of no activity intermingle.

From Newborn Stepping to Mature Walking 65

Application to EMG biofeedback training

Infant independent walking at 1 year of age

To develop an index of gait instability from electromyographic (EMG) information, we made observations on infants from the time they first began to walk independently at about 1 year of age until around 3 years of age. From our findings we obtained the following criteria. (1) Very unstable gait: As seen in a child within the first month of learning to walk, the vastus medialis is active in the latter half of swing phase, the tibialis anterior and rectus femoris are active during stance phase, and activity of the vastus medialis is continuous. These EMG characteristics are not usually seen in subsequent childhood gait or in adult gait, and they serve as markers of very unstable gait. (2) Unstable gait: Activity of the gastrocnemius in the latter half of swing phase is generally noted only within the first 3 months after the child learns to walk, and that activity is interpreted as a sign of unstable gait. (3) Slightly unstable gait: Activity of the gastrocnemius in the first half of stance phase and the continuous activities of the biceps femoris and gluteus maximus from initial contact with the floor until push off are found in children until 3 years of age. These activities are considered EMG markers of slightly unstable gait.

Generally a baby becomes able to perform bipedal upright walking at about 1 year of age. Compared to quadrupedal crawling, walking involves maintenance of a mechanically unstable upright position and keeping one's balance while transporting the body's center of gravity. This requires a highly developed antigravity mechanism and operative balance reactions. Thelen et al. (1989) noted that independent walking emerges when a threshold has been reached for muscle strength and ability to balance, but the baby who has just become able to walk independently exhibits a pattern notably different from adult gait. McGraw (1940) and Okamoto et al. (1985, 2001, 2003), for example, have both noted that, although initial contact of the heel on the floor can be found in babies first learning to walk, as might be found in adult gait, they also often contact the floor first with the forefoot, which is not characteristic of adult gait at all. Novice walkers have been noted to have several characteristics that differ from adult gait, such as increased cadence, decreased step length, excessive rotation or flexion in stance phase, a pattern of circumduction in place of the hip and knee flexion that arises (in the adult) at the very beginning of stance phase, a lack of accompanying arm movement, or strong co-contraction or other form of muscle hyperactivity. McGraw (1940), Burnett et al. (1971), and Okamoto et al. (1985, 2001, 2003) have analyzed the development of independent walking and noted a regular progression from an initially wide base at the feet and a "high guard" position of abducted arms as evidence of instability in walking, giving way to a slightly narrower base and "medium guard" posture as the walking becomes more stable, finally to an adult-like "low guard" posture associated with stable gait. The mechanism of the motor development of babies has usually been studied in relation to postural development, often with, for example, the motor development evaluation form of Milani-Comparetti et al. (1967), which relates primitive reflexes to postural and motor development. To study gait in babies using electromyography (EMG) , goniometers, and force plates as is done with adults is very difficult, although some investigators have used film analysis to study the development of gait. Longitudinal study of motor development, emphasized by Touwen (1976) as very important, has been rare. McGraw (1940) studied the relations between several reflexes and the development of motor behavior, and Touwen (1971, 1976) has clarified the interactions between reflexes and the development of motor

70 Application to Gait Analysis and Evaluation

behavior. Cross-sectional kinesiological EMG studies on the development of infant walking have been performed by Sutherland et al. (1980), Forssberg (1985), and Thelen et al. (1987), but we have not seen much in the way of longitudinal EMG study on the acquisition of gait outside of that by Okamoto et al. (1972, 1983, 1985, 2001, 2003).

TA

LG

RF

SF

.EMG patterns began to decrease or disappear at about 1-3 months after learning to walk •

EMG patterns began to decrease or disappear at about 3 years of age

D

EMG patterns began to appear at about 3 years of age

IMMATIURE CHILD WALKING PATTERN

.+11]+0 11]+0

MATURE ADULT WALKING PATTERN

0

IMMATURE INFANT WALKING PATTERN

Fig. 4-1. Schematic diagram of EMG activity as indication of unstable walking. TA: tibialis anterior, LG: lateral gastrocnemius, VM: vastus medialis, RF: rectus femoris, BF: biceps femoris, GM: gluteus maximus.

An Index of Gait Instability 71

Knowing the developmental changes in the EMG features of the process of acquiring gait in normal babies should be helpful in better understanding the diagnosis and treatment of abnormal gait in developmentally delayed children. EMG enables one to view the ability to walk and to maintain balance and control during gait in ways that would not be noticed otherwise. Even for the elderly individual, the stages of motor development can be considered in relation to therapeutic exercise for maintaining gait function. To examine the role of muscle activity in the developmental process of normal gait in babies, we used EMG with surface electrodes to longitudinally study babies from the very unstable stage of first learning to walk at 1 year of age to the attainment of adult-like stability in gait at about 3 years of age. In addition to reconfirming the forward lean with shuffling in the early stages of gait and the more erect gait with decisive push off in stable adult-like gait, we were able to derive from these longitudinal EMG observations an index for the instability of gait based on developmental processes (Figs. 4-1 and 4-2) .

••

••+

·· ..

···· ···

......

+•

INFANT PATTERN

++.

.

••

·:......_-..A IMMATURE INFANT WALKING

IMMATURE CHILD WALKING

MATURE ADULT WALKING

Fig. 4-2. Refinement of excessive muscular activity during the development of gait. Excessive muscle activities in infant walking are considered to express weak muscle strength and an immature balancing system. As months and years pass, the muscles become stronger and balance matures, obviating the need for so much myoelectric activity.

72 Application to Gait Analysis and Evaluation

EMG findings during the development of gait The subjects were three babies who first began to walk independently at about 1 year of age (at 306 days, 375 days, and 385 days after birth). We made longitudinal observations on these three children from the time they first began to walk independently at about 1 year of age until a stable adult-like walking pattern was achieved at around 3 years of age. EMGs were recorded every 1 or 2 weeks in the initial period of independent walking and after that every 1 or 2 months. For purposes of comparison, these observations were supplemented with cross-sectional observations taken from fifteen babies in the first year of age (five infants at the onset of independent walking, five at 1 month after learning to walk, and five at 3 months after learning to walk), five babies in the second year, five more in the third year of age, and five adults. To more precisely search for EMG characteristics of gait stability, we also observed maintenance of standing posture in infants in the initial period of learning to walk.

1. Learning process of walking Fig. 4-3 shows longitudinal changes in EMG activity patterns in the learning process of walking for subject A, who began to walk on the 385th day after birth for the first time. Fig. 4-4 shows EMGs in the learning process of walking for subject B, who began to walk on the 306th day. In all subjects tested, excessive muscular activities appeared from the initial period of learning to walk until around 3 years of age, as compared with the corresponding muscular activities of adult walking. In the description that follows, we focus attention on the peculiar EMG activity patterns seen in the infant that deviate from normal adult walking. We looked for EMG patterns that gradually changed and were refined in the transition between first learning to walk and achieving an adult-like pattern, finding such refinements principally in stance phase and in the latter part of swing phase. (See chapters 2 and 3, Figs. 4-1 and 4-2) (Okamoto et al. 1972, 1983, 1985,2001,2003).

An Index of Gait Instability 73

TO

TO

12.5 months

Fe

Fe

1 year 1 month

TA (Tibialis anterior) LG (Gastrocnemius) VM (Vastus medialis) RF (Rectus femoris) BF (Biceps femoris) GM (Gluteus maximus) +--4I+
ST

ST

1 sec

I 0.5 mv

TO

He

it

He

"

T A '- I'--NIH'I' LG

,.......-\IrIoI~'f­

~1IlM-v,~II'II-f---..,-llMi\YiI

2 years 1 month

11

_----'1'-',:::,0'---------110.5 mv 1 month after learning to walk (A-2)

1st day of learning to walk (A-l)

TO

SW

He

TO

l

2 years 11 months

I--'

L

RF ........O-+\\I>.-.....-Ho!--.,.."..\O__

BF

ifi---I---IIU/If,-I-/IfM'-

GM

WIir"---t~
ST 1 sec

j,t

'['I

T

I'

'"

SW 10.5 mv

3 months after learning to walk (A-3)

T

[I" ST

SW

_ _1:.;':::'::.. 0 _ --,10.5 mv 2 years 1 month (A-4)

ST

_ --,Ic:,::::.o, ---,p.S mv 2 years 11 months (A-5)

Fig. 4-3. EMGs of learning process of infant walking (subject A). Top: Left (A-1); 1st day of learning to walk (at 12,5 months), Right (A-2); 1 month after learning to walk (at 1 year 1 month), Bottom: Left (A-3); 3 months after learning to walk (at 1 year 3 months), Center (A-4); at 2 years 1 month, Right (A-5); at 2 years 11 months.

74 Application to Gait Analysis and Evaluation

11 months ~'"'I..J - -

T A (Tibialis anterior) LG (Gastrocnemius) VM (Vastus medialis) RF (Rectus femoris) B F (Biceps femoris) GM (Gluteus maxim us) -+--'~""'i\III~~~-t---"""+-1h'-"""--l SW

ST

ST

_---',-" ,,"'c_

SW

2 weeks after learning to walk (8-1)

1 month after learning to walk (B-2)

TO He

3 years 5 months

Jil.

TA 1-lI\I\It~~.~~ LG

10.5 mv

1 sec

-'10.5 mv

1""1 -,.~"

1!t\l't-l'/li{W-+~/Wo--r-+

1".,11.. II'r'

r,1I'

'I'~j'

,.",1. "",.

VM -+~;"'--+---f¥--+­ RF

SW

ST

_--,-I="~C_~I 0.5 mv

3 months after learning to walk (8-3)

SW

ST

SW

ST

1 sec

_-----"-.::,=.:'c'-----'10.5 mv 1 year 9 months (8-4)

/0.5 mv

3 years 5 months (8-5)

Fig, 4-4. EMGs of learning process of infant walking (subject B). Top: Left (B-1); 2 weeks after learning to walk (at 10.5 months), Right (B-2); 1 month after learning to walk (at 11 months). Bottom: Left (B-3); 3 months after learning to walk (at 1 year 1 month), Center (B-4); at 1 year 9 months, Right (B-5); at 3 years 5 months.

An Index of Gait Instability 75

2. Standing posture at the onset of independent walking

SF --''''*~''''''''''''ri-+1...../ -.....................'-~-tlf',""'-Ifc.. ~\I\,~,~~,,,,,,,,~,,,,\",,I''1'1<~'''''''''---~-+-

~~\Wlu.'tI~lil'

GM

KNEEEXT.~_--_ _ _ _ _ _ _ _ _ - - -FLEX.

-------

.-----,

FC--------------" (Foot Contact)

Foot Flat

~

Toe Contact t sec

Foot Flat

I 0.5 mv

Heel Contact 1 sec

I 0.5 mv

STANDING POSTURE

1st day of learning to walk

2 weeks after learning to walk

Fig, 4-5, EMGs of standing posture at the onset of independent walking, Left: standing posture with pronounced squat on the 1st day of independently walking (subject B, at 10 months), Right: standing posture with slight squat at 2 weeks after learning to walk (subject B, at 10,5 months), Foot Flat: foot flat with the body erect, Toe Contact: toe contact with the body inclined forward, Heel Contact: heel contact with the body inclined backward,

1) On the 1st day of independent walking

Fig. 4-5 (left panel) shows EMGs of standing posture on the 306th day after birth, just as the baby was starting to walk for the first time (subject B, same as in Fig. 4-4). During maintenance of standing posture, at the ankle, alternative bursts between the TA and LG generally showed a reciprocal pattern. At the knee, the VM showed continuous strong activity during the maintenance of standing posture. At the hip and knee, three discharge patterns (reciprocal, reversed reciprocal, and co-contraction) between biarticular muscles (the RF and BF) could be seen, similar to those reported for stance phase at the onset of independent walking (Figs. 4-3 A-1 and 4-4 B-1).

76 Application to Gait Analysis and Evaluation

2) At about 2 weeks after learning to walk

Fig. 4-5 (right panel) shows EMGs of standing posture at about 2 weeks after learning to walk, on the 318th day after birth (subject B, same as in Fig. 4-4). During standing, at the ankle, alternative reciprocal activity between the TA and LG began to decrease or disappear. More specifically, strong continuous discharge patterns of the TA tended to decrease or disappear, whereas the LG showed continuous activity. At the knee, activity in the VM tended to decrease or disappear. At the hip and knee, whereas the reciprocal pattern increased, the reversed reciprocal and co-contraction patterns tended to decrease or even disappear compared to what was seen at the onset of independent walking. Discharges of the RF tended to decrease or disappear while those of the BF generally increased. The GM showed continuous discharges, similar to the pattern on the 1st day of independent walking. However, when the infant maintained balance with the body inclined backward, losing her body balance control momentarily, strong bursts of the TA and RF emerged (Fig. 4-5, right panel, just before heel contact). KNEE EXTENDED

KNEE JOINT

BACKWARD

ERECT

KNEE FLEXED

FORWARD

BACKWARD

ERECT

FORWARD

STANDING

POSTURE

L

'--

Tibialis anterior

(+)

(- )

Gastrocnemius

(-)

(-)

(+)

(+),H

(-)

H.C±l

(+)

(-)

Vastus med ialis

(+)

(+),H

(+)

(-l.C±)

(- )

(+)

(+l.C±)

(+)

Rectus femoris

(+)

(-)

(-)

(+)

(+),(±)

(-)

Biceps femoris

(-)

(-l.C±)

(+)

(-)

(-)

(+)

Gluteus maximus

(+)

(+)

(+)

(+)

(+)

(+)

Fig. 4-6. EMG patterns in leg muscles of standing posture at the onset of infant independent walking. (+): noticeable activity, (-): no activity, (±): slight activity, (+),(-): instances of noticeable

activity and of no activity intermingled.

An Index of Gait Instability 77

3. Supported walking

TO Fe

TC FC ~ 1'j1-1"

""\...J"""'"""r-'""r'""

r-~ ~\.

TA

W,

~1

LG

....... ~I" Ji.!..

10.5 months ~

. ~

tuh. 111",

1,1

1m '"

VM RF ~1.

,I

SW

ST

ST

SW

SW

ST 1 sec!

2 weeks after learning to walk

INDEPENDENT GAIT

111/

l'

0.5 mv

SUPPORTED GAIT

SUPPORTED GAIT

(FORWARD SWAY)

(ERECT POSTURE)

Fig. 4-7. EMGs of independent and supported walking at the onset of learning to walk (subject B, at 10.5 months). Left: independent walking at 2 weeks after learning to walk, Center: supported walking with the body inclined forward (immature child walking pattern), Right: supported walking with the body erect (mature adult walking pattern).

Fig. 4-7 (center panel) shows EMGs of supported walking with the body inclined forward. The right panel of this figure shows supported walking with the body erect at about 2 weeks after learning to walk in a 10 month old baby (subject B, same as in Fig. 4-4) . When the infant walked with external support, strong discharges of the LG and VM disappeared in the latter half of swing phase, and continuous discharges of the TA, VM, and RF decreased or disappeared during stance phase. With forward sway of the trunk, the excessive activities in the LG, BF, and GM were seen during stance phase, and these discharge patterns closely resembled the independent walking pattern of a child usually seen from 3 months to 3 years of age. With the body erect, discharges of the LG in the first part of stance phase as well as activities of the BF and GM in stance phase tended to decrease or disappear, closely resembling the stable adult walking pattern (Fig. 4-7, right panel).

78 Application to Gait Analysis and Evaluation

EMG activity in unstable walking Based on EMG studies on the development of independent gait in babies, Okamoto et al. (1972, 1983, 1985, 2001, 2003) (Fig. 4-1), Kazai et al. (1976) and Sutherland et al. (1980) have reported that specific changes can be observed at certain times in the course of that development. In the present study we have confirmed, by means of both longitudinal and cross-sectional EMG and cinematographic findings, that during the first month of independent walking, a baby squats slightly while leaning forward and takes steps with strong active extension of the legs, exhibiting considerable instability. After about 3 months of independent walking, a baby exhibits increased stability with the body tilted slightly forward, and by 3 years of age the body is upright as in adult walking. The activities of muscles during these different stages show a similar transition from instability to stability. At first, when the baby is just beginning to walk, characteristic discharge patterns can be seen that are excessive when compared to the corresponding patterns in adults. As the baby matures, these excesses gradually become refined until, at about 3 years of age, they very much resemble muscle activities of adults. As mentioned at the beginning of this chapter, we have been able to derive an index of instability in walking based on our EMG findings and previous studies (Okamoto et al. 1972, 1983, 1985, 2001, 2003) on the developmental process of the acquisition of gait. We did this by identifying EMG patterns found in the very unstable walking of babies first learning to walk but not seen in the stable walking of adults. We looked for EMG patterns that gradually changed and were refined in the transition between first learning to walk and achieving an adult-like pattern, finding such refinements principally in stance phase and in the latter part of swing phase.

1. Stance Phase ST-TA, ST-RF: Bursts or continuous activity of the tibialis anterior

during stance phase (ST-TA) and activity of the rectus femoris in synchrony with the tibialis anterior (ST-RF) are normally seen during the very unstable period of the first month after beginning to walk (Figs. 4-3 A-1 and 4-4 B-1), but not thereafter (Figs. 4-3 A-2 and 4-4 B-2).

An Index of Gait Instability 79

When enough external support is given to the infant during stance phase at the onset of independent walking, strong discharges of the TA and RF tend to decrease or disappear (Fig.4-7, same subject as in Fig. 4-4 B-1). On the other hand, we noticed that when one infant maintained balance with the body inclined backward, losing her body balance control momentarily during stationary standing at the initial period of independent walking, strong bursts of the TA and RF emerged (Fig.4-5, right: just before heel contact). These findings suggest that the ST-TA and ST-RF patterns help control displacement of the body's center of mass by participating in maintenance of posterior inclination of the trunk. ST-VM: Continuous discharges of the VM during stance phase are seen until around 1 month after learning to walk in many instances, as well as those of the TA and RF (Figs. 4-3 A-1 and 4-4 B-1), as mentioned above. After the first month of walking, continuous discharges of the VM tend to decrease or disappear (Figs. 4-3 A-2 and 4-4 B-2). That such activity is related to instability is further suggested by disappearance of the ST-VM pattern when enough external support is given to the baby to make the slightly squatted position unnecessary (Fig. 4-7, same subject as in Fig. 4-4 B-1), although the ST-VM pattern is seen in the initial period of independent walking when support is withdrawn (Fig. 4-4 B-1). Considering stationary standing, the VM is continuously active as the baby stands fairly squatted on the 1st day of independent walking (Fig. 4-5, left), whereas 2 weeks later the squatting is much shallower and activity of the VM is minimal or absent (Fig. 4-5, right). This agrees with observations by Okamoto et al. (1985, 2001, 2003) that the load at the knees decreases as strength and balance develop. The ST-VM activity seen at the onset of independent gait thus appears to contribute to holding a posture with slight knee flexion, permitting the body's center of gravity to be lowered so that balance is easier to maintain. ST-LG, ST-BF, ST-GM: Activity of the gastrocnemius in the first half of stance phase (ST-LG) and continuous discharges of the biceps femoris and gluteus maximus during stance phase (ST-BF, ST-GM) are found in the slightly unstable gait typically seen from 1 month after having learned to walk until about the middle of 2 years of age (Fig. 4-3 A-2, A-3, A-4, 4-4 B-2, B-3, and B-4). After that these activities are no longer seen, as the pattern of walking closely resembles that of an adult (Figs. 4-3 A-5 and 4-4 B-5). A look at muscle activities when external support is given to a child who has just begun to walk

80 Application to Gait Analysis and Evaluation

independently (Fig. 4-7) reveals that when the support is provided with the trunk inclined forward, the ST-LG, ST-BF, and ST-GM patterns appear as they would in an independently walking child with at least 3 months of walking experience but not yet 3 years of age (Fig. 4-7, center). If support is provided so that the trunk is upright, on the other hand, the activities found with the trunk inclined forward are not present and the results look more like mature adult gait (Fig. 4-7, right). These findings suggest that the ST-LG, ST-BF, and ST-GM patterns help control displacement of the body's center of mass by participating in maintenance of anterior inclination of the trunk. Before strength and balance have matured to the point that push off can be effectively used with the trunk upright, as in adult gait, these three muscle activation patterns are considered to be necessary for gait with an anteriorly inclined trunk (Figs. 4-5, 4-6, 4-7, and 4-8). We have found that excessive muscular activity in the stance phase of gait in a child who has just begun to independently walk strongly resembles lower limb activity during maintenance of an upright standing posture for the same period of development (Figs. 4-3 A-l, 4-4 B-1, 4-5 left, and 4-6), suggesting that a common mechanism operates both in standing and in the initiation of gait. From a mechanical point of view, at this very early stage, both activities require a low center of gravity and a wide base of support to assure maximum stability. Generally these tasks can be accomplished, even though strength and balance are yet undeveloped by spreading the legs apart to widen the base of support and by maintaining the knees in flexion to lower the center of gravity. In our study, the role of the uniarticular vastus medialis for maintaining stability became clear as a slightly squatted position was used to lower the center of gravity. Another important factor to consider is keeping the vertical projection of the body's center of gravity well within the bounds of the base of support. In our study, babies who had just begun to walk independently exhibited two-way control over inclination of the trunk during walking or standing, thus keeping the center of gravity within the base of support, by orderly patterns of activity in the biarticular rectus femoris and biceps femoris (Figs. 4-3 A-l and 4-4 B-1). While these two muscles act at the hip and knee, Nashner et al. (1985) have pointed out that ankle strategy is the most efficient for returning the body's center of mass to its initial position. Indeed, in our study the gastrocnemius and tibialis anterior exhibited reciprocal patterns of activity, thus affording anteroposterior control over the center of An Index of Gait Instability 81

gravity to help maintain upright stability. The fact that this EMG pattern becomes attenuated in standing after 2 weeks' experience of walking (Fig. 4-5) further suggests that it is a characteristic feature of balance control when the baby takes steps for the very first time. In some instances these mutual antagonists co-contract, indicative of maintaining balance at the ankle by strongly stabilizing the ankle, as pointed out by McGraw (1940) (Fig. 4-4 B-1). These findings illustrate how intricately the muscles across the ankle, knee, and hip joints contribute to maintenance of balance during stance phase in the initial period of independent walking (See ST in Fig. 4-10). During this most unstable period, the anteriorly located muscles of the lower limb (TA, VM, RF) are just as active as the posteriorly located muscles (LG, BF, GM), but after a full month of walking, balance has matured to the point that activity of the posterior muscles tends to become more dominant. This suggests that excessive activity of the anterior muscles should indicate marked instability, whereas excessively strong activity of the posterior muscles should be associated with a lesser degree of instability. When gait is performed without much activity even from the posterior muscles, a high degree of intrinsic stability can be inferred to be present.

TO

TA

He

TO

,.

He

~

.il.

.,111.

LG

"1"

-"I

VM RF

BF

~I

~~ ,

I

GM sw

ST

I !

""1 u/u>

'"""T"

sw

ST

ERECT POSTURE

sw

ST

_ _' ,_ec_ --,'

ADULT WALK

FORWARD

0.5

FORWARD WITH KNEE FLEXED

Fig. 4-8, EMGs of adult walking under differing conditions,

82 Application to Gait Analysis and Evaluation

mv

2. Swing Phase SW-LG, SW-VM: Up to the first month of walking, the vastus medialis is active from the middle of swing phase until subsequent foot contact (SW-VM, Figs. 4-3 A-1, A-2, and 4-4 B-1) in many instances. The gastrocnemius is likewise active in this part of swing phase for about the first 3 months of independent gait (SW-LG, Figs. 4-3 and 4-4) . As seen in studies by Okamoto et al (1985) and by Kazai et al (1976) as well as in the present investigation, even in the most unstable period of the onset of independent walking (Fig. 4-4 B-1) the provision of external support turns off the SW-LG and SW-VM patterns (Fig. 4-7), implying that these two patterns of muscle activity are definitely associated with intrinsic instability in gait. Compared to the situation of standing on both feet, these patterns occur when only the contralateral leg is providing a very small base of support, and the airborne foot is being actively plantarilexed (SW-LG) while the knee is being actively extended (SW-VM), suggestive of the operation of the protective parachute reflex to prevent falling (Fig. 4-9).

j{/i~jJffM'Ai 1

2

@

4

5

STANCE

6

7

8

9

10

11

SWING

LG --;-'.

KNEEFLEX.

FC~L--li

VTR -

I

HC FF

I

1

HO

I

@

+

STUMBLE

TO

FF HO

lit-

5

7

+

10

+11

FOOT CONTACT

Fig. 4-9. EMGs of leg extensors after a stumble at 3 years of age. Activities of the LG (an ankle plantarflexor) and VM (a knee extensor) in the latter part of swing phase (before foot contact) tended to increase after a stumble (VTR @).

An Index of Gait Instability 83

Criteria for Instability

ST-TA

VERY UNSTABLE

UNSTABLE

TRANSITION TO ADULT WALKING PATTERN

II

EMG patterns began to decrease or disappear at about 1month after learning to walk

EMG patterns began to decrease or disappear at about 3 month s after learning to walk

D

EMG patterns began to decrease or disappear at about 3 years of age

Fig. 4-10. Schematic diagram of EMG activity as indication of unstable walking. ST: stance phase, SW: swing phase, TA: tibialis anterior, LG: lateral gastrocnemius, VM: vastus medialis, RF: rectus femoris, BF: biceps femoris, GM: gluteus maximus.

84 Application to Gait Analysis and Evaluation

It thus becomes clear that when a baby first begins to walk, muscle

activity plays a relatively great role in providing stability to maintain posture and to keep the body's center of gravity low and within the base of support. As months and years pass, the muscles become stronger and balance matures, obviating the need for so much myoelectric activity. Thus some patterns of EMG activity can be identified in the early stages of walking that are no longer present in adult gait, and they can be validly associated with intrinsic instability. More specifically, these "excessive" patterns of myoelectric activity can be classified into the following categories of association with instability in gait (Fig. 4-10, Table 4-1). (1) Very unstable gait: As seen in a child within the 1st month of

learning to walk the vastus medialis is active in the latter half of swing phase, the tibialis anterior and rectus femoris are active during stance phase, and activity of the vastus medialis is continuous. These EMG characteristics are not usually seen in subsequent childhood gait or in adult gait, and they serve as markers of very unstable gait. (2) Unstable gait: Activity of the gastrocnemius in the latter half of

swing phase is generally noted only within the first 3 months after the child learns to walk, and that activity is interpreted as a sign of unstable gait. (3) Slightly unstable gait: Activity of the gastrocnemius in the first half

of stance phase and continuous activities of the biceps femoris and gluteus maximus from initial contact with the floor until push off are found in children until 3 years of age . These activities are considered EMG markers of slightly unstable gait. As a practical application, we have been able to apply this EMG index of gait instability to EMG patterns noted during recovery of walking in an elderly man after cerebral infarction, demonstrating the validity and usefulness of this index of gait instability. This index of instability is thus proposed as a basic tool for analyzing the developmental process of gait as well as for electro myographically assessing clinical progress in acquisition of the ability to walk.

An Index of Gait Instability 85

An Index of Gait Instability Table 4-1. EMG activity in unstable walking Region

Code ST-TA

Ankle

SW-LG ST-LG SW-VM

Knee ST-VM ST-RF Knee & Hip ST-BF Hip

ST-GM

Interpretation Activity of the TA in stance phase Activity of the LG in the latter part of swing phase Activity of the LG in the first half of stance phase Activity of the VM in the latter part of swing phase Activity of the VM in stance phase Activity of the RF in stance phase Activity of the BF in stance phase Activity of the GM in stance phase

Activity decreases or disappears at 1 month after learning to walk 3 months after learning to walk 3 years of age 1 month after learning to walk 1 month after learning to walk 1 month after learning to walk 3 years of age 3 years of age

Indication Very unstable Unstable Slightly unstable Very unstable Very unstable Very unstable Slightly unstable Slightly unstable

ST : stance phase, SW: swing phase, TA: tibialis anterior, LG: lateral gastrocnemius, VM : vastus medialis, RF : biseps femoris, BF : biseps femoris, GM : gluteus maximus.

Conclusion To obtain an index of gait instability from EMG information, we made longitudinal observations on three children from the time they first began to walk independently at about 1 year of age until a stable adult-like walking pattern was achieved at around 3 years of age. For purposes of comparison, these observations were supplemented with cross-sectional observations taken from fifteen babies in the first year of age, five babies in the second year, and five more in the third year of age. From all of these observations we were able to construct an index of gait instability (Table 4-1) . As seen in a child within the first month of learning to walk, the vastus medialis is active in the latter half of swing phase, the tibialis anterior and rectus femoris are active during stance phase, and activity of the vastus medialis is continuous. These EMG characteristics are not usually seen in subsequent childhood gait or in adult gait, and they serve as markers of very unstable gait.

86 Application to Gait Analysis and Evaluation

Activity of the gastrocnemius in the latter half of swing phase is generally noted only within the first 3 months after the child learns to walk, and that activity is interpreted as a sign of unstable gait. Activity of the gastrocnemius in the first half of stance phase and continuous activities of the biceps femoris and gluteus maximus from initial contact with the floor until push off are found in children until 3 years of age. These activities are considered EMG markers of slightly unstable gait. We propose this index of instability as a way to analyze gait in terms of developmental processes and also as a way to electromyographically assess clinical progress in gait training.

Fig. 4-11 . Very unstable infant walking at 1 year of age.

An Index of Gait Instability 87

When does a baby feel instability from?

The purpose of this study was to see whether an electromyographic (EMG) index of gait instability is applicable to the developmental process of supported walking in normal neonates and infants. In six neonates ranging in age from 14 to 26 days after birth, EMGs of stepping were recorded at approximately from 1 to 4 week intervals until around 4 months. Additionally, longitudinal EMGs of one subject were recorded at 1 or 2 week intervals until just before independent walking. EMG patterns of the lateral gastrocnemius (an ankle plantar flexor) and vastus medialis (a knee extensor) in the latter part of swing phase indicating unstable walking, not seen in the neonatal period up to the first postnatal month, tended to increase in young infants at around 3 postnatal months. These results suggest the addition of voluntary infant stepping to reflex neonate stepping from around 3 months. From 6 to 12 months, these marked activities tended to decrease, gradually coming to resemble adult stable walking through development of strength, balance, and postural control. In conclusion, muscular activities of the lateral gastrocnemius and vastus medialis in the latter part of swing phase indicate unstable walking, findings which are applicable to developmental changes during newborn stepping and infant supported walking.

Human gait without support is associated with muscle strength, balance, and postural control. Although some people have to resort to supported walking by virtue of immaturity, aging, postoperative status, or disease, supported walking generally resembles the normal adult walking pattern except for the substitution of missing postural and balance elements with support. However, to decide whether gait is stable or not is very difficult from motion analysis alone, for example, by comparison with a normal pattern. Although supported walking may show no perceptible change during rehabilitation, the underlying EMG pattern may be changing during the recovery of gait. To evaluate gait stability during supported walking, it is therefore very important to examine muscle function. We have determined signs of unstable walking from EMG activity patterns based on the developmental process of normal infant walking and normal adult walking (Chapter 4), and have used them to analyze gait in terms of development processes and also as a way to electro myographically assess clinical process in gait training. They have reported that the vastus medialis is active in the latter half of swing phase, the tibialis anterior and rectus femoris are active during stance phase, and that activity of the vastus medialis is continuous, as seen in a child within the first month of learning to walk. These EMG characteristics are not usually seen in subsequent childhood gait or in adult gait, and they serve as markers of very unstable gait. Activity of the gastrocnemius in the latter half of swing phase is generally noted only within the first 3 months after the child learns to walk, and that activity is interpreted as a sign of unstable gait. Activity of the gastrocnemius in the first half of stance phase and continuous activities of the biceps femoris and gluteus maximus from initial contact with the floor until push off are found in children until 3 years of age. These activities are considered EMG markers of slightly unstable gait. We were interested in applying the index of gait instability to the developmental process of supported walking in normal neonates and infants. Newborn stepping has been an object of study for a long time. McGraw (1940) and Zelazo et al. (1972) have discussed the significance of early stepping movements for development of adult gait. Only a few attempts so far, however, have been made to study developmental changes of stepping by EMG, particularly in relation to gait instability. In this study, we applied the idea of the index of gait instability to developmental changes during newborn stepping and infant supported walking.

90 Application to Gait Analysis and Evaluation

Four male and two female infants were observed at 14, 18, 19, 22, 23, and 26 days after birth. We initially observed developmental process of supported walking in neonates and infants using an index of gait instability. In interpreting the EMG pattern of supported walking in all subjects tested, we focused on the discharge pattern in the latter part of swing phase and in stance phase. From this, we could see developmental changes in the EMG patterns of swing phase and relatively wide variations in those of stance phase. Most of the figures in this chapter are from longitudinal representative EMG patterns in subject A (Fig. 5-1; 22 days after birth).

Fig. 5-1. Stepping at 22 days after birth.

Application of an Index of Gait Instability 91

Until the 1st month of age

ST

(Al

rtrtftft ft sw

I~

TA

ST

.....

~~

"0'

'.

LG ' VM

I'!

RF SF .l,

J

,ul"I,li

,~~

''''''1'

GM (U

TA"

-

I

.L,wI,,~.

"",rr"1tr~

LG

'.;.~fI/I~\1i!I, ~"'14,....·

RF SF

""""-

..

#-' SWING

\~~o\iiI~i;.<""'I" ~/~---

STANCE 1 sec

I

0.5 m~

22 days NEONATAL STEPPING

Fig, 5-2, EMGs of newborn stepping at 22 days after birth (same subject as in Fig, 5-1), ST: stance phase, SW: swing phase, (R): right leg, (L): left leg, TA: tibialis anterior, LG: lateral gastrocnemius, VM: vastus medialis, RF: rectus femoris, SF: biceps femoris, GM: gluteus maximus,

Figure 5-2 shows representative EMG patterns of stepping at 22 days after birth. This is one of the more rhythmical walking-like patterns that we obtained in this neonate. The discharge patterns of leg muscles during newborn stepping are generally more irregular than in adult walking. The walking in this period was characterized by quick hip and knee flexion in which the thigh became almost horizontal in the first part of swing phase. The foot was raised forward and dorsiflexed strongly. Then the foot began to reach the floor slowly, the knee extending passively along with the hip. The foot usually contacted the floor with the heel first, but in a few instances the forefoot made initial contact (Fig. 5-3) . A squatting posture was often observed during stance phase. At

92 Application to Gait Analysis and Evaluation

around 1 month, such squatting began to become less pronounced. As the ankle and knee were extending, activities in the LG (an ankle plantarflexor) and VM (a knee extensor) were hardly seen in the second half of swing phase. In stance phase continuous activities were seen in the ankle, knee, and hip muscles, especially in single stance. The discharge patterns of many leg muscles (TA, LG, VM, RF, BF, and GM) indicated unstable walking showing reciprocal and co-contraction patterns, with wider variations and more excessive activity of discharge patterns than in adult walking. The TA and LG showed reciprocal patterns on the right and co-contraction on the left. Activity of the TA decreased or disappeared while that of the LG increased in many instances. Conversely, in some of the subjects activity of the LG decreased or disappeared while that of the TA increased. In the knee and hip muscles, the RF and BF showed both reciprocal and cocontraction patterns. In some subjects, the reciprocal pattern was one in which activity of the RF decreased or disappeared while that of the BF increased, whereas in other subjects the converse reciprocal pattern prevailed. Among the knee muscles, continuous activity of the VM was often observed, but sometimes no activity was seen at all.

~

~~ 20 days after birth

Fig. 5-3. Foot prints of neonatal stepping at 20 days after birth.

Application of an Index of Gait Instability 93

From 1 to 4 months of age

ST

(R)

~ffl:fff ST

SW

TA 'fi\>:iI

f'it>l

~r\{"

LG '!ri,"

VM~'Mt~..

~,~,""

1 1 -

,.'#' .,!"",~

"""

,

01<

..

RF BF~

,- , -

GM

(L)i~lUl

'1"1

TA

LG 't-'/"

J,

I

RF '~

BF--

r. SWING

t

--"'@t·~ -:.: f1 I"

........

STANCE

1 sec

I

O.5mv

44 days YOUNG INFANT STEPPING

Fig. 5-4. EMGs of newborn stepping at 44 days after birth (same subject as in Figs, 5-1 and

5-2).

Figures 5-4 and 5-5 show representative EMG patterns of stepping at 44 and 105 days after birth in the same infant. We found in six infants tested that we could induce a stepping pattern after 1 month of age, but not so easily as before 1 month. After 1 month, leg flexion was performed strongly in the first part of swing phase as in the neonatal period, but the degree of hip flexion tended to decrease slightly. We found mostly plantarflexion before floor contact rather than dorsiflexion which was more prevalent in the newborn period. From 3 to 4 months, the toe initially contacted the floor for the most part. Knee extension was performed more actively than in the neonatal period. After around 1 month, a half-squatting posture during stance phase tended to increase.

94 Application to Gait Analysis and Evaluation

5W

5T

(Rl TA---4--~---4~--~~--~--~-r--~

LG __~~~~~~~~~~--~~~-­ VM~~-r~~'~~~~~~~__+-~~_ RF--~--~~-~~--_4~_.--~_r----

,.."," ""--+--t-...~--+-+,~-

GM

';~ : )" 'f ::~'~I=~:n~ll~~:n SWING

STANCE

1 sec

I O.S mv

105 days YOUNG INFANT STEPPING

Fig, 5-5. EMGs of newborn stepping at 105 days after birth (same subject as in Figs. 5-1,

5-2, and 5-4).

From 1 to 3 months, the LG and VM began to show activity before the foot actually touched the floor (Fig. 5-4). From 3 to 4 months, such activity become more pronounced (Fig. 5-5). On the other hand, we found wide variations in the discharge patterns of many leg muscles (TA, LG, VM, RF, BF, and GM) during stance phase (Figs. 5-4 and 5-5) similar to the neonatal period (Fig. 5-2). The stepping did not cease at around 4 or 5 months after birth, as shown in the longitudinal observations of subject A.

Application of an Index of Gait Instability 95

From 6 to 12 months of age

(R)

ST

SW

LG

~lilii -.

_....,

",'

"I'

TA

J.L •. I.,

VM

'1'~Il'f1

RF SF

.I~"

..lJ1l, ''1~1

-'I

'''1''

GM ( L)

TA

II

1J,

k'l

T

If" r~'!""r

LG

~, L!a...

1,11..11.,[1

" II"

.J,

RF

.L" "'1

SF SWING

"'.

STANCE

1 sec

I

O.5 mv

351 days INFANT SUPPORTED WALKING

Fig. 5-6, EMGs of supported walking at 351 days after birth, 34 days before independent walking (same subject as in Figs, 5-1 , 5-2, 5-4, and 5-5),

During infant supported walking (6-12 months of age), step frequency was more regular than during the newborn period. The relatively pronounced flexion in the hip and knee seen in the previous period was slightly reduced (Fig. 5-7). Figure 5-6 shows representative EMG patterns of supported walking at 351 days after birth, but before independent walking was achieved. From 11 to 12 months after birth, the infant became able to stand by herself and to walk with one-handed supported walking. The femur was not actively lifted up in the first part of swing phase, and active ankle plantarflexion and knee extension tended to disappear in the latter part of swing phase. The heel usually contacted the floor first. This subject began to walk without support at 385 days after birth.

96 Application to Gait Analysis and Evaluation

Strong myoelectric activity of the LG and VM seen in earlier periods (Figs. 5-4 and 5-5) tended to disappear in the latter part of swing phase in this period (Fig. 5-6). The marked variations in the discharge patterns of both these muscles seen previously were no longer seen during infant supported walking, and more closely resembled adult walking. Co-contraction patterns of ankle, knee, and hip muscles seen in newborn stepping also tended to decrease or disappear during stance phase, but reciprocal patterns of mutual antagonists generally remained. Nevertheless we did find some instances of excessive muscular activity in many leg muscles during stance phase, similar to what was seen in earlier periods (Figs. 5-2, 5-4, and 5-5).

Fig. 5-7. Stable infant supported walking before independent walking.

Application of an Index of Gait Instability 97

Developmental changes in EMG patterns

ffftJttftl r···~·' I

TO

I

STANCE (ST)

SWING (SW)

1 sec

14 days

0.5 mv

NEONATAL STEPPING

~fllff~i ~:="I =t,~:It:: TO

SW

FC

ST

83 days

1 sec

I

0.5 mv

YOUNG INFANT STEPPING

Fig. 5-8. EMGs of newborn stepping at 14 and 83 days after birth (subject 8).

Figure 5-8 shows the EMG patterns of the ankle and knee extensors before foot contact during stepping at 14 and 83 days after birth (subject B), illustrating the differences between neonate and young infant periods. In the neonatal period (up to 1 month), we could not see strong discharges of the LG and VM in the latter part of swing phase (Fig. 5-8, top). However, in the young infant period (1-4 months), we could see strong discharges of both these muscles in the same phase in many instances, as mentioned above (Fig. 5-8, bottom) . These developmental changes in muscle activity from the neonatal period to the young infant period were similar to those of subject A (Figs. 5-2, 5-4, and 5-5) .

98 Application to Gait Analysis and Evaluation

Application of an index of gait instability to supported walking in babies Table 5-1. EMG index of gait instability Joint

Sign of instability

Interpretation

Ankle

SW-LG

Activity in the LG during SW (+)

Unstable

Knee

SW-VM

Activity in the VM during SW (+)

Unstable

SW: latter part of swing phase, LG : lateral gastrocnemius, VM: vastus medialis, (+): noticeable activity.

Table 5-1 shows the EMG index of gait instability in the latter part of swing phase (Chapter 4, Fig. 5-9). Table 5-2 shows the results of evaluating developmental changes in LG and VM muscular activities during newborn stepping of each subject using the criteria for gait instability (Table 5-1). Table 5-3 shows developmental changes of muscular activities in the LG and VM before foot contact, from newborn stepping to supported walking just prior to independent walking in subject A. From the results obtained in the swing phase, it was discovered that we were able to apply the idea of the index of gait instability to developmental changes during newborn stepping and also infant supported walking.

Ankle plantar flexion

Knee extension

Activity of the LG in the latter part of SW

Activity of the VM in the latter part of SW

Fig. 5-9. EMG index of gait instability in the latter part of swing phase. SW: swing phase, LG: lateral gastrocnemius, VM : vastus medialis, +: noticeable activity.

Application of an Index of Gait Instability 99

Table 5-2. Activities of the lateral gastrocnemius (LG) and vastus medialis (VM) in the latter part of swing phase of stepping induced during very early development

Months after birth Subject

Muscle

B

C

0

E

F

0-1

1-3

3-4

VM

H H

(+), H (+),H

(+), partly H (+), partly H

LG

LG

(-)

VM

H

H, partly (+) (+), H

(+), (-) (+), partly H

LG

(-) (-)

H, partly (+) (-), partly (+)

(+), (-) (+), (-)

H (-)

(+), H (+), (-)

(+), partly H (+), partly (-)

H (-)

(+), H (+), (-)

(+), partly H (+), partly (-)

VM

LG VM

LG VM

LG : lateral gastrocnemius, VM : vast us medialis, (+) : noticeable activity, (-) : no activity, (+), (-): instances of noticeable activity and of no activity intermingled.

Table 5-3. Activities of the lateral gastrocnemius (LG) and vastus medialis (VM) in the latter part of swing phase of stepping during the first year of development in subject A

Months after birth Muscle 0-1

1-3

3-4

LG

H

VM

(-)

(+), H (+), (-)

(+), partly H (+), partly (-)

6-12

H, partly H , partly

(+) (+)

(+): noticeable activity, (-): no activity, (+), (-): instances of noticeable activity and of no

activity intermingled.

100 Application to Gait Analysis and Evaluation

Discussion McGraw (1940) reported that infant stepping can be elicited shortly after birth and during the first months, and that thereafter it usually disappears. Thelen et al. (1987) and Forssberg (1985) pointed out from movement patterns and EMGs that the locomotor pattern of the newborn differs markedly from that of an adult. Usually the leg muscle activities of newborn stepping are irregular and include a high degree of co-activation compared with the adult walking pattern. We will first interpret the meaning of wide variations in stance phase, especially in the EMG patterns of mutual antagonists of the leg muscles during stepping (Figs. 5-2, 5-4, and 5-5). Diminution of activity in the TA and RF accompanied by greater activity in the LG and BF may have been due to forward leaning of the body. When, conversely, activity in the LG and BF decreased or disappeared while that in the TA and RF increased, this may have resulted from leaning backward. Co-contraction of these muscles was also observed in many instances, probably related to maintaining a standing posture with the body erect or to stabilizing of the ankle, knee, and hip joints. These variations of leg muscle activities may be caused by the changing posture during supported newborn stepping. During stepping, sustained discharges of the VM are probably attributable to bearing body weight with the knee flexed. When VM activity was low or absent during stance phase, the manner of bearing weight on the knee joint may have resulted in a smaller load. We thus suggest that the degree of activity in the VM may be regarded as an indication of magnitude of load on the knee joint. Even if the EMG patterns of leg muscles in stance phase appear to indicate unstable walking during stepping, we hesitate to consider them as reliably consistent signs of instability because variations in discharge patterns of the leg muscles are probably closely related to magnitude of joint load, which can be influenced by many factors other than instability. We thus find it undesirable to apply the EMG index of gait instability to discharge patterns in stance phase, during which considerable variations are seen in newborn stepping. Secondly, we will focus on developmental changes in the EMG patterns of supported walking during swing phase. Up to the first month of age, muscular activities of the LG and VM were hardly seen in the latter part of swing phase (Figs. 5-2, 5-8 top, Tables 5-2, 5-3), as

Application of an Index of Gait Instability 101

also reported by Thelen et al. (1987) and Okamoto et al. (2001, 2003) . The leg extends passively and the foot contacts the floor usually with the heel first, as mentioned in chapter 1. These findings clearly show that muscular activities for knee extension and ankle plantarflexion are not observed in this period. As mentioned above (Table 5-1), we determined that activity of the leg extensors before floor contact indicates gait instability, but it is risky to judge the presence of gait stability from an absence of activity in these two muscles in the neonatal period, because stepping in the neonatal period is a reflex movement performed under the control of lower (spinal) levels of the central nervous system (eNS), and equilibrium reflexes are yet immature.

Neonatal newborn stepping ( until 1 month after birth)

~j~~ll Slow leg extension

Young infant stepping ( 1-3 months after birth)

fflll Fast leg extension

Fig. 5-10. Developmental changes in newborn stepping. Strong muscle activities of leg extensors (LG and VM) due to a parachute reaction of the legs before floor contact, not seen in the neonatal period, began to appear in the young infant period from 1 month of age to 3 months.

102 Application to Gait Analysis and Evaluation

From 1 to 3 months, the leg extensors begin to show some activity before the foot reaches the floor (Figs. 5-2, 5-8 bottom, Tables 5-2, 5-3) and the forefoot begins to contact the floor first more often. When the infant begins to actively perform knee extension and ankle plantarflexion, the strong muscle activities of leg extensors begin to participate before floor contact in this period (Fig. 5-10). The marked discharge patterns of these two muscles are similar to those of unstable supported walking and very unstable independent walking for the first time. From these findings, we believe that the activities of the leg extensors before floor contact in this period indicate lack of stability, as identified by Okamoto et al. (2001, 2003). We thus believe that infants in this period begin to feel instability (Fig. 5-11). From around 3 to 4 months old, marked activity of the leg extensors are observed (fables 5-2 and 5-3). Active ankle plantarflexion and knee extension before foot contact become the mode of performance in this period. Milani-Comparetti (1967) reported that the parachute reaction of the legs begins to appear at about 4 months after birth. The parachute reaction is an equilibrium reflex performed under control of higher (cortical) levels of the CNS. Strong muscle activities of the leg extensors before the floor contact may arise as parachute reactions (self-protection) with the maturation of the CNS. Although McGraw (1940) pointed out that it is difficult sometimes to tell whether the active leg extension before floor contact is deliberate or of reflex quality, we would suggest that it is the beginning of voluntary infant stepping (supported walking) added upon reflex neonate stepping at this period. From 6 to 12 months, absence of activity in these muscles before floor contact becomes the rule, and marked activity tends to disappear in the remaining instances (Table 5-3). Lack of activity in these muscles indicates that leg extension before foot contact is not performed actively in this period, in contrast with the active leg extension of around 3 or 4 months. The developmental change of infant stepping in this period is due mainly to maturation of the equilibrium system and development of strength. We interpret the absence of activity in the leg extensors in the latter part of swing phase in this period to resemble adult stable walking and thus to have resulted from development of strength, balance, and postural control.

Application of an Index of Gait Instability 103

Conclusion We made longitudinal observations on six normal neonates to see whether an EMG index of gait instability derived from the developmental process of normal infant walking is applicable to EMG patterns of supported walking in neonates and infants. Muscular activities of the LG (an ankle plantarflexor) and VM (a knee extensor) in the latter part of swing phase indicate unstable walking, findings which are applicable to developmental changes during newborn stepping and infant supported walking. In stepping during the first month, muscular activities were not seen in the LG or VM. It would be misleading to consider such gait to be "stable" simply because these muscles are inactive in the neonatal period, especially since stepping in the neonatal period is under the control of lower levels of the central nervous system. At around the third postnatal month, the LG and VM showed strong activity just before the foot reached the floor, suggesting that muscular activities participating in active ankle plantarflexion and knee extension act as a parachute reaction (Figs. 5-10 and 5-11). This may be the beginning of superimposition of voluntary infant stepping on top of reflex neonate stepping. From 6 to 12 months, when the infant was becoming able to maintain standing without support, marked activities of the LG and VM before floor contact tended to disappear as in the adult pattern. We thus presume that absence of activity in the LG and VM at that time suggests stable walking, reflecting development of strength, balance and postural control during that period. In summary, our observations of developmental changes in newborn stepping and infant supported walking, combined with an EMG index of gait instability based on subsequent stages of development, lead us to believe that activities of the LG and VM in the latter part of swing phase can be interpreted in terms of a scheme of early development of stability in walking.

104 Application to Gait Analysis and Evaluation

Young infant period

Neonatal period 1.------+ ••

4----------------------------------+.

VM: vastus medialis (Knee extensor)

Frequency of occurrence

LG: lateral gastrocnemius (Ankle plantar flexor)

(EMG)

/-----'/ '/ '/ '/

"'1....

'/ '/ '/

~--.~ Birth

1 month

..... •••••

TA: tibialis anterior

··.............~~~.~I~.~~:~!~;.~~:~. 2 months 3 months 4 months after birth

Fig. 5-11. Developmental changes of EMGs in leg muscles before floor contact of newborn stepping. EMG patterns of the LG (an ankle plantarflexor) and VM (a knee extensor) in the latter part of swing phase indicating unstable walking, not seen in the neonatal period up to the first postnatal month, tended to increase in young infants at around the third postnatal month. These results suggest the addition of voluntary infant stepping to reflex neonate stepping from around 3 months after birth.

Application of an Index of Gait Instability 105

Young infant stepping

We believe that infants at around 3 months after birth begin to feel instability.

To study the recovery of walking in an 85 year old man who had right hemiplegia after suffering a cerebral infarction, electromyograms (EMGs) were recorded from his leg muscles. We used signs for instability that we derived from EMG patterns seen in the developmental process of normal infants. The myoelectric activity at 1 month after the stroke showed many patterns indicative of unstable walking, closely resembling activity patterns seen in very unstable independent gait of a 1 year old baby in the first month of learning to walk. 7 months later, these patterns indicating unstable gait tended to have decreased or disappeared, although some marked activity betraying instability still remained. However, when the patient walked with the support of a hand cart and was able to hold his trunk upright, these excessive muscular activities decreased or disappeared, closely resembling the stable adult walking pattern. We recommend further study of the evaluation of recovery of walking after stroke by comparing the patient's EMG patterns to those not only of normal adult human gait, but also of the development of human walking in early childhood.

To recover gait in hemiplegic patients, it is desirable to start rehabilitation as soon as possible after the stroke. Gait analysis after stroke has been studied by film, temporal patterns, footprints, gas metabolism, electromyography (EM G) , and other methods of mechanical, anatomical, and physiological analysis. However, few studies have thus far been done on EMG evaluation of recovery of walking after stroke. We have studied EMG activity in normal infants from the time they first walk independently until they display adult-like walking patterns. We have become able to determine signs of unstable walking from EMG activity patterns based on the developmental process of normal infants (Chapter 4, Table 6-1). We were interested in examining gait in elderly persons, especially those undergoing a process of rehabilitation in walking, and we wanted to consider the appropriateness of using our signs for instability derived from the EMG patterns seen in the developmental process. The purpose of this study was to consider electro myographically recovery of walking in an elderly man undergoing rehabilitation after suffering a cerebral infarction.

Fig. 6-1. Form of walking in an elderly man and an infant. Gait patterns in the recovery of walking after the stroke closely resembled gait patterns seen in the very unstable independent gait at 1 year of age.

108 Application to Gait Analysis and Evaluation

The patient was an 85 year old man who had right hemiplegia after suffering a cerebral infarction. Slight spasticity was seen at 1 month after onset of the stroke, but coordination in walking was almost normal. He usually displayed heel contact and a heel-to-toe pattern in gait. His walking posture was characterized by a markedly forward lean of the trunk and a slight squat (Fig. 6-1). After the stroke he had slight dementia, so it was hazardous to let him live by himself. He needed assistance or supervision in his daily life. In the first month of recovery, the patient was given training to enhance activity of the tibialis anterior. He was instructed to do heel walking with support. Training also included trunk function and balance. At 2-3 weeks after the stroke, he could perform supported walking using a parallel bar, and by 1 month he had recovered independent walking. We recorded EMGs of the independent walking at 1 month after the stroke. To more closely examine the EMG characteristics of the walking stability in the patient, he was instructed to walk with a hand cart for support, keeping his trunk upright. Table 6-1. EMG activity in unstable walking Region

Code ST-TA

Ankle

SW-LG ST-LG SW-VM

Knee ST-VM ST-RF Knee & Hip ST-SF Hip

ST-GM

Interpretation Activity of the T A in stance phase Activity of the LG in the latter part of swing phase Activity of the LG in the first half of stance phase Activity of the VM in the latter part of swing phase Activity of the VM in stance phase Activity of the RF in stance phase Activity of the SF in stance phase Activity of the GM in stance phase

Activity decreases or disappears at 1 month after learning to walk 3 months after learning to walk 3 years of age 1 month after learning to walk 1 month after learning to walk 1 month after learning to walk 3 years of age 3 years of age

Indication Very unstable Unstable Slightly unstable Very unstable Very unstable Very unstable Slightly unstable Slightly unstable

ST : stance phase, SW : swing phase, TA : tibialis anterior, LG : lateral gastrocnemius, VM: vastus medialis, RF: rectus femoris , SF: biseps femoris, GM: gluteus maximus.

Application of an Index of Gait Instability 109

1 month after the stroke

SW

fffr{ ST

(Rl

TA

r-J11M~'I'flIIl\l1

GM '---I-'tI--Mi'M.m. (Ll

TA~~~~~--~~~~~~~-r--~WHW~~~--~~

GM ST

SW

1 sec

0.5 mv

85.0 years

Fig. 6-2. EMGs of walking at 1 month after stroke. (R): Right leg; affected limb, (L): Left leg; normal limb, ST: stance phase, SW: swing phase, TA : tibialis anterior, LG : lateral gastrocnemius, VM: vastus medialis, RF : rectus femoris, SF : biceps femoris , GM : gluteus maximus.

110 Application to Gait Analysis and Evaluation

The subject had difficulty walking, writing, and eating by himself, but recovered independent walking by 1 month after the stroke. Fig. 6-2 shows representative EMGs of walking at 1 month after the stroke. The walking velocity was about 50 - 55 m/min. His walking form was characterized by a marked forward lean of the trunk, a slight squat, and a short step length. He generally contacted the floor with the heel first, barely clearing the floor with the toes, but his foot occasionally hit the ground with the toes first or with the entire sole (Fig. 6-3). In swing phase, the LG of either leg showed activity in the latter half of that phase, but in some instances this activity was absent on either side. At the knee, the VM on the right affected side also showed strong activity in the latter half of swing phase. On the sound left side, however, the VM did not exhibit such activity, but rather it showed discharges just prior to heel contact as in normal adult walking. In stance phase, at either ankle, the TA was continuously active from heel contact until push off. The LG was frequently active, on either side, in the first half of stance phase, but instances of no discharge were also seen on either side. At the knee and hip, the VM, RF, BF, and GM bilaterally showed strong co-contractile patterns in stance phase.

Fig. 6-3. Form of walking at 1 month after the stroke (left) and at 1 month after learning to walk (right).

Application of an Index of Gait Instability 111

7 months after the stroke

fff/f ( R)

TA LG VM

RF SF GM ( L)

TA LG VM

RF SF GM ST

sw 1 sec

85.6 years

Fig. 6-4. EMGs of walking at 7 months after stroke. (R): Right leg; affected limb, (L): Left leg; normal limb.

112 Application to Gait Analysis and Evaluation

0.5 mv

Fig. 6-4 shows representative EMGs of walking at 7 months after the stroke. The gait pattern was similar to that at 1 month after the stroke, but heel contact became very consistent, with rare instances of initial contact by the entire sole or toes. In swing phase, the LG of the affected right leg continued to exhibit excessive activity in the latter half of the phase, but the corresponding myoelectric discharge pattern of the left LG had begun to disappear. The strong burst of activity in the VM of right leg observed at 1 month after the stroke was no longer seen in the latter half of swing phase, but instead a burst was observed just prior to heel contact as in normal adult walking. In stance phase, the strong continuous discharge patterns of the TA seen in both legs 1 month after the stroke tended to decrease or even disappear. These discharge patterns were similar to the normal adult pattern, but marked activity could still occasionally be seen in the TA during stance phase on either side. The LG on both sides showed strong activity at the beginning of stance phase. The VM, BF, and GM on either side still exhibited marked co-contractions throughout stance phase, but the excessive continuous discharge patterns of the RF on either side seen at 1 month after the stroke tended to decrease.

Fig. 6-5. Diagram illustrating the differences between an adult and an elderly man during walking. LG: lateral gastrocnemius, VM: vastus medialis, BF: biceps femoris, GM: gluteus maxim us, (+): noticeable activity.

Application of an Index of Gait Instability 113

1 year 7 months after the stroke

ST

SW

(R)

(L)

TA~~~~1~~~~~~~n~~~~~~;H~~~~~-

SW

ST 1 sec

86.6 years

Fig. 6-6. EMGs of walking at 1 year 7 months after stroke. (R): Right leg; affected limb, (L): Left leg; normal limb.

114 Application to Gait Analysis and Evaluation

0.5 mv

Fig. 6-6 shows representative EMGs of walking at 1 year 7 months after the stroke. The gait pattern resembled that at 7 months after the stroke. In swing phase, the LG of the affected right leg still showed discharges in the latter half of swing phase as in the recordings of the previous two occasions, but the magnitude of that activity was decreased or sometimes even absent. On the other hand, activity of the LG on the sound left side was not usually seen, thus resembling the pattern seen in normal adult gait. In stance phase, excessive discharge of the LG seen at the beginning of the phase on either side looked like the pattern at 1 month and at 7 months after the stroke. The VM, BF, and GM on either side, as before, showed excessive continuous co-contractile patterns in stance phase.

fl ~

f

86.6 years

TA LG VM RF SF GM

KNEE

EXTENSION

*

FLEXION

SmlNG

1 sec

SQATTING

I

0.5 mv

STANDING

Fig. 6-7. EMGs of the right leg during standing at 1 year 7 months after stroke (same subject as in Fig. 6-6). EMG patterns of squatting posture before standing closely resemble the excessive continuous patterns of activity in the LG, VM , BF, and GM associated with the forward sway of the trunk with the knees flexed in stance phase of elderly walking.

Application of an Index of Gait Instability 115

EMG evaluation of walking stability

Table 6-2. EMG evaluation of walking stability

Right leg: affected limb EMG pattern of unstable walking

1 month

7 months

after stroke

after stroke

ST-TA

(++)

SW-LG

(+), partly (-)

(+), partly (-)

(+), partly (-)

(+)

(+)

H, partly

(+)

1 year 7 months after stroke

H , partly

ST-LG

(++), partly (-)

SW-VM

(+)

H

H

ST-VM

(++)

(++)

(++)

(+)

ST-RF

(+)

(+)

(+)

ST-BF

(++)

(++)

(++)

ST-GM

(++)

(++)

(++)

1 month after stroke

7 months after stroke

1 year 7 months

unstable walking ST-TA

(++)

SW-LG

(+), partly (-)

Left leg : normal limb EMG pattern of

ST-LG

(++), partly

H, partly H

(+)

after stroke

H, partly

(-)

(-)

(++)

(++)

SW-VM

(-)

H

(-)

ST-VM

(++)

(++)

(++)

ST-RF

(++)

(+)

(+)

ST-BF

(++)

(++)

(++)

ST-GM

(++)

(++)

(+)

(+)

(++): marked activity, (+): noticable activity, (-): no activity.

See Table 6-1 for descriptions of the EMG patterns.

Table 6-2 shows in brief the results of evaluating the patient's gait based on our criteria for walking stability. The myoelectric activity at 1 month after the stroke showed many patterns indicative of unstable walking. 7 months later, these patterns indicating unstable gait tended to decrease or disappear, but the overall picture was still one of a more unstable walking pattern than normal adult gait.

116 Application to Gait Analysis and Evaluation

Discussion One purpose of gait evaluation is to describe how a patient's performance differs from "normal" gait. Generally, normal adults show regular reciprocal discharge patterns of agonists and antagonists during the gait cycle. If an individual cannot maintain dynamic stability because of a central nervous disorder or aging, some EMG activities not usually seen in the adult pattern should appear. In our case study, excessive muscular activities appeared early in the recovery period after a stroke (Fig. 6-2). In the latter half of swing phase, the VM and LG showed activities not ordinarily seen in the adult. These discharge patterns were similar to those seen at the onset of very unstable independent walking in early child development. That is, activity of the VM is observed in the latter half of swing phase until about 1 month after first learning to walk, whereas activity from the LG in the latter half of swing phase continues until about 2 or 3 months after learning to walk. As Okamoto et al. (1985, 2001, 2003) have pointed out, these patterns may be considered to come from the knee extending and the ankle plantartlexing to prevent falling. These excessive muscular activities decrease or disappear when the child is given external support, so these EMG patterns in the latter half of swing phase suggest unstable walking. In a patient who has suffered a stroke, excessive plantartlexion on the affected side is one typical problem. In our patient, activity of the LG was observed in the latter half of swing phase bilaterally 1 month after the stroke. At 7 months after the stroke, the excessive activity was still found in the right LG, but no longer in the left LG. Stability had thus returned to the extent that indirect effects of the stroke on the sound side had disappeared and the patient was experiencing disability more from the direct effects on the affected side alone. In stance phase, the TA showed strong activity not usually seen except at the beginning and end of the phase in the adult. This discharge pattern was also similar to that seen in the first month of independent walking by a baby. This activity might be interpreted as an effort to maintain balance with the toes gripping the surface. The LG also exhibited strong activity not seen at the beginning of the phase in normal adult gait, similar to what occurs in small children until around 3 years of age. Since the co-contraction of the TA and LG in stance phase decreases or disappears in stable supported walking

Application of an Index of Gait Instability 117

of a baby and the patient (Fig. 6-8), those patterns of excessive myoelectric activity are indicative of an unstable walking pattern. To efficiently control body balance in various postures, the TA and LG should operate in reciprocal patterns. Co-contraction of these two muscles suggests effort to stabilize the ankle joint. In our patient strong continuous discharges of the TA were observed bilaterally in stance phase 1 month after the stroke. At 7 months after the stroke, such excessive patterns of the TA during stance phase, suggesting very unstable walking, rarely appeared on either side, indicating that walking at this stage was much closer to normal. The elderly subject in this study had a forward posture with slightly flexed knees, a tendency pointed out by Crithley (1956). In stance phase, activity of the VM not seen in normal adult gait was strong and continuous, resembling the pattern seen in a baby during the very first month of independent walking. This muscle activity might be interpreted as helping to maintain balance with slightly flexed knees. Similarly, strong continuous activity of the LG, BF, and GM in our patient not seen in the normal adult were like the activities of these muscles seen in small children until around 3 years of age. Such activities of these antigravity muscles might be attributable to the patient's forward posture (Figs. 6-5 and 6-7). When the patient used support while walking and attained an upright posture, the excessive activities of these muscles tended to decrease or disappear, looking instead like the stable normal adult pattern (Fig. 6-8). These excessive discharge patterns of the VM, LG, BF, and GM in stance phase thus suggest an unstable walking pattern. Just as Finley et al. (1969) suggested that the elderly persons in their study were trying to preserve stability, the increased EMG of our patient not seen in normal adult walking may reflect an attempt to control progression and stability or the age-related changes in gait posture. "Age-related" changes in walking behavior may be caused either by the normal aging process or by disease, but it is difficult to determine which changes in gait occur as part of the normal aging process and which occur as the sequelae of pathological processes. In either case, it is necessary to evaluate all age-related changes in gait and to compare them with normal human gait, including the development of walking.

118 Application to Gait Analysis and Evaluation

Our patient showed excessive muscular activities compared with normal adult walking at 1 month after a stroke. At 7 months after the stroke, the muscle activities indicative of unstable walking tended to decrease or disappear, although signs of excessive muscular activity still remained. When stable supported (hand cart) walking with the trunk upright was performed, the EMG patterns were closely similar to those of a stable adult walking pattern (Fig. 6-8). From the EMG data in this study, we saw that we could elicit a more normal adult walking pattern by substituting for missing postural and balance elements. These results suggest the validity and usefulness of evaluating the recovery of walking after a stroke by comparing EMG data with patterns seen in normal human gait, including the development of walking in early childhood.

~

(

~ HC

TO

TA

LG VM RF

~

BF

GM sw

ST

1 sec

CANE

I 0.5 mv

sw

ST

1 sec

I

HAND CART

sw

ST

1 sec

0.5 mv

I

0.5 mv

HAND CART (ERECT POSTURE)

Fig. 6-8. EMGs of the right leg during supported walking at 1 year 7 months after stroke (same subject as in Fig. 6-6). TO: toe off, He: heel contact, SW: swing phase, ST: stance phase, TA: tibialis anterior, LG: lateral gastrocnemius, VM: vastus medialis, RF: rectus femoris,

SF: biceps femoris, GM:

gluteus maximus.

Application of an Index of Gait Instability 119

Conclusion To study the recovery of functional mechanisms of the leg muscles of an elderly man following a cerebral infarction, we studied the activities of his leg muscles during gait at 1 month, 7 months, and 1 year 7 months after his stroke. The muscle activities at 1 month after the stroke closely resembled activity patterns seen in the very unstable independent gait of a 1 year old baby in the first month of learning to walk. At 7 months after the stroke, these abnormal patterns of activity in the tibialis anterior (TA) and lateral head of the gastrocnemius (LG) tended to decrease or disappear, although activity of the LG in the affected right leg during the latter half of swing phase still remained. Muscle activities associated with flexed knees in the vastus medialis (VM) and with a forward inclined trunk in the LG, biceps femoris (BF), and gluteus maximus (GM) were still marked in stance phase bilaterally. These EMG patterns were similar to slightly unstable walking seen in small children until around 3 years of age. When the patient walked with the support of a hand cart and was able to hold his trunk upright, the strong burst of activity in the LG in the latter half of right swing phase disappeared, and the excessive continuous patterns of activity in the LG, VM, BF, and GM associated with forward sway of the trunk with the knees flexed in stance phase decreased or disappeared, closely resembling the stable adult walking pattern. We recommend further study of the evaluation of recovery of walking after stroke by comparing patients' EMG patterns to those not only of normal adult human gait, but also those of the development of human walking in early childhood.

120 Application to Gait Analysis and Evaluation

References Basmajian, J. v., & Deluca, C. J. (1985). Human locomotion. In Muscles Alive (pp.367-388). Baltimore: Williams & Wilkins. Burnett, C. N., & Johnson, E. W. (1971). Development of gait in childhood. Part II. Develop. Med. Child. Neurol., 13(2),207-215. Crithley, M. (1956). Neurologic changes in the aged. ]. Chron. Dis., 3, 459-477. Finley, F. R; Cody, K A; & Finizie, R V. (1969). Locomotion patterns in elderly women. Arch. Phys. Med., 50, 140-146. Forssberg, H. (1985). Ontogeny of human locomotor control. 1. Infant stepping, supported locomotion and transition to independent locomotion. Exp. Brain. Res., 57, 480-493. Kazai, N.; Okamoto, T.; & Kumamoto, M. (1976) . Electromyographic study of supported walking in infants in the initial period of learning to walk. In P. V. Komi (Ed.), Biomechanics V-A (pp. 311-318) . Baltimore: University Park Press. McGraw, M. B. (1940). Neuromuscular development of the human infant as exemplified in the achievement of erect locomotion. ]. Pediat., 17,747-771. Milani-Comparetti, A, & Gidoni, E. A (1967). Routine developmental examination in normal and retarded children. Develop. Med. Child. Neurol., 9, 631-638. Nashner, L. M., & McCollum, G. (1985). The organization of human postural movements: A formal basis and experimental synthesis. Behavior Brain Sci., 8, 135-172. Okamoto, T.; Okamoto, K; & Andrew, P. D. (2003). Electromyographic developmental changes in one individual from newborn stepping to mature walking. Gait and Posture, 17, 18-27. Okamoto, T.; Okamoto, K; & Andrew, P. D. (2001). Electromyographic study of newborn stepping in neonates and young infants. Electromyogr. Clin. Neurophysiol., 41, 289-296. Okamoto, T., & Okamoto, K (2001). Electromyographic characteristics at the onset of independent walking in infancy. Electromyogr. Clin. Neurophysiol., 41, 33-41.

121

Okamoto. T.; Tsutsumi, H.; Goto, Y.; & Andrew, P. D. (1987). A simple procedure to attenuate artifacts in surface electrode recordings by painlessly lowering skin impedance. Electromyogr. Clin. Neurophysiol., 27, 173-176. Okamoto, T., & Goto, Y. (1985). Human infant pre-independent and independent walking. In S. Kondo (Ed.), Primate Morphophysiology, Locomotor Analyses and Human Bipedalism (pp. 25-45) . Tokyo: University of Tokyo Press. Okamoto, T.; Goto, Y.; & Kumamoto, M. (1983). Electromyographic study of the bifunctional leg muscles during the learning process in infant walking. In H. Matsui & K. Kobayashi (Eds.), Biomechanics VJll-A (pp 419-422). Illinois: Human Kinetics Publishers. Okamoto, T., & Kumamoto, M. (1972). Electromyographic study of the learning process of walking in infants. Electromyography, 12, 149-158. Sutherland, D. H.; Olshen, R.; Cooper, L.; & Woo, S. L. (1980). The development of mature gait. ]. Bone. Joint. Surg., 62-A, 3, 336-353. Thelen, E.; Ulrich, B. D.; & Jensen, ]. L. (1989). The developmental origins of locomotion. In M. H. Woollacott & A Shumway-Cook (Eds.), Development of Posture and Gait Across the Life Span (pp. 25-47). South Carolina: University of South Carolina Press. Thelen, E., & Cooke D. W. (1987). Relationship between newborn stepping and later walking: A new interpretation. Develop. Med. Child. Neurol., 29, 380-393. Thelen, E., & Fisher, D. M. (1982). Newborn stepping: An explanation for a "disappearing" reflex. Dev. Psychol., 18, 760-775. Touwen, B. C. L. (1976). Neurological Development in Infancy. London: Spastics International Heinemann. Touwen, B. C. L. (1971). A study on the development of some motor phenomena in infancy. Develop. Med. Child. Neurol., 13, 435-446. Winter, D. A (1991). The Biomechanics and Motor Control of Human Gait. Ontario: University of Waterloo Press. Zelazo, P. R.; Zelazo, N. A; & Kolb, S. (1972). "Walking" in the newborn. Science, 176,314-315.

122

The books which were quoted

Primale Morphopbysiology, LocomOior Anal)' cs and Human Bipedali m

1

2

4

3

5

1) Basmajian, ]. V. (1974) . Muscles Alive. Baltimore : Williams & Wilkins. 2) Kondo, S. (Ed.) (1985). Primate Morphophysiology, Locomotor Analyses and Human Bipedalism. Tokyo: University of Tokyo Press. 3) Lois, B. (1994) . Motor Skills Acquisition in the First Year. Texas: Therapy Skill Builders. 4) Leonard, C. T. (1998). The Neuroscience of Human Movement. St. Louis: Mosby-Year Book. 5) Woollacott, M. H., & Shumway-Cook, A. (Eds.) (1989) . Development of Posture and Gait Across the Life Span. South Carolina: University of South Carolina Press.

123

Mature adult walking pattern

The development and degradation of gait

TO Fe

TO

1.9 years

He

""' I'"'"'"

r.J

+-H

~r'

TO

He

~

3 years

r---r--'"~r-

~..~ .

.I"

.,0, ""..

" /"

UI" r1l"

.L

I~ ,,,-

'~'r

VM -Hl<\lJo1rn!1/1J.+JI~~-...f.II RF~~__-m~__~~

..~

SF '*'*~'*""ftIJl OM -t--Mi!I>---h-+'!flII' sw

ST

SW

~O.5mv

INFANT WALKING

.....

n ST

'T I

~ O.5 mv

TA

He

29 y ears

TO

ST

~ O.5mv

IMMATURE CHILD WALKING.....

kU~k A~jtA TO

SW

He

60 years

TO

MATURE WALKING

ffff{ He

85 y ellr s

/'""

I

JiI,

LO

nl~

VM RF SF

'II 1.11

OM sw

ST I sec

I 0.5 mv

ADULT WALKING

sw

ST

I sec

I 0.5 mv

SW

ST

~ O.5 mv

ELDERLY WALKING

TO: toe off, FC: foot contact, HC: heel contact, SW: swing phase, ST: stance phase, TA: tibialis anterior, LG: lateral gastrocnemius, VM : vastus medialis, RF: rectus femoris , SF: biceps femoris, GM : gluteus maximus.

126 A ppendix

IMMATURE INFANT WALKING PATTERN

IMMATURE CHILD WALKING PATTERN

MATURE ADULT WALKING PATTERN

[[}srn~rn\ [1J ADULT WALKING

r:i,~~~NI+1 E1RI~;1:r~N UNSTABLE

SUGHTLY UNSTABLE

I)

STABLE

EMG activity in unstable walking Region

Ankle

Code

Interpretation

Indication

ST-TA

Activity of the TA in stance phase

Very unstable

SW-LG

Activity of the LG in the latter part of swing phase

Unstable

ST-LG

Activity of the LG in the first half of stance phase

Slightly unstable

SW-VM

Activity of the VM in the latter part of swing phase

Very unstable

ST-VM

Activity of the VM in stance phase

Very unstable

ST-RF

Activity of the RF in stance phase

Very unstable

ST-BF

Activity of the BF in stance phase

Slightly unstable

ST-GM

Activity of the GM in stance phase

Slightly unstable

Knee

Knee & Hip

Hip

ST: stance phase, SW: swing phase, TA : tibialis anterior, LG : lateral gastrocnemius, VM: vastus medialis, RF: rectus femoris , BF : biseps femoris, GM : gluteus maximus.

Appendix 127

EMG experiment of infant walking EMG signals were recorded via silver electrodes coated with silver chloride. The electrodes were 5 mm in diameter. Muscles chosen for recording, based on our previous studies of gait (Okamoto et aI., 1972, 1985, 2001, 2003), were the long head of the biceps femoris (BF) , gluteus maximus (GM), lateral head ofthe gastrocnemius (LG), rectus femoris (RF), tibialis anterior (TA), and vastus medialis (VM). Two electrodes were placed on each muscle 5 mm apart from edge to edge of the electrode collars in the direction of the muscle fibers, midway between the proximal and distal ends of the superficially palpable part of the muscle belly. A reference electrode was placed over the patella. The skin at each electrode locus was scratched lightly with a needle, reducing the resistance between pairs of electrodes to less than 5000Q (Okamoto et aI., 1987). The signals were amplified one million times by biological amplifiers and recorded on a pen-writing electroencephalograph with the paper speed set at 60 mm/s and an amplitude of 12 mm representing a voltage change of 0.5 mV. The upper limit of the frequency response of the pens was 120 Hz at -3 dB. The child's gait was recorded with a video camera at 60 frames/s, and a voltage pulse synchronizing each frame was recorded with the EM Gs. From the video recordings we determined when and, if feasible, how the foot contacted the floor and made qualitative observations about joint motions. With this instrumentation we were able to divide gait into stance and swing phases and judge presence and absence of muscle activity. We did not generally need to assess relative intensity of activity for a given muscle, and when that was done it was only for such obvious distinctions as "large" versus "small". At the beginning of each session, clean baselines were established for each channel of EMG. Because of the low skin impedance at each electrode locus, no artefact appeared due to movement of electrode cables (Okamoto et aI., 1987). An artefact was producible only by striking the electrode itself with a flick of the finger, a procedure used to confirm proper channel assignments for the muscles recorded. Small voltage changes beyond the baseline were generally interpreted as cross talk from nearby muscles. Activity from the target muscle was crisp and large in contrast, and accorded with reasonable expectations in terms of conceivable motor activity. A stricter criterion would have been difficult to establish, especially when the child was in the earlier stages of development. 128 Appendix

Certain technical limitations in our study are noteworthy. We used a pen-writer to record EMG, so the higher frequency components of the myoelectric signal were lost. Since our observations were qualitative rather than quantitative and since the significant content of EMG was contained below 100 Hz, we believe that our assertions based on data thus recorded are fundamentally valid. Another potential problem in this study was the possibility of cross talk. Had this been an obvious problem it would have been apparent in correlated patterns among the numerous recordings examined. Muscles were deemed to be active only if crisp and clear waveforms were recorded rather than lowamplitude voltages with rounded peaks. Since the observations of the present study are based on a few babies, our findings should be confirmed in longitudinal studies by other investigators, hopefully with other muscles recorded as well and with more precise methods of kinematic recording.

My first EMG experiment in infant walking (1967) The subject is my daughter (co-author), and my wife is providing strong cooperation.

Appendix 129

An infant's first steps alone are an achievement in the struggle for the development of balance, postural control , and strength during the first of year of life.

Acknowledgements In 1967, with the help and cooperation of my wife, I decided to make my two daughters the subjects of study. I learned that no cross-sectional and longitudinal electro myographic studies of infant walking were being carried out anywhere in the world and I decided to use my daughters, two-year old Kayoko and six-month old Emi, as they then were, as the subjects of my study, and I succeeded in creating electromyographic records of the development process in which infants acquired independent walking and developed further skills. My studies attracted almost no attention in Japan. However, Dr. John V. Basmajian, Professor at the School of Medicine, Emory University, Atlanta, U.S.A., the world authority on gait, recognized the value of my studies when I presented a report at the International Congress of Electromyography held in Brussels in 1971, and he invited me to join the gait research project at the Center for Rehabilitation Medicine at Emory University. Since then I have made electromyographical recordings of the walking of more than a thousand people, ranging from the newborn to the elderly, and accumulated much valuable data on gait. This time too, thanks to the help of my wife and daughters, I was able to analyze this data and to gather the results into a book that would introduce them for the first time to the world. This study of gait carried out by one Japanese family may seem a very small step in the eyes of the world, but I firmly believe that our study will contribute the development of gait studies in the next generation. I would like to express my gratitude to all the children and their parents, together with many supporting staff, who took part in and played a vital role in the difficult experiments involving the electromyographical recording of infant walking. I would also like to express my sincere gratitude to my wife and daughters who supported my studies right from the first experiment up to the publication of this book, and who will continue, I am sure, to give me their support.

131

Last but not least, I would like to express my gratitude and affection to my darling grandchildren, who appear in this book and who will be the bearers of the next generation. I would like you to follow your own chosen paths with full confidence into the future, just as I plan to continue in "my way" until the end of my own life.

Tsutomu Okamoto

"Let's leave something wonderful for the children!"

132

About the Authors

Tsutomu Okamoto, Ph.D. Director of Walking Development Group Professor Emeritus of Kansai Medical University

Dr. Tsutomu Okamoto has researched electromyographical gait studies for forty years. He was a member of the gait research project at the Center for Rehabilitation Medicine at Emory University, Atlanta, U. S. A. His papers have been published in many professional journals and his electromyographical data on infant walking is cited widely internationally. He won the All Japan Canoe Championships (1961), served as one of the Tokyo Olympics Special Coaches and the President of the All Japan Intercollegiate Canoe Fderation (1990-2002). He has received many awards, including the Minister of Education, Culture, Sports, Science and Technology Award, Prince Chichibu Memorial Award and the Osaka Governor Award.

133

Kayoko Okamoto, Ph.D. Researcher of Walking Development Group Part-time Lecturer at Kyoto University

Dr. Kayoko Okamoto became the first subject of her father's electromyographical experiments when she was one year old. She has continued gait research and development together with her father, Dr. Tsutomu Okamoto, and won the Incentive Prize from the Japanese Society of Biomechanics. She has researched and developed a unique walking method and exercise program for preventing falls and avoiding becoming bed-ridden based on her own grandfather's experience, who regained independent walking power with rehabilitation training from a bedridden condition caused by cerebral infarction. Her lectures on anti-aging walking are extremely popular among middle-aged and elderly persons. She teaches the theory and practice of "exercise for a heaithy life" at university.

134

EMG

WALKING

Walking Development Group

The Walking Development Group aims through its lectures, educational activities, and publications to spread knowledge of international research into walking in an easily understandable form and hopes through its activities to increase the number of people in the world who are overflowing with youth and health. Our message is "Let's leave something wonderful for the children!" We aim to create a society where the children who will carry the world forward are able actively to pursue their aims in a spirit of vitality.

Director of Walking Development Group Tsutomu Okamoto, Ph.D.

Walking Development Group

~lTIm~liJf~pJf

G-804 Tenno 2-6, Ibaraki-shi, Osaka 567-0876, JAPAN http://www13.ocn.ne.jpFhokou/

135

Development of Gait

Same subject as cover photograph (at 3 years of age)

The children acquire mature adult walking pattern, which indicates strong push off motions of the foot, with the body erect, at about 3 years of age. This foundation will provide them with a basis to continue walking across the life span.

Okamoto & Okamoto