Professional Documents
Culture Documents
by Electromyography
Application to Gait Analysis and Evaluation
Tsutomu Okamoto, Ph.D.
Kayoko Okamoto, Ph.D.
Published by
llJ
WALKING
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
iv
Contents
Preface
Contents
Part I
iii
v
8
12
16
18
20
25
28
30
32
34
36
38
40
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
Part II
67
69
73
79
84
86
89
References
Appendix
Acknowledgements
About the Authors
vi
110
112
114
116
117
121
125
131
133
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.
Fig. 1-1.
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
(Hip extensor)
Mutual antagonist: SF
SF : biceps femoris
(Knee flexor, Hip extensor)
Mutual antagonist: RF
LG : lateral gastrocnemius
TA : tibialis anterior
(Ankle dorsiflexor)
Mutual antagonist: TA
Mutual antagonist : LG
6 Development of Gait
0.5
k~j~AA:j~A
326
mv]
333
339
346
356
365
372
378
388
Plantar flexion
Dorsiflexion
Knee
Extension
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).
ST
sw
ST
~ 'I'\'fJtW~~t;~i~
(R)
TA
",""
LG
,
VM
RF
BF
GM
,--
(l)
TA
LG
VM ----~~~--------~------+_----------~~~~~------+_
BF
-------------------+------+-------------------~------~.
STANCE (ST)
SWING (SW)
--'I
___
1_
s e_c __
0.5 mv
1 week (Y.T.)
NEONATAL STEPPING
8 Development of Gait
ST
(R)
SW
TA
'i~
LG--________
ST
" " , ,
.. , I
......r+~
r-__~--~--------------------+_--+_-----
VM
RF
BF
GM
( L)
TA-""""\
LG
RF
BF
I
SW
ST
1 sec
10 Development of Gait
0.5 mv
Fig. 1-7. Foot contact of stepping in neonatal period (1-2 weeks after birth).
(-): no activity.
11
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
12 Development of Gait
0.5 mv
--
SW
hS-rT-r-r--rr-h h----r-r-r--r-,,-...,-,
(R)~
TA
II"
1ft.
""
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
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)
Fig. 1-10. Foot contact of stepping in neonatal period (3-4 weeks after birth).
(-): no activity, (+): noticeable activity.
14 Development of Gait
15
SW
ST
U ' , ,, ,
( R)
TA
BF
GM~--...
(Ll
TA'IIA(/~""",~-""'t"''''r'TM~~.~~i''J,rIII..~~.....~IJ1fIt4>~Ji'.
ST
SW
1 sec
0.5 mv
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
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.
17
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.
Vastus medialis
(Knee extensor)
Fig. 1-16. Foot contact of stepping in young infant period (3-4 months after birth).
(+): noticeable activity, (-): no activity.
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
Neonatal period
(Early. 1- 2 weeks)
~~~~j
2 weeks (A.I.)
Neonatal period
(Early. 1-2 weeks)
jjlttf1
3 weeks (H.Y.)
Neonatal period
(Late. 3-4 weeks)
fttfttttt
1.5
Infant period
(Onset. 1-2 months)
months (H.Y.)
frftfflf
3
Infant period
(Initial. 3-4 months)
iiw~~
3.5
Infant period
(Initial. 3-4 months)
months (A.I.)
months (H.Y.)
rl~~~(f
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
1-2 months
after birth
Lateral border of forefoot
with fast leg extension
3-4 months
after birth
Forefoot
with fast leg extension
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.
26 Development of Gait
fffffr
(R)
TA
LG
VM
RF
BF
GM
(U
TA
LG
VM
RF
BF
GM
sw
ST
( R)
KNEE~
(L)
KNEE
EXT. ~
FLEX.
-----------
V
t sec
0.5 mv
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
Fig. 2-3. Foot prints on the 1st day of independently walking (at 1 year 1 month).
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
Fig. 2-4. EMGs at 2 weeks after learning to walk (at 10.5 months).
30 Development of Gait
0.5 mv
11 months
( R)
TA~~~~~~~~~~~~mM~~~~~~-+~WM~
GM+-~~~~~~~~~~--~--~~~-r~~~~+-~~~
(L)
TA~~~~NH~~~~~~~~~~~~~~~~~
( R)
KNEE
EXT. ..
( Ll FLEX.
KNEE ---~
/---1 sec
32 Development of Gait
0.5 mv
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).
( R)
!il"~
~11
1'1'
11\'
ld'llLtk
LG ,~.
VM
RF
GM
.I.
l' \
LG 1.1
I.
VM
111
3'I.lj~h'.
hUll,
1'1'
'If"
.L IdlL~.!
"I fU""llr
~II"
!'I
I.
.1,lhlLJ
"11.1.
"1'''
I'
JL
.~.
"
.~ 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"
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
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
1!111 I~.
TA
SF
12 months
SW
ST
~""""'\J
'IIITI'H
,r.I~I~1
ItJ,
.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
34 Development of Gait
lJ,,,
,I.
~~
'1'
j~lI..lI..1c
'It
'''I
TO
sw
Fe
1 year
ST
TO
SW
1 sec
FC
1 year 3 months
ST
0.5 mv
1 sec
0.5 mv
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,
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.
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
36 Development of Gait
IMMATURE
INFANT WALKING
PATTERN
..
IMMATURE
CHILD WALKING
PATTERN
up to 3 months
after learning to walk
3 months - 2 years
after learning to walk
..
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.
GM~~~n~~~--~~~~~~~~MM~~~~~~~~~
( L)
TA
::~~~
GM..
... 11,
~~.
",., ..
"'-1111-'....._ _ __
( R)
KNEE-------."
( L)
EXT .
FLEX.
K N E E - - - - -......
( R)
FC
( L)
FC
FF
HC
TC
1 sec
0 .5 mv
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.
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
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
Ankle
EMG pattern
1st day
2 wks
1 mon
2-3 mons
(+)
(+)
(++)
(++)
(++)
(+)
(-)
(-)
(+)
(+)
(+)
()
Continuous (VM+)
(++)
(++)
()
()
(+)
(+)
(++)
(++)
()
()
(-)
(-)
(+)
(+)
()
()
Continuous (GM + )
(++)
(++)
(++)
(++)
Knee
Hip
sw
ST
sw
ST
sw
ST
1 sec
0.5 mv
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.
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
46 Development of Gait
t1tft~f frffff
Itlrftfi
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 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).
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).
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
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.
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~
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
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
621
385 days
0.5 mv
TA
LG
VM
RF
8F
GM
KNEE
Ext.
Flex.
FC~
600
700
Fig. 3-6. EMGs on the 1st day of independent walking (at 1 year of age).
SUPPORT: supported walking, INDEPENDENT: independent walking.
Infant walking
(from 1 week to 2 months after learning to walk)
j~R'~~ i~A~~'
TO
Fe
jl9Ajil
TO
Fe
"I.
TA
LG
1'1
VM
RF
""
~Ii.
BF~~\~~~~~-+~'
GM~~~~~~~--~~
sw
ill.
"~:~',r
sw
ST
1 sec
0.5 mv
1 year
HI
111I~I'r
UI.._
TI~"
.11,
'I,'
sw
ST
1 sec
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).
54 Development of Gait
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
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
LG+
Terminal SW
Terminal SW
VM+
Ak ~rt
0-
H M+
LG+, BF+
~F+
TA+
BFl
LO+
Squat
Backward
Forward
++)
(++ )
(++)
(++ )
(+ )
(_)
(+)
(_)
(_)
(++)
(-)
(-)
(-)
(-)
(++)
U -------n-------D.-------.[}-------. ----
TA+, RF+
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.
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
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
After birth
Development 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
1 year
1.1 years
Infant walking
1.3 years
It
3 years
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.
Neonatal
st epping
Young infant
stepping
Infant supported
walking
Infant
walking
:: ~ ~:-
Reciprocal pattern
:: ~~:
Co-contraction 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,
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.
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.
After birth
Birth
RF+
Initial SW
Dorsiflexion
I".
Hip flexion
(++)
(++)
(++)
(++)
(++)
(++)
(++)
(+),(-)
(++)
(++)
(++)
(++)
(++)
(+),(-)
(++)
(+),(-)
Development of gait
TA+
Initial SW
~~.
Neonatal stepping
1 month
2 months
................................ _.....
1.1 years
Infant walking
1.3 years
......................... _.......
3 years
Mature walking
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.,
TA+
VM+
TA+, RF+
LG+, BF+
~,~
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.
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.
TA
LG
RF
SF
.EMG patterns began to decrease or disappear at about 1-3 months after learning to walk
.+11]+0
11]+0
..
......
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.
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+</-t-"'-M.w.-'-I'{o/'I'IHI'+~#I.t~
sw
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)
TO
SW
He
TO
2 years 11 months
I--'
RF ........O-+\\I>.-.....-Ho!--.,.."..\O__
BF
ifi---I---IIU/If,-I-/IfM'-
GM
WIir"---t~<fr'iI'IIr-t--'fMVrSW
ST
1 sec
j,t
'['I
I'
'"
SW
10.5 mv
[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)
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
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
SW
ST
SW
ST
1 sec
_-----"-.::,=.:'c'-----'10.5 mv
1 year 9 months (8-4)
/0.5 mv
~~\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
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).
KNEE JOINT
BACKWARD
ERECT
KNEE FLEXED
FORWARD
BACKWARD
ERECT
FORWARD
STANDING
POSTURE
'--
Tibialis anterior
(+)
(- )
Gastrocnemius
(-)
(-)
(+)
(+),H
(-)
H.Cl
(+)
(-)
(+)
(+),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
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!
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).
1. Stance Phase
ST-TA, ST-RF: Bursts or continuous activity of the tibialis anterior
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
~~ ,
GM
sw
ST
I
!
""1
u/u>
'"""T"
sw
ST
ERECT POSTURE
sw
ST
ADULT WALK
FORWARD
0.5
FORWARD
WITH KNEE FLEXED
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
STANCE
10
11
SWING
LG --;-'.
KNEEFLEX.
FC~L--li
VTR -
HC FF
HO
STUMBLE
TO
FF HO
lit-
10
+11
FOOT CONTACT
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
It thus becomes clear that when a baby first begins to walk, muscle
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
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.
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.
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).
ST
(Al
rtrtftft ft
sw
I~
TA
ST
.....
~~
"0'
'.
LG '
VM
I'!
RF
SF .l,
,ul"I,li
,~~
''''''1'
GM
(U
TA"
.L,wI,,~.
"",rr"1tr~
LG
'.;.~fI/I~\1i!I, ~"'14,....
RF
SF
""""-
..
#-'
SWING
\~~o\iiI~i;.<""'I"
~/~---
STANCE
1 sec
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,
~~
20 days after birth
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,
RF '~
BF--
r.
SWING
--"'@t~
-:.: f1
I"
........
STANCE
1 sec
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).
5W
5T
(Rl
TA---4--~---4~--~~--~--~-r--~
LG __~~~~~~~~~~--~~~-
VM~~-r~~'~~~~~~~__+-~~_
RF--~--~~-~~--_4~_.--~_r----
,.."," ""--+--t-...~--+-+,~-
GM
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,
(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
If" r~'!""r
LG
~, L!a...
1,11..11.,[1
" II"
.J,
RF
.L" "'1
SF
SWING
"'.
STANCE
1 sec
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),
ffftJttftl
r~' I
TO
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
0.5 mv
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) .
Sign of instability
Interpretation
Ankle
SW-LG
Unstable
Knee
SW-VM
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.
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.
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
Muscle
0-1
1-3
3-4
VM
H
H
(+), H
(+),H
(+), partly H
(+), partly H
LG
LG
(-)
VM
H, partly (+)
(+), H
(+), (-)
(+), partly H
LG
(-)
(-)
H, partly (+)
(-), partly (+)
(+), (-)
(+), (-)
H
(-)
(+), H
(+), (-)
(+), partly H
(+), partly (-)
H
(-)
(+), H
(+), (-)
(+), partly H
(+), partly (-)
VM
LG
VM
LG
VM
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
1-3
3-4
LG
VM
(-)
(+), H
(+), (-)
(+), partly H
(+), partly (-)
6-12
H, partly
H , partly
(+)
(+)
(+): noticeable activity, (-): no activity, (+), (-): instances of noticeable activity and of no
activity intermingled.
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
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
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.
Young infant
period
Neonatal
period
1.------+
4----------------------------------+.
Frequency
of
occurrence
(EMG)
/-----'/
'/
'/
'/
"'1....
'/
'/
'/
~--.~
Birth
1 month
.....
.............~~~.~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.
We believe that infants at around 3 months after birth begin to feel instability.
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
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. 6-3. Form of walking at 1 month after the stroke (left) and at 1 month after
learning to walk (right).
fff/f
( R)
TA
LG
VM
RF
SF
GM
( L)
TA
LG
VM
RF
SF
GM
ST
sw
1 sec
85.6 years
0.5 mv
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.
ST
SW
(R)
(L)
TA~~~~1~~~~~~~n~~~~~~;H~~~~~-
SW
ST
1 sec
86.6 years
0.5 mv
fl ~
86.6 years
TA
LG
VM
RF
SF
GM
KNEE
EXTENSION
FLEXION
SmlNG
1 sec
SQATTING
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.
1 month
7 months
after stroke
after stroke
ST-TA
(++)
SW-LG
(+)
(+)
H, partly
(+)
1 year 7 months
after stroke
H , partly
ST-LG
SW-VM
(+)
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
ST-LG
(++), partly
H, partly
H
(+)
after stroke
H, partly
(-)
(-)
(++)
(++)
SW-VM
(-)
(-)
ST-VM
(++)
(++)
(++)
ST-RF
(++)
(+)
(+)
ST-BF
(++)
(++)
(++)
ST-GM
(++)
(++)
(+)
(+)
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.
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
~
HC
TO
TA
LG
VM
RF
BF
GM
sw
ST
1 sec
CANE
I 0.5 mv
sw
ST
1 sec
HAND CART
sw
ST
1 sec
0.5 mv
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,
gluteus maximus.
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.
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
Primale Morphopbysiology,
LocomOior Anal)' cs
and Human Bipedali m
123
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
/'""
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
Ankle
Code
Interpretation
Indication
ST-TA
Very unstable
SW-LG
Unstable
ST-LG
Slightly unstable
SW-VM
Very unstable
ST-VM
Very unstable
ST-RF
Very unstable
ST-BF
Slightly unstable
ST-GM
Slightly unstable
Knee
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
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
132
133
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
~lTIm~liJf~pJf
135
Development of Gait