JOURNALOFNEUROPHYSIOLOGY Vol. 68, No. 2, August 1992. Printed

in U.S.‘4.

Development of Anticipatory Postural Adjustments During Locomotion in Children H. HIRSCHFELD AND H. FORSSBERG Karolinska Institute, The Nobel Institute&v Neurophysiology SUMMARY

AND

CONCLUSIONS

1. Anticipatory postural adjustments were studied in children (6- 14 yr of age) walking on a treadmill while pulling a handle. Electromyographs ( EMGs) and movements were recorded from the left arm and leg. 2. Postural activity in the leg muscles preceded voluntary arm muscle activity in all age groups, including the youngest children (6 yr of age). The latency to both leg and arm muscle activity, from a triggering audio signal, decreased with age. 3. In older children the latency to both voluntary and postural activity was influenced by the phase of the step cycle. The shortest latency to the first activated postural muscle occurred during single support phase in combination with a long latency to arm muscle activity. 4. In the youngest children, there was no phase-dependent modulation of the latency to the activation of the postural muscles. The voluntary activity was delayed during the beginning of the support phase resulting in a long delay between leg and arm muscle activity. 5. The postural muscle activation pattern was modified in a phase-dependent manner in all children. Lateral gastrocnemius (LG) and hamstring muscles (HAM) were activated during the early support phase, whereas tibialis anterior (TA) and quadriceps (Q) muscles were activated during the late support phase and during the swing phase. However, in the 6-yr-old children, LG was also activated in the swing phase. LG was activated before the HAM activity in the youngest children but after HAM in 14-yrold children and adults. 6. The occurrence of LG activity in postural responses before heel strike suggests an immature (nonplantigrade) gating of postural activity. A similar shift of LG activity is seen earlier during the maturation of the locomotor pattern into a plantigrade gait. 7. The considerable temporal reorganization of voluntary and postural activity in relation to locomotor activity and the development of phase-dependent modulation of the postural activation pattern implies a continuously increasing integration of these motor control systems with age.

INTRODUCTION

Humans use a large repertoire of innate and acquired motor acts (programs), often integrated to form complex motor behaviors. Before a voluntary movement, humans induce postural adjustments that compensate for the subsequent perturbation of the equilibrium (Belenkii et al. 1967; Bouisset and Zattara 198 1; Cordo and Nashner 1982; Massion 1984). During locomotion these postural adjustments are integrated with the locomotor activity (Hirschfeld and Forssberg 199 1; Nashner and Forssberg 1986). The onset, temporal sequence, and amplitude of the postural activity in various muscles are modulated, creating specific muscu542

and Department

ofpediatrics,

S-104 01 Stockholm,

Sweden

lar activation patterns in each phase of the step cycle (Hirschfeld and Forssberg 199 1) . This phase-dependent modulation is required to produce functional adjustments during the continuously changing biomechanical condition. For example, a pull on a handle causing a forward body sway is preceded by postural activity in the gastrocnemius and hamstring muscles in the beginning of the support phase, but in the tibialis anterior and quadriceps muscles during the end of the support phase. The first response resists the forward rotation of the tibia and the latter reduces the forward propulsion exerted by the calf muscle activity. It has been suggested that the anticipatory postural activity in conjunction with voluntary movements is controlled by frontal areas of the cortex, by cerebellum and by the basal ganglia (Birjukova et al. 1989; Gurfinkel and Elner 1988; Paltsev and Elner 1967; Traub et al. 1980). Studies on the phase-dependent modulation of cutaneous reflex responses imply that modulation of the transmission in reflex pathways can be exerted by locomotor generating circuits in the spinal cord (Andersson et al. 1978; Crenna and Frigo 1984; Dietz et al. 1987; Drew and Rossignol 1985; Duysens et al. 1990; Forssberg et al. 1977, 1979; Garrett et al. 1984; Gossard et al. 1990). The development of phase-dependent anticipatory postural responses thus likely reflects the maturation of both high-level motor centers and spinal cord mechanisms, as well as the integration among the neural systems involved in the control of voluntary movements, postural adjustments, and locomotion. Motor tasks requiring interaction between different systems seem to develop late during childhood (e.g., Forssberg and Nashner 1982), whereas motor activities controlled by one system, such as locomotion or postural responses to external perturbations, mature earlier (Forssberg 1985; Forssberg and Nashner 1982; Shumway-Cook and Woollacott 1985; Sutherland et al. 1980). The purpose of this study was to investigate the development of the integration among voluntary, postural, and locomotor systems. Children (6- 14 yr of age) were instructed to pull on a handle while walking on a treadmill. The temporal relationship of the activity from the different motor systems and the modulation of the postural responses in relation to the step cycle was recorded by means of electromyography (EMG) and movement analysis. METHODS

Subjects Thirteen children with normal in the study. They were divided

motor development participated into three groups: 6-yr-old (~2 =

0022-3077192 $2.00 Copyright 0 1992 The American Physiological Society

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on September 19, 2018. Copyright © 1992 American Physiological Society. All rights reserved.

DEVELOPMENT

OF

ANTICIPATORY

POSTURAL

5 ) , lo-yr-old ( y2= 5 ) , and 14-yr-old ( y2= 3 ) children. Eight adults were investigated in an earlier study, with the use of the same experimental setup (Hirschfeld and Forssberg 199 1). Their data were used for comparison.

ADJUSTMENTS

DURING

during the 100 ms after from the absolute latency lay of onset. The analysis matched pairs test were

LOCOMOTION

543

the first activated muscle, i.e., 100 ms in all muscles independent on their deof variance (ANOVA) and Wilcoxon used to assess statistical significance

(P < 0.05). Protocol The children walked at a comfortable speed on a treadmill (0.91.2 m/s), holding a handle in the left hand with the elbow flexed ~90”. They were instructed to pull on the handle immediately following a tone signal. The handle was free to be moved in the frontal plane but could not be moved in the anterior-posterior direction. Force transducers under the shoes monitored foot contact and were used to trigger the tone signal. The tone was presented in each IOO-ms interval of the step cycle, and 10 trials were performed in each interval. More than seven unperturbed step cycles separated successive tone signals. The children were instructed to apply a low steady pulling force (9 N), which brightened up the left eye of a stuffed animal connected to the load cell of the handle. After the tone signal, they were instructed to pull as fast as possible above 50 N and then to release the pull. If they pulled hard enough to include the trial for analysis (50 N), both eyes brightened up (for further details see Hirschfeld and Forssberg 1991).

Data recording and analysis EMGs were recorded by surface silver/ silver chloride electrodes placed over the muscles of the left arm and leg: biceps brachii (BIG), lateral gastrocnemius (LG), tibialis anterior (TA), biceps femoris (HAM), vastus lateralis (VL), and rectus femoris (RF). The EMG signals were amplified 100 times by differential amplifiers, attached to the skin near the electrodes, and thereafter bandpass filtered between 50 Hz and 1 kHz and full-wave rectified. Leg movements were recorded by a Selspot System with the use of six light-emitting diodes (LEDs) attached to the left leg to allow the ankle, knee, and hip joint angles to be calculated from vectors drawn between adjacent LEDs. The coordinates of the LEDs were sampled at 156 Hz and led to a minicomputer and stored for later off-line analysis together with EMG recordings and signals from foot transducers, tone signal, and handle force; all sampled at 312 Hz. Step cycle phases and events were determined by left and right shoe force transducers and Selspot kinematic data. Step cycle events were as follows: LHS, left heel strike; RTO, right toe off; RHS, right heel strike; LTO, left toe off; TRST, transition to stance, between 100 and 0 ms before LHS; DS 1, double support 1, from LHS to RTO; SS 1, single support 1, from RTO until the right ( swing) leg passes the left (support) leg at midstance of the support leg; SS2, single support 2, from midstance of the left leg until RHS; DS2, double support 2, from RHS to LTO; SW 1, swing 1, before the swing leg passes the support leg; SW2, swing 2, after passing the support leg (see Fig. 1, Hirschfeld and Forssberg 199 1).

Analysis A graphics terminal was interactively used to mark the events during a trial, i.e., tone signal, EMG bursts, and phases of the step cycle. The onset of the EMG burst was marked at the time when the activity exceeded +2 SD of the base level activity. A postural adjustment was categorized to the phase in which it appeared (i.e., not when the stimulus occurred). The temporal coordination of the muscle activation pattern was analyzed by determining the delay between the onset of activated leg muscles. The absolute latency was defined as the latency from the audio signal to the first activated muscle. The intensity was estimated by calculation of the area under the full-wave rectified and filtered EMG envelone

RESULTS

Integration oj’voluntary, postural, and locomotor activity

All children activated postural muscles before initiating the pull. The postural activity was superimposed on the ongoing locomotor activity. The latency from the tone signal to both the leg and arm muscle activity decreased with age (P < 0.05 for both latencies, Table 1). The mean latency to the onset of arm muscle activity decreased between 6 yr and 10 yr of age, and the mean latency to the onset of postural muscle activity decreased only after 10 yr of age (P < 0.05 in both cases). In older children ( 10 and 14 yr of age) the latency to both arm and leg muscle activity varied during different phases of the support phase. The shortest latency to the onset of postural activity was combined with a relatively long latency to the arm muscle activation, resulting in a long delay between the onset of leg and arm muscle activity in the beginning of the single support phase ( SS 1). This delay was significantly shorter during DSl and DS2 than it was during SSl, because of a relatively longer latency to the postural muscle activity and a significantly shorter latency to the arm muscle activity (P < 0.05 ). In the 6-yr-old children, the latency to the postural muscle activity was constant during the entire support phase (Table 1). Instead, the latency to the arm muscle activity varied to be longest when the postural muscles were activated during DS 1, i.e., the pull occurred in later parts of the support phase. The arm muscle latency gradually decreased during subsequent phases to become shortest during DS2 (P < 0.05, DSl, SSl, and SS2 relative to DSZ). The difference was almost two-fold. This resulted in a decreasing delay between leg and arm muscle initiation as the support phase progressed. The short delay during DS2 was in the TABLE

1.

Latency qfthe 1st activated leg and arm muscle DSl

SSl

ss2

DS2

Mean

Leg, ms 6yrold 10yrold 14yrold Adults Arm, ms 6 yrold 10yrold 14yrold Adults Delay, ms 6 yr old 10 yr old 14 yr old Adults

313-t 291-t 256-t 2271:

48 48 58 64

581 ~fr 365.k 343 -t 285t

131 59 58 61

268 t 74+ 87 AI 58 f

99 32 25 49

314+ 245+ 216+ 199t-

43 49 32 45

312+ 2782 2875 263k

80 71 61 82

452 -t 66 399-t 54 379 + 127 3OOt90

410 -t 104 391 -t 66 372 + 41 3282 94

138 + 154k 163 IL 101 zk

98 IL 113+ 85 k 50 +

Values are means -t SD. Latencies during different phases of the support the support and swing phase (Mean) support: SS, single sutx~0r-t.

56 58 59 57

90 63 49 38

312H33 362+78 225+32 212k41

307+ 15 304t-38 251+25 233k31

368 -t 87 423+91 321 + 57 265k66

420 + 85 384~24 353 k 25 280+25

56-t28 61 AI 16 96 + 55” 53 AI 49

121 80 102 47

-t77 -t 40 + 30 + 32

of the 1st leg and arm muscle activity phase and the combined latencies for are shown for all groups. DS, double

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on September 19, 2018. Copyright © 1992 American Physiological Society. All rights reserved.

544

H. HIRSCHFELD

TRST

AND H. FORSSBERG

DS I

LG

HAM

RF ms

BIC I

I

?

LHS RTO

RHS

LHS RTO RHS

LHS RTO

++

RHS

RHS

tt

?

LTO LHS

RHS

t

LTO LHS

FIG. 1. Ensemble average of filtered and rectified electromyographs (EMGs) of 8 unperturbed step cycles (fine lines) compared with the average of 8 perturbed step cycles (heavy lines) for a 6-yr-old child during various phases of the step cycle. The vertical line with T on the top indicates the tone signal, and the following vertical line denotes onset of the 1st postural muscle activity, i.e., when the EMG of the perturbed step cycles deviated +2 SD from that of the unperturbed step cycles. Arrows show step cycle events. The postural activity, including both anticipatory and compensatory muscle activity, are highlighted by the use of hatched areas for the dorsal leg muscles and dotted areas for the ventral leg muscles and grain area for biceps brachii (arm). Note that the magnitude of the anticipatory muscle activity was calculated only during the 1st 100 ms after the 2nd vertical line. Stick figures above the EMGs illustrate the approximate orientation of the body and muscles involved in the postural response. TRST, transition to stance; DS, double support; SS, single support; SW, swing; LHS, left heel strike; RTO, right toe off; RHS, right heel strike; LTO, left toe off.

same range as in adults, whereas it was four to five times longer than in adults during DS 1 (P < 0.05 ). The shift of the latency to the arm activity in the youngest children reflected a tendency to initiate the arm pull before TRST

1OY

the contralateral foot placement (i.e., in DS2) even though the postural response had already occurred in the beginning of the support phase (DS 1). In older children and adults, there was another tendency, although not as pronounced,

ss I

DS I

DS 2

LG TA

4 LHS

RTO

RHS

LHS

RTO

RHS

LHS

RTO

RHS

RTO

RHS

LTO

1

44 RHS

I

LTO

4 LHS

FIG. 2. Ensemble average of filtered and rectified electromyographs of 8 unperturbed step cycles (fine lines) compared with the average of 8 perturbed step cycles (heavy lines) for a lo-yr-old child during various phases of the step cycle (for details see Fig. 1).

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on September 19, 2018. Copyright © 1992 American Physiological Society. All rights reserved.

DEVELOPMENT

OF

ANTICIPATORY

POSTURAL

ADJUSTMENTS

DURING

LOCOMOTION

545

DS 2

14Y LG TA

HAM

RF

I

I I

I

4

44 LHS

RTO

4 RHS

LHS

FIG. 3.

RTO

RHS

LHS

Ensemble average of filtered and rectified electromyographs with the average of 8 perturbed step cycles (heavy lines) for a 14-yr-old details see Fig. 1).

RTO

ofthe postural

RHS

LTO

LHS

LTO

LHS

RTO

of 8 unperturbed step cycles (fine lines) compared child during various phases of the step cycle (for

to initiate the pull around either the ipsilateral or contralatera1 foot placement with relatively short latencies during the end of swing (SW2, TRST; not shown) and the end of support (DS2). In summary, there was an overall decrease of the latencies to both postural and voluntary activity with age. The latency to the postural activity was modified in phase with the locomotor activity in older children and adults but not in the youngest children. The 6-yr-old children had an “asymmetric” delay of arm muscle activity during the beginning of the ipsilateral support phase, whereas older children had a “symmetric” biphasic modulation of the latency to arm muscle activity that was reciprocal to that of the postural activity. Spatial organization

RHS

activity

The combination of muscle activity in the various phases of the step cycle was mainly the same as previously reported in adults (Hirschfeld and Forssberg 199 1) ; namely HAM and LG during the first part of the support phase and TA, VL, and RF during the later part of the support phase and the swing phase (Figs. l-3). TA and LG were coactivated during DS 1 and SS 1 in the 6- and lo-yr-old children but not in older children (Fig. 4). The amplitude of the muscle activity was modified depending on the phase of the step cycle in all subjects, but there was an age-related phase shift of the maximal amplitudes (Fig. 4). Six-year-old children activated LG maximally during TRST, with gradually decreasing amplitude in DS 1 and SS 1. In the same age group, there was no HAM activity in TRST but maximal HAM activation in DS 1 and SS 1. In the older age groups and adults, the relation was the opposite. LG was maximally activated during DS 1, and the HAM resnonse was largest during TRST (Fig. 4). In the

6Y

%

100

.‘f 1 5o ;jshq$-+< $q+ w*‘r-----0I --------~---------------------_____+-___ ------------------:----1QY loo 1

I,

4

5o++i:,i-‘:::i”\;.’ 0:3 ,__ __________ I------------f ---------‘-------------------------------_-_ ____ ----_ ---14 Y

TRST

f LHS

DSI

1 RTO

SSI

ss2

t RHS

DS2

t LTO

SWI

SW2

t LHS

FIG. 4. Electromyographic (EMG) intensity measured for all groups (6, 10, and 14 yr old: mean -+ SD of means) during various phases of the step cycle. The area between the EMG traces of the perturbed and unperturbed step cycles was calculated during the 1st 100 ms after the 1st activated postural muscle. The reference value ( 100% ) is equal to-the average EMG during the 100-ms interval when each muscle had its maximal intensity during unperturbed locomotion. The symbols for different muscles are shown in the bottom right corner. Vastus lateralis (VL) is used for analysis during transition to stance (TRST) to single support 1 (SS 1) and rectus femoris (RF) durjng single support 2 (SS2) to swing 2 (SW2). Step cycle chases are indicated on the x-axis.

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on September 19, 2018. Copyright © 1992 American Physiological Society. All rights reserved.

546

H. HIRSCHFELD 6Y

10

HAM

Y

AND H. FORSSBERG

Temporal

HAM

organization

of the muscle activation pattern

Inthetwoyoungestage groups, LG was activated before HAM, during TRST and DS 1 (P < 0.05; Fig. 5). In these phases the LG onset was delayed in relation to HAM with increasing age and was initiated after HAM in the 14-yr-old children (P > 0.05) and adults (P < 0.05, Figs. 3 and 5). Later in the support phase (SS 1 ), the temporal relation switched, i.e., LG was activated after HAM in the two youngest age groups (P > 0.05) and before HAM in the older children and adults (P < 0.05 ). There was no age-related change of the order of muscle activation during the later part of the step cycle.

SEQUENCEOFMUSCLEACTIVATION.

14 Y

adult

HAM

HAM

ONSET OF LG ACTIVITY 20 ms/div

FIG. 5. Temporal relation of lateral gastrocnemius (LG) and biceps femoris (HAM) onset (mean + SD of means) is plotted for the various phases for all groups. HAM is used as the reference, and LG activity preceding HAM is plotted to the Zeft.Note time scale.

middle of the support phase (SSl, SS2), the two combinations (TA-VL,RF and LG-HAM) were activated in conjunction. In lo-yr-old children, LG and VL,RF were usually small, resulting in a TA-HAM response. The relationship between TA and VL,RF varied in the later part of the support phase. RF was maximally activated during the swing phase in all age groups, whereas the activation of additional muscles shifted with age. Six-year-old children activated LG and HAM as a part of the anticipatory swing response (Figs. 1A and 4). This never occurred in the older children. In addition to the anticipatory postural activity, the leg muscles were active after the initiation of the pull (i.e., cornpensatory postural activity) (see Hirschfeld and Forssberg 199 1). TA was activated during later parts of the support phase after anticipatory responses during early support phase, either as a continuation of the anticipatory activity (6 and 10 yr) or as new activity ( 14 yr, Figs. l-3). The 6and IO-yr-old children exhibited large compensatory LG activity during the swing phase after anticipatory adjustments in late support or early swing. Fourteen-year-old children and adults never had any postural activity in LG during the swing phase (see below). All children and adults had compensatory activation of LG in the beginning of the support phase. SW

LIIS

‘IX S’I

IN RELATION

TO FOOT PLACEMENT.

With increasing age, the onset of LG activity was delayed in relation to foot contact (Fig. 6). All children activated LG before heel strike in the anticipatory TRST response (P < 0.05), whereas LG was activated after heel strike in adults (P -C 0.05 ). LG was activated before foot contact (255 t 96 ms, mean t SD) as a part of the anticipatory swing response only in the 6-yr-old children. In the IO-yr-old children, LG was also activated before foot contact ( 118 t 23 ms) but now as a part of the compensatory activity, i.e., after the initiation of the pull. Adults and 14-yr-old children activated LG as a compensatory activity at or after heel strike. In summary, even the youngest children used patterns of muscle activation similar to these seen in adults. There was, however, a major shift in the postural response of the onset of the LG muscle with age, from onsets early in the swing phase and TRST, to onsets only after heel strike. LG was also delayed in relation to the HAM activity. Kinematics The change o f the postural activity during de velopment was reflected by changes of the kinematics. The activation of the calf muscles before foot contact reduced the ordinary dorsiflexion of the ankle in 6- and lo-yr-old children (Fig. 7, TRST response). This prevented a prominent heel strike with the forepart of the foot placed on the ground. Fourteen-year-old children and adults had a regular dorsiflexion of the ankle followed by a heel strike. After foot contact, the dorsiflexion of the ankle was reduced in all subjects (Fig. 7), i.e., the forward rotation of the tibia was decreased corresponding to the calf muscle activation during the beginning of the support phase. There was also an increased LIIS

LIIS

6YV 10 Y 14 Y Adult

I

I

I

I

I

I

I

I

I 50 msfdiv

FIG. 6. Temporal relation of lateral gastrocnemius (LG) onset and heel strike plotted in swing (SW) and transition to stance (TRST) responses for all groups (mean + SD of means). Heel strike is used as a reference, and LG activation before left heel strike (LHS) is plotted to the left. In the graph to the right, the onset of LG activity during unperturbed locomotion is plotted in relation to heel strike.

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on September 19, 2018. Copyright © 1992 American Physiological Society. All rights reserved.

DEVELOPMENT deg

OF ANTICIPATORY

POSTURAL

ADJUSTMENTS

T 180

f

l-l0

l

LOCOMOTION

ss 2

OS I

TRST

DURING

B

T

180

DS 2 T

B

180,

l-10

140.

180’

180

180’

140 -

140

150 -

lS0

1’0’

110

110

110 +

l

547

B

. Lo

10 Y

T

‘HZ

T

B

ho

L”:

i”S

;TOA”:

B

Lo

T

L”)S iTO R”2 ho

T

B

180 -

180 -

l-IO-

140 l-

180’

180 *

140-

140 -

lfo110

150

150 *

110 w

110.

B

M L;oL”:

T

iTO

T

B

i”S

LHZ ;ro

RH:

T

B

B

Lo

LHi

AT0 R”:

Lo

B

T

180 110 -

150 -

Lto

L”!

ko

;“S

LH:

iTO R”S Lo

t LHS

t

RTO

t RHS

t LTO

FIG. 7. Ensembled average of kinematics of hip, knee, and ankle movements are plotted for the different unperturbed (thin lines) and perturbed support phases (thick lines; n = 8) for 1 subject of each age group. Step cycle events are marked on the x-axis. The vertical line with T on the top indicates the tone signal, and the following vertical line ( B) indicates biceps brachii (arm) activity onset.

“knee flexion wave” during single support, in agreement with postural activity in the knee flexor muscles, e.g., HAM and LG. Postural responses elicited during DS 1 also reduced the dorsiflexion of the ankle during the succeeding support phase in all age groups (Fig. 7, DS 1). In addition, there was a reduced hip extension in the two youngest age groups. In all age groups the ankle extension during the end of the support phase was reduced after TRST and DS 1 responses, corresponding to the compensatory postural activity in pretibia1 ankle flexors and decreased activity in ankle extensor muscles (Figs. l-3 and 7). After SS2 and DS2 responses the ankle extension before lift-off was reduced. In the 6-yr-old children the reduction was smaller than in older children, but in addition the rate of extension was lower and the lift-off delayed (Fig. 7). Because this ankle extension against the ground is one of the main propulsive movements during human plantigrade gait (Elftman and Manter 1935 ) , the reduced ankle exten-

sion reflects a decreased forward propulsion of the body. In the 6-yr-old children the ankle usually continued to extend during the swing phase in combination with maintained knee and hip flexion, whereas older children and adults maintained the reduced ankle extension during the swing phase. Pull responses during the swing phase were accompanied by an increased hip flexion in all age groups. This produced an increased forward swing of the leg and a longer stride length, enhancing the deceleration of the body after foot placement. In the 6-yr-old and in two of the IO-yr-old children, the dorsiflexion of the ankle was reduced and turned into extension before foot contact. This caused the foot to be placed on the forepart, as after the TRST response. DI sc u s sI o N

Natural and familiar motor tasks as well as advanced athletic skills often require an interaction between various

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on September 19, 2018. Copyright © 1992 American Physiological Society. All rights reserved.

548

H. HIRSCHFELD

AND H. FORSSBERG

motor control systems. The coordinative structures controlling equilibrium during locomotion when voluntary movements are performed have to consider the continuously changing biomechanics of the body. This process involves I) programming of the voluntary movement, 2) calculation of the resulting perturbation of the equilibrium, 3) programming of appropriate postural activity to compensate for the perturbation (Belenkii et al. 1967; Bouissett and Zattara 198 1; Crenna et al. 1987; Massion 1984; Nashner and Cordo 198 1) that is properly adapted to the actual phase of the step cycle (Hirschfeld and Forssberg 199 1; Nashner and Forssberg 1986), and 4) execution of the programmed voluntary and postural activity in an appropriate temporal sequence (Cord0 and Nashner 1982). The present study demonstrates that even 6-yr-old children produce postural activity before arm movement and that this muscle activity is added to the locomotor muscle activity and modified depending on the phase of the step cycle. However, the phase-dependent modulation of the postural activity and the timing of the voluntary and postural activity are not mature. The presence of anticipatory postural activity during locomotion in 6-yr-old children corresponds with earlier studies reporting anticipatory activity as early as 18 mo in a similar task during standing (Forssberg and Nashner 1982). Haas et al. ( 1989) have described anticipatory postural activity in the pretibial muscles, shifting the center of mass forward, before tip toe rising in 4-yr-old children. It was impossible to define the emergence of the anticipatory behavior in any of these tasks because of difficulties instructing young children to adequately perform the complex movements. In contrast, this has been possible in a lifting task. During the second year, children had already developed an anticipatory strategy for programming isometric finger forces to the weight of an object (Forssberg et al. 1992). The brain needs access to an internal representation of the biomechanical characteristics of the body (structural, kinematic, and dynamic) to program postural activity properly (Clement et al. 1984; Clement and Lestienne 1988; Gurfinkel and Levik 1978; Lestienne and Gurfinkel 1988a). The phasic modulation of postural activity during the locomotor cycle suggests a continuous update of such an internal representation. The location and mechanisms of this phasic modulation were discussed in a previous paper (Hirschfeld and Forssberg 199 1) . One possibility is a programming of the final postural activity in cortical and subcortical circuits, in direct conjunction with the programming of the voluntary activity (Gahery and Massion 198 1; Massion 1984). This would imply a continuous input of information to these circuits regarding leg movements (sensory) and neuronal activity of the spinal cord circuits (motor). Another alternative is a distributed organization of the internal representation involving spinal cord mechanisms. This latter suggestion is based on studies of phase-dependent modulation of cutaneous reflexes during cat locomotion (Duysens and Loeb 1980; Duysens et al. 1990; Forssberg et al. 1975, 1977; Rossignol et al. 1988) due to central gating of the transmission in reflex pathways (e.g., presynaptic inhibition) by the locomotor generating

circuits in the spinal cord (Andersson et al. 1978; Gossard et al. 1990). Phasic gating of transmission in descending rubro- and vestibulospinal tracts in the locomoting cat has also been shown (Orlovsky 1972; Russel and Zajac 1979; Arshavsky and Orlovsky 1985). The presence of similar phasic gating mechanisms in the spinal cord during human locomotion is supported by studies on phase-dependent cutaneous (segmental) reflexes (Capaday and Stein 1986, Crenna and Frigo 1984; Drew and Rossignol 1985; Duysens et al. 1990; Garrett et al. 1984). It was therefore suggested that postural activity is initiated at a cortical-subcortical level and later modulated and tuned in the spinal cord by locomotor generating circuits and segmental reflex pathways (see Hirschfeld and Forssberg 199 1) . The postural activation patterns of the 6-yr-old children differ from those of the adults, especially during the late swing and early support phases. In responses during these phases, the ankle extensor (LG) is activated before foot contact and is relatively larger and earlier than the HAM activity. The delay of the LG activity until after foot contact and the reduction in responses during early support phase is similar to a shift of the LG locomotor burst earlier during development [ 2-4 yr of age (Fig. 6)]. This delay of the LG-locomotor activity has been suggested to be one of the main determinants of plantigrade gait (Saunders et al. 1953; Forssberg 1985; Forssberg et al. 199 1) because it allows the foot to be placed with a heel strike and the late activity produces a major propulsive force during the end of the support phase (Elftman and Manter 1935; Winter and Robertson 1978, 1983). There are different theories on the mechanisms underlying the transformation of the locomotor pattern during development (Forssberg 1985; Thelen et al. 1987; Thelen and Whitley-Cooke 1987). These results suggest that the neural locomotor networks in the spinal cord continue to gate the postural responses according to the earlier nonplantigrade pattern several years after transformation to plantigrade gait ( cf. Forssberg 198 5 ) . The temporal coordination of voluntary and postural movements in relation to locomotion changes considerably with age, implying a gradually increasing integration of the involved motor control systems. A general reduction of the latencies with age is in agreement with shorter central processing in reaction time tasks (Connolly and Elliott 1962). There is no phase-dependent influence on the latency to the first postural muscle activation in the younger children. With increasing age the latency during the single support phase is gradually decreased compared with the other phases of the step cycle. The relative reduction is combined with longer latency to the arm activity, allowing longer time for the postural activity to stabilize the body before the pull. This development of a “biphasic” modulation of the latenties to both arm and postural activity seems to be a functional adaptation to differences in stability during different phases of the step cycle. The period around foot contact seems then to be the most stable, with long latencies to postural and short latencies to arm activity. This is also the period during which subjects pace their pull -without a trigger signal (Nashner and Forssberg 1986). Although this biphasic modulation of the latency to arm muscle activity is small in the 6-yr-old children, they have a “monophasic” modulation with long latencies during DS 1

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on September 19, 2018. Copyright © 1992 American Physiological Society. All rights reserved.

DEVELOPMENT

OF ANTICIPATORY

POSTURAL

followed by a gradual reduction during the progressing support phase. At the same time, the latency to the postural activity is constant. This may have several explanations. The anticipatory postural activity might not be sufficient to allow a pull during early support phase. This is supported by the lack of effects on the leg kinematics after the postural LX1 response. It might also be a mechanical constraint for the young child to induce an arm pull in the beginning of the ipsilateral support phase, opposing the forward rotation of the thorax and the swinging arm. Probably there is also a conflict between neural processes intending to activate the arm muscle in different patterns. As in the cat (Miller et al. 1975 ), propriospinal pathways from rhythm-generating circuits in the lumbar spinal cord are probably involved in the homolateral out-of-phase coordination between leg and arm muscles. This neural activity underlying interlimb coordination constrains the timing of the pull. In 6-yr-old children, leg-arm coordination has been well developed for several years (B. C. L. Touwen and G. Cioni, unpublished observations). Although arm swings are not present when holding the handle, the interlimb neural activity might still influence the motor performance (cf. Fukuda 196 1; Hellebrandt et al. 1956). Although adults and older children manage to integrate the activity from locomotor generating circuits and voluntary motor activity, younger children seem to delay the release of the arm activity until it falls into phase with the locomotor activity, i.e., when the arm starts to swing back. Whatever the mechanisms, this monophasic modulation of the latency to the pull activity creates an asymmetric behavior in the youngest children compared with older subjects. Although no differences were found between postural adjustments in ipsi- and contralateral legs in adults (cf. Hirschfeld and Forssberg 199 1 ), this could likely be the case in the youngest children. Hence this study has shown that the motor development leads to an increased flexibility of motor performance by an increased integration between various motor control systems, modifying the motor output to the requirements of the environment. We thank Dr. G. Orlovsky for valuable comments on the manuscript, Dr. S. J. Schotland and Ch. T. Leonard for correcting the English text, and V. P. Stokes for technical assistance. This study was supported by the Swedish Medical Research Council ( 4X-5925 ) , Norrbacka-Eugenia Stiftelsen, Stiftelsen Solstickan and Omsorgsnamnden in Stockholm. Address for reprint requests: H. Hirschfeld, Dept. of Pediatrics, Motor Control Laboratories Q4, Karolinska Hospital, S- 104 0 1 Stockholm, Sweden. Received 1 July 199 1; accepted in final form 12 March 1992. REFERENCES O., FORSSBERG, H., GRILLNER, S., AND LINDQUIST, M. Phasic gain control of the transmission in cutaneous reflex pathways to motoneurones during ‘fictive’ locomotion. Brain Res. 149: 503-507, 1978. ARSHAVSKY, Y. I. AND ORLOVSKY, G. N. Role of the cerebellum in the control of rhythmic movements. In: Neurobiology qf Vertebrate Locomotion, edited by S. Grillner, G. S. Stein, G. Douglas, H. Forssberg, and R. M. Herman. Stockholm: Wenner-Gren International Symposium Series, 1985, vol. 45, p. 677-689. ANDERSSON,

ADJUSTMENTS

DURING

LOCOMOTION

549

V. Y., GURFINKEL, V. S., AND PALTSEV, Y. I. Elements of control of voluntary movements. Bio3zika 12: 135- 14 1, 1967. BIRJUKOVA, E. V., DUFOSS~, M., FROLOV, A. A., IOFFE, M. E., AND MASSION, J. Role of the sensorimotor cortex in postural adjustments accompanying a conditioned paw lift in the standing cat. Exp. Brain Res. 78: 588-596, 1989. BOUISSET, S. AND ZATTARA, M. A sequence of postural movements precedes voluntary movement. Neurosci. Lett. 22: 263-270, 198 1. CAPADAY, C. AND STEIN, R. B. Amplitude modulation of the soleus H-reflex in the human during walking and standing. J. Neurosci. 6: 13081313, 1986. CLEMENT, G., GURFINKEL, V. S., LESTIENNE, F., LIPSHITS, M. I., AND POPOV, K. E. Adaptation of postural control to weightlessness. ZZxp. Brain Res. 57: 6 1-72, 1984. CLEMENT, G. AND LESTIENNE, F. Adaptive modifications of postural attitude in conditions of weightlessness. Exp. Brain Rex 72: 38 l-389, 1988. CONNOLLY, K. J. AND ELLIOTT, J. The evolution and ontogeny of hand function. In: Etiological Studies Of’Child Behavior, edited by C. BlurtonJones. Cambridge, UK: Cambridge Univ. Press, 1962, p. 329-383. CORDO, P. J. AND NASHNER, L. M. Properties of postural adjustments associated with rapid arm movements. J. Neurophysiol. 47: 287-302, 1982. CRENNA, P. AND FRIGO, C. Evidence of phase-dependent nociceptive reflexes during locomotion in man. Exp. Neural. 85: 336-345, 1984. CRENNA, P., FRIGO, C., MASSION, J., AND PEDOTTI, A. Forward and backward axial synergies in man. Exp. Brain Res. 65: 538-548, 1987. DIETZ, V., QUINTERN, J., AND SILLEM, M. Stumbling reactions in man: significance of proprioceptive and pre-programmed mechanisms. J. Physiol. Lond. 386: 149-163, 1987. DREW, T. AND ROSSIGNOL, S. Forelimb responses to cutaneous nerve stimulation during locomotion in intact cats. Brain Res. 329: 323-328, 1985. DUYSENS, J. AND LOEB, G. E. Modulation of ipsi- and contralateral reflex responses in unrestrained walking cats. J. Neurophysiol. 44: 1024- 1037, 1980. DUYSENS, J., TRIPPEL, M., HORSTMANN, G. A., AND DIETZ, V. Gating and reversal of reflexes in ankle muscles during human walking. Exp. Brain Res. 82: 35 l-358, 1990. ELFTMAN, H. AND MANTER, J. Chimpanzee and human feet in bipedal walking. Am. J. Phys. Anthropol. 20: 69-79, 1935. FORSSBERG, H. Stumbling corrective reaction: a phase-dependent compensatory reaction during locomotion. J. Neurophysiol. 42: 936-953, 1979. FORSSBERG, H. Ontogeny of human locomotor control. I. Infant stepping, supported locomotion and transition to independent locomotion. Exp. Brain Res. 57: 480-493, 1985. FORSSBERG, H., GRILLNER, S., AND ROSSIGNOL, S. Phase dependent reflex reversal during walking in chronic spinal cats. Brain Res. 85: 103- 107, 1975. FORSSBERG, H., GRILLNER, S., AND ROSSIGNOL, S. Phasic gain control of reflexes from the dorsum of the paw during spinal locomotion. Brain Res. 132: 121-139, 1977. FORSSBERG, H., HIRSCHFELD, H., AND STOKES, V. P. Development of human locomotor mechanisms. In: Neurobiological Basis of Human Locomotion, edited by M. Shimamura, S. Grillner, and V. R. Edger-ton. Tokyo: Japan Scientific Societies Press, 199 1, p. 259-273. FORSSBERG, H., KINOSHITA, H., ELIASSON, A. C., JOHANSSON, R. S., WESTLING, G., AND GORDON, A. M. Development of human precision grip II: anticipatory control of isometric forces targeted for object’s weight. Exp. Brain Res., 1992. In press. FORSSBERG, H. AND NASHNER, L. M. Ontogenetic development of postural control in man: adaptation to altered support and visual conditions during stance. Neuroscience 2: 545-552, 1982. FUKUDA, T. Studies on human dynamic postures from the viewpoint of postural reflexes. Acta Oto-Laryngol. Stockh. Suppl 16 1: l-52, 196 1. GAH~RY, Y. AND MASSION, J. Co-ordination between posture and movement. Trends Neurosci. 4: 199-202, 198 1. GARRETT, M., IRELAND, A., AND LUCKWILL, R. G. Changes in-excitability of the Hoffman reflex during walking in man (Abstract). J. Physiol. Lond. 23P, 1984. GOSSARD, J. P., CABELGUEN, J. M., AND ROSSIGNOL, S. Phase-dependent modulation of primary afferent depolarization in single cutaneous primary afferents evoked by peripheral stimulation during fictive locomotion in the cat. Brain Res. 537: 14-23, 1990. BELENKII,

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on September 19, 2018. Copyright © 1992 American Physiological Society. All rights reserved.

550

H. HIRSCHFELD

AND H. FORSSBERG

V. S. AND ELNER, A. M. Participation of secondary motor area of the frontal lobe in organization of postural components of voluntary movements in man. Neurophysiology 20: 7-14, 1988. GURFINKEL, V. S. AND LEVIK, Y. S. Sensory complexes and sensorimotor integration. Fiziol. Chel. 5 13: 399-414, 1978. HAAS, G., DIENER, H. C., RAPP, H., AND DICHGANS, J. Development of feedback and feedforward control of upright stance. Dev. Med. Child Neural. 31: 481-488, 1989. HELLEBRANDT, F. A., HOUTZ, S. J., PARTRIDGE, M. J., AND WALTERS, C. E. Tonic neck reflexes in exercises of stressin man. Awt. J. Phys. Med. 35: 144-159, 1956. HIRSCHFELD, H. AND FORSSBERG, H. Phase-dependent modulations of anticipatory postural activity during human locomotion. J. Neurophysiol. 66: 12-19, 1991. LESTIENNE, F. G. AND GURFINKEL, V. S. Posture as an organizational structure based on a dual process: a formal basis to interpret changes of posture in weightlessness. Prog. Brain Res. 76: 307-3 13, 1988a. LESTIENNE, F. G. AND GURFINKEL, V. S. Postural control in weightlessness: a dual process underlying adaptation to an unusual environment. Trends Neurosci. 11: 359-363, 1988b. MASSION, J. Postural changes accompanying voluntary movements. Normal and pathological aspects. Hum. Neurobiol. 2: 26 l-267, 1984. MILLER, S., VAN DER BUG, J., AND VAN DER MECH~, F. G. A. Coordination of movements of the hindlimbs and forelimbs in different forms of locomotion in normal and decerebrate cats. Brain Res. 9 1: 2 17-237, 1975. NASHNER, L. M. AND CORDO, P. J. Relation of automatic postural responses and reaction-time voluntary movements of human leg muscles. Exp. Brain Res. 43: 395-405, 198 1. NASHNER, L. M. AND FORSSBERG, H. Phase-dependent organization of postural adjustments associated with arm movements while walking. J. Neurophysiol. 55: 1382- 1394, 1986. ORL,OVS~Y; G. N. Activity of vestibulospinal neurons during locomotion. Brain Res. 46: 85-98, 1972. GURFINKEL,

Y. I. AND ELNER, A. M. Preparatory and compensatory period during voluntary movement in patients with involvement of the brain of different localization. Biophysics 12: 16 1- 168, 1967. ROSSIGNOL, S., LUND, J. P., AND DREW, T. The role of sensory inputs in regulating patterns of rhythmical movements in higher vertebrates. A comparison between locomotion, respiration and mastication. In: Neural Control of Rhythmic Movements in Vertebrates, edited by A. H. Cohen, S. Rossignol, and S. Grillner. New York: Wiley, 1988, p. 20 l284. RUSSELL, D. F. AND ZAJAC, F. E. Effects of stimulating Deiters’ nucleus and medial longitudinal fasciculus on the timing of the fictive locomotor rhythm induced in cats by DOPA. Brain Res. 177: 588-592, 1979. SAUNDERS, M., INMAN, V., AND EBERHART, H. The major determinants in normal and pathological gait. J. Bone Joint Surg. [Am] 35: 543-558, 1953. SHUMWAY-COOK, A. AND WOOLLACOTT, M. H. The growth of stability: postural control from a developmental perspective. J. Motor Behav. 17: 131-147, 1985. SUTHERLAND, D. H., OLSHEN, R. A., COOPER, L., AND Woo, S. Y. The development of Mature Gait. J. Bone Jt. Surg. Am. 62: 336-353, 1980. THELEN, E., KELSO, J. A. S., AND FOGEL, A. Self-organizing systems and infant motor development. Dev. Rev. 7: 39-65, 1987. THELEN, E. AND WHITLEY-COOKE, D. Relationship between newborn stepping and later walking: a new interpretation. Dev. Med. Child Neural. 29: 380-393, 1987. TRAUB, M. M., ROTHWELL, J. C., AND MARSDEN, C. D. Anticipatory postural reflexes in Parkinson’s disease and other akinetic-rigid syndromes and in cerebellar ataxia. Brain 103: 393-4 12, 1980. WINTER, D. A. Biomechanical motor patterns in normal walking. J. Mot. Behav. 15: 302-330, 1983. WINTER, D. A. AND ROBERTSON, D. G. E. Joint torque and energy patterns in normal gait. Biol. Cybern. 29: 137-142, 1978. PALTSEV,

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (192.236.036.029) on September 19, 2018. Copyright © 1992 American Physiological Society. All rights reserved.