- Email: [email protected]
Development of speed of repetitive movements in children is determined by structural changes in corticospinal efferents
Neuroscience Letters, 144 (1992) 57 60 © 1992 Elsevier Scientific Publishers Ireland Ltd. All rights reserved 0304-3940/92/$ 05.00
57
NSL 08919
Development of speed of repetitive movements in children is determined by structural changes in corticospinal efferents K. M~iller a a n d V. H 6 m b e r g b aDepartment of Pediatrics, bNeurological Therapy Center, Heinrich-Heine-University, Dfisseldorf (FRG) (Received 15 April 1992; Revised version received 9 June 1992; Accepted 10 June 1992) Key words: Child; Repetitive movement; Corticospinal efferent This study was aimed to determine the relationship between the maturation of corticospinal efferents, determined by transcranial stimulation of motor cortex, and the development of fastest repetitive voluntary motor activity in children. The development of fastest repetitive voluntary motor activity was assessed for 3 different types of movements including fastest repetitivetapping movements,aiming movementsand a pegboard transportation task. These 3 motor activitieswere chosen as they were different as to their dependence on detailed sensory guidance. Despite these differences the speed of all 3 movements showed a very similar developmentalprofile, which was matched, however, by the developmental slope of the fastest cortico-motoneuronal efferents. Hence the development of central conduction times determines the speed of repetitive movements in children. In contrast, we could not observe significant effectsof repetitive training on speed of these movements.We show for the first time that the development of fastest voluntary movements is a structure-bound phenomenon, being independent from learning.
In 1892 Bryan was the first to show quantitatively that older children show faster voluntary movements than younger children [2]. Since, numerous papers have addressed this issue of the development of m o t o r speed in children; an exponential course of the developmental profile for fastest voluntary movements has been described for hand tapping movements [1, 6, 16], as well as for foot taps, arm-pronation-supination, and fist opening-closing movements [3]. It has, however, remained a mystery if the observed increase in movement speed with age is determined by a structural change in the central nervous or neuromuscular system or is dependent on training or m o t o r learning [9, 12]. The use of transcranial magnetoelectrical stimulation (TMS) of m o t o r cortex now provides an elegant means to describe non-invasively the maturational pattern of fastest corticospinal efferents. This technique has had already numerous applications in clinical neurology for the assessment of the integrity of the corticospinal system in various disorders such as demyelinating disease [8], stroke [e.g. 11] or extrapyramidal disorders [e.g. 10]. As it can also be applied safely in children, it has the power to describe the development of fastest cortico-motoneuronal efferents also in man. This technique has recently Correspondence." K. Mt~ller, Kinderklinik der Heinrich-Heine-Universitgtt D~isseldorf,Moorenstral3e 5, D-4000 Dtisseldorf I, FRG.
been used to describe the maturational pattern of corticospinal efferents in children [5, 13]. The main finding was that adult values were not reached before the age of 8-10 years [13]. In this study we looked at the covariation of the maturational pattern of fastest cortico-motoneuronal efferents with the developmental profile of fastest voluntary alternating m o t o r activities. For this purpose a variety of m o t o r acts was chosen: as an example of movement with 'open loop' control a repetitive tapping task was used. Also movements dependent on detailed sensory guidance ('closed loop') were applied. Magnetic stimulation was performed in 68 neurologically normal children in the age range from 2-13 years. For TMS of m o t o r cortex a Cadwell MES10 magnetoelectrical stimulator was used, which produces a maxim u m magnetic field of 2.2 T with a bipolar shape of the resulting electrical field. During stimulation of m o t o r cortex the children were lying supine as relaxed as possible in order to avoid latency differences possibly caused by different preinnervational levels due to voluntary m o t o r activity. It has been shown in adults [7] as well as in children [14] that even slight preinnervation m a y cause significant latency changes. Therefore, to obtain an optimal estimate of the developmental profile of cortico-motoneuronal conduction times, the analysis was restricted to the best controllable situation of complete relaxation.
58
ms/m
B00 ms CCT
200~
tsw S
TAP r
r
20-
ms
BOO.
AIM r
PEG r
1500-
ioo-
...,.
I0-
400
i000-
200
50~
,,
5-
'
~o 16o i~o "-~o
0I
do
:(So Bo'~o
~o
16o i~o"~o PEG I
AIM 1
TAP 1
CCT 1 20-
0( 2000- ms
800 ms
ms/m
i
i500.
600
i00. t5
400
iOOO
200-
500
#
t0
50
5 0
go
i6o
i~o"-Ko
o
~o
16o t~o "-~o
~o 16o Bo "-Ko
~o
:6o l~o'~D
ACE (WONTHS) Fig. 1. Maturational profiles for central conduction times (CCT), fastest tapping movements (TAP), fastest repetitive aiming movements (AIM), and pegboard transportation movements (PEG), separated for right (upper panels) and left hand (lower panels) performance. Adult values (AD) are reached at the end of the first decade of life only. Note the almost identical exponential slopes of all 3 types of movement, going in parallel with the maturation of the central conduction time.
The stimulator coil was positioned over the vertex. Latency measurements of muscle responses were obtained at the full 2.2 T intensity. E M G electrodes were positioned bilaterally over the thenar eminence (musculus abductor pollicis). In addition to cortical stimulation, stimulation of peripheral nerve roots was performed by positioning the coil in the midline over the 7th cervical vertebrae. Central conduction times (CCT) were esti-
TABLE I COEFFICIENTS OF THE ADAPTED EXPONENTIAL FUNCTION OF THE MATURAT1ONAL PROFILE OF CCT AND DIFFERENT MOTOR TASKS (TAP, AIM, PEG) FOR THE RIGHT AND LEFT HAND Coefficients of the adapted exponential functions (a.e ,,x), the development of central conduction times (CCT), and movement times of 3 different types of repetitive movements (TAP, AIM, PEG). The slopes of the adapted functions are very similar.
Right hand Left hand
CCT
TAP
-0.0083 -0.0076
-0.0070 -0.0076
AIM 0.0090 0.0088
PEG -0.0073 -0.0081
mated by taking the latency differences between cortical and cervical root stimulation. Various motor activities were recorded in 71 righthanded children in the age range between 2 and 13 years. All have a normal neurological examination and history. Three different kinds of voluntary movements were used on a standardized motor performance series. In the tapping task (TAP) children were instructed to tap as fast as possible with a stylus repetitively on a contact plate (4 x 4 cm size)• The stylus was connected to a digital time keeping device on a microcomputer• The number of repetitive taps was counted over a period of 10 s. The fastest tapping rate was estimated as the average repetition frequency over two runs. These two runs were separated by at least 5 min breaks to avoid effects of peripheral neuromuscular fatigue. For aiming movements (AIM) the children were instructed to move the stylus from a starting position over 20 consecutive targets, separated by 1 cm, to a target zone over a total distance of 32 cm. The movement time per target was calculated in milliseconds by dividing the total time needed to hit all necessary targets in series by the number of finally aimed targets. Data were averaged over two successive runs, separated
59
m8
350, 300.
--
TAP1.1 - - - TAP2,5
[
',
~
\
I -
AIM1.1 - - - A I M 2 ' 5
800,
',,,
250,
900,
.
]
80.
I
--
REGI.1 - - - PEG2.3 I
70.
700, 60.
.
......
"',
600, 50,
200,
500 150,
400
40.
0
0
.4
s
0
5
6
7
8
Age (years)
9
10
5
6
7
8
Age (years)
9
10
5
6
7
8
9
10
Age (years)
Fig. 2. Group means of tapping (TAP),aiming (AIM) and pegboard(PEG) movementsfor all age groups, separated for the first trial on the first day (solid lines)and the final trial on the secondday (broken lines). For the TAP movementsperformancedeclineswith training, for the AIM movements there is no statistically significantdifferenceover training. Only the PEG task shows a consistent improvementwith repetitivetraining.
by breaks of at least 5 min. In the pegboard transportation task (PEG) the children were asked to move small pegs (length 5 cm, diameter 2 mm) from a home matrix, containing 25 pegs in an array of 9 x 9 cm, into a vertical line of target holes (diameter 4.9 mm) with the home matrix positioned 15 cm away from the mid-position of the target line. Time measures were made by an electronic circuit, started by inserting the first peg into the first hole and stopped when the final peg was inserted into the last hole. Pegboard transportation time (in seconds) is the time used to transfer pegs from the matrix into the vertical line of target holes. For comparison stimulation of motor cortex as well as the described motor tasks were also performed in 10 young adults (5 male, 5 female) under the same conditions. The developmental profiles for central conduction times of fastest corticospinal efferents, TAP, AIM and PEG movements were worked out separately for right and left hand. To determine the maturational slopes, exponential functions were fitted to each set of parameters. Statistical comparisons between movement parameters and conduction times were performed by using an analysis of covariance on the B M D P statistical analysis package to test for equality of slopes of the different resulting functions [4]. Fig. 1 depicts the maturational profiles of central conduction times normalized for changes in body dimensions and the developmental profiles for TAP, AIM and P E G movements on the right or left side, respectively. The maturational profiles could best be fitted by exponential functions. There were no significant left/right side differences for the slope of these profiles, although for
this group of right-handed children all movement times tended to be faster on the dominant right than on the left side. There were no asymmetries between the central conduction times on right and left sides. Although the 3 types of movements were differentially dependent on the sensory inputs, the observed maturational profiles did not show significant differences between the 'open loop' TAP and the 'closed loop' AIM and P E G movements, which are critically dependent on visual guidance. Table I summarizes the exponential coefficients of the fitted exponential functions. Slopes are very similar for all of the 3 movement parameters and the central conduction times. An analysis of covariance after logarithmic transformation of the data did not reveal a significant difference between the slopes of all resulting functions (F-1.74, df=7,514, P>0.315). From these data it appears that all 3 types of movements used, irrespective of their relative dependence on detailed sensory guidance, show a similar maturational profile. That makes it likely that the shape of the profile is primarily dependent on a factor common to all of these movements. The identical maturational profiles of fastest corticospinal efferents, as estimated by magnetoelectrical stimulation of motor cortex and peripheral nerve roots, makes it most likely that this common factor in determining speed of movement repetition is the maturation of the fastest corticospinal efferents. As it is not yet clear if magnetic stimulation of motor cortex excites downstream pyramidal tract neurons directly or presynaptically, it cannot be definitely decided if the observed maturational trend indicates the increase
60
of myelination of downstream cortico-motoneuronal fibers or reflects the development of the premotor neuronal circuitry [11]. To assess the influence of repetitive training on the speed of each of the 3 described movements, a group of 47 children (at least 5 in each age group between 5 and 10 years) was instructed to do 5 (for the pegbord task 3) repetitive runs for each movement task on two consecutive days. A possible improvement of movement speed was analyzed by an analysis of variance with replicate measurements over the first trials on day 1 and the final trials on day 2. To avoid fatigue, the replicate movements were separated by intervals of at least 5 rain. Fig. 2 illustrates the impact of repetitive training sessions on movement speed for the 3 movement parameters for either hand. Although there are minute changes with repetitive training of these tasks, comparing the first trial on day 1 and the final trial on day 2, the maturational slopes among them remain fairly identical. A twoway analysis of variance with repeated measurements over training sessions did not show significant benefits from training for the AIM movement (F-2.89: df-l,41: P=0.096). The TAP movements showed even a significant decline of motor speed with training (F=17.02; df=l,41; P<0.001). Only for the PEG task a consistent improvement was found (F=27.0; df= 1,41; P<0.001). From these data it becomes obvious that maturation of fastest corticospinal efferents, indicating a structural change in the central nervous system, has a much more powerful impact than training for the development of movement speed. These data for the first time give quantitative support for the view that the maturation of motor cortex and downstream corticospinal efferents is the main determinator for speed of repetitive movements in children. This means in consequence that all attempts are useless to try to make a child of a given age move faster just by training. This is different for tasks with more complex trajectories (such as the PEG task used here). Children move only as fast as the maturation of corticospinal efferents allows them to do.
I Annett, M., The growth of manual prefercncc and speed, Br. I Psychol., 61 (1970) 545 558. 2 Bryan, L., On the development of voluntary molor ability, Am. J Psychol.,2(1892) 125 171. 3 Denckla, M.B., Development of motor co-ordination in normal children, Dev. Med. Child. Neurol., 16 (1973) 729 741. 4 Dixon, W.J. (Ed.), BMDP Statistical Software Manual, Vol. 2, University of California, Los Angeles, 1988, p. 1121. 5 Eyre, J.A., Miller, S. and Ramesh, V., Constancy of central conduction delays during development in man: investigation of motor and somatosensory pathways, J. Physiol., 434 (1991) 441 551. 6 Goodenough, F.L.. A further study of speed of tapping in early childhood, J. Appl. Psychol., 19 (1935) 309-315. 7 Hess, C.W., Mills, K.R. and Murray, N.M.F., Magnetic stimulation of the human brain: the effects of voluntary muscle activity, J. Physiol., 378 (1986) 37R 8 Hess, C.W. Mills, K.R., Murray, N.M.F. and Schriefer, T.N., Magnetic brain stimulation: central motor conduction studies in multiple sclerosis, Ann. Neurol., 22 (1987) 744 752. 9 Hicks, J.A., The acquisition of motor skill in young children, Child. Dev.,1(1930) 90 103. 10 HOmberg, V. and Lange, H.W., Central motor conduction to hand and leg muscles in Huntington's disease, Movement Dis., 5 (1990) 214 218. 11 HOmberg, V., Stephan, K.M. and Netz, J., Transcranial stimulation of motor cortex in upper motor neuron syndrome: its relation to the motor deficit, Electroencephalogr., Clin. Neurophysiol., 81 (1991) 377 -388. 12 Kerr. R., Movement control and maturation in elementary-grade children, Percept. Motor Skills, 41 (1975) 151-154. 13 Mtiller, K., HOmberg, V. and Lenard, H.-G., Magnetoelectrical stimulation of motor cortex and nerve roots in children. Maturation of cortico-motoneural projections, Electroencephalogr. Clin. Neurophysiol., 81 (1991) 63 70. 14 MOiler, K., HOmberg, V., Aulich, A. and Lenard, H.G., Magnetoelectrical stimulation of motor cortex in children with motor disturbances, Electroencephalogr. Clin. Neurophysiol., 85 (1992) 86-94. 15 Netz, J. and HOmberg, V., Normal responses after magnetic stimulation of motor cortex in patients with severe tetraspasticity after supratentoric lesions. In RA. Anderson, D.J. Hobart and J.V. Danoff (Eds.), Electromyographical Kinesiology, Elsevier, 1991, pp. 397 399. 16 Schellekens, J.M.H., Kalverboer, A.F. and Scholten, C.A., The micro-structure of tapping movements in children, J. Motor Behav., 16 (1984) 20 39.
57
NSL 08919
Development of speed of repetitive movements in children is determined by structural changes in corticospinal efferents K. M~iller a a n d V. H 6 m b e r g b aDepartment of Pediatrics, bNeurological Therapy Center, Heinrich-Heine-University, Dfisseldorf (FRG) (Received 15 April 1992; Revised version received 9 June 1992; Accepted 10 June 1992) Key words: Child; Repetitive movement; Corticospinal efferent This study was aimed to determine the relationship between the maturation of corticospinal efferents, determined by transcranial stimulation of motor cortex, and the development of fastest repetitive voluntary motor activity in children. The development of fastest repetitive voluntary motor activity was assessed for 3 different types of movements including fastest repetitivetapping movements,aiming movementsand a pegboard transportation task. These 3 motor activitieswere chosen as they were different as to their dependence on detailed sensory guidance. Despite these differences the speed of all 3 movements showed a very similar developmentalprofile, which was matched, however, by the developmental slope of the fastest cortico-motoneuronal efferents. Hence the development of central conduction times determines the speed of repetitive movements in children. In contrast, we could not observe significant effectsof repetitive training on speed of these movements.We show for the first time that the development of fastest voluntary movements is a structure-bound phenomenon, being independent from learning.
In 1892 Bryan was the first to show quantitatively that older children show faster voluntary movements than younger children [2]. Since, numerous papers have addressed this issue of the development of m o t o r speed in children; an exponential course of the developmental profile for fastest voluntary movements has been described for hand tapping movements [1, 6, 16], as well as for foot taps, arm-pronation-supination, and fist opening-closing movements [3]. It has, however, remained a mystery if the observed increase in movement speed with age is determined by a structural change in the central nervous or neuromuscular system or is dependent on training or m o t o r learning [9, 12]. The use of transcranial magnetoelectrical stimulation (TMS) of m o t o r cortex now provides an elegant means to describe non-invasively the maturational pattern of fastest corticospinal efferents. This technique has had already numerous applications in clinical neurology for the assessment of the integrity of the corticospinal system in various disorders such as demyelinating disease [8], stroke [e.g. 11] or extrapyramidal disorders [e.g. 10]. As it can also be applied safely in children, it has the power to describe the development of fastest cortico-motoneuronal efferents also in man. This technique has recently Correspondence." K. Mt~ller, Kinderklinik der Heinrich-Heine-Universitgtt D~isseldorf,Moorenstral3e 5, D-4000 Dtisseldorf I, FRG.
been used to describe the maturational pattern of corticospinal efferents in children [5, 13]. The main finding was that adult values were not reached before the age of 8-10 years [13]. In this study we looked at the covariation of the maturational pattern of fastest cortico-motoneuronal efferents with the developmental profile of fastest voluntary alternating m o t o r activities. For this purpose a variety of m o t o r acts was chosen: as an example of movement with 'open loop' control a repetitive tapping task was used. Also movements dependent on detailed sensory guidance ('closed loop') were applied. Magnetic stimulation was performed in 68 neurologically normal children in the age range from 2-13 years. For TMS of m o t o r cortex a Cadwell MES10 magnetoelectrical stimulator was used, which produces a maxim u m magnetic field of 2.2 T with a bipolar shape of the resulting electrical field. During stimulation of m o t o r cortex the children were lying supine as relaxed as possible in order to avoid latency differences possibly caused by different preinnervational levels due to voluntary m o t o r activity. It has been shown in adults [7] as well as in children [14] that even slight preinnervation m a y cause significant latency changes. Therefore, to obtain an optimal estimate of the developmental profile of cortico-motoneuronal conduction times, the analysis was restricted to the best controllable situation of complete relaxation.
58
ms/m
B00 ms CCT
200~
tsw S
TAP r
r
20-
ms
BOO.
AIM r
PEG r
1500-
ioo-
...,.
I0-
400
i000-
200
50~
,,
5-
'
~o 16o i~o "-~o
0I
do
:(So Bo'~o
~o
16o i~o"~o PEG I
AIM 1
TAP 1
CCT 1 20-
0( 2000- ms
800 ms
ms/m
i
i500.
600
i00. t5
400
iOOO
200-
500
#
t0
50
5 0
go
i6o
i~o"-Ko
o
~o
16o t~o "-~o
~o 16o Bo "-Ko
~o
:6o l~o'~D
ACE (WONTHS) Fig. 1. Maturational profiles for central conduction times (CCT), fastest tapping movements (TAP), fastest repetitive aiming movements (AIM), and pegboard transportation movements (PEG), separated for right (upper panels) and left hand (lower panels) performance. Adult values (AD) are reached at the end of the first decade of life only. Note the almost identical exponential slopes of all 3 types of movement, going in parallel with the maturation of the central conduction time.
The stimulator coil was positioned over the vertex. Latency measurements of muscle responses were obtained at the full 2.2 T intensity. E M G electrodes were positioned bilaterally over the thenar eminence (musculus abductor pollicis). In addition to cortical stimulation, stimulation of peripheral nerve roots was performed by positioning the coil in the midline over the 7th cervical vertebrae. Central conduction times (CCT) were esti-
TABLE I COEFFICIENTS OF THE ADAPTED EXPONENTIAL FUNCTION OF THE MATURAT1ONAL PROFILE OF CCT AND DIFFERENT MOTOR TASKS (TAP, AIM, PEG) FOR THE RIGHT AND LEFT HAND Coefficients of the adapted exponential functions (a.e ,,x), the development of central conduction times (CCT), and movement times of 3 different types of repetitive movements (TAP, AIM, PEG). The slopes of the adapted functions are very similar.
Right hand Left hand
CCT
TAP
-0.0083 -0.0076
-0.0070 -0.0076
AIM 0.0090 0.0088
PEG -0.0073 -0.0081
mated by taking the latency differences between cortical and cervical root stimulation. Various motor activities were recorded in 71 righthanded children in the age range between 2 and 13 years. All have a normal neurological examination and history. Three different kinds of voluntary movements were used on a standardized motor performance series. In the tapping task (TAP) children were instructed to tap as fast as possible with a stylus repetitively on a contact plate (4 x 4 cm size)• The stylus was connected to a digital time keeping device on a microcomputer• The number of repetitive taps was counted over a period of 10 s. The fastest tapping rate was estimated as the average repetition frequency over two runs. These two runs were separated by at least 5 min breaks to avoid effects of peripheral neuromuscular fatigue. For aiming movements (AIM) the children were instructed to move the stylus from a starting position over 20 consecutive targets, separated by 1 cm, to a target zone over a total distance of 32 cm. The movement time per target was calculated in milliseconds by dividing the total time needed to hit all necessary targets in series by the number of finally aimed targets. Data were averaged over two successive runs, separated
59
m8
350, 300.
--
TAP1.1 - - - TAP2,5
[
',
~
\
I -
AIM1.1 - - - A I M 2 ' 5
800,
',,,
250,
900,
.
]
80.
I
--
REGI.1 - - - PEG2.3 I
70.
700, 60.
.
......
"',
600, 50,
200,
500 150,
400
40.
0
0
.4
s
0
5
6
7
8
Age (years)
9
10
5
6
7
8
Age (years)
9
10
5
6
7
8
9
10
Age (years)
Fig. 2. Group means of tapping (TAP),aiming (AIM) and pegboard(PEG) movementsfor all age groups, separated for the first trial on the first day (solid lines)and the final trial on the secondday (broken lines). For the TAP movementsperformancedeclineswith training, for the AIM movements there is no statistically significantdifferenceover training. Only the PEG task shows a consistent improvementwith repetitivetraining.
by breaks of at least 5 min. In the pegboard transportation task (PEG) the children were asked to move small pegs (length 5 cm, diameter 2 mm) from a home matrix, containing 25 pegs in an array of 9 x 9 cm, into a vertical line of target holes (diameter 4.9 mm) with the home matrix positioned 15 cm away from the mid-position of the target line. Time measures were made by an electronic circuit, started by inserting the first peg into the first hole and stopped when the final peg was inserted into the last hole. Pegboard transportation time (in seconds) is the time used to transfer pegs from the matrix into the vertical line of target holes. For comparison stimulation of motor cortex as well as the described motor tasks were also performed in 10 young adults (5 male, 5 female) under the same conditions. The developmental profiles for central conduction times of fastest corticospinal efferents, TAP, AIM and PEG movements were worked out separately for right and left hand. To determine the maturational slopes, exponential functions were fitted to each set of parameters. Statistical comparisons between movement parameters and conduction times were performed by using an analysis of covariance on the B M D P statistical analysis package to test for equality of slopes of the different resulting functions [4]. Fig. 1 depicts the maturational profiles of central conduction times normalized for changes in body dimensions and the developmental profiles for TAP, AIM and P E G movements on the right or left side, respectively. The maturational profiles could best be fitted by exponential functions. There were no significant left/right side differences for the slope of these profiles, although for
this group of right-handed children all movement times tended to be faster on the dominant right than on the left side. There were no asymmetries between the central conduction times on right and left sides. Although the 3 types of movements were differentially dependent on the sensory inputs, the observed maturational profiles did not show significant differences between the 'open loop' TAP and the 'closed loop' AIM and P E G movements, which are critically dependent on visual guidance. Table I summarizes the exponential coefficients of the fitted exponential functions. Slopes are very similar for all of the 3 movement parameters and the central conduction times. An analysis of covariance after logarithmic transformation of the data did not reveal a significant difference between the slopes of all resulting functions (F-1.74, df=7,514, P>0.315). From these data it appears that all 3 types of movements used, irrespective of their relative dependence on detailed sensory guidance, show a similar maturational profile. That makes it likely that the shape of the profile is primarily dependent on a factor common to all of these movements. The identical maturational profiles of fastest corticospinal efferents, as estimated by magnetoelectrical stimulation of motor cortex and peripheral nerve roots, makes it most likely that this common factor in determining speed of movement repetition is the maturation of the fastest corticospinal efferents. As it is not yet clear if magnetic stimulation of motor cortex excites downstream pyramidal tract neurons directly or presynaptically, it cannot be definitely decided if the observed maturational trend indicates the increase
60
of myelination of downstream cortico-motoneuronal fibers or reflects the development of the premotor neuronal circuitry [11]. To assess the influence of repetitive training on the speed of each of the 3 described movements, a group of 47 children (at least 5 in each age group between 5 and 10 years) was instructed to do 5 (for the pegbord task 3) repetitive runs for each movement task on two consecutive days. A possible improvement of movement speed was analyzed by an analysis of variance with replicate measurements over the first trials on day 1 and the final trials on day 2. To avoid fatigue, the replicate movements were separated by intervals of at least 5 rain. Fig. 2 illustrates the impact of repetitive training sessions on movement speed for the 3 movement parameters for either hand. Although there are minute changes with repetitive training of these tasks, comparing the first trial on day 1 and the final trial on day 2, the maturational slopes among them remain fairly identical. A twoway analysis of variance with repeated measurements over training sessions did not show significant benefits from training for the AIM movement (F-2.89: df-l,41: P=0.096). The TAP movements showed even a significant decline of motor speed with training (F=17.02; df=l,41; P<0.001). Only for the PEG task a consistent improvement was found (F=27.0; df= 1,41; P<0.001). From these data it becomes obvious that maturation of fastest corticospinal efferents, indicating a structural change in the central nervous system, has a much more powerful impact than training for the development of movement speed. These data for the first time give quantitative support for the view that the maturation of motor cortex and downstream corticospinal efferents is the main determinator for speed of repetitive movements in children. This means in consequence that all attempts are useless to try to make a child of a given age move faster just by training. This is different for tasks with more complex trajectories (such as the PEG task used here). Children move only as fast as the maturation of corticospinal efferents allows them to do.
I Annett, M., The growth of manual prefercncc and speed, Br. I Psychol., 61 (1970) 545 558. 2 Bryan, L., On the development of voluntary molor ability, Am. J Psychol.,2(1892) 125 171. 3 Denckla, M.B., Development of motor co-ordination in normal children, Dev. Med. Child. Neurol., 16 (1973) 729 741. 4 Dixon, W.J. (Ed.), BMDP Statistical Software Manual, Vol. 2, University of California, Los Angeles, 1988, p. 1121. 5 Eyre, J.A., Miller, S. and Ramesh, V., Constancy of central conduction delays during development in man: investigation of motor and somatosensory pathways, J. Physiol., 434 (1991) 441 551. 6 Goodenough, F.L.. A further study of speed of tapping in early childhood, J. Appl. Psychol., 19 (1935) 309-315. 7 Hess, C.W., Mills, K.R. and Murray, N.M.F., Magnetic stimulation of the human brain: the effects of voluntary muscle activity, J. Physiol., 378 (1986) 37R 8 Hess, C.W. Mills, K.R., Murray, N.M.F. and Schriefer, T.N., Magnetic brain stimulation: central motor conduction studies in multiple sclerosis, Ann. Neurol., 22 (1987) 744 752. 9 Hicks, J.A., The acquisition of motor skill in young children, Child. Dev.,1(1930) 90 103. 10 HOmberg, V. and Lange, H.W., Central motor conduction to hand and leg muscles in Huntington's disease, Movement Dis., 5 (1990) 214 218. 11 HOmberg, V., Stephan, K.M. and Netz, J., Transcranial stimulation of motor cortex in upper motor neuron syndrome: its relation to the motor deficit, Electroencephalogr., Clin. Neurophysiol., 81 (1991) 377 -388. 12 Kerr. R., Movement control and maturation in elementary-grade children, Percept. Motor Skills, 41 (1975) 151-154. 13 Mtiller, K., HOmberg, V. and Lenard, H.-G., Magnetoelectrical stimulation of motor cortex and nerve roots in children. Maturation of cortico-motoneural projections, Electroencephalogr. Clin. Neurophysiol., 81 (1991) 63 70. 14 MOiler, K., HOmberg, V., Aulich, A. and Lenard, H.G., Magnetoelectrical stimulation of motor cortex in children with motor disturbances, Electroencephalogr. Clin. Neurophysiol., 85 (1992) 86-94. 15 Netz, J. and HOmberg, V., Normal responses after magnetic stimulation of motor cortex in patients with severe tetraspasticity after supratentoric lesions. In RA. Anderson, D.J. Hobart and J.V. Danoff (Eds.), Electromyographical Kinesiology, Elsevier, 1991, pp. 397 399. 16 Schellekens, J.M.H., Kalverboer, A.F. and Scholten, C.A., The micro-structure of tapping movements in children, J. Motor Behav., 16 (1984) 20 39.