Development of the monosynaptic reflex pathway in the human spinal cord

Developmental Brain Research, 42 (1988) 239-246

239

Elsevier BRD 50786

Development of the monosynaptic reflex pathway in the human spinal cord Susumu Hakamada 1, Fumio Hayakawa 2, Kuniyoshi Kuno 2 and Reisaku Tanaka 3 1Departmentof Pediatrics, Okazaki City Hospital, Aichi (Japan), 2Departmentof Pediatrics, Anjo-Kosei Hospital, Aichi (Japan) and 3Department of Neurobiology, Tokyo Metropolitan Institutefor Neuroscience, Tokyo (Japan) (Accepted 29 March 1988)

Key words: H-reflex; Development; Premature baby; Infant; Triceps surae; Hypothenar reflex

Development of the monosynaptic reflex pathway of the spinal cord was investigated in 96 neurologically normal infants with ages ranging from 25 weeks in postconceptional age (PCA) to 24 months after full-term delivery (PDA) by examining H-reflexes from the triceps surae and hypothenar muscles in terms of their incidence, latency and maximal size in reference to the maximal M-wave. The triceps H-reflex was evoked in all cases, and the latency was longest (26 ms) in the youngest case of 25 weeks (PCA). It gradually shortened until full-term gestation, reaching the shortest value of 17 ms (mean). The H-reflex size initially increased until full-term gestation, reaching the maximum value of 70% and then reducing gradually to the plateau level of about 30% at 12 months (PDA). The hypothenar H-reflex could not be elicited until 32 weeks (PCA). The time course of changes in its latency and size was similar to those of the triceps H-reflex, except that it could not be elicited after 12 months (PDA). Thus, the monosynaptic reflex pathway is already functioning at the age of 25 weeks (PCA) in man. The significance of the systematic change in latency and excitability of the Hreflex with age is discussed. INTRODUCTION The tendon reflex, which is regarded as a representative of the proprioceptive reflex pathway, would provide a very suitable means for investigation of a developmental aspect of the central nervous system in human subjects because of its very simple structure with principally a monosynaptic connection in the spinal cord. It is well known that this tendon reflex can be elicited in various muscles of infants at the perinatal stage. Reports on preterm infants, however, are only fragmentary and ambiguous. For instance, Minkowsky 2° reported a patellar reflex in a fetus of 7 cm (estimated age of 12 weeks gestation), but it was combined with movements of the other leg and both arms. More recently, Schulte et al. 21 described a patellar reflex in the youngest preterm case (25 weeks) which, however, consisted of a burst of unsynchronized E M G with variable latencies. They indicated that the typical response with a single compound E M G activity as usually seen in newborn in-

fants was obtained only in infants beyond 30 weeks gestation. It would be very difficult in very premature babies to discriminate tendon reflexes from exteroceptive reflexes which could also be activated by tendon tap stimuli. It is well known that, in humans and other mammalians, the exteroceptive reflex pathway develops much earlier than the p r o p r i o c e p tive pathway 15,17,24. Several technical limitations of this method, such as obtaining a constant mechanical tap stimulation, the tested limb posture, and the level of subject consciousness, make the systematic investigation of the proprioceptive reflex pathway difficult. Furthermore, the fact that the elicitation of tendon reflexes also depends greatly upon the sensitivity of the muscle spindle system in addition to the efficacy of the Iamotoneuronal synapse, makes the interpretation of obtained results rather complex. The time course of maturation might be subtly different in both factors, and it would be difficult to distinguish them in this type of experiment.

Correspondence: S. Hakamada, Department of Pediatrics, Okazaki City Hospital, 2-2 Wakamiya, Okazaki, Aichi 444, Japan. 0165-3806/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)

24~ The H-reflex is regarded as a proprioceptive reflex in which activation of muscle spindles is substituted by direct electrical stimulation of group la afferents t~ and may thus provide a better tool for analysis of the spinal monosynaptic linkage. Thomas and Lambert 22 were the first to introduce this method to the field of pediatrics. They reported the existence of H-reflex in the hypothenar muscles of preterm babies older than 34 weeks gestation. Many investigators followed suit and extended the test to various muscles 2't°'~2. The results for preterm age, however, have remained fragmentary and qualitative compared with those for perinatal and postdelivery ages. An interesting observation here is that, whereas the triceps H-reflex can be evoked from newborn to aged people '-1~'1~'23, the hypothenar H-reflex was elicited only at the perinatal stage and not beyond the age of one year in normal cases 6'1°'22. From this finding, the revelation of the hypothenar H-reflex in infants of more than one year old has been regarded as pathological 12-~2. Thus, study of the exact time course of the appearance of the hypothenar H-reflex at preterm is to be desired. The present study aimed at investigating the development of H-reflexes systematically in neurologically normal premature and mature infants whose age ranged from 25 weeks gestation to two years after full-term delivery. The typical time course of Hreflex development was demonstrated. SUBJECTS A N D M E T H O D S

The subjects were 96 infants (44 males and 52 females) without pathological involvement in the central and peripheral nervous system. The age of subjects was calculated as the post-conceptional age (PCA) in weeks after the last menstrual period of their mothers or as the post-delivery age ( P D A ) in months after the expected full-term delivery. The distribution of subject age at testing is shown in Table I. The preterm babies had been delivered uneventfully except for their short gestation. Informed consent was obtained from their parents after the aim and procedures of the present investigation had been explained. The test was performed when the infants were fully awake and quiet. A few cases were tested while they were moving or crying, which increased their muscle tone (see Results). The triceps surae and the hypoth-

TABLE 1

Number of subjects and the incidem'e ql tt-relle.v itz ,'ac'h u~,,(: Incidence of the H-reflex in relation to subject age. Numerals in parentheses in the 3rd and 4th columns indicate the number of cases showing H-reflexes against that of tested cascs.

Age

No. of Case5'

Postconceptional (weeks) 25-28 5 29-32 9 33-36 6 37-40 7 41-44 10 Postdelivery (months) 1-3 11 4-6 17 7-9 13 10- I2 11 13-14 7 Total

96

lncidence cff H re/le.~ Triceps ~urae

l(J0f,7 (5/5) 100% (9/9) 100% (4/4) I(t0~4 (6/6) 100% (6/6)

H) pothenar

0';k (0/5) 22% (2*/9) 86% (6/7**) 10(V)~ (6/6) 10()~ (9/9)

100% (11/11) 100% (ll)/10) 100f~ (1 l / l l ) 1110% (7/7) 100% (4/4)

91(.2 110/11) $2c'~ (14/17) 46c"~ (6/13) 36'~ (4/11 ) 0(:i (0/7)

/73/73)

(57/95)

* Cases showing H-reflexes only during movement. '~* onc

case which was tested twice at different days counted doubly (see text).

enar muscle were used for testing. Surface silver-cup electrodes (O 5 mm) were used for both stimulation and recording. Recording electrodes were placed longitudinally about 2 cm apart over the bellies of the tested muscles. Evoked E M G responses were amplified and recorded with a modular electromyograph machine ( N I H O N KOD E N Neuropack MEM-3102; frequency band, 30 Hz to 3 kHz). H-reflex of the triceps surae muscle was evoked by stimulating the tibial nerve at the popliteal fossa. The subjects were held in a prone position. The ulnar nerve was stimulated at the levels of the elbow joint for evoking H-reflexes of the hypothenar muscle. In this case, subjects were held in a supine position. The subjects's face was kept in the midline position during recording to avoid the effect of tonic neck reflex. Rectangular stimuli of 0.2 or 0.5 ms duration were delivered at an interval of 5 s through a constant voltage isolator. Although longer intervals would be more effective in evoking larger H-reflexes, the present interval was chosen as a compromise to make the experiment as short as possible. The stimulus intensity was expressed in multiples of the threshold intensity of the direct motor response M-wave (xMT). Peak-to-peak amplitudes of H-re-

241 flexes were measured, and the maximum H-reflex size (H-max) was expressed as a percentage of the maximum M-wave amplitudes. It is generally accepted that the H-reflex in the triceps surae muscle is derived mainly from the soleus muscle in adult subjects 11't4. Because of the very small size of the subjects, selective recording from the soleus muscle was not attempted. We

do not know whether or not the gastrocnemius muscle contributes significantlyto the triceps H-reflex. Temperature of the testing room was adjusted so as to maintain the subject skin temperature at more than 34 °C, RESULTS

Identification of H-reflex Fig. 1 shows specimen records and a recruitment curve of E M G responses of the triceps surae muscle

Intensity

(A)

of the stimuli(V) 62

.....

5m.~/C),' 5 ~ V

/ |

|

|

.

i

i

=

m

!

!

w

by tibial nerve stimulation from a newborn near-term baby. The first response with about 3 ms latency in Fig. 1A represents the M-wave. The second response was regarded as H-reflex, since it had a lower threshold than that of the M-wave and, while initially increasing its size as the stimulus intensity increases, it was abolished by the development of M-wave (Fig. 1B). These characteristics have been used by previous authors to identify the H-reflex 2'22"23, as is the case in adult subjects 14. Latency of the second response was 17 ms and matched that of a previous study with a corresponding age TM. Although the latency in infants is fairly short compared with that of adult cases, it has been taken a priori to be due to much shorter conduction distance and slower conduction velocity of the responsible nerve in infants. The time course of the latency change with age provided a good support for H-reflex identification in more preterm babies, in whom the threshold of the second response tended to be higher than that of the M-wave (see Discussion for details). By the same token, the second response in the hypothenar muscle of the same subject was also identified as the H-reflex (Fig. 2).

72 AMPLITUDE(mv) 3.0- (B)

80

i

85

!

v

i

!

|

i

i

|

!

|

/I III v'-"

0

; H-reflex

•

: M-wave

,e.,e

2.0-

90

1.0-

95

0

5'0

l

150

1.0xMT

I

Intensity of the stimuli (volt) 5mS/D,

100

IVV

V

Fig. 1. Specimen records of evoked EMGs (A) and H-reflex/Mresponse recruitment curve (B) from the triceps surae muscle of one subject with age of 39 weeks (PCA). Each arrow indicates the initiation of H-reflex and M-response.

242 Intensity of the stimuli(V) (A)

H-REFLEX LATENCY

i r l i i i i i i i i

-Tr,ceps Surae-

5mS/D, 200~V

v~ 049x+382.q

(R- 089)

51 A _

60

"

2O

c

.

"..

o e

\'~o e

i

~

!

i

i

i

r

i

r

i

eUP en que o

5 m S / D , 500/z V

70

-"~-M"~

80

-"v'-3/-~~

8

emmo m lid

°e

?

m • • •

eo

oe

•

•

eo

• •

:i --~u>, tOi 25

v i i i i i I i i i 5 m S / D , 1 mV

i

90

~ - -

Weeks (PCA)

M o n t h s (PDA)

Fig. 3. Changes in H-reflex latency in the triceps surae muscle with increasing age. The linear equation: Y = -0.49 x +38.25 (R = -0.89).

100 120

AMPLITUDE

ent ages. Fig. 3 shows changes in the H-reflex latency in this muscle with age. It distributed b e t w e e n 22 and 26 ms (mean: 25 ms) at 2 5 - 2 8 weeks ( P C A ) . It reduced almost linearly with age, reaching the shortest latency of about 17 ms (mean) just after full-term delivery. A f t e r birth, it t e n d e d to increase but the change was scarcely significant for a year, confirming a previous r e p o r t TM. As shown in Fig. 4, the Hmax of this muscle was 33.6% of the Mmax (ranging between 18% and 55%) at 2 5 - 2 8 weeks ( P C A ) and; gradually increasing over 3 months, reached the m a x i m u m 70% (range, 2 8 % - 9 1 % ) at the term of n o r m a l delivery. It r e d u c e d gradually thereafter, and reached a plateau level of 30% after one year of age.

(mV)

3.0(8)

2.0-

1.0-

H-reflex in hypothenar muscle 0

The incidence of the h y p o t h e n a r H-reflex is also shown in Table I. In contrast to the case of the triceps

50

l

100

| .0xMT I n t e n s i t y of t h e stimuli (volt)

Fig. 2. Specimen records of evoked EMGs (A) and H-reflex/Mresponse recruitment curve (B) from the hypothenar muscle of the same subject as in Fig. 1. Each arrow indicates the initiation of H and M responses.

CHANGES OF H-MAX/M-MAX AS A FUNCTION OF AGE

2E 75c

50. 25"

H-reflex in soleus muscle The incidence of the triceps H-reflex in terms of age is shown in Table I. H-reflex was e v o k e d from all subjects, from the youngest case of 25 weeks ( P C A ) to 24 months ( P D A ) . T h e r e were, however, significant variations in latency and a m p l i t u d e with differ-

--Triceps surae--

% 100-

E I

~

~

3'6 4'0 4~

Weeks (PCA)

]

~

~

1'2 2'4

Months (PDA)

Fig. 4. Changes in size of maximal triceps surae H-reflexes, expressed as a percentage of maximal M-response size with increasing age. Vertical bars indicate 1 S.D.

243 H-REFLEX 25-

LATENCY --Hypothenar -y"-0.61x+39.78 (R--0.84)

one year of age. The latency change along the age scale is shown in Fig. 5. It was about 22 ms at 32 weeks and reduced gradually, reaching the shortest latency of about 15 ms at full-term delivery. Fig. 6 shows Hmax/Mmax from the hypothenar muscle along the age scale. It increased from 32 weeks (PCA) up to normal delivery, reaching the size of about 30%, thereafter gradually decreasing to less than 7.8% at 12 months (PCA) and leveling off at one year of age.

eli

E 20-

c_

_o

.i. i.=:!

"

+m%

15

3'0

25

3'5

4'o

4'5

Weeks ( P C A )

i ~ §i~2:1

DISCUSSION

M o n t h s (PDA)

Fig. 5. Changes in H-reflex latency in the hypothenar muscle with increasing age. The linear equation: Y = -0.61 x +39.78 (R = -0.84).

muscle, it was only after 33 weeks (PCA) that the hypothenar H-reflex could be evoked while the subjects remained awake and quiet. The H-reflex was constantly elicited at ages between 34 weeks (PCA) and 6 months (PDA), but not evoked at all in infants younger than 31 weeks (PDA), even with increased postural tone when moving or crying. It was evoked for the first time in two infants at 32 weeks (PCA), but only while they were moving or crying. One infant of 33 weeks (PCA) did not show any H-reflex even during moving and crying, but did so in the resting state in re-examination at 34 weeks. Another infant of 33 weeks (PCA) showed H-reflex at rest. Therefore, the age of 32-33 weeks seemed to be critical for elicitation of the hypothenar H-reflex. The incidence of the H-reflex reduced gradually after full-term delivery and was seldom elicited after

CHANGES OF H-MAX/M-MAX AS A FUNCTION OF AGE -- Hypothenar

%

"~50" E

~. 40-

"6

30-

.E

20.

.I

~,

~o-

E :I::

O-

/5

2'8

tI+. 3'2

3~

4'0 4i

Weeks (PCA)

~

~

§

1'2

2'4

M o n t h s (PDA)

Fig. 6. Changes in size of maximal hypothenar H-reflexes expressed as a percentage of maximal M-response size with increasing age. Vertical bars indicate 1 S.D.

Onset of H-reflex elicitation The existence of H-reflex indicates that the monosynaptic reflex pathway is already active. The earliest age reported previously for H-reflex elicitation in premature babies was 33 weeks (PCA), for the median nerve stimulation 5. Earlier ages were apparently not tested. The present study indicated a little earlier appearance of the hypothenar muscle at 32 weeks (PCA), and a much earlier appearance of the triceps surae muscle at 25 weeks (PCA). This difference between the two muscles was clearly more than 7 weeks, and its significance will be discussed later. The first appearance of the triceps surae H-reflex must be determined. A histological study on the development of the muscle spindle system in the elbow flexor biceps muscle of human fetuses revealed that, while differentiation of the muscle spindle structure from the extrafusal muscle fiber seems to start at 11 weeks, its typical structure with innervation of thickly myelinated nerve fibers appears to be attained between the 24th and 31st week 7. It was also reported that some of the collaterals of sensory fibers could reach the ventral horn of the cervical cord at the 8th gestational week, but that the development of sensory fiber connection in the lumbar cord is somewhat delayed 24. According to the studies on the spinal cord of the rat fetuses, while primary afferent terminals first reach the ventral horn of the lumbar segments on the 17.5th embryonic day 17, differentiation of the muscle spindle structure in unidentified shank muscles starts on the 19.5th day£9; the monosynaptic reflex activity in the triceps surae muscle also starts on this day 16. If we assume that the spindle system of the triceps surae muscle has a time course of development not much different from that of the biceps brachii muscle, and sequences of nervous and muscular

24,1 developments are qualitatively the same in human and rat fetuses, the first appearance of the triceps surae H-reflex could be at 24 weeks (PCA) or earlier. Further studies are expected to determine the exact term.

Latency Latency distribution of the H-reflex in both muscles matched well the previous results at the corresponding ages 5"6"1°'12"18"22 and supplemented them with continuous sampling, particularly for the preterm age. The most significant feature revealed here is that the latency decreases almost linearly with the increase in the gestational age. This continuity is especially important when determining the onset age of H-reflex appearance. Kudo and Yamada ~7 reported that the latency of ventral root response to the dorsal root stimulation in the rat fetus abruptly shortened at the embryonic stage when the monosynaptic reflex pathway started to function, which means that the previous response with a longer latency is exteroceptive. Thus the linear decrease of latency of the recorded responses supports that they are indeed H-reflexes and not of exteroceptive origin, and that the onset of H-reflex appearance is prior to 25 weeks PCA.

Size of H-reflex The change in the maximal H-reflex size with age showed two phases: a gradual increase from its appearance to the expected full-term delivery and a subsequent decrease over one or two years after normal birth. For the following discussion on the development of the spinal monosynaptic pathway, we presume that the Hma x represents the excitability level of this pathway. Strictly speaking, the Hma x does not necessarily represent the maximum number of motoneurons excited by the maximum group Ia volley as usually interpreted, because a part of the efferent motor impulses may be blocked by the antidromic motor volley when the stimulus intensity employed is stronger than 1 xMT. However, we consider the above presumption to be workable for the following reasons. The threshold difference between la and a - m o t o r fibers and the motoneuron excitability is the major factor for determining Hma x. The threshold of the H-reflex is lower than that of the M-wave in many preterm

infants as shown ill the present and previous studies -~2, indicating that the la threshold is lower than that of a - m o t o r fibers. This is supported by the finding that the conduction velocity of la fibers in preterm and newborn infants, as estimated by the H-reflex method, is faster than that of a - m o t o r fibers by about 11)%5-~-~, as is already established by the more precise neurogram method in adult subjects ~s. These findings indicate that the relationship between la and a-motor fibers in infants is the same as that of adults. The increase of H-reflex size from the preterm period to the time of the full-term delivery is demonstrated in this study for the first time. Such an increase was recently reported in rat fetuses by Kudo and Yamada 17. Several mechanisms responsible for the H-reflex increase in this period involve both increase in sensitivity of motoneurons and efficacy of the afferent inflow. The first possibility is that the number of synaptic boutons in the Ia afferent terminals connecting to motoneurons increases, augmenting the synaptic efficacy in activation of motoneurons. Kudo and Yamada 17 demonstrated this feature histologically in rat fetuses, but also noticed that the increase in magnitude of the monosynaptic reflex response with age appeared to be less marked than that in the number of boutons. Although no compatible morphological data are available for human fetuses, it is likely that this factor is partly responsible. The second possibility is a change in the input resistance of motoneurons with age. Fulton and Walton ~) reported that motoneurons of newborn rats have a higher input resistance than those of more mature cells, and it is well known that EPSP size is somehow larger in motoneurons with higher input resistance 3. Kudo and Yamada 17, however, stated that the input resistance of motoneurons is much higher in rat fetuses. This possibility, therefore, is unlikely to contradict the present findings, but might provide a partial explanation for the problem raised in connection with the first possibility. The third possibility is the development in thickness of the nerve fibers as revealed in the increase of their conduction velocity. This would increase the efficacy of synaptic transmission by more synchronizing Ia inputs to motoneurons that contribute to increase of EPSP size. Increase in the motor conduction velocity would also serve to further synchronize E M G responses.

245 The fourth possibility is the development of the fusimotor control system which may increase the background excitability of motoneurons. The fifth is the development in long descending tracts from the upper structures. These two possibilities remain only hypothetical at present. The decrease of the triceps surae H-reflex after normal delivery was already reported is. The decrease of the hypothenar H-reflex keeps pace with that of the H-reflex incidence (compare Fig. 6 and Table I). Its very high incidence at the neonatal period, and later decrease resulting in virtual extinction after one year confirms the findings of previous. studies 6'12'22. This H-reflex decrease has been attributed to a postnatal development of central mechanisms of inhibition 2'13'22. The detailed mechanisms underlying this presumed inhibition are not known yet. Other mechanisms could also contribute to the H-reflex decrease. A decrease in the input resistance mentioned above, which is supposed to be due to an increase in the size of motoneurons with age 9, would contribute in the perinatal stage. Kudo and Yamada 17 suggested that the development of dendrites would also be important by shifting Ia-synapses more distally and thus making them less effective in exciting the motoneurons. Postnatal elimination of synapses on motoneurons 4 might also encourage the decrease in the reflex response after birth, although the origin and function of these synapses are unknown. Delayed appearance and weaker connection of hypothenar H-reflex One explanation for the delayed appearance of the hypthenar H-reflex is that the development of the muscle spindle system is delayed in this muscle.

REFERENCES 1 Abbruzzese, M., Ratto, S., Abbruzzese, G. and Faval, E., Electroneurographic correlates of the monosynapticreflex: experimental studies and normative data, J. Neurol. Neurosurg. Psychiat., 48 (1985) 434-444. 2 Blom, S., Hagbarth, K.E. and Skoglund, S., Post tetanic potentiation of H-reflexes in human infants, Exp. Neurol., 9 (1964) 198-211. 3 Burke, R.E. and Rudomin, P., Spinal neurons and synapses. In E.R. Kandel (Ed.), Handbook of Physiology, Section I: Nervous System. Vol. I. Cellular Biology of Neurons, American Physiological Society, Bethesda, MD, 1977, pp. 877-944. 4 Conradi, S. and Ronnevi, L.O., Spontaneous elimination

While we await a histological study for the triceps surae and hypothenar muscles, we consider this possibility less likely because the delayed development of more than 7 weeks in the hypothenar spindles seems too long. A better explanation would be a relatively low level of overall excitability of the proprioceptive reflex pathway belonging to this muscle, which is mainly determined by the sensitivity of motoneurons and the efficacy or density of Ia inflows. The similar time course of increase in H-reflex magnitude of the triceps surae and hypothenar muscle supports this idea. The increase in excitability of the monosynaptic reflex pathway towards the full-term gestation should occur systematically in every muscle. Then, this mechanism brings up some motor pools, in which the electrically setup Ia inflow remains subliminal in the more mature state, above the threshold for H-reflex elicitation around the perinatal stage. The delay of H-reflex appearance may represent the time required to bring the reflex excitability to the liminal level. By the same token, the disappearance at more than one year of age can be explained as corresponding to the decrease in excitability below the subliminal level a year after normal delivery. ACKNOWLEDGEMENTS The authors are grateful to Dr. Y. Hirata (Tokyo Metropolitan Institute for Neuroscience) for guiding us to important references concerning histology in human materials and to Dr. N. Kudo (Tsukuba University) for valuable comments on the manuscript. We also thank Ms. K. Nomura for secretarial assistance.

of synapses on cat motoneurons after birth. Do half of the synapses on the cell bodies disappear?, Brain Res., 92 (1975) 505-510. 5 Cruz-Martinez, A., Peraz Conde, M.C. and Ferrer, M.T., Motor conduction velocity and H-reflex in infancy and childhood. 1. Study in newborns, twin and small-for-dates, Electromyogr. Clin. Neurophysiol., 17 (197"7)493-505. 6 Cruz-Martinez, A., Ferrer, M.T., Perez Conde, M.C. and Bernacer, M., Motor conduction velocity and H-reflex in infancy and childhood. II. Intra and extrauterine maturation of the nerve fibers, Electromyogr. Clin. Neurophysiol., 18 (1978) 11-27. 7 Cuajunco, F., Development of the neuromuscular spindle in human fetuses, Contrib. Embryol., (Carnegie Inst. Wash. Publ.), 173 (1940) 95-128.

246 8 Fukushima, Y., Yamashita, N. and Shimada, Y.. Facilitation of H-reflex by homonymous Ia-afferent fibers in man, J. Neurophysiol., 48 (1982) 1079-1088. 9 Fulton, B.P. and Walton, K., Electrophysiological properties of neonatal rat motoneurones studied in vitro, J. Physiol. (Lond.), 370 (1986) 651-678. 10 Gamstorp, I., Normal conduction velocity of ulnar, median and peroneal nerves in infancy, childhood and adolescence, Acta Paediatr., Suppl. 146 (1963) 68-76. 11 Hagbarth, K.E., Post-tetanic potentiation of myotactic reflexes in man, J. Neurol. Neurosurg. Psychiatr., 25 (1962) 1 - 10.

12 Hodes, R. and Gribetz, I., H-reflex in normal human infants: depression of these electrically induced reflexes (EIR's) in sleep, Proc. Soc. Exp. Biol. Med., 110 (1962) 577-580. 13 Hodes, R., Gribetz, I. and Hodes, H.L., Abnormal occurrence of the ulnar nerve-hypothenar muscle H-reflex in Sydenham's chorea, Pediatrics, 30 (1962) 49-56. 14 Hugon, M., Methodology of the Hoffmann reflexes in man. In J.E. Desmedt (Ed.), New Developments in Electromyography and Clinical Neurophysiology, Vol. 3, Karger, Basel, 1973, pp. 277-293. 15 Humphrey, T., Some correlations between the appearance of human fetal reflexes and the development of the nervous system, Progr. Brain Res., 4 (1964) 93-135. 16 Kudo, N. and Yamada, T., Development of the monosynaptic stretch reflex in the rat: an in vitro study, J. Physiol.

(Lond.), 369 (1985) 127-144. 17 Kudo, N. and Yamada, T., Morphological and physiological studies of development of the monosynaptic reflex pathway in the rat lumbar spinal cord. J. Physiot. (Lond.), 389 (1987) 441-46(/. 18 Mayer, R.F. and Mosser, R S . , Excitability ol motoneurons in infants, Neurology, 19 (1969) 932-945. 19 Milburn, A.. The early development of muscle spindles in the rat, J. Cell. Sci., 12 (1973) 175-195. 20 Minkowski, M., Sur les mouvements, les reflexes c t l e s reactions musculaires du foetus humain de 2 /~ 5 mois ct leurs relations avec le syst6me nerveux foetal, Rev. Neurol., 37 (1921) 1105-1118. 21 Schulte, F.J., Linke, I., Michaelis, R. and Nolte, R., Excitation, inhibition, and impulse conduction in spinal motoneurons of preterm, term and small-for-dates newborn infants. In R.J. Robinson (Ed.), Brain and Earl)' Behavior. Academic, London, 1969, pp. 87-114. 22 Thomas, J.E. and Lambert, E.H., Ulnar nerve conduction velocity and H-reflex in infants and children, J. Appl. Physiol., 14 (1960) 1-9. 23 Vecchierini-Blineau, M.F. and Guiheneuc, P., Excitability of the monosynaptic reflex pathway in the child from birth to four years of age, J. Neurol. Neurosurg. Psvchiatr., 44 (1981) 309-314. 24 Windle, W.F. and Fitzgerald, J.E., Development of the spinal reflex mechanism in human embryos, J. Comp. Neurol., 67 (1937) 493-509.