The cutaneous withdrawal reflex in human neonates: sensitization, receptive fields, and the effects of contralateral stimulation

PAIN

02117

The cutaneous withdrawal reflex in human neonates: sensitization, receptive fields, and the effects of contralateral stimulation

The threshold of a cutaneous withdrawal reflex, elicited by calibrated von Frey hairs applied to the Summary foot and leg, has been used to study the development of spinal sensory processing in a group of SO preterm and age (PCA). Data sets (108) were collcctcd on full-term infants ranging from 27.5 to 42.5 weeks postconceptional initial threshold, the effects of repeated innocuous stimuli, the receptive field of the withdrawal reflex, and the effect of a contralateral stimulus. As reported previously (Fitzgerald et al. 1988, 19X9), thcrc was a correlation between PCA and initial threshold. The mean threshold at 29 weeks was 0.237 g (S.E.M. 0.042L whereas the mean threshold at 41 weeks was 0.980 g (S.E.M. 0.134). Kepeated stimulation with von Frey hairs led to a significant lowering of threshold or “sensitization” of the reflex in infants of up to 35 weeks PCA. Thercaftcr, the decrease in threshold was not signiFi~~~nt, and habituation was observed. From 37.5 weeks PCA, it was possible to elicit the withdrawal retlex from the whole limb as far up as the top of the thigh and buttock. Below 30 weeks PCA. the thresholds within this receptive field were uniform, but after 30 weeks a gradient of thresholds was observed increasing progressively from the sole of the foot towards the knee. The application of a maintained stimulus to the contralateral limb significantly inhibited withdrawal reflex responses to ipsilateral von Frey hair stimulation, across all age bands. These results illustrate postnatal changes in sensory processing within the human spinal cord. Low thresholds ovt’r a large receptive field and sensitization on repeated stimulation suggest a lack of some. but not all, inhibit~~ry ~(~nn~~ti~~ns in the preterm neonate since ~(~ntr~~~~~ter~~l inhibiti~~n is well ~st~iblished at birth. Key words:

Neonate:

Withdrawal

reflex;

initial

threshold;

Introduction Cutaneous sensation in neonates is an important arca of study. However, research into the development of somatosensory function in general has been relatively neglected with respect to other areas. such as the auditory. visual and motor systems. Recently, this area of study has assumed greater importance because of the increasing number of preterm neonates surviving, and undergoing traumatic procedures (Fitzgerald 199l; Craig et al. 1993). However, before we can address the specific issue of the

Sensitization;

Receptive

field; Contralateral

stimulation

pain experienced by these infants. it is necessary to have a more fundamental und~rst~~ndil~g of underlying somatoscnsory processes. While it is important to find pain indicators, we still know little about neonatal sensitivity to low intensity mechanical stimuli (Fitzgerald et al. 1988). Fitzgerald and coworkers t 19Sti. I9891 previ~msly reported that the flexion withdrawal reflex of the limb in the neonate is evoked with l~)w-iI~tens~ty mechanical stimuli to the foot and has a threshold that is much lower than that of the nociceptive tlcxion reflex in the adult. It can therefore be evoked by tactile as well as noxious stimuli, and is not a specific nociccptivc reflex as in the adult. Nevertheless, it is a specific response resulting in withdrawal of the limb from the stimulus. which is ~IccoInp~~nied by activity in flexor muscles (Fitzgerald et al. 198X). In order to avoid confusion

Yh

with the adult nociceptive flexor reflex (RI111 (Hugon 1973). we will call the reflex we have observed in the neonate a cutaneous withdrawal reflex. The threshold of the nociceptive flexor reflex in the adult has been demonstrated to be a sensitive measure of spinal cord excitability, and nociceptive activity of the kind generated postoperatively or in intensive care can lower the threshold of the reflex indicating a “hyperalgesic response” (Woolf and Wall 1986; Woolf 1983, 1984). In spite of the fact that in the neonate the withdrawal reflex already has a low threshold, it too appears to be a useful measure of such central excitability. Evidence for this was provided by Fitzgerald and coworkers (19891, who demonstrated a fall in threshold of the reflex in neonates in the presence of injury. Study of the threshold of this withdrawal reflex in neonates can therefore be used firstly as an index of the development of somatosensory function and, secondly, as a measure of spinal cord excitability changes that may result from intensive care. A thorough knowledge of the properties of the withdrawal reflex, including a careful study of its development at different postconceptional ages (PCA), is needed in order to fully realise its value as a test for sensory function in neonates. This research aimed to investigate the threshold and receptive field of the reflex, the effects of repetitive innocuous stimulation, and the effects of a maintained contralateral stimulus on the ipsilateral withdrawal reflex. A further aim was to investigate the development of these properties with age. Sensitization of the reflex has previously been reported to occur with repeated stimulation in infants below 30 weeks PCA, as measured by an increase in the amplitude and number of responses with decreasing PCA (Fitzgerald et al. 1988). Above approximately 32 weeks PCA, habituation rather than sensitization is generally observed (Fitzgerald et al. 1988). The present study aimed to quantify this effect and measure the effects of repeated stimulation on thresholds as well as on the number of responses. In the adult rat the nociceptive flexor reflex has a receptive field which encompasses the area from toes to hindquarter and tail (Woolf and Swett 1984). Here we investigate whether the neonatal withdrawal reflex has an equally large receptive field, whether thresholds are uniform within the field, and whether there are changes with gestational age. Contralateral inhibition is a clear and easily measured example of segmental inhibition of a reflex (Sherrington 1910), and has been observed clearly in neonatal animals (Ekholm 1967; Fitzgerald and Koltzenburg 1986), and also in adult animals, though less consistently (Woolf and Swett 1984). Evidence for this kind of segmental inhibition in the human neonate

lacking and would provide important information on the development of spinal cord inhibitory processing. is

Methods All infants included in the study were born at University College Hospital (UCH), London. Prior to the commencement of the project. permission for the experiments was obtained from the Ethics Committee at UCH. Written parental consent was obtained for each infant, and this was sought no earlier than 72 h after birth in the case of very sick neonates on the neonatal intcnsivc care unit (NICLJ). This waiting period was not deemed necrssary fhr healthy preterm and full-term neonates. The properties of the cutaneous withdrawal reflex in each infant were tested using calibrated von Frey hairs. The strength of hairs followed an arithmetic progression, each hair requiring approximately 1.7 times more weight before it bent. The range of weights applied varied from 0.08 g to 12.6 g. The infants were tested only when they were being disturbed for some other reason, such as an X hourly “all-care” procedure. They were tested in whatever position was most comfortable for them, usually either prone or on their sides. They were either quietly awake or lightly asleep when tested. The entire testing session on each infant took approximately 20 min. Each experiment was performed on 1 leg only. The initial threshold for the withdrawal reflex was determined by applying a von Frey hair perpendicularly on the plantar surface of the foot towards the outer aspect of the heel. A very fine hair was used to begin with, and working upwards in force, the initial threshold was considered to have been reached when a clear. brisk, withdrawal of the foot away from the stimulus was obtained. Repeated innocuous stimuli were applied immediately after ertablishing the initial threshold. In exactly the same place on the foot, the von Frey hair 2 steps above threshold (3.4 times more weight applied) was applied in the same way, al IO-set intervals (see Armstrong-James 1975) 10 times. The response was noted after each stimulation. The threshold following repeated stimuli was then determined starting with the von Frey hair at 1 grade of intensity below that used to obtain the initial threshold. In order to ascertain the area over which the withdrawal reflex could be evoked, 11 points on the lower leg along distal-proximal lines were tested, beginning on the plantar surface of the foot with the grade of von Frey hair at which the initial threshold had heen evoked. An interval of at least IO-15 set was left between each stimulation. Where possible. the thigh and buttock were also stimulated. A contralateral stimulus was applied by gently holding and stroking the contralateral foot, while stimulaling the ipsilateral foot with the grade of von Frey hair used for the initial threshold. The infant was always left for a few minutes following the last experiment before this test was performed. The foot was stimulated 3 times without holding the contralateral foot, then there was a S-mm waiting period before the ipsilateral foot was tested 3 times more. this time while the other foot was being held and stroked.

Results

Data sets (108) for initial threshold, sensitization, reflex area, and the effects of a contralateral stimulus were obtained from 50 infants of both sexes born at 23-43 weeks PCA. Thirty-three of the infants were patients on the NICU, and 17 infants were on the postnatal wards having been born after 37 weeks PCA.

TABLE

1

POSTCONCEPTlONAL

AGE

Band

BANDS

Age range (weeks

gestation)

27.5-30.0 30.1-32.5 3X-35.0 35.1-37s 37.6-40.0 40.1-42s

PC1 PC2 PC3 PC4 PC5 PC6

PC5

PC1

The postnatal age (PNA) of the infants when tested ranged from less than 24 h to 15 weeks. Of the 34 infants on the NICU, 21 of them were either receiving positive pressure ventilation through an endotracheal tube or oxygen therapy during the period of the study. The full-term infants on the postnatal wards were tested onty once, but the preterm infants on the NICU were usually tested 3 or 4 times. Mean values for initial thresholds were obtained according to 6 PCA bands from the whole sample. The PCA bands were divided as shown in Table 1. Initial threshold and age The effects of PCA on initial threshoId have already been extensively examined by Fitzgeraid and coworkers (1988, 1989) so are not reported in detail here. The results from this study were in agreement with those from previous work, in that there was an overall increase in threshold with increasing PCA (confirmed by l-way ANOVA, and 2-sample t test) (see Fig. 1). The largest increase in threshold occurred between PCA bands 4 (X.1-37.5 weeks) and 5 (37.6-40.0 weeks) 1.2

1.0

1

2

3

4

5

6

PC AGE EtAND Figure

I. Postconceptional

without

leg injury. Error

significance

of differences

age versus initial bars denote

S.E.M.

in threshold

11.05. * * * -P

threshold, Stars denote

between

< O.001 ).

in infants levels of

age bands (* -p <

PC6

PC AGE BAND

Figure

7. Effect

denote

S.E.M.

threshold

of repeated Stars denote

before

and after

stimulation

on threshold.

levels of significance repeated

stimulation

( *-p <:0.05,* * * -P

Error

of differences within

bars in

age hands

< 0,OOl).

(P < 0.001). PCA bands l-4 contained all the infants who had received intensive care. whereas a large majority of infants in PCA bands 5 and 6 were on the postnatal wards and had received no intensive care. Babies who had leg injuries were excluded from this analysis. Repeated st~~ulatiorl A total of 104 tests on 49 infants were used for the analysis of the effects of repeated stimulation. The data were divided into the 6 PCA bands. Mean thresholds (with S.E.M.) before and after a period of repeated stimulation (see Methods) were obtained. Sensitization, resulting in significant decreases in threshold following repeated stimulation, occurred in PCA band 1 (P < 0.05, paired r test), and in PCA bands 2 and 3 (P < 0.001, paired t test) (see Fig. 2). There were no significant decreases in threshold f~)ilowing repeated stimulation in PCA bands 4-6, indicating that scnsitization had not occurred at this older age. In I‘act, habituation was observed in the older infants in PCA bands 5 and 6, as denoted by a significant decrease in the numbers of responses to repeated stimulation (I’ < 0.01, Mann-Whitney iJ test). Receptire fields In order to look at the receptive field area of the withdrawal reflex, data were analysed from IO5 tests on 47 infants, divided as before into 6 PCA bands. Usually, only the part of the leg below the knee was tested because the thigh was often covered by a napkin, and in many cases, its removal would have meant a much greater disturbance to the infant. However, in u few instances, it was possible to test up to the top of the thigh and buttock. Mean thresholds (with S.E.M.) were obtained for each of the 11 sites on the lower limb, and were tabulated separately for each age band. There was an increase in threshold progressing up the leg

from toe to knee in all age bands, although this wa:, less noticeable in PCA bands I and 2 (see Fig. 3). The increase was significant in PCA bands 2-h ( I-’ c 0.00 1: age bands 2, 3, 5, and 6; P < 0.05: age band 3; I-way ANOVA, and Z-sample f test) and the gradient of the increase became steeper with increasing PCA until PCA band 5, with larger differences between thresholds on the plantar surface of the foot and those at the knee. There was also a genera1 increase in threshold at all sites with increasing PCA until PCA band 5. which agrees with the results on initial threshold. Only 6 tests on the thigh of 4 infants were performed. There was 1 infant in each of the first 4 PCA bands. Thresholds from the top of the thigh ranged from 0.08 g to 0.6 g, but there was no pattern across the 4 age bands. The withdrawal reflex was tested and elicited from the buttock in 2 infants: the threshold was 0.22 g in both cases. Both subjects were receiving intensive care. Contrulateral stimulation Data were analysed from 45 tests on 32 infants for the effects of contralateral stimulation. Testing for these effects involved 3 ipsilateral control stimuli using the von Frey hair required to evoke initial threshold, followed by a 5-min waiting period, after which 3 more stimuli were given while gently holding and stroking the contralateral foot. Changes in threshold were not measured; only the presence or absence of a withdrawal reflex was recorded. A failure to respond to stimulation was assumed to indicate inhibition. The number of withdrawal reflex responses to stimulation for each test before (control) and during contralateral stimulation were compared. The results were divided into 2 categories: those tests which showed no decrease in the number of withdrawal reflex responses compared to the control, and those tests which did show a decrease compared to the control. Of 45 tests, 32 showed a decrease in response on contralateral stimu-

1

2

3

4

5

6

7

8

9 1011

SITE OF STIMULUS

Figure 3. Effect of increasing PC age and site of stimulus on threshold. Error bars are not inserted as the S.E.M.s are so small. The numbers marked on the abscissa correspond to the sites of stimulation shown on the leg.

m

PC1

Pa

PC4

PC4

PC5

DECREASE

Fc6

PC AGE BAND

Figure 4. Effect of a contralateral stimulus on ipsilateral response stimulation. In PC age band 4. every test showed a decrease ipsilateral response on contralateral stimulation.

to In

lation, and 13 showed no change. Contralateral stimulation led to a significant failure of response to ipsilatera1 von Frey hair stimulation (chi-squared, Y < 0.01) (see Fig. 4). No increase was found in any of the tests. No significant differences as a consequence of contralateral stimulation were found between PCA bands (Fisher’s exact test).

Discussion The results from this study demonstrate that the initial threshold for withdrawal to a tactile stimulus in preterm human neonates is very low in the youngest neonates and increases gradually with increasing PCA. This is consistent with previous findings of Fitzgerald and coworkers (Fitzgerald and Gibson 1984; Fitzgerald et al. 1988). Newborn rats demonstrate similar lowthreshold reflex withdrawal responses following cutaneous stimulation observed by behavioural methods (Fitzgerald et al. 1988), and on direct recording from flexor motor axons (Fitzgerald and Gibson 1984). Other researchers have reported exaggerated reflexes in preterm human infants (Issler and Stephens 1983; Myklebust et al. 1986; Eyre et al. 1989), and kittens (Ekholm 1967). In previous studies on the development of the cutaneous withdrawal reflex from this laboratory (Fitzgerald et al. 1988, 19891, we have called this reflex the “flexor reflex”. This may have caused confusion because the flexor reflex in the adult is generally associated with nociception (Wilier 1983, 1985; Woolf 1984). However, studies using electrical stimulation have shown that the RI1 component of the adult flexor reflex (Hugon 1973) has an A@ fibre low-threshold input (Wilier 1977; Woolf and Swett 1984), but that this input is not physiologically functional in the adult (i.e., not evoked by natural stimulation). It seems likely that this AP input is functional in the neonate, which accounts for

the low threshold of flexion withdrawal. However, this remains to be clearly established, and the need to avoid confusion with the adult nociceptive flexor reflex is the reason why we are now calling this reflex in the neonate the “withdrawal reflex”. The low-threshold withdrawal reflexes seen in the neonate appear to reflect a high level of excitability within the developing spinal cord (Ekholm 1067; Fitzgerald et al. 198X). Physiological evidence of such an increased excitability comes from a study of dorsal horn cells in the neonatal rat (Fitzgerald lY851, showing that single mechanical stimuli can cause long-lasting after-discharges in these cells, the duration and amplitude of which decrease with age. Furthermore, the cutaneous receptive fields of these cells arc larger than those of older animals. It has been proposed that such changes arc due to the late maturation of intcrncuronal pathways (Fitzgerald 1985), including those involved in descending inhibition from the brainstem (Fitzgerald and Koltzenburg 1986). There is evidence in a number of developing sensory systems to suggest that excitatory pathways develop before inhibitory ones. for example in the olfactory bulb (Mair et al. 10X2), and in the hippocampus (Michelson and Lothman 1989). It seems likely that similar changes occur in the human neonate, despite the different developmental time course (Fitzgerald et al. 19X8, 1989). Against this background of developmental changes, WC also observe effects of external stimulation on neonatal thresholds. The fact that there is such a large difference in threshold between those infants in PCA bands l-4 who received intensive care and those in PCA bands 5 and 6. most of whom did not. suggests that intensive care in itself may have an effect on threshold, over and above the developmental effect. This is consistent with the findings of Fitzgerald et al. (1989) who observed long-term lowering of cutaneous thresholds in infants receiving intensive care, in response to repeated heel stabs for obtaining blood samples. One of the mechanisms responsible for this drop in threshold may be a change in the excitability of the spinal cord, which has been shown using the adult flexor reflex (Woolf lYX.3, 1984). Inputs from unmyelinated C-affcrent fibres activated by peripheral injury trigger changes in the response properties of spinal neurons, rendering them more excitable, which in turn brings about a prolonged facilitation in the flexor reflex (Woolf 1992). This may be due in part to the presence of particular neuropcptides in C-afferent fibres, such as substance P and calcitonin gene-related peptidc (CGRP), both of which have been shown to increase the excitability of the flexor reflex for prolonged periods (Woolf and Wiesenfeld-Hallin 1986). The activity of I.-glutamate via the NMDA receptor also has an important role in the production of C fibre-induced changes in central excitability and central components

of hyperalgesia (Woolf and Thompson IYYI), and its general role in the development and plasticity of connections in the immature CNS adds an interesting dimension here (see Garthwaite lY8Y). Similar excitability changes may take place in the immature spinal cord, albeit against a background of increased excitability in the neonate. Despite their delayed functional maturation. C-fibre afferents in the neonatal rat are capable of producing a subthreshold increase in excitability in dorsal horn cells before they evoke postsynaptic spikes. This may cause a lowering of threshold to subsequent stimuli, even from non-noxious sensory inputs (Fitzgerald 1YYI ). The idea of increased excitability on exccssivc stimulation was examined here in human neonates by subjecting them to repeated stimulation. and testing to see if this produces a lowering of withdrawal reflex threshold. Previous studies have indicated such an increase in withdrawal reflex excitability using amplitude and number of limb withdrawals to repeated mechanical stimuli but not measuring threshold (Fitzgerald ct al. 1988, 1089), as was done here. The results show a significant drop in threshold of lo-20% in the more preterm infants below 35 weeks PCA following repeated stimulation. This is not as great as the drop in threshold following repeated stimulation in neonatal rats (about 60% in rats of up to 20 postnatal days, Fitzgerald et al. lY88), but the difference is likely to be due to their differing developmental time courses. The number of withdrawal rctlcx rcsponses to repeated stimulation was also significantly higher in the infants below 37.5 weeks. Both findings indicate that “sensitization” is occurring in prctcrm infants. This “sensitization” is particularly striking since in adults, habituation of reflex responses occurs with rcpeated stimulation at regular intervals (Dimitrijevic and Nathan lY70; Dimitrijevic et al. 1972: Wickclgren 1967a,b). characterized by a decrease in their amplitude and duration. The sensitization changes to habituation at approximately 35 weeks PCA in this study, confirming previous reports (Fitzgerald et al. IYXX). In the older infants (37.6-42.5 weeks PCA) repeated stimulation resulted in a decrease in the number of responses to repeated stimuli (after 2-3 stimuli, no response could be evoked), and no change in threshold following repeated stimulation. Further studies on neonatal laboratory animals arc needed to examine the mechanisms underlying this change from sensitization to habituation of the withdrawal reflex with age. It is unlikely to result from changes in cutaneous receptor properties (Fitzgerald 1985; Payne et al. 1991). Nor is this phenomenon likely to be due to the propcrtics of immature muscle and neuromuscular junction, since this would result in scnsitization being a general feature of all reflexes at this

age: yet the stretch reflex, although highly synchronized does not sensitize (Myklebust et al. 1986). The sensitization is therefore most likely to be due to a property of central circuitry within the spinal cord of neonates. The lack of functional segmental or descending inhibitory pathways has already been discussed (Ekholm 1967; Fitzgerald 1985; Fitzgerald and Koltzenburg 1986; Fitzgerald et al. 1988). Further possible factors are transient synapses and receptor and transmitter expression in the spinal cord during development (Charlton and Helke 1986; Tribollet et al. 199 11, and differing functional properties of immature synapses (Fitzgerald 1985). In this study the receptive field for the cutaneous withdrawal reflex in the human neonate has been shown to be the same as that of the flexor reflex in the adult cat and rat, extending from the toes to the top of the thigh and buttock (Sherrington 1910; Woolf and Swett 1984). There is, however, a developmental change of threshold distribution within the receptive field. Although the field occupies a large area, it is by no means homogeneous. As has been found for the flexor reflex in the adult cat (Sherrington 1910) and rat (Woolf and Swett 19841, there is a “hot spot” of maximal sensivity on the foot, particularly on the distal plantar surface, which constitutes the “central zone” of the receptive field, and a clear decrease in sensitivity progressing up the leg towards the knee. However the changes in threshold are not the same across all age bands. The youngest neonates displayed far more uniformity of threshold within the receptive field than the older ones. This is also accompanied by an overall increase in thresholds within the receptive field as the infants mature. These characteristics are a further example of the relative lack of inhibition in the youngest neonatal spinal cord. All cutaneous inputs to the flexor reflex motoneurone pool are apparently equally effective in the youngest group. This is consistent with widespread excitatory input of spinal reflexes observed in isolated fetal rat spinal cord (Saito 1979). However, in older neonates segmental inhibition tends to decrease the effectiveness of inputs at the edge of the receptive field (Fitzgerald 19851. This indicates maturation of specific inhibitory circuits as well as a generalized reduction in the excitable properties of spinal cord neurons. Whatever central mechanisms underlie the excitability of the neonatal spinal cord, they are not without selectivity since the neonatal withdrawal reflex was clearly inhibited by contralateral stimulation. Contralateral inhibition was shown in the youngest infants tested in this study (27.5 weeks PCA), indicating that, even at this early age, some inhibitory interneuronal connections are fully mature. The similarity of responses across PCA bands in this study, suggests that this function was already well organised in very preterm

infants. Contralateral inhibition involves a segmental mechanism as it is clearly observed in spinal adult rat:, (Fitzgerald 19821, but it may also be amplified by supraspinal mechanisms (Wilier et al. 19841. In the dorsal horn it is effective on responses of wide-dynamic-range but not low-threshold cells (Fitzgerald 1982). It may be a means of both short-term “focussing” of afferent input or of longer-term organization of dorsal horn activity (Fitzgerald 19821. In the adult, contralateral inhibition due to both noxious stimuli involving A6 and C fibres, and non-noxious stimuli involving AP fibres has been reported (see Fitzgerald 19821. The former type of inhibition appears to be more powerful in the adult (Fitzgerald 1982). but this seems not to be the case in the neonate. Functional consequences

The present study highlights how many of the phenomena normally preferentially or specifically evoked by noxious stimuli in the adult can be evoked by non-noxious tactile stimuli in the neonate. Flexion withdrawal in the neonate is evoked by tactile stimulation (rather than pinch as for the flexor reflex in the adult), sensitization can be produced by repeated tactile stimulation (rather than injury), and contralateral inhibition is achieved by tactile stimulation (rather than pinch). The AP connections that are present but apparently ineffective on adult flexor motoneurones (Cook and Woolf 198.5) may well be effective in the neonate, and only gradually become suppressed by inhibition during development. This does not mean that stimuli which are not painful in the adult (i.e., low-threshold stimuli) are painful in the neonate. Rather, it may mean that a pathway which can be excited by a wide range of stimulus intensities in the neonate gradually becomes restricted to high intensity stimuli as the infant matures. Despite this increased excitability (Ekholm 19671, the existence of defined and graded receptive fields, and the presence of contralateral inhibition provide evidence of clear organization in spinal sensory processing in preterm infants. Finally, this withdrawal reflex, although not a direct index of pain, is a useful measure of the state of the somatosensory system in neonates in the clinical setting.

Acknowledgements

We would like to thank the Wellcome Trust for their support, Alan Ainsworth for technical advice and for making the von Frey hairs, and Patrick Wall for reading the manuscript and for helpful comments. We would also like to thank the staff of the Neonatal Unit at University College Hospital, particularly Osmund Reynolds, for their help and cooperation in this study.

101

Lastly, we would like to thank the parents of all the babies who were included in the study for their consent and cooperation.

References Armstrong-James. M.. The functional status and columnar organization of single cells responding to cutaneous stimulation in neonatal rat homatosensory cortex Sl, J. Physiol., 246 (1975) 501-538. Charlton, C.G. and Helke, C.J.. Ontogeny of substance P receptors in rat spinal cord: quantitative changes in receptor number and differential expression in specific loci. Dev. Brain Res., 20 (19X6) x1-91. Cook. A.J. and Woolf, C.J., Cutaneous receptive field and morphological properties of hamstring flexor a-motoneurons in the rat. J. Physiol., 364 (1985) 249-263. Craig. K.D., Whitfield, M.F., Grunau, R.V.E.. Linton, J. and Hadjistavropoulos. lI.D., Pain in the preterm neonate: behavioural and physiological indices, Pain, 52 (1993) 2X7-299. Dimitrijevic, M.R. and Nathan. P.W., Studies of spasticity in man. 4. Changes in flexion reflex with repetitive cutaneous stimulation in spinal man, Brain. 93 (1970) 743-768. Dimitrijevic. M.R., Faganel. J., Gregoric, M., Nathan, P.W. and Trontelj, J.K.. Habituation: effects of regular and stochastic stimulation. J. Neural. Neurosurg. Psychiat., 35 (1972) 234-242. Ekholm, J., Postnatal changes in cutaneous reflexes and in the discharge pattern of cutaneous and articular sense organs, Acta Physiol. Stand. Suppl. 297 (1967) I-130. Eyre. J.A.. Miller, S., O’Sullivan, M.C. and Ramesh, V.. Radiation of the phasic stretch reflex in biceps brachii to muscles of the arm and shoulder in the human neonate, and its restriction with increasing age. J. Physiol., 417 (1989) 104P. Fitzgerald, M., The contralateral input to the dorsal horn of the spinal cord in the decerebrate spinal rat, Brain Res., 236 (19X2) 275-2X7. Fitzgerald. M., The postnatal development of cutaneous afferent fibre input and receptive field organization in the rat dorsal horn. J. Physiol.. 364 (1985) I-IX. Fitzgerald. M.. The developmental neurobiology of pain. In: M.R. Bond. J.E. C’harlton and C.J. Woolf (Eds.), Proc. Vlth World Congress on Pain. Elsevier, Amsterdam. 1991, pp. 253-261. Fitrgerald. M.. The neurobiology of fetal and neonatal pain. In: P.D. Wall and R. Melzack (Eds.). A Textbook of Pain. 3rd edn.. Churchill Livingstone, Edinburgh, in press. Fitzgerald. M. and Gibson. S.. The postnatal physiological and neurochemic‘d development of peripheral sensory C fibres. Neurosci.. I3 (19X4)933-Y44. Fitzgerald. M. and Koltzenburg, M., The functional development of descending inhibitory pathways in the dorsolateral funiculus of the newborn rat spinal cord, Dev. Brain Res., 24 (1986) 261-770. Fitzgerald. M., Shaw, A. and Macintosh, N., The postnatal development of the cutaneous flexor reflex: comparative study of preterm infants and newborn rat pups, Dev. Med. Child Neural.. 30 (19X.8) 520-526. Fitzgerald. M.. Millard. C. and Macintosh, N., Cutaneous hypersensitivity following peripheral tissue damage in newborn infants and its reversal with topical anaesthesia. Pain, 39 (1989) 31-36. Garthwaite, J.. NMDA receptors, neuronal development and neurodegeneration. In: J.C. Watkins and G.L. Collinridge (Eds.), The NMDA receptor. IRL Press, Oxford, 1989, pp. 187-205. Holmqviat. B.. Crossed spinal reflex actions evoked by volleys in somatic afferents, Acta Physiol. Stand.. S2 (Suppl. 1X1) (1061) I-67. Hugon. M.. Exteroceptive reflexes to stimulation of the sural nerve

in normal man. In: J.E. Desmedt (Ed.). New Developmenta in Electromyography and Clinical Neurophy\iology. Vol. 3. Karger. Base1 lY7.1. pp. 713-729. lssler. H. and Stephens. J.A.. The maturation of cutaneous rrllcxes studied in the upper limb in man. J. Phyaiol.. 33.5 (10x3) 643-654. Mair. R.G.. Grllman, R.L. and Gesteland. R.C.. Postnatal prollferation and maturation of olfactory bulb ncurona in the rat. Ncuroscience. 7 (1982) 3105~~3110. Michelson, H.B. and Lothman. E.W.. An In viva clectl-ophysiologlc;ll study of the ontogeny of excitatory and lnhihitory proccsse, In the rat hippocampus, Dev. Brain Res.. 37 (IYXY) I 1% 122. Myklebust. B.M., Gottlieb. G.L. and Argawal. Ci.(‘.. Stretch reflexes of the normal infant. Dev. Med. Child Neural.. 2X ( IYXh) 440~J4Y. Payne. J.. Middleton. J. and Fitzgerald. M.. The pattern and timing of cutaneous hair follicle innervation in the rat pup and human fetus. Dev. Brain Res.. 61 (IYYI) 173.-1511. Saito, K., Development of reflexes in the rat fctuh studied in vitro. J. Physiol. (Land.), 294 (lY7Y) SYI -5Y4. Sherrington. C.S.. Flexion reflex of the limb. crohsrd extension reflex. and reflex stepping and standing. J. Physiol. (Lontl.). 40 (1410) 7x-121. Tribollet, E.. Goumaz. M., Raggenbass. M.. Dubois-Dauphin. M. and Dreifuss. J.J.. Early appearance and transirnl exprehslon ol vaaopressin receptors in the braln of rat fetus and infant. An autoradiographical and electrophysiological study. Dev. Brain Res.. 5X (1991) 13-24. Wickelgren. B.G., Habituation of spinal motoneurons. J. Neurtrphysiol.. 30 (lY67a) 1404-1423. Wickelgren. B.G.. Habituation of spinal interneurons. J. Neurophysiol.. 30 (lYh7b) 1424-1438. Wilier, J.C., Comparative study of perceived pain and nocicrptlvr flexion reflex in man, Pain. 3 ( 1977) hY-X0. Wilier. J.C.. Nociceptive flexion reflexes as a tool for pain research in man. In: J.E. Drsmedt (Ed.), Motor Control Mechanisms in Health and Disease. Raven Press, New York, IYX3. pp. X0%X27. Wilier. J.C.. Studies on pain: effects of morphine on a spinal nociceptivc flexion reflex and related pain sensation in man. Brain Res.. 331 (IYXS) 105-l 14. Wilier. J.C.. Roby. A. and Le Bars. D., Psychophysical and rlectrophysiological approaches to the pain-relieving effect\ of heterotopic nociceptive stimuli. Brain. 107 (lYX4) IOY5- I1 12. Woolf. C.J.. Evidence for a central component of post-injur) pain hypersensitivity, Nature. 306 (19X3) hXh-6XX. Woolf. C.J., Long-term alterations in the excitability ot the flexion reflex produced by peripheral tissue injury in the chronic decerebrate rat, Pain, 18 (I 984) 325-343. Woolf, C.J., Functional plasticity of the flexor withdrawal reflex in the rat following peripheral tissue injury. Adv. Pain Res. Ther., Y (19%) lY3-201. Woolf. C.J., Excitability changes in central neurons following peripheral damage. Role of central sensitization in the pathogencsis ol pain. In: W.D. Willis (Ed.). Hyperalgesia and Allodynia. Raven Press. New York, 1992. pp. 221-243. Woolf. C.J. and Swett. J., The cutaneous contrlhution to the hamstring flexor reflex in the rat. An electrophysiological and anatomical study, Brain Res.. 303 ( lYX4) 290-312. Woolf. C.J. and Thompson. S.W.N.. The induction and maintcnancr of central sensitization is dependant on N-methyl-o-aspartic acid receptor activation: implications for the treatment of post-injury pain hypersensitivity states. Pain. 44 ( IYYl) 2Y3-2YY. Woolf. C.J. and Wall, P.D.. The relative effectivenr!,s of C primary afferent fibres of different origins in evoking a prolonged facilitation of the flexor reflex in the rat, J. Neurosci.. 6 (1Y86) 143% 1443. Woolf, C.J. and Wiesenfeld-Hallin, Z.. Substance P and calcitonin gene-related peptide synergistically modulate the gain of the nociceptive flexor withdrawal reflex in the rat. Neurosci. Lett.. 66 ( IYXh) 226-230.