32
Somatosensory Evoked Potentials in Term and Preterm Infants in Relation to Postconceptional Age and Birth Weight ByJ.J. Bongers-Schokking1, E.J. Colon2 , R. A. Hoogland,J. L. Van den Brande 3 and C.J. de Groot4 1Department of Pediatrics, Zuiderziekenhuis, Rotterdam, 2Department of Psychiatrics, Delta Hospital, Rotterdam, 3Department of Pediatrics, University Hospital for Children and Youth "Het Wilhelmina Kinderziekenhuis", Utrecht, 4Department of Pediatrics, Academical Medical Centre, Amsterdam, The Netherlands
The effect of age on the maturation of median nerve somatosensory evoked potentials (SEPs) was studied in 103 normal neonates (24 preterm, 79 term) at the postconceptional age (PCA) of 36-48 weeks. The influence of birth weight was evaluated in 44 term neonates, aged 0-7 days, according to their gestational age (GA) stratified into three groups: A: 38-39 weeks (n = 15); B: 39.5-40.5 we.eks (n = 15); C: 41-43 weeks (n = 14). The mean birth weight was not different in the three groups. For aIl infants the N13latency recorded at cervical (CS2-Fz) level as weIl as the N 19 onset and peak latency at cortical (C3'/C4'-Fz) level we.re measured. For these parameters and for the Nl3latency divided by arm length (N 13/AL) and the N 19 ascending time (NI9AT) the P50, P97 and P3 were calculated as a function of PCA. They were all decreasing in the 3648 weeks period, fast before 40 weeks and slowly thereafter. The SEP values of groups, A, Band Cwere not different, but in each group the wave pattern of the cortical SEPs was more mature in the larger than in the smaller infants. For the groups A, Band C together birth weight (in SDS) correlated inversely with the NI3/AL and the N19AT (both in SDS) (r = 0.73 and 0.52 respectively, p < 0.001). Our data indicate that the progression of maturation of the various SEP components in the period of 36-48 PCA is non-linear. The peripheral and central nervous system of smalI, term infants is less mature than that of large newborns of the same gestational age. The SEP method can sensitively record this gradually increasing maturation in a group of infants with normal birth weight.
Several authors have emphasized the importanc.e of postconceptional age (PCA) in relation to the neurophysiological maturation of the CNS of the newborn as measured with SEPs (7, 9, 10, 11, 32). The few studies on preterm infants (9, 10, 11) show linear regressions with age for the various SEP components from 32 to 40 weeks. Gallai (9) studied the influence of weight on the SEP maturation, but concluded that age and not weight had a significant impact on the ripening of the SEP. Yet we know from recent behavioural studies (23, 30) that a delay in the normal maturation of behavioural states is observed in growth retarded fetuses and neonates. Biochemical (18) and pathological (17) studies demonstrate that a delayed myelination of the CNS can be expected in small-forgestational age (SGA) term infants compared with appropriatefor-gestational age (AGA) term neonates. Brainstem auditory evoked potentials (BAEPs) of malnourished or SGA infants are less mature than those of AGA neonates of the same gestational age (3, 13, 20). However, several reports on the peripheral nerve conduction velocity (pNCV) of the ulnar or median nerve measured from wrist to elbow (6, 22, 25, 31) in SGA and AGA term infants did not demonstrate a difference between the two groups. This study was started to evaluate the usefulness of SEPs as method to measure peripheral and central nerve conduction in the newborn and to investigate the influence of birth weight on the maturation of the peripheral and central pathways.
Methods and subjects Keywords SEP methodology
Neonatal SEPs - Birth weight - Peripheral nerve conduction velocity - Central conduction time
Received March 6, 1989; accepted July 4, 1989 Neuropediatrics 21 (1990) 32-36 © Hippokrates Verlag Stuttgart
Cervical (CS2-Fz) and cortical (C3'/C4'-Fz) SEPs were recorded using a standard procedure described earlier (5). The median nerve was stimulated at the wrist at a frequency of 0.5/s. The strength of the pulses (duration 0.2 ms) was j ust above twitch level of the abductor pollicis brevis museIe. A near nerve temperature of 37-38 °C was ensured by a thermoelement above the arms and a thermosensor fixed in the palm of one of the hands. Recordings were made with a 2-100 Hz filter bandpass (-6dB/oct), appropriate to record both cortical and cervical SEPs in newborns (5). Analysis times were 200 ms for
Downloaded by: Universite Laval. Copyrighted material.
Introduction Abstract
Neonatal SEPs and Birth Weight
N europediatrics 21 (1990)
the cortical and 50 ms for the cervicalleads. For each recording the results of 25-250 stimuli were averaged and every recording was repeated to test reproducibility. Movement induced artefacts were automaticaIly rejected. The infants were either awake or in irregular sleep. Latencies from the beginning of the stimulus to the top of the first negativity in the cervicalleads (N 13), as weIl as to the onset and peak of the first negativity in the cortical leads (N 19) were measured. The N 19 ascending time (N 19AT) was the N 19 onset to peak latency. The N 13/AL was the N 13 latency divided by arm length which was half of the distance from wrist to wrist (caput ulnae) with the infant in ventral supine. position with his arms in 90° abduction. For each parameter the results of the right and left side obtained after separate stimulation were averaged.
33
tion taok place during the first week of life. The GA ranged from 38 to 43 weeks (40.0 ± 1.4 weeks, M ± ISD) and the birth weight from 2700 to 4200 gram (3311 ± 402 gram). The GA was estimated from the last menstrual period of the mother and in 39 infants confirmed by one or more fetal ultrasound investigations during pregnancy. For this group of term infants birth weight did not correlate with GA (r = 0.17, N.S.). According GA three groups were formed: A: 38-39 weeks (n = 15), B: 39.5-40.5 weeks (n= 15), C: 41-43weeks (n= 14). The groups were comparable in birth weight and body and arm length (Table 1), while the head circumference of group C was larger than that of group A. The SEP investigations took place during the first week (median age 1 day) and the frequency distributions of the days at which the investigation was performed were not different for the three studied groups (X2 test; p > 0.10). For statistical analysis of means the Student-t-
Relation with PCA In 103 infants with a PCA of 36-48 weeks SEP investigations were performed in order to obtain reference values for the various SEP parameters with age. The group contained 24 preterm (gestational age [GA] 25.5-36.5 weeks) and 79 term infants (GA 38-43 weeks) with 54 boys and 49 girls. The pregnancy and delivery of these infants were uneventful, the 1 min and 5 min Apgar score 6 or higher. On preterm infant (25.5 weeks) had artificial respiration during the first week of life, but was thereafter in a good clinical condition. The infants had no obvious neurological defects or visible malformations. The highest bilirubin value was below 250 Jlmol/l. Informed consent for the investigation was obtained from the parents. The infants were divided in six age groups of two weeks. Since within these groups the values for the SEP parameters were not normaIly distributed, the median value of each parameter was determined for each group. The 2SD above and below these median values were separately calculated using a mirror technique with the median as reflection point. Through these medians, -2SD and +2SD points curve fitting was performed using at least squares methode In this way an estimation was obtained of the 50th, 97th and 3rd percentile of the N 19 and N 13 latencies, the N 19AT and the N 13/AL for normal infants with a PCA of 36-48 weeks. Relation with birth weight From the group of 103 infants 44 neonates were selected who were born at term and whose SEP investiga-
Results Relation with age As can be seen in Figure 1 the N 19 peak latency, NI9AT, N13latency and N13/AL were aIl decreasing in the period of 36 to 48 weeks PCA, but the decrease was not linear in time. For the N 19 latency the decreasing rate was 3.0 ms/week for the period of 36 to 40 weeks, while it was 0.5 ms/week from 40 to 48 weeks. For the N 19AT these values amounted to 1.6 and 0.4 ms/week respectively, for the N13 latency 0.6 and 0.2 ms/week and for the N 13/AL 3 and 1.4 ms/m/week. The large variance present in the curves of the cortical components in the younger infants decreased with age. Relation with birth weight The values of the various SEP parameters for group A, Band C yielded no significant differences (Table 1). In order to evaluate the relation of birth weight with the SEP parameters for group A, Band C together with minimal interference of the factor age both birth weight as weIl as the SEP parameters were expressed in SDS as a function of the PCA. As reference values for birth weight served the growth charts of Kloosterman (15); for the SEP parameters our own references were used. The obtained correlations for birth weight
Table 1 Clinieal data and mean values (M ± 1 SO) of the SEP investigations of the three term groups. * signifieant differenee with value of group A (p < 0.001). Group (n)
A
Gestat. age (weeks)
Birthweight (gram)
Arm length (em)
Head eire. (em)
N 13 peak lat. (ms)
N13jAL
(msjm)
N 19 onset lat. (ms)
N 19 peak lat. (ms)
N19AT (ms)
±
385 0.5
±
3260 399
±
19.4 1.0
±
339 1.3
±
10.8 0.7
±
55.4 4.0
±
180 1.2
±
29.6 2.9
±
11.4 2.8
±
40.0 0.2
±
3229 378
±
19.3 1.4
±
34.5 1.1
±
10.6 0.9
±
54.2 3.9
±
18.2 0.8
±
30.6 3.9
±
12.4 2.9
(14)
±
41.6 0.7
±
3456 480
±
20.1 1.3
±
35.4* 1.4
±
10.7 0.8
±
53.4 4.2
±
18.0 3.0
±
29.1 5.3
±
11.1 3.8
Total (44)
±
40.0 1.4
±
3311 402
±
19.6 1.3
±
34.6 1.4
±
10.7 0.8
±
54.3 4.0
±
18.1 1.9
±
29.8 3.8
±
11.6 3.1
(15)
B (15)
C
Downloaded by: Universite Laval. Copyrighted material.
test was used.
Neuropediatrics 21 (1990)
].]. Bongers-Schokking et al
~\
I
~ ~
~~38-39
40
~
~
~95-405
14
~ ~ ~41_43
~ 70
(jj
6
r- ~
60
...J
< _
50
Cf)
Z
Cii
6
~ ~
Somatosensory Evoked Potentials in Term and Preterm Infants in Relation to Postconceptional Age and Birth Weight ByJ.J. Bongers-Schokking1, E.J. Colon2 , R. A. Hoogland,J. L. Van den Brande 3 and C.J. de Groot4 1Department of Pediatrics, Zuiderziekenhuis, Rotterdam, 2Department of Psychiatrics, Delta Hospital, Rotterdam, 3Department of Pediatrics, University Hospital for Children and Youth "Het Wilhelmina Kinderziekenhuis", Utrecht, 4Department of Pediatrics, Academical Medical Centre, Amsterdam, The Netherlands
The effect of age on the maturation of median nerve somatosensory evoked potentials (SEPs) was studied in 103 normal neonates (24 preterm, 79 term) at the postconceptional age (PCA) of 36-48 weeks. The influence of birth weight was evaluated in 44 term neonates, aged 0-7 days, according to their gestational age (GA) stratified into three groups: A: 38-39 weeks (n = 15); B: 39.5-40.5 we.eks (n = 15); C: 41-43 weeks (n = 14). The mean birth weight was not different in the three groups. For aIl infants the N13latency recorded at cervical (CS2-Fz) level as weIl as the N 19 onset and peak latency at cortical (C3'/C4'-Fz) level we.re measured. For these parameters and for the Nl3latency divided by arm length (N 13/AL) and the N 19 ascending time (NI9AT) the P50, P97 and P3 were calculated as a function of PCA. They were all decreasing in the 3648 weeks period, fast before 40 weeks and slowly thereafter. The SEP values of groups, A, Band Cwere not different, but in each group the wave pattern of the cortical SEPs was more mature in the larger than in the smaller infants. For the groups A, Band C together birth weight (in SDS) correlated inversely with the NI3/AL and the N19AT (both in SDS) (r = 0.73 and 0.52 respectively, p < 0.001). Our data indicate that the progression of maturation of the various SEP components in the period of 36-48 PCA is non-linear. The peripheral and central nervous system of smalI, term infants is less mature than that of large newborns of the same gestational age. The SEP method can sensitively record this gradually increasing maturation in a group of infants with normal birth weight.
Several authors have emphasized the importanc.e of postconceptional age (PCA) in relation to the neurophysiological maturation of the CNS of the newborn as measured with SEPs (7, 9, 10, 11, 32). The few studies on preterm infants (9, 10, 11) show linear regressions with age for the various SEP components from 32 to 40 weeks. Gallai (9) studied the influence of weight on the SEP maturation, but concluded that age and not weight had a significant impact on the ripening of the SEP. Yet we know from recent behavioural studies (23, 30) that a delay in the normal maturation of behavioural states is observed in growth retarded fetuses and neonates. Biochemical (18) and pathological (17) studies demonstrate that a delayed myelination of the CNS can be expected in small-forgestational age (SGA) term infants compared with appropriatefor-gestational age (AGA) term neonates. Brainstem auditory evoked potentials (BAEPs) of malnourished or SGA infants are less mature than those of AGA neonates of the same gestational age (3, 13, 20). However, several reports on the peripheral nerve conduction velocity (pNCV) of the ulnar or median nerve measured from wrist to elbow (6, 22, 25, 31) in SGA and AGA term infants did not demonstrate a difference between the two groups. This study was started to evaluate the usefulness of SEPs as method to measure peripheral and central nerve conduction in the newborn and to investigate the influence of birth weight on the maturation of the peripheral and central pathways.
Methods and subjects Keywords SEP methodology
Neonatal SEPs - Birth weight - Peripheral nerve conduction velocity - Central conduction time
Received March 6, 1989; accepted July 4, 1989 Neuropediatrics 21 (1990) 32-36 © Hippokrates Verlag Stuttgart
Cervical (CS2-Fz) and cortical (C3'/C4'-Fz) SEPs were recorded using a standard procedure described earlier (5). The median nerve was stimulated at the wrist at a frequency of 0.5/s. The strength of the pulses (duration 0.2 ms) was j ust above twitch level of the abductor pollicis brevis museIe. A near nerve temperature of 37-38 °C was ensured by a thermoelement above the arms and a thermosensor fixed in the palm of one of the hands. Recordings were made with a 2-100 Hz filter bandpass (-6dB/oct), appropriate to record both cortical and cervical SEPs in newborns (5). Analysis times were 200 ms for
Downloaded by: Universite Laval. Copyrighted material.
Introduction Abstract
Neonatal SEPs and Birth Weight
N europediatrics 21 (1990)
the cortical and 50 ms for the cervicalleads. For each recording the results of 25-250 stimuli were averaged and every recording was repeated to test reproducibility. Movement induced artefacts were automaticaIly rejected. The infants were either awake or in irregular sleep. Latencies from the beginning of the stimulus to the top of the first negativity in the cervicalleads (N 13), as weIl as to the onset and peak of the first negativity in the cortical leads (N 19) were measured. The N 19 ascending time (N 19AT) was the N 19 onset to peak latency. The N 13/AL was the N 13 latency divided by arm length which was half of the distance from wrist to wrist (caput ulnae) with the infant in ventral supine. position with his arms in 90° abduction. For each parameter the results of the right and left side obtained after separate stimulation were averaged.
33
tion taok place during the first week of life. The GA ranged from 38 to 43 weeks (40.0 ± 1.4 weeks, M ± ISD) and the birth weight from 2700 to 4200 gram (3311 ± 402 gram). The GA was estimated from the last menstrual period of the mother and in 39 infants confirmed by one or more fetal ultrasound investigations during pregnancy. For this group of term infants birth weight did not correlate with GA (r = 0.17, N.S.). According GA three groups were formed: A: 38-39 weeks (n = 15), B: 39.5-40.5 weeks (n= 15), C: 41-43weeks (n= 14). The groups were comparable in birth weight and body and arm length (Table 1), while the head circumference of group C was larger than that of group A. The SEP investigations took place during the first week (median age 1 day) and the frequency distributions of the days at which the investigation was performed were not different for the three studied groups (X2 test; p > 0.10). For statistical analysis of means the Student-t-
Relation with PCA In 103 infants with a PCA of 36-48 weeks SEP investigations were performed in order to obtain reference values for the various SEP parameters with age. The group contained 24 preterm (gestational age [GA] 25.5-36.5 weeks) and 79 term infants (GA 38-43 weeks) with 54 boys and 49 girls. The pregnancy and delivery of these infants were uneventful, the 1 min and 5 min Apgar score 6 or higher. On preterm infant (25.5 weeks) had artificial respiration during the first week of life, but was thereafter in a good clinical condition. The infants had no obvious neurological defects or visible malformations. The highest bilirubin value was below 250 Jlmol/l. Informed consent for the investigation was obtained from the parents. The infants were divided in six age groups of two weeks. Since within these groups the values for the SEP parameters were not normaIly distributed, the median value of each parameter was determined for each group. The 2SD above and below these median values were separately calculated using a mirror technique with the median as reflection point. Through these medians, -2SD and +2SD points curve fitting was performed using at least squares methode In this way an estimation was obtained of the 50th, 97th and 3rd percentile of the N 19 and N 13 latencies, the N 19AT and the N 13/AL for normal infants with a PCA of 36-48 weeks. Relation with birth weight From the group of 103 infants 44 neonates were selected who were born at term and whose SEP investiga-
Results Relation with age As can be seen in Figure 1 the N 19 peak latency, NI9AT, N13latency and N13/AL were aIl decreasing in the period of 36 to 48 weeks PCA, but the decrease was not linear in time. For the N 19 latency the decreasing rate was 3.0 ms/week for the period of 36 to 40 weeks, while it was 0.5 ms/week from 40 to 48 weeks. For the N 19AT these values amounted to 1.6 and 0.4 ms/week respectively, for the N13 latency 0.6 and 0.2 ms/week and for the N 13/AL 3 and 1.4 ms/m/week. The large variance present in the curves of the cortical components in the younger infants decreased with age. Relation with birth weight The values of the various SEP parameters for group A, Band C yielded no significant differences (Table 1). In order to evaluate the relation of birth weight with the SEP parameters for group A, Band C together with minimal interference of the factor age both birth weight as weIl as the SEP parameters were expressed in SDS as a function of the PCA. As reference values for birth weight served the growth charts of Kloosterman (15); for the SEP parameters our own references were used. The obtained correlations for birth weight
Table 1 Clinieal data and mean values (M ± 1 SO) of the SEP investigations of the three term groups. * signifieant differenee with value of group A (p < 0.001). Group (n)
A
Gestat. age (weeks)
Birthweight (gram)
Arm length (em)
Head eire. (em)
N 13 peak lat. (ms)
N13jAL
(msjm)
N 19 onset lat. (ms)
N 19 peak lat. (ms)
N19AT (ms)
±
385 0.5
±
3260 399
±
19.4 1.0
±
339 1.3
±
10.8 0.7
±
55.4 4.0
±
180 1.2
±
29.6 2.9
±
11.4 2.8
±
40.0 0.2
±
3229 378
±
19.3 1.4
±
34.5 1.1
±
10.6 0.9
±
54.2 3.9
±
18.2 0.8
±
30.6 3.9
±
12.4 2.9
(14)
±
41.6 0.7
±
3456 480
±
20.1 1.3
±
35.4* 1.4
±
10.7 0.8
±
53.4 4.2
±
18.0 3.0
±
29.1 5.3
±
11.1 3.8
Total (44)
±
40.0 1.4
±
3311 402
±
19.6 1.3
±
34.6 1.4
±
10.7 0.8
±
54.3 4.0
±
18.1 1.9
±
29.8 3.8
±
11.6 3.1
(15)
B (15)
C
Downloaded by: Universite Laval. Copyrighted material.
test was used.
Neuropediatrics 21 (1990)
].]. Bongers-Schokking et al
~\
I
~ ~
~~38-39
40
~
~
~95-405
14
~ ~ ~41_43
~ 70
(jj
6
r- ~
60
...J
< _
50
Cf)
Z
Cii
6
~ ~
