Embed Size (px)
Citation preview
Europ. J. Obstet. Gynec. reprod. Biol., 1980, 10/3,187-201 0 ElsevierlNorth-Holland Biomedical Press
187
THE EFFECT OF DIAZEPAM ADMINISTRATION DURING PREGNANCY OR LABOR ON THE HEART RATE VARIABILITY OF THE NEWBORN INFANT *
H.P. van GEIJN, H.W. JONGSMA, W.H. DOESBURG, W.A.J.G. LEMMENS, J. de HAAN ** and T.K.A.B. ESKES
Department of Obstetrics and Gynecology and Department for Statistical Consultation, Catholic University, Nijmegen, The Netherlands
Accepted for publication 15 November 1979
Van GEIJN, H.P., JONGSMA, H.W., DOESBURG, W.H., LEMMENS, W.A.J.G., de HAAN, J. and ESKES, T.K.A.B. (1980): The effect of diazepam administration during pregnancy or labor on the heart rate variability of the newborn infant. Europ. J. Obstet. Gynec. reprod. Biol, 1013, 187-201.
Maternal diazepam medication during labor reduces beat-to-beat variability of the fetal heart rate. In this study, the prolongation of the effect was examined in the new- born. The mother received diazepam: (A) during labor as a tranquillizer, (B) daily in low doses at the end of pregnancy, or (C) in high doses i.v. for (pre-)eclampsia. A control group had no analgesic or sedative during pregnancy or labor. The electrocardiogram (ECG) was recorded daily from each newborn during the first 6 days after birth. After preprocessing of the ECG, the median R-R interval, the long-term irregularity index (LTI index) and the interval difference index (ID index) were calculated. A cubic spline interpolation method was applied to compare the various groups with respect to these heart rate parameters. The median R-R interval showed no particular differences in trend for all groups. The LTI index was decreased in the first days after birth in the chronic diazepam (B), and in the diazepam infusion groups (C), but statistical significance was not reached. The ID index was significantly decreased in the first l-2 days in the newborns of the acute and chronic diazepam group. In the diazepam infusion group, an even longer-lasting effect was observed. Maternal diazepam medication affects the beat-to- beat variability in the newborn. The duration of the effect is dependent on dosage and route of administration.
diazepam; pregnancy; labor; heart rate variability; newborn infant
* Supported by Grant 06273, Catholic University, Nijmegen, The Netherlands. **Present address: State University Limburg, Ziekenhuis St. Annadal, Dept. of Obste-
trics and Gynecology, Maastricht, The Netherlands. Reprint requests: Herman P. van Geijn, M.D., Department of Obstetrics and Gynecology, St. Radboudziekenhuis, Geert Grooteplein Zuid 16, Nijmegen, The Netherlands.
188
INTRODUCTION
Diazepam is widely used during pregnancy and labor, either to induce sedation (Hi&r, 1968; Friedman et al., 1969) or for the management of (pre-)eclampsia (Joyce and Kenyon, 1972; Kawathekar et al., 1973). Diaze- pam is a lipid-soluble drug with a small molecular weight, and thus crosses the placenta rapidly. The drug is detectable in the fetal blood 5 min after maternal administration (Erkkola et al., 1973).
In the fetus, diazepam causes a reduction of beat-to-beat heart rate vari- ability (de Haan et al., 1971b; Yeh et al., 1974). This effect of diazepam on fetal heart rate variability is visible on the heart rate tracing within minutes of the administration of the drug to the mother (Fig. 1). The current study examines the prolongation of this effect of diazepam in the newborn infant. The neonatal electrocardiogram (ECG) was recorded daily during the first 6 days after birth. For the analysis of heart rate variability, quantitative indices were applied which discriminated long-term and short-term variabil- ity.
SUBJECTS
Sixty-three newborns were included in the study: 28 were born from mothers who used diazepam during pregnancy or labor; the other 35 were born from mothers who used no analgesics or sedatives during pregnancy or labor.
F.H.R. (b.p.m.)
I’ 10 , T 8 r ' 8 '4 5 ' 7 7
I.D.index
180
60-
LU. P (mm Hg)
187 227 - I , 174 , 175
t - 605
10mg Valium @ i.v.
-600 minutes
S-L : arch. 29745
Fig. 1. The effect of diazepam administered i.v. to the mother on the fetal heart rate variability. F.H.R. = fetal heart rate in h.p.m.; I.U.P. = intra-uterine pressure; I.D. index = interval difference index (see text); M = Montevideo units.
TA
BLE
I
NE
WB
OR
NS
OF
T
HE
D
IAZE
PA
M
GR
OU
PS:
MA
TE
RN
AL
AG
E,
GR
AV
IDIT
Y
(G),
PA
RIT
Y
(P)
AN
D
DIA
ZE
PA
M
ME
DIC
AT
ION
Gro
up
Infa
nt
Mat
ernal
G
p
Tota
l dose
age
per
day
(YF)
(mg)
A
1
f 4 5
6
7
8
9
10
B
1
2
3
4
5
6’
7*
8
9
10
* 11
* 12
13
14
15
C
1
2
3
-
23
1
1
10
(1)
25
1
1
10
(1)
23
1
1
10
(1)
28
4
3
10
(1)
16
1
1
10
(1)
25
1
1
10
(1)
26
1
1
15
(2)
22
1
1
20
(2)
23
1
1
25
(3)
19
1
1
30
(3)
26
5
2
6
(3)
29
1
1
7.5
(2
) 27
4
2
15
(3)
31
2
2
16
(3)
22
1
1
15
(3)
41
4
3
15
(3)
41
4
3
15
(3)
15
2
2
15
(3)
29
1
1
15
(3)
30
2
2
20
(3)
30
2
2
20
(3)
26
2
1
30
(3)
18
2
1
30
(3)
36
2
2
40
(4)
30
2
1
60
(6)
24
1
1
65
26
1
1
170
26
1
1
420
DO
W?
No.
of
(mg/
Wday
) day
s
0.1
3
0.1
2
0.1
7
0.1
7
0.1
2
0.1
3
0.1
8
0.3
6
0.3
4
0.3
7
0.0
9
0.0
8
0.1
8
0.2
0
0.1
9
0.1
9
0.1
9
0.2
7
0.2
6
0.2
2
0.2
2
0.3
3
0.3
5
0.5
1
0.8
0
1.0
3.0
6.0
1
1
1
1
1
1
1
1
1
1
>20
>20
>20
>20
12
220
>20
>20
>20
>20
>20 9
3
>20
>20 1
1
2
Adm
inis
trat
ion
i.v.
i.
m.
i.v.
i.
m.
i.v.
i.
m.
iv.
i.v.
i.
v.
i.v.
p.0
. p.0
. p.0
. P.O
. p-0
. p.0
. p.0
. p.0
. p.0
. p.0
. p.0
. p.
0.
p.0
. p.0
. p.0
.
i.v.
i.
v.
i.v.
Las
t in
ject
ion
(h
befo
re
del
iver
y)
-2
-3
-4
-3
-3
-3
-1
-2
-1
-7
Dia
zepam
le
vel
s m
easu
red
- - +
- +
- +
+
+
+ +
+
+
+
+ +
+
+
- - +
+
- +
+
* T
win
p
reg
na
nd
ies.
D
osa
ge
sch
edu
le
in p
are
nth
eses
: n
o.
of
ad
min
istr
ati
on
s to
tall
y
(gro
up
A
) or
per
day
(g
roup
B).
i.m
. =
in
tra
mu
scu
larl
y;
5
i-v
. =
in
tra
ven
ou
sly
; p
.0.
= o
rall
y.
+ =
ma
tern
al
an
d
neo
nat
al
blood
sam
ple
s w
ere
coll
ecte
d
for
det
erm
inat
ion
of
dia
zepam
an
d
des
met
hyld
iaze
pam
; -
= m
ater
nal
an
d/o
r n
eon
atal
bl
ood
sam
ple
s w
ere
not
coll
ecte
d.
TA
BL
E
II
DE
SC
RIP
TIO
N
OF
TH
E G
RO
UP
S
z
Non
-sed
ated
gr
oups
(n
= 3
5)
Acu
te
Ch
ron
ic
diaz
epam
di
azep
am
Bro
UP
gr
oup
(n=
10
)(A
) (n
= 1
5)
(B)
Dia
zepa
m
infu
sion
gr
oup
(n =
3)
(C)
1 2
3
Mat
ern
al a
ge (
yr)
Gra
vidi
ty
pari
ty
No.
of
twin
s W
eigh
t ga
in i
n p
regn
ancy
(k
g)
Mea
n m
inim
um
an
d m
axim
um
di
asto
lic
BP
(m
m
Hg)
: O
-20
wk
20
-40
wk
V
agin
al d
eliv
ery
(n (
a))
1st
stag
e of
lab
or
(h)
2nd
stag
e of
lab
or
(min
) C
ord
com
plic
atio
ns
(n (
%))
G
esta
tion
al
age
(day
s)
Bir
th w
eigh
t (g
) S
ex (
mal
e : f
emal
e)
Apg
ar s
core
: 1
min
5
min
U
mb.
art
ery:
pH
~02
&Pa)
PC
GZ
(k
Pa)
B
E
Obs
tetr
ic
scor
e (P
rech
tl,
1968
) H
obel
sc
ores
: pr
enat
al
intr
apar
tum
n
eon
atal
Dat
a pr
esen
ted
as m
ean
s *
SD
.
27.3
f
4.3
1.9
f 1.
2 1.
7 f
1.0
0 9.3
+
3.5
6617
7 69
190
35 (
100)
8.
9 f
7.8
32
5 28
17
(48
) 27
9 f
9 35
50
+ 5
10
21:1
4 8.
1 -+
1.
0 8.
9 +
0.
3 7.
23
f 0.
05
2.9
f 1.
1 5.
9 f
1.2
-8.5
+
3.
0 39
.2
f 1.
7 10
*
8 11
-+
10
2
+
1.5
23
f 3.
5 1.
3 +
0.
9 1.
2 f
0.6
0 8.4
f 1.
6
6417
8 65
/92
10 (
100)
13
.3
* 7.
8 37
+
17
4
(40)
28
5 z!
z 8
3580
+
450
8:
2 7.
5 -+
1.
6 8.
7 +
0.
7 7.
19
+
0.06
2.
8 f
1.0
6.1
+
1.4
-8.8
f
1.8
38.9
f
1.8
12
+
4 11
_+
5
3 +
3
29.4
+
6.
3 24
2.
4 +
1.
2 1
1.8
* 0.
7 1
2 -
7.2
+
4.3
9.7
7018
7 73
1104
12
(80
) 9.
2 +
6.
2 20
f
12
l(8)
27
1 +
14
29
35
+
485
7:8
6.8
f 1.
9 8.
5 +
0.
8 7.
25
* 0.
07
2.3
* 0.
6 5.
9 f
1.2
-7.5
f
3.2
38.1
*
1.5
18
f 7
10
-+
4 6
+
5
6018
5 60
190
yes 24
25
n
o 290
3630
5 8 7.23
4.
8 6.
0 -8
.2
37
15
15
16
26 1 1 - -
130/
140
no
- - -
238
1500
m
ale 4 6 7.
09
2.5
5.9
-15.
2 35
20
30
71
26 1 1 - 15.5
6517
5 90
1120
ye
s 8
10
no 25
4 26
00
mal
e 2 8 7.21
0.
7 8.
7 -5
.1
34
16
20
26
191
The 28 newborns in the diazepam group were arranged in the following way (Table I):
(A) An acute diazepam group (n = IO). The mothers of these newborns received diazepam as a sedative i.v. or i.m. during labor. The dose varied from 10 to 30 mg, administered in l-3 injections.
(B) A chronic diazepam group (n = 15). The mothers used diazepam at the end of pregnancy until the day of delivery. The daily oral dose varied from 6 to 60 mg, and amounted to 15-30 mg in 11 of the 15 cases. The diazepam m.edication was maintained for at least 3 days in a row. In a few cases the diazepam medication had been changed or interrupted during the course of pregnancy, but not in the last 20 days before delivery. Considering the long duration of use, steady-state levels of diazepam and its metabolites could be expected in the maternal plasma.
(C) A diazepam infusion group. Three patients received diazepam in high doses i.v. during the last 2 days of pregnancy or during labor for the manage- ment of eclampsia or severe preeclampsia.
The 28 diazepam subjects were collected over a period of 2 yr and com- prised nearly all of the pregnant patients receiving diazepam on our service during this time period. The 35 newborns constituting the non-sedated group were collected over the same time period. These patients were selected before or during labor primarily on practical grounds, such as the availability of the recording equipment during the delivery or the days after birth. If conditions with a known influence on beat-to-beat variability were present (e.g. severe fetal distress, diabetes mellitus or prematurity), the newborn was not included in the study. During labor and again on the first day after delivery, the patients were asked about the use of sedatives, tranquillizers or analgesics during pregnancy or labor. All patients of the non-sedated gro,up denied the use of any of these drugs or similar compounds obtainable without a prescription. Informed consent for the study was obtained in advance from both parents. Table II summarizes the most important obstetrical and neonatal data of the non-sedated and the diazepam groups.
METHODS
The neonatal electrocardiogram (ECG) was recorded for 20-30 min once or twice daily during the first 6 days after birth. The total number of record- ings was 473. The number of recordings per infant varied from 5 to 9. Three surface electrodes were attached precordially to the infant’s skin. During this procedure the infants remained in their cribs in the position in which they were found. The shirt-button was lifted cautiously and the electrodes were applied, after which the shirt button and the blanket were repositioned. In the rare instance that the newborn woke up and began crying, the infant received 10-20 ml of glucose lo%, and the recording was not started until the baby was asleep again. The behavioral state of the infant present during most of the recording time was noted, applying the criteria proposed by Prechtl (1974). Most of the recordings were obtained in the morning
192
between 10 and 12 a.m. (53%), or in the afternoon between 4 and 6 p.m. (22%), at times halfway between two feedings. All heart rate recordings were made in the same room, except for a few infants who were lying in an incubator.
The ECG signal was recorded (Hewlett-Packard ECG, Heart Rate Module, Model 78203 A) and stored on magnetic tape (HP 3960 A instrumentation recorder). After amplification, the analog ECG signal was passed through a band-pass filter (15-60 Hz) and led to a level detector. The R-R interval length was digitized in time units of 0.8 msec (PDP-11 computer). A special artefact detection procedure was applied (see Appendix A). Every 30 set the median R-R interval length (RR), the long-term irregularity index (LTI index) (de Haan et al., 1971a) and the interval difference index (ID index) (Jongsma et al., 1978; van Geijn et al., 1979a) were calculated. The defini- tions of the indices .are given in Appendix B. The values obtained over these 30-see periods were used to calculate medians and semi-interquartile ranges of the heart rate parameters (RR, LTI index, ID index) over the total length of the recording. PDP-11 computers were used for the computations.
STATISTICAL ANALYSIS
Each heart rate recording obtained in the neonatal period was character- ized by the number of 30-see intervals, the medians and semi-interquartile ranges of the heart rate parameters and the time in hours since the moment of birth. Heart rate recordings were obtained at relatively fixed hours of a day, while the moment of birth naturally was spread out over a 24-h day. Thus the time intervals ( tl , t2 . . . ti) between birth and the moment of the recording varied between the infants. The actual values of the heart rate parameters were therefore not suitable to compare between the various groups and to examine factors within the non-sedated group that could influence the heart rate parameters.
An interpolation method (cubic splines, see Appendix C) was applied as an extension of the method of measurement (Wold, 1974). This method enables comparisons between various groups, applying for each child inter- polated values of the heart rate parameters at a set of fixed time points after birth. The data set subjected to statistical analysis (Hotelling test for two samples, see Appendix D) comprised the time points 12, 24,48,72,96 and 120 h after birth. Exceptionally a spline value of a heart rate parameter was obtained by extrapolation. The spline values corresponded very well to the actual values of the heart rate parameters at the same hour after birth.
The number of infants who had at least daily heart rate recordings done between 12 and 120 h after birth is given in Table III, last column. The same table describes the few instances where extrapolation was applied. In these cases the time interval between the actual measurement and the spline value was less than 3 h. The high drop-out of infants from the non-sedated group was due to relatively early discharge of mother and infant from the hospital. The values of the heart rate parameters for the remaining subjects within
193
TABLE III
NUMBER OF NEWBORNS RECORDED AND NUMBER OF NEWBORNS INCLUDED IN THE STATISTICAL ANALYSIS
Group No. of newborns recorded
No. of infants with a complete set of spline values for ti = 12, 24, 48, 72, 96 and120h
Non-sedated 35 ** 22 * Acute diazepam 10 8 Chronic diazepam 15 14 * Diazepam infusion 3 2
* Extrapolated values were used for 3 newborns in the non-sedated group and 1 newborn in the chronic diazepam group. The actual measurements were done in these cases within 3 h before ti = 12 or after ti = 120 h. ** One infant was not included in the statistical analysis because of unexplained extreme high values of the ID index in the first 24 h after birth.
each group were highly representative for all infants in the particular group, i.e. the non-sedated, acute diazepam and chronic diazepam groups respec- tively.
A second analysis examined factors within the subjects of the non-sedated group (n = 22) which could influence the heart rate parameters. For this purpose, multiple correlation coefficients were calculated between the examined variable and the values of the heart rate parameter at ti (12, 18, 24, 36,43,72,96,120 h).
RESULTS
The trends of the RR interval length, the LTI index and the ID index are demonstrated in Figure 2A for the non-sedated group. Figures 2B, C and D show the trends of the heart rate parameters in the 3 diazepam groups and compare these with the results of the non-sedated group. Figures 2A-D are composed from all recorded data in each of the groups. The RR interval length in the non-sedated group was around 500 msec in the first 12 h after birth, increased between 24 and 96 h, and then returned to about 500 msec for the rest of the testing period. The LTI index in this group showed a steep rise in the first 50 h after birth and remained rather constant thereafter. The ID index increased gradually in the non-sedated group during the first 150 h after birth.
The acute diazepam and the non-sedated group showed rather similar trends for the RR interval length and the LTI index over the first 150 h after birth (Fig. 2B). The pattern of the ID index, however, differed signif- icantly between these two groups (Hotelling test: F = 2.586, df = 6.23, P < 0.05). The ID index was initially lower in the acute diazepam group than in the untreated group. At 24 h the groups were similar, and, after this time,
194
the ID index of the acute diazepam infants exceeded that of the unmedi- cated infants.
Similarly, the ID index was the only heart rate parameter which exhibited trends that were significantly different between the chronic diazepam and the non-sedated group (Hotelling test: F = 3.876, df = 6.29, P < 0.01). The ID index was lowered in the chronic diazepam group for about the first 48 h after birth, then was almost equal in both groups between 48 and 96 h, and next was slightly elevated in the chronic diazepam group for the rest of the observation period (Fig. 2C).
R-R interval (msec) pL.T. I. index I. D. index I
700 600
600
500
0
NO SEDATION ( n = 35) A
400
300
200
, , I I 1 I
50 100 150 Hours after birth
R-R interval (m set ) -L.T. I. index I. D. index 1
600
DIAZEPAM MEDICATION DURING LABOR ( n = 10 ) B
50 100 150 Hours after birth
.,...... .a..
~~~~ _ no sedation
R-R interval (msec)----- L.T. I. index I.D. index
400
700
600
500
0
~ 2oo
-0
195
1
0 50 loo 150 0 50 100 150 0 50 100 150 Hours after birth
C CHRONIC DIAZEPAM MECIICATION ( n = 15 1
R R Interval ( m set)- LT.1. index I.D. index
. . . ,.,. ..:.:.. ,.:.:.,.:. ::::: = no sedation
700 600
600
500
0 I I! I IO 0 50 100 150
400
, 300 :”
: : ,, : : . . . . . . .
: .: :“:.’ :.
.. I. :y 2oo Y P / l \.,/
100 i cZ&
! I, 1 IO
50 100 150 0 50 100 150 Hours after birth
DIAZEPAM INFUSION THERAPY (n = 3) D .::: :.: : : : :.: :.>:_:.:_: .:y::: = no sedation
Fig. 2A-D. Means and SEM of the RR interval length, the LTI index and the ID index during the first 150 h after birth in the non-sedated (A), the acute diazepam (B), the chronic diazepam (C) and the diazepam infusion (D) groups. The shaded areas in B-D represent the values of the non-sedated group. The original values of the LTI index and the ID index have been multiplied by a factor of 10 in these figures.
The results from the diazepam infusion group are demonstrated in Figure 2D. The diazepam infusion group showed shorter RR intervals and both a lower LTI index and lower ID index than the non-sedated group over the first 150 h after birth. The data obtained from the diazepam infusion group were not subjected to statistical analysis because of the small number and non-homogeneity of the subjects.
196
Figure 3A shows the concentration curves of diazepam and N-desmethyl- diazepam in infant no. 3 from the diazepam infusion group (see Table II). Figure 3B shows the course of the heart rate parameters for this same infant over the first 18 days after birth. As the concentrations of diazepam and desmethyldiazepam decrease, the values of the ID index increase.
Factors that contributed to a decrease of the ID index in the non-sedated group were a duration of the first stage of labor of more than 5 h, the presence of umbilical cord complications and a Prechtl obstetric score (Prechtl, 1968) of less than 40. In these instances, the ID index was sig- nificantly lowered in the non-sedated group for the first 12 h after birth. Other factors such as maternal age, parity, gestational age, duration of the second stage of labor, values of blood gas analyses of the umbilical artery and sex of the infant did not have a statistically significant influence on the heart rate parameters under study.
Plasma cont. (ng I ml)
PARTUS
+
.
U.C.
i \
o \ *A\, U.C. \
l F”
7 b
“o\O ‘\ q
h, ‘\ 0
‘\
‘L \ D‘s 8
0.f. desmethyl diazepam
b s ‘b‘P
t?j 8Oh
b ‘b\
‘\ 0’
‘\ \
maternal dose of diozepam
us.= umbilical cord blood a.f. = amniotic fluid
A -2 6 12 16 20 24 28 Days before/after birth
197
R-R interval length (msec)
550 $59(
. I
20
i
** ‘* #s
R
+: *
+ *+
*
;* *
l jr* a.
1 * 10 z* * * ST. 1. index
2D
1 l
I.D. index
301 .
B 2 4 6 8 10 12 14 16 18 Days after birth
Fig. 3. A: plasma concentration curves of diaxepam and desmethyldiaxepam of infant no. 3 of the diazepam infusion group. Determination of diazepam and desmethyldiazepam by means of high-pressure liquid chromatography (HPLC), performed in the Department of Clinical Pharmacy, Catholic University, Nijmegen. B: the values of the RR interval length, the LTI index, the ST1 index (de Haan et al., 1971a) and the ID index over the first 18 days after birth for the same infant.
DISCUSSION
The results of the present study confirm that, after maternal diazepam medication during pregnancy or labor, the beat-to-beat heart rate variability
198
is decreased in the newborn infant for the first day(s) after birth. The dura- tion of the effect is dependent on dose and route of administration of diazepam.
A reduction of heart rate variability in the newborn has also been observed in other conditions. A ‘flat’ heart rate pattern has been associated with a bad prognosis (Urbach et al., 1965), respiratory distress syndrome (Kero, 1974) and ‘deep sleep’ (Stemmann et al., 1974). None of the infants of the non-sedated, acute diazepam or chronic diazepam group suffered from the respiratory distress syndrome (RDS). The slightly lower gestational age present in the chronic diazepam group (Table II) can hardly explain the diminished short-term variability present during the first 2 days after birth in this group. One infant (C2) of the diazepam infusion group suffered from severe RDS and died on day 5 after birth. The ID index measured on the 2 previous days were respectively 5.6 and 0.7.
The incidence of state 1 (quiet sleep) observed in the current study was 22% for the non-sedated, 23% for the acute diazepam and 55% for the chronic diazepam group. However, the increased incidence of state 1 in the chronic diazepam group in the days following birth is not the explanation for the diminished short-term variability in this group during the first 2 days after birth. In comparison with active sleep, more short-term variability instead of less is present during state 1 (van Geijn et al., 1979b). Long-term variability is the component of heart rate variability that is decreased during state 1 - not short-term variability.
Besides diazepam, another factor which has to be considered is the condi- tion of the infant at the moment of birth. Although the 1-min Apgar score was lower in the diazepam groups than in the unmedicated group, the condi- tion at birth cannot be the explanation for the diminished short-term vari- ability in the diazepam groups for the first l-2 days. The umbilical cord blood gas values were similar in the acute, chronic diazepam and the non- sedated groups. Several factors were found to be significantly associated with a low ID index in the non-sedated group during the first 12 h after birth. Among these factors, the duration of the first stage of labor might be a con- tributing factor for the acute diazepam infants, but not for the chronic diazepam group.
The period of decreased beat-to-beat variability was followed by a period of increased beat-to-beat variability in both the acute and chronic diazepam groups. The ID index in the acute diazepam group exceeded that in the unmedicated group after approximately 30 h. In the chronic diazepam group, this occurred after about 100 h. A differing distribution of sleep states between the various groups at the time of recording does not seem to be a likely explanation for these differences in heart rate variability. Although polygraphic analysis of sleep state was not carried out, the apparent sleep state during the recording was noted on the basis of observa- tional criteria, and, during the periods in question, in both the control and diazepam groups most of the registrations were made during state 2 (active sleep). A possible explanation for this rebound increase in beat-to-beat heart
199
rate variability is a transient disturbance in the neural regulation of the heart rate, resulting from the withdrawal of either diazepam itself or one of its metabolites. This concept is supported by the finding that the period of increased ID index occurred earlier in the acute diazepam group, which had been exposed to relatively low total loses of diazepam, than in the chronic diazepam group. Although no period of increased ID index was detected in the diazepam infusion group, the duration of observation after birth may have been too short to detect a rebound increase in beat-to-beat variability in these infants.
Several other effects on the newborn have been reported from diazepam administered during pregnancy or labor. These effects are low Apgar scores (Flowers et al., 1969), hypothermia (Owen et al., 1972), apnea, hypo- reflexia and poor sucking (Cree et al., 1973). These phenomena are mostly qualitative, the results of subjective observation and dependent on environ- mental factors. The current study demonstrates quantitatively an effect on the newborn from diazepam administered before birth. Although the reduc- tion of beat-to-beat heart rate variability in itself might not be alarming, it shows objectively that a tranquillizer administered during pregnancy or labor can exert its influence on the newborn for several days.
APPENDIX
A. Artefuct detection The intervals ti+l are accepted for which applies:
ti-0.43 di < tf+l < ti + dj di = t,--300.
The minimum value for cl, is 20 msec. For acceptance of ti+l, a minimum of 3 intervals that qualify according to this formula must be present in succession.
B. The LTI and ID indices are defined over 30-set periods. In the following, ti is the length in msec over the R-R interval (i = 1, . . . , n).
Definition of the long-term irregularity index (LTI index; de Haan et al., 1971a): The LTI index is defined as the interquartile range of:
l/“;F (i=2,...,n)
Definition of the interval difference index (ID index; Jongsma et al., 1978; van Geijn et al., 1979a): The ID index is defined as the interquartile range of:
gi(ti-tj_1) (i = 2, . . . . n) In this formula gi is a weighting factor, depending on the mean interval length.
gi = (t, t8.!j20),05 with< = ti + ti _ 1
2
200
For ?i < 381 msec: gi = 5 C. Cubic splines (Wold, 1974; the method of Reinsch and Greville)
A cubic spline S(t) has been applied under the following conditions:
n = number of recordings per infant ti = the hour of recording after birth Xi = median value of the heart rate parameter for the recording at ti ri = semi-interquartile range of the heart rate parameter for the recording
at ti Pi = number of 30-see intervals in the total recording at ti S( ti ) = adapted spline function value at ti
D. The Hotelling two-sample test is a multivariate generalization of the two- sample Student test. This test is described in Multivariate Statistical Methods, Donald F. Morrison (ed.), McGraw-Hill, New York, Ch. 4.3, pp. 125-129,1967.
ACKNOWLEDGEMENTS
The authors are grateful to Theo Janssen and Machiel van Duuren, medical students at the time of the study, for their assistance in recording the neo- natal heart rate data. In addition the authors wish to thank Professor Chester B. Martin for his review of the manuscript and Mrs. G. Theunissen for secretarial assistance.
REFERENCES
Cree, J.E., Meyer, J. and Hailey, D.M. (1973): Diazepam in labour: its metabolism and effect on the clinical condition and thermogenesis of the newborn. Brit. med. J., 4, 251.
de Haan, J., van Bemmel, J.H., Versteeg, B., Veth, A.F.L., Stolte, L.A.M., Janssens, J. and Eskes, T.K.A.B. (1971a): Quantitative evaluation of fetal heart rate patterns. I. Processing methods. Europ. J. Obstet. Gynec. reprod. Biol., 3, 95.
de Haan, J., van Bemmel, J.H., Stolte, L.A.M., Janssens, J., Eskes, T.K.A.B., Versteeg, B., Veth, A.F.L. and Braaksma, J.T. (1971b): Quantitative evaluation of fetal heart rate patterns. II. The significance of the fixed heart rate during pregnancy and labor. Europ. J. Obstet. Gynec. reprod. Biol., 3, 103.
Erkkola, R., Kangas, L. and Pekkarinen, A. (1973): The transfer of diazepam across the placenta during labour. Acta obstet. gynaec. stand., 52, 167.
Flowers, C.E., Rudolph, A.J. and Desmond, M.M. (1969): Diazepam (Valium) as an adjunct in obstetric analgesia. Obstet. Gynec., 34, 68.
Friedman, E.A., Niswander, K.R. and Sachtleben, M.R. (1969): Effect of diazepam on labor. Obstet. Gxnec., 34, 82.
Huter, J. (1968): Uber Wirksamkeit, Nebenwirkungen, Indikationen und Kontra-indika- tionene von Diazepam (Valium) in der Geburtshilfe und Frauenheilkunde. Klin. Wschr., 46, 681.
Jongsma, H.W., van Geijn, H.P. and de Haan, J. (1978): The analysis of heart rate variabil-
201
ity in the perinatal period. In: Computerdiagnostik in der Geburtsmedizin, pp. 249- 260. Editor: W. Krause. Friedrich Schiller Universitiit, Jena.
Joyce, D.N. and Kenyon, V.G. (1972): The use of diazepam and hydralazine in the treatment of severe preeclampsia. J. Obstet. Gynaec. Brit. Cwlth, 79, 250.
Kawathekar, P., Anusuya, S.R., Sriniwas, P. and Lagali, S. (1973): Diazepam (Calm- pose) in eclampsia: a preliminary report of 16 cases. Curr. ther. Res., 15, 845.
Kero, P. (1974): Heart rate variation in infants with the respiratory distress syndrome. Acta paediatr. stand., Suppl. 250, 1.
Owen, J.R., Irani, S.F. and Blair, A.W. (1972): Effect of diazepam administered to mothers during labour on temperature regulation of the neonate. Arch. Dis. Child., 4 7, 107.
Prechtl, H.F.R. (1968): Neurological findings in newborn infants after pre- and paranatal complications. In: Aspects of Prematurity and Dysmaturity, pp. 303-321. Editors: J.H.P. Jonxis, H.K.A. Visser and J.A. Troelstra. Stenfert Kroese, Leiden.
Prechtl, H.F.R. (1974): The behavioural states of the newborn infant (a review). Brain Res., 76, 185.
Stemmann, E.A., Kosche, F., Icks, U., Albrecht, H., Lemburg, P. and Schmidt, E. (1974): Registrierung der momentanen Herzfrequenz, eine neue Methode zur Uberwachung gefiihrdeter Kinder. Mschr. Kinderheilk., 122, 596.
Urbach, J.R., Phuvchit, B., Zweizig, H., Millican, E., Carrington, E.R., Loveland, M., Williams, J.M., Lambert, R.L., Duncan, A.M., Farrell, S.L., Simons, P.O. and Spur- geon, I.L. (1965): Instantaneous heart rate patterns in newborn infants. Amer. J. Obstet. Gynec., 93, 965.
van Geijn, H.P., Jongsma, H.W., de Haan, J. and Eskes, T.K.A.B. (1979a): Analysis of heart rate and beat-to-beat variability: the interval difference index (ID index). Amer. J. Obstet. Gynec., Submitted for publication.
van Geijn, H.P., Jongsma, H.W., de Haan, J., Eskes, T.K.A.B. and Prechtl, H.F.R. (1979b): Heart rate as an indicator of the behavioral state: studies in the newborn and prospects for fetal heart rate monitoring. Amer. J. Obstet. Gynec., in press.
Wold, S. (1974): Spline functions in data analysis. Technometrics, 16, 1. Yeh, S.Y., Paul, R.H., Cordero, L. and Hon, E.H. (1974): A study of diazepam during
labor. Obstet. Gynec., 43, 363.
