- Email: [email protected]
Numeric analysis of heart rate variation in intrauterine growth-retarded fetuses: A longitudinal study
Numeric analysis of heart rate variation in intrauterine growth -retarded fetuses: A longitudinal study Rosalinde J.M. Snijders, MSc, Lucia S.M. Ribbert, MD, Gerard H.A. Visser, MD, PhD, and Eduard J.H. Mulder, MSc, PhD Groningen, The Netherlands OBJECTIVE: We attempted to determine changes occurring with time in fetal heart rate and its variation in fetuses with intrauterine growth retardation in whom late antepartum fetal heart rate decelerations eventually develop. STUDY DESIGN: Thirteen fetuses with inlrauterine growth retardation were studied over a median period of 25 days. One-hour fetal heart rate records were made two to five times per week and were analyzed numencally. Fetal movements were recorded by the women. RESULTS: On average long-term fetal heart rate variation decreased gradually with time and fell below the norm (30 milliseconds) at about the same time decelerations appeared. Mean heart rate showed a slight but statistically significant increase after the occurrence of decelerations. There were large interfetal differences in all parameters studied. CONCLUSION: In fetuses with intrauterine growth retardation a decrease in long-term fetal heart rate variation is a rather late sign of impairment that coincides with the occurrence of late decelerations. In the surveillance of the fetus with intrauterine growth retardation it might be most appropriate to use each fetus as its own control. (AM J OBSTET GVNECOL 1992;166:22-7.)
Key words: Intrauterine growth retardation, fetal heart rate monitoring, fetal hypoxemia, computer analysis In cross-sectional studies of fetuses with intrauterine growth retardation (IUGR), long-term heart rate variation is usually within the normal range in the absence of antepartum late fetal heart rate (FHR) decelerations, but it is reduced in the presence of late decelerations. 1. 2 Reduced FHR variation and the occurrence of decelerations coincide with hypoxemia, as determined in cord blood obtained by cordocentesis or at elective cesarean section,2°' and are assumed to precede the "terminal" state, in which fetal acidemia is present. 5 . 6 These data suggest that a reduction in heart rate variation is a rather late sign of impairment of the fetal condition. The cross-sectional design of these studies has not allowed determination of the actual sequence of changes in FHR patterns with progressive deterioration . In this study therefore heart rate and its variation were measured longitudinally in fetuses with IUGR to determine changes occurring with time and progressive deterioration.
From the Department of Obstetrics and Cynaecology, Unzverszty Hospztal Gronmgen. Supported by a grant from the Dutch Pnnses Beatnx Fonds (grant No . 87-2684). R eceIved f01' publication November 8, 1990; revtsed May 1-1. 1991; accepted May 28, 1991 . R epnnt r equests: Prof Dr. C .H .A . Visser, Department of Dbstetnes and Gynaecology, UnIVersity Hospital, P O . . Box 85500,3508 CA Utrecht, The Netherlands. 611 131262
22
Patients and methods The study group consisted of 13 women admitted to the obstetric ward. They were selected retrospectively on these criteria: (1) Computerized analyses of I-hour FHR records were available before and after the occurrence of antepartum late decelerations; (2) cesarean section had been performed because of recurrent late decelerations; (3) the infants had a birth weight below the 10th percentile as corrected for sex and parity 7 and had no congenital malformations. All of the women had been admitted because of the clinical diagnosis of IUGR. Nine had pregnancy-induced hypertension; proteinuria was detected in four women. Four patients were treated with antihypertensive drugs, and two received 80 mg of acetylsalicylic acid daily. No other medication was prescribed. The clinical data of the infants are shown in Table I. Most were delivered before 35 weeks of gestation. All but one (case number 12) had abnormal umbilical artery blood velocity waveforms, and in seven there was absence of end-diastolic frequencies above filter level (150 Hz). In one there was reversed flow during diastole. In the seven fetuses in whom blood velocity waveforms were measured in the internal carotid artery, the pulsatility index (Peak systolic velocity - Peak diastolic velocity / Mean velocity) was decreased, which probably indicated reduced cerebral vascular resistance. The blood velocity waveforms in the umbilical artery and internal carotid were related to observations made on normal fetuses.
Heart rate variation in fetuses with JUGR
Volume 166 Number 1. Part I
23
Table I. Clinical data of 13 fetuses with IUGR* GestatIOnal age Case No.
2 3 4 5 6 7 8 9 10 II
12 13
First deceleration
29 wk 0 days 30 wk I day 30 wk 5 days 30 wk 4 days 32 wk 6 days 31 wk 6 days 31wk 4 days 33 wk 6 days 3 1 wk 2 days 33 wk I day 34 wk 2 days 34 wk 5 days 35 wk 5 days
Pulsatility zndex Internal carotid artery
At cesarean sectIOn
Head / abdomen ratIO
Umbilical artery
29 wk 3 days
1.24
2.93t
30 wk 6 days
1.18t
2.05t
0.92
31 wk 0 days
1.26
2.93§
0.96
31 wk 3 days
1.26
32 wk 6 days
Birth weight
Mean minute range (msec)
gm
34.3
Umbilical artery P02 (kPa)
%
pH
975
<5
7 .11
18.7
900
< 2.3
7.17
12.9
900
< 2.3
7.25
2.22
1.93t
19.1
890
< 2.3
7.21
0.97
l.l2t
2.7It
33.8
1245
<5
7.22
33 wk 0 days
1.1lt
2.35t
22.0
1035
< 2.3
7.22
1.35
33 wk 6 days
l.l8
1.71
1.08
33.2
1130
< 10
7.21
3.44
33 wk 6 days
1.16
1.59
0.93
40.1
1310
< 2.3
7.28
2.70
34 wk 0 days
1.09t
2.71
0.91
23.2
1I60
< 2.3
7. 16
1.05
34 wk 4 days
l.l7
2.62t
0.75
~i.4
1I50
< 2.3
7.30
0.62
35 wk 0 days
1.12
2.10
24.9
1490
<5
7.30
2.23
36 wk 5 days
1.26
1.10
25.3
1750
< 2.3
7.26
1.70
37 wk 6 days
1.24
1.44
59.2
1735
< 2.3
7.22
0.97
*Data on head/abdomen ratio, pulsatility index (absolute values), and mean minute range refer to the last measurement before cesarean section. tAbsence of end-diastolic velocities. tSymmetric IUGR. §Reversed flow.
Before birth, nine fetuses exhibited asymmetric growth retardation; four were symmetrically growth retarded, according to the head /abdomen circumference ratios.s At birth, nine infants weighed <2.3rd percentile. Five of the nine fetuses with measurements had hypoxe mia (umbilical arterial P0 2 < 2.00 kPa) and in 10 of the 13 fetuses the pH was within the normal range (umbilical artery pH > 7.20)" There was one neonatal death . Other complications included intraventricular hemorrhage (n = I), necrotizing enterocolitis (n = I), and hyaline membrane disease (n = 4) . In the 13 fetuses 200 I-hour FHR records were made with a Hewlett Packard model 8040A cardiotocograph over a median period of 25 days (range 4 to 44 days) two to five times per week between 10 AM and 6 PM. The records were analyzed numerically with an on-line microcomputer system 9 adapted for use with a personal computer (Apricot Xen). This computer program for FHR analysis is identical to that of the commercially available Sonicaid System 8000. With this system a baseline is fitted to the FHR trace. Accelerations (2: 10 beats / min lasting 2: 15 seconds) and decelera-
tions (2: 10 beats/min lasting 2: 60 seconds or 2: 20 beats/ min lasting 2: 30 seconds) are then identified. FHR variation is measured as the range of mean pulse intervals about the baseline calculated minute by minute, with data from decelerations excluded; it is averaged to give a measure of long-term variation ("mean minute range" in milliseconds) . Episodes of high and low heart rate variation are identified when 5 of 6 successive minutes have a range of pulse intervals> 32 or < 30 milliseconds, respectively. Basal FHR is calculated as the mean FHR in episodes of low variation or, if no such episodes are present, as the mean pulse interval expressed as beats per minute. The fetal movements felt by the women were recorded with an event marker. No more than one movement every 1/16 minute was identified . An example of a I-hour recording is shown in Fig. 1. The number of records differed among the fetuses; therefore data were standardized per fetus per week (Figs. 2 and 3). It has been shown previously that FHR variation is > 30 milliseconds in 95% to 97% of normal third-
24 Snijders et al.
J anuary 1992 Am J Obstet Gynecol
RECORD
-
NGI_558
DATE
3-Aug-1988 Ge.tation 33 week. Ti. . of record inc 8859
Variation
High Low
Moves
23
32 1
Taco
,"
I
-
,
~--rL , 0j... , 35 25 30 15 20 0 5 10 Advice CONTINUE at 60 min Normal ANALYSIS at 60 min . Values
0
)
>12 <160 ) >8/hr 0 )
( >4 )
)
-~
21
42
22
20
A
,
I
i
40
"F
45
I
,
50
IIIII
60 55 min
2.3 1. SIONAL LOSS ( % ) .. , . . ....... .. .... . ... . 1 2 . CONTRACTION PEAKS . . .. ... . .. . .. . . ...... . .. . 30 3 . FETAL MOVEMENTS (per hour) . .... . .... .. .. . per min in High 0 . 00 in Low 0.38 150 4 . BASAL HEART RATE (bpm) . .... ... . . . . . ... .. . .. 3 5 . ACCELERATIONS >10 bpm & 15 sec . .. .. 1 6. DECELERATIONS > 20 lost beats .. ... . . . .. . .. Signal Variation msee Lost Length Troutrh Time (bpm) 3' before after loss % (min) beats (min) 11 . ., 22.7 20.3 52 1.8 6b 18 *
* *
7. HIOH EPISODES (min) Threshold 32 msee .... LOW EPISODES (min) ....
6
37
* *
8 . VARIATION OVERALL (Mean range
mllee) .... . .
26.6
INTERPRETATION OF VARIATION
Normal Questionable Abnormal
> 30 msee 20 - 30 msee < 20 msee
Fig. 1. Example of computer printout of I-hour record at 33 weeks. Heart rate variation is below normal range (26.6 milliseconds); there is one large deceleration.
trimester fetuses. 9 This value was taken as the lower limit of normality. Because neither this lower limit nor that of the incidence of fetal body movements 2 changes during the third trimester of pregnancy, the data were grouped according to the first occurrence of late decelerations without regard to the differences in gestational age. The occurrence of antepartum late decelerations was assessed visually from the original FHR traces 5 • 6 by clinicians not aware of the data from the numeric analysis. The women were delivered of their infants between < 1 day and 19 days after the first deceleration occurred (median interval 6 days). On average, visually assessed decelerations were present in 38% of the records made after the occurrence of the first deceleration.
Results The changes with time in basal heart rate, FHR variation, and the percentage of time that episodes of high FHR variation were present are shown in Fig. 2. There was a slight but significant increase in basal FHR when
decelerations occurred (Wilcoxon test, p < 0.05). On average, FHR variation decreased gradually with time . The median value fell below the normal range (30 milliseconds) at about the time at which decelerations a ppeared. A similar gradual decline occurred in the duration of episodes of high FHR variation (Fig. 2) and in the incidence of fetal movements as felt by the women (Fig. 3). The number of accelerations per hour was low initially and remained constant (Fig. 3). There were no differences among fetuses with symmetric and asymmetric IVCR except for the incidence of movements felt subjectively, which was somewhat higher in the fetuses with symmetric IVCR. Large interfetal differences were present (Figs. 2 and 3). In Figs. 4 and 5 basal heart rate, FHR variation, and episodes of high variation are shown for two fetuses delivered at 37 weeks and for two delivered at 31 weeks. In spite of similar birth weights there were large differences in FHR variation, which was already low in one fetus weeks before the onset of decelerations; in
Heart rate variation In fetuses with IUGR
Volume 166 Number 1, Part 1
HEART RATE (beats.min.-1)
NUMBER OF MOVEMENTS PER HOUR
'~h
160] 140 120
::A]" RA; VAjATlON:
moo
:l
m"" .,,"
NUMBER OF ACCELERATIONS PER HOUR
20 EPISODES OF HIGH VARIATION (% of time)
100
60
I 4J
t-
20
-3 (8
-2 I
(9 I
-1 112 I
25
0 1131
t
I
~1---+-t4 I
I
I
-3
-2
-1
o
181
191
(121
1131
I
+1 weeks 1121 (nl
t
FIRST DE[ELERA TlON
+1 weeks 1121 In I
Fig. 3. Changes with time in number of movements felt by women and in number of accelerations ( ~ 10 beats/min lasting ~ 15 milliseconds) in 13 growth-retarded fetuses. Data are grouped according to first occurrence of late decelerations and are presented as medians and interquartile ranges.
FIRST DE[ELERATION
Fig. 2. Changes with time in basal heart rate, heart rate variation, and percentage of time in which episodes of high heart rate variation are present in 13 growth-retarded fetuses. Data are grouped according to first occurrence oflate decelerations and are presented as medians and interquartile ranges. Numbers in parentheses are number of fetuses included. the other it was still high after the onset of decelerations (Fig. 4) or decreased when the first deceleration occurred (Fig. 5). All four fetuses in Figs. 4 and 5 had asymmetric growth retardation; no antihypertensive drugs were administered. There was a gradual increase in the pulsatility index of the umbilical artery (Fig. 6). The pulsatility index was increased ( > 2 SD above the mean) in 7 of 10 cases studied 3 weeks before the decelerations occurred, in 7 of 12 cases studied at - 2 weeks, in 10 of 13 cases at -1 week, and in 11 of 13 cases at the occurrence of decelerations. Comment
The results of this longitudinal study demonstrate that in growth-retarded fetuses, who are eventually delivered because of repetitive FHR decelerations, there is a gradual decline in long-term FHR variation. However. variation often remains within the normal range ( > 30 milliseconds) until (or even after) late decelerations occur. This finding is in agreement with data from earlier cross-sectional studies l ,2 and implies that often the FHR variation of fetuses with IUGR cannot be distinguished from that of normal fetuses before late decelerations appear. Furthermore, the coincidence of the decline in FHR variation and the occur-
HEART RATE (beats.mtn- 1)
160j 140 120
~r RAT~~',m"" EPISODES OF HIGH VARIATION (% of time)
1::~ 20j I
I
o t
-40
I 10
20I days
FIRST DECELERATION
• 37+ 6 9 1735gr
o
36 +5/1750 gr.
Fig. 4. Basal fetal heart rate, heart rate variation, and occurrence of episodes of high variation in two fetuses with similar birth weights who were delivered at 36 and 37 weeks' gestation, respectively. Data are grouped according to first occurrence of late decelerations. Note interfetal differences. rence of late decelerations indicates that both parameters should be considered as a rather late sign of fetal impairment, because antepartum FHR decelerations are associated with adverse neonatal neurologic outcome. 10
26
Snijders et al.
:1
January 1992 Am J Obstet Gynecol
10.5
HEART RATE(beats min.-1)
8.5
120
~r
RATE
6.5 4.5
~~;;,m'~1
2.5 0.5
I
-3 (10)
~ mSODES OF ~ t;~1
I -40
I -30
I -20
I -10
I
o t
I
10
I
20 days
I
-2 (12)
-1
I
o
(13)
(13)
I
I
+ 1weeks
Fig. 6. Changes with time in pulsatility index of umbilical artery in 13 growth-retarded fetuses. Pulsatility index decreases with gestation; therefore data are expressed as number of standard deviations by which measured values differed from appropriate normal mean for gestation . Data are grouped according to first occurrence of late decelerations and are presented as medians and interquartile ranges. Numbers in parentheses are number of fetuses included.
FIRST DECELERATION
• 31+ 3 9
890 gr.
Fig. 5. Basal fetal heart rate, heart rate variation, and occurrence of episodes of high variation in two fetuses with similar birth weights who were delivered at 31 weeks' gestation. Data are grouped according to first occurrence of late decelerations.
In the rhesus monkey it has been found that with progressive deterioration of the fetal condition, late heart rate decelerations first appeared. II This condition was associated with hypoxemia. Heart rate variation (accelerations) decreased later, when the arterial pH also fell. In the human fetus decreased FHR variation occurs (on average) when decelerations appear. Decelerations are, however, dependent on the amount of uterine activity (frequency, duration, and intensity of Braxton Hicks contractions), and with recordings of longer duration or with induced contractions a different time relationship may be found. In this respect it should be noted that in our study decelerations were only present in about 40% of records made after the occurrence of the first deceleration. It is likely that the initial reduction in FHR variation is mediated by fetal hypoxemia and that evident acidemia is absent until terminal FHR patterns develop. Evidence for this assumption has been provided by reports on associations between decreased FHR variation or development of antepartum FHR decelerations and fetal hypoxemia diagnosed at cordocentesis or at elective cesarean section. 2-4. 12 In addition, maternal hyperoxygenation in pregnancies complicated by IUGR has demonstrated that improved fetal oxygenation results in an increase of FHR variation," whereas variation is temporarily decreased directly after "hypox-
emic" events (late decelerations) .14 In fetal acidemia more severe reductions in FHR variation have been described, with simultaneous (shallow) decelerations following each Braxton Hicks contraction (terminal FHR pattern).<-6 In this study the interval between the last computerized FHR record and cesarean section varied considerably (1 to 48 hours). The correlation coefficient between FHR variation of the last records and umbilical arterial P02 at cesarean section was, however, in the same range (r = 0.56, n = 9; not significant; see Table I) as described by others (r = 0.52 and r = 0.66, respectively).3. 12 In our study and in that of Smith et al. 3 FHR variation was not significantly correlated with pH at cesarean section. In a series including severely compromised fetuses, a significant correlation between mean I-minute range and pH at cordocentesis has been described (r = 0.69).12 Although the evolution of the FHR patterns was similar in most fetuses, in some cases decelerations were preceded by reduced FHR variation for up to 5 weeks; in others the variation remained within the normal range even in the presence of decelerations. Differences in the trends in FHR changes could not be explained by differences in head/abdomen circumference ratios, birth weight, or abnormalities in Doppler velocity waveforms. However, they might reflect variations in adaptation to chronic compromise. Further evidence for adaptation to an adverse intrauterine environment can be derived from the finding that the development of decelerations coincided with a small but significant increase in basal FHR, a finding also reported in a previous cross-sectional study.2 However, perhaps prolonged low FHR variation in the absence of decelerations should be interpreted as
Volume 166 Number I, Part I
evidence for an inherently low FHR variation,I3 In normal fetuses there are large interindividual differences in FHR variation; within the same individual variation remains within a much narrower range. I5 16 The data of this study indicate that this premise also applies to fetuses with IUGR (Figs. 4 and 5). In spite of interindividual differences, most fetuses had a progressive decline in FHR variation. Thus, in the surveillance of the fetus with IUGR, it is better to use each fetus as its own control. 17 This is facilitated by an objective numeric analysis of FHR variation. In fetuses with IUGR, changes in umbilical artery waveforms usually precede FHR decelerations. In studies in which the latter were taken as an end point, abnormal waveforms had been present for at least 9 d aysI8 or had been present 17 days on average. 19 A study from our institute showed that the duration of this interval differs considerably (0 to 60 days) and that FHR decelerations occasionally occur when Doppler recordings are still normal. I9 In the present study there were two cases with normal waveforms at the occurrence of decelerations and one with a normal pattern still present just before delivery. To a large extent the considerable differences in time interval between first abnormal waveform and occurrence of FHR abnormalities can be explained by differences in gestational age. In early pregnancy this interval appears to be much longer than it is at late gestation. I9 The variable interval between the first abnormal Doppler recording and abnormality in the FHR indicates, however, that Doppler recordings of the umbilical artery are of limited value with respect to the timing of delivery and that a normal umbilical artery pusatility index cannot be considered a definite proof of fetal well-being. In summary, with progressive deterioration of the condition of the fetus with IUGR, there is a gradual fall in heart rate variation. Both antepartum decelerations and/ or reduceed FHR variation are associated with fetal hypoxemia and/or acidemia. The main additional values of numeric analysis of FHR variation are the identification of those fetuses, in whom reduced variation proceeds the development of decelerations and the provision of objective and longitudinal data on progressive deterioration in fetuses with apparently normal FHR variation or in fetuses with abnormal patterns in whom further maturation in utero is considered. REFERENCES 1. Visser GHA. Antenatal cardiotocography in the evaluation of fetal well-being. Aust N Z ] Obstet Gynaecol 1984;24:80-5.
Heart rate variation
In
fetuses with IUGR 27
2. Bekedam D], Visser GHA, Mulder EJH, PoelmannWeesjes G. Heart rate variation and movement incidence in growth retarded fetuses: the significance of antenatal late heart rate decelerations. AM ] OBSTET GYNECOL 1987; 157: 126-33. 3. Smith ]H, Anand K]S, Cotes PM, et al. Antenatal fetal heart rate variation in relation to the respiratory and metabolic status of the compromised human fetus. Br] Obstet Gynaecol 1988;95:980-9. 4. Visser GHA. Sadovsky G, Nicolaides KH. Antepartum heart rate patterns in small-for-gestational-age thirdtrimester fetuses: correlations with blood gas values obtained at cordocentesis. AM ] OBSTET GYNECOL 1990; 162:698-703. 5. Visser GHA, Huisjes H]. Diagnostic value of the unstressed antepartum cardiotocogram. Br] Obstet Gynaecol 1977;84:321-6. 6. Visser GHA. Redman CWG, Huisjes H]. Turnbull AC. Nonstressed antepartum heart rate monitoring: implications of decelerations after spontaneous contractions. AM ] OBSTET GYNECOL 1980;138:429-35. 7. Kloosterman GJ. On intrauterine growth. Int] Gynaecol Obstet 1970;8:895-912. 8. Campbell S, Thomas A. Ultrasound measurement of fetal head to abdomen circumference ratio in the assessment of growth retardation. Br] Obstet GynaecoI1977;84: 16574. 9. Dawes GS, Redman CWG, Smith ]H. Improvements in the registration and analysis of fetal heart rate records at the bedside. Br] Obstet Gynaecol 1985;92:317-25. 10. Visser GHA. Antepartum fetal heart rate patterns and subsequent handicap. In: Patel N, ed. Antenatal and perinatal causes of handicap. Bailleres Clin Obstet Gynaecol 1988;2: 117-24. 11. Murata Y, Martin CB, Ikenone T. et al. Fetal heart rate accelerations and late decelerations during the course of intrauterine death in chronically catheterized rhesus monkeys. AM] OBSTET GYNECOL 1982;144:218-23. 12. Ribbert LSM, Snijders RJM, Nicolaides KH, Visser GHA. Relation of fetal blood gases and data from computerassisted analysis of fetal heart rate patterns in small for gestation fetuses. Br J Obstet Gynaecol 1991 ;98:820-3. 13. Bekedam DJ. Fetal heart rate and movement patterns in growth retardation [PhD thesis]. University of Groningen, Groningen, The Netherlands: 1989:45-56. 14. Bekedam D], Visser GHA. Effects of hypoxemic events on breathing, body movements, and heart rate variation: a study in growth-retarded human fetuses. AM] OBSTET GYNECOL 1985;153:52-6. 15. Smith ]H, Dawes GS. Redman CWG. Low human fetal heart rate variation in normal pregnancy. Br ] Obstet Gynaecol 1987;94:656-64. 16. Visser GHA. Dawes GS, Redman CWG. Numerical analysis of the normal human antenatal fetal heart rate. Br J Obstet Gynaecol 1981 ;88:792-802. 17. Devoe LD, Castillo R, McKenzie], et al. Sequential nonstress testing with use of each fetus as its own control. AM ] OBSTET GYNECOL 1986;154:931-6. 18. Reuwer P]HM, Sijmons EA, Rietman GW, et al. Intrauterine growth retardation: prediction of fetal distress by Doppler ultrasound. Lancet 1987;2 :415-18. 19. Bekedam D], Visser GHA, van der Zee AG], et al. Abnormal velocity waveforms of the umbilical artery in growth retarded fetuses: relationship to antepartum late heart rate decelerations and outcome. Early Hum Dev 1990;24:79-89.
Key words: Intrauterine growth retardation, fetal heart rate monitoring, fetal hypoxemia, computer analysis In cross-sectional studies of fetuses with intrauterine growth retardation (IUGR), long-term heart rate variation is usually within the normal range in the absence of antepartum late fetal heart rate (FHR) decelerations, but it is reduced in the presence of late decelerations. 1. 2 Reduced FHR variation and the occurrence of decelerations coincide with hypoxemia, as determined in cord blood obtained by cordocentesis or at elective cesarean section,2°' and are assumed to precede the "terminal" state, in which fetal acidemia is present. 5 . 6 These data suggest that a reduction in heart rate variation is a rather late sign of impairment of the fetal condition. The cross-sectional design of these studies has not allowed determination of the actual sequence of changes in FHR patterns with progressive deterioration . In this study therefore heart rate and its variation were measured longitudinally in fetuses with IUGR to determine changes occurring with time and progressive deterioration.
From the Department of Obstetrics and Cynaecology, Unzverszty Hospztal Gronmgen. Supported by a grant from the Dutch Pnnses Beatnx Fonds (grant No . 87-2684). R eceIved f01' publication November 8, 1990; revtsed May 1-1. 1991; accepted May 28, 1991 . R epnnt r equests: Prof Dr. C .H .A . Visser, Department of Dbstetnes and Gynaecology, UnIVersity Hospital, P O . . Box 85500,3508 CA Utrecht, The Netherlands. 611 131262
22
Patients and methods The study group consisted of 13 women admitted to the obstetric ward. They were selected retrospectively on these criteria: (1) Computerized analyses of I-hour FHR records were available before and after the occurrence of antepartum late decelerations; (2) cesarean section had been performed because of recurrent late decelerations; (3) the infants had a birth weight below the 10th percentile as corrected for sex and parity 7 and had no congenital malformations. All of the women had been admitted because of the clinical diagnosis of IUGR. Nine had pregnancy-induced hypertension; proteinuria was detected in four women. Four patients were treated with antihypertensive drugs, and two received 80 mg of acetylsalicylic acid daily. No other medication was prescribed. The clinical data of the infants are shown in Table I. Most were delivered before 35 weeks of gestation. All but one (case number 12) had abnormal umbilical artery blood velocity waveforms, and in seven there was absence of end-diastolic frequencies above filter level (150 Hz). In one there was reversed flow during diastole. In the seven fetuses in whom blood velocity waveforms were measured in the internal carotid artery, the pulsatility index (Peak systolic velocity - Peak diastolic velocity / Mean velocity) was decreased, which probably indicated reduced cerebral vascular resistance. The blood velocity waveforms in the umbilical artery and internal carotid were related to observations made on normal fetuses.
Heart rate variation in fetuses with JUGR
Volume 166 Number 1. Part I
23
Table I. Clinical data of 13 fetuses with IUGR* GestatIOnal age Case No.
2 3 4 5 6 7 8 9 10 II
12 13
First deceleration
29 wk 0 days 30 wk I day 30 wk 5 days 30 wk 4 days 32 wk 6 days 31 wk 6 days 31wk 4 days 33 wk 6 days 3 1 wk 2 days 33 wk I day 34 wk 2 days 34 wk 5 days 35 wk 5 days
Pulsatility zndex Internal carotid artery
At cesarean sectIOn
Head / abdomen ratIO
Umbilical artery
29 wk 3 days
1.24
2.93t
30 wk 6 days
1.18t
2.05t
0.92
31 wk 0 days
1.26
2.93§
0.96
31 wk 3 days
1.26
32 wk 6 days
Birth weight
Mean minute range (msec)
gm
34.3
Umbilical artery P02 (kPa)
%
pH
975
<5
7 .11
18.7
900
< 2.3
7.17
12.9
900
< 2.3
7.25
2.22
1.93t
19.1
890
< 2.3
7.21
0.97
l.l2t
2.7It
33.8
1245
<5
7.22
33 wk 0 days
1.1lt
2.35t
22.0
1035
< 2.3
7.22
1.35
33 wk 6 days
l.l8
1.71
1.08
33.2
1130
< 10
7.21
3.44
33 wk 6 days
1.16
1.59
0.93
40.1
1310
< 2.3
7.28
2.70
34 wk 0 days
1.09t
2.71
0.91
23.2
1I60
< 2.3
7. 16
1.05
34 wk 4 days
l.l7
2.62t
0.75
~i.4
1I50
< 2.3
7.30
0.62
35 wk 0 days
1.12
2.10
24.9
1490
<5
7.30
2.23
36 wk 5 days
1.26
1.10
25.3
1750
< 2.3
7.26
1.70
37 wk 6 days
1.24
1.44
59.2
1735
< 2.3
7.22
0.97
*Data on head/abdomen ratio, pulsatility index (absolute values), and mean minute range refer to the last measurement before cesarean section. tAbsence of end-diastolic velocities. tSymmetric IUGR. §Reversed flow.
Before birth, nine fetuses exhibited asymmetric growth retardation; four were symmetrically growth retarded, according to the head /abdomen circumference ratios.s At birth, nine infants weighed <2.3rd percentile. Five of the nine fetuses with measurements had hypoxe mia (umbilical arterial P0 2 < 2.00 kPa) and in 10 of the 13 fetuses the pH was within the normal range (umbilical artery pH > 7.20)" There was one neonatal death . Other complications included intraventricular hemorrhage (n = I), necrotizing enterocolitis (n = I), and hyaline membrane disease (n = 4) . In the 13 fetuses 200 I-hour FHR records were made with a Hewlett Packard model 8040A cardiotocograph over a median period of 25 days (range 4 to 44 days) two to five times per week between 10 AM and 6 PM. The records were analyzed numerically with an on-line microcomputer system 9 adapted for use with a personal computer (Apricot Xen). This computer program for FHR analysis is identical to that of the commercially available Sonicaid System 8000. With this system a baseline is fitted to the FHR trace. Accelerations (2: 10 beats / min lasting 2: 15 seconds) and decelera-
tions (2: 10 beats/min lasting 2: 60 seconds or 2: 20 beats/ min lasting 2: 30 seconds) are then identified. FHR variation is measured as the range of mean pulse intervals about the baseline calculated minute by minute, with data from decelerations excluded; it is averaged to give a measure of long-term variation ("mean minute range" in milliseconds) . Episodes of high and low heart rate variation are identified when 5 of 6 successive minutes have a range of pulse intervals> 32 or < 30 milliseconds, respectively. Basal FHR is calculated as the mean FHR in episodes of low variation or, if no such episodes are present, as the mean pulse interval expressed as beats per minute. The fetal movements felt by the women were recorded with an event marker. No more than one movement every 1/16 minute was identified . An example of a I-hour recording is shown in Fig. 1. The number of records differed among the fetuses; therefore data were standardized per fetus per week (Figs. 2 and 3). It has been shown previously that FHR variation is > 30 milliseconds in 95% to 97% of normal third-
24 Snijders et al.
J anuary 1992 Am J Obstet Gynecol
RECORD
-
NGI_558
DATE
3-Aug-1988 Ge.tation 33 week. Ti. . of record inc 8859
Variation
High Low
Moves
23
32 1
Taco
,"
I
-
,
~--rL , 0j... , 35 25 30 15 20 0 5 10 Advice CONTINUE at 60 min Normal ANALYSIS at 60 min . Values
0
)
>12 <160 ) >8/hr 0 )
( >4 )
)
-~
21
42
22
20
A
,
I
i
40
"F
45
I
,
50
IIIII
60 55 min
2.3 1. SIONAL LOSS ( % ) .. , . . ....... .. .... . ... . 1 2 . CONTRACTION PEAKS . . .. ... . .. . .. . . ...... . .. . 30 3 . FETAL MOVEMENTS (per hour) . .... . .... .. .. . per min in High 0 . 00 in Low 0.38 150 4 . BASAL HEART RATE (bpm) . .... ... . . . . . ... .. . .. 3 5 . ACCELERATIONS >10 bpm & 15 sec . .. .. 1 6. DECELERATIONS > 20 lost beats .. ... . . . .. . .. Signal Variation msee Lost Length Troutrh Time (bpm) 3' before after loss % (min) beats (min) 11 . ., 22.7 20.3 52 1.8 6b 18 *
* *
7. HIOH EPISODES (min) Threshold 32 msee .... LOW EPISODES (min) ....
6
37
* *
8 . VARIATION OVERALL (Mean range
mllee) .... . .
26.6
INTERPRETATION OF VARIATION
Normal Questionable Abnormal
> 30 msee 20 - 30 msee < 20 msee
Fig. 1. Example of computer printout of I-hour record at 33 weeks. Heart rate variation is below normal range (26.6 milliseconds); there is one large deceleration.
trimester fetuses. 9 This value was taken as the lower limit of normality. Because neither this lower limit nor that of the incidence of fetal body movements 2 changes during the third trimester of pregnancy, the data were grouped according to the first occurrence of late decelerations without regard to the differences in gestational age. The occurrence of antepartum late decelerations was assessed visually from the original FHR traces 5 • 6 by clinicians not aware of the data from the numeric analysis. The women were delivered of their infants between < 1 day and 19 days after the first deceleration occurred (median interval 6 days). On average, visually assessed decelerations were present in 38% of the records made after the occurrence of the first deceleration.
Results The changes with time in basal heart rate, FHR variation, and the percentage of time that episodes of high FHR variation were present are shown in Fig. 2. There was a slight but significant increase in basal FHR when
decelerations occurred (Wilcoxon test, p < 0.05). On average, FHR variation decreased gradually with time . The median value fell below the normal range (30 milliseconds) at about the time at which decelerations a ppeared. A similar gradual decline occurred in the duration of episodes of high FHR variation (Fig. 2) and in the incidence of fetal movements as felt by the women (Fig. 3). The number of accelerations per hour was low initially and remained constant (Fig. 3). There were no differences among fetuses with symmetric and asymmetric IVCR except for the incidence of movements felt subjectively, which was somewhat higher in the fetuses with symmetric IVCR. Large interfetal differences were present (Figs. 2 and 3). In Figs. 4 and 5 basal heart rate, FHR variation, and episodes of high variation are shown for two fetuses delivered at 37 weeks and for two delivered at 31 weeks. In spite of similar birth weights there were large differences in FHR variation, which was already low in one fetus weeks before the onset of decelerations; in
Heart rate variation In fetuses with IUGR
Volume 166 Number 1, Part 1
HEART RATE (beats.min.-1)
NUMBER OF MOVEMENTS PER HOUR
'~h
160] 140 120
::A]" RA; VAjATlON:
moo
:l
m"" .,,"
NUMBER OF ACCELERATIONS PER HOUR
20 EPISODES OF HIGH VARIATION (% of time)
100
60
I 4J
t-
20
-3 (8
-2 I
(9 I
-1 112 I
25
0 1131
t
I
~1---+-t4 I
I
I
-3
-2
-1
o
181
191
(121
1131
I
+1 weeks 1121 (nl
t
FIRST DE[ELERA TlON
+1 weeks 1121 In I
Fig. 3. Changes with time in number of movements felt by women and in number of accelerations ( ~ 10 beats/min lasting ~ 15 milliseconds) in 13 growth-retarded fetuses. Data are grouped according to first occurrence of late decelerations and are presented as medians and interquartile ranges.
FIRST DE[ELERATION
Fig. 2. Changes with time in basal heart rate, heart rate variation, and percentage of time in which episodes of high heart rate variation are present in 13 growth-retarded fetuses. Data are grouped according to first occurrence oflate decelerations and are presented as medians and interquartile ranges. Numbers in parentheses are number of fetuses included. the other it was still high after the onset of decelerations (Fig. 4) or decreased when the first deceleration occurred (Fig. 5). All four fetuses in Figs. 4 and 5 had asymmetric growth retardation; no antihypertensive drugs were administered. There was a gradual increase in the pulsatility index of the umbilical artery (Fig. 6). The pulsatility index was increased ( > 2 SD above the mean) in 7 of 10 cases studied 3 weeks before the decelerations occurred, in 7 of 12 cases studied at - 2 weeks, in 10 of 13 cases at -1 week, and in 11 of 13 cases at the occurrence of decelerations. Comment
The results of this longitudinal study demonstrate that in growth-retarded fetuses, who are eventually delivered because of repetitive FHR decelerations, there is a gradual decline in long-term FHR variation. However. variation often remains within the normal range ( > 30 milliseconds) until (or even after) late decelerations occur. This finding is in agreement with data from earlier cross-sectional studies l ,2 and implies that often the FHR variation of fetuses with IUGR cannot be distinguished from that of normal fetuses before late decelerations appear. Furthermore, the coincidence of the decline in FHR variation and the occur-
HEART RATE (beats.mtn- 1)
160j 140 120
~r RAT~~',m"" EPISODES OF HIGH VARIATION (% of time)
1::~ 20j I
I
o t
-40
I 10
20I days
FIRST DECELERATION
• 37+ 6 9 1735gr
o
36 +5/1750 gr.
Fig. 4. Basal fetal heart rate, heart rate variation, and occurrence of episodes of high variation in two fetuses with similar birth weights who were delivered at 36 and 37 weeks' gestation, respectively. Data are grouped according to first occurrence of late decelerations. Note interfetal differences. rence of late decelerations indicates that both parameters should be considered as a rather late sign of fetal impairment, because antepartum FHR decelerations are associated with adverse neonatal neurologic outcome. 10
26
Snijders et al.
:1
January 1992 Am J Obstet Gynecol
10.5
HEART RATE(beats min.-1)
8.5
120
~r
RATE
6.5 4.5
~~;;,m'~1
2.5 0.5
I
-3 (10)
~ mSODES OF ~ t;~1
I -40
I -30
I -20
I -10
I
o t
I
10
I
20 days
I
-2 (12)
-1
I
o
(13)
(13)
I
I
+ 1weeks
Fig. 6. Changes with time in pulsatility index of umbilical artery in 13 growth-retarded fetuses. Pulsatility index decreases with gestation; therefore data are expressed as number of standard deviations by which measured values differed from appropriate normal mean for gestation . Data are grouped according to first occurrence of late decelerations and are presented as medians and interquartile ranges. Numbers in parentheses are number of fetuses included.
FIRST DECELERATION
• 31+ 3 9
890 gr.
Fig. 5. Basal fetal heart rate, heart rate variation, and occurrence of episodes of high variation in two fetuses with similar birth weights who were delivered at 31 weeks' gestation. Data are grouped according to first occurrence of late decelerations.
In the rhesus monkey it has been found that with progressive deterioration of the fetal condition, late heart rate decelerations first appeared. II This condition was associated with hypoxemia. Heart rate variation (accelerations) decreased later, when the arterial pH also fell. In the human fetus decreased FHR variation occurs (on average) when decelerations appear. Decelerations are, however, dependent on the amount of uterine activity (frequency, duration, and intensity of Braxton Hicks contractions), and with recordings of longer duration or with induced contractions a different time relationship may be found. In this respect it should be noted that in our study decelerations were only present in about 40% of records made after the occurrence of the first deceleration. It is likely that the initial reduction in FHR variation is mediated by fetal hypoxemia and that evident acidemia is absent until terminal FHR patterns develop. Evidence for this assumption has been provided by reports on associations between decreased FHR variation or development of antepartum FHR decelerations and fetal hypoxemia diagnosed at cordocentesis or at elective cesarean section. 2-4. 12 In addition, maternal hyperoxygenation in pregnancies complicated by IUGR has demonstrated that improved fetal oxygenation results in an increase of FHR variation," whereas variation is temporarily decreased directly after "hypox-
emic" events (late decelerations) .14 In fetal acidemia more severe reductions in FHR variation have been described, with simultaneous (shallow) decelerations following each Braxton Hicks contraction (terminal FHR pattern).<-6 In this study the interval between the last computerized FHR record and cesarean section varied considerably (1 to 48 hours). The correlation coefficient between FHR variation of the last records and umbilical arterial P02 at cesarean section was, however, in the same range (r = 0.56, n = 9; not significant; see Table I) as described by others (r = 0.52 and r = 0.66, respectively).3. 12 In our study and in that of Smith et al. 3 FHR variation was not significantly correlated with pH at cesarean section. In a series including severely compromised fetuses, a significant correlation between mean I-minute range and pH at cordocentesis has been described (r = 0.69).12 Although the evolution of the FHR patterns was similar in most fetuses, in some cases decelerations were preceded by reduced FHR variation for up to 5 weeks; in others the variation remained within the normal range even in the presence of decelerations. Differences in the trends in FHR changes could not be explained by differences in head/abdomen circumference ratios, birth weight, or abnormalities in Doppler velocity waveforms. However, they might reflect variations in adaptation to chronic compromise. Further evidence for adaptation to an adverse intrauterine environment can be derived from the finding that the development of decelerations coincided with a small but significant increase in basal FHR, a finding also reported in a previous cross-sectional study.2 However, perhaps prolonged low FHR variation in the absence of decelerations should be interpreted as
Volume 166 Number I, Part I
evidence for an inherently low FHR variation,I3 In normal fetuses there are large interindividual differences in FHR variation; within the same individual variation remains within a much narrower range. I5 16 The data of this study indicate that this premise also applies to fetuses with IUGR (Figs. 4 and 5). In spite of interindividual differences, most fetuses had a progressive decline in FHR variation. Thus, in the surveillance of the fetus with IUGR, it is better to use each fetus as its own control. 17 This is facilitated by an objective numeric analysis of FHR variation. In fetuses with IUGR, changes in umbilical artery waveforms usually precede FHR decelerations. In studies in which the latter were taken as an end point, abnormal waveforms had been present for at least 9 d aysI8 or had been present 17 days on average. 19 A study from our institute showed that the duration of this interval differs considerably (0 to 60 days) and that FHR decelerations occasionally occur when Doppler recordings are still normal. I9 In the present study there were two cases with normal waveforms at the occurrence of decelerations and one with a normal pattern still present just before delivery. To a large extent the considerable differences in time interval between first abnormal waveform and occurrence of FHR abnormalities can be explained by differences in gestational age. In early pregnancy this interval appears to be much longer than it is at late gestation. I9 The variable interval between the first abnormal Doppler recording and abnormality in the FHR indicates, however, that Doppler recordings of the umbilical artery are of limited value with respect to the timing of delivery and that a normal umbilical artery pusatility index cannot be considered a definite proof of fetal well-being. In summary, with progressive deterioration of the condition of the fetus with IUGR, there is a gradual fall in heart rate variation. Both antepartum decelerations and/ or reduceed FHR variation are associated with fetal hypoxemia and/or acidemia. The main additional values of numeric analysis of FHR variation are the identification of those fetuses, in whom reduced variation proceeds the development of decelerations and the provision of objective and longitudinal data on progressive deterioration in fetuses with apparently normal FHR variation or in fetuses with abnormal patterns in whom further maturation in utero is considered. REFERENCES 1. Visser GHA. Antenatal cardiotocography in the evaluation of fetal well-being. Aust N Z ] Obstet Gynaecol 1984;24:80-5.
Heart rate variation
In
fetuses with IUGR 27
2. Bekedam D], Visser GHA, Mulder EJH, PoelmannWeesjes G. Heart rate variation and movement incidence in growth retarded fetuses: the significance of antenatal late heart rate decelerations. AM ] OBSTET GYNECOL 1987; 157: 126-33. 3. Smith ]H, Anand K]S, Cotes PM, et al. Antenatal fetal heart rate variation in relation to the respiratory and metabolic status of the compromised human fetus. Br] Obstet Gynaecol 1988;95:980-9. 4. Visser GHA. Sadovsky G, Nicolaides KH. Antepartum heart rate patterns in small-for-gestational-age thirdtrimester fetuses: correlations with blood gas values obtained at cordocentesis. AM ] OBSTET GYNECOL 1990; 162:698-703. 5. Visser GHA, Huisjes H]. Diagnostic value of the unstressed antepartum cardiotocogram. Br] Obstet Gynaecol 1977;84:321-6. 6. Visser GHA. Redman CWG, Huisjes H]. Turnbull AC. Nonstressed antepartum heart rate monitoring: implications of decelerations after spontaneous contractions. AM ] OBSTET GYNECOL 1980;138:429-35. 7. Kloosterman GJ. On intrauterine growth. Int] Gynaecol Obstet 1970;8:895-912. 8. Campbell S, Thomas A. Ultrasound measurement of fetal head to abdomen circumference ratio in the assessment of growth retardation. Br] Obstet GynaecoI1977;84: 16574. 9. Dawes GS, Redman CWG, Smith ]H. Improvements in the registration and analysis of fetal heart rate records at the bedside. Br] Obstet Gynaecol 1985;92:317-25. 10. Visser GHA. Antepartum fetal heart rate patterns and subsequent handicap. In: Patel N, ed. Antenatal and perinatal causes of handicap. Bailleres Clin Obstet Gynaecol 1988;2: 117-24. 11. Murata Y, Martin CB, Ikenone T. et al. Fetal heart rate accelerations and late decelerations during the course of intrauterine death in chronically catheterized rhesus monkeys. AM] OBSTET GYNECOL 1982;144:218-23. 12. Ribbert LSM, Snijders RJM, Nicolaides KH, Visser GHA. Relation of fetal blood gases and data from computerassisted analysis of fetal heart rate patterns in small for gestation fetuses. Br J Obstet Gynaecol 1991 ;98:820-3. 13. Bekedam DJ. Fetal heart rate and movement patterns in growth retardation [PhD thesis]. University of Groningen, Groningen, The Netherlands: 1989:45-56. 14. Bekedam D], Visser GHA. Effects of hypoxemic events on breathing, body movements, and heart rate variation: a study in growth-retarded human fetuses. AM] OBSTET GYNECOL 1985;153:52-6. 15. Smith ]H, Dawes GS. Redman CWG. Low human fetal heart rate variation in normal pregnancy. Br ] Obstet Gynaecol 1987;94:656-64. 16. Visser GHA. Dawes GS, Redman CWG. Numerical analysis of the normal human antenatal fetal heart rate. Br J Obstet Gynaecol 1981 ;88:792-802. 17. Devoe LD, Castillo R, McKenzie], et al. Sequential nonstress testing with use of each fetus as its own control. AM ] OBSTET GYNECOL 1986;154:931-6. 18. Reuwer P]HM, Sijmons EA, Rietman GW, et al. Intrauterine growth retardation: prediction of fetal distress by Doppler ultrasound. Lancet 1987;2 :415-18. 19. Bekedam D], Visser GHA, van der Zee AG], et al. Abnormal velocity waveforms of the umbilical artery in growth retarded fetuses: relationship to antepartum late heart rate decelerations and outcome. Early Hum Dev 1990;24:79-89.