Fetal oxygenation at cordocentesis, maternal smoking and childhood neuro-development

European Journal of Obstetrics & Gynecology and Reproductive Biology 59 (1995) 21-24

ELSEVIER

GYNECOU

Fetal oxygenation at cordocentesis, maternal smoking and childhood neuro-development Peter W. Soothill *a, Richard A. Ajayi b, Stuart Campbell b, Euan M. Ross c, Kypros H. Nicolaides b aDepartment of Obstetrics and Gynaecology, Bristol University, St. Michael's Hospital, Bristol B52 8EG. UK bHarris-Birthright Research Centre of Fetal Medicine, King's College Hospital, Denmark Hill. London SE5 9RS, UK CDepartment of Community Paediatrics, King's College School of Medicine and Dentistry. St Giles' Hospital. London SE5, UK

Accepted 7 November 1994

Abstract Objective: To assess the relationship between subsequent neuro-development and measurements in small for gestational age (SGA) fetuses, including oxygenation at cordocentesis. Design: A longitudinal study of a cohort of SGA fetuses using those with normal oxygenation as controls. Subjects: 65 chromosomally normal children aged 12-66 months who, as fetuses, had been small for gestational age, had undergone cordocentesis and had been delivered after 32 weeks' gestation. Results: Griffiths's developmental quotient (DQ) was significantly associated with fetal blood pH (r= 0.41, P = 0.0008), Po 2 (r = 0.25, P = 0.047) and Pco2 (r = -0.28, P = 0.027) at cordocentesis but not with the severity of growth retardation (r = 0.18, P = 0.17). However, the significant correlations between DQ and both Po 2 and Pco 2 were explained by their co-correlations with pH and neither remained significant after removing the effect of pH by analysis of variance. The association between DQ and pH was still significant after removing the effect of confounding variables. Maternal smoking during pregnancy was significantly associated with DQ (F to remove = 6.89, P = 0.011) even after allowing for the effects of possible confounding variables including fetal blood pH. Conclusions: Both fetal pH at cordocentesis and maternal smoking correlate significantly with subsequent neuro-development. Further investigation into the causes of these associations is required. Keywords: Small for gestational age; Cordocentesis; Griffiths's developmental quotient; Po2; Pco2; pH

1. Introduction

Perinatal mortality has fallen so substantially over the past decade that reducing morbidity caused by events in fetal life has become an increasingly important aim of obstetric care [1]. However, it has been difficult to demonstrate any beneficial effect of fetal monitoring and obstetric intervention in labour on conditions such as cerebral palsy [2]. Furthermore, there is very little association between assessments at birth, such as the pH of cord blood, and subsequent impaired neurodevelopment [3]. Nevertheless, there is a widely held view, and supporting evidence, that a link exists between complications of pregnancy and impaired neurodevelopment in childhood [4-8]. The likely explanation * Corresponding author.

for this apparent discrepancy is that acute changes during labour in previously healthy fetuses may be common but produce little if any long-term damage while chronic deprivation during fetal development may be important. Fetal blood sampling before labour has given us an objective way of assessing placental function [91. In small for gestational age (SGA) fetuses it allows classification into normal small, hypoxic/starved small, and abnormal small fetuses (with genetic disease or congenital infections). Abnormal blood gas results reflect utero-placental dysfunction which must have been present sufficiently long for the reduced growth rate to be detected. Unlike scalp blood gas measurements in labour, abnormalities of blood gases at cordocentesis in SGA fetuses should not be regarded as a ~snap-shot' measurement with little relevance to changes over time but as a means of determining different pathological

0028-2243/95/$09.50 © 1995 Elsevier Science Ireland Ltd. All rights reserved SSDI 0028-2243(94)02013-5

22

P. Iv. Soothill et aL /European Journal of Obstetrics & Gynecology and Reproductive Biology 59 (1995) 21-24

groups reflecting longer term events and the aetiology of the reduced growth. Fetuses can survive long periods of acidaemia and in one study, after stratification by Doppler results, blood gases at cordocentesis did not predict perinatal death [10]. If blood gases in SGA fetuses also do not predict morbidity it would be inappropriate to use these results to guide obstetric management or assess other less invasive tests [11]. The aim of this study was to test the null hypothesis that chronic hypoxaemia or acidosis in fetal life has no effect on subsequent neurodevelopment. 2. Subjects and methods

A Griffiths neuro-developmental assessment was carried out on chromosomally normal children, who as fetuses had been SGA (detected by ultrasound and confirmed by a birth weight < - 2 S.D.s), had undergone cordocentesis for assessment of the cause of the reduced growth and had been delivered after 32 weeks' gestation. All were singletons and neuro-development was assessed at the age of 12-66 months (mean, 28.5 months). Therefore, although the study was prospectively planned it was retrospective to the cordocentesis data. The requirement that they should be born after 32 weeks' gestation was in order to avoid any effect of the complications of prematurity (such as intraventricular haemorrhage) on neuro-development. Although l l0 fetuses seen in our unit fulfilled the above criteria, 68 were seen because eight died in infancy, 20 families could not be traced and 14 declined to participate. Three children who came for follow-up (who were thought to be developing normally) could not be fully assessed; two spoke languages other than English, impairing our ability to assess hearing, speech and practical reasoning, and one would not co-operate. There was no significant difference in maternal smoking or prenatal findings between the 65 children who were assessed and those who were not. For example, regarding fetal pH in multiples of S.D., children assessed had a mean pH = -1.62 S.D.s and those not assessed had a mean pH -1.81 S.D.s; (t = 0,49, n = 110, NS). The antenatal diagnosis of intrauterine growth retardation was based on a fetal abdominal circumference < - 2 S.D.s for gestational age. Cordocentesis was done as described previously [9] and the results of all investigations were made available to the referring obstetricians, who decided the timing and mode of delivery which was in a large number of different hospitals. The neuro-developmental assessments were done according to the Griffiths scales [12,13] by a certified Griffiths assessor (R.A.A.) who was strictly blind to the cordocentesis results. The following additional information was recorded: (1) social class according to occupation; (2) race, which was Caucasian, Afro-Caribbean or

Asian; (3) maternal smoking during the pregnancy; (4) interval between cordocentesis and delivery; (5) gestational age at delivery; (6) birth weight; and (7) the child's age at neuro-developmental assessment. Data about neonatal condition were not collected since assessments at birth such as Apgar scores and cord pH at delivery have been shown to be poor predictors of neurodevelopment [3]. The study was approved by the Hospital's Ethics Committee and an information letter and request to contact the parents were sent to the families" general practitioners. In normal pregnancies fetal blood Po2, Pco~ and pH [14] and birth weight [15] change with gestational age. Therefore, the measured values were expressed as multiples of standard deviation from the appropriate normal mean for gestational age. Comparison of means was by the unpaired t-test, the significance of associations between two variables was assessed by Pearson's correlation coefficient and that between multiple variables was determined by analysis of variance. 3. Results

Before our assessment 62 of the 65 children were thought to be developing normally but three (4%) had been diagnosed as having cerebral palsy. One of these children had been the most severely acidemic fetus in the group, one had had a blood pH below the 5th centile and another above the 95th centile (see Fig. 1). The correlations between DQ and fetal blood gases described below remained significant even if these three cases with cerebral palsy were excluded from the analysis (e.g. for pH, r = 0.38, n = 63, P = 0.002). Subsequent DQ correlated significantly with fetal

pH (SD's)

-2

.... i_ ~ ~ .~ .- . -. .

-4

J

x

× ~K

×

X~: X

-6

X "~

-8 -10

-12 - 14 . . . . . . 70

~ 80

Griffiths

~ -

90

Developmental

1 O0

. . . . . . . . . . . .

110

:. . . . . . . . .

120

130

Quotient

Fig. 1. Griffiths'sneuro-developmentalquotient (DQ) and fetal blood pH (expressed in multiples of S.D. from the normal mean) at cordocentesis in chromosomallynormal SGA fetuses (r = 0.41o n = 65, P = 0.0008). The square dots indicate the results from the three children with cerebral palsy.

P. 14I. Soothill et al./ European Journal of Obstetrics & Gynecology and Reproductive Biology 59 (1995) 21-24

blood pH (r = 0.41, P = 0.0008; Fig. 1), Po 2 (r = 0.25, P = 0.047) and Pc02 (r = - 0 . 2 8 , P = 0.027) at cordocentesis but not the severity of growth retardation (birth weight expressed in multiples of S.D. from the normal mean) (r = 0.18, P = 0.17). However, the significant correlations between DQ and both P02 and Pc02 were explained by their co-correlations with pH, and neither remained significant after removing the effect of pH by analysis of variance (for P02, F to remove = 0.001, P = 0.99; for PCO2, F to remove 0.001, P = 0.98). No significant correlation was found between gestational age and DQ (r = 0.2, n = 62, NS, Fig. 2). This supports our 32 weeks' gestation cut off to avoid the complications of prematurity resulting from changed management following the blood gas results. There was also no significant correlation between fetal head circumference at the time of cordocentesis (expressed in multiples of S.D.) and subsequent DQ (r = 0.19, n = 65, NS). Analysis of variance was used to allow for the possible confounding effects of social class, racial group, maternal smoking, fetal blood PO2 and Pco2, cordocentesis to delivery interval, gestational age at delivery, birth weight and the age of the child at assessment, but the association between DQ and pH remained significant even after allowing for all of these (F to remove = 4.17, P = 0.046). After removing the effects of all the other variables listed above and fetal blood pH, whether the mother smoked or not in pregnancy significantly predicted DQ (F to remove = 6.89, P = 0.011). Although the correlation between the reported number of cigarettes smoked and DQ was not statistically significant (r = -0.125, n = 24, P = 0.56) the mean DQ of the

Developental Q u o t i e n t 130

x 120

.

x 110

X x

100 >(

x

.

X x x Xx

X X

x

90

.

.

X

x X

×

XXxX

X

X X

XxxX -~

~xx x x x ~ x

X

X

xX

× x x.

x x x X

x

X

x

I

80 70 32

x

BX

x

33

34

35

36

37

38

39

40

Gestational age (Wks)

Fig. 2. There was no significant correlation between gestational age at delivery and Griffiths's neuro-developmental quotient (DQ) (r = 0.19, n = 65, NS). This suggests that by including only cases delivered over 32 weeks' gestation we reduced any effect of the complications of prematurity on DQ. The square dots indicate the results from the three children with cerebral palsy.

23

children of mothers who smoked in pregnancy (n = 24) was 96.1 compared with 103.1 when they did not (n = 38).

4. Discussion

The finding that fetal blood pH significantly predicted subsequent neuro-development suggests that SGA fetuses with utero-placental dysfunction severe enough to cause chronic acidosis may not achieve their full genetic neuro-developmental potential even in the absence of obvious handicap. This supports the hypothesis that blood gases at cordocentesis in SGA fetuses reflect chronic utero-placental function and so the cause of the reduced growth. They are not a 'snap-shot' of fetal wellbeing in a rapidly changing condition and any comparison with the use of fetal scalp pH in labour is misleading. It is also very important not to extrapolate these results to fetal acidaemia in association with other conditions, whether acute (e.g. labour, maternal hemorrhage or hypotension) or chronic (e.g. maternal diabetes) [16] or severe fetal anaemia [17]. The results of this study suggest that acidaemia may be too late in the disease process to be used as an ideal indication for delivery. The association described may be because fetal acidaemia is a marker of the severity of utero-placental dysfunction rather than any direct causative effect of acidaemia on brain development. Inadequate placental function severe enough to stunt fetal growth can produce many other fetal nutritional and metabolic effects such as hypoglycaemia, erythroblastosis, thrombocytopaenia, leucopaenia, hypertriglyceridaemia, hypoaminoacidaemia and hypothyroidism (reviewed by Soothill et al. [18]) which in themselves could impair neuro-development. However, since all of these are associated with acidaemia, pH provides a simple indicator of disease severity and it is likely that, if maturity allows, delivery at an early stage in the asphyxial process may limit damage. The observation that Po2 failed to predict DQ independently of pH may be useful if utero-placental dysfunction can be identified before damage to future neuro-development has occurred. However, no change in management sl~ould be instituted until this hypothesis has been tested prospectively with an intervention trial. Three of the 65 children (4%) had been diagnosed as having cerebral palsy and, although this is a large proportion, the pH results demonstrate the heterogeneity of this condition. One was severely acidaemic as a fetus, one was borderline and the other had an unusually high blood pH. Clearly the aetiologies of cerebral palsy are very complicated and utero-placental dysfunction and fetal acidosis will be associated with only some of these cases. It was possible that the magnitude of the relationship between blood gases at cordocentesis and neuro-

24

P.W. Soothill et al. /European Journal of Obstetrics & Gynecology and Reproductive Biology 59 (1995) 21-24

development may have been underestimated by studying a group who were all SGA at birth. Some children who had normal results at cordocentesis could have been neuro-developmentally delayed as the result of other, unidentified, pathologies associated with their unusually small size. However, in our group this seems not to have been the case since those children had a mean DQ of around 100. It must be remembered, however, that our patients were highly selected both by referral to our unit and by our decision to undertake cordocentesis. We cannot determine whether the association between maternal smoking and subsequent neurodevelopment was the result of a harmful effect in utero or a hostile post-natal environment. Although it is impossible in this study to allow for all factors likely to influence neuro-development, the persistence of the correlation between DQ and chronic fetal pH after removing the effects of social class and race makes an intrauterine effect more likely. Recently a histological placental abnormality associated with smoking has been described even in the first trimester of pregnancy, which might affect placental function [19]. Since our findings demonstrate that the effect of smoking was not mediated by abnormal fetal blood gases, research into other factors may be useful. The association between pH at cordocentesis in SGA fetuses and neuro-development indicates that chronic fetal acidaemia may be important. It supports the use of blood gases at cordocentesis as an end-point in the study of non-invasive tests of fetal well-being. Since the tests currently used (e.g. biophysical profile including cardiotocography) become abnormal with fetal acidaemia rather than hypoxaemia [20], the ideal timing of delivery of growth retarded fetuses is likely to be earlier than it is in present antenatal practice. Acknowledgement This study was supported by a grant from South East Thames Regional Health Authority, (LORS grant 90/11). References [1] Soothill PW. Cordocentesis and fetuses that are small for gestational age. N Engl J Med 1993; 328: 728-729.

[2] Stanley F J, Watson L. The cerebral palsies o1 western Australia: trends, 1968 to 1981. Am J Obstet Gynecol 1988; 158: 89-93. [3] Ruth VJ, Ravio KO. Perinatal brain damage: predictive value of metabolic acidosis and the Apgar score. Br Med J 1988: 297: 24-27 [4] Freud S. (1897) Infantile cerebral paralysis. Translator: Russin LA, University of Miami Press 1968: 142. [5] Low JA, Pancbam SR, Worthington D, Boston RW. Clinical characteristics of pregnancies complicated by intrapartum fetal asphyxia. Am J Obstet Gynecol 1975: 121: 452-455. [6] Soothill PW, Ajayi RA, Campbell S, Ross EM. Candy DCA, Snijders, Nicolaides KH. Relationship between fetal acidemia at cordocentesis and subsequent neuro-development. Ultrasound Obstet Gynecol 1992; 2: 80-83. [7] Naeye R, Peters EC. Antenatal hypoxia and low IQ values. Am J Dis Child 1987~ 141: 50-54. [8] Breart G, Poisson-Salomon A. Intrauterine growth retardation and mental handicap: epidemiological evidence. Bailliere's Clin Obstet Gynaecol, 1988: 2: 92-93, [9] Soothill PW, Nicolaides KH, Campbell S. Prenatal asphyxia, hyperlacticaemia, hypoglycaemia and erythroblastosis in growth retarded fetuses. Br Med J 1987: 294: 1051-1053. [10] Nicolini U, Nicolaides P, Fisk NM et al. Limited role of fetal blood sampling in prediction of outcome in intrauterine growth retardation. Lancet 1990; ii: 768-772 [11] Soothill PW~ Nicolaides KH, Rodeck CH, Campbell S. Effect of gestational age on fetal and intervillous blood gas and acid base values in human pregnancy, Fetal Ther 1986: !: 168-175. [12] Griffiths R. The abilities of babies. A study in mental measurement. University of London Press, London, 1954. Reprinted: The Test Agency Ltd, Amersham, UK 1986: 100. [13] Griffiths R. The abilities of young children. A comprehensive system of mental measurement for the first 8 years of life. Association for Research in Child Development 1970. Reprinted: The Test Agency Ltd, Amersham, UK 1984, [14] Soothill PW, Nicolaides KH, Bilardo CM, Campbell S, Relationship of fetal hypoxia in growth retardation to mean blood velocity in the fetal aorta. Lancet 1986: ii: I 118-t 120. [15] Yudkin P L Aboualfa JA, Eyre JA, Redman CWG, Wilkinson AR. New birthweight and head circumference centiles for gestational ages 24-42 weeks, Early Hum Dev 1987: 15: 45-52, [16] Salverson D, Brudenell M, Nicolaides K H Fetal polycythemia and thrombocytopenia in pregnancies complicated by maternal diabetes. Am J Obstet Gynecol ~in press). [17] Soothill PW, Nicolaides KH, Rodeck C H The effect of anaemia of fetal acid-base status. Br .I Obstet Gynaecol 1987: 94: 880-883 [18] Soothill PW, Ajayt RA, Nicolaides KH. Fetal biochemistry in growth retardation. Early Hum Dev 1992: 29: 91-97, [19] Jaunieax E. Burton G J. The effect of smoking in pregnancy on placental morphology, Obstet Gynecol 1992: 79: 645-648. [20] Ribbert LSM, Snidjers RJM. Nicolaides KH, Vlsser GHA, Relationship between fetal biophysical profile and blood gas values at cordocentesis in severely growlh retarded fetuses. Am J Obstet Gynecol 1990: 163: 569-57t