The use of intrapartum fetal blood lactate measurements for the early diagnosis of fetal distress

The use of intrapartum fetal blood lactate measurements for the early diagnosis of fetal distress A. Eguiluz, Murcia,

Spain,

A. L@ez and

Bernal,

Oxford,

K. McPherson,

J. J. Parrilla,

and L. Abad

England

Lactate concentrations were measured during labor and at delivery in blood samples from the fetal presenting part and from the umbilical cord with the use of a rapid electrochemical technique. The value of these measurements to discriminate between normal and distressed fetuses was compared to that of pH, base excess, Pco2 and PO, measurements in the same blood samples. The fetuses were divided into three groups, normal, prepathologic, and pathologic, according to the presence and severity of fetal distress as evaluated by Apgar score, intrapartum cardiotocography, meconium staining of the amniotic fluid, and cord arterial pH at birth. Lactate and pH provided the best parameters to distinguish between groups, with lactate having the most discriminating power at least in early labor and midlabor. The prospective value of discriminant functions derived from lactate and pH data was good when the fetuses were allocated into the normal group but poor when an attempt was made to allocate the fetuses into prepathologic and pathologic groups, with a high false negative rate. However, the discriminating ability was improved when prepathologic and pathologic fetuses were included into one single abnormal group. These results confirm the potential use of rapid fetal blood lactate measurements for the early diagnosis of intrapartum fetal distress. (AM. J. OBSTET. GYNECOL. 147:949, 1983.)

A major aim in modern obstetrics is the early diagnosis of fetal distress. The use of intrapartum fetal cardiotocography and pH measurements in blood samples obtained from the fetal presentation is well established clinically. However, the interpretation of cardiotocography tracings is difficult and, while a normal tracing almost always ensures a good fetal outcome, the reverse does not hold true; thus, abnormal fetal heart rate patterns may occur in the absence of fetal distress.‘-” Furthermore, pH values from distressed and nondistressed fetuses overlap considerably and provide little information as to the severity of fetal distressdm6 The level of blood lactate has been considered a good parameter for evaluating the presence and severity of fetal distress since lactate levels rise rapidly in response to a variety of stimuli, such as catecholamine release, respiratory or metabolic alkalosis, and, particularly, hypoxemia.‘, * Lactate is synthesized from pyruvate in a reaction catalyzed by lactate dehydrogenase, a key enzyme in anaerobic metabolism. In conditions of tissue hypoxia, lactate dehydrogenase is activated and lactate From the Department

of Obstetrics and Gynaecology, Ciudad Sanitariu, “Virgen de la Arrkzca,” University of Murcia, the Nufjeld Departknt of Obstetrics and Gynaecolo&>ohn Radclz;ffe Ho&al. and the Debartment of Communitv Medicine, Radcliffe*< In&ma& Universiti of Oxfonl Received for publication January 31, 1983. Revised June 17, 1983. Accepted August I, 1983. Reprint requests: A. L6pez Bern&, Nuffield Department of Obstetrics and Gynaecology, John Radcliffe Hospital, Headington, Oxford, England OX3 9011.

production increases in a close relationship to the degree of oxygen deficit. 7, ’ Many workers have measured lactate levels in umbilical cord blood in relation to fetal distress and demonstrated a direct relationship between fetal hypoxia and excess lactate.g-” This relationship has been confirmed by lactate measurements in blood obtained from the fetal presenting part during labor.“* “* “j However, these workers measured lactate with a spectrophotometric technique which is timeconsuming and, therefore, limits the clinical application of the results for the study of the fetus during birth. The introduction of a rapid electrochemical technique has recently allowed accurate lactate measurements within 1 minute of obtaining a blood sample of less than 100 ~1.‘~ This has enabled the obstetrician to measure lactate in fetal blood samples obtained during labor. I53 I6 The aim of this study was to determine whether rapid lactate measurements in fetal scalp blood samples are useful in discriminating between normal and distressed fetuses during labor and to compare these measurements with other parameters, such as pH, base excess, Pco?, and PO,, in common use for the evaluation of fetal well-being. Material

and methods

A total of 102 patients were studied. All women were in labor at term (37 to 42 weeks) with a single fetus presenting by the vertex. The study was approved by the ethics committee of the “Virgen de la Arrixaca” Hospital. Full and informed consent was obtained from 949

950

Table

Eguiluz

et al.

I. Clinical

December Am. J. Obstet.

data of patients

croup Normal (group 1; n = 52) Prepathologic (group 2; n = 26) Pathologic (group 3; n = 24)

15, 1983 Gynecol.

in study* Maternal (Yd

age

Gestational IwW

Parity

Onset of labor

age

Spontaneous

Induced

26.3

2 1.1

0.9 2 0.2

39.9

2 0.3

42

10

26.9

? 1.7

0.9 t 0.3

40.8

2 0.4

18

8

27.9

-t 1.6

1.1 5 0.3

40.2

2 0.3

19

5

*See text for explanation of groups. Figures are means 2 SEM. TTime from the beginning of the active phase of labor to delivery. &S = Spontaneous vaginal delivery: V = vacuum extraction; F = forceps;

all patients. Labor was of spontaneous onset in 79 women and was induced in the remaining 23 women. All women received an intravenous infusion of oxytotin, for either the augmentation or the induction of labor. Further clinical details of the patients are given in Table I. In every case the fetal heart rate was monitored throughout labor with a fetal scalp electrode which was inserted in early labor before the first fetal blood sample was collected. An intramuscular analgesic, most commonly meperidine, was administered as required. Cesarean sections were performed with the patients under general anesthesia. Maternal peripheral blood samples were taken in early labor and immediately after delivery. Fetal blood samples were obtained from the scalp in early labor, when the cervix was 3 to 4 cm dilated, in midlabor, at a cervical dilatation of 5 to 7 cm, and during secondstage labor. Umbilical cord arterial and venous blood samples were also collected at delivery. Whole blood (200 ~1) was used for analyses. Lactate was measured by a radiochemical method” with the use of Roche lactate analyzer 640. Blood gases and parameters of acid-base balance were measured with a Radiometer BMS 3MK2. The patients were classified into three groups as follows: Group 1 consisted of 52 patients with clear amniotic fluid and normal cardiotocographic tracings. The neonates had a 5-minute Apgar score of 9 to 10 and the pH in arterial cord blood at delivery was ~7.25. This was considered the normal group. Group 2 consisted of 26 patients with meconiumstained amniotic fluid and/or abnormal cardiotocographic tracings (late or variable decelerations, decreased baseline variability, bradycardia, or tachycardia). The neonates had a 5-minute Apgar score of 7 to 10 and an arterial cord blood pH ~7.20. This group was considered prepathologic. Group 3 consisted of 24 patients with meconiumstained amniotic fluid and/or abnormal cardiotocographic tracings. The neonates had a 5-minute Apgar

CS = cesarean

section

score of s6 and a pH (7.20 in arterial cord blood. This was considered the pathologic group. It should be noted that almost 50% of the patients were classified as “abnormal” (groups 2 and 3), but this is probably a reflection of the fact that the patients studied included a relatively greater proportion with high-risk pregnancies compared to the hospital population as a whole. In group 3, three women had preeclampsia, two patients developed intrapartum pyrexia, and three were delivered of small-for-dates infants. In two instances the cord was tightly wound around the baby’s neck at delivery. Nevertheless, in most cases of fetal distress the etiology was unknown. There were two perinatal deaths in group 3. One baby with congenital heart disease died soon after birth. Another baby with severe meconium aspiration developed a pulmonary hemorrhage and sepsis and died 5 days after birth. In group 2 there were two small-for-dates babies and two macrosomic infants born to diabetic mothers. A further two women in this group had preeclampsia. In group 1 two babies were small for dates but there was no other maternal or fetal pathology. Statistical analysis. Statistical differences were assessed by paired t test and one-way analysis of variance. Since not all infants had data obtained at all points, a series of paired t tests were used on existing data to detect differences within groups among the three stages of labor and cord arteriovenous differences. One-way analysis of variance was used at each stage to test for differences between groups. Differences were considered statistically significant at p < 0.05. In order to select the parameters that differentiate best between groups and provide the most information to predict the fetal outcome, the data were studied further by means of discriminant function analysis. The idea was to estimate a linear function of these parameters which distinguished most successfully between the groups. We used a stepwise approach as

Volume Number

Early

147 8

Duration of labor? (k)

Mode S

V

F

CS

Newborn weight (kg)

4.1 ‘- 0.4

38

6

2

6

3.3 * 0.1

4.7 2 0.7

15

5

1

5

3.5 k 0.1

4.9 2 0.8

8

8

2

6

3.3

diagnosis

951

of fetal distress

of delivery$

k 0.1

provided in the Statistical Package for the Social Sciences suite of programsI which selects that parameter which distinguishes best first and subsequently chooses parameters according to their extra discriminating power. Separate analyses were performed on the data from early labor and midlabor. Moreover, at each stage we tried to discriminate between the three groups defined above in addition to combining the prepathologic (group 2) and pathologic (group 3) groups into a single group denoted as abnormal. The value of the discriminant function so derived was evaluated by applying it prospectively to the data from patients whose group was known but who had not been used in its calculation. It was not possible to perform this analysis on data from second-stage labor because of the relatively small number of samples. Results Fig. 1 shows the mean fetal and maternal lactate levels in the three groups of patients at different stages of labor. In group 1, lactate concentrations in scalp blood rose from 0.98 + 0.04 mmol/L (SEM) in early labor to 1.68 -C 0.06 mmol/L at the beginning of the second stage of labor. The highest values, 2.41 f 0.11 mmol/L, were obtained in cord arterial blood immediately after delivery. The increase in lactate levels observed at each stage (early labor, midlabor, secondstage labor, and cord arterial levels) was highly significant. There was also a significant cord arteriovenous difference and a significant difference between maternal samples obtained in early labor and those obtained post partum. Although the absolute values were higher, a similar pattern of lactate levels was observed in groups 2 and 3. In group 2 lactate levels rose from 2.03 -C 0.19 mmol/L in early labor to 3.04 + 0.18 mmol/L in the second stage and 3.7 1 2 0.13 mmol/L in cord arterial blood at delivery. The corresponding values in group 3 were 2.2 + 0.22, 4.31 & 0.47, and 5.89 2 0.52 mmol/L, respectively. Again, the increase at each stage was significant. Cord arteriovenous differences were not significant in group 2 but were significant in group 3. There was also a significant in-

n J ,

Early labor

I

Midlabor

1

Second stage

I

Umb. artery

I

Umb. vein

Fig. 1. Fetal lactate levels during labor and at delivery. The open symbols represent maternal values. Results are means * SEM (vertical bars, not drawn when smaller than the symbols). Circles = Normal group (n = 42 to 52). Squares = Prepathologic group (n = 13 to 26). 7‘tiangles = Pathologic group (n = 12 to 24).

crease in maternal lactate concentrations between samples obtained in early labor and those obtained post partum in both groups 2 and 3. In group 1 lactate concentrations in early labor were significantly higher in maternal blood than in fetal scalp blood. This difference was lost in groups 2 and 3. However, after delivery, lactate levels in cord arterial blood were much higher than in maternal blood in all three groups. From the data displayed in Fig. 1, it was evident that there were systematic differences between groups at each stage. This was confirmed by analysis of variance (p < 0.001 for all stages, including cord venous blood samples). The mean fetal and maternal values for pH, Pco2, base excess, and PO, are shown in Fig. 2. In general, these values followed the patterns described previo~sly.‘~, Is As expected, pH values in fetal scalp blood tended to fall with the progress of labor. This fall was observed in all three groups but was particularly steep in group 3. In this group, the mean pH values fell significantly from 7.29 + 0.01 in early labor to 7.25 * 0.01 in midlabor and 7.18 + 0.02 in the second stage. In all groups pH values in cord arterial blood samples at delivery were significantly lower than pH values in fetal scalp blood samples at the second stage of labor. It was, again, evident that there were systematic differences in mean pH values between groups (p < 0.005 to
952

Eguiluz et al.

December Am. J. Obstet.

15, 1983 Gynecol.

Table IIA. Cases allocated by discriminant function analysis-Early labor: Group 1, group 2, and group 3 Allocation

7.2

Actual

group*

Normal

Normal Prepathologic Pathologic

Prepathologic

8 3 3

Pathologic

0 3 0

1 4 6

*These data are derived from patients whose group was known but whose values for fetal lactate and pH had not been used to calculate the discriminanr functions used for the allocation. Table IIB. Cases allocated by discriminant function analysis-Early labor: Group 1 and group 2 plus group 3 Allocation

/I F

*Abnormal

15I

Midlabor

I

second rtage

1

lJmb. artery

Normal

9 5 3

group = Prepathologic

Abnormal* 0 5 6 group plus pathologic

group.

25-

I

group

Normal Prepathologic Pathologic

f

T-----I

Early labor

A&al

I 5

I-

I

Umb. vein

Fig. 2. Fetal acid-base status and blood gases during labor and at delivery. The open symbols represent maternal values. Results are means -C SEM (vertical bars). Circles = Normal group (n = 28 to 42). Squures = Prepathological group (n = 11 to 23). Triangles = Pathologic group (n = 12 to 21).

spiratory components. This is supported by the PCO* and base excess values shown in Fig. 2. Groups 1 and 2 had similar PCO, values in fetal scalp and umbilical cord blood throughout labor. However, after early labor, Pcoz levels were significantly higher in group 3 and remained so until delivery. Despite considerably more overlap, the data on base excess in Fig. 2 show a pattern symmetrical to the lactate values in Fig. 1. This suggests that high lactate levels were responsible for the increased metabolic acidosis in groups 2 and 3 as compared to group 1. PO, values in groups 2 and 3 were similar throughout labor, but they were consistently

higher in group 1. This difference was statistically significant in fetal scalp samples at all stages of labor but was not significant in either cord arterial or venous blood samples. Discriminant function analysis. In the discriminant analysis the only two variables selected were, first, lactate and, second, pH. After these were included, no extra discriminating power was achieved by the addition of other variables. Moreover, as can be seen in Tables HA and IIIA, the ability of the best discriminant function to discriminate between group 2 (prepathologic) and group 3 (pathologic) is limited in practice. However, with the use of just two groups, normal (group 1) and abnormal (group 2 plus group 3), we could quite successfully predict a normal outcome (Tables IIB and IIIB). The functions are as follows, where a positive result predicts abnormality and a negative result predicts normality. Early labor: 36.6 + 1.13 X lactate (millimoles per liter) - 5.22 X pH; midlabor: 28.7 + 0.957 x lactate (millimoles per liter) 4.20 x pH. Thus, in early labor (Table IIB), all normal patients were correctly classified, but there was a 50% false negative rate in the prepathologic group and a 33.3% false negative rate in the pathologic group, indicating that these fetuses had a combination of lactate and pH in the normal range. In midlabor, however, the prediction appeared to be improved although the data were more sparse (Table IIIB). In this case only one of

Volume Number

Early

147 8

Table IIIA. Cases allocated function analysis-Midlabor: and group 3

by discriminant Group 1, group

2,

Table IIIB. Cases allocated function analysis-Midlabor: plus group 3

diagnosis

of fetal distress

by discriminant Group 1 and group

953

2

Allocation Actual

group*

Normal

Prepathologic

Normal 14

0

Prepathologic

2 4

Pathologic

Actual

0 0

*These data are derived from patients whose group was known but whose values for fetal lactate and pH had not been used to calculate the discriminant functions used for the allocation.

four prepathologic fetuses was incorrectly classified. These functions compared favorably with the use of pH alone, for which, in early labor, the best discrimination allocated three of the nine normal fetuses as abnormal, five of the 10 prepathologic fetuses as normal, and three of the nine pathologic fetuses as normal. In midlabor, the prediction with the use of pH alone was rather poor, with two of the 16 normal fetuses allocated as abnormal and all four prepathologic fetuses classified as normal. Comment The measurement of lactate, pH, base excess, Pco,, and PO,in blood samples from the fetal presenting part provides much information on the acid-base status of individual fetuses with its metabolic and respiratory components. However, when an attempt is made to distinguish between groups with and without fetal distress as assessed by Apgar score, cardiotocographic tracings, presence of meconium in amniotic fluid, and cord arterial pH at birth, lactate and pH are the parameters with the best discriminating power. Differences in lactate and pH levels between the normal, prepathologic, and pathologic groups were highly significant at all stages of labor, with lactate having the most discriminating power, at least in early labor and midlabor. It is interesting to note that in early labor lactate levels in the prepathologic and pathologic groups were already significantly elevated when compared to levels in the normal group. At the same time, pH levels were significantly decreased in the same groups, but the average values were well above 7.25, which is considered a normal pH by most authors. This confirms the view that in distressed fetuses there is a considerable delay before pH falls below what is usually considered a physiologic leve15* ’ The prospective value of discriminant functions derived from lactate and pH data was assessed in early labor and midlabor by applying these functions to patients belonging to a known group but whose data had not been used to calculate the functions. The dis-

g+roup

Normal

Prepathologic *Abnormal

group

I&-

16 1 = Prepathologic

0

3 group

plus

pathologic

group.

criminating ability of these functions was good when patients were allocated into group 1, since virtually all patients in the normal group were correctly allocated in both early labor and midlabor. However, the discriminating ability was relatively poor in the abnormal group (group 1 plus group 2) with a high false negative rate. The discriminating ability was impaired further when abnormal patients were allocated into separate prepathologic and pathologic groups. Thus, discriminant functions derived from lactate and pH data are useful to predict normal newborn infants, but the practical value of the measurements is limited by the possibility of missing some fetuses with signs of fetal distress. The analysis of the data is made difficult by the lack of a specific end point to which the classification of patients can be referred. It must be emphasized that Apgar score, cardiotocography, meconium staining of the amniotic fluid, and cord pH do not provide a perfect way of determining the presence and severity of fetal distress. For example, the validity of the Apgar score to indicate the presence of fetal distress (as defined by fetal acidosis, i.e., low umbilical arterial pH and base excess) has been questioned in a recent study in which, in a large number of patients, only a relatively small percentage of acidotic babies had a low Apgar score (~7) and, vice versa, a high percentage of acidotic babies had Apgar scores >7.” Similar criticisms have been expressed about the accuracy and reliability of intrapartum cardiotocographic monitoring and pH measurements.“, ‘9 ‘I In our series a high lactate level was usually associated with abnormal cardiotocography, but episodes of abnormal cardiotocography also occurred in fetuses with low levels of lactate. On the other hand, the definition of fetal distress itself is rather vague, since it includes fetal acidosis, usually of metabolic origin, and other hemodynamic, infectious, or traumatic conditions that may result in the birth of a depressed infant.“’ It is, therefore, necessary to establish more specific criteria for the evaluation of fetal outcome before a better selection of patients can be obtained and the predictive value of the biochemical information obtained during birth can be increased.

954

Eguiluz

This

et al.

study

December 15, 1983 Am. J. Obstet. Gynecol.

has confirmed

measurements in blood ing part.4, 12, 13 Lactate lation.’

The

infants

is approximately

slow index

half-life

clearing of fetal

reflected fetal

the

potential

samples is cleared of circulating 35

lactate

hours.*

time, lactate levels hypoxia, although

when within

can

ratio.7

be elevated

in newborn

Because

early

Lactate

in

as opposed

tuations

in Pcop

levels.

ference

found

in

fetomaternal fetal

the

mainly

are not

positive

indicate

represent

although

some

conclusion,

we

as pH, lactate

affected

cord

are con-

by fluc-

and

that lactate

placental

in

at a time

arteriovenous

umbilical

gradient

blood

fetus,

to pH, The

this

levels

labor,

other acid-base parameters, such the “normal” range. Furthermore,

centrations,

of

provide a cumulative this is more accurately

by the lactatelpyruvate blood

use of lactate

from the fetal presentslowly from the circu-

difthe

lactate

high

levels

production

in

by the

contribution

seems

likely.g” In fetal the

acid-base

parameters.

diagnosis

of the

tion, lect

lactate fetuses of

dictive

We value

better

creasing *Eguiluz,

rapid

helps state

intrapartum and improve from other

to establish of the

uncertain

negative

of fetal

outcome.

This

metabolic

is still

false

episodes

ing

that

complement obtained

a better

fetus.

In

addi-

and pH measurements can be used to sein distress. The predictive value of these

measurements ber

believe

lactate measurements biochemical information

distress

are

criteria number

and

to define

not trying

pH

they

necessarily

a numindicate

a bad

fetal

the

pre-

to enhance

measurements fetal

give

they may

and

but

currently

of lactate

the

since

results

outcome

by adoptand

by

in-

of patients.

A.: Unpublished

observations.

REFERENCES 1. Goodlin, R. C.: Fetal cardiovascular responses to distress, Obstet. Gvnecol. 49:371, 197’7. 2. Zuspan, F: P., Quilligan,.E. J., lams, J. D., and Van Geijn, H. P.: Predictors of intrapartum fetal distress: The role of electronic fetal monitoring, AM. J. OBSTET. GYNECOL. X35:287, 1979. 3. Low, J. A., Cox, M. J., Karchmar, E. J.. McGrath, M. J., Pancham, S. R., and Piercy, W. N.: The prediction of intrapartum fetal metabolic acidosis by fetal heart rate mon&oring, AM. J. OBSTET. GYNECOL. kk299, 1981. 4. Yoshioka. T.. and Roux. 1. F.: Correlation of fetal scaln blood p& glucose, lac& and pyruvate concentrado& with cord blood determinations at time of delivery and cesarean section, J. Reprod. Med. 5:209, 1970.

5. Beard, R. W.: The detection of fetal asphyxia in labor, Pediatrics 53:157, 1974. 6. Wood, C.: Diagnostic and therapeutic implications of intrapartum fetal pH measurement, Acta Obstet. GynecoI. Stand. 57:13, 1978. 7. Huckabee, W. E.: Relationships of pyruvate and lactate during anaerobic metabolism. I. Effects of infusion of pyru&e or glucose and of hyperventilation, J. Clin. lnvest. 37:244, 1958. 8. Huckabee, w. E.: Relationships of pyruvate and lactate during anaerobic metabolism. Ill. Effect of breathing low-oxygen gases, J. Clin. Invest. 37:264, 1958. 9. Derom, R.: Anaerobic metabolism in the human fetus, AX J. OBSTET. GYNECOL. 89~241, 1964. 10. Low, J. A., Pancham, S. R., Worthington, D., and Boston, R. W.: Acid-base, lactate, and pyruvate characteristics of the normal obstetric patient and fetus during the intrapartum period, AM. J. OBSTET. GYNECOL. 120:862, 1974. 11. Gordmark, S., Gennser, G., Jacobson, L., Rooth, G., and Thorell, J.: Influence on fetal carbohydrate and fat metabolism and on acid-base balance of glucose administration to the mother during labour, Biol. Neonate 26:129, 1975. 12. Fioretti, P., Bonzani, A., and Rondinelli, M.: L’acido lattico e l’acido piruvico ematici fetali nel travaglio di parto normale, Riv. Ital. Ginecol. 51:299, 1967. 13. Schmid, J.: Glukose, Laktat und Pyruvat in der Schwangerschaft und unter der Geburt, Fortschr. Geburtshilfe Gynaekol. 50:1, 1973. 14. Racine, P., Engelhardt, R., Higelin, J. C., and Mindt, R.: An instrument for the rapid determination of L-lactate in biological fluids, Med. lnstrum. 9:11, 1975. 15. Smith, N., Quinn, M., Soutter, W., and Sharp, F.: Rapid whole blood lactate measurement in the fetus and mother during labour, Br. J. Obstet. Gynaecol. 86:251, 1979. 16. Eguiluz, A., Parrilla, J. J., Molina, E., Abad, L., and Lopez Bernal, A.: Blood lactate levels in fetal scalp samples taken intrapartum, in Proceedings of the Blair Bell Research Society, London, England, October 16, 1980. 17. Klecka, W. R.: Discriminant analysis, in Nie, N. H., Hull, C. H., Jenkins, J. G., Steinberger, K., and Bent, D. H., editors: Statistical Package for the Social Sciences, New York, 1975, McGraw-Hill, Inc., pp. 434-467. 18. Saling, E.: Die Blutgasverhaltnisse und der Sgure-BasenHausholt des Feten bei ungestortem Geburtsablauf, Z. Geburtshilfe Gynaekol. 161:262, 1964. 19. Beard, R. W., and Morris, E. D.: Foetal and maternal acid-base balance during normal labour, J. Obstet. Gynaecol. Br. Commonw. 72:496, 1965. 20. Sykes, G. S., Johnson, P., Ashworth, F., Molloy, P. M., Gu, W., Stirrat, G. M., and Turnbull, A. C.: Do Apgar scores indicate asphyxia? Lancet 1:494, 1982. 2 1. Quilligan, E. J.: Monitoring the fetus using fetal acid base status, Clin. Obstet. Gynaecol. 6:309, 1979. 22. Steer, P. J.: Has the expression “fetal distress” outlived its usefulness? Br. 1. Obstet. Gynaecol. 89:690, 1982. 23. Burd, L. I., Jon&, M. D., Jr.; Simmons, M. A., Makowski, E. L., Meschia, G., and Battaglia, F. C.: Placental production and foetal utilisation of lactate and pyruvate, Nature 254710, 1975.