Relationship of Amniotic Fluid Index and Cord Blood Erythropoietin Levels in Small for and Appropriate for Gestational Age Fetuses


Obstetrics & Gynecology:
Original Research

Objective: To determine whether low amniotic fluid index (AFI) reflects an increase in cord blood erythropoietin levels and is a marker of antenatal fetal hypoxia in small for gestational age (SGA) and appropriate for gestational age (AGA) fetuses in the third trimester.

Methods: Erythropoietin levels in cord blood were measured in 134 high-risk obstetric patients, including 40 with low AFI (7 cm or less) and 94 with normal AFI (greater than 7 cm), which were sampled at elective cesarean delivery of fetuses with a gestational age of 32 to 39 weeks. The infants were divided into the SGA and AGA groups based on birth weight. Partial correlations of AFI with cord blood erythropoietin levels were evaluated by Pearson correlation coefficient after logarithmic transformation of erythropoietin levels in each group.

Results: Cord blood erythropoietin levels in SGA fetuses with low AFI (n = 24) were significantly higher than those in SGA fetuses with normal fluid volume (n = 16) (171.6 ± 207.4 mIU/dL compared with 36.1 ± 24.1 mIU/dL, P < .001). Conversely, cord blood erythropoietin levels in AGA fetuses with low AFI (n = 16) were not significantly different than those in AGA fetuses with normal fluid volume (n = 78) (32.1 ± 18.7 mIU/dL compared with 29.5 ± 15.3 mIU/dL). A significant partial correlation between AFI and erythropoietin levels was demonstrated only within the SGA group (P < .001, r = 2.67).

Conclusion: Low AFI could indicate the degree of antenatal fetal hypoxia in SGA fetuses. The impact of a reduced amniotic fluid volume on antenatal fetal condition might be less severe in AGA fetuses than in SGA fetuses.

Assessment of amniotic fluid (AF) volume has been an important component in antepartum fetal testing in high-risk populations, eg, for the biophysical profile1 and modified biophysical profile,2 based on the concept that reduced AF volume reflects relatively long-term placental insufficiency. Oligohydramnios is a common phenomenon in fetal growth restriction (FGR) and post-term pregnancy, both of which have been associated with increased perinatal morbidity and mortality. Many studies sought to determine the cause of oligohydramnios from the perspective of chronic fetal hypoxemia.3–6 Although oligohydramnios has been advocated as an indication for delivery regardless of FGR in recent obstetric treatment,7 the possible implications of reduced AF volume as an independent marker for antenatal fetal compromise are still unknown. Third-trimester isolated oligohydramnios together with normal fetal growth without any structural anomalies has not been well investigated to determine whether it might be related to antenatal hypoxia. Furthermore, it is not known whether the severity of oligohydramnios is associated with the degree of fetal hypoxia.

Increased cord blood erythropoietin levels, which stimulate fetal erythropoiesis during hypoxia, have been reported to indicate chronic intrauterine hypoxia in FGR,8 post-term pregnancy,9 and pregnancies with maternal smoking.10 Tissue hypoxia is the only known stimulus of erythropoietin production in which an increase is observed at least several hours after an initial exposure to hypoxia.11 A previous study12 reported that cord blood erythropoietin levels at delivery might be elevated even in response to spontaneous labor. Therefore, cord blood samples in nonlaboring subjects should be used to investigate chronic hypoxia in utero. In the current study, we evaluated erythropoietin levels in umbilical cord blood sampled at cesarean delivery before the onset of labor as an index of antepartum fetal oxygenation to determine indirectly whether the amniotic fluid index (AFI) correlates with the degree of chronic hypoxia in small for gestational age (SGA) and appropriate for gestational age (AGA) fetuses.

In Brief

Low amniotic fluid index correlates with the degree of elevation in cord blood erythropoietin levels in small for gestational age fetuses but not in appropriate for gestational age fetuses.

Author Information

Department of Obstetrics and Gynecology, Chiba University School of Medicine, 1-8-1 Inohana, Chuo-ku, Chiba City, Chiba, Japan.

Address reprint requests to: Shigeharu Doi, MD, Department of Obstetrics and Gynecology, Chiba University School of Medicine, 1-8-1 Inohana, Chuo-ku, Chiba City, Chiba, 260-0856, Japan; E-mail:

Received January 4, 1999. Received in revised form May 11, 1999. Accepted May 21, 1999.

Article Outline
Back to Top | Article Outline

Materials and Methods

One hundred eighty-two cord blood samples were collected consecutively from high-risk obstetric patients at elective cesarean delivery at the Division of Obstetrics and Gynecology of Chiba University Hospital between April 1, 1993, and September 30, 1998. This study was approved by the Chiba University Hospital Ethics Committee. All subjects were Japanese women with singleton pregnancy who had at least one medical complication or who developed a complication during pregnancy. Women with diabetes (n = 13), preterm ruptured membranes (n = 6), major fetal anomalies (n = 15), congenital viral infection (n = 1), or chromosomal abnormalities (n = 2) were excluded from the study because of the possibility of these factors affecting variables such as birth weight and AF volume. Deliveries before 32 weeks' gestation (n = 11) were also excluded because of unsuccessful blood sampling at delivery for subsequent analysis. A total of 134 subjects whose gestation was 32 to 39 weeks and who had preeclampsia or chronic hypertension (n = 56), renal disease (n = 12), hyperthyroidism (n = 9), collagen disease (n = 25), cardiac disease (n = 7), cerebrovascular disease (n = 4), severe fetal growth restriction (FGR) diagnosed by ultrasound (n = 23), previous fetal death from unknown causes (n = 9), or advanced maternal age (n = 18) were enrolled in this study. Among the maternal indications for cesarean were severe preeclampsia with uncontrolled blood pressure, worsening of renal function, pulmonary edema or visual symptoms, placenta previa, previous cesarean delivery, and obstructive huge uterine fibroid. The fetal indications were severe FGR with no interval growth, nonreassuring fetal testing, breech presentation, and cephalopelvic disproportion. In all cases, cesarean was done before the onset of labor and there was no evidence of acute asphyxial insults before or during the operation.

Amniotic fluid index was determined by the four-quadrant technique13 of ultrasonography by a single ultrasonographer. Measurement of AFI was done at least twice within 3 days before the cesarean, and mean AFI was used in this study. It is difficult to define reduced or low AFI as a single value through 32–39 weeks' gestation because of the physiologic decrease in AF volume with advancing gestational age, and there is disagreement on the definition of oligohydramnios. In this study, low AFI was defined as the cutoff value equal to or less than 7 cm. This cutoff value is lower than the fifth percentile of AFI measurement to the gestational week–specific nomogram for near-term and term gestation.14 The intraobserver variability, which was previously tested in 30 cases of oligohydramnios, was approximately 14%.15 Small for gestational age is defined as less than the tenth percentile of birth weight in the Japanese population.16 Of the 134 high-risk patients, low AFI occurred in 40, and the remaining 94 served as the normal AFI group. They were further divided into SGA and AGA groups based on birth weight.

At delivery, umbilical arterial blood was taken by needle puncture for acid-base analysis (Blood Gas Analyzer 288; Ciba Corning, Tokyo, Japan), and umbilical venous blood was collected into heparinized syringes. After centrifugation, plasma was stored at −20C. Cord blood erythropoietin levels were measured in duplicate by a double-antibody radioimmunoassay technique.17 The intraassay coefficient of variation was approximately 8% and interassay coefficient of variation was about 10%.

Statistical analysis was done using Fisher exact test for proportional data and Kruskal-Wallis test followed by Dunn test for comparisons between groups. Correlation between erythropoietin levels with AFI, birth weight, and gestational age at delivery were assessed using Pearson correlation coefficient after logarithmic transformation of erythropoietin levels in the SGA and AGA groups. P < .05 was considered statistically significant.

Back to Top | Article Outline


The clinical characteristics of the patients are presented in Table 1. There were no significant differences in maternal age, parity, and gestational age at delivery among the four groups. In the AGA group, but not in the SGA group, the rate of preeclampsia was significantly higher in women with low AFI than in those with normal AFI (P < .001). Cord arterial pH was lower in SGA fetuses with low AFI compared with SGA fetuses with normal AFI (7.22 ± .08 compared with 7.29 ± .05, P < .01), but not in the AGA group. Cord arterial base deficit was also significantly lower in SGA fetuses with low AFI compared with SGA fetuses with normal AFI (−5.8 ± 3.1 compared with −3.1 ± 2.1, P < .05), but not in the AGA group. Birth weights were not significantly different in the SGA group, but they were significantly lower in AGA fetuses with low AFI than in AGA fetuses with normal AFI (2521 ± 358 g compared with 2882 ± 345 g, P < .01). Erythropoietin levels in cord blood were significantly higher in SGA fetuses with low AFI than in SGA fetuses with normal AFI (171.6 ± 207.4 mIU/dL compared with 36.1 ± 24.1 mIU/dL, P < .001). There were no significant differences in erythropoietin levels in the AGA groups.

Figures 1 and 2 show the logarithm of erythropoietin levels in relation to AFI in the SGA and AGA fetuses, respectively. In the SGA group, there was a significant correlation between log (erythropoietin level) and AFI, as well as birth weight (r = −.68, P < .001; r = −.55, P < .01, respectively), but not between log (erythropoietin level) and gestational age at delivery (P = .09). After adjusting for gestational age, partial correlation was still significant between log (erythropoietin level) and AFI in the SGA group, (r = −.67, P < .001). Conversely, there were no significant correlations between log (erythropoietin level) and any other variables in the AGA group.

Back to Top | Article Outline


The strong relationship between ultrasonographic estimates of oligohydramnios in high-risk subjects and antenatal fetal acidosis was previously reported by Vintzileos et al18 and Manning et al19 as determined by cord blood sampling at elective cesarean delivery and antenatal funicentesis, respectively. Both studies found a high incidence (46–53%) of fetal acidosis in cases of severe oligohydramnios. However, none of the reported studies compared the significance of oligohydramnios on the fetal condition, by taking the differences between SGA and AGA fetuses into account. This distinction between the two groups seems to be important from the perspective of previous hypoxia as a cause of reduced AF volume. In this study, cord blood erythropoietin levels correlated inversely with AFI in the SGA group but not in the AGA group. This finding suggests that low AFI might indicate the degree of in utero hypoxia in SGA fetuses and might forecast FGR resulting from chronic hypoxia among SGA fetuses. The significance of low AFI in AGA fetuses as a measure of chronic hypoxia still remains questionable. The impact of reduced AF volume on antenatal fetal condition might be less severe in AGA fetuses than in SGA fetuses.

There are few reports of perinatal outcomes regarding isolated oligohydramnios without either FGR or structural anomalies in the third trimester. Recently, Garmel et al20 reported that normal-size fetuses with oligohydramnios were not at increased risk of adverse outcome. Although our results cannot be attributed to intrapartum events associated with low AFI, they do not conflict with their findings. Uteroplacental insufficiency that causes blood flow redistribution away from nonvital organs such as the kidney has been proposed as a major process of oligohydramnios, but causes of reduced AF volume, except for fetal structural malformations, include not only hypoxia but also maternal dehydration.21 In our study, preeclampsia, which is a typical model of maternal intravascular fluid depletion, occurred in 50% of AGA fetuses with low AFI. The birth weights of those fetuses were also significantly lower than those of AGA fetuses with normal AFI. Recently, Bar-Hava et al5 reported that the cause of oligohydramnios in post-term pregnancy was related to lower birth weight, not to redistribution of fetal blood flow, as determined by Doppler echocardiography. This finding as a possible cause of reduced AF volume also does not conflict with our results with AGA fetuses with low AFI. It is unknown whether these fetuses would have had FGR or postdates if cesarean delivery had not been done. However, fetal oxygenation might still have been well maintained in these fetuses despite the low AFI, based on the lack of elevation in cord blood erythropoietin levels. We found that the mean cord blood erythropoietin level in AGA fetuses with low AFI was only 9% higher than that in AGA fetuses with normal AFI. Although it is not clear what constitutes a clinically significant difference, our study had a power of 80% at one-sided α of .05 to detect a difference of 30% between groups.

Alternatively, it could be argued that the low AFI does not reflect the actual reduced AF volume by the semiquantitative technique, as demonstrated by the broad range of actual AF volume for any criteria of ultrasonographic measures.22,23 This inaccuracy could provide one reason for the disparity found in predicting adverse outcome using AFI, with some studies showing an increased risk in oligohydramnios24–26 and others showing no difference between oligohydramnios and normal fluid volume.27–29 However, no ultrasonographic methods of assessing AF volume available to date have been consistently accurate.

Back to Top | Article Outline


1. Manning FA, Morrison I, Lange IR, Harman CR, Chamberlain PF. Fetal assessment based on fetal biophysical profile scoring: Experience in 12,620 referred high-risk pregnancies. I. Perinatal mortality by frequency and etiology. Am J Obstet Gynecol 1985;151:343–50.
2. Nageotte JP, Towers CV, Asrat T, Freeman RK. Perinatal outcome with the modified biophysical profile score. Am J Obstet Gynecol 1994;170:1672–6.
3. Nicolaides KH, Peters MT, Vyas S, Rabinowitz R, Rosen DJD, Campbell S. Relation of rate of urine production to oxygen tension in small for gestational age fetuses. Am J Obstet Gynecol 1990;162:387–91.
4. Campbell S, Wladimiroff JW, Dewhurst CJ. The antenatal measurement of fetal urine production. Br J Obstet Gynaecol 1973;80:680–6.
5. Bar-Hava I, Divon MY, Sardo M, Barnhard Y. Is oligohydramnios in postterm pregnancy associated with redistribution of fetal blood flow? Am J Obstet Gynecol 1995;173:519–22.
6. Cock ML, McCrabb GJ, Wlodek ME, Harding R. Effects of prolonged hypoxemia on fetal renal function and amniotic fluid volume in sheep. Am J Obstet Gynecol 1997;176:320–6.
7. Manning FA. Amniotic fluid: Genesis, regulation, composition, and clinical application. In: Manning FA, ed. Fetal medicine: Principles and practice. 1st ed. Norwalk, Connecticut: Appleton & Lange, 1995:204.
8. Snijders RJ, Abbas A, Melby O, Ireland DM, Nicolaides KH. Fetal plasma erythropoietin concentration in severe growth retardation. Am J Obstet Gynecol 1993;168:615–9.
9. Jazayeri A, Tsibris JCM, Spellacy W. Elevated umbilical cord erythropoietin levels in prolonged pregnancies. Obstet Gynecol 1998;92:61–3.
10. Jazayeri A, Tsibris JCM, Spellacy W. Umbilical cord plasma erythropoietin levels in pregnancies complicated by maternal smoking. Am J Obstet Gynecol 1998;178:433–5.
11. Clemons GK, Fitzsimmons SL, DeManincor D. Immunoreactive erythropoietin concentrations in fetal and neonatal rats and the effects of hypoxia. Blood 1986;68:892–9.
12. Widness JA, Phillips AF, Clemons GK, Garcia JF, Oh W, Schwartz R. Increased immunoreactive erythropoietin in cord serum after labor. Am J Obstet Gynecol 1984;148:194–7.
13. Phelan JP, Smith CV, Broussard P, Small M. Amniotic fluid volume assessment with four-quadrant technique at 36–42 weeks' gestation. J Reprod Med 1987;32:540–2.
14. Moore TR, Cayle JE. The amniotic fluid index in normal human pregnancy. Am J Obstet Gynecol 1990;162:1168–73.
15. Doi S, Osada H, Seki K, Sekiya S. Effect of maternal hydration on oligohydramnios: A comparison of three volume expansion methods. Obstet Gynecol 1998;92:525–9.
16. Sato A, Akama M, Yamanobe H, Hoshi K, Suzuki M. Intrauterine growth of live-born Japanese infants between 28 and 42 weeks of gestation. Acta Obstet Gynecol Jpn 1982;34:1535–8.
17. Garcia JF, Ebbe SN, Hollander L, Cutting HO, Miller ME, Cronkite EP. Radioimmunoassay of erythropoietin: Circulating levels in normal and polycythemic human beings. J Lab Clin Med 1982;99:624–35.
18. Vintzileos AM, Lettier L, Tsapanos V, Campbell WA. The relationship between combined fetal biophysical activities, oligohydramnios, and fetal acid base status. J Ultrasound Med 1993;12:519–22.
19. Manning FA, Snijders R, Harman CR, Nicolaides K, Menticoglou S, Morrison I. Fetal biophysical profile score. Correlation with ante-partum umbilical venous fetal pH. Am J Obstet Gynecol 1993;169:755–63.
20. Garmel SH, Chelmow D, Sha SJ, Roan JT, D'Alton ME. Oligohydramnios and the appropriately grown fetus. Am J Perinatol 1997;14:359–63.
21. Goodlin RC, Anderson JC, Gallagher TF. Relationship between amniotic fluid volume and maternal plasma volume expansion. Am J Obstet Gynecol 1983;146:505–11.
22. Magann EF, Nolan TE, Hess LW, Martin RW, Whitworth NS, Morrison JC. Measurement of amniotic fluid volumes: Accuracy of ultrasonographic techniques. Am J Obstet Gynecol 1992;167:1533–7.
23. Chauhan SP, Magann EF, Morrison JC, Whitworth NS, Hendrix NW, Devoe LD. Ultrasonographic assessment of amniotic fluid does not reflect actual amniotic fluid volume. Am J Obstet Gynecol 1997;177:291–7.
24. Rutherford SE, Phelan JP, Smith CV, Jacob N. The four-quadrant assessment of amniotic fluid volume: An adjunct to antepartum fetal testing. Obstet Gynecol 1987;70:353–60.
25. Craigo SD, Beach ML, Harvey-Wilkes KB, D'Alton ME. Ultrasound predictors of neonatal outcome in intrauterine growth restriction. Am J Perinatol 1996;13:465–71.
26. Sarno JP Jr, Ahn MO, Phelan JP. Intrapartum amniotic fluid index at term: Association of ruptured membranes, oligohydramnios and increased risk. J Reprod Med 1990;35:719–23.
27. Robson SC, Crawford RA, Spencer JA, Lee A. Intrapartum amniotic fluid index and its relationship to fetal distress. Am J Obstet Gynecol 1992;166:78–82.
28. Chauhan SP, Cowan BD, Magann EF, Roberts WE, Morrison JC, Martin JN Jr. Intrapartum amniotic fluid index. A poor diagnostic test for adverse perinatal outcome. J Reprod Med 1996;41:860–6.
29. Schucker JL, Mercer BM, Audibert FA, Lewis RL, Friedman SA, Sibai BM. Serial amniotic fluid index in severe preeclampsia: A poor predictor of adverse outcome. Am J Obstet Gynecol 1996;175:1018–23.
© 1999 The American College of Obstetricians and Gynecologists