- Email: [email protected]
The relationship between umbilical artery Doppler velocimetry and fetal biometry
Meconium and birth asphyxia
Volume 165 Number 4, Part 1
plicated term vaginal deliveries. AM ] OBSTET GVNECOL 1985; 151 :798-80 1. 18. Thorp ]A, Sampson ]E, Parisi VM, Creasy RK. Routine umbilical cord gas determinations. AM] OBSTET GVNECOL 1989;161:600-5. 19. Gilstrap LC, Leveno K], Burris], Williams ML, Little BB. Diagnosis of birth asphyxia on the basis of fetal pH, Apgar score, and newborn cerebral dysfunction. AM] OBSTET GVNECOL 1989; 161 :825-30.
20. Freeman ]M, Nelson KB. Intrapartum asphyxia and cerebral palsy. Pediatrics 1988;82:240-4. 21. Weitzner ]S, Strassner HT, Rawlins RG, et al. Objective assessment of meconium content of amniotic fluid. Obstet Gynecol 1990;76: 1143-4. 22. Yeomans ER, Gilstrap LC, Leveno K], Burris ]S. Meconium in the amniotic fluid and fetal acid-base status. Obstet Gynecol 1989;73: 175-8.
The relationship between umbilical artery Doppler velocimetry and fetal biometry William E. Scorza, MD, Deborah Nardi, RT(R), RDMS, Anthony M. Vintzileos, MD, Alfred D. Fleming, MD, John F. Rodis, MD, and Winston A. Campbell, MD Farmington, Connecticut The relationship between peak-systolic/ end-diastolic ratio of the umbilical artery waveform and fetal biometry was studied in 127 uncomplicated pregnancies with established dates between 20 and 40 weeks' gestation. At each ultrasonographic examination fetal biometry included measurement of the biparietal diameter, head circumference, abdominal circumference, and femur length. The peak-systolic/end-diastolic ratio was measured by either a continuous or a pulsed-wave method. There were significant linear negative correlations between all the biometric parameters, as well as between the ultrasonographically estimated fetal weight and peak-systolic/end-diastolic ratio. Of the individual ultrasonographic parameters the femur length (for gestations <30 weeks) and the abdominal circumference (for gestations 2:30 weeks) were found to be best correlated with the peak-systolic/end-diastolic ratio. Regression curves, including the 10th and the 90th percentile, were developed between each biometric parameter (biparietal diameter, head circumference, abdominal circumference, and femur length), as well as between estimated fetal weight and peak-systolic/end-diastolic ratio. The estimated fetal weight nomogram had the best sensitivity (48%) in predicting intrauterine growth retardation. These nomograms should prove most useful in assessing downstream placental vascular resistance in high-risk patients with unknown dates. (AM J OBSTET GVNECOL 1991 ;165:1013-9.)
Key words: Umbilical artery velocimetry, fetal biometry Continuous or pulsed-wave Doppler ultrasonography of the fetal umbilical artery has been increasingly used in clinical obstetrics in the management of highrisk pregnancy. I·' The measurement of the peak-systolic/ end-diastolic ratio of the umbilical artery flow velocity waveform is the most commonly used method. Abnormal ratios have been associated with an increased incidence of neonatal morbidity, mortality, and intrauterine growth retardation. 1 . 5 The abnormal waveFrom the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Connecticut Health Center. Presented at the Eleventh Annual Meeting of the Society of Perinatal Obstetricians, San Francisco, California, january 28-February 2, 1991. Reprint requests: William E. Scorza, MD, Department of Obstetrics and Gynecology, Division of Maternal-Fetal Medicine, Mount Sinai Hospital, 500 Blue Hills Ave., Hartford, CT 06112. 6/6/30898
forms are believed to be the result of increased downstream placental vascular resistance and therefore of impaired placental perfusion. 6 The proper use of the umbilical artery peak-systolic/ end-diastolic ratio, however, requires knowledge of the length of gestation, which is unknown or unavailable in 20% to 40% of pregnant patients. 7 When the last menstrual period is unknown or uncertain, extrapolation of gestational age on the basis of fetal biometric data may lead to erroneous interpretation of the peak-systolic/ end-diastolic ratio results, because there are no data regarding the relationship between the umbilical artery peak-systolic/ end-diastolic ratio and fetal biometry. Therefore the purpose of the present report was: (1) to study the relationships and generate regression curves between the peak-systolic/ end-diastolic ratio and the most commonly used fetal biometric parameters (biparietal di1013
1014 Scorza et al.
October 1991 Am J Obstet Gyneco1
5.5
y=-.185X+4.132
5.0
.5! iii...
4.5 4.0
0
en >...
.!"-
III III
u
•
3.5
~
• •
3.0
••
2.5
• •
J:I
E
•
2.0
I
••
•
•
•
• • •
1.5 1.0 5
7
6
8
9
10
Biparietal diameter(cm)
Fig. 1. Scattergram illustrating relationship (along with 10th, 50th, and 90th percentile lines) between umbilical artery peak-systolic lend-diastolic ratio (SID) and biparietal diameter (R = 0.395, P = 0.0001).
ameter, head circumference, abdominal circumference, femur length, and estimated fetal weight); (2) to assess which fetal biometric parameters, if any, are significantly correlated with the peak-systolic/ end-diastolic ratio; and (3) to determine the sensitivity of the fetal biometric parameter nomograms to predict intrauterine growth retardation (IUGR). Material and methods Patients were recruited from those referred to our antepartum fetal evaluation unit at the University of Connecticut Health Center. We used 127 uncomplicated pregnancies between 20 and 40 weeks of gestation (dates confirmed by early ultrasonography) for a single ultrasonographic assessment. The patients were referred for confirmation of clinical dates or fetal size. Patients were selected if they had initiated prenatal care early in pregnancy, if they had regular menstrual cycles with reliable menstrual dates, and if the results of ultrasonographic examination were consistent with their menstrual dates. All patients had intact membranes and singleton pregnancies. Patients were excluded if fetal anomalies were detected or if the estimated fetal weight was < - 2 SD or >2 SD from the mean for gestational age. s The ultrasonographic examinations were performed, using a model 128 (Acuson, Mountain View, Calif.) or a model 3000 or 2600 (General Electric, Rancho Cordova, Calif.). A 3.5 or 5 MHz linear-array trans-
ducer was used. All machines had freeze-frame capabilities and electronic calipers for measurement. The ultrasonographic velocity was 1540 m/sec. The study was cross-sectional, and each patient was included only once. Routine measurements (biparietal diameter, head circumference, abdominal circumference, and femur length) were obtained and recorded on each patient. The biparietal diameter and head circumference measurements were obtained from an axial scanning plane at the level of the thalami. The fetal abdominal circumference was taken perpendicular to the long axis of the fetal body at the level of the ductus venosus complex. The abdominal circumference was measured directly with electronic calipers or predominantly by using the formula Dl + D2 X 1.57, where D is diameter (both methods have been shown to give equivalent results).9 The femur length was measured according to the technique described by O'Brien and Queenan. 10 The ultrasonographic estimated fetal weight was calculated by averaging the estimated fetal weight obtained by both Shepard's (using biparietal diameter and abdominal circumference)" and Hadlock's (using abdominal circumference and femur length)l2 formulas. The umbilical artery peak-systolic/ end-diastolic ratio was obtained transabdominally by continuous (model 5000A, Multigon, Mount Vernon, N.Y.) or pulsed-wave (model 128, Acuson) Doppler ultrasonography. All studies were performed during fetal apnea. The peak-systolic/ end-diastolic ratios of three waveforms from three
Umbilical artery velocimetry and fetal biometry
Volume 165 Number 4 , Pan I
1015
5.5 y=-.059X+4.364
5.0 ~
iii
4.5
~
0
4 .0
>-
3.5
in ~
!
~
as
3.0
ii u
o
2.5
.Q
E
~
. ....
:
0
00
2.0
00
..-.
1.5 1.0 20
22
24
26
28
30
32
34
Head clrcumference(cm)
Fig. 2. Scattergram illustrating relationship (along with 10th, 50th, and 90th percentile lines) between umbilical artery peak-systolic I end-diastolic (SID) ratio and head circumference (r = 0.422, P = 0.0001).
different sampling sites were computed, and the mean ratio was used in the analysis. Scattergrams of the peak-systolic! end-diastolic ratio values versus gestational age and fetal biometry were developed. Regression analyses were used to establish nomograms, including the 10th, 50th, and 90th percentile lines. To assess which fetal biometric para meteres), if any, best correlated with the peak-systolic/ enddiastolic ratio, multiple regression analyses and stepwise regression analyses were used. A p value < 0.05 was considered statistically significant. The statistical analysis of the data was performed by computer (Macintosh II) with the Stat View 512 + (Abacus Concepts, Berkeley, Calif.) and CricketGraph (CricketGraph, Malvern, Pa.) statistical packages. During the study period there were 25 patients with known dates (based on last menstrual period and confirmed by early ultrasonography) who had growthretarded fetuses (estimated fetal weights < - 2 SD from the mean for gestational age).s The gestational ages of these fetuses ranged from 26 to 38 weeks. The fetuses were used to determine the sensitivity of the newly constructed nomograms on the basis of fetal biometry, to predict growth retardation. In all 25 fetuses the diagnosis of IVGR was confirmed after birth.
Results The study population consisted of 127 patients with gestational age (mean ± SD) of 31 ± 4 weeks (median,
31.1). Of these , 104 (81 %) were white, 12 (10%) black, and 11 (9%) Hispanic. Seventy patients (55 %) were nulliparous and 57 (45%) were multiparous. The maternal age (mean ± SD) was 26.3 ± 5.6 years (range, 15 to 40). The best-fit curves between peak-systolic/ enddiastolic ratio and biparietal diameter, head circumference, abdominal circumference, femur length, and estimated fetal weight were linear. The r values were 0.395 , 0.422,0.461,0.453, and 0.460, respectively (all p values were = 0.0001). The scattergrams, including the 10th, 50th, and 90th percentile boundaries, and the regression equations are illustrated in Figs. 1 to 5 . The estimated fetal weight had the highest linear negative correlation with the peak-systolic/ end-diastolic ratio (r = 0.461). The best-fit curve between the peaksystolic/ end-diastolic ratio and gestational age also was linear (Fig. 6). To determine the origins and to assess which fetal biometric parameters contribute to this negative correlation between estimated fetal weight and the peaksystolic/ end-diastolic ratio, the data were analyzed according to two gestational age groups, <30 weeks' gestation (46 patients) and ~30 weeks' gestation (81 patients). In both gestational age groups there was a significant linear negative correlation between estimated fetal weight and the peak-systolic! end-diastolic ratio (p < 0.05 and p < 0.0005, respectively). To assess which ultrasonographic parameter(s) significantly correlated with the peak-systolic/ end-diastolic ratio within
1016
Scorza et al.
October 199 1 Am J Obstet Gynecol
5.5
y=-.05X+4.05
5.0
.2 iii ...
c in >...
!... CD
iii
4.5 4.0 3.5 3.0
u
2.5
:0 E
2.0
~
•
•
••
•
•
•
•
." .
1.5
15
19
17
23
21
29
27
25
31
33
35
Abdominal clrcumference(cm) Fig. 3. Scattergram illustrating relation ship (along w ith 10th , 50th, and 90th perce ntile lines) between umbilical artery peak·systolic/end-diastolic (S I D) ratio and abdominal circumference (r = 0 .460. P = 0.000 1).
5.5 y=-.235X+4.089
5.0 0
:;::
...
4.5
0
4.0
...>...
3.5
III
en ~
III
iii (,,)
•
E
•
3.0 2.5
••
.Q ~
•
• •• •• • •
•
•
2.0
•••
•
• •••
•
I
. . • •• .. - •.
1.5 1.0 4 .0
4 .5
5.5
5.0
6.0
6.5
7.0
7 .5
Femur lenllth(cm) Fig. 4. Scattergram illustrating relationship (along with 10th, 50th, a nd 90th percentile lines) betwee n umbilical artery peak-systolic I end-diastolic (SID ) ratio and femur length (r '" 0.453, P '" 0.0001 ).
each gestational age group. multiple regression analyses and stepwise regression analyses were used with the peak-systolic lend-diastolic ratio used as the dependent (outcome) variable and biparietal diameter, head
circumference, abdominal circumference, and femur length used as the inde pendent (predictors) variables. Multiple regression analysis o f he t group < 30 weeks' gestation revealed that the overall association was sig-
Umbilical artery velocimetry and fetal biometry
Volume 165 Number 4, Part I
5.5
y=-3.020E-4x+3.234
5.0
..... 0
en >... ... III
..
+
+ +
4.5
III
c
1017
4.0
+ + •
+
+ •
3.5
G)
III
3.0
.!:!
2.5
.c E
2.0
:)
•
.
•
••
1.5 1.0 500
1000
1500
2000
2500
3000
3500
4000
Estimated fetal weight (in gms) Fig. 5. Scattergram illustrating relationship (along with 10th, 50th, and 90th percentile lines) between umbilical artery peak-systolic I end-diastolic (SID) ratio and estimated fetal weight (r = 0.461, P = 0.0001). Plus sign, Growth-retarded fetuses. 5.5 y=-.045X+4.115 5.0 0
~ c
en ~
~ as
as
u
4.5 4.0
•
3.5 3.0
.c
E
:)
•
..
22
24
2.5 2.0
...
. . ... .,- .- ..-...•-.•-... ...... • •
• • •
1.5
20
26
28
30
32
34
36
38
40
Gestational age (weeks) Fig. 6. Scattergram illustrating relationship (along with 10th, 50th, and 90th percentile lines) between umbilical artery peak-systoliclend-diastolic ratio and gestational age (r = 0.391, P = 0.0001).
nificant (p < 0.05). Of th~ four independent variables, only femur length was significantly correlated with the peak-systolic! end-diastolic ratio (p < 0.05; 13 coefficient, -0.477; standard error, 0.211; standard coeffi-
cient, - 0.673; t value, 2.262). Stepwise regression analysis of this group of fetuses of <30 weeks' gestation also indicated that the femur length was the only independent variable to enter the equation (r = 0.477,
1018 Scorza et al.
W = 0.227,adjustedR 2 ;= 0.21). Similar analyses were performed for the group of fetuses with gestational ages ~30 weeks. With multiple regression analysis it was found that the overall correlation of the independent variables to the peak-systolic! end-diastolic ratio was significant (p < 0.0005). Of the individual independent variables only abdominal circumference was significantly associated with the peak-systolic/ end-diastolic ratio (p < 0.005; 13 coefficient, - 0.086; standard error, 0.03; standard coefficient, - 0.534; t value, 2.896). Stepwise regression analysis also was performed, and only the abdominal circumference entered the equation (r = 0.525, R2 = 0.276, adjusted R2 = 0.267). The same analyses were repeated with gestational age included as one of the independent variables, but the results were the same. Even when gestational age was used as one of the independent variables, only femur length for fetuses <30 weeks' gestation and only abdominal circumference for fetuses ~30 weeks' gestation again were the only variables significantly correlated with the peak-systolic/ enddiastolic ratio. With the established nomograms between the peaksystolic! end-diastolic ratio and the fetal biometric parameters, the sensitivity of each parameter to detect IUGR was calculated. The gestational age of the fetuses with IUGR was 32.7 (± 3.6) weeks (median, 33). Two of the fetuses had absent end-diastolic velocities and were excluded from calculations of sensitivity. The sensitivity of detecting growth retardation by using the biparietal diameter nomogram was 35% (8/23), head circumference 35% (8/23), abdominal circumference 35% (8/23), femur length 30% (7/23), and estimated fetal weight 48% (11123). Of these ultrasonographic parameters, the estimated fetal weight versus the peaksystolic/ end-diastolic ratio had the best sensitivity. The sensitivity by using the gestational age versus the peaksystolic/end-diastolic ratio nomogram was 52% (12/23).
Comment In 1978 McCallum et al. 13 suggested that umbilical artery velocity waveforms could be recorded by ultrasonographic Doppler probes. Subsequently various investigators suggested that the calculation of peak-systolic/ end-diastolic ratio of the umbilical artery velocity waveform reflects placental vascular resistance. 6 Elevated ratios have been associated with an increased incidence of adverse perinatal outcome, such as growth-retarded infants who are prone to hypoxic complications, twin-to-twin transfusion, and preeclampsia'" 5.14. 15 It has been clearly shown from previous investigations that there is a steady decline in the peak-systolic/ end-diastolic ratio with advancing gestational age,16, 17 the length of gestation should be known for the accurate interpretation of the ratio result. How-
October 1991
Am
J Obslet Gynccol
ever, the last menstrual period is unknown or uncertain in approximately 20% to 40% of pregnant patients. 7 This is the first report to evaluate the relationships and establish nomograms between each of the most commonly used ultrasonographic parameters (biparietal diameter, head circumference, abdominal circumference, femur length) and the estimated fetal weight and peak-systolic/ end-diastolic ratio. Statistically significant relationships were found between the ratio and each of the ultrasonographic parameters as well as the estimated fetal weight. There was an inverse relationship between the ratio and each ultrasonographic parameter. The use of these gestational age-independent nomograms may prove useful for evaluating placental resistance in patients with unknown dates or late registrants with uncertain dates. In the 23 fetuses with IUGR encountered during the study period, the sensitivity of the fetal biometry versus the peak-systolic/end-diastolic ratio nomograms to detect IUGR was not significantly different from that with the gestational age versus the peak-systolic/ end-diastolic ratio nomogram. The best sensitivity was achieved by the estimated fetal weight versus the ratio nomogram (48%), which compared quite favorably with the sensitivity obtained by using gestational age (52%). The detection of an abnormally high peak-systolic/ end-diastolic ratio in growth-retarded fetuses depends on whether small fetal size is the result of impaired uteroplacental blood flow or non placental causes. Furthermore, even with utero placental insufficiency the ratio still may be normal in some cases. Therefore it is not surprising that the sensitivity of detecting IUGR by our method is approximately 50% because many of the fetuses that weigh < 10th percentile for gestational age may suffer lesser degrees of uteroplacental insufficiency or the low weight may be the result of non placental causes. Of interest were the results of the multiple and stepwise regression analyses that used the peak-systolie/end-diastolic ratio as the dependent (outcome) variable and the remaining ultrasonographic biometric parameters as the independent variables. One of the most interesting results of this study was the only significant correlation of femur length to the peak-systolic/ enddiastolic ratio for fetuses <30 weeks and of abdominal circumference to peak-systolic/ end-diastolic ratio for fetuses ~30 weeks. These findings persisted even when the gestational age was included as one of the independent variables. The explanation for these findings lies in the fact that femur length is an indirect measure of fetal length and abdominal circumference is an indirect measure of fetal body weight. It is already known that the fetus exhibits two growth spurts, one in fetal length during the second trimester and the other in body weight during the third trimester. 18, 19 A breakpoint of 30 weeks was chosen in the analysis of the data because, as has been previously shown,18. 19 the fetal
Volume 165 Numbe r 4, Part 1
growth spurt in length has been almost completed while the growth spurt in weight has just begun by 30 weeks of gestation. Therefore it is not surprising to find a significant correlation between femur length, which represents fetal length, and the peak-systolic/ end-diastolic ratio at gestational ages <30 weeks and between abdominal circumference, which represents fetal body weight, and the peak-systolic / end-diastolic ratio for gestational ages ~30 weeks. It can be speculated that the lower the ratio, the lower the vascular placental resistance and thus the higher the utero placental blood flow, resulting in increased fetal growth, which is reflected by increased fetal length (or femur length) during the second trimester and increased body weight (increased abdominal circumference) during the third trimester. The finding of best correlation between the peak-systolic! end-diastolic and the femur length in gestations <30 weeks (and between the peak-systolic / enddiastolic ratio versus abdominal circumference in gestations 230 weeks) should not be interpreted that the femur nomograms should be used at <30 weeks and the abdominal circumference nomograms at ~30 weeks. Our data simply suggest that the estimated fetal weight nomogram is similar to the gestational age nomogram, which is used widely, and that the greatest sensitivity for IUGR is achieved when the estimated fetal weight nomogram is used . As far as the clinical relevance of the other nomograms, (biparietal diameter, head circumference, abdominal circumference, and femur length vs the peak-systolic/ end-diastolic ratio), these may be used in patients with fetal congenital anomalies that preclude the use of the estimated fetal weight nomogram (i.e., hydrocephalus, skeletal dysplasias, hydrops, etc.). In such cases aj udgment should be made as to which biometric parameter is not affected by the fetal malformation , and the corresponding nomogram should be used. In summary, this study supports the use of gestational age-independent nomograms with fetal biometry versus the peak-systolic / end-diastolic ratio used to detect fetuses with increased placental vascular resistance. The use of these nomograms should prove useful in evaluating patients with unknown dates. It also sheds some light into the relationship between fetal biometric parameters and placental vascular resistance. Placental vascular resistance as determined by the peak-systolic! end-diastolic ratio is more closely correlated with fetal size parameters than with gestational age. The fact that appropriately grown fetuses of the same gestational age may have different peak-systolic! end-diastolic ratios may be due to differences in placental vascular resistance and therefore nutrition. To investigate this possibility the next logical step should be an assessment of the correlation between fetal biometry and the peak-systolic/ end-diastolic ratio in fetuses of the same gestational age. The value of the fetal biometric
Umbilical artery velocimetry and fetal biometry
1019
nomograms to assess IUGR is currently being investigated at our institution in a prospective manner.
REFERENCES 1. Trudinger BJ, Cook CM, Giles WB, Connelly A, Thompson RS. Umbilical artery flow velocity waveforms in highrisk pregnancy. Randomized controlled trial. Lancet 1987;1: 188-90. 2. Lombardi SJ, Rosemond R, Ball R, Entman SS, Boehm FH. Umbilical artery velocimetry as a predictor of adverse outcom.e in pregnancies complicated by oligohydramnios. Obstet Gynecol 1989;74:338-41. 3. Soothill PW, Nicolaides KH, Bilardo K, Hatchett GA, Campbell S. Uteroplacental blood velocity resistance index and umbilical versus p02, pC0 2, pH, lactate and erythroblast count in growth-retarded fetuses. Fetal Ther 1986; I: 176-9. 4. Hackett GA, Campbell S, Gunson H , Cohen-Overbeek T, Pearce JMF. Doppler studies in the growth retarded fetus and prediction of neonatal necrotizing enterocolitis, haemorrhage and neonatal morbidity. BMJ 1987;294: 13-6. 5. Rochelson BL, Schulman H, Fleischer A, et al. The clinical significance of Doppler umbilical velocimetry in the smallfor-gestational age fetus. AM J OBSTET GYNECOL 1987; 156: 1223-6. 6. Giles WB, Trudinger BJ, Baird PJ. Fetal umbilical artery flow velocity waveforms and placental resistance: pathological correlation. Br 1 Obstet Gynaecol 1985;92:31-8. 7. Dewhurst CJ, Beazley JM, Campbell S. Assessment of the fetal maturity and dysmaturity. AM J OBSTET GYNECOL 1972; 1l3:141-6. 8. Usher R, McLean F. Intrauterine growth of liveborn Caucasians at sea level: standards obtained from measurements in seven dimensions of infants born between 25 and 44 weeks gestation. J Pediatr 1969;74:901-6. 9. Hadlock FP, Kent WR, Loyd JL, et al. An evaluation of two methods for measuring fetal head and body circumference. J Ultrasound Med 1982; I :359-64. 10. O'Brien GD, QueenanJT. Growth of the ultrasound fetal femur length during normal pregnancy. Part 1. AM J OBSTET GYNECOL 1981;833: 141-5. 11. Sh~pard MJ, Richards VA, Berkowitz RL, et at. An evaluation of two equations for predicting fetal weight by ultrasound. AM J OBSTET GYNECOL 1982; 142:47-54. 12. Had lock FP, Harrist RB, Carpenter RJ , Deter RL, Park SK. Sonographic estimation of fetal weight. The value of femur length in addition to head and abdomen measurements. Radiology 1984;150:535-40. 13. McCallum WD, Williams CS, Nape1 S, Daigle RE. Fetal blood velocity waveforms. AM J OBSTET GYNECOL 1978; 132 :425-9. 14. Farmakides G, Schulman A, Saldama L, Bracero LA, Fleisher AS, Rochelson B. Surveillance of twin pregnancy with umbilical artery velocimetry. AM 1 OBSTET GYNECOL 1985; 153:789-92. 15. Fleischer A, Schulman H, Farmakides G, et at. Uterine arte ry Doppler velocimetry in pregnant women with hypertension. AM J OBSTET GYNECOL 1986;154:806-13. 16. Schulman H , Fleischer A, Stern W, Farmakides G,Jagani N, Blattner P. Umbilical velocity wave ratios in human pregnancy. AM J OBSTET GYNECOL 1984;148:985-90. 17. Trudinger BJ, Warwick BG, Cook CM, Bombardieri J , Collins L. Fetal umbilical artery flow velocity waveforms and placental resistance: clinical significance. AMJ OBSTET GYNECOL 1985;92:23-30. 18. Villar J, Belizan 1M. The timing factor in the pathophysiology of intrauterine growth retardation syndrome. Obstet Gynecol Sur 1982;37:499-506. 19. Tanner JM , Growth before birth. In: Fetus into man. Cambridge, Massachusetts: Harvard University Press, 1978:3.
Volume 165 Number 4, Part 1
plicated term vaginal deliveries. AM ] OBSTET GVNECOL 1985; 151 :798-80 1. 18. Thorp ]A, Sampson ]E, Parisi VM, Creasy RK. Routine umbilical cord gas determinations. AM] OBSTET GVNECOL 1989;161:600-5. 19. Gilstrap LC, Leveno K], Burris], Williams ML, Little BB. Diagnosis of birth asphyxia on the basis of fetal pH, Apgar score, and newborn cerebral dysfunction. AM] OBSTET GVNECOL 1989; 161 :825-30.
20. Freeman ]M, Nelson KB. Intrapartum asphyxia and cerebral palsy. Pediatrics 1988;82:240-4. 21. Weitzner ]S, Strassner HT, Rawlins RG, et al. Objective assessment of meconium content of amniotic fluid. Obstet Gynecol 1990;76: 1143-4. 22. Yeomans ER, Gilstrap LC, Leveno K], Burris ]S. Meconium in the amniotic fluid and fetal acid-base status. Obstet Gynecol 1989;73: 175-8.
The relationship between umbilical artery Doppler velocimetry and fetal biometry William E. Scorza, MD, Deborah Nardi, RT(R), RDMS, Anthony M. Vintzileos, MD, Alfred D. Fleming, MD, John F. Rodis, MD, and Winston A. Campbell, MD Farmington, Connecticut The relationship between peak-systolic/ end-diastolic ratio of the umbilical artery waveform and fetal biometry was studied in 127 uncomplicated pregnancies with established dates between 20 and 40 weeks' gestation. At each ultrasonographic examination fetal biometry included measurement of the biparietal diameter, head circumference, abdominal circumference, and femur length. The peak-systolic/end-diastolic ratio was measured by either a continuous or a pulsed-wave method. There were significant linear negative correlations between all the biometric parameters, as well as between the ultrasonographically estimated fetal weight and peak-systolic/end-diastolic ratio. Of the individual ultrasonographic parameters the femur length (for gestations <30 weeks) and the abdominal circumference (for gestations 2:30 weeks) were found to be best correlated with the peak-systolic/end-diastolic ratio. Regression curves, including the 10th and the 90th percentile, were developed between each biometric parameter (biparietal diameter, head circumference, abdominal circumference, and femur length), as well as between estimated fetal weight and peak-systolic/end-diastolic ratio. The estimated fetal weight nomogram had the best sensitivity (48%) in predicting intrauterine growth retardation. These nomograms should prove most useful in assessing downstream placental vascular resistance in high-risk patients with unknown dates. (AM J OBSTET GVNECOL 1991 ;165:1013-9.)
Key words: Umbilical artery velocimetry, fetal biometry Continuous or pulsed-wave Doppler ultrasonography of the fetal umbilical artery has been increasingly used in clinical obstetrics in the management of highrisk pregnancy. I·' The measurement of the peak-systolic/ end-diastolic ratio of the umbilical artery flow velocity waveform is the most commonly used method. Abnormal ratios have been associated with an increased incidence of neonatal morbidity, mortality, and intrauterine growth retardation. 1 . 5 The abnormal waveFrom the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Connecticut Health Center. Presented at the Eleventh Annual Meeting of the Society of Perinatal Obstetricians, San Francisco, California, january 28-February 2, 1991. Reprint requests: William E. Scorza, MD, Department of Obstetrics and Gynecology, Division of Maternal-Fetal Medicine, Mount Sinai Hospital, 500 Blue Hills Ave., Hartford, CT 06112. 6/6/30898
forms are believed to be the result of increased downstream placental vascular resistance and therefore of impaired placental perfusion. 6 The proper use of the umbilical artery peak-systolic/ end-diastolic ratio, however, requires knowledge of the length of gestation, which is unknown or unavailable in 20% to 40% of pregnant patients. 7 When the last menstrual period is unknown or uncertain, extrapolation of gestational age on the basis of fetal biometric data may lead to erroneous interpretation of the peak-systolic/ end-diastolic ratio results, because there are no data regarding the relationship between the umbilical artery peak-systolic/ end-diastolic ratio and fetal biometry. Therefore the purpose of the present report was: (1) to study the relationships and generate regression curves between the peak-systolic/ end-diastolic ratio and the most commonly used fetal biometric parameters (biparietal di1013
1014 Scorza et al.
October 1991 Am J Obstet Gyneco1
5.5
y=-.185X+4.132
5.0
.5! iii...
4.5 4.0
0
en >...
.!"-
III III
u
•
3.5
~
• •
3.0
••
2.5
• •
J:I
E
•
2.0
I
••
•
•
•
• • •
1.5 1.0 5
7
6
8
9
10
Biparietal diameter(cm)
Fig. 1. Scattergram illustrating relationship (along with 10th, 50th, and 90th percentile lines) between umbilical artery peak-systolic lend-diastolic ratio (SID) and biparietal diameter (R = 0.395, P = 0.0001).
ameter, head circumference, abdominal circumference, femur length, and estimated fetal weight); (2) to assess which fetal biometric parameters, if any, are significantly correlated with the peak-systolic/ end-diastolic ratio; and (3) to determine the sensitivity of the fetal biometric parameter nomograms to predict intrauterine growth retardation (IUGR). Material and methods Patients were recruited from those referred to our antepartum fetal evaluation unit at the University of Connecticut Health Center. We used 127 uncomplicated pregnancies between 20 and 40 weeks of gestation (dates confirmed by early ultrasonography) for a single ultrasonographic assessment. The patients were referred for confirmation of clinical dates or fetal size. Patients were selected if they had initiated prenatal care early in pregnancy, if they had regular menstrual cycles with reliable menstrual dates, and if the results of ultrasonographic examination were consistent with their menstrual dates. All patients had intact membranes and singleton pregnancies. Patients were excluded if fetal anomalies were detected or if the estimated fetal weight was < - 2 SD or >2 SD from the mean for gestational age. s The ultrasonographic examinations were performed, using a model 128 (Acuson, Mountain View, Calif.) or a model 3000 or 2600 (General Electric, Rancho Cordova, Calif.). A 3.5 or 5 MHz linear-array trans-
ducer was used. All machines had freeze-frame capabilities and electronic calipers for measurement. The ultrasonographic velocity was 1540 m/sec. The study was cross-sectional, and each patient was included only once. Routine measurements (biparietal diameter, head circumference, abdominal circumference, and femur length) were obtained and recorded on each patient. The biparietal diameter and head circumference measurements were obtained from an axial scanning plane at the level of the thalami. The fetal abdominal circumference was taken perpendicular to the long axis of the fetal body at the level of the ductus venosus complex. The abdominal circumference was measured directly with electronic calipers or predominantly by using the formula Dl + D2 X 1.57, where D is diameter (both methods have been shown to give equivalent results).9 The femur length was measured according to the technique described by O'Brien and Queenan. 10 The ultrasonographic estimated fetal weight was calculated by averaging the estimated fetal weight obtained by both Shepard's (using biparietal diameter and abdominal circumference)" and Hadlock's (using abdominal circumference and femur length)l2 formulas. The umbilical artery peak-systolic/ end-diastolic ratio was obtained transabdominally by continuous (model 5000A, Multigon, Mount Vernon, N.Y.) or pulsed-wave (model 128, Acuson) Doppler ultrasonography. All studies were performed during fetal apnea. The peak-systolic/ end-diastolic ratios of three waveforms from three
Umbilical artery velocimetry and fetal biometry
Volume 165 Number 4 , Pan I
1015
5.5 y=-.059X+4.364
5.0 ~
iii
4.5
~
0
4 .0
>-
3.5
in ~
!
~
as
3.0
ii u
o
2.5
.Q
E
~
. ....
:
0
00
2.0
00
..-.
1.5 1.0 20
22
24
26
28
30
32
34
Head clrcumference(cm)
Fig. 2. Scattergram illustrating relationship (along with 10th, 50th, and 90th percentile lines) between umbilical artery peak-systolic I end-diastolic (SID) ratio and head circumference (r = 0.422, P = 0.0001).
different sampling sites were computed, and the mean ratio was used in the analysis. Scattergrams of the peak-systolic! end-diastolic ratio values versus gestational age and fetal biometry were developed. Regression analyses were used to establish nomograms, including the 10th, 50th, and 90th percentile lines. To assess which fetal biometric para meteres), if any, best correlated with the peak-systolic/ enddiastolic ratio, multiple regression analyses and stepwise regression analyses were used. A p value < 0.05 was considered statistically significant. The statistical analysis of the data was performed by computer (Macintosh II) with the Stat View 512 + (Abacus Concepts, Berkeley, Calif.) and CricketGraph (CricketGraph, Malvern, Pa.) statistical packages. During the study period there were 25 patients with known dates (based on last menstrual period and confirmed by early ultrasonography) who had growthretarded fetuses (estimated fetal weights < - 2 SD from the mean for gestational age).s The gestational ages of these fetuses ranged from 26 to 38 weeks. The fetuses were used to determine the sensitivity of the newly constructed nomograms on the basis of fetal biometry, to predict growth retardation. In all 25 fetuses the diagnosis of IVGR was confirmed after birth.
Results The study population consisted of 127 patients with gestational age (mean ± SD) of 31 ± 4 weeks (median,
31.1). Of these , 104 (81 %) were white, 12 (10%) black, and 11 (9%) Hispanic. Seventy patients (55 %) were nulliparous and 57 (45%) were multiparous. The maternal age (mean ± SD) was 26.3 ± 5.6 years (range, 15 to 40). The best-fit curves between peak-systolic/ enddiastolic ratio and biparietal diameter, head circumference, abdominal circumference, femur length, and estimated fetal weight were linear. The r values were 0.395 , 0.422,0.461,0.453, and 0.460, respectively (all p values were = 0.0001). The scattergrams, including the 10th, 50th, and 90th percentile boundaries, and the regression equations are illustrated in Figs. 1 to 5 . The estimated fetal weight had the highest linear negative correlation with the peak-systolic/ end-diastolic ratio (r = 0.461). The best-fit curve between the peaksystolic/ end-diastolic ratio and gestational age also was linear (Fig. 6). To determine the origins and to assess which fetal biometric parameters contribute to this negative correlation between estimated fetal weight and the peaksystolic/ end-diastolic ratio, the data were analyzed according to two gestational age groups, <30 weeks' gestation (46 patients) and ~30 weeks' gestation (81 patients). In both gestational age groups there was a significant linear negative correlation between estimated fetal weight and the peak-systolic! end-diastolic ratio (p < 0.05 and p < 0.0005, respectively). To assess which ultrasonographic parameter(s) significantly correlated with the peak-systolic/ end-diastolic ratio within
1016
Scorza et al.
October 199 1 Am J Obstet Gynecol
5.5
y=-.05X+4.05
5.0
.2 iii ...
c in >...
!... CD
iii
4.5 4.0 3.5 3.0
u
2.5
:0 E
2.0
~
•
•
••
•
•
•
•
." .
1.5
15
19
17
23
21
29
27
25
31
33
35
Abdominal clrcumference(cm) Fig. 3. Scattergram illustrating relation ship (along w ith 10th , 50th, and 90th perce ntile lines) between umbilical artery peak·systolic/end-diastolic (S I D) ratio and abdominal circumference (r = 0 .460. P = 0.000 1).
5.5 y=-.235X+4.089
5.0 0
:;::
...
4.5
0
4.0
...>...
3.5
III
en ~
III
iii (,,)
•
E
•
3.0 2.5
••
.Q ~
•
• •• •• • •
•
•
2.0
•••
•
• •••
•
I
. . • •• .. - •.
1.5 1.0 4 .0
4 .5
5.5
5.0
6.0
6.5
7.0
7 .5
Femur lenllth(cm) Fig. 4. Scattergram illustrating relationship (along with 10th, 50th, a nd 90th percentile lines) betwee n umbilical artery peak-systolic I end-diastolic (SID ) ratio and femur length (r '" 0.453, P '" 0.0001 ).
each gestational age group. multiple regression analyses and stepwise regression analyses were used with the peak-systolic lend-diastolic ratio used as the dependent (outcome) variable and biparietal diameter, head
circumference, abdominal circumference, and femur length used as the inde pendent (predictors) variables. Multiple regression analysis o f he t group < 30 weeks' gestation revealed that the overall association was sig-
Umbilical artery velocimetry and fetal biometry
Volume 165 Number 4, Part I
5.5
y=-3.020E-4x+3.234
5.0
..... 0
en >... ... III
..
+
+ +
4.5
III
c
1017
4.0
+ + •
+
+ •
3.5
G)
III
3.0
.!:!
2.5
.c E
2.0
:)
•
.
•
••
1.5 1.0 500
1000
1500
2000
2500
3000
3500
4000
Estimated fetal weight (in gms) Fig. 5. Scattergram illustrating relationship (along with 10th, 50th, and 90th percentile lines) between umbilical artery peak-systolic I end-diastolic (SID) ratio and estimated fetal weight (r = 0.461, P = 0.0001). Plus sign, Growth-retarded fetuses. 5.5 y=-.045X+4.115 5.0 0
~ c
en ~
~ as
as
u
4.5 4.0
•
3.5 3.0
.c
E
:)
•
..
22
24
2.5 2.0
...
. . ... .,- .- ..-...•-.•-... ...... • •
• • •
1.5
20
26
28
30
32
34
36
38
40
Gestational age (weeks) Fig. 6. Scattergram illustrating relationship (along with 10th, 50th, and 90th percentile lines) between umbilical artery peak-systoliclend-diastolic ratio and gestational age (r = 0.391, P = 0.0001).
nificant (p < 0.05). Of th~ four independent variables, only femur length was significantly correlated with the peak-systolic! end-diastolic ratio (p < 0.05; 13 coefficient, -0.477; standard error, 0.211; standard coeffi-
cient, - 0.673; t value, 2.262). Stepwise regression analysis of this group of fetuses of <30 weeks' gestation also indicated that the femur length was the only independent variable to enter the equation (r = 0.477,
1018 Scorza et al.
W = 0.227,adjustedR 2 ;= 0.21). Similar analyses were performed for the group of fetuses with gestational ages ~30 weeks. With multiple regression analysis it was found that the overall correlation of the independent variables to the peak-systolic! end-diastolic ratio was significant (p < 0.0005). Of the individual independent variables only abdominal circumference was significantly associated with the peak-systolic/ end-diastolic ratio (p < 0.005; 13 coefficient, - 0.086; standard error, 0.03; standard coefficient, - 0.534; t value, 2.896). Stepwise regression analysis also was performed, and only the abdominal circumference entered the equation (r = 0.525, R2 = 0.276, adjusted R2 = 0.267). The same analyses were repeated with gestational age included as one of the independent variables, but the results were the same. Even when gestational age was used as one of the independent variables, only femur length for fetuses <30 weeks' gestation and only abdominal circumference for fetuses ~30 weeks' gestation again were the only variables significantly correlated with the peak-systolic/ enddiastolic ratio. With the established nomograms between the peaksystolic! end-diastolic ratio and the fetal biometric parameters, the sensitivity of each parameter to detect IUGR was calculated. The gestational age of the fetuses with IUGR was 32.7 (± 3.6) weeks (median, 33). Two of the fetuses had absent end-diastolic velocities and were excluded from calculations of sensitivity. The sensitivity of detecting growth retardation by using the biparietal diameter nomogram was 35% (8/23), head circumference 35% (8/23), abdominal circumference 35% (8/23), femur length 30% (7/23), and estimated fetal weight 48% (11123). Of these ultrasonographic parameters, the estimated fetal weight versus the peaksystolic/ end-diastolic ratio had the best sensitivity. The sensitivity by using the gestational age versus the peaksystolic/end-diastolic ratio nomogram was 52% (12/23).
Comment In 1978 McCallum et al. 13 suggested that umbilical artery velocity waveforms could be recorded by ultrasonographic Doppler probes. Subsequently various investigators suggested that the calculation of peak-systolic/ end-diastolic ratio of the umbilical artery velocity waveform reflects placental vascular resistance. 6 Elevated ratios have been associated with an increased incidence of adverse perinatal outcome, such as growth-retarded infants who are prone to hypoxic complications, twin-to-twin transfusion, and preeclampsia'" 5.14. 15 It has been clearly shown from previous investigations that there is a steady decline in the peak-systolic/ end-diastolic ratio with advancing gestational age,16, 17 the length of gestation should be known for the accurate interpretation of the ratio result. How-
October 1991
Am
J Obslet Gynccol
ever, the last menstrual period is unknown or uncertain in approximately 20% to 40% of pregnant patients. 7 This is the first report to evaluate the relationships and establish nomograms between each of the most commonly used ultrasonographic parameters (biparietal diameter, head circumference, abdominal circumference, femur length) and the estimated fetal weight and peak-systolic/ end-diastolic ratio. Statistically significant relationships were found between the ratio and each of the ultrasonographic parameters as well as the estimated fetal weight. There was an inverse relationship between the ratio and each ultrasonographic parameter. The use of these gestational age-independent nomograms may prove useful for evaluating placental resistance in patients with unknown dates or late registrants with uncertain dates. In the 23 fetuses with IUGR encountered during the study period, the sensitivity of the fetal biometry versus the peak-systolic/end-diastolic ratio nomograms to detect IUGR was not significantly different from that with the gestational age versus the peak-systolic/ end-diastolic ratio nomogram. The best sensitivity was achieved by the estimated fetal weight versus the ratio nomogram (48%), which compared quite favorably with the sensitivity obtained by using gestational age (52%). The detection of an abnormally high peak-systolic/ end-diastolic ratio in growth-retarded fetuses depends on whether small fetal size is the result of impaired uteroplacental blood flow or non placental causes. Furthermore, even with utero placental insufficiency the ratio still may be normal in some cases. Therefore it is not surprising that the sensitivity of detecting IUGR by our method is approximately 50% because many of the fetuses that weigh < 10th percentile for gestational age may suffer lesser degrees of uteroplacental insufficiency or the low weight may be the result of non placental causes. Of interest were the results of the multiple and stepwise regression analyses that used the peak-systolie/end-diastolic ratio as the dependent (outcome) variable and the remaining ultrasonographic biometric parameters as the independent variables. One of the most interesting results of this study was the only significant correlation of femur length to the peak-systolic/ enddiastolic ratio for fetuses <30 weeks and of abdominal circumference to peak-systolic/ end-diastolic ratio for fetuses ~30 weeks. These findings persisted even when the gestational age was included as one of the independent variables. The explanation for these findings lies in the fact that femur length is an indirect measure of fetal length and abdominal circumference is an indirect measure of fetal body weight. It is already known that the fetus exhibits two growth spurts, one in fetal length during the second trimester and the other in body weight during the third trimester. 18, 19 A breakpoint of 30 weeks was chosen in the analysis of the data because, as has been previously shown,18. 19 the fetal
Volume 165 Numbe r 4, Part 1
growth spurt in length has been almost completed while the growth spurt in weight has just begun by 30 weeks of gestation. Therefore it is not surprising to find a significant correlation between femur length, which represents fetal length, and the peak-systolic/ end-diastolic ratio at gestational ages <30 weeks and between abdominal circumference, which represents fetal body weight, and the peak-systolic / end-diastolic ratio for gestational ages ~30 weeks. It can be speculated that the lower the ratio, the lower the vascular placental resistance and thus the higher the utero placental blood flow, resulting in increased fetal growth, which is reflected by increased fetal length (or femur length) during the second trimester and increased body weight (increased abdominal circumference) during the third trimester. The finding of best correlation between the peak-systolic! end-diastolic and the femur length in gestations <30 weeks (and between the peak-systolic / enddiastolic ratio versus abdominal circumference in gestations 230 weeks) should not be interpreted that the femur nomograms should be used at <30 weeks and the abdominal circumference nomograms at ~30 weeks. Our data simply suggest that the estimated fetal weight nomogram is similar to the gestational age nomogram, which is used widely, and that the greatest sensitivity for IUGR is achieved when the estimated fetal weight nomogram is used . As far as the clinical relevance of the other nomograms, (biparietal diameter, head circumference, abdominal circumference, and femur length vs the peak-systolic/ end-diastolic ratio), these may be used in patients with fetal congenital anomalies that preclude the use of the estimated fetal weight nomogram (i.e., hydrocephalus, skeletal dysplasias, hydrops, etc.). In such cases aj udgment should be made as to which biometric parameter is not affected by the fetal malformation , and the corresponding nomogram should be used. In summary, this study supports the use of gestational age-independent nomograms with fetal biometry versus the peak-systolic / end-diastolic ratio used to detect fetuses with increased placental vascular resistance. The use of these nomograms should prove useful in evaluating patients with unknown dates. It also sheds some light into the relationship between fetal biometric parameters and placental vascular resistance. Placental vascular resistance as determined by the peak-systolic! end-diastolic ratio is more closely correlated with fetal size parameters than with gestational age. The fact that appropriately grown fetuses of the same gestational age may have different peak-systolic! end-diastolic ratios may be due to differences in placental vascular resistance and therefore nutrition. To investigate this possibility the next logical step should be an assessment of the correlation between fetal biometry and the peak-systolic/ end-diastolic ratio in fetuses of the same gestational age. The value of the fetal biometric
Umbilical artery velocimetry and fetal biometry
1019
nomograms to assess IUGR is currently being investigated at our institution in a prospective manner.
REFERENCES 1. Trudinger BJ, Cook CM, Giles WB, Connelly A, Thompson RS. Umbilical artery flow velocity waveforms in highrisk pregnancy. Randomized controlled trial. Lancet 1987;1: 188-90. 2. Lombardi SJ, Rosemond R, Ball R, Entman SS, Boehm FH. Umbilical artery velocimetry as a predictor of adverse outcom.e in pregnancies complicated by oligohydramnios. Obstet Gynecol 1989;74:338-41. 3. Soothill PW, Nicolaides KH, Bilardo K, Hatchett GA, Campbell S. Uteroplacental blood velocity resistance index and umbilical versus p02, pC0 2, pH, lactate and erythroblast count in growth-retarded fetuses. Fetal Ther 1986; I: 176-9. 4. Hackett GA, Campbell S, Gunson H , Cohen-Overbeek T, Pearce JMF. Doppler studies in the growth retarded fetus and prediction of neonatal necrotizing enterocolitis, haemorrhage and neonatal morbidity. BMJ 1987;294: 13-6. 5. Rochelson BL, Schulman H, Fleischer A, et al. The clinical significance of Doppler umbilical velocimetry in the smallfor-gestational age fetus. AM J OBSTET GYNECOL 1987; 156: 1223-6. 6. Giles WB, Trudinger BJ, Baird PJ. Fetal umbilical artery flow velocity waveforms and placental resistance: pathological correlation. Br 1 Obstet Gynaecol 1985;92:31-8. 7. Dewhurst CJ, Beazley JM, Campbell S. Assessment of the fetal maturity and dysmaturity. AM J OBSTET GYNECOL 1972; 1l3:141-6. 8. Usher R, McLean F. Intrauterine growth of liveborn Caucasians at sea level: standards obtained from measurements in seven dimensions of infants born between 25 and 44 weeks gestation. J Pediatr 1969;74:901-6. 9. Hadlock FP, Kent WR, Loyd JL, et al. An evaluation of two methods for measuring fetal head and body circumference. J Ultrasound Med 1982; I :359-64. 10. O'Brien GD, QueenanJT. Growth of the ultrasound fetal femur length during normal pregnancy. Part 1. AM J OBSTET GYNECOL 1981;833: 141-5. 11. Sh~pard MJ, Richards VA, Berkowitz RL, et at. An evaluation of two equations for predicting fetal weight by ultrasound. AM J OBSTET GYNECOL 1982; 142:47-54. 12. Had lock FP, Harrist RB, Carpenter RJ , Deter RL, Park SK. Sonographic estimation of fetal weight. The value of femur length in addition to head and abdomen measurements. Radiology 1984;150:535-40. 13. McCallum WD, Williams CS, Nape1 S, Daigle RE. Fetal blood velocity waveforms. AM J OBSTET GYNECOL 1978; 132 :425-9. 14. Farmakides G, Schulman A, Saldama L, Bracero LA, Fleisher AS, Rochelson B. Surveillance of twin pregnancy with umbilical artery velocimetry. AM 1 OBSTET GYNECOL 1985; 153:789-92. 15. Fleischer A, Schulman H, Farmakides G, et at. Uterine arte ry Doppler velocimetry in pregnant women with hypertension. AM J OBSTET GYNECOL 1986;154:806-13. 16. Schulman H , Fleischer A, Stern W, Farmakides G,Jagani N, Blattner P. Umbilical velocity wave ratios in human pregnancy. AM J OBSTET GYNECOL 1984;148:985-90. 17. Trudinger BJ, Warwick BG, Cook CM, Bombardieri J , Collins L. Fetal umbilical artery flow velocity waveforms and placental resistance: clinical significance. AMJ OBSTET GYNECOL 1985;92:23-30. 18. Villar J, Belizan 1M. The timing factor in the pathophysiology of intrauterine growth retardation syndrome. Obstet Gynecol Sur 1982;37:499-506. 19. Tanner JM , Growth before birth. In: Fetus into man. Cambridge, Massachusetts: Harvard University Press, 1978:3.