Hypertension in Pregnancy, 27:237–245, 2008

ISSN: 1064-1955 print / 1525-6065 online

DOI: 10.1080/10641950802000901

Prediction of Maternal
1525-6065
1064-1955
LHIP
Hypertension in Pregnancy
Pregnancy, Vol. 27, No. 3, Jun 2008: pp. 0–0

Predisposition to Preeclampsia
MaternaletPredisposition
Emonts al. to Preeclampsia
Downloaded from informahealthcare.com by CDL-UC Davis on 01/17/15. For personal use only.

Patrick Emonts,1 Sontera Seaksan,1 Laurence Seidel,2

Henri Thoumsin,1 Ulysse Gaspard,1 Adelin Albert,2
and Jean-Michel Foidart1
1
Department of Obstetrics and Gynaecology, University of Liège, CHR Citadelle, Liège,
Belgium
2
Hypertens Pregnancy 2008.27:237-245.

Department of Biostatistics, University of Liège, CHU Sart Tilman, Liège, Belgium

Objective: To derive a prediction index based on the most salient patient history,
laboratory, and clinical parameters for identifying women at high risk for developing
preeclampsia (PE). Methods: Nonpregnant women with a history of PE (n = 101)
were compared with nonpregnant parous women with a history of one or more suc-
cessful normotensive pregnancies (n = 50) but with comparable age, gestation, and
parity profiles. The parameters included a medical examination (demographics,
patient history, family history, and clinical and obstetrical findings), laboratory
investigations (hemostasis, coagulation, and vitamins), and morphological and func-
tional tests (cardiovascular and renal functions). Stepwise logistic regression analysis
was applied to develop a three-step PE prediction index based on the most discrimi-
nant parameters. Results: Patients with and without PE differed significantly (p < 0.05)
with respect to 1) maternal history of chronic hypertension, body mass index, and
blood pressure; 2) APTT, PT, activated factor VIII, homocystein, free protein S and
vitamin B1; and 3) relative plasma volume. Based on these three sets of parameters,
a three-step PE prediction index was developed. The likelihood ratio of a positive
index score was equal to 3.4, 7.3, and 8.8, respectively. Thus, assuming a PE preva-
lence (or prior probability) of 5%, a patient’s chances of developing PE when present-
ing with a positive score on the three-step prediction index were 15%, 28%, and 32%,
respectively. Discussion: In the absence of welldefined pre-pregnancy screening
guidelines for PE, the present study attempts to proceed in a stepwise fashion by
looking at medical examination data first, requesting, if necessary, specific hemostasis
and coagulation tests next, and finally measuring the relative plasma volume for con-
firmatory purposes. This approach offers a satisfactory positive predictive value and
cost efficiency ratio.

Keywords Preeclampsia, Hypertension, Thrombophilia, Screening, Prenatal care.

This article is not subject to United States copyright law.

Address correspondence to Dr. Patrick Emonts, Department of Obstetrics and Gynae-
cology, University of Liège, Centre Hospitalier Régional de la Citadelle, Boulevard 12e
Ligne 1, 4000 Liège, Belgium. E-mail: [email protected]
 238 Emonts et al.

INTRODUCTION
Preeclampsia (PE) is a major cause of maternal and fetal mortality and mor-
bidity (1–2). It is a pregnancy-specific multisystemic disorder, occuring most
commonly in primigravida and characterized by the development of hyperten-
sion and proteinuria in the second half of pregnancy (3). The pathogenesis of
PE is not fully understood, but it is believed that genetic predisposition and
Downloaded from informahealthcare.com by CDL-UC Davis on 01/17/15. For personal use only.

immune deficiency lead to placental ischemia and perturbation of the mater-

nal vascular endothelium (3). It is associated with abnormal vasoregulation,
platelet hyperaggregability, activation of the coagulation system, and endot-
helial-cell dysfunction (4). Clinical findings associated with PE can manifest
as a maternal syndrome, such as severe PE, HELLP (hemolysis, elevated liver
enzymes, and low platelets) syndrome or eclampsia, and retro-placental
hematoma (RPH). The fetal consequences of PE include intra-uterine growth
retardation (IUGR) and intra-uterine fetal death (IUFD) (5).
Hypertens Pregnancy 2008.27:237-245.

In the absence of reliable predictive markers, these is no effective prevention

strategy or curative treatment for PE (6). Since PE is a heterogeneous condi-
tion with multiple pathophysiological pathways, it has been difficult to
develop reliable predictive measures of PE for routine use in clinical practice (7).
In other words, there is no single best test for predicting PE (8). It has been
recently suggested that multivariate analysis of clinical, biochemical, and
biophysical parameters could improve the screening efficacy of past PE (7).
The present study aimed to identify among several clinical, laboratory, and
functional investigations those parameters that best discern nonpregnant
women with a past history of PE from women with uneventful pregnancies. A
three-step PE predictive index was developed, first based on simple patient
history findings, next upgraded with carefully selected laboratory tests, and
finally optimized by including patients’ relative plasma volume.

METHODS
The study protocol was approved by the Human Research Ethics Committee
of the University teaching hospital.
One hundred and one women hospitalized between 1999 and 2002 with a
diagnosis of severe PE were included in the study. Preeclampsia was defined
as severe if at least one of the following conditions was met (9): systolic blood
pressure ≥160 mm Hg and/or diastolic blood pressure ≥110 mm Hg, pro-
teinuria ≥3.5g/24 h, platelet count <100,000/μL, elevated serum transaminase
levels ( >3SD), oliguria (diuresis <500mL/24 h), pulmonary edema, epigastric
pain, or cerebral or visual disturbances. Preeclampsia was complicated in 21
paticnts by the HELLP syndrome, namely a Weinstein acronym (10) charac-
terized by hemolysis (decrease of hemoglobin or hematocrit >10%, total biliru-
bin >1mg/dL, lactate deshydrogenase >600 U/L), elevated liver enzymes
 Maternal Predisposition to Preeclampsia 239

(serum transaminase levels >3SD), and low platelet count (<100.000/μL).

Further, some PE patients had IUFD (n = 11), severe IUGR (<5th percentile)
(n = 56), or RPH (complete or incomplete placental abruption with conse-
quential effects on the mother and/or the fetus) (n = 9). A group of 50 parous
women (without PE) with a past history of successful term delivery after a
normotensive pregnancy was selected and matches on average to the
patients with PE with respect to gestation, parity, age, smoking history,
Downloaded from informahealthcare.com by CDL-UC Davis on 01/17/15. For personal use only.

and time since pregnancy. Written, informed consent was obtained from all
patients.
The parameters recorded for each subject were classified into three cate-
gories. The first encompassed a detailed medical examination, including age,
height, weight, body mass index (BMI), smoking, history, systolic and diastolic
blood pressures measured at least 18 months after pregnancy with patients at
rest in the sitting position, medications, patient history of insulin-dependent
Hypertens Pregnancy 2008.27:237-245.

diabetes mellitus, chronic hypertension, deep venous thromboembolism

(VTE), glomerulopathy, and of cardiac insufficiency, paternal and maternal
medical history of insulin dependent diabetes mellitus, chronic hypertension,
VTE, and of renal disease, and patient obstetrical history (parity, gestation,
abortion, miscarriage, other pregnancy related disorders, gestational age at
birth, mode of delivery, birth weight, and apgar score). The second category
included an extensive clinical laboratory investigation (performed in the same
registered laboratory) of hemostasis and coagulation, including platelets
count and platelet activating factor (PFA), activated partial thromboplastin
time (APTT), prothrombin time (PT), thrombotest, activated factors II, VII,
VIII and XI, plasma antithrombin III, activated protein C, free protein S, acti-
vated protein C resistance and factor V Leiden, factor II mutation G20210A,
lupus anticoagulant, anticardiolipin antibodies, anti-β2 glycoprotein-I antibodies,
homocysteine (baseline and 4 hours before ingestion of 0.1kg/kg of methion-
ine), and vitamin B (B1, B6 and B12). The third included a morphological and
functional evaluation of the cardiovascular function (electrocardiography to
identify repolarization abnormalities, cardiac ultrasound to evaluate the
ejection fraction and the thickness of the interventricular septum), renal
parameters (albuminuria and creatinine clearance), and determination of
plasma volume (total plasmatic volume determined by administration of albumin
marked by radioactive iodine isotope and relative plasmatic volume in relation to
weight).
Results are expressed as means ± standard deviations (SD) for quantita-
tive variables and as frequencies and proportions (%) for categorical variables.
Mean values for patients with and without PE were compared by Student
t-test and by nonparametric Wilcoxon test. Proportions were compared by the chi-
square test. The two groups of patients were also compared from a multivariate
standpoint while accounting for matching factors by logistic regression analy-
sis. A stepwise variable selection was used to retain the most discriminating
 240 Emonts et al.

variables. Specifically, logistic regression was applied in three steps: 1) on

medical examination data only, 2) on laboratory data only, but including the
score derived in Step 1, and 3) on morphological and functional data only, but
including the score derived in Step 2. The efficacy of the three predictive
scores was assessed by sensitivity, specificity, and positive and negative likeli-
hood ratios. The likelihood ratio (LR) is defined as the ratio of the probability
of a positive (negative) score in the patients with PE and the probability of a
Downloaded from informahealthcare.com by CDL-UC Davis on 01/17/15. For personal use only.

positive (negative) score in the patients without PE. All results were consid-
ered to be significant at the 5% critical level (p < 0.05). Statistical analyses
were carried out using SAS (version 9.1 for Windows) and S-PLUS (version
7.0 for Windows) software packages.

RESULTS
Hypertens Pregnancy 2008.27:237-245.

The clinical data recorded in patients with and without PE are displayed in
Table 1. A stepwise logistic regression analysis was applied to these data,
except for those parameters related to first pregnancy. As a result, history of
chronic hypertension in the patient’s mother (p = 0.0051), body mass index
(p = 0.023), systolic (p = 0.022) and diastolic blood pressure (p = 0.022) mea-
sured at least 18 months after pregnancy were found to be jointly associated
with PE. All remaining variables either did not significantly contribute to the
discrimination of patients with from those without PE or entailed redundant
information. A first PE prediction index (R1) was established on the selected
variables and the matching parameters, namely R1=-3.72+0.030 × age
(years)-0.50 × parity + 0.15 × gestation + 1.89 × HTA in patient’s mother (yes = 1/
no = 0) + 0.14 × BMI (kg/m2) + 0.079 × SBP (mm Hg)-0.13 × DBP (mm Hg).
Positive scores indicate a high risk for PE, whereas negative scores tend to
correspond to an uneventful pregnancy. In the PE group, 67% were positive
(sensitivity) and in the non-PE group, 80% were negative (specificity), yielding
a positive LR of 3.4. Thus, a positive R1 score is over three times more likely in
patients with PE than in those without the disorder. The corresponding negative
LR was equal to 0.41.
The distributions of laboratory tests in patients with and without PE are sum-
marized in Table 2. Significant differences were found for APTT, activated factor
VIII, free protein S, anti-cardiolipin antibodies (IgG and IgM), homocysteine lev-
els and vitamins B1, B6, and B12. When combining the first PE prediction index
R1 and the biological parameters into a stepwise logistic regression, it appeared
that only APTT (p = 0.0007), prothrombin time (p = 0.037), activated factor VIII
(p = 0.0018), free protein S (p = 0.021), homocysteine levels (p = 0.0002), and vita-
min B1 (p = 0.0095) levels improved the discrimination between patients with and
without PE. A second PE prediction index R2 was derived as follows: R2 = 30 +
0.92 × R1 -0.28 × APTT (s) - 0.15 × PT (%) − 2.6 × activated factor VIII (UI)-0.046
× free protein S (%) + 0.11 × homocystein 4h (μmol/L) - 0.026 × vitamin B1(μg/L).
 Maternal Predisposition to Preeclampsia 241

Table 1: Anamnestic and clinical examination parameters recorded in 101 women

with PE and 50 matched patients without PE. Results are expressed as mean ± SD
or frequencies (%).

Parameters PE patients Non-PE patients p Value

Number of patients 101 50 —

Age (yrs) 29.9 ± 4.5 29.8 ± 4.9 0.82
Height (cm) 164 ± 6.4 165 ± 8.3 0.55
Downloaded from informahealthcare.com by CDL-UC Davis on 01/17/15. For personal use only.

Weight (kg) 68.9 ± 13.6 63.3 ± 10.0 0.010

BMI (kg/m2) 25.7 ± 5.1 23.3 ± 2.8 0.0024
Smoking 44 (44) 23 (46) 0.86
Patient history
Diabetes 5 (5) 1 (2) 0.66
Hypertension 17 (17) 5 (10) 0.33
Thrombosis 3 (3) 0 (0) 0.55
Paternal history
Diabetes 9 (9) 2 (4) 0.26
Hypertension 32 (32) 7 (14) 0.018
Hypertens Pregnancy 2008.27:237-245.

Thrombosis 6 (6) 2 (4) 0.72

Maternal history
Diabetes 13 (13) 3 (6) 0.27
Hypertension 25 (25) 3 (6) 0.0039
Thrombosis 11 (11) 3 (6) 0.39
Obstetrical history
Parity 1.5 ± 0.9 1.6 ± 1.1 0.64
Gestation 2.2 ± 1.7 2.1 ± 1.3 0.94
Abortion 0.24 ± 0.55 0.24 ± 0.56 0.95
No. of early miscarriage 0.49 ± 1.11 0.24 ± 0.56 0.24
First pregnancy
Infant’s birth weight (g) 1870 ± 971 3258 ± 354 <0.0001
Caesarean section 62 (62) 4 (8) <0.0001
Birth gestational age (wks) 33.9 ± 5.0 38.8 ± 1.1 <0.0001
Blood pressure
SBP (mm Hg) 129 ± 11.3 122 ± 8.5 0.022
BDP (mm Hg) 73.5 ± 7.9 72.8 ± 5.3 0.56

The sensitivity of this index increased to 88% and the specificity to 88%, yielding a
positive LR of 7.3 and a negative LR of 0.14.
At the last step, patients with and without PE were compared with respect to
cardiac and renal function tests (Table 3). Significant differences were found in
the proportions of echocardiographic abnormalities, such as ejection fraction and
heart size (32% of PE patients and 12% of non-PE patients, p < 0.05). Renal func-
tion also different significantly as revealed by a higher albuminuria, although the
range remained within normal limits. The creatinine clearance rate was higher in
women with PE, whereas the plasma volume (both total and relative) was signifi-
cantly lower. When adding morphological and functional parameters to the sec-
ond PE prediction index in a stepwise logistic regression analysis, only the
relative plasma volume (RPV) led to a significant improvement in PE prediction
(p = 0.0075). The final PE prediction index was: R3 = 10.0 + 1.05 × R2 - 0.27 × RPV
(ml/kg). Its sensitivity remained 88%, while its specificity rose to 90%, yielding a
positive LR of 8.8 and a negative LR of 0.13.
 242 Emonts et al.

Table 2: Laboratory parameters measured in 101 women with PE and 50 matched

patients without PE. Results are expressed as mean ± SD.

PE patients Non PE patients

Parameters n = 101 n = 50 p Value

Platelets (103/mm3) 250 ± 52 267 ± 64 0.11

Platelet function 103 ± 22 95.6 ± 21 0.04
aPTT (s) 32.6 ± 3.8 37.1 ± 14 <0.0001
Downloaded from informahealthcare.com by CDL-UC Davis on 01/17/15. For personal use only.

Prothrombin time (%) 95.5 ± 11 97.2 ± 3.8 0.41

Thrombotest 1.01 ± 0.09 1.03 ± 0.10 0.43
Factor II a (IU) 1.10 ± 0.23 1.12 ± 0.20 0.43
Factor VII a (IU) 1.10 ± 0.24 1.16 ± 0.19 0.11
Factor VIII a (IU) 1.26 ± 0.35 1.53 ± 0.35 <0.0001
Factor XI a (IU) 1.09 ± 0.23 1.12 ± 0.24 0.58
Antithrombin III (%) 110 ± 11 107 ± 11 0.12
Activated protein C (%) 104 ± 17 109 ± 21 0.14
Free protein S (%) 84.4 ± 16 92.9 ± 14 0.0010
APCR 2.24 ± 0.35 2.20 ± 0.37 0.54
9.5 ± 5.6 7.4 ± 4.7
Hypertens Pregnancy 2008.27:237-245.

Anticardiolipin antibody IgG (UPGL/ml) 0.017

IgM (UPGL/ml) 7.5 ± 5.9 5.2 ± 3.8 0.012
Anti-β2 glycoprotein antibodies (U/mL) 5.5 ± 4.8 4.6 ± 3.0 0.65
Homocystein (μmol/L)
Baseline 11.4 ± 5.5 9.8 ± 3.3 0.046
4 hours aft Meth 37.7 ± 18.9 24.5 ± 9.0 <0.0001
Vitamin B (μg/l)
B1 38 ± 31 59 ± 2.3 <0.0001
B6 9.1 ± 12 9.6 ± 4.4 0.029
B12 358 ± 200 567 ± 235 <0.0001

Table 3: Cardiovascular and renal functional assessment in 101 PE women (data

missing for 10 patients) and 50 matched patients without PE. Results are expressed
as means ± SD or frequencies (%).

PE patients Non PE patients

Parameters n = 101(a) n = 50(b) p Value

ECG abnormalities 5 (5) 0 (0) 0.17

Cardiac ultrasonography abnormalities 30 (32) 6 (12) 0.014
Diuresis (L) 1.24 ± 0.32 1.25 ± 0.20 0.84
Albuminuria (mg/24h) 14.9 ± 14 9.2 ± 7.0 0.0095
Creatinine clearance (mL/min) 104 ± 32 88 ± 11 0.0011
Total plasma volume (l) 2.19 ± 0.35 2.47 ± 0.32 <0.0001
Relative plasma volume (ml/kg) 32.6 ± 6.7 39.4 ± 3.2 <0.0001
(a)
data
missing for 9 patients.
(b)
data missing for 3 patients.

If one assumes that the prevalence of PE is 5%, a positive score of R1 will

lead to a positive predictive value (PVV) of 15%. In other words, the patient’s
risk for developing PE has been approximately multiplied threefold. For posi-
tive scores of R2 and R3, the PVV increases to 28% and 32%, respectively.
Thus, a patient presenting with a positive score R3 has a ≥30% chance of
developing PE.
 Maternal Predisposition to Preeclampsia 243

DISCUSSION
The present study aimed to predict the occurrence of severe PE from specific
maternal features outside pregnancy. In this regard, our results confirm find-
ings previously reported by others. Women with thrombophilic disorders are
at increased risk for placental vascular pathology (11). Several studies have
documented an association between specific genetic or acquired thrombophilic
Downloaded from informahealthcare.com by CDL-UC Davis on 01/17/15. For personal use only.

disorders with vascular complications, such as PE, RPH, IUGR or fetal

demise(12–21). Women with elevated homocysteine levels are considered to be
at higher risk not only for thrombosis but also for PE (22). Our findings
confirm this opinion. Vitamins B levels are closely related to hyperhomocys-
teinemia because they are cofactors for enzymes involved in the degradation
of homocysteine (23,24). It is therefore no surprise that the present study
emphasized their alterations in women with previous PE.
Expansion of plasma volume in normal pregnancy improves the placental
Hypertens Pregnancy 2008.27:237-245.

perfusion (25,26), but PE is associated with a relative decrease in plasma

volume and plasma volume expansion (27). The pathophysiological basis of
this observation remains unclear but could be the consequence of a mild
chronic endothelial dysfunction. Our study documented the significance of a
relative decrease in plasma volume. It is therefore tempting to speculate that
a low plasma volume outside pregnancy is an important contributing risk
factor for the development of PE.
In this study, we proposed a three-step PE prediction index to be used
poor to pregnancy. Less than 1% of the nullipare population develops severe
PE. The low incidence of a clinically significant disease may preclude the use
of expensive tests in such a low-risk population (28). Identification of women
who may develop severe PE is then most desirable in terms of reducing neona-
tal and maternal morbidity and mortality as well as subsequent social
expenses (28). Considered individually, risk factors such as blood pressure,
BMI, and family history are not discriminant enough. However, when com-
bined with each other but also with laboratory and functional tests, they can
be highly predictive for PE. In daily practice, the physician may proceed a
follows. At the initial prenatal care visit, a first PE prediction index based on
simple and rapidly available anamnestic and clinical data is computed. A
highly negative score of this index may halt further investigation and hence
lower costs, whereas a highly positive score may lead to immediate intervention
and consequently reduce morbidity. Borderline (near 0) and positive scores,
however, may always prompt the attending physician to request additional
specific laboratory and functional investigations for confirmation. We found
that by measuring APTT, PT, activated factor VIII, free protein S, homocystein,
and vitamin B1, the patient’s risk for presenting PE can again be assessed. A
large positive (negative) score will comfort the physician about the high (low)
patient’s chances of developing PE. In case of a borderline or positive score,
 244 Emonts et al.

measurement of the relative plasma volume may become necessary to make a

final decision about the risk for developing PE. It is believed that the three-step
decision-making process developed here provides a cost effective way to assess
the individual risk for PE of future mothers and in particular to improve
our ability of identifying as precisely as possible patients at highest risk for
preeclampsia.
Downloaded from informahealthcare.com by CDL-UC Davis on 01/17/15. For personal use only.

REFERENCES
1. Confidential Enquiries into Maternal Deaths. Why mothers die 1997–1999. The
fifth report of the confidential enquiries into maternal deaths in the United Kingdom.
London: Royal College of Obstetricians and Gynaecologists Press, 2001.
2. Confidential Enquiry into Stillbirths and Deaths in Infancy. 8th annual report.
London: Maternal and Child Health Research Consortium, 2001.
3. Walker ID. Thromphilias and preeclampsia In: Pre-eclampsia, Etiology and Prac-
Hypertens Pregnancy 2008.27:237-245.

tical Practice, Fiona Lyall and Michael Belfort, eds. Cambridge University Press,
2007, p. 183.
4. Sibai BM, Dekker G, Kupferminc M. Pre-eclampsia. Lancet 2005; 365:785–799.
5. Ness RB, Roberts JM. Heterogeneous causes constituting the single syndrome of
pre-eclampsia: a hypothesis and its implications. Am J Obstet Gynecol 1996;
175:1365–1370.
6. Audibert Fr. Screening for pre-eclampsia: the quest for the Holy Grail? Lancet
2005;365:1367–1379.
7. Tjoa ML, Oudejans CBM, Van Vugt JMG, Blankenstein MA, Van Wijk IJ. Markers
for presymptomatic prediction of pre-eclampsia and intrauterine growth restriction.
Hypertens Pregnancy 2004;23(2):171–189.
8. Conde-Agudelo A, Villar J, Lindheimer M. World Health Organization systematic
review of screening tests for preeclampsia. Obstet Gynecol 2004;104(6): 1367–1391.
9. Sibai BM, Mercer B, Sarinoglu C. Severe preeclampsia in the second trimester:
recurrence risk and long-term prognosis. Am J Obstet Gynecol 1991 Nov;165(5 Pt 1):
1408–1412.
10. Weinstein L. Syndrome of hemolysis, elevated liver enzymes, and low platelet
count: a severe consequence of hypertension in pregnancy. Am J Obstet Gynecol
1982;142:159–167.
11. Walker ID. Thrombophilia in pregnancy. J Clin Pathol 2000;53:573–580.
12. Gates S. Thromboembolic disease in pregnancy. Cur Opin Obstet Gynecol
2000;12:117–122.
13. Walker JJ. Pre-eclampsia. Lancet 2000;356:1260–1265.
14. Alfirevic Z, Robets D, Martlew V. How strong is the association between maternal
thrombophilia and adverse pregnancy outcome? A systematic review. Eur J
Obstet Gynecol Reprod Biol 2002;101(1):6–14.
15. Empson M, Lassere M, Craig JC, Scott JR (2002) Recurrent pregnancy loss with
antiphospholipid antibody: a systematic review of therapeutic trials. Obstet
Gynecol 2002;99;135–144.
16. Ghee CB, Burrows RF. Prothrombin G20210A mutation is not associated with
recurrent miscarriages. Aust N Z J Obstet Gynecol 2002;42:167–169.
 Maternal Predisposition to Preeclampsia 245

17. Morrison ER, Miedzybrodzka ZH, Campbell DM, Haites NE, Wilson BJ, Watson
MS. Prothrombotic genotypes are not associated with pre-eclampsia and gesta-
tional hypertension: results from a large population-based study and systematic
review. Thromb Haemost 2002;87:779–785.
18. Greer IA. Thrombophilia: implications for pregnancy outcome. Thromb Res
2003;2152:1–9.
19. Rey E, Kahn SR, David M, Shrier I. Thrombophilic disorders and fetal loss: a
meta-analysis. Lancet 2003;361:901–908.
Downloaded from informahealthcare.com by CDL-UC Davis on 01/17/15. For personal use only.

20. Lin J. Genetic Thrombophilias and pre-eclampsia: a meta-analysis. Am Coll

Obstet Gynecol 2005;105:182–192.
21. Robertson L & coll. Thrombophilia in pregnancy: a systematic review. Br J
Haematol 2006;132:171–186.
22. Cotter A, Molloy A, Scott JM, Daly SF. Elevated plasma homocysteine in early
pregnancy: A risk factor for the development of severe preeclampsia. Presented at
the 21st Annual Meeting of the Society of Maternal-Fetal Medicine, Reno,
Nevada. February 5–10, 2001.
Hypertens Pregnancy 2008.27:237-245.

23. Sibai BM. Hypertensive Disorders in Women. Philadelphia: W.B. Saunders

Company 2001;1001.
24. Duckitt K, Harrington D. Risk factors for pre-eclampsia at antenatal booking:
systematic review of controlled studies. BMJ 2005;330:565–571.
25. Chesley LC. Plasma and red cell volumes during pregnancy. Am J Obstet Gynecol
1972;112:440–450.
26. Silver HM, Seebeck MA, Carlson R. Comparison of total volume in normal, pre-
eclamptic, and non proteinuric gestational hypertensive pregnancy by subcutane-
ous measurement of red blood cell and plasma volume. Am J Obstet Gynecol
1998;179:87–93.
27. Cunningham FG, Gant NF, Leveno KJ, GilstrapLC, Hauth JC, Wenstrom KD.
Hematological Changes: Maternal adaptations to pregnancy (chapter 8) : Physiology
of Pregnancy (section II); In Williams Obstetrics, 21st Edition, New York:
McGraw-Hill. 2001, pp 177–178.
28. Myatt L, Carpenter LB. Prediction of preeclampsia, In: Pre-eclampsia, Etioloy and
Clinical Practice, E. Fiona Lyall and Michael Belfort, eds. Chapter 15, Cambridge:
Cambridge University Press, 2007; p. 226.
 Hypertens Pregnancy 2008.27:237-245.
Downloaded from informahealthcare.com by CDL-UC Davis on 01/17/15. For personal use only.