Mean Arterial Pressure at 11ⴙ0 to 13ⴙ6 Weeks

in the Prediction of Preeclampsia

Leona C.Y. Poon, Nikos A. Kametas, Ivilina Pandeva, Catalina Valencia, Kypros H. Nicolaides

Abstract—This study aimed to determine the performance of screening for preeclampsia (PE) by maternal medical history and
mean arterial pressure (MAP) at 11⫹0 to 13⫹6 weeks. In 5590 women with singleton pregnancies attending for routine care
at 11⫹0 to 13⫹6 week’s gestation we recorded maternal variables and measured the MAP. We excluded 397 because they had
missing outcome data or the pregnancies resulted in miscarriage or termination. In 104 patients there was subsequent
development of PE, 97 developed gestational hypertension, 574 delivered small-for-gestational-age newborns, and 4418 were
unaffected by PE, gestational hypertension, or small for gestational age. A multivariate Gaussian model was fitted to the
distribution of log multiple of the median MAP in the PE and unaffected groups. Likelihood ratios for log multiple of the
median MAP were computed and used together with maternal variables to produce patient-specific risks for each case.
Detection rates and false-positive rates were calculated by taking the proportions with risks above a given risk threshold. In
the unaffected group, log MAP was influenced by maternal age, ethnic origin, smoking, family and personal history of PE,
and fetal crown-rump length. In the prediction of PE, significant contributions were provided by log multiple of the median
MAP, ethnic origin, body mass index, and personal history of PE. The detection rate of PE by log multiple of the median MAP
and maternal variables was 62.5% for a false-positive rate of 10%. Maternal variables, together with MAP, at 11⫹0 to 13⫹6
weeks identify a group at high risk for development of PE. (Hypertension. 2008;51:1027-1033.)
Key Words: first trimester 䡲 mean arterial pressure 䡲 pregnancy 䡲 preeclampsia 䡲 screening

reeclampsia (PE), which affects ⬇2% of pregnancies, is

P a major cause of perinatal and maternal morbidity and
mortality.1–3 Attempts at prevention of PE by prophylactic
methods in selection of the screened population, measure-
ment of BP, cutoffs used in defining the screen-positive
group, and definitions of PE. There are 2 first-trimester
Downloaded from http://ahajournals.org by on September 10, 2023

interventions from midgestation have been largely unsuc- screening studies for PE. The first study used an automated
cessful.4 –7 It is uncertain whether interventions starting from device to measure BP in 983 women at 9 to 12 weeks and
the first rather than the second trimester would prove to be reported that, with a cutoff of 90 mm Hg in mean arterial
more effective in the prevention of PE, but before this could pressure (MAP), the DR of PE was 62%, for an FPR of
be investigated, it is essential to develop a method of effective 38%.15 However, the definition of PE used in this study is not
and early identification of the high-risk group. accepted by any professional organization, because it was
The likelihood of developing PE is increased by a based on the development of gestational hypertension (GH)
number of factors in the maternal history, including with either weight gain or a reading of only 1⫹ of protein on
Afro-Caribbean ethnicity, nulliparity, high body mass dipstick analysis on 1 occasion. The second study was a
index (BMI), and previous or family history of PE.8,9 retrospective one in which the medical charts of pregnant
However, screening by maternal history alone will detect women attending for routine prenatal care were examined to
only ⬇30% of those who will develop PE, for a false- identify the BP measurements taken by mercury sphygmo-
positive rate (FPR) of 10%.9 manometers before 20 weeks (mean: 13.7 weeks) from 1655
The diagnosis of PE is based on the demonstration of high women.20 At a cutoff of 92 mm Hg in MAP, the DR of PE
blood pressure (BP) and significant proteinuria during the was 25%, for an FPR of 10%.
second half of pregnancy in previously normotensive women. In current clinical practice, the use of mercury sphygmo-
Several second-trimester studies have reported on the use of manometers remains the gold standard for noninvasive BP
BP measurement as a screening method for subsequent monitoring, but there are concerns for both the clinical
development of PE. These studies have reported contradic- performance and safety of these instruments.21,22 These prob-
tory results with FPR ranging from 7% to 52% and detection lems have been largely overcome by the use of automated BP
rates (DRs) ranging from 8% to 93% (Table 1).10 –19 These devices, but so far only 1 of these has been validated for use
differences are likely to be the consequence of the varied both in pregnancy and in PE.23

Received November 7, 2007; first decision November 12, 2007; revision accepted November 19, 2007.
From the Harris Birthright Research Centre for Fetal Medicine, King’s College Hospital, London, United Kingdom.
Correspondence to Nikos A. Kametas, Harris Birthright Research Centre for Fetal Medicine, King’s College Hospital, Denmark Hill, London SE5 9RS,
United Kingdom. E-mail [email protected]
© 2008 American Heart Association, Inc.
Hypertension is available at http://hyper.ahajournals.org DOI: 10.1161/HYPERTENSIONAHA.107.104646

1027
 1028 Hypertension April 2008 Part II

Table 1. Second-Trimester Studies That Assessed the Use of BP Measurement as a

Screening Method for Hypertensive Disorders in Pregnancy
Prevalence, MAP Cutoff, DR, FPR,
Author Year N % mm Hg % %
Fallis et al10* 1963 113 35 90 82 12
Page et al11† 1976 14 833 3 90 44 13
Phelan et al12‡ 1977 96 16 85 60 12
Quaas et al13§ 1982 801 20 90 63 11
Öney et al14II 1983 200 15 90 93 34
Moutquin et al15¶ 1985 983 8 85 85 52
Marya et al *16 1988 200 14 90 32 16
Villar et al17# 1989 700 20 90 8 7
Ales et al18** 1989 730 5 85 88 16
Conde-Agudelo et al19†† 1993 580 15 85 61 21
*Hypertensive disease in pregnancy was not defined.
†PE is hypertension (the mean of the 2 highest recordings in MAP ⱖ110 mm Hg in the third trimester) with
proteinuria (ⱖ2⫹ on dipstick) and/or generalized edema.
‡GH is BP ⬎140/90 mm Hg with a rise in diastolic BP by 20 mm Hg.
§GH is BP persistently above 135/85 mm Hg or an increase in diastolic pressure of 20 mm Hg above what was
observed on a previous occasion.
IIGH is systolic BP ⱖ140 mm Hg and/or diastolic BP ⱖ90 mm Hg on 2 occasions 6 hours apart after 24 weeks
gestation. PE is hypertension (as for GH) and proteinuria (ⱖ300 mg/dL in a 24-hour urine collection).
¶GH is BP ⱖ140/90 mm Hg after 20 weeks that disappears at the postpartum visit. PE is hypertension (BP
ⱖ140/90 mm Hg after 20 weeks on 2 occasions 6 hours apart and BP ⱕ120/80 mm Hg at 6 weeks postpartum),
with proteinuria (ⱖ1⫹ on dipstick on 2 occasions 6 hours apart after 20 weeks) and/or edema (weight gain of ⬎1
kg/week) or elevated serum urate levels (ⱖ4.6 mg/dL).
#GH is systolic BP ⱖ140 mm Hg and/or diastolic BP ⱖ90 mm Hg on 2 occasions 6 hours apart. PE is hypertension
(as for GH) and proteinuria (not defined).
**GH is BP ⱖ140/90 mm Hg or MAP ⱖ107 mm Hg or an increase in systolic BP by ⱖ30 mm Hg and in diastolic
Downloaded from http://ahajournals.org by on September 10, 2023

BP by ⱖ15 mm Hg on 2 occasions 6 hours apart after 24 weeks’ gestation. PE is hypertension (as for GH) and
proteinuria (ⱖ2⫹ on dipstick in a random urine specimen or ⱖ300 mg/dL in a 24-hour urine collection).
††GH is systolic BP ⱖ140 mm Hg and/or diastolic BP ⱖ90 mm Hg on 2 occasions 6 hours apart after 20 weeks’
gestation. PE is hypertension (as for GH) and proteinuria (ⱖ2⫹ on dipstick in ⱖ2 random urine specimens 6 hours
apart or ⱖ300 mg/dL in a 24-hour urine collection).

We used a validated automated device to prospectively beyond 23 weeks), obstetric history (including previous pregnancy
measure the MAP at 11⫹0 to 13⫹6 weeks in 5590 singleton with PE), and family history of PE (sister, mother, or both). The
pregnancies attending for routine pregnancy care.23 The aim maternal weight and height were measured, and the BMI was
calculated in kilograms per meter squared.
of our study was to determine the performance of screening The BP was taken by automated devices (3BTO-A2, Microlife),
for PE by maternal characteristics and measurement of MAP which were calibrated before and at regular intervals during the
at 11⫹0 to 13⫹6 weeks. study. The recordings were made by doctors who had received
appropriate training on the use of these machines. The women were
Methods in the seated position, their arms were supported at the level of the
This was a prospective screening study for PE in singleton pregnan- heart, and a small (⬍22-cm), normal (22- to 32-cm), or large (33- to
cies. All of the women were attending our center for routine 42-cm) adult cuff was used depending on the midarm circumfer-
assessment of risk for chromosomal abnormalities by measurement ence.26 After rest for 5 minutes, BP was measured in both arms
of fetal nuchal translucency thickness and maternal serum-free simultaneously, and a series of recordings were made at 1-minute
␤-human chorionic gonadotropin and pregnancy-associated plasma intervals until variations between consecutive readings fell within
protein A at 11⫹0 to 13⫹6 weeks of gestation between March and 10 mm Hg in systolic and 6 mm Hg in diastolic BP in both arms.27
December 2006.24,25 Gestational age was derived from the fetal When this point of stability was reached, we calculated the MAP of
crown-rump length (CRL). Written informed consent was obtained each arm as the average of the last 2 stable measurements, and, as
from the women agreeing to participate in the study, which was recommended, we took the arm with the highest final MAP for the
approved by King’s College Hospital Ethics Committee. subsequent analysis of results.27
Patients were asked to complete a questionnaire on maternal age, The MAP, ultrasound findings, and woman’s characteristics,
ethnic origin (white, Afro-Caribbean, Indian or Pakistani, Chinese or including demographic data and obstetric and medical history,
Japanese, or mixed), cigarette smoking during pregnancy (yes or no), were entered into a computer database. Data on pregnancy
alcohol intake during pregnancy (yes or no), drug abuse during outcome were collected from the hospital maternity records or
pregnancy (yes or no), medical history (including chronic hyperten- their general medical practitioners. The obstetric records of all of
sion, diabetes mellitus, antiphospholipid syndrome, thrombophilia, the women with pre-existing or pregnancy associated hyperten-
HIV infection, and sickle cell disease), medication (including anti- sion were examined to determine whether the condition was
hypertensive, antidepressant, antiepileptic, anti-inflammatory, anti- chronic hypertension, PE, or GH. Similarly, for quality control,
retroviral, antithyroid, aspirin, betamimetic, insulin, lithium, ste- we examined the records of 500 randomly selected patients
roids, or thyroxin), parity (parous or nulliparous if no delivery without pregnancy-associated hypertension.
 Poon et al First-Trimester Blood Pressure in Preeclampsia 1029

Outcome Measures diastolic BP in 4524 subjects (87.1%). It became necessary to

The outcome measures were PE and GH with or without small for have 3 recordings (10.1%) in 525, 4 (1.7%) in 89, 5 (0.7%) in
gestational age. The group of patients with PE included those with 36, and 6 or 7 (0.4%) in 19.
PE superimposed on chronic hypertension.
The definitions of PE and GH were those of the International
Society for the Study of Hypertension in Pregnancy.28 In patients Log MAP in the Unaffected Group
with GH, the diastolic BP should be ⱖ90 mm Hg on ⱖ2 occasions In the multiple regression model for log MAP, significant
4 hours apart developing after 20 weeks of gestation in previously independent contributions were provided by maternal ethnic
normotensive women in the absence of significant proteinuria, and in origin, age, BMI, previous history of PE, maternal history of
patients with PE there should be GH with proteinuria of ⱖ300 mg in PE, cigarette smoking, and fetal CRL.
24 hours or 2 readings of at least ⫹⫹ on dipstick analysis of
midstream or catheter urine specimens if no 24-hour collection is Log MAP⫽1.8666⫺0.0002⫻CRL (in mm)⫹(⫺0.0035 if
available. In chronic hypertension there should be a history of Afro-Caribbean, ⫺0.0075 if Indian or Pakistani, ⫺0.0132 if
hypertension before conception or the presence of hypertension at Chinese or Japanese, ⫺0.0077 if mixed, or 0 if white)⫹
the booking visit before 20 weeks of gestation in the absence of 0.0005⫻age (in years)⫹0.0024⫻BMI (in kg/m2)⫹(0.0057 if
trophoblastic disease. In PE superimposed on chronic hypertension,
significant proteinuria (as defined above) should develop after 20
woman’s mother had PE or 0 if she did not)⫹(⫺0.008 if parous
weeks of gestation in women with known chronic hypertension.28 without previous PE, 0.009 if parous with previous PE or 0 if
nulliparous)⫹(⫺0.0078 if smoker or 0 if not smoker; R2⫽0.120;
Statistical Analysis P⬍0.001).
The following 8 steps were taken. First, the women were subdivided
into 3 groups depending on pregnancy outcome: PE, GH, and Distributions of log MoM MAP
unaffected by PE or GH and delivering babies with birth weight
In each woman we, first, logarithmic transformed the mea-
above the 10th percentile after correction for gestation at delivery
and sex of the newborn, maternal ethnic origin, weight, height, and sured MAP (log-observed MAP), second, used the formula
parity.29 Second, the distribution of MAP was made Gaussian after above for log MAP in the unaffected group to calculate the
logarithmic transformation. Third, multiple regression analysis was log expected MAP, and, third, calculated the ratio of the
used to determine which of the factors among the maternal charac- observed-to-expected values: log (observed/expected)⫽log
teristics, medical and obstetric history, and gestation (see Table 2)
were significant predictors of log MAP in the unaffected group. MoM MAP⫽log observed⫺log expected.
Fourth, the distribution of log MAP, expressed as multiples of the The mean (95% CI of the mean) MoM MAP was 1.0
median (MoMs) of the unaffected group, was determined in the PE (1.0008 to 1.0055) MoM in the unaffected group, 1.0840
and GH groups. Fifth, multiple regression analysis was used to (1.0649 to 1.1032) MoM in the PE group, and 1.0646 (1.0452
determine which of the factors among the maternal characteristics,
to 1.0839) MoM in the GH group (Figure 1). Therefore, the
medical and obstetric history and gestation (see Table 2), had a
Downloaded from http://ahajournals.org by on September 10, 2023

significant contribution in explaining the a priori risk for PE and GH. mean MoM MAP in both the PE and GH groups was
Sixth, likelihood ratios were computed from the fitted distributions significantly higher than in the unaffected group. The mean
of log MoM values in the unaffected pregnancies and in each of the MoM MAP did not change significantly with gestation at
2 groups with pregnancy complications. Seventh, patient-specific delivery in either the PE (r⫽0.184; P⫽0.059) or the GH
risks for each complication were derived by multiplying the appro-
priate a priori risk with the likelihood ratio. Eighth, the DR and FPR group (r⫽0.155; P⫽0.130).
were calculated as the respective proportions of PE and GH (DR) and
unaffected pregnancies (FPR) with MoM values above given cutoffs. Likelihood Ratios for PE and GH
The statistical software package SPSS 15.0 (SPSS Inc) was used for The overlapping Gaussian distributions of log MoM MAP in
all of the data analyses. the unaffected group and each of the PE and GH groups were
used to calculate the likelihood ratios for each pregnancy
Results complication (Table 3).
Study Population
First-trimester screening was carried out in 5590 consecutive The A Priori Risk for PE and GH
singleton pregnancies with a live fetus at 11⫹0 to 13⫹6 weeks. The a priori risk for each pregnancy complication is calcu-
We excluded 397 (7.1%) because they had missing outcome lated from the following formula: odds/(1⫹odds), where
data (n⫽289), the pregnancies resulted in fetal death or odds⫽eY and Y is derived from multiple regression analysis
miscarriage before 24 weeks of gestation (n⫽50), or the of maternal characteristics, medical and obstetric history. For
pregnancies were terminated for fetal abnormalities (n⫽49) PE: Y⫽⫺6.311⫹(1.299 if Afro-Caribbean or 0 if other
or social reasons (n⫽9). In the remaining 5193 subjects there ethnic origin)⫹0.092⫻BMI (in kg/m2)⫹(0.855 if woman’s
were 104 (2.0%) who developed PE, including 6 subjects mother had PE or 0 if she did not)⫹(⫺1.481 if parous without
with PE superimposed on chronic hypertension, 97 (1.9%) previous PE, 0.933 if parous with previous PE, or 0 if nullipa-
who developed GH, 574 (11.0%) who did not develop PE or rous; R2⫽0.153; P⬍0.001). For GH: Y⫽⫺5.967⫹0.092⫻BMI
GH but delivered small-for-gestational-age newborns, and (in kg/m2)⫹(⫺0.822 if parous without previous PE or 0 if
4418 (85.1%) subjects who were unaffected by PE, GH, or parous with previous PE or nulliparous; R2⫽0.051; P⬍0.001).
small for gestational age. In the quality control assessment of
the 500 subjects with reported normal outcome, there was 1 Patient-Specific Risk for PE and GH
subject with GH. The characteristics of the 3 outcome groups The likelihood ratios for PE and GH from the log MoM MAP
are summarized in Table 2. are shown in Table 3. For example, in an Afro-Caribbean
The difference in BP between the first 2 measurements was woman in her first pregnancy, with no family history of PE,
⬍10 mm Hg for the systolic BP and ⬍6 mm Hg for the who is 28 years old, has a BMI of 20 kg/m2, does not smoke,
 1030 Hypertension April 2008 Part II

Table 2. Maternal Characteristics, Medical and Obstetric History, and Gestation in the
3 Groups of Pregnancy Outcomes
Maternal Characteristic Unaffected (n⫽4418) PE (n⫽104) GH (n⫽97)
Maternal age, mean (range), y 31.9 (16.0 to 46.0) 31.8 (17.0 to 49.0) 32.3 (17.0 to 46.0)
BMI, mean (range), kg/m2 25.5 (15.4 to 52.6) 28.8 (18.9 to 46.4)* 28.1 (19.0 to 53.0)*
CRL, mean (range), mm 64.4 (45.0 to 84.0) 63.9 (47.0 to 84.0) 63.7 (47.0 to 84.0)
Ethnicity, n (%)
White 3232 (73.2) 46 (44.2)‡ 72 (74.2)
Afro-Caribbean 755 (17.1) 45 (43.3)‡ 19 (19.6)
Indian or Pakistani 216 (4.9) 6 (5.8) 1 (1.0)
Chinese or Japanese 72 (1.6) 3 (2.9) 1 (1.0)
Mixed 143 (3.2) 4 (3.8) 4 (4.1)
Parity, n (%)
Nulliparous 2054 (46.5) 66 (63.5)‡ 57 (58.8)*
Parous, no previous PE 2258 (51.1) 21 (20.2)‡ 32 (33.0)‡
Parous, previous PE 106 (2.4) 17 (16.3)‡ 8 (8.2)†
Cigarette smoker, n (%) 317 (7.2) 5 (4.8) 8 (8.2)
Alcohol drinker, n (%) 41 (0.9) 1 (1.0) 0
Drug abuser, n (%) 16 (0.4) 0 1 (1.0)
Family history of PE
Mother, n (%) 170 (3.8) 12 (11.5)† 7 (7.2)
Sister, n (%) 81 (1.8) 3 (2.9) 0
Conception, n (%)
Spontaneous 4303 (97.4) 100 (96.2) 94 (96.9)
Ovulation drugs 48 (1.1) 1 (1.0) 0
In-vitro fertilization 67 (1.5) 3 (2.9) 3 (3.1)
Downloaded from http://ahajournals.org by on September 10, 2023

Medical history, n (%)

None 4303 (97.4) 95 (91.3)† 95 (97.9)
Chronic hypertension 41 (0.9) 6 (5.8)‡ 0
Diabetes mellitus 38 (0.9) 1 (1.0) 2 (2.1)
Antiphospholipid syndrome 9 (0.2) 1 (1.0) 0
Thrombophilia 17 (0.4) 1 (1.0) 0
Sickle cell 6 (0.1) 0 0
HIV infection 4 (0.1) 0 0
Medication during pregnancy, n (%)
None 4098 (92.8) 90 (86.5)* 85 (87.6)
Antihypertensive agents 29 (0.7) 4 (3.8)† 0
Insulin 34 (0.8) 1 (1.0) 2 (2.1)
Steroids 8 (0.2) 1 (1.0) 0
␤-mimetics 61 (1.4) 2 (1.9) 2 (2.1)
Combined asthma medications 40 (0.9) 2 (1.9) 2 (2.1)
Thyroxin 57 (1.3) 2 (1.9) 2 (2.1)
Aspirin 40 (0.9) 1 (1.0) 1 (1.0)
Antithyroid medication 3 (0.1) 0 0
Antiepileptic 18 (0.4) 0 2 (2.1)
Lithium 2 (0.05) 0 0
Antidepressants 22 (0.5) 1 (1.0) 1 (1.0)
Antiretroviral 1 (0.02) 0 0
Anti-inflammatory 5 (0.1) 0 0
Comparison with unaffected group: *P ⬍ 0.05; †P⬍0.01; ‡P⬍0.001.
 Poon et al First-Trimester Blood Pressure in Preeclampsia 1031

Figure 2. Receiver operating characteristic curves of history,

MAP, and the combination of the 2 in the prediction of PE.

Figure 1. Box-whisker plot of MAP MoM of the 3 pregnancy 0.0024⫻20 (BMI in kg/m2)⫹0 (no family history of PE)⫹0
outcome groups: unaffected, PE, and GH. (nulliparous)⫺0 (non smoker)⫽1.9121
● Log MoM MAP⫽1.9294⫺1.9121⫽0.0173
is at 12 weeks of gestation (CRL: 65 mm), with an MAP of ● Likelihood ratio for log MoM MAP of 0.0173 (see Table
85 mm Hg, the risk of developing PE is 3.9%. 3)⫽0.98
For the a priori risk for PE:
A posteriori risk⫽a priori risk⫻likelihood ratio:
● Y⫽⫺6.311⫹1.299 (Afro-Caribbean)⫹0.092⫻20 (BMI in
kg/m2)⫹0 (no family history of PE)⫹0 (nulliparous)⫽ ● risk for PE⫽0.04023312⫻0.98⫽0.039 or 3.9%
⫺3.172
● Odds⫽ey⫽0.04191967 If the same woman had had a previous pregnancy with PE
● A priori⫽odds/(1⫹odds)⫽0.04023312 and her BMI was 35 kg/m2, her risk for PE would have
been 29.2%.
Downloaded from http://ahajournals.org by on September 10, 2023

For the likelihood ratio for PE:

Performance of Screening
● Measured MAP⫽85, log observed MAP⫽1.9294 The DR of PE and GH for different FPRs in screening by
● Log-expected MAP⫽1.8666⫺0.0002⫻65 (CRL in mm)⫺
maternal factors only, MAP only, and by the combination of
0.0035 (Afro-Caribbean)⫹0.0005⫻28 (age in years)⫹
the 2 are given in Figures 2 and 3. The areas under the
Table 3. The Likelihood Ratios for PE and GH From the Log receiver operating characteristic curves (AROCs) for the
MAP MoMs detection of PE were significantly higher with the combined
model (AROC: 0.852) than with either history alone (AROC:
Likelihood Ratio (95% CI)
0.801; P⫽0.017) or MAP alone (AROC: 0.734; P⬍0.001).
Log MoM MAP PE GH Similarly, for the detection of GH, the AROC for the
⫺0.05 0.27 (0.27 to 0.27) 0.38 (0.35 to 0.41) combined model was significantly higher (AROC: 0.743)
⫺0.04 0.30 (0.29 to 0.32) 0.43 (0.42 to 0.44)
than with either history alone (AROC: 0.682; P⫽0.030) or
⫺0.03 0.35 (0.34 to 0.36) 0.47 (0.44 to 0.49)
⫺0.02 0.40 (0.38 to 0.42) 0.52 (0.49 to 0.55)
⫺0.01 0.49 (0.45 to 0.54) 0.61 (0.58 to 0.63)
0 0.63 (0.59 to 0.67) 0.72 (0.67 to 0.76)
0.01 0.76 (0.67 to 0.85) 0.84 (0.77 to 0.92)
0.02 0.98 (0.88 to 1.07) 1.03 (0.92 to 1.14)
0.03 1.27 (1.08 to 1.46) 1.27 (1.13 to 1.41)
0.04 1.68 (1.44 to 1.93) 1.59 (1.42 to 1.77)
0.05 2.32 (1.92 to 2.72) 2.05 (1.83 to 2.27)
0.06 3.14 (2.80 to 3.48) 2.65 (2.37 to 2.97)
0.07 4.55 (3.92 to 5.19) 3.75 (3.19 to 4.29)
0.08 7.73 (5.05 to 7.71) 5.11 (4.69 to 5.54)
0.09 8.74 (7.57 to 9.92) 7.40 (7.40 to 7.40)
0.10 14.45 (13.04 to 15.87) 10.27 (9.58 to 10.97)
Figure 3. Receiver operating characteristic curves of history,
0.11 22.64 (22.10 to 23.18) 14.80 (14.75 to 14.85)
MAP, and the combination of the 2 in the prediction of GH.
 1032 Hypertension April 2008 Part II

MAP alone (AROC: 0.680; P⫽0.006). At a 10% FPR, the development of chronic hypertension.36,37 The association
DR of PE was 43.3% for history alone, 37.5% for MAP alone, between black race and increased risk of PE is well docu-
and 62.5% for combined testing, and the respective values for mented.38 The surprising finding in our study was that, in
GH were 27.8%, 32.0%, and 41.2%. women not developing PE or GH, the MAP was lower in
women of Afro-Caribbean origin than in whites. These racial
Discussion differences are the subject of further investigation.
In this screening study for hypertensive disorders of preg- The underlying mechanism for PE is thought to be im-
nancy we did the following: (1) prospectively examined a paired trophoblastic invasion of the maternal spiral arteries
large population of pregnant women attending for routine and their conversion from narrow muscular vessels to wide
care in a well-defined gestational age, which is now widely nonmuscular channels. There is a wide spectrum in such
used for screening for chromosomal defects24,30; (2) we used impaired placentation and consequent clinical presentation of
a validated automated device and appropriately trained doc- the disease. Pathological studies reported that the prevalence
tors to measure BP23; (3) we used strict criteria to define of placental lesions in women with PE is inversely related to
MAP, PE, and GH27,28; and (4) we applied a statistical the gestational age at delivery.39,40 Similarly, Doppler ultra-
approach that is widely accepted in screening for trisomy 21 sound studies of the uterine arteries have demonstrated that
to examine the performance of screening and calculate the prevalence of increased impedance to flow in women
patient-specific risks.31 developing PE is inversely related to gestation at delivery.9,41
We found that combined first-trimester testing, for a 10% In contrast to Doppler, we found that the mean MoM MAP at
FPR, can identify ⬇60% of those who will develop PE 11⫹0 to 13⫹6 weeks in those developing PE did not change
several months later and 40% of those who will develop GH. significantly with gestation at delivery, and, therefore, mea-
Taking the maternal history and recording BP are the cheap- surement of MAP is equally effective in screening for early
est and most ubiquitously accessible screening tools. We and late disease. Such a finding provides some support for the
chose 11⫹0 to 13⫹6 weeks as the gestation for screening,
emerging evidence that there may be different etiologies for
because this is emerging as the first hospital visit of pregnant
early and late-onset PE, with the first being primarily because
women at which combined sonographic and biochemical
of impaired placentation and the second because of maternal
testing for chromosomal and other major defects is carried
hemodynamic maladaptation and/or impaired glucose metab-
out.24 At this visit, first, a record is made of maternal
olism.42,43 The extent to which measurement of BP could be
characteristics; second, an ultrasound scan is carried out to
combined with other sonographic and biochemical markers
determine the number of fetuses, confirm the gestation from
for more effective screening of both early and late PE remains
Downloaded from http://ahajournals.org by on September 10, 2023

the fetal CRL, exclude major defects, and measure the nuchal
to be determined.
translucency thickness and other first-trimester markers of
chromosomal defects; and, third, maternal blood is taken for
Perspectives
measurement of free ␤-human chorionic gonadotropin and
This study has established a methodology for the develop-
pregnancy-associated plasma protein A. It would be easy to
ment of a screening model in the detection of PE. Maternal
measure the MAP of women in this same visit and use the
history and BP constitute the cornerstones of prenatal care
same methodology to calculate the patient-specific risk for
and the foundation of any future methods for the estimation
both chromosomal defects and PE. Essentially, factors from
of patient-specific risk for the development of PE.
the maternal characteristics and history are used to calculate
the a priori risk, which is multiplied by the likelihood ratio
associated with biophysical and biochemical measurements
Source of Funding
This study was supported by a grant from the Fetal Medicine
to derive patient-specific risks. Foundation (United Kingdom charity No. 1037116).
In the unaffected group, which did not develop PE or GH,
MAP decreased with gestation, increased with maternal age Disclosures
and BMI, was lower in cigarette smokers and in all of the None.
ethnic groups other than in whites, and was higher in those
with a family or personal history of PE. The risk for References
developing PE increased with BMI and was higher in those of 1. World Health Organization. Make Every Mother and Child Count. World
Afro-Caribbean origin than in other ethnic groups and in Health Report, 2005. Geneva, Switzerland: World Health Organization;
2005.
those with a family or personal history of PE. The finding of 2. Lewis G, ed. Why Mothers Die 2000 –2002: The Sixth Report of Confi-
increased MAP with age and BMI and decreased MAP in dential Enquiries Into Maternal Deaths in the United Kingdom. London,
smokers is in agreement with previous reports in nonpregnant United Kingdom: RCOG Press; 2004.
individuals.32–34 The finding that a family and personal 3. American College of Obstetricians and Gynecologists Committee on
Practice Bulletins–Obstetrics. ACOG practice bulletin: diagnosis and
history of PE increases the risk of developing this disorder is management of pre-eclampsia and eclampsia: number 33, January 2002.
in agreement with previous studies reporting a 7-fold and Obstet Gynecol. 2002;99:159 –167.
3-fold increase in risk for PE in women with a respective 4. Villar J, Abdel-Aleem H, Merialdi M, Mathai M, Ali MM, Zavaleta N,
personal and family history of the disease.35 The additional Purwar M, Hofmeyr J, Nguyen TN, Campódonico L, Landoulsi S, Carroli
G, Lindheimer M; World Health Organization Calcium Supplementation
finding that women with such history have a higher MAP for the Prevention of Preeclampsia Trial Group. World Health Organi-
even if they do not develop PE is compatible with the zation Calcium Supplementation for the Prevention of Preeclampsia Trial
emerging evidence that a history of PE predisposes to the Group. Am J Obstet Gynecol. 2006;194:639 – 649.
 Poon et al First-Trimester Blood Pressure in Preeclampsia 1033

5. Rumbold AR, Crowther CA, Haslam RR, Dekker GA, Robinson JS, fetal nuchal translucency thickness at 10 –14 weeks of gestation. Lancet.
ACTS Study Group. Vitamins C and E and the risks of preeclampsia and 1998;351:343–346.
perinatal complications. N Engl J Med. 2006;354:1796 –1806. 25. Nicolaides KH. Nuchal translucency and other first-trimester sonographic
6. Poston L, Briley AL, Seed PT, Kelly FJ, Shennan AH. Vitamins in markers of chromosomal abnormalities. Am J Obstet Gynecol. 2004;191:
Pre-eclampsia (VIP) Trial Consortium. Lancet. 2006;367:1145–1154. 45– 67.
7. Askie LM, Duley L, Henderson-Smart DJ, Stewart LA, on behalf of the 26. Pickering TG, Hall JE, Appel LJ, Falkner BE, Graves J, Hill MN, Jones
PARIS Collaborative Group. Antiplatelet agents for prevention of pre- DW, Kurtz T, Sheps SG, Roccella EJ; Subcommittee of Professional and
eclampsia: a meta-analysis of individual patient data. Lancet. 2007;369: Public Education of the American Heart Association Council on High
1791–1798. Blood Pressure Research. Recommendations for blood pressure mea-
8. National Collaborating Centre for Women’s and Children’s Health. Com- surement in humans and experimental animals: part 1: blood pressure
missioned by the National Institute for Clinical Excellence. Antenatal measurement in humans: a statement for professionals from the Subcom-
Care Routine Care for the Healthy Pregnant Woman. Clinical Guideline. mittee of Professional and Public Education of the American Heart
October 2003. United Kingdom: National Collaborating Centre for Association Council on High Blood Pressure Research. Hypertension.
Women’s and Children’s Health; 2003. 2005;45:142–161.
9. Yu CK, Smith GCS, Papageorghiou AT, Cacho AM, Nicolaides KH. An 27. National Heart Foundation of Australia. Hypertension management guide
integrated model for the prediction of pre-eclampsia using maternal for doctors 2004. Available at: http://www.heartfoundation.org.au.
factors and uterine artery Doppler velocimetry in unselected low-risk Accessed April 1, 2006.
women. Am J Obstet Gynecol. 2005;193:429 – 436. 28. Davey DA, MacGillivray I. The classification and definition of the
10. Fallis NE, Langford HG. Relation of second trimester blood pressure to hypertensive disorders of pregnancy. Am J Obstet Gynecol. 1988;158:
toxemia of pregnancy in the primigravid patient. Am J Obstet Gynecol. 892– 898.
1963;87:123–125. 29. Gardosi J, Francis A. Software program for the calculation of customized
11. Page EW, Christianson R. The impact of mean arterial pressure in the birth weight percentiles. Version 6.2, 2000 –2007. Available at: http://
middle trimester upon the outcome of pregnancy. Am J Obstet Gynecol. www.gestation.net. Accessed May 14, 2007.
1976;125:740 –746. 30. United Kingdom National Screening Committee. Antenatal screening
12. Phelan JP. Enhanced prediction of pregnancy-induced hypertension by service for Down syndrome in England: 2001. A report to the United
combining supine pressor test with mean arterial pressure of middle Kingdom National Screening Committee. August 2002. Available at:
trimester. Am J Obstet Gynecol. 1977;129:397– 400. http://www.nsc.nhs.uk. Accessed October 2, 2007.
13. Quaas L, Robrecht D, Kaltenbach FJ. The mean arterial pressure versus 31. Royston P, Thompson SG. Model-based screening by risk with appli-
roll-over test as predictors of hypertensions in pregnancy. In: Samour MB, cation to Down’s syndrome. Stats in Med. 1992;11:257–268.
Symonds EM, Zuspan FP, El-Tomi N, eds. Pregnancy Hypertension. Cairo, 32. Rosenthal T, Oparil S. Hypertension in women. J Hum Hypertens. 2000;
Egypt: Ain Shams University Press; 1982:145–149. 14:691–704.
14. Öney T, Kaulhausen H. The value of the mean arterial blood pressure in the 33. Stamler R, Stamler J, Riedlinger WF, Algera G, Roberts RH. Weight and
second trimester (MAP-2 value) as a predictor of pregnancy-induced hyper- blood pressure. JAMA. 1978;240:1607–1610.
tension and preeclampsia. A preliminary report. Clin Exp Hypertens B. 34. Penney DG, Howley JW. Is there a connection between carbon monoxide
1983;2:211–216. exposure and hypertension? Environ Health Perspect. 1991;95:191–198.
15. Moutquin JM, Rainville C, Giroux L, Raynauld P, Amyot G, Bilodeau R, 35. Duckitt K, Harrington D. Risk factors for pre-eclampsia at antenatal
Pelland N. A prospective study of blood pressure in pregnancy: prediction booking: systematic review of controlled studies. BMJ. 2005;330:
Downloaded from http://ahajournals.org by on September 10, 2023

of preeclampsia. Am J Obstet Gynecol. 1985;151:191–196. 565–572.

16. Marya RK, Rathee S, Mittal R. Evaluation of three clinical tests for 36. Sibai BM, El-Nazer A, Gonzalez-Ruiz A. Severe preeclampsia-eclampsia
predicting pregnancy-induced hypertension. Am J Obstet Gynecol. 1988; in young gravid women: subsequent pregnancy outcome and remote
158:683– 684. prognosis. Am J Obstet Gynecol. 1986;155:1011–1016.
17. Villar MA, Sibai BM. Clinical significance of elevated mean arterial 37. Manten GT, Sikkema MJ, Voorbij HA, Visser GH, Bruinse HW, Franx
blood pressure in second trimester and threshold increase in systolic or A. Risk factors for cardiovascular disease in women with a history of
diastolic blood pressure during third trimester. Am J Obstet Gynecol. pregnancy complicated by preeclampsia or intrauterine growth restriction.
1989;160:419 – 423. Hypertens Pregnancy. 2007;26:39 –50.
18. Ales KL, Norton ME, Druzin ML. Early prediction of antepartum hyper- 38. Eskenazi B, Fenster L, Sidney S. A multivariate analysis of risk factors
tension. Obstet Gynecol. 1989;73:928 –933. for preeclampsia. JAMA. 1991;266:237–241.
19. Conde-Agudelo A, Belizán JM, Lede R, Bergel EF. What does an 39. Moldenhauer JS, Stanek J, Warshak C, Khoury J, Sibai B. The frequency
elevated mean arterial pressure in the second half of pregnancy predict- and severity of placental findings in women with pre-eclampsia are
gestational hypertension or preeclampsia? Am J Obstet Gynecol. 1993; gestational age dependent. Am J Obstet Gynecol. 2003;189:1173–1177.
169:509 –514. 40. Egbor M, Ansari T, Morris N, Green CJ, Sibbons PD. Morphometric
20. Miller RS, Rudra CB, Williams MA. First-trimester mean arterial placental villous and vascular abnormalities in early- and late-onset pre-
pressure and risk of preeclampsia. Am J Hypertens. 2007;20:573–578. eclampsia with and without fetal growth restriction. BJOG. 2006;113:
21. Mion D, Pierin AM. How accurate are sphygmomanometers? J Hum 580 –589.
Hypertens. 1998;12:245–248. 41. Plasencia W, Maiz N, Bonino S, Kaihura C, Nicolaides KH. Uterine
22. Markandu ND, Whitcher F, Arnold A, Carney C. The mercury sphyg- artery Doppler at 11⫹0 to 13⫹6 weeks in the prediction of pre-eclampsia.
momanometer should be abandoned before it is proscribed. J Hum Ultrasound Obstet Gynecol. 2007;30:742–749.
Hypertens. 2000;14:31–36. 42. Bosio PM, McKenna PJ, Conroy R, O’Herlihy C. Maternal central he-
23. Reinders A, Cuckson AC, Lee JT, Shennan AH. An accurate automated modynamics in hypertensive disorders or pregnancy. Obstet Gynecol.
blood pressure device for use in pregnancy and pre-eclampsia: the 1999;94:978 –984.
Microlife 3BTO-A. BJOG. 2005;112:915–920. 43. D’Anna R, Baviera G, Corrado F, Giordano D, De Vivo A, Nicocia G, Di
24. Snijders RMJ, Noble P, Sebire N, Souka A, Nicolaides KH. UK multi- Benedetto A. Adiponectin and insulin resistance in early- and late-onset
centre project on assessment of risk of trisomy 21 by maternal age and pre-eclampsia. BJOG. 2006;113:1264 –1269.