UvA-DARE (Digital Academic Repository) … · Hemodynamics and maternal characteristics prior to...

UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Hemodynamics and maternal characteristics prior to hypertensive disorders ofpregnancyVollebregt, K.C.

Link to publication

Citation for published version (APA):Vollebregt, K. C. (2011). Hemodynamics and maternal characteristics prior to hypertensive disorders ofpregnancy

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: http://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 01 Aug 2018

http://dare.uva.nl/personal/pure/en/publications/hemodynamics-and-maternal-characteristics-prior-to-hypertensive-disorders-of-pregnancy(349805d2-f2b8-4649-b354-c173c6cd660f).html

Hemodynamics and maternal characteristics prior to hypertensive disorders of pregnancy

The printing of this thesis was sponsored by:

Department of Obstetrics and Gynecology, AMC, Amsterdam, The Netherlands

BMEYE, Amsterdam, The Netherlands


Thesis, University of Amsterdam, The Netherlands

Copyright© 2010. K.C. Vollebregt, Amsterdam, The Netherlands. All rights reserved.

No parts of this thesis may be reproduced or transmitted in any form or by any means,

without prior permission of the author.

Cover: Kees Boer “Het relatieve belang van spiralen voor de zwangerschap”

Lay-Out: Chris Bor, Medische Fotografie en Illustratie, AMC, Amsterdam,

The Netherlands

Printed by: Buijten & Schipperheijn, Amsterdam, The Netherlands

ISBN: 978-90-9025939-0


ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor

aan de Universiteit van Amsterdam

op gezag van de Rector Magnificus

prof. dr. D.C. van den Boom

ten overstaan van een door het college

voor promoties ingestelde commissie,

in het openbaar te verdedigen

in de Agnietenkapel op

vrijdag 4 maart 2011

te 12:00 uur

door

Karlijn Corien VollebregtGeboren te Bussum

Promotiecommissie

Promotor: Prof. dr. J.A.M. van der Post

Co-promotores: Dr. K. Boer

Dr. H. Wolf

Overige leden: Prof. dr. E.T. van Bavel

Dr. J.M. Karemaker

Prof. dr. B.W.J. Mol

Prof. dr. B.J.M. Mulder

Dr. M.E.A. Spaanderman

Prof. dr. E.A.P. Steegers

Faculteit der Geneeskunde

CONTENTS

Chapter 1 General introduction and outline of the thesis 7

Chapter 2 Accuracy of mean arterial pressure and blood pressure measurements in predicting preeclampsia: systematic review and meta-analysis. BMJ 2008;336(7653):1117-20

19

Chapter 3 Limited accuracy of the hyperbaric index, ambulatory blood pressure and sphygmomanometry measurements in predicting gestational hypertension and preeclampsia.Journal of Hypertension 2010;28(1):127-134

53

Chapter 4 The association of three different techniques to measure blood pressure in the first trimester of pregnancy with later preeclampsia.Submitted

69

Chapter 5 Arterial stiffness at 8 0/7-11 0/7 weeks of pregnancy and the association with later preeclampsia.Submitted

85

Chapter 6 Is psychosocial stress in first ongoing pregnancies associated with preeclampsia and gestational hypertension?BJOG 2008;115(5):607-615

99

Chapter 7 Does physical activity in leisure time early in pregnancy reduce the incidence of preeclampsia or gestational hypertension?Acta Obstetricia et Gynecologica Scandinavica 2010;89(2):261-269

117

Chapter 8 Sensitivity of spontaneous baroreflex control of the heart and hemodynamic parameters are not influenced by the menstrual cycle.Hypertension in Pregnancy 2006;25(3):159-167

131

Chapter 9 Summary, conclusion and recommendations for future researchSamenvatting, conclusies en aanbevelingen

143161

Appendices Authors’affiliations 179

Bibliography 183

Dankwoord 185

Curriculum vitae 191

11 Introduction

Chapter 1

9

General Introduction

INTRODUCTION

Preeclampsia and gestational hypertension are major causes of maternal and fetal

morbidity as well as mortality worldwide.1;2 In the Netherlands 8.3% of all pregnancies in

2006 was affected by preeclampsia or gestational hypertension.3 The majority of these

women were pregnant with their first child. During pregnancy or postpartum, women

with hypertensive disorders of pregnancy and their babies are at risk of complications

like elective preterm birth, fetal growth retardation, perinatal death, HELLP syndrome,

eclampsia, renal insufficiency, cerebral bleeding or maternal mortality.4 Additionally

women with previous preeclampsia have an elevated risk for cardiovascular disease in

later life.5;6 Being small for gestational age at birth is associated with future cardiovascular

and metabolic disorders.7

Hypertensive disorders of pregnancy present a large variation in phenotype and severity

of disease. These vary from gestational hypertension with moderate hypertension after

20 weeks of gestation in a previously normotensive woman, to preeclampsia (gestational

hypertension with proteinuria), hemolysis, elevated liver enzymes, low platelets (HELLP)

syndrome to eclampsia.8 Generally the disease is more severe if it occurs early in

pregnancy and if it is combined with physical signs and symptoms that are associated

with preeclampsia.4 At present preeclampsia is unpredictable in onset and progression.

The only effective treatment is termination of pregnancy.4 If preeclampsia starts early in

pregnancy, maternal risks have to be balanced against neonatal risks caused by elective

preterm birth.

Despite extensive research the pathogenesis of preeclampsia remains unclear. A widely

accepted hypothesis is that a disturbed placental function early in pregnancy causes a

systemic inflammatory response characterized by generalized endothelial dysfunction

leading to the maternal syndrome in the second half of pregnancy.9 Inadequate placentation

may be the result of vascular, immunological, genetic or environmental factors.

Prevention of preeclampsia is hardly possible. Based on large meta-analyses there is

evidence that low-dose aspirin reduces the risk by approximately 10%.10;11 This treatment

is recommended in women with moderate to high risk for preeclampsia.12 In women with

low baseline calcium intake calcium supplementation reduces the risk of preeclampsia.13

Trials investigating possible strategies to reduce the incidence of severe preeclampsia and

gestational hypertension are hampered by the low incidence of the disease. Predictive

tests identifying high risk women could facilitate such research. Because placentation

10

starts early in pregnancy it seems likely that early interventions will be the most effective.

Identification of a high risk group early in pregnancy is therefore relevant. Up to now

there is no early validated predictive test with sufficient accuracy, that is useful in daily

clinical practice.14

Several risk factors have been identified with respect to preeclampsia. These risk factors

can be pregnancy specific, like parity or multiple pregnancy, or not pregnancy specific

and are shared with risk factors of adult cardiovascular disease. Among these are

maternal physical characteristics like obesity, age and pre-existing vascular diseases like

hypertension and diabetes.4 Other known risk factors for cardiovascular disease like

psychosocial stress might also influence the incidence of preeclampsia.15;16 It is assumed

that psychosocial stress causes excessive sympathetic nervous system activation.17Several

studies describe a higher sympathetic activity in women with preeclampsia, compared

to women with an uncomplicated pregnancy.18;19 However, it is not clear if sympathetic

hyperactivity in preeclampsia is influenced by psychosocial stress.20-22

Protective or disease modifying interventions that reduce the risk of cardiovascular disease

might be useful for the prevention of hypertensive disorders in pregnancy. Physical activity

is one such potential factor, known to prevent non-pregnant cardiovascular disease and

to reduce its symptoms. Its precise mechanism is unclear. An acceptable hypothesis

is that it reduces oxidative stress, which in turn influences the immune response and

improves endothelial function.23-26 These processes are also involved in the development

of preeclampsia and gestational hypertension.4 Literature regarding physical activity and

preeclampsia shows conflicting results.27-29

By studying the maternal cardiovascular system in pregnancy, one understands why factors

that regulate blood pressure might be predictive of preeclampsia. Already in the very

beginning of pregnancy, before 12 weeks gestational age, the maternal cardiovascular

system changes. Cardiac output increases and peripheral vascular resistance decreases.

Total circulating plasma volume increases by 40% compared with non-pregnant

controls. Blood pressure decreases until mid-pregnancy and rises again till the end of

pregnancy.30;31 Women with severe preeclampsia not only have hypertension but also a

reduced cardiac output, an increased vascular resistance, an increased arterial stiffness

and a diminished circulating plasma volume.32;33

Together with proteinuria an increased blood pressure is the obligatory symptom of the

syndrome preeclampsia. Women predisposed to develop preeclampsia already have an

increased blood pressure early in pregnancy compared to women who do not develop

preeclampsia or gestational hypertension, long before clinical symptoms arise. Although

Chapter 1

11


they do not meet the criteria for hypertension yet, their blood pressure measured in the

beginning of pregnancy can be used to predict preeclampsia.34-36 Blood pressure can be

measured using different methods and devices. In the first half of the eighteenth century

blood pressure was measured for the first time invasively in a mare by Stephan Hales. Until

1905 it was not possible to measure blood pressure non-invasively with a simple device

suitable for daily clinical practice. It was Nicolai Korotkoff who combined the mercury

manometer developed by Riva-Rocci and a stethoscope to register the occurrence and

disappearance of vascular sounds on the pressure release in an occluding upper arm

cuff.37 This method is considered as golden standard for non-invasive blood pressure

measurements and criteria have been described to standardize the procedure.38 Mercury

sphygmomanometry is used to diagnose preeclampsia and gestational hypertension.8

In the Netherlands the use of mercury sphygmomanometers in hospitals is not allowed

because of mercury toxicity. They are replaced by aneroid sphygmomanometers.

Nowadays several automated devices can be used to measure blood pressure, based

on registration of vascular oscillations or vascular sounds during cuff pressure release or

peripheral vascular pressure. However, many of these devices have not been validated

for the use in pregnancy.

Sphygmomanometry as described by Korotkoff and oscillometry register vascular

oscillations during the gradual relaxation of vascular occlusion in the upper arm. The

appearance and disappearance of vascular sounds determine systolic and diastolic

blood pressure and the mean arterial pressure is calculated as ((2x DBP) + SBP)/3.

In oscillometry the mean arterial pressure is estimated primarily by the pressure where

oscillations are largest. Systolic and diastolic blood pressures are derived from this

point. Sphygmomanometry registers systolic from one and diastolic blood pressure

from another heart beat cycle. In oscillometry the envelope of all oscillations during cuff

pressure release is registered, but the single maximum oscillation denotes mean arterial

pressure.39 Both techniques are influenced by arm diameter and vascular deformability.

Oscillometry is often used in automated devices suitable for ambulatory blood pressure

measurement (ABPM). In this way information about out of office blood pressure and

heart rate can be obtained. A more sophisticated approach is the calculation of the

Hyperbaric Index (HBI) using ABPM of at least 24 hours. The HBI is the individual amount

of blood pressure excess during the measurement period above a 90% tolerance limit.40

Non-invasive finger arterial waveform registration enables continuous blood pressure

registration. One of the advantages of this method is that not only blood pressure

but also heart rate, cardiac output, total peripheral resistance, arterial stiffness and

autonomic cardiovascular control can be estimated. It is not clear if there are differences

in predictive capacity between the different techniques and devices.

12

The outcome of studies using blood pressure for the prediction of hypertensive disorders

is generally disappointing. Addition of maternal characteristics and other variables like

arterial stiffness seems to improve the predictive accuracy for preeclampsia.41-43 Of

these maternal characteristics, the most predictive variable is the outcome of a former

pregnancy. Nearly half of all pregnant women are nulliparous and 2/3 of all cases of

preeclampsia occur in nulliparous women. Therefore this parameter cannot be used to

predict preeclampsia in these women.

AIM OF THIS THESIS

To investigate if differences in cardiovascular parameters can be observed early in

pregnancy between women, who will develop preeclampsia or gestational hypertension

later in pregnancy and women with an uncomplicated pregnancy. Moreover, if these

differences could be used to develop a screening strategy with a focus on nulliparous

women. Additionally we wanted to establish if modifiable risk factors for cardiovascular

disease like psychosocial stress and physical activity are also associated with preeclampsia.

Specifically we wanted to answer the following questions:

l Can blood pressure measured in the beginning of pregnancy be used as a screening

test for preeclampsia?

l Is there a difference in predictive capacity between different methods of measuring

blood pressure?

l Is arterial stiffness increased in the beginning of pregnancy in women who will develop

preeclampsia later in pregnancy and is this associated with the disease?

l Is psychosocial stress in the first half of pregnancy associated with later preeclampsia?

l Can physical exercise in leisure time in the first half of pregnancy reduce the incidence

of preeclampsia?

l Does the menstrual cycle influence blood pressure and the spontaneous baroreflex

control of the heart?

Chapter 1

13


OUTLINE OF THE THESIS

Part 1. Blood pressure measurement as a screening test

A meta-analysis of the value of blood pressure measurements as a screening test for

preeclampsia is presented (Chapter 2). The two following chapters explore if automated

blood pressure measurements have a stronger association with later hypertensive

disorders of pregnancy than conventional sphygmomanometry measurement. In a

prospective cohort of healthy nulliparous women and women with a previous pregnancy

complicated by preeclampsia blood pressure is measured in the first trimester with 3

different techniques; conventional sphygmomanometry, non-invasive continuous finger

arterial pressure waveform registration and ABPM. In chapter 3 we prospectively evaluate

the Hyperbaric Index, calculated from a 48 hour ABPM as a possible screening test in

the first trimester for preeclampsia or gestational hypertension. In addition we compare

the predictive capacity of the HBI with conventional sphygmomanometry and a 48 hour

ABPM. In chapter 4 the associations between blood pressure and later preeclampsia or

gestational hypertension are estimated for the three measurement techniques.

Part 2. Association of preeclampsia with arterial stiffness

Using the same cohort as in chapter 3, arterial stiffness was measured in healthy

nulliparous women and multiparous former preeclamptic women with continuous finger

arterial pressure waveform analysis. We analyze if this could enable the selection of

women at risk for developing preeclampsia later in pregnancy (Chapter 5).

Part 3. Maternal characteristics and their association with preeclampsia

Between January 2003 and March 2004 all pregnant women who received their

antenatal care by a midwife or gynecologist in Amsterdam were asked to participate

in the Amsterdam Born Children and their Development Study (ABCD-study). All

participants filled in a questionnaire on sociodemographic characteristics, medical

history, psychosocial stress and lifestyle. We explore data of all nulliparous women with

a singleton pregnancy in this cohort, to assess the association of psychosocial stress

(chapter 6) or physical activity in leisure time (chapter 7) with preeclampsia. Calculations

are adjusted for a number of well known risk factors like age, body mass index, smoking

and pre-existing medical conditions.

14

Part 4. The influence of the menstrual cycle on hemodynamics and cardiovascular control

Although differences in hemodynamics and cardiovascular regulation in pregnancy are

found between uncomplicated pregnancies and pregnancies with hypertensive disorders

not much is known about measurements before pregnancy and the individual alterations

in women. The influence of the menstrual cycle on the spontaneous baroreflex control of

the heart and hemodynamic parameters is investigated (Chapter 8).

Chapter 1

15


Reference List 1. Khan KS, Wojdyla D, Say L, Gulmezoglu AM, Van Look PF. WHO analysis of causes of maternal

death: a systematic review. Lancet 2006;367:1066-74. 2. Schutte JM, Steegers EA, Schuitemaker NW, Santema JG, de BK, Pel M et al. Rise in maternal

mortality in the Netherlands. BJOG. 2010;117:399-406. 3. Stichting Perinatale Registratie Nederland. Perinatale zorg in Nederland 2006. Utrecht: Stichting

Perinatale Regsitratie Nederland, 2008. 4. Steegers EA, von DP, Duvekot JJ, Pijnenborg R. Pre-eclampsia. Lancet 2010;376:631-44. 5. Manten GT, Sikkema MJ, Voorbij HA, Visser GH, Bruinse HW, Franx A. Risk factors for

cardiovascular disease in women with a history of pregnancy complicated by preeclampsia or intrauterine growth restriction. Hypertens.Pregnancy. 2007;26:39-50.

6. Lykke JA, Langhoff-Roos J, Sibai BM, Funai EF, Triche EW, Paidas MJ. Hypertensive pregnancy disorders and subsequent cardiovascular morbidity and type 2 diabetes mellitus in the mother. Hypertension 2009;53:944-51.

7. Barker DJ. Maternal nutrition, fetal nutrition, and disease in later life. Nutrition 1997;13:807-13.

8. Brown MA, Lindheimer MD, de Swiet M, Van Assche A, Moutquin JM. The classification and diagnosis of the hypertensive disorders of pregnancy: statement from the International Society for the Study of Hypertension in Pregnancy (ISSHP). Hypertens.Pregnancy. 2001;20:IX-XIV.

9. Sibai B, Dekker G, Kupferminc M. Pre-eclampsia. Lancet 2005;365:785-99. 10. Askie LM, Duley L, Henderson-Smart DJ, Stewart LA. Antiplatelet agents for prevention of

pre-eclampsia: a meta-analysis of individual patient data. Lancet 2007;369:1791-98. 11. Duley L, Henderson-Smart DJ, Meher S, King JF. Antiplatelet agents for preventing pre-eclampsia

and its complications. Cochrane.Database.Syst.Rev. 2007;CD004659. 12. NICE clinical guideline. Hypertension in pregnancy: the management of hypertensive disorders

during pregnancy. National Collaborating Centre for Women’s and Children’s Health 2010. 13. Hofmeyr GJ, Lawrie TA, Atallah AN, Duley L. Calcium supplementation during pregnancy

for preventing hypertensive disorders and related problems. Cochrane.Database.Syst.Rev. 2010;CD001059.

14. Cnossen JS, Ter RG, Mol BW, van der Post JA, Leeflang MM, Meads CA et al. Are tests for predicting pre-eclampsia good enough to make screening viable? A review of reviews and critical appraisal. Acta Obstet.Gynecol.Scand. 2009;88:758-65.

15. Rozanski A, Blumenthal JA, Kaplan J. Impact of psychological factors on the pathogenesis of cardiovascular disease and implications for therapy. Circulation 1999;99:2192-217.

16. Anand SS, Islam S, Rosengren A, Franzosi MG, Steyn K, Yusufali AH et al. Risk factors for myocardial infarction in women and men: insights from the INTERHEART study. Eur.Heart J. 2008;29:932-40.

17. Lucini D, Di FG, Parati G, Pagani M. Impact of chronic psychosocial stress on autonomic cardiovascular regulation in otherwise healthy subjects. Hypertension 2005;46:1201-06.

18. Schobel HP, Fischer T, Heuszer K, Geiger H, Schmieder RE. Preeclampsia -- a state of sympathetic overactivity. N.Engl.J.Med. 1996;335:1480-85.

19. Greenwood JP, Stoker JB, Walker JJ, Mary DA. Sympathetic nerve discharge in normal pregnancy and pregnancy-induced hypertension. J.Hypertens. 1998;16:617-24.

20. Paarlberg KM, Vingerhoets AJ, Passchier J, Dekker GA, van Geijn HP. Psychosocial factors and pregnancy outcome: a review with emphasis on methodological issues. J.Psychosom.Res. 1995;39:563-95.

21. Klonoff-Cohen HS, Cross JL, Pieper CF. Job stress and preeclampsia. Epidemiology 1996;7:245-49. 22. Marcoux S, Berube S, Brisson C, Mondor M. Job strain and pregnancy-induced hypertension.

Epidemiology 1999;10:376-82.

16

23. Mora S, Cook N, Buring JE, Ridker PM, Lee IM. Physical activity and reduced risk of cardiovascular events: potential mediating mechanisms. Circulation 2007;116:2110-18.

24. Dickinson HO, Mason JM, Nicolson DJ, Campbell F, Beyer FR, Cook JV et al. Lifestyle interventions to reduce raised blood pressure: a systematic review of randomized controlled trials. J.Hypertens. 2006;24:215-33.

25. Kojda G, Hambrecht R. Molecular mechanisms of vascular adaptations to exercise. Physical activity as an effective antioxidant therapy? Cardiovasc.Res. 2005;67:187-97.

26. Moyna NM, Thompson PD. The effect of physical activity on endothelial function in man. Acta Physiol Scand. 2004;180:113-23.

27. Sorensen TK, Williams MA, Lee IM, Dashow EE, Thompson ML, Luthy DA. Recreational physical activity during pregnancy and risk of preeclampsia. Hypertension 2003;41:1273-80.

28. Magnus P, Trogstad L, Owe KM, Olsen SF, Nystad W. Recreational physical activity and the risk of preeclampsia: a prospective cohort of Norwegian women. Am.J.Epidemiol. 2008;168:952-57.

29. Osterdal ML, Strom M, Klemmensen AK, Knudsen VK, Juhl M, Halldorsson TI et al. Does leisure time physical activity in early pregnancy protect against pre-eclampsia? Prospective cohort in Danish women. BJOG. 2009;116:98-107.

30. Duvekot JJ, Peeters LL. Maternal cardiovascular hemodynamic adaptation to pregnancy. Obstet.Gynecol.Surv. 1994;49:S1-14.

31. Spaanderman ME, Meertens M, van BM, Ekhart TH, Peeters LL. Cardiac output increases independently of basal metabolic rate in early human pregnancy. Am.J.Physiol Heart Circ.Physiol 2000;278:H1585-H1588.

32. Visser W, Wallenburg HC. Central hemodynamic observations in untreated preeclamptic patients. Hypertension 1991;17:1072-77.

33. Spasojevic M, Smith SA, Morris JM, Gallery ED. Peripheral arterial pulse wave analysis in women with pre-eclampsia and gestational hypertension. BJOG. 2005;112:1475-78.

34. Moutquin JM, Rainville C, Giroux L, Raynauld P, Amyot G, Bilodeau R et al. A prospective study of blood pressure in pregnancy: prediction of preeclampsia. Am.J.Obstet.Gynecol. 1985;151:191-96.

35. Hermida RC, Ayala DE, Mojon A, Fernandez JR, Silva I, Ucieda R et al. Blood pressure excess for the early identification of gestational hypertension and preeclampsia. Hypertension 1998;31:83-89.

36. Sibai BM, Ewell M, Levine RJ, Klebanoff MA, Esterlitz J, Catalano PM et al. Risk factors associated with preeclampsia in healthy nulliparous women. The Calcium for Preeclampsia Prevention (CPEP) Study Group. Am.J.Obstet Gynecol. 1997;177:1003-10.

37. Booth J. A short history of blood pressure measurement. Proc.R.Soc.Med. 1977;70:793-99. 38. O’Brien E, Asmar R, Beilin L, Imai Y, Mallion JM, Mancia G et al. European Society of

Hypertension recommendations for conventional, ambulatory and home blood pressure measurement. J.Hypertens. 2003;21:821-48.

39. Ramsey M, III. Noninvasive automatic determination of mean arterial pressure. Med.Biol.Eng Comput. 1979;17:11-18.

40. Halberg F, Ahlgren A, Haus E. Circadian systolic and diastolic hyperbaric indices of high school and college students. Chronobiologia 1984;11:299-309.

41. Duckitt K, Harrington D. Risk factors for pre-eclampsia at antenatal booking: systematic review of controlled studies. BMJ 2005;330:565.

42. Poon LC, Karagiannis G, Leal A, Romero XC, Nicolaides KH. Hypertensive disorders in pregnancy: screening by uterine artery Doppler imaging and blood pressure at 11-13 weeks. Ultrasound Obstet Gynecol. 2009;34:497-502.

43. Khalil A, Cowans NJ, Spencer K, Goichman S, Meiri H, Harrington K. First-trimester markers for the prediction of pre-eclampsia in women with a-priori high risk. Ultrasound Obstet.Gynecol. 2010;35:671-79.

12Jeltsje S. Cnossen Karlijn C. Vollebregt Nynke de Vrieze Gerben ter Riet Ben W.J. Mol Arie FranxKhalid S. Khan Joris A.M. van der Post

BMJ. 2008; 336(7653):1117-20

Accuracy of mean arterial pressure and blood pressure measurements in predicting preeclampsia: systematic review and meta-analysis

20

ABSTRACT

Objective To determine the accuracy of using systolic and diastolic blood pressure, mean arterial

pressure, and increase of blood pressure to predict preeclampsia.

Design

Systematic review with meta-analysis of data on test accuracy.

Data sources

Medline, Embase, Cochrane Library, MEDION, reference list checking of included articles

and reviews, contact with authors.

Review methods

Without language restrictions, two reviewers independently selected the articles in

which the accuracy of blood pressure measurement during pregnancy was evaluated to

predict preeclampsia. Data were extracted on study characteristics, quality and results

to construct 2x2 tables. Summary receiver operating characteristic curves and likelihood

ratios were generated for the various levels and their thresholds.

Results

34 studies, testing 60,599 women (3341 cases of preeclampsia) were included. In

women at low risk for preeclampsia, the areas under the summary receiver operating

characteristic curves for blood pressure measurement in the second trimester were

0.68 (95% confidence interval 0.64 to 0.72) for systolic blood pressure, 0.66 (0.59 to

0.72) for diastolic blood pressure, and 0.76 (0.70 to 0.82) for mean arterial pressure.

Findings for the first trimester showed a similar pattern. Second trimester mean arterial

pressure of 90 mmHg or more showed a positive likelihood ratio of 3.5 (95% confidence

interval 2.0 to 5.0) and a negative likelihood ratio of 0.46 (0.16 to 0.75). In women

deemed to be at high risk, a diastolic blood pressure of 75 mmHg or more at 13 to 20

weeks’ gestation best predicted preeclampsia: positive likelihood ratio 2.8 (1.8 to 3.6),

negative likelihood ratio 0.39 (0.18 to 0.71). Additional subgroup analyses did not show

improved predictive accuracy.

Conclusion

When blood pressure is measured in the first or second trimester of pregnancy, the

mean arterial pressure is a better predictor for preeclampsia than systolic blood pressure,

diastolic blood pressure or an increase of blood pressure.

Chapter 2

21

Blood pressure and preeclampsia; a m

eta-analysis

INTRODUCTION

Preeclampsia is an important disorder of pregnancy, with potentially severe consequences

for mother and child.1;2 The frequency of preeclampsia varies between 2% and 7% in

healthy nulliparous women.3-5 In preeclampsia, the interaction between the placenta

and maternal constitution is influenced by genetic and environmental factors, causing a

hypertensive inflammatory response.6 The gestosis is associated with low birth weight and

preterm delivery but can also develop at term, during labour, or even postpartum.2 It has

been shown that women destined to develop preeclampsia have higher mean arterial

pressures in the first and second trimester and even before pregnancy than women with

normal pregnancies.7;8

Blood pressure measurement is a screening test that is routinely used in antenatal

care to detect or predict hypertensive disease.2 Accurate prediction of women at risk

for preeclampsia is crucial to judicious allocation of monitoring resources and use of

preventive treatment9, with the prospect of improving maternal and neonatal outcome.

Studies investigating the predictive accuracy of blood pressure measurement report

conflicting results. In view of these conflicting reports, it is uncertain whether blood

pressure measurement should be used routinely as a predictive test or used only to

diagnose hypertensive disorders in pregnancy once they are suspected. We carried

out a systematic review to investigate the accuracy of blood pressure measurement for

prediction of preeclampsia in pregnant women.

METHODS

This review was based on our previously published protocol.10 Librarians carried out

two electronic searches covering Medline, Embase, Cochrane Library (2006;4), and

Medion (www.mediondatabase.nl/) from inception to February 2007. The search strategy

(Appendix 1) consisted of MeSH or key terms related to the disease (preeclampsia) and

population (pregnant women) combined with methodological filters for identification

of studies on diagnostic tests and causal factors.11-13 A second search strategy was

done to ascertain retrieval of relevant papers because ‘blood pressure’ is a term both

related to our index test blood pressure measurement and the outcome preeclampsia.

We checked reference lists of relevant studies to identify cited articles not captured by

electronic searches and contacted authors of primary studies who had e-mail addresses

available. No language restrictions were applied. We included studies that reported on

any technique to measure blood pressure in pregnant women in any healthcare setting

22

and of any level of risk for preeclampsia. We included test accuracy studies allowing

generation of 2x2 tables.

Trained reviewers independently screened titles and abstracts for relevance (J.S.C. and

K.C.V.), and full papers for inclusion and data extraction (JSC and NdV). An explanation

of extracted clinical, methodological and statistical data has been published.10 Studies

were assessed by one reviewer (J.S.C.) for methodological quality against the quality

assessment of diagnostic accuracy studies (QUADAS) criteria14 and randomly checked

by a second reviewer (NdV). How we assessed items is summarised in Appendix 2. We

also assessed application of preventive treatment. For multiple publication of one data

set we included only the most recent or complete study. Disagreements were resolved by

consensus or a third reviewer.

Reference standards for preeclampsia were a persistent systolic blood pressure of 140

mmHg or more or diastolic blood pressure of 90 mmHg or more, or both, with proteinuria

of 0.3 g/ day or more or a dipstick result of + or more (30 mg/dl in a single urine

sample), or both, new after 20 weeks of gestation. Severe preeclampsia was defined as

systolic blood pressure of 160 mmHg or more or diastolic blood pressure of 110 mmHg

or more, or both, with proteinuria of 2.0 g/ day or more or a dipstick result of +++ or

more, or both, or of early onset (< 34 weeks) gestation. Superimposed preeclampsia

was defined as the development of proteinuria of 0.3 g/ day or more or a dipstick result

of + or more after 20 weeks of gestation in chronically hypertensive patients. Chronic

hypertension was defined as hypertension present before pregnancy or detected before

20 weeks gestation and not resolving within three months after delivery.15

Data synthesis

From the 2x2 tables we calculated sensitivity and specificity and plotted their results

in receiver operating characteristics plots. We pooled results among groups of studies

measuring similar blood pressure variables (systolic, diastolic, mean arterial, or increase)

and similar outcome. We used a bivariate regression model that takes into account the

negative correlation between sensitivity and specificity. This method has been extensively

described elsewhere and is recently recommended for meta-analysis of diagnostic

tests.16;17 Briefly, rather than using a single outcome measure per study, such as the

diagnostic odds ratio, the bivariate model preserves the two-dimensional nature of

diagnostic data in a single model. This model incorporates the correlation that may exist

between sensitivity and specificity within studies owing to possible differences in threshold

between studies. When necessary, the bivariate model uses a random effects approach for

both sensitivity and specificity, allowing for heterogeneity beyond chance due to clinical

or methodological differences between studies. In addition the model acknowledges the

Chapter 2

23


eta-analysis

difference in precision by which sensitivity and specificity have been measured in each

study. This means that studies with a larger number of women with preeclampsia receive

more weight in the calculation of the pooled estimate of sensitivity, whereas studies with

more women without preeclampsia are more influential in the pooling of specificity.

To estimate a summary receiver operating characteristic curve and an area under that

curve, we used the results of the model with the smallest Akaike’s Information Criterion

(a measure of the goodness of fit of an estimated statistical model).18 This model best

accounts for heterogeneity between studies. We calculated areas under the curves in the

first trimester for the complete summary receiver operating characteristic curve, which

meant extension beyond the available data. We also calculated pooled sensitivities and

specificities and derived likelihood ratios thereof.19

We carried out the following subgroup analyses, defined a priori: outcome (severe

preeclampsia), sample (low-risk sample v high-risk sample), number of readings (multiple

measurements v single measurement, and gestational age at testing (first trimester v

second trimester). Sensitivity analyses were done for application of preventive treatment

and study quality. We considered studies of high quality when they scored positive on

at least four of the following items: prospective design with consecutive recruitment,

appropriate reference standard, follow-up greater than 90%, adequate description of the

index test, and reporting of preventive treatment.

All statistical analyses were done using SAS 9.1 for Windows (Proc NLMixed in the bivariate

model). We used STATA/SE 9.0 0 for drawing the receiver operating characteristic plots.

RESULTS

Figure 1 summarises the flow of studies through the review. Overall, 34 studiesw1-w34

screening 60,599 women (3341 preeclamptic women) were included, which comprised

28 cohorts, three randomised controlled trials and three case-control studies resulting in

135 2x2 tables. None of 12 authors contacted about unclear information in their articles

provided additional information. The main reason given was “data not available”.

Twenty-eight studies were prospective and six retrospective. The risk profiles of the women

included in the studies are shown in Appendix 3. Sample sizes ranged from 22 to 22,582

women. The incidence rates for preeclampsia ranged from 0.8 to 40.8% (median

incidence 6.3%; case control studies excluded). Seven studies reported on “high risk”

women (range for incidence rates 2.0 to 28.7%). Since the criteria for high risk varied

from women with an abnormal uterine artery detected by Doppler ultrasonography to

women with chronic hypertension or diabetes, this may have caused the great variation

24

in incidence. Fourteen studies explicitly reported exclusion of chronic hypertension and

four studies reported inclusion of chronic hypertension (women at high risk). Twenty

eight studies reported on blood pressure measurement in the second trimester. Eighteen

studies reported on mean arterial pressure, defined as (diastolic blood pressure + one

third x (systolic blood pressure – diastolic blood pressure)) or automated mean arterial

pressure readings. Six studies reported on systolic blood pressure, 11 on diastolic blood

pressure, and three on increase of systolic blood pressure or increase of diastolic blood

pressure, or both. One studyw33 reported on diastolic midline estimating statistic of rhythm

(MESOR) and two studiesw28;w31 reported on absolute blood pressure levels. Since both

randomised controlled trials in women at low risk investigated calcium supplementation,

which has not been proved effective,20 and one only included the placebo group, we

analysed the trials together with the cohort studies. Four studies were excluded from the

final meta-analysis (three case control studiesw2;w15;w20 to further enhance validity and

one study without threshold for the index testw26).

Quality assessment

Figure 2 summarises the results of quality assessment. Over 70% of studies met the

following items for quality assessment: period between tests, avoidance of partial and

Primary articles retrieved for detailed evaluation (n= 186) - from electronic searches (n= 182) - from reference lists (n= 4)

Articles excluded (n= 152) - not prediction (n= 32) - reviews, letters, comments, or editorials (n= 16) - pre-eclampsia not separate outcome or pregnancy induced

hypertension only (n= 49) - no test accuracy (n= 19) - insufficient data to construct 2x2 table (n= 31) - data duplication (n= 1) - other (n= 4)

Primary articles in cluded in systematic review (n= 34)

Potentially relevant citations identified from electronic searches to capture primary articles on all tests used in prediction of pre-eclampsia (n= 19 548)

Potentially relevant citations on blood pressure measurement (n= 4 000)

References excluded after screening titles or abstracts (n= 3 818)

Figure 1. Literature identification and study selection

Chapter 2

25


eta-analysis

differential verification, independent reference test, blind assessment of index test,

and clinical data available. Less than 30% of studies scored positive on the items for

adequate patient spectrum, adequate descriptions of selection criteria, index test, and

reference test, and blind assessment of reference test. Seven publications reported on the

application of preventive treatment.

3

13

12

34

3

30

5

8

28

34

34

31

15

9

7

4

3

2

1

16

11

12

2

15

27

18

20

30

4

13

15

6

2

7

23

12

0% 20% 40% 60% 80% 100%

Preventive treatment

Withdrawals explained

Uninterpretable results reported

Clinical data available

Blinding of reference test

Blinding of index test

Adequate description reference test

Adequate description index test

Independent reference test

Differential verification bias

Partial verification bias

Time period between tests

Appropriate reference test

Selection criteria described

Representative patient spectrum

Compliance to quality item

Yes; N/A No Unclear

Figure 2. Summary of quality assessment

Data analysis

Figure 3 shows the receiver operating characteristic plots for the results for systolic blood

pressure, diastolic blood pressure, mean arterial pressure and increases of systolic

blood pressure or diastolic blood pressure according to risk and trimester. Some studies

are shown several times with several thresholds. Figure 4 shows the summary receiver

operating characteristic curves for systolic blood pressure, diastolic blood pressure, and

mean arterial pressure in low-risk populations in the second trimester. The area under

26

the curve was 0.68 (95% confidence interval 0.64 to 0.72) for systolic blood pressure,

0.66 (0.59 to 0.72) for diastolic blood pressure, and 0.76 (0.70 to 0.82) for mean

arterial pressure. For a specificity of 90%, the sensitivities of diastolic blood pressure

and mean arterial pressure were both 35%, whereas for systolic blood pressure the

sensitivity was only 24%. For the six studies with a mean arterial pressure of 85 mmHg

or more the pooled sensitivity was 52% (95% confidence interval 28% to 75%) and the

pooled specificity 84% (95% confidence interval 75% to 94%); derived positive likelihood

ratio 3.3 (95% confidence interval 2.2 to 4.3) and negative likelihood ratio 0.57 (95%

confidence interval 0.35 to 0.80). A mean arterial pressure of 90 mmHg or more showed

a pooled sensitivity of 62% (35% to 89%) and a pooled specificity of 82% (72% to 92%);

020

4060

8010

0se

nsiti

vity

(%)

020406080100specificity

SBP

020

4060

8010

0se

nsiti

vity

(%)


DBP

020

4060

8010

0se

nsiti

vity

(%)


MAP0

2040

6080

100

sens

itivi

ty (%

)


increase SBP/DBP

Figure 3. Receiver operating characteristic plots for all results according to timing of test during gestation. Hollow circles represent 2nd trimester testing in low risk populations; solid circles 2nd trimester in high risk populations; and hollow triangles 1st trimester in low risk populations.

Chapter 2

27


eta-analysis

derived positive likelihood ratio 3.5 (CI 2.0 to 5.0) and negative likelihood ratio 0.46

(0.16 to 0.75). The three studies on increase of systolic blood pressure or diastolic

blood pressure, or both, showed poor predictive accuracy (figure 3.3). Poor predictive

accuracy was shown for an increase of systolic blood pressure or diastolic blood pressure

compared with baselinew16;w27 and an increase of blood pressure compared with later

gestational ages.w32

Subgroup and sensitivity analysis

Severe preeclampsia was reported by only one cohort study w25 (first trimester: positive

likelihood ratio 2.9, 1.9 to 4.2; negative likelihood ratio 0.74, 0.59 to 0.88) and one

case control study w2 (positive likelihood ratio 4.3, 2.2 to 8.9; negative likelihood ratio

0.54, 0.45 to 0.70).

In high-risk populations a diastolic blood pressure of 75 mmHg or more at 13 to 20

weeks’ gestation best predicted preeclampsia: (positive likelihood ratio 2.8, 1.8 to

3.6, negative likelihood ratio 0.39, 0.18 to 0.71).w26 Another study showed a positive

likelihood ratio 2.8 and negative likelihood ratio of 0.00, but this study had only one

case of preeclampsia (sensitivity 100%).

In 24 studies it was unclear whether single (in doctor’s surgery) or multiple (ambulatory

blood pressure monitoring during the daytime) readings were reported. When it was

assumed that these studies carried out single measurements, the areas under the curves

for testing in low risk populations in the second trimester remained 0.68 for systolic

0

0,2

0,4

0,6

0,8

1

0 0,2 0,4 0,6 0,8 1

1-specificity

sensitivity

SBP

DBP

MAP

Figure 4. Summary receiver operating characteristic curves with 95% confidence intervals for systolic blood pressure, diastolic blood pressure and mean arterial pressure in population of women at low risk of pre-eclampsia tested in second trimester

28

blood pressure, 0.66 for diastolic blood pressure, and 0.76 for mean arterial pressure.

Areas under the curves for blood pressure measurement in the first trimester were 0.66

for systolic blood pressure, 0.67 for diastolic blood pressure, and 0.79 for mean arterial

pressure.

Studies that excluded women with chronic hypertension showed positive likelihood ratios

for systolic blood pressure of 1.9 (95% confidence interval 0.0 to 3.7), diastolic blood

pressure 2.7 (0.0 to 5.5), and mean arterial pressure 2.6 (1.5 to 3.7). Negative likelihood

ratios were, respectively, 0.6 (0.4 to 0.6), 0.6 (0.4 to 0.7), and 0.5 (0.3 to 0.8).

Sensitivity analysis on preventive treatment was considered impractical because 25

studies (80%) did not report on it. None of the studies met the criteria for high quality.

DISCUSSION

We reviewed the literature on the accuracy of blood pressure measurement during

pregnancy to predict preeclampsia. The mean arterial pressure predicted preeclampsia

fairly well (area under the curve between 0.70 and 0.80) whereas systolic blood pressure,

diastolic blood pressure, and an increase of systolic blood pressure or increase of diastolic

blood pressure, or both, predicted preeclampsia poorly (area under the curve < 0.70).

In low-risk populations a mean arterial pressure in the second trimester best predicted

preeclampsia, with an area under the curve of 0.76. An increased mean arterial pressure

of 90 mmHg or more in the second trimester, however, showed a small increase of the

likelihood of developing preeclampsia (positive likelihood ratio 3.5; negative likelihood

ratio 0.46). In high-risk populations a diastolic blood pressure of 75 mmHg or more

at 13 to 20 weeks’ gestation best predicted preeclampsia, although the accuracy of

prediction was modest (positive likelihood ratio 2.8, negative likelihood ratio 0.39).

Subgroup analyses were limited and did not improve predictive accuracy.

Strengths and weaknesses of the review

We carried out extensive literature searches without language restrictions, assessed the

quality of the studies and the reporting, and used contemporary statistical methods.

Many studies did not distinguish between preeclampsia and other hypertensive disorders

in pregnancy (figure 3.1), nor did many report sufficient information to construct a 2x2

table. Contacting authors did not increase the number of included studies, resulting

in a potential loss of relevant data. Quality assessment and subgroup analyses were

hindered by unclear reporting in many studies, which is a common problem in diagnostic

reviews. Previous studies reported that poor study design and conduct can affect

estimates of diagnostic accuracy21;22 but it is not entirely clear how individual aspects

Chapter 2

29


eta-analysis

of quality may effect this and to what magnitude in this particular area. Many strategies

to account for differences in quality have been applied but none led to estimates that

were systematically less optimistic than ignoring quality in meta-analyses of test accuracy

studies.23;24 As a result of unclear reporting it was not possible to carry out multivariate

subgroup analysis on the basis of individual quality criteria, therefore we reported the

overall results. Poor reporting occurred in details of the index test and reference standard,

patient selection criteria, and blinding. Definitions of preeclampsia have changed over

time, with previous definitions including oedema and increases in blood pressure. The

measurement of blood pressure was poorly reported, underestimating the importance

of recording diastolic blood pressure with Korotkoff phase V as this is more reliably

recorded and more closely reflects direct measurement of diastolic blood pressure.25-27

Poor reporting of the device used underestimates the importance of its validation for

blood pressure measurement during pregnancy.28;29 Poor reporting of patient selection

criteria may partly explain the great variability of incidence rates of preeclampsia not

only between but also within the categories of populations considered high risk or low

risk. In some of the studies the population that was intended to be recruited differed from

the population enrolled. The selection criteria for high risk varied from abnormal uterine

artery detected by Doppler ultrasonography to pre-existent disease or mixtures thereof,

such as chronic hypertension or diabetes. Large cohort studies (> 300,000 women)

reflecting unselected populations however showed incidence rates of preeclampsia

between 0.8 and 5.1%.30-32 This suggests that small studies of low risk or unselected

populations within this review may have been prone to selection bias. Studies that tested

in mid-trimester should have taken into account that the predictive accuracy at 14 weeks

may differ from that at 27 weeks because by definition no one develops preeclampsia

in the first half of pregnancy.

Strengths and weaknesses in relation to other studies

A published review in this area was restricted to evaluating risk factors for preeclampsia

at the first antenatal visit.33 This review included four studies on blood pressure and

concluded that the risk of preeclampsia was increased in women with a raised diastolic

blood pressure (> 80 mmHg) at the first antenatal visit (relative risk 1.48, 95% confidence

interval 1.0 to 1.9). Other published reviews did not apply appropriate methods for

systematic reviews of screening tests or did not distinguish between different hypertensive

disorders.34-39 Some of these reviews concluded that an increased mean arterial pressure

(≥ 85 or ≥ 90 mmHg) predicted transient hypertension rather than preeclampsia 34-36 and that pregnant women with diastolic blood pressures of 70 mmHg or less, or

30

mean arterial pressure of 80 mmHg or less in the second trimester have a small risk of

developing preeclampsia.38

Unanswered questions, future research and implications

When deciding on whether a predictive test should be applied in clinical practice several

things need to be considered: the prevalence of the disease and the predictive accuracy of

the test, the cost of the test and the acceptability to patients, and the treatments available

for the disease in question. Preeclampsia is a disease with relatively low prevalence. A

clinically useful test would need to have a great area under the curve (preferably > 0.80)

or high positive likelihood ratio (> 10) and low negative likelihood ratio (< 0.10).40 From

the results of this review mean arterial pressure shows the greatest predictive accuracy in

both first and second trimesters, with a relatively small likelihood ratio (positive likelihood

ratio 3.5; negative likelihood ratio 0.46). In clinical practice measurement of mean

arterial pressure at the first antenatal visit may improve the accuracy for estimating risk

of preeclampsia. Although it will probably not make a clinical impact in isolation, it is

highly likely that the prediction of preeclampsia will evolve through the development of

algorithms that possibly include clinical, biophysical and biochemical markers. Recently,

the authors of a large prospective study concluded that maternal variables, together

with mean arterial pressure at 11+0 to 13+6 weeks, identify a group at high risk for

preeclampsia.41 Our data cannot rationalise current obstetrical practice of repeated

blood pressure measurements during the first and second trimester in healthy women

with a normal blood pressure at the first antenatal visit. A formal cost utility analysis is

needed.

Women can experience unnecessary anxiety when being identified at risk of preeclampsia

after an antenatal test. At present no pharmacological treatment or management strategy

(for example, regular ultrasound scanning, early delivery) has been shown to effectively

prevent the development of preeclampsia. Early antihypertensive treatment has been

shown to only prevent severe hypertension, not any other complication. Research into

aspirin as a treatment has, however, shown a modest preventive effect (relative risks of

0.9 for preeclampsia and 0.9 for fetal growth restriction42) in the absence of any serious

side effects. Aspirin is a cheap and readily available preventive treatment. In this instance

a false negative test result is potentially more harmful than a false positive test result.

It is imperative to differentiate between mild and severe disease, because early or severe

preeclampsia is associated with raised rates of maternal morbidity and mortality and with

pronounced risks for the fetus, such as severe fetal growth restriction.2;43 Future research

should also concentrate on the development of algorithms that combine biochemical

and biophysical markers, including blood pressure measurement – a diagnostic process

Chapter 2

31


eta-analysis

used in clinical care. These may help improve the predictive accuracy of the tests to

clinically important values.

Acknowledgements

We thank Faridi van Etten for help with the searches, Kees Boer and Hans Wolf for

comments on the project and manuscript, and Koos Zwinderman for help with statistical

analyses.

32

Reference List (1) Khan KS, Wojdyla D, Say L, Gulmezoglu AM, Van Look PF. WHO analysis of causes of maternal

death: a systematic review. Lancet 2006; 367:1066-1074. (2) Sibai B, Dekker G, Kupferminc M. Pre-eclampsia. Lancet 2005; 365(9461):785-799. (3) Vatten LJ, Skjaerven R. Is Preeclampsia More Than One Disease? Obstet Gynecol Surv 2004;

59(9):645-646. (4) Sibai BM. Diagnosis and management of gestational hypertension and preeclampsia. Obstet

Gynecol 2003; 102(1):181-192. (5) Hauth JC, Ewell MG, Levine RJ, Esterlitz JR, Sibai B, Curet LB et al. Pregnancy outcomes in

healthy nulliparas who developed hypertension. Calcium for Preeclampsia Prevention Study Group. Obstet Gynecol 2000; 95(1):24-28.

(6) Redman CW, Sargent IL. Latest advances in understanding preeclampsia. Science 2005; 308(5728):1592-1594.

(7) Easterling TR, Benedetti TJ, Schmucker BC, Millard SP. Maternal hemodynamics in normal and preeclamptic pregnancies: a longitudinal study. Obstet Gynecol 1990; 76(6):1061-1069.

(8) Rang S, Wolf H, Van Montfrans GA, Karemaker JM. Serial assessment of cardiovascular control shows early signs of developing pre-eclampsia. J Hypertens 2004; 22(2):369-376.

(9) Duley L, Henderson-Smart DJ, Knight M, King JF. Antiplatelet agents for preventing pre-eclampsia and its complications. Cochrane Database Syst Rev 2004;(1):CD004659.

(10) Cnossen JS, van der Post JA, Mol BW, Khan KS, Meads CA, Ter RG. Prediction of pre-eclampsia: a protocol for systematic reviews of test accuracy. BMC Pregnancy Childbirth 2006; 6:29.

(11) Yale University School of Medicine: Finding the evidence on PubMed. http://info med yale edu/library/reference/publications/pubmed/ [cited 2003 Dec. 10]

(12) Bachmann LM, Coray R, Estermann P, Ter Riet G. Identifying diagnostic studies in MEDLINE: reducing the number needed to read. J Am Med Inform Assoc 2002; 9(6):653-658.

(13) Haynes RB, Wilczynski N, McKibbon KA, Walker CJ, Sinclair JC. Developing optimal search strategies for detecting clinically sound studies in MEDLINE. J Am Med Inform Assoc 1994; 1(6):447-458.

(14) Whiting P, Rutjes AW, Reitsma JB, Bossuyt PM, Kleijnen J. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 2003; 3(1):25.

(15) Brown MA, Lindheimer MD, de Swiet M, Van Assche A, Moutquin JM. The classification and diagnosis of the hypertensive disorders of pregnancy: statement from the International Society for the Study of Hypertension in Pregnancy (ISSHP). Hypertens Pregnancy 2001; 20(1):IX-XIV.

(16) Harbord RM, Deeks JJ, Egger M, Whiting P, Sterne JA. A unification of models for meta-analysis of diagnostic accuracy studies. Biostatistics 2007; 8(2):239-251.

(17) Reitsma JB, Glas AS, Rutjes AW, Scholten RJ, Bossuyt PM, Zwinderman AH. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005; 58(10):982-990.

(18) Akaike H. A new look at the statistical model identification. IEEE Transactions on Automatic Control 1974; 19(6):716-723.

(19) Zwinderman AH, Bossuyt PM. We should not pool diagnostic likelihood ratios in systematic reviews. Stat Med 2008; 27(5):687-697.

(20) Meads CA, Cnossen JS, Meher S, Juarez_Garcia A, ter Riet G, Duley L et al. Methods of prediction and prevention of pre-eclampsia - systematic reviews of accuracy and effectiveness literature with economic modelling. 12. 2008. Health Technology Assessment.

(21) Lijmer JG, Mol BW, Heisterkamp S, Bonsel GJ, Prins MH, van der Meulen JH et al. Empirical evidence of design-related bias in studies of diagnostic tests. JAMA 1999; 282(11):1061-1066.

Chapter 2

33


eta-analysis

(22) Rutjes AW, Reitsma JB, Di Nisio M, Smidt N, van Rijn JC, Bossuyt PM. Evidence of bias and variation in diagnostic accuracy studies. CMAJ 2006; 174(4):469-476.

(23) Leeflang M, Reitsma J, Scholten R, Rutjes A, Dinisio M, Deeks J et al. Impact of Adjustment for Quality on Results of Metaanalyses of Diagnostic Accuracy. Clin Chem 2007; 53(2):164-172.

(24) Whiting P, Harbord R, Kleijnen J. No role for quality scores in systematic reviews of diagnostic accuracy studies. BMC Med Res Methodol 2005; 5:19.

(25) Brown MA, Reiter L, Smith B, Buddle ML, Morris R, Whitworth JA. Measuring blood pressure in pregnant women - a comparison of direct and indirect methods. Am J Obstet Gynecol 1994; 171(3):661-667.

(26) Brown MA, Hague WM, Higgins J, Lowe S, McCowan L, Oats J et al. The detection, investigation and management of hypertension in pregnancy: full consensus statement. Aust N Z J Obstet Gynaecol 2000; 40(2):139-155.

(27) Shennan A, Gupta M, Halligan A, Taylor DJ, deSwiet M. Lack of reproducibility in pregnancy of Korotkoff phase IV as measured by mercury sphygmomanometry. Lancet 1996; 347(8995):139-142.

(28) Brown MA, Robinson A, Buddle ML. Accuracy of automated blood pressure recorders in pregnancy. Aust N Z J Obstet Gynaecol 1998; 38(3):262-265.

(29) Pomini F, Scavo M, Ferrazzani S, De Carolis S, Caruso A, Mancuso S. There is poor agreement between manual auscultatory and automated oscillometric methods for the measurement of blood pressure in normotensive pregnant women. J Matern Fetal Med 2001; 10(6):398-403.

(30) Conde-Agudelo A, Althabe F, Belizan JM, Kafury-Goeta AC. Cigarette smoking during pregnancy and risk of preeclampsia: a systematic review. Am J Obstet Gynecol 1999; 181(4):1026-1035.

(31) Sebire NJ, Jolly M, Harris JP, Wadsworth J, Joffe M, Beard RW et al. Maternal obesity and pregnancy outcome: a study of 287,213 pregnancies in London. Int J Obes Relat Metab Disord 2001; 25(8):1175-1182.

(32) Sebire NJ, Jolly M, Harris J, Regan L, Robinson S. Is maternal underweight really a risk factor for adverse pregnancy outcome? A population-based study in London. BJOG 2001; 108(1):61-66.

(33) Duckitt K, Harrington D. Risk factors for pre-eclampsia at antenatal booking: systematic review of controlled studies. BMJ 2005; 330(7491):565.

(34) Chesley LC, Sibai BM. Clinical significance of elevated mean arterial pressure in the second trimester. Am J Obstet Gynecol 1988; 159(2):275-279.

(35) Conde-Agudelo A, Lede R, Belizan J. Evaluation of methods used in the prediction of hypertensive disorders of pregnancy. Obstet Gynecol Surv 1994; 49(3):210-222.

(36) Dekker GA, Sibai BM. Early detection of preeclampsia. Am J Obstet Gynecol 1991; 165(1):160-172.

(37) O’Brien WF. Predicting preeclampsia. Obstet Gynecol 1990; 75(3 Pt 1):445-452. (38) O’Brien WF. The prediction of preeclampsia. Clin Obstet Gynecol 1992; 35(2):351-364. (39) Visser W, Wallenburg HC. Prediction and prevention of pregnancy-induced hypertensive

disorders. Baillieres Best Pract Res Clin Obstet Gynaecol 1999; 13(1):131-156. (40) Deeks JJ, Altman DG. Diagnostic tests 4: likelihood ratios. BMJ 2004; 329(7458):168-169. (41) Poon LCY, Kametas NA, Pandeva I, Valencia C, Nicolaides KH. Mean arterial pressure at 11+0

to 13+6 weeks in the prediction of preeclampsia. Hypertension 2008; 51:1-7. (42) Askie LM, Duley L, Henderson-Smart DJ, Stewart LA. Antiplatelet agents for prevention of

pre-eclampsia: a meta-analysis of individual patient data. Lancet 2007; 369:1791-1798. (43) Report of the National High Blood Pressure Education Program Working Group on High Blood

Pressure in Pregnancy. Am J Obstet Gynecol 2000; 183(1):S1-S22.

34

Appendix 1

Search strategies to identify articles on blood pressure measurement used to predict preeclampsiaThe overall search strategy on prediction of preeclampsia is previously published in a

protocol.1 For this review we identified citations on blood pressure measurement in the

established Reference Manager 11.0 database using the following terms in keywords,

titles, and abstract: {dinamap} OR {dynamap} OR OR {mercury} OR {oscill\*} OR {Auscultat\*} OR {Automat\*}

OR {Sphygmomanom\*} OR {Mercury} OR {Ambulatory Monitoring} OR OR OR {finapres}

OR {finger arterial pressure waveform} OR {Blood Pressure Determination} OR {Blood Pressure

Measurement} OR {Blood Pressure Monitoring\*} OR {Blood Pressure Monitor}

Additional search in PubMED

1. (((“Sphygmomanometers”[MeSH]) OR (“Blood Pressure Determination”[MeSH]) OR (“Blood Pressure”[MeSH:noexp])) AND ((“Hypertension, Pregnancy-Induced”[MeSH]) OR (“Pregnancy Complications, Cardiovascular”[MeSH] AND Humans[Mesh]))) AND (predict*[tiab] OR predictive value of tests[mh] OR scor*[tiab] OR observ*[tiab] OR observer variation[mh])

2. Search (((“Sphygmomanometers”[MeSH]) OR (“Blood Pressure Determination”[MeSH]) OR (“Blood Pressure”[MeSH:noexp])) AND ((“Hypertension, Pregnancy-Induced”[MeSH]) OR (“Pregnancy Complications, Cardiovascular”[MeSH] AND Humans[Mesh]))) NOT (((((“Sphygmomanometers”[MeSH]) OR (“Blood Pressure Determination”[MeSH]) OR (“Blood Pressure”[MeSH:noexp])) AND ((“Hypertension, Pregnancy-Induced”[MeSH]) OR (“Pregnancy Complications, Cardiovascular”[MeSH] AND Humans[Mesh]))) AND Letter[ptyp]) OR ((((“Sphygmomanometers”[MeSH]) OR (“Blood Pressure Determination ”[MeSH]) OR (“Blood Pressure”[MeSH:noexp])) AND ((“Hypertension, Pregnancy-Induced”[MeSH]) OR (“Pregnancy Complications, Cardiovascular”[MeSH] AND Humans[Mesh]))) AND Editorial[ptyp]) OR ((((“Sphygmomanometers”[MeSH]) OR (“Blood Pressure Determination”[MeSH]) OR (“Blood Pressure”[MeSH:noexp])) AND ((“Hypertension, Pregnancy-Induced”[MeSH]) OR (“Pregnancy Complications, Cardiovascular”[MeSH] AND Humans[Mesh]))) AND Review[ptyp]) OR ((((“Sphygmomanometers”[MeSH]) OR (“Blood Pressure Determination”[MeSH]) OR (“Blood Pressure”[MeSH:noexp])) AND ((“Hypertension, Pregnancy-Induced”[MeSH]) OR (“Pregnancy Complications, Cardiovascular”[MeSH] AND Humans[Mesh]))) AND Practice Guideline[ptyp]))

3. #2 NOT #1

Reference List

(1) Cnossen JS, van der Post JA, Mol BW, Khan KS, Meads CA, Ter RG. Prediction of pre-eclampsia:

a protocol for systematic reviews of test accuracy. BMC Pregnancy Childbirth 2006; 6:29.

Chapter 2

35


eta-analysis

Appendix 2

QUADAS (quality assessment of diagnostic accuracy studies) criteria used to assess the methodological quality of studies of blood pressure measurement used to predict pre-eclampsiaQUADAS question Study characteristics and methods required to

meet criterion

1. Was the spectrum of patients representative of the patients who will receive the test in practice? (spectrum bias)

Pregnant women, consecutively recruited; study has prospective design. Patients with chronic hypertension, insulin dependent diabetes mellitus, chronic renal disease, autoimmune disorders (AID), multiple pregnancies, or history of pre-eclampsia; so called ‘high risk’ patients

2. Were patient selection criteria clearly described? (selection bias)

Information is available on at least 5 of the following factors: chronic hypertension, insulin dependent diabetes mellitus, chronic renal disease, parity, singleton/ multiple pregnancies, previous pre-eclampsia, and age

3. Is the reference standard likely to correctly classify the target condition?

lPre-eclampsia: systolic blood pressure ≥ 140 mmHg or diastolic blood pressure ≥ 90 mmHg plus proteinuria measured as ≥ 0.3 g of protein in a 24-hour urine collection or dipstick test result ≥ 1+ (equivalent to 30 mg/dL in a single urine sample)

lSuperimposed pre-eclampsia: proteinuria measured as ≥ 0.3 g of protein in a 24-hour urine collection or dipstick test result ≥ 1+ after 20 weeks of gestation in patients with chronic hypertension

lSevere pre-eclampsia: systolic blood pressure ≥ 160 mmHg or diastolic blood pressure ≥ 110 mmHg plus proteinuria measured as ≥ 2.0 g of protein in a 24-hour urine collection or dipstick test result ≥ 3+

4. Is the time period between reference standard and index test short enough to be reasonably sure that the target condition did not change between the 2 tests? (disease progression bias)

Not applicable

5. Did the whole sample, or a random selection of the sample receive verification using a reference standard of diagnosis? (partial verification bias)

All patients or a random selection received verification with reference standard (even if reference standard not the same for all patients)

6. Did patients receive the same reference standard regardless of the index test result? (differential verification bias)

All patients received same reference test (this is likely because the index test is non-invasive).

7. Was the reference standard independent of the index test (i.e. the index test did not form part of the reference standard)? (incorporation bias)

The results of the index test are not incorporated in the definition of pre-eclampsia

8. Was the execution of the index test described in sufficient detail to permit its replication?

Description includes at least seven of the following: linstrument used, (manufacturer), calibration of

instrument, appropriate cuff, position of patient, left or right arm, state of mind (rest), Korotkoff sound (IV or V) for DBP, number of measurements (single or serial), measurement parameter/ cutoff level

9. Was the execution of the reference standard described in sufficient detail to permit its replication?

Description includes:lBlood pressure: instrument, position of patient,

Korotkoff sound for diastolic blood pressure lProteinuria: 24-hour urine collection or use of

dipstick with cutoff point

36

10a. Were the index test results interpreted without knowledge of the results of the reference standard? (review bias)

Always fulfilled, reference test results not yet available when index test is performed (prediction)

10b. Were the reference standard results interpreted without knowledge of the results of the index test? (review bias)

Relevant statement is included in text (e.g., “Assessors were blind to index test results”)

11. Were the same clinical data available when tests results were interpreted as would be available when the test is used in practice?

Any information to the patient obtained by direct observation (age, symptoms, body mass index) normally available when test is interpreted in practice was also available when the test was interpreted in the study, or data unavailable in practice also unavailable during interpretation

12. Were uninterpretable/ intermediate test results reported?

All test results are reported, including uninterpretable/ intermediate results

13. Were withdrawals from the study explained? It is clear what happened to all patients in study (e.g., flow diagram [follow-up])

Supplementary question

14. Was there any preventive intervention? After blood pressure measuerment, patients received any of the following: acetylsalicylic acid, low-molecular-weight heparin, vitamin C or E, antihypertensive medication, saline infusion, oxygen

38

Table 1. Characteristics of studies of blood pressure measurement used to predict preeclampsia in pregnant women at low or unspecified riskFirst author (year) Country

Gestation(weeks)

N(% preeclampsia)

SensitivityTP/(TP+FN)

Specificity TN/(FP+TN)

Index test Threshold(mmHg)

Details index test Reference standard preeclampsia

Population (study design)*

Ales1

(1989)USA

2nd trimester 730 (0.8) 4/6 589/724 MAP ≥ 85 Automated ultrasound device with digital readout, seated, average both arms, 5min rest, K5

RR ≥ 140/90 mmHg, MAP ≥ 107 mmHg or increase > 30/15 mmHg, twice 6h apart after 24 weeks; proteinuria ++ dipstick or > 0.3g/24h

IN: all pregnant women

Atterbury2 (1996)USA

2nd trimester 114 (50.0) 30/57 50/57 MAP ≥ 85 Sphygmomanometer, Baumonometer, seated, K5

Severe preeclampsia: SBP > 170 or DBP > 110 mmHg, twice 6h apart; proteinuria ++ dipstick; (oliguria, cerebral or visual disturbances, pulmonary oedema, epigastric or right upper quadrant pain, HELLP)

IN (cases): gestational age over 24 weeks, complete records, midtrimester blood pressure, protein assay 18-22 weeks, severe preeclampsia; IN (controls): normotensive, delivery after 37 weeks; EX: from control group for any medical, obstetrical or fetal complications.(case control)

Conde-Agudelo3 (1993)Argentina

202631

580 (4.0) 11/2310/239/23

334/557434/557479/557

MAP > 81MAP > 85MAP > 89

4 random-zero sphygmomanometers rotating among centers, calibrated every 3 weeks seated, left arm, 10 min rest, mean of 5 measurements, K5

SBP ≥ 140 or DBP ≥ 90 mmHg, twice 6h apart; proteinuria 0.3g/L

IN: placebogroup of trial on calcium supplementation (same population as Lopez et al.) EX: chronic medical illnesses

Fallis4

(1963) USA2nd trimester 113 (34.5) 32/39 65/74 MAP > 90 Hypertension + proteinuria IN: primigravid women

Friedman5 (1977)USA

17-26 22,582 (3.9) 571/888 13458/21694 MAP ≥ 90 Indirect sphygmomanometer, K4 or K5

Hypertension + proteinuria IN: unselected population (Collaborative Perinatal Project)

Higgins6 (1997)Ireland

18-24 984 (2.3) 4/233/234/235/234/232/231/232/235/235/234/236/23

920/961940/961914/961916/961919/961924/961934/961926/961933/961923/961931/961919/961

MAP > 86 (ABPM)MAP > 86 (clinic)

MAP > 86 (nighttime)MAP > 86 (daytime)SBP > 119 (ABPM)

SBP > 119 (nighttime)SBP > 119 (clinic)

SBP > 119 (daytime)DBP > 71 (ABPM)

DBP > 71 (nighttime)DBP > 71 (clinic)

DBP > 71 (daytime)

ABPM, SpaceLabs 90207, every 30 min, non-dominant armOffice blood pressure: appropriate cuff, seated, relaxed, arm at heart level, non-dominant arm

DBP > 90 mmHg twice 4h apart or DBP > 110 mmHg once; ≥ + dipstick twice 4h apart, no urinary tract infection

IN: healthy primigravid women; EX: history of hypertension, renal and cardiac disease, diabetes mellitus

Iwasaki7

(2002)Japan

1st trimester 1599 (2.8) 44/4541/4540/4529/45

300/1554621/1554951/15541253/1554

MAP ≥ 74MAP ≥ 80MAP ≥ 86MAP ≥ 92

Non-invasive automated measurements (BP-203RV, Nippon Colin Co, Tokyo), seated, right arm held at heart level

SBP ≥ 140 or DBP ≥ 90 mmHg after second trimester; proteinuria ≥ 0.3g/24h

IN: singleton pregnancies

Knuist8

(1998)Netherlands

at booking 2413 (1.4) 16/3425/34

1684/23791107/2379

SBP > 120DBP > 70

SBP at bookingDBP at booking

DBP > 90 mmHg twice 4h apart, after 20 weeks; ≥ ++ dipstick twice 4h apart without urinary tract infection

IN: all nulliparous, singleton pregnancies; EX: pre-existing disease (such as diabetes, severe hypertension, renal disease) or obstetric abnormality (such as multiple gestation)

Appendix 3

Chapter 2

39


eta-analysis

Table 1. Characteristics of studies of blood pressure measurement used to predict preeclampsia in pregnant women at low or unspecified riskFirst author (year) Country

Gestation(weeks)

N(% preeclampsia)



Index test Threshold(mmHg)



Ales1

(1989)USA

2nd trimester 730 (0.8) 4/6 589/724 MAP ≥ 85 Automated ultrasound device with digital readout, seated, average both arms, 5min rest, K5

RR ≥ 140/90 mmHg, MAP ≥ 107 mmHg or increase > 30/15 mmHg, twice 6h apart after 24 weeks; proteinuria ++ dipstick or > 0.3g/24h

IN: all pregnant women

Atterbury2 (1996)USA

2nd trimester 114 (50.0) 30/57 50/57 MAP ≥ 85 Sphygmomanometer, Baumonometer, seated, K5

Severe preeclampsia: SBP > 170 or DBP > 110 mmHg, twice 6h apart; proteinuria ++ dipstick; (oliguria, cerebral or visual disturbances, pulmonary oedema, epigastric or right upper quadrant pain, HELLP)

IN (cases): gestational age over 24 weeks, complete records, midtrimester blood pressure, protein assay 18-22 weeks, severe preeclampsia; IN (controls): normotensive, delivery after 37 weeks; EX: from control group for any medical, obstetrical or fetal complications.(case control)

Conde-Agudelo3 (1993)Argentina

202631

580 (4.0) 11/2310/239/23

334/557434/557479/557

MAP > 81MAP > 85MAP > 89

4 random-zero sphygmomanometers rotating among centers, calibrated every 3 weeks seated, left arm, 10 min rest, mean of 5 measurements, K5

SBP ≥ 140 or DBP ≥ 90 mmHg, twice 6h apart; proteinuria 0.3g/L

IN: placebogroup of trial on calcium supplementation (same population as Lopez et al.) EX: chronic medical illnesses

Fallis4

(1963) USA2nd trimester 113 (34.5) 32/39 65/74 MAP > 90 Hypertension + proteinuria IN: primigravid women

Friedman5 (1977)USA

17-26 22,582 (3.9) 571/888 13458/21694 MAP ≥ 90 Indirect sphygmomanometer, K4 or K5

Hypertension + proteinuria IN: unselected population (Collaborative Perinatal Project)

Higgins6 (1997)Ireland

18-24 984 (2.3) 4/233/234/235/234/232/231/232/235/235/234/236/23

920/961940/961914/961916/961919/961924/961934/961926/961933/961923/961931/961919/961

MAP > 86 (ABPM)MAP > 86 (clinic)

MAP > 86 (nighttime)MAP > 86 (daytime)SBP > 119 (ABPM)

SBP > 119 (nighttime)SBP > 119 (clinic)

SBP > 119 (daytime)DBP > 71 (ABPM)

DBP > 71 (nighttime)DBP > 71 (clinic)

DBP > 71 (daytime)

ABPM, SpaceLabs 90207, every 30 min, non-dominant armOffice blood pressure: appropriate cuff, seated, relaxed, arm at heart level, non-dominant arm

DBP > 90 mmHg twice 4h apart or DBP > 110 mmHg once; ≥ + dipstick twice 4h apart, no urinary tract infection

IN: healthy primigravid women; EX: history of hypertension, renal and cardiac disease, diabetes mellitus

Iwasaki7

(2002)Japan

1st trimester 1599 (2.8) 44/4541/4540/4529/45

300/1554621/1554951/15541253/1554

MAP ≥ 74MAP ≥ 80MAP ≥ 86MAP ≥ 92

Non-invasive automated measurements (BP-203RV, Nippon Colin Co, Tokyo), seated, right arm held at heart level

SBP ≥ 140 or DBP ≥ 90 mmHg after second trimester; proteinuria ≥ 0.3g/24h

IN: singleton pregnancies

Knuist8

(1998)Netherlands

at booking 2413 (1.4) 16/3425/34

1684/23791107/2379

SBP > 120DBP > 70

SBP at bookingDBP at booking

DBP > 90 mmHg twice 4h apart, after 20 weeks; ≥ ++ dipstick twice 4h apart without urinary tract infection

IN: all nulliparous, singleton pregnancies; EX: pre-existing disease (such as diabetes, severe hypertension, renal disease) or obstetric abnormality (such as multiple gestation)

40

Kyle9

(1993)UK

181828282828

145 (11.7) 11/174/1715/1711/176/172/17

65/128115/12854/128104/128115/128125/128

MAP ≥ 80MAP ≥ 85MAP ≥ 80MAP ≥ 85DBP ≥ 75DBP ≥ 80

ABPM: TM2420 (Japan), cuff size determined by upper arm circumference each visit, seated, left arm, every 15 min awake/ every 30min asleep, at least 20/ 8 valid readings

increase DBP ≥ 25 mmHg to a maximum of DBP ≥ 90 mmHg or more; proteinuria ≥ + dipstick twice

IN: normotensive, no renal disease and not taking medication

Lopez10

(1994)Argentina

3434

1167 (3.3) 16/3810/38

926/11291039/1129

DBP K4DBP K5

4 random-zero sphygmomanometers rotating monthly among centers, calibration every 3 weeks, 10 min rest, lateral position, 5 measurements in each position (lateral/ supine/ seated), K4 and K5, standardised

DBP ≥ 90 mmHg twice 6h apart; proteinuria ≥ 0.3g/L twice 6h apart, random urine specimens

IN: nulliparous, singleton pregnancies, free of clinical/ laboratory evidence of current/previous diseases, not taking any treatment, no history of diabetes mellitus, chronic renal disease, and normal oral glucose tolerance test results (RCT)

Mahanna11 (1983)Italy

17-25 210 (4.8) 9/10 185/200 MAP > 90 Twice n.r. IN: unselected; EX: chronic hypertension

Moutquin12 (1986)Canada

9-1213-1617-2021-2425-289-1213-1617-2021-2425-289-1213-1617-2021-2425-289-1213-1617-2021-2425-28

983 (5.0) 23/4922/4916/4919/4922/4915/4915/498/4910/499/4923/4928/4921/4920/4921/4915/4921/499/4914/4914/49

786/934839/934850/934872/934856/934869/934902/934910/934914/934915/934704/934758/934797/934805/934772/934814/934842/934873/934867/934854/934

DBP > 80DBP > 80DBP > 80DBP > 80DBP > 80DBP > 85DBP > 85DBP > 85DBP > 85DBP > 85SBP > 130SBP > 130SBP > 130SBP > 130SBP > 130SBP > 135SBP > 135SBP > 135SBP > 135SBP > 135

Automated random-zero sphygmomanometer (Dinamap 845), right arm, seated, 5 min rest, at least twice, calibration daily, oscillometric, K4

RR ≥ 140/90 mmHg twice ≥ 6h apart; proteinuria ≥ + albustix twice 6h apart or oedema or serum urate levels (>4.6 mg/dL); RR < 120/80 mmHg 6 weeks postpartum

IN: patients of 3 private obstetricians


9-1213-1617-2021-2425-28

224 (6.3) 6/146/145/141/144/14

200/210202/210206/210206/210208/210

DBP ≥ 89DBP ≥ 89DBP ≥ 89DBP ≥ 89DBP ≥ 89

Automated sphygmomanometer (Dinamap 845*, USA), calibrated, digital, seated, 5 min rest, twice 2 min. interval, K4

RR > 140/90 mmHg or increase 30/15 mmHg twice 6h apart after 20 weeks; proteinuria 0.3g/L twice 6h apart

IN: patients of 3 private obstetricians; singleton, primi/multiparity

Chapter 2

41


eta-analysis

Kyle9

(1993)UK

181828282828

145 (11.7) 11/174/1715/1711/176/172/17

65/128115/12854/128104/128115/128125/128

MAP ≥ 80MAP ≥ 85MAP ≥ 80MAP ≥ 85DBP ≥ 75DBP ≥ 80

ABPM: TM2420 (Japan), cuff size determined by upper arm circumference each visit, seated, left arm, every 15 min awake/ every 30min asleep, at least 20/ 8 valid readings

increase DBP ≥ 25 mmHg to a maximum of DBP ≥ 90 mmHg or more; proteinuria ≥ + dipstick twice

IN: normotensive, no renal disease and not taking medication

Lopez10

(1994)Argentina

3434

1167 (3.3) 16/3810/38

926/11291039/1129

DBP K4DBP K5

4 random-zero sphygmomanometers rotating monthly among centers, calibration every 3 weeks, 10 min rest, lateral position, 5 measurements in each position (lateral/ supine/ seated), K4 and K5, standardised

DBP ≥ 90 mmHg twice 6h apart; proteinuria ≥ 0.3g/L twice 6h apart, random urine specimens

IN: nulliparous, singleton pregnancies, free of clinical/ laboratory evidence of current/previous diseases, not taking any treatment, no history of diabetes mellitus, chronic renal disease, and normal oral glucose tolerance test results (RCT)

Mahanna11 (1983)Italy

17-25 210 (4.8) 9/10 185/200 MAP > 90 Twice n.r. IN: unselected; EX: chronic hypertension


9-1213-1617-2021-2425-289-1213-1617-2021-2425-289-1213-1617-2021-2425-289-1213-1617-2021-2425-28

983 (5.0) 23/4922/4916/4919/4922/4915/4915/498/4910/499/4923/4928/4921/4920/4921/4915/4921/499/4914/4914/49

786/934839/934850/934872/934856/934869/934902/934910/934914/934915/934704/934758/934797/934805/934772/934814/934842/934873/934867/934854/934

DBP > 80DBP > 80DBP > 80DBP > 80DBP > 80DBP > 85DBP > 85DBP > 85DBP > 85DBP > 85SBP > 130SBP > 130SBP > 130SBP > 130SBP > 130SBP > 135SBP > 135SBP > 135SBP > 135SBP > 135

Automated random-zero sphygmomanometer (Dinamap 845), right arm, seated, 5 min rest, at least twice, calibration daily, oscillometric, K4

RR ≥ 140/90 mmHg twice ≥ 6h apart; proteinuria ≥ + albustix twice 6h apart or oedema or serum urate levels (>4.6 mg/dL); RR < 120/80 mmHg 6 weeks postpartum

IN: patients of 3 private obstetricians


9-1213-1617-2021-2425-28

224 (6.3) 6/146/145/141/144/14

200/210202/210206/210206/210208/210

DBP ≥ 89DBP ≥ 89DBP ≥ 89DBP ≥ 89DBP ≥ 89

Automated sphygmomanometer (Dinamap 845*, USA), calibrated, digital, seated, 5 min rest, twice 2 min. interval, K4

RR > 140/90 mmHg or increase 30/15 mmHg twice 6h apart after 20 weeks; proteinuria 0.3g/L twice 6h apart

IN: patients of 3 private obstetricians; singleton, primi/multiparity

42


9-12

13-1617-2021-2425-2829-3233-369-1213-1617-2021-2425-2829-3233-369-1213-1617-2021-2425-2829-3233-369-1213-1617-2021-2425-2829-3233-36

751 (5.9) 12/44

27/4430/4427/4428/4429/4415/449/4419/4424/4423/4421/4424/448/448/4416/4423/4417/4416/4416/447/443/4410/448/445/447/448/443/44

625/707

449/707365/707363/707304/707467/707535/707658/707558/707491/707531/707532/707590/707637/707677/707598/707559/707591/707602/707626/707655/707703/707679/707670/707680/707687/707692/707693/707

increase DBP > 10(every 4 weeks reference

period)increase DBP > 10increase DBP > 10increase DBP > 10increase DBP > 10increase DBP > 10increase DBP > 10increase DBP > 15increase DBP > 15increase DBP > 15increase DBP > 15increase DBP > 15increase DBP > 15increase DBP > 15increase SBP > 20increase SBP > 20increase SBP > 20increase SBP > 20increase SBP > 20increase SBP > 20increase SBP > 20increase SBP > 30increase SBP > 30increase SBP > 30increase SBP > 30increase SBP > 30increase SBP > 30Increase SBP > 30

Automated random-zero sphygmomanometer with digital dysplay (Dinamap 845, USA), oscillometric, daily calibration, seated, 5 min rest, twice 2 min. intervals, K4

RR > 140/90 mmHg twice 6h apart; proteinuria ≥ + dipstick twice 6h apart or generalised oedema; RR < 120/80 mmHg 6 weeks postpartum

IN:primiparae; EX: gestational hypertension, chronic hypertension, liver disease and hypertension,

Odegard15 (2000)Norway

first visit (< 18)first visit (< 18)first visit (< 18)first visit (< 18)first visit (< 18)first visit (< 18)

906 (34.2) 274/310201/31085/310304/311237/31193/311

125/596323/596527/59632/596387/596455/596

SBP > 110SBP > 120SBP > 130DBP > 60DBP > 70DBP > 80

SBP/ DBP DBP > 90 mmHg in addition to increase > 25 mmHg or when baseline DBP > 90 mmHg then increase > 15 mmHg; proteinuria ≥ + dipstick (= 0.3g/L)

IN: Norwegian Medical Birth Registry, delivered in Rogaland Central Hospital (nested case control)

Ohkuchi16 (2003)Japan

1st trimester 1498 (7.3) 36/110 1155/1388 Increase SBP/DBP >30/15 Noninvasive automated measurement (BP-203RV II, Japan), seated, right arm at heart level, K5

SBP ≥ 140 or DBP ≥ 90 mmHg twice; proteinuria + dipstick twice or ++ dipstick once or > 0.3g/24h

IN: normotensive, non-proteinuric

Oney17 (1983)Germany

18-26 200 (10.5) 19/21 113/179 MAP ≥ 90 Auscultatory blood pressure mearsurements, 10 min rest

SBP ≥ 140 or DBP ≥ 90 mmHg twice 6h apart after 24 weeks gestation; proteinuria ≥ 0.3g/24h

IN: normotensive, nulliparous women; EX: cardiovascular disease, urinary infection


18-26 108 (13.9) 14/15 46/93 MAP > 90 n.r. Gestose index n.r.

Page19

(1976)2nd trimester2nd trimester2nd trimester2nd trimester2nd trimester

14833 (2.7) 375/399337/399265/399174/399100/399

3353/144346500/1443410008/1443412594/1443413848/14434

MAP ≥ 75MAP ≥ 80MAP ≥85MAP ≥ 90MAP ≥ 95

n.r. Hypertension; proteinuria ≥ ++ dipstick or oedema

IN: patients of Child Health and Development Studies (California)

Reiss20

(1987)USA

2nd trimester2nd trimester

60 (50.0) 6/304/30

28/3028/30

MAP > 75MAP > 80

Right arm, resting in left lateral recumbent position, K4

rise DBP > 15 mmHg; proteinuria ≥ + dipstick

IN: hypertensive cases, controls matched for age, race gravidity, parity, an approximate delivery date (case control)

Robrecht21 (1980)Germany

14-2814-28

285 (19.6) 50/5635/56

156/229204/229

MAP ≥ 85MAP > 90

n.r. Gestose index ≥ 1: RR > 140/90 mmHg, proteinuria ≥ 0.5g/24h, pretibial oedema

n.r.

Chapter 2

43


eta-analysis


9-12

13-1617-2021-2425-2829-3233-369-1213-1617-2021-2425-2829-3233-369-1213-1617-2021-2425-2829-3233-369-1213-1617-2021-2425-2829-3233-36

751 (5.9) 12/44

27/4430/4427/4428/4429/4415/449/4419/4424/4423/4421/4424/448/448/4416/4423/4417/4416/4416/447/443/4410/448/445/447/448/443/44

625/707

449/707365/707363/707304/707467/707535/707658/707558/707491/707531/707532/707590/707637/707677/707598/707559/707591/707602/707626/707655/707703/707679/707670/707680/707687/707692/707693/707

increase DBP > 10(every 4 weeks reference

period)increase DBP > 10increase DBP > 10increase DBP > 10increase DBP > 10increase DBP > 10increase DBP > 10increase DBP > 15increase DBP > 15increase DBP > 15increase DBP > 15increase DBP > 15increase DBP > 15increase DBP > 15increase SBP > 20increase SBP > 20increase SBP > 20increase SBP > 20increase SBP > 20increase SBP > 20increase SBP > 20increase SBP > 30increase SBP > 30increase SBP > 30increase SBP > 30increase SBP > 30increase SBP > 30Increase SBP > 30

Automated random-zero sphygmomanometer with digital dysplay (Dinamap 845, USA), oscillometric, daily calibration, seated, 5 min rest, twice 2 min. intervals, K4

RR > 140/90 mmHg twice 6h apart; proteinuria ≥ + dipstick twice 6h apart or generalised oedema; RR < 120/80 mmHg 6 weeks postpartum

IN:primiparae; EX: gestational hypertension, chronic hypertension, liver disease and hypertension,

Odegard15 (2000)Norway

first visit (< 18)first visit (< 18)first visit (< 18)first visit (< 18)first visit (< 18)first visit (< 18)

906 (34.2) 274/310201/31085/310304/311237/31193/311

125/596323/596527/59632/596387/596455/596

SBP > 110SBP > 120SBP > 130DBP > 60DBP > 70DBP > 80

SBP/ DBP DBP > 90 mmHg in addition to increase > 25 mmHg or when baseline DBP > 90 mmHg then increase > 15 mmHg; proteinuria ≥ + dipstick (= 0.3g/L)

IN: Norwegian Medical Birth Registry, delivered in Rogaland Central Hospital (nested case control)

Ohkuchi16 (2003)Japan

1st trimester 1498 (7.3) 36/110 1155/1388 Increase SBP/DBP >30/15 Noninvasive automated measurement (BP-203RV II, Japan), seated, right arm at heart level, K5

SBP ≥ 140 or DBP ≥ 90 mmHg twice; proteinuria + dipstick twice or ++ dipstick once or > 0.3g/24h

IN: normotensive, non-proteinuric


18-26 200 (10.5) 19/21 113/179 MAP ≥ 90 Auscultatory blood pressure mearsurements, 10 min rest

SBP ≥ 140 or DBP ≥ 90 mmHg twice 6h apart after 24 weeks gestation; proteinuria ≥ 0.3g/24h

IN: normotensive, nulliparous women; EX: cardiovascular disease, urinary infection


18-26 108 (13.9) 14/15 46/93 MAP > 90 n.r. Gestose index n.r.

Page19

(1976)2nd trimester2nd trimester2nd trimester2nd trimester2nd trimester

14833 (2.7) 375/399337/399265/399174/399100/399

3353/144346500/1443410008/1443412594/1443413848/14434

MAP ≥ 75MAP ≥ 80MAP ≥85MAP ≥ 90MAP ≥ 95

n.r. Hypertension; proteinuria ≥ ++ dipstick or oedema

IN: patients of Child Health and Development Studies (California)

Reiss20

(1987)USA

2nd trimester2nd trimester

60 (50.0) 6/304/30

28/3028/30

MAP > 75MAP > 80

Right arm, resting in left lateral recumbent position, K4

rise DBP > 15 mmHg; proteinuria ≥ + dipstick

IN: hypertensive cases, controls matched for age, race gravidity, parity, an approximate delivery date (case control)

Robrecht21 (1980)Germany

14-2814-28

285 (19.6) 50/5635/56

156/229204/229

MAP ≥ 85MAP > 90

n.r. Gestose index ≥ 1: RR > 140/90 mmHg, proteinuria ≥ 0.5g/24h, pretibial oedema

n.r.

44

Rogers22 (1994)HongKong

18-26 220 (6.8) 14/15 125/205 MAP > 68 Dinamap 845A (Critikon Inc), 5-10 min rest in left lateral recumbent position, SBP, DBP and MAP were measured in right arm, resting on patients side, 1 min. intervals until stabilisation of blood pressure

RR ≥ 140/90 mmHg twice 4h apart; proteinuria ≥ ++ dipstick twice (=0.3g/L)

IN: normotensive, primigravid women. EX: congenital malformations

Shaarawy23 (2000)Egypt

2nd trimester 76 (40.8) 16/31 22/45 MAP > 80 Mild preeclampsia: SBP ≥ 140 or increase ≥ 30 mmHg systolic or DBP ≥ 90 or increase > 15 mmHg over baseline in first 20 weeks, proteinuria ≥ 0.3g/24h. Severe preeclampsia: SBP ≥ 160 or DBP ≥ 110 mmHg, proteinuria ≥ 5.0g/24h (oliguria, cerebral or visual disturbances, epigastric pain, pulmonary oedema)

IN: random selection; EX: history of chronic hypertension, diabetes, renal disease, RR > 140/90 mmHg at time of registration, evidence of proteinuria

Sibai24

(1997)USA

13-2113-2113-2113-2113-2113-2113-2113-2113-2113-2113-21

4314 (7.6) 315/326281/326230/326147/32675/32634/326291/326264/326189/326122/32637/326

451/39881116/39881937/39882847/39883385/39883797/3988675/39881148/39882416/39883121/39883714/3988

SBP > 95SBP > 100SBP > 105SBP > 110SBP > 115SBP > 120DBP > 50DBP > 55DBP > 60DBP > 65DBP > 70

Mercury sphygmomanometer, seated, 2 measurements, 1 min apart, mainly K5

DBP ≥ 90 mmHg; proteinuria ≥ + dipstick or ≥ 0.3g/24h

IN: healthy nulliparous women, no medical or obstetrical complications (RCT)

Stamilio25 (2000)USA

1st trimester24-28

1998 (2.5) 17/4914/49

1717/19491731/1949

MAP > 90MAP > 90

Screening Severe preeclampsia: SBP > 160 or DBP < 110 mmHg; proteinuria ≥ +++ dipstick or ≥ 0.3g/24h (oliguria, cerebral or visual disturbances, epigastric pain, pulmonary oedema/ cyanosis)

IN: all singleton pregnancies who underwent 2nd trimester triple screening

Tranquilli26 (2000)Italy

24-30 75 (18.7) 14/14 27/61 ABPM positive SpaceLabs 90207 (USA), 24h, oscillometric, every 30 minutes

RR > 140/90 mmHg; significant proteinuria

IN: primigravid women, normotensive; EX: history of chronic hypertension, fetal/ chromosomal abnormalities

Villar27 (1989)USA

13-2713-2713-2713-27

700 (19.6)700

554 (18.1)700

53/13730/13723/10011/137

450/563503/563422/563527/563

increase DBP ≥ 15increase SBP ≥ 30

increase SBP/DBP ≥ 30/15MAP ≥ 90

n.r. RR ≥ 140/90 mmHg twice ≥ 6h apart or increase 30/15 mmHg; proteinuria.

IN: young normotensive primigravid women

* Studies are cohort studies unless otherwise stated (case control or randomised controlled trial (RCT)). TP true positives; FP false positives; FN false negatives; TN true negatives; SBP systolic blood pressure; DBP diastolic blood pressure; MAP mean arterial pressure (calculated as (SBP+2*DBP)/3); ABPM ambulatory blood pressure monitoring; h hour; IN included; EX excluded.

Chapter 2

45


eta-analysis

Rogers22 (1994)HongKong

18-26 220 (6.8) 14/15 125/205 MAP > 68 Dinamap 845A (Critikon Inc), 5-10 min rest in left lateral recumbent position, SBP, DBP and MAP were measured in right arm, resting on patients side, 1 min. intervals until stabilisation of blood pressure

RR ≥ 140/90 mmHg twice 4h apart; proteinuria ≥ ++ dipstick twice (=0.3g/L)

IN: normotensive, primigravid women. EX: congenital malformations

Shaarawy23 (2000)Egypt

2nd trimester 76 (40.8) 16/31 22/45 MAP > 80 Mild preeclampsia: SBP ≥ 140 or increase ≥ 30 mmHg systolic or DBP ≥ 90 or increase > 15 mmHg over baseline in first 20 weeks, proteinuria ≥ 0.3g/24h. Severe preeclampsia: SBP ≥ 160 or DBP ≥ 110 mmHg, proteinuria ≥ 5.0g/24h (oliguria, cerebral or visual disturbances, epigastric pain, pulmonary oedema)

IN: random selection; EX: history of chronic hypertension, diabetes, renal disease, RR > 140/90 mmHg at time of registration, evidence of proteinuria

Sibai24

(1997)USA

13-2113-2113-2113-2113-2113-2113-2113-2113-2113-2113-21

4314 (7.6) 315/326281/326230/326147/32675/32634/326291/326264/326189/326122/32637/326

451/39881116/39881937/39882847/39883385/39883797/3988675/39881148/39882416/39883121/39883714/3988

SBP > 95SBP > 100SBP > 105SBP > 110SBP > 115SBP > 120DBP > 50DBP > 55DBP > 60DBP > 65DBP > 70

Mercury sphygmomanometer, seated, 2 measurements, 1 min apart, mainly K5

DBP ≥ 90 mmHg; proteinuria ≥ + dipstick or ≥ 0.3g/24h

IN: healthy nulliparous women, no medical or obstetrical complications (RCT)

Stamilio25 (2000)USA

1st trimester24-28

1998 (2.5) 17/4914/49

1717/19491731/1949

MAP > 90MAP > 90

Screening Severe preeclampsia: SBP > 160 or DBP < 110 mmHg; proteinuria ≥ +++ dipstick or ≥ 0.3g/24h (oliguria, cerebral or visual disturbances, epigastric pain, pulmonary oedema/ cyanosis)

IN: all singleton pregnancies who underwent 2nd trimester triple screening

Tranquilli26 (2000)Italy

24-30 75 (18.7) 14/14 27/61 ABPM positive SpaceLabs 90207 (USA), 24h, oscillometric, every 30 minutes

RR > 140/90 mmHg; significant proteinuria

IN: primigravid women, normotensive; EX: history of chronic hypertension, fetal/ chromosomal abnormalities

Villar27 (1989)USA

13-2713-2713-2713-27

700 (19.6)700

554 (18.1)700

53/13730/13723/10011/137

450/563503/563422/563527/563

increase DBP ≥ 15increase SBP ≥ 30

increase SBP/DBP ≥ 30/15MAP ≥ 90

n.r. RR ≥ 140/90 mmHg twice ≥ 6h apart or increase 30/15 mmHg; proteinuria.

IN: young normotensive primigravid women


46

Table 2 Characteristics of studies of blood pressure measurement used to predict preeclampsia in pregnant women at high risk First author (year) Country

Gestation(weeks)

N(% preeclampsia)



Index test Threshold (mmHg)



Bar28

(1999)Israel

14-28 60 (26.7) 13/16 34/44 ABPM RR ≥135/85

(mean daytime)

ABPM: SpaceLabs 90202 (USA), non-dominant arm, every 30 min;White coat hypertension: standard mercury sphygmomanometer, seated, larger bladder for arms larger than 32 cm, mean of 3 measurements, right arm, being supine for 10min, K5

Increase > 25 mmHg with a maximum value of DBP > 90 mmHg; proteinuria ≥ 0.3g/24h

IN: women with high blood pressure were referred from community clinics; EX: women with normal blood pressure levels

Caritis29 (1998)USA

13-2613-26

2503 (19.8) 445/495319/485

459/20181154/2018

MAP > 75 (baseline)MAP > 85(baseline)

Seated, K5 SBP ≥ 140 or DBP ≥ 90 mmHg twice > 4h apart; ≥ ++ dipstick twice or ≥ 0.3g/24h without urinary tract infection

IN: type 1 diabetes, chronic hypertension, multiple pregnancies, previous preeclampsia (RCT)

Konijnenberg30 (1997)Netherlands

13-20 244 (7.0) 12/17 170/227 DBP ≥ 75 first antenatal DBP, standard sphygmomanometer

DBP ≥ 90 mmHg; proteinuria ≥ 0.3g/24h

IN: women with increased risk: multiple pregnancies, known pre-existing diseases such as hypertension, diabetes mellitus, vasculitis, history of preeclampsia or fetal growth restriction not caused by congenital malformations or infectious fetal diseases

Lauszus31 (2001)Denmark

1st trimester 87 (28.7) 17/25 43/62 RR >122/74 SpaceLabs 90207 (Redmond USA) oscillometry, cuff insufflation every 20 min from 6-23 hrs and every hour during night, non-dominant arm, average of 3 oscillometric and 3 auscultatory (random-zero sphygmomanometer (Haweeksley, Lancing UK)) measurements. Mean of day- and nighttime measurements


IN: pregestational type 1 diabetes patients

Penny32

(1998)UK

> 20 (median 36)> 20 (median 36)> 20 (median 36)> 20 (median 36)

347 (3.7)270 (3.3)347 (3.7)270 (3.3)

4/134/95/135/9

260/334176/261234/334156/261

SBP > 140SBP > 135DBP > 90DBP > 85

Standardised obstetric day unit (ODU) assessment with up to 5 measurements, trained observer, 30 min intervals, conventional mercury sphygmomanometer, K5, seated, encouraged to rest. ABPM: SpaceLabs 90207, 24hr, every 30 min ODU ABPM

RR > 140/90 mmHg; proteinuria ≥ ++ dipstick or ≥ 0.5g/24hRR> 135/85 mmHg; proteinuria ≥ ++ dipstick or ≥ 0.5g/24hRR > 140/90 mmHg; proteinuria ≥ ++ dipstick or ≥ 0.5g/24hRR > 135/85 mmHg; proteinuria ≥ ++ dipstick or ≥ 0.5g/24h

IN: chronic hypertensive patients without antihypertensive treatment (proteinuria >2 weeks after index test measurement)

Perini33 (2004)Italy

21-35,4 22 (27.3) 5/6 9/16 MESOR diastolic ≥ 68

24h, every 30 min, diastolic midline estimating statistic of rhythm (MESOR)

SBP ≥ 140 or DBP ≥ 90 mmHg; proteinuria ≥ 0.3g/24h

IN: 17 patients with increased resistance index (> 0.62); 5 patients with history of preeclampsia

Valensise34 (1995)Italy

21-29 (mean 24,6)

48 (2.0) 1/1 30/47 DBP > 68 ABPM, Space labs 90207 (USA), appriopriate cuff, oscillometric, self-calibrating, every 15 min 8-22h/ every 30min 20-8h, mean diastolic blood pressure (M24h DBP)

DBP > 90 mmHg twice consecutive 4h apart; proteinuria ≥ 0.3g/24h

IN: healthy at recruitment, normotensive, high uterine artery resistance index; EX: chronic hypertension, diabetes, renal disease, fetal abnormality, fetal growth restriction, oligohydramnios


Chapter 2

47


eta-analysis

Table 2 Characteristics of studies of blood pressure measurement used to predict preeclampsia in pregnant women at high risk First author (year) Country

Gestation(weeks)

N(% preeclampsia)



Index test Threshold (mmHg)



Bar28

(1999)Israel

14-28 60 (26.7) 13/16 34/44 ABPM RR ≥135/85

(mean daytime)

ABPM: SpaceLabs 90202 (USA), non-dominant arm, every 30 min;White coat hypertension: standard mercury sphygmomanometer, seated, larger bladder for arms larger than 32 cm, mean of 3 measurements, right arm, being supine for 10min, K5

Increase > 25 mmHg with a maximum value of DBP > 90 mmHg; proteinuria ≥ 0.3g/24h

IN: women with high blood pressure were referred from community clinics; EX: women with normal blood pressure levels

Caritis29 (1998)USA

13-2613-26

2503 (19.8) 445/495319/485

459/20181154/2018

MAP > 75 (baseline)MAP > 85(baseline)

Seated, K5 SBP ≥ 140 or DBP ≥ 90 mmHg twice > 4h apart; ≥ ++ dipstick twice or ≥ 0.3g/24h without urinary tract infection

IN: type 1 diabetes, chronic hypertension, multiple pregnancies, previous preeclampsia (RCT)

Konijnenberg30 (1997)Netherlands

13-20 244 (7.0) 12/17 170/227 DBP ≥ 75 first antenatal DBP, standard sphygmomanometer


IN: women with increased risk: multiple pregnancies, known pre-existing diseases such as hypertension, diabetes mellitus, vasculitis, history of preeclampsia or fetal growth restriction not caused by congenital malformations or infectious fetal diseases

Lauszus31 (2001)Denmark

1st trimester 87 (28.7) 17/25 43/62 RR >122/74 SpaceLabs 90207 (Redmond USA) oscillometry, cuff insufflation every 20 min from 6-23 hrs and every hour during night, non-dominant arm, average of 3 oscillometric and 3 auscultatory (random-zero sphygmomanometer (Haweeksley, Lancing UK)) measurements. Mean of day- and nighttime measurements


IN: pregestational type 1 diabetes patients

Penny32

(1998)UK

> 20 (median 36)> 20 (median 36)> 20 (median 36)> 20 (median 36)

347 (3.7)270 (3.3)347 (3.7)270 (3.3)

4/134/95/135/9

260/334176/261234/334156/261

SBP > 140SBP > 135DBP > 90DBP > 85

Standardised obstetric day unit (ODU) assessment with up to 5 measurements, trained observer, 30 min intervals, conventional mercury sphygmomanometer, K5, seated, encouraged to rest. ABPM: SpaceLabs 90207, 24hr, every 30 min ODU ABPM

RR > 140/90 mmHg; proteinuria ≥ ++ dipstick or ≥ 0.5g/24hRR> 135/85 mmHg; proteinuria ≥ ++ dipstick or ≥ 0.5g/24hRR > 140/90 mmHg; proteinuria ≥ ++ dipstick or ≥ 0.5g/24hRR > 135/85 mmHg; proteinuria ≥ ++ dipstick or ≥ 0.5g/24h

IN: chronic hypertensive patients without antihypertensive treatment (proteinuria >2 weeks after index test measurement)

Perini33 (2004)Italy

21-35,4 22 (27.3) 5/6 9/16 MESOR diastolic ≥ 68

24h, every 30 min, diastolic midline estimating statistic of rhythm (MESOR)

SBP ≥ 140 or DBP ≥ 90 mmHg; proteinuria ≥ 0.3g/24h

IN: 17 patients with increased resistance index (> 0.62); 5 patients with history of preeclampsia

Valensise34 (1995)Italy

21-29 (mean 24,6)

48 (2.0) 1/1 30/47 DBP > 68 ABPM, Space labs 90207 (USA), appriopriate cuff, oscillometric, self-calibrating, every 15 min 8-22h/ every 30min 20-8h, mean diastolic blood pressure (M24h DBP)

DBP > 90 mmHg twice consecutive 4h apart; proteinuria ≥ 0.3g/24h

IN: healthy at recruitment, normotensive, high uterine artery resistance index; EX: chronic hypertension, diabetes, renal disease, fetal abnormality, fetal growth restriction, oligohydramnios


48

Reference List (1) Ales KL, Norton ME, Druzin ML. Early prediction of antepartum hypertension. Obstet Gynecol

1989; 73(6):928-933. (2) Atterbury JL, Groome LJ, Baker SL. Elevated midtrimester mean arterial blood pressure in

women with severe preeclampsia. Appl Nurs Res 1996; 9(4):161-166. (3) Conde-Agudelo A, Belizan JM, Lede R, Bergel EF. What does an elevated mean arterial pressure

in the second half of pregnancy predict--gestational hypertension or preeclampsia? Am J Obstet Gynecol 1993; 169(3):509-514.

(4) Fallis NE, Langford HG. Relation of second trimester blood pressure to toxemia of pregnancy in the primigravid patient. Am J Obstet Gynecol 1963; 87(1):123-125.

(5) Friedman EA, Neff RK. Systolic and mean arterial blood pressure. In: Friedman EA, Neff RK, editors. Pregnancy Hypertension. A systematic evaluation of clinical diagnostic criteria. Littleton, Massachusetts: PSG Publishing Company Inc; 1977. 212-219.

(6) Higgins JR, Walshe JJ, Halligan A, O’Brien E, Conroy R, Darling MR. Can 24-hour ambulatory blood pressure measurement predict the development of hypertension in primigravidae? Br J Obstet Gynaecol 1997; 104(3):356-362.

(7) Iwasaki R, Ohkuchi A, Furuta I, Ojima T, Matsubara S, Sato I et al. Relationship between blood pressure level in early pregnancy and subsequent changes in blood pressure during pregnancy. Acta Obstet Gynecol Scand 2002; 81(10):918-925.

(8) Knuist M, Bonsel GJ, Zondervan HA, Treffers PE. Risk factors for preeclampsia in nulliparous women in distinct ethnic groups: a prospective cohort study. Obstet Gynecol 1998; 92(2):174-178.

(9) Kyle PM, Clark SJ, Buckley D, Kissane J, Coats AJ, de Swiet M et al. Second trimester ambulatory blood pressure in nulliparous pregnancy: a useful screening test for pre-eclampsia? Br J Obstet Gynaecol 1993; 100(10):914-919.

(10) Lopez MC, Belizan JM, Villar J, Bergel E. The measurement of diastolic blood pressure during pregnancy: which Korotkoff phase should be used? Am J Obstet Gynecol 1994; 170(2):574-578.

(11) Mahanna I, Algeri T, Cigarini C, Zinelli G. [Arterial pressure, MAP and dynamic tests in the monitoring of pregnancy]. Ann Ostet Ginecol Med Perinat 1983; 104(4):248-255.

(12) Moutquin JM, Rainville C, Giroux L, Raynauld P, Amyot G, Bilodeau R et al. A prospective study of blood pressure in pregnancy: prediction of preeclampsia. Am J Obstet Gynecol 1985; 151(2):191-196.

(13) Moutquin JM, Bilodeau R, Raynault P, Amyot G, Blair JF, Labelle L et al. [Prospective study of arterial pressure during pregnancy. Prediction of hypertensive complications]. J Gynecol Obstet Biol Reprod (Paris) 1982; 11(7):833-837.

(14) Moutquin JM, Rainville C, Giroux L, Raynauld P, Bilodeau R, Amyot G et al. Is a threshold increase in blood pressure predictive of preeclampsia? A prospective cohort study. Clin Exp Hypertens B 1990; 9(2):225-235.

(15) Odegard RA, Vatten LJ, Nilsen ST, Salvesen KA, Austgulen R. Risk factors and clinical manifestations of pre-eclampsia. BJOG 2000; 107(11):1410-1416.

(16) Ohkuchi A, Iwasaki R, Ojima T, Matsubara S, Sato I, Suzuki M et al. Increase in systolic blood pressure of > or = 30 mm Hg and/or diastolic blood pressure of > or = 15 mm Hg during pregnancy: is it pathologic? Hypertens Pregnancy 2003; 22(3):275-285.

(17) Oney T, Kaulhausen H. The value of the mean arterial blood pressure in the second trimester (MAP-2 value) as a predictor of pregnancy-induced hypertension and preeclampsia. A preliminary report. Clin Exp Hypertens B 1983; 2(2):211-216.

(18) Oney T, Balogh A, Kaulhausen H. The predictive value of blood pressure and weight gain during pregnancy for the early diagnosis of gestosis/preeclampsia. Preliminary report. [German]. Fortschritte der Medizin 1982; 100(7):277-280.

Chapter 2

49


eta-analysis

(19) Page EW, Christianson R. The impact of mean arterial pressure in the middle trimester upon the outcome of pregnancy. Am J Obstet Gynecol 1976; 125(6):740-746.

(20) Reiss RE, O’Shaughnessy RW, Quilligan TJ, Zuspan FP. Retrospective comparison of blood pressure course during preeclamptic and matched control pregnancies. Am J Obstet Gynecol 1987; 156(4):894-898.

(21) Robrecht D, Schriever M, Rasenack R, Steiner H, Kaltenbach FJ. [The mean blood pressure in the second trimester (map-2) as a valuable aid in the early recognition of the pregnancies with a risk of hypertension (author’s transl)]. Geburtshilfe Frauenheilkd 1980; 40(2):121-124.

(22) Rogers MS, Chung T, Baldwin S. A reappraisal of second trimester mean arterial pressure as a predictor of pregnancy induced hypertension. J Obstet Gynaecol 1994; 14(4):232-236.

(23) Shaarawy M, Abdel-Magid AM. Plasma endothelin-1 and mean arterial pressure in the prediction of pre-eclampsia. Int J Gynaecol Obstet 2000; 68(2):105-111.

(24) Sibai BM, Ewell M, Levine RJ, Klebanoff MA, Esterlitz J, Catalano PM et al. Risk factors associated with preeclampsia in healthy nulliparous women. The Calcium for Preeclampsia Prevention (CPEP) Study Group. Am J Obstet Gynecol 1997; 177(5):1003-1010.

(25) Stamilio DM, Sehdev HM, Morgan MA, Propert K, Macones GA. Can antenatal clinical and biochemical markers predict the development of severe preeclampsia? Am J Obstet Gynecol 2000; 182(3):589-594.

(26) Tranquilli AL, Giannubilo SR, Bezzeccheri V, Garbati E. [The relative weight of Doppler of the uterine artery and 24-h ambulatory blood pressure monitoring in predicting hypertension in pregnancy and preeclampsia]. Acta Biomed Ateneo Parmense 2000; 71 Suppl 1:351-355.

(27) Villar MA, Sibai BM. Clinical significance of elevated mean arterial blood pressure in second trimester and threshold increase in systolic or diastolic blood pressure during third trimester. Am J Obstet Gynecol 1989; 160(2):419-423.

(28) Bar J, Maymon R, Padoa A, Wittenberg C, Boner G, Ben Rafael Z et al. White coat hypertension and pregnancy outcome. J Hum Hypertens 1999; 13(8):541-545.

(29) Caritis S, Sibai B, Hauth J, Lindheimer M, Vandorsten P, Klebanoff M et al. Predictors of pre-eclampsia in women at high risk. National Institute of Child Health and Human Development Network of Maternal-Fetal Medicine Units. Am J Obstet Gynecol 1998; 179(4):946-951.

(30) Konijnenberg A, van der Post JA, Mol BW, Schaap MC, Lazarov R, Bleker OP et al. Can flow cytometric detection of platelet activation early in pregnancy predict the occurrence of preeclampsia? A prospective study. Am J Obstet Gynecol 1997; 177(2):434-442.

(31) Lauszus FF, Rasmussen OW, Lousen T, Klebe TM, Klebe JG. Ambulatory blood pressure as predictor of preeclampsia in diabetic pregnancies with respect to urinary albumin excretion rate and glycemic regulation. Acta Obstet Gynecol Scand 2001; 80(12):1096-1103.

(32) Penny JA, Halligan AW, Shennan AH, Lambert PC, Jones DR, de Swiet M et al. Automated, ambulatory, or conventional blood pressure measurement in pregnancy: which is the better predictor of severe hypertension? Am J Obstet Gynecol 1998; 178(3):521-526.

(33) Perini R, Fisogni C, Bonera R, Rovetto B, Frusca T, Lojacono A et al. [Role of Doppler velocimetry of uterine arteries and ambulatory blood pressure monitoring in detecting pregnancies at risk for preeclampsia]. Minerva Ginecol 2004; 56(2):117-123.

(34) Valensise H, Tranquilli AL, Arduini D, Garzetti GG, Romanini C. Screening pregnant women at 22-24 weeks for gestational hypertension or intrauterine growth retardation by Doppler ultrasound followed by 24-hour blood pressure recording. Hypertens Pregnancy 1995; 14(3):351-359.

Chapter 2

51


eta-analysis

13Karlijn C. VollebregtJanneke GisolfIlja GuelenKees BoerGert A. van MontfransHans Wolf

Journal of Hypertension 2010;28(1):127-134

Limited accuracy of the hyperbaric index, ambulatory blood pressure and sphygmomanometry measurements in predicting gestational hypertension and preeclampsia

54

ABSTRACT

Objective The aim of this study was to validate the hyperbaric index (HBI) for first trimester prediction

of preeclampsia and gestational hypertension.

Methods

Participants were low-risk and high-risk nulliparous women and high-risk multiparous

women, and were recruited between April 2004 and June 2006. At a gestational age of

9 weeks (range 8 – 11 weeks), blood pressure was measured first by sphygmomanometry

and thereafter by ambulatory blood pressure measurement (ABPM) for 48 h. The first

90 low-risk women who had an uneventful pregnancy formed the reference group for

calculation of a time-specified tolerance interval with 90% confidence limits. In the

validation group, consisting of the remaining women, the HBI was calculated as the

time-specified blood pressure excess over this tolerance limit for SBP, DBP and mean

arterial pressure.

Results

The validation group contained 101 women. Fifteen developed preeclampsia and 13

gestational hypertension. For preeclampsia the maximum HBI had the best predictive

capacity with a sensitivity of 73 % and a specificity of 86 %. However, the difference

with standard ABPM measurement or sphygmomanometry was small with a sensitivity

between 75 % and 73 % and a specificity between 86 % and 95 %. The predictive efficacy

for gestational hypertension predictive capacity was poor with all methods (sensitivity

between 54 % and 77 %, specificity between 41 % and 78 %).

Conclusion

Standardized sphygmomanometry, ABPM measurement and the hyperbaric index

calculated from 48 h ABPM had a comparable, restricted predictive efficacy. The

high predictive value of hyperbaric index as observed in earlier studies could not be

reproduced.

Chapter 3

55

The Hyperbaric Index

INTRODUCTION

Preeclampsia is one of the main causes of maternal and fetal morbidity and mortality

worldwide. 1 The exact pathophysiology is unknown despite extensive research. Many

hypotheses have been proposed and rejected. Although preeclampsia is usually

diagnosed in the second half of pregnancy, it is hypothesized that the disease originates

early in pregnancy with defective placentation.2 Early selection of women at high risk

for this condition could offer the opportunity to target care at those most likely to benefit

and to evaluate or design preventive strategies. However, currently available diagnostic

tests have limited predictive efficacy with likelihood ratio’s for a positive test of 2-5 and

likelihood ratio for a negative test of 0.2–0.8.2-4

Blood pressure measurement as a predictive tool has been the focus of research for many

years. Although women who will develop preeclampsia later in pregnancy have a slightly

higher blood pressure in the beginning of pregnancy5-7 the discriminative capacity of blood

pressure is disappointing, irrespective of blood pressure measurement technique. 4;6;7

Ambulatory blood pressure measurement (APBM) is a well known technique for the

diagnosis and treatment evaluation of hypertension in nonpregnant women. ABPM

correlates better with hypertension-related organ damage and cardiovascular events

than office blood pressure.8 Using ABPM, a higher blood pressure was observed in

mid-pregnancy in women who later experienced the development of preeclampsia or

gestational hypertension. Although the predictive efficacy of ABPM was higher than

standard blood pressure measurement the predictive capacity appeared to low to be

useful in daily clinical practice.6;9-11

Most studies regarding the prediction of preeclampsia or gestational hypertension

present 24-h means. Although some studies report that diurnal differences can be used

for assessing disease severity, usefulness for prediction has not been reported.12;13 A

more sophisticated technique for evaluation of ABPM was presented by Hallberg et al.,

using the hyperbaric index (HBI).14 HBI was calculated as the amount of blood pressure

excess during the measurement period above a 90% tolerance limit expressed as mmHg

multiplied by hour. Studies by Hermida et al. evaluated HBI as a test for prediction

of preeclampsia and gestational hypertension.15-18 These studies observed a sensitivity

of 94% and a specificity of 100% using HBI calculation from 48-h ABPM. Another

group studied HBI, calculated in a different way, in a hospital-based research setting

and observed a sensitivity of 80% and specificity of 77% at 8-16 weeks for gestational

hypertension and preeclampsia.19

56

Because the results from the studies by Hermida et al. seemed promising and the research

setting between all studies differed we decided to perform a validation study in our own

population.15-19

METHODS

Subjects This research is part of a prospective study, which investigates cardiovascular parameters

in the beginning of pregnancy for the prediction of preeclampsia and fetal growth

restriction. Healthy nulliparous women at low risk and women with elevated risk for

preeclampsia or fetal growth restriction were included between April 2004 and June

2006. Low-risk participants were recruited by advertisements in local newspapers and

flyers at midwife practices near our hospital. Women having diabetes, renal disease,

clotting disorders or a previous pregnancy complicated by preeclampsia were defined

as high risk. They were recruited from our antenatal clinic (Academic Medical Center,

Amsterdam, The Netherlands) and by advertisements in local newspapers. At inclusion,

all women had normal blood pressure (less than 140 mmHg SBP and 90 mmHg DBP,

measured auscultatory) and none used antihypertensive medication. Only women with

a singleton pregnancy were included. Three women who delivered an infant with major

congenital malformation were excluded.

After written informed consent was obtained, all participants underwent identical

study protocols. At 8-11 weeks, blood pressure was measured first by conventional

sphygmomanometry and thereafter for 48 h using an ambulatory blood pressure device

(see below). Gestational age was defined by first trimester ultrasound. The results of

the ambulatory blood pressure measurements were not available for the clinicians/

midwives responsible for prenatal care. Pregnancy outcome was classified as ‘gestational

hypertension’, ‘preeclampsia’ or ’normal’ using blood pressure measurements that

were obtained during routine clinical care. Two obstetricians (H.W. and K.B.) who were

unaware of the results of the early gestational blood pressure measurements reviewed

medical files for classification.

The Medical Ethical Committee of the Academic Medical Center approved the study.

Definitions

Gestational hypertension was defined as a systolic blood pressure (SBP) of at least

140, diastolic blood pressure (DBP) of at least 90 mmHg or both after 20 weeks in a

previously normotensive woman, measured twice with an interval of at least 6 h. Blood

pressure was measured under routine clinical conditions. Preeclampsia was defined as

Chapter 3

57


the combination of gestational hypertension and proteinuria of at least 0.3 g/ 24 h or

dipstick of at least ++ after 20 weeks gestation according to the International Society for

the Study of Hypertension in Pregnancy (ISSHP) guidelines.20

Instruments

Blood pressure was measured by aired sphygmomanometry in supine position at the

nondominant arm after at least 15 minutes of rest in a quiet room with an ambient

temperature between 20-22° C between 0700 and 1200 h. Korotkoff V was used to

determine DBP. The average of three measurements with an interval of at least 1 minute

was used for calculations. The mean arterial pressure (MAP) was calculated as ((2x DBP)

+ SBP)/3. ABPM was performed by oscillometry (Spacelab 90207, Redmond, Wasington,

USA). An appropriately sized cuff was placed on the non-dominant arm. Blood pressure

was measured every 30 minutes between 06:00 h till midnight and every 60 minutes

between midnight and 0600 h. All women started the measurements between 0700 and

1200 h and continued for 48 consecutive hours. If blood pressure results were missing

for more than 2 h the participant was excluded from further analysis.21 The participants

were instructed to maintain their daily routine and to complete a diary to identify time of

sleeping and awakening. Mean 24-h blood pressure, daytime blood pressure and, night-

time blood pressure using fixed intervals (between 0600 h till midnight and between

midnight and 0600 h) and night to day ratio were calculated.22 Women were defined as

‘dippers’ when their night to day ratio SBP was 0.90 or less.23All women were instructed

and measured by two trained researchers.

Computation of time-specified tolerance interval of blood pressure and hyperbaric index

The calculation of nonparametric tolerance intervals for hybrid time series and HBI was

performed according to the method described by Hermida et al.17;18 Measurement data

from the first 90 nulliparous women who had an uneventful pregnancy and delivered

at term of an infant with normal birthweight24 were used for calculation of the time-

specified tolerance limits (reference group). The remaining 101 women, including those

with gestational hypertension or preeclampsia, constituted the validation group.

On the basis of self-reported sleeping and waking times, data were synchronized at

mid-sleep. Because the time of blood pressure registration differed between participants,

2-hour time classes were defined at each complete hour of day (running from one before

to one hour after each complete hour), resulting in overlapping time classes. Individual

data were averaged per time class, producing one blood pressure result per individual

per time class. For each time class, the 90% confidence limits were calculated across all

participants. Because the reference group was small and normality and symmetry of the

58

data could not be assumed, a nonparametric approach was chosen using bootstrapping,

as described by Hermida et al.17 The tolerance interval was calculated to include at least

a specified proportion of the population (1-β) with a stated confidence (1-α). We chose

to compute for given time class a tolerance interval which included at least 90% of the

population with a confidence of 90% (β= 0.1 and α = 0.1). The assumption was that the

sample was a reasonable representation of the underlying population. Data in each time

class was bootstrapped (resampled with replacement); the bootstrapped data contained

the same number of records as the original sample. Bootstrapping was repeated until

the standard error of the mean (SEM) was less than 1% of the mean (the Monte Carlo

stopping rule).

For computing tolerance intervals, the mean and variance of the sample and of each

of the bootstrapped resample were calculated. For each resample a lower (ß/2, e.g.

5%) and upper (1- ß/2, e.g. 95%) percentile was determined, thus creating a large

number of upper and lower limits for each time class (one for each bootstrap resample).

Because we were interested in the proportion of the population with confidence 1- α,

from all the (bootstrap) lower percentiles, the 1-√(1- α) percentile was determined: the

lower tolerance limit. From the upper percentiles, the √(1- α) percentile was determined:

the upper tolerance limit. There is √(1- α) probability that a sample from the specified

population lies over the upper tolerance limit; there is also √(1- α) probability that the

130

120

110

100

90

0 500 1000 1500

100

90

80

70

600 500 1000 1500

110

100

90

80

70

60

500 500 1000 1500

80

70

60

50

0 500 1000 1500

SBP

(mm

Hg)

MAP

(mm

Hg)

Relative time of day [min] Relative time of day [min]

Relative time of day [min] Relative time of day [min]

DBP

(mm

Hg)

Hea

rt ra

te (b

pm)

Figure 1. Time-specified tolerance intervals with 90% confidence limits for SBP, DBP, mean arterial pressure and heart rate in the reference population (n=90) synchronized at midsleep.

Chapter 3

59


sample lies under the lower tolerance limit: the probability that the sample lies between

both tolerance limits is therefore 1- α (rule of conditional probability). Upper and lower

tolerance intervals were computed for all time classes, see figure 1.

The size of the reference population was determined by stepwise increment of the

reference population by 10 women and calculation of the standard error (SE) of the

90th percentile tolerance limit. After a group size of 70, further increase resulted in a

variation of the SE of less than 2 mmHg. Therefore calculations of the tolerance limits

were discontinued at a group size of 90.

The HBI was calculated for the SBP, DBP and MAP as the individual excess above the

reference threshold (the upper limit of the tolerance interval) using numerical integration

(for an example see figure 2). HBImax was defined as the maximum value of either the HBI

90

80

70

MAP

(mm

Hg)

100

0 500 1000 1500Relative time of day (min)

Figure 2. Plot of the mean arterial pressure of a women with elevated hyperbaric index. The dashed line is the upper 90th percentile tolerance limit of the MAP. The curve starts at midsleep. The continious line is the individual pressure. Marked area is the hyperbaric index.

of SBP, DBP or MAP for each individual. Time-specified tolerance intervals and HBI were

calculated using Matlab® R2007b (The Mathworks, Inc, Natick, Massachusetts, USA).

One-way analysis of variance was used to determine differences between the outcome

groups and multiple post hoc comparisons were performed by Bonferroni t-test or

Kruskal -Wallis test as appropriate. Receiver Operator Characteristics (ROC) curves

were constructed to test the discriminate value of sphygmomanometry measurements,

chronobiological blood pressure parameters and the HBI in predicting gestational

hypertension, preeclampsia or both. We considered an area under the curve less than

0.75 as poor discriminative capacity. Between 0.75-0.90 was considered as fair and

between 0.90-0.97 was considered as good. The optimal cut-off value was chosen using

the maximum Youden index (sensitivity + specificity - 1).25 Statistical calculations were

performed with SPSS 16.0.2(SPSS Inc.Chicago, Illinois, USA). The level of significance

used was a P value of les than 0.05.

Study size calculation was based on the test characteristics observed by Hermida et

al.16;17 We expected that HBI could predict preeclampsia with a sensitivity and specificity

60

of 0.9 or higher. One hundred women with an incidence of preeclampsia of 7 % would

be necessary to determine a statistically significant predictive effect with an α equal to

0.05 and a β equal to 0.8, calculated by chi square test, tested two-sided. Exclusion of

the first 90 women with uneventful pregnancies, who were selected for calculation of

tolerance limits, was expected to increase the incidence of preeclampsia, which would

elevate the power of our analysis. Sensitivity, specificity and area under the ROC curve

are independent of incidence and therefore not influenced by this selection.

Table 1.Characteristics of the reference group and study populationReference population

n=90

Normotensive women n=73

Gestational hypertension

n=13Preeclampsia

n=15

Age(years) 30.2 ± 3.7 30.5 ± 4.1 32.4 ± 2.9 29.1 ± 4.5

BMI(kg/m2) 23.7 ± 3.2 23.8 ± 3.9 25.6 ± 3.0 27.3 ± 4.2*

High risk women(%) 0% 37% 31% 53%

History of preeclampsia(%) 0% 26% 23 % 47 %

Nulliparous(%) 100% 67 % 69 % 47 %

Sphygmomanometry

SBP(mmHg) 104 ± 9 103 ± 10 110 ± 9 114 ± 15*

DBP(mmHg) 61 ± 7 61 ± 7 66 ± 7 69 ± 12*

MAP(mmHg) 76 ± 7 75 ± 7 81 ± 8‡ 84 ± 13*

Mean ABPM

SBP(mmHg) 110 ± 12 111 ± 7 114 ± 8 119 ± 14*

DBP(mmHg) 65 ± 4 65 ± 5 69 ± 6 72 ± 9*

MAP(mmHg) 80 ± 5 83 ± 5 86 ± 7 90 ± 10*

HBI

SBP(mmH x hour) 4.1 ± 17 6.0 ± 22 4.9 ± 9 43.0 ± 54*†

DBP(mmH x hour) 2.9 ± 9 8.0 ± 40 26.2 ± 54 46.7 ± 52*

MAP(mmH x hour) 2.8 ± 9 7.3 ± 36 20.6 ± 41 49.0 ± 57*

Dippers 56% 48% 54% 60%

Gestational age at delivery(weeks) 40.1 ± 1.1 38.7 ± 2.6 39.4 ± 1.6 34.6 ± 3.6*†

Delivery < 37 weeks 0% 16,4% 7.7% 53.3% *†

Birthweight (g) 3616 ± 420 3122 ± 607 3316 ± 602 1959 ± 828*†

Small for gestational age(%) 0 24.7 30.8 73.3*†

Variables are expressed as mean ± SD or percentage. All measurements were in the first trimester of pregnancy except pregnancy outcome GH: gestational hypertension. PE: preeclampsia, SBP: systolic blood pressure, DBP: diastolic blood pressure, MAP: mean arterial pressure, Mean ABPM: mean 24h blood pressure measured with ambulatory blood pressure monitoring.*uncomplicated vs. preeclampsia P<0.05, † gestational hypertension vs. preeclampsia P<0.05, ‡ uncomplicated vs. gestational hypertension P<0.05

Chapter 3

61


RESULTS

Two hundred and nineteen women participated in the study. The mean gestational age

at the beginning of the measurements was 9 weeks and 3 days (SD 6 days). Data of

28 women were not usable because of missing measurements or failure of the device.

The main reason for missing measurements was discontinuation of wearing the device

because women could not sleep with the device. After calculation of the time specified

tolerance intervals with 90% confidence limits using measurements in the first 90

nulliparous women the data of the remaining 101 women were available for evaluation

of the method. Patient characteristics of the study population are given in table 1. In

the validation group 38 high risk women were included; 36 women had preeclampsia

in a previous pregnancy, 1 woman had diabetes mellitus, and 1 woman had severe

fetal growth retardation in a previous pregnancy. Seventy-three women had normal

blood pressure throughout pregnancy, 13 developed gestational hypertension and 15

preeclampsia (Table 1).

30 40 50 60 70 80 90 100

HBI max

HBI MAP

HBI DBP

HBI SBP

ABPM MAP

ABPM DBP

ABPM SBP

MAP

DBP

SBP

20

Sensitivity 95% Cl

Specificity 95% Cl

40 60 80 100

HBI max

HBI MAP

HBI DBP

HBI SBP

ABPM MAP

ABPM DBP

ABPM SBP

MAP

DBP

SBP

Preeclampsia(a) (b) Gestational Hypertention

Figure 3. (a) Sensitivity and specificity of sphygmomanometry, ambulatory blood pressure monitoring and Hyperbaric index for predicting preeclampsia with 95% confidence interval. (b) Sensitivity and specificity of sphygmomanometry, ABPM and HBI for predicting gestational hypertension with 95% CI. ABPM, ambulatory blood pressure monitoring; CI, confidence interval; DBP, diastolic blood pressure; HBI, hyperbaric index; MAP, mean arterial pressure; SBP, systolic blood pressure.

62

BMI, blood pressure measurements with sphygmomanometry and ABPM, and HBI differed

significantly between normotensive women and women who developed preeclampsia

(Table 1). Only sphygmomanometry MAP differed significantly between normotensive

women and women who developed gestational hypertension. The ratio of day time and

night time blood pressure and the percentage of ‘dippers’ were not significantly different

between the groups.

Sphygmomanometry, mean ABPM and HBI had a comparable discriminative efficacy for

gestational hypertension (sensitivity between 54 and 77%, specificity between 41 and

78%), see figure 3B. Only MAP and DBP had a statistically significant effect. The area

under the curve for all parameters was between 0.57 and 0.66 which signifies poor test

characteristics.

The results of the different measurement techniques for the prediction of preeclampsia

did not vary largely (sensitivity between 53 and 73%, specificity between 86 and 95%),

see figure 3A. All results reached statistical significance. HBImax had the best discrimination

of preeclampsia with an area under the curve of 0.77(95% CI 0.61-0.92), which can be

classified as fair. The effect of all parameters was statistically significant and the area under

the curve was between 0.71 and 0.77. Differences between the parameters were only

small; in figure 4 and 5 ROC curves for the best parameter of each method are shown.

Preeclampsia

1 - Specificity

0,0 0,2 0,4 0,6 0,8 1,0

Sens

itivi

ty

0,0

0,2

0,4

0,6

0,8

1,0

SBPMAPHBI

MAPMAX

Gestational hypertension

1 - Specificity0,0 0,2 0,4 0,6 0,8 1,0

Sens

itivi

ty

0,0

0,2

0,4

0,6

0,8

1,0

HBIABPMMAP

MAXDBP

Figure 4. Receiver operator characteristic curve for the parameter with the highest area under the curve for predicton of preeclampsia for each method. ABPM, ambulatory blood pressure monitoring; HBI, hyperbaric index; MAP, mean arterial pressure; SBP, systolic blood pressure.

Figure 5. Receiver operator characteristic curve for the parameter with the highest area under the curve for predicton of preeclampsia for each method. ABPM, ambulatory blood pressure monitoring; HBI, hyperbaric index; MAP, mean arterial pressure; SBP, systolic blood pressure.

Chapter 3

63


The optimum cut-off value for HBI was a HBImax of 17.6 mmHg x hour resulting in

a sensitivity of 73% (95% CI 48-89%) and a specificity of 86% (95% CI 77-92). The

optimum cut-off value for sphygmomanometry was a SBP of 118 mmHg corresponding

with a sensitivity of 60% (95% CI 36-81%) and a specificity of 92% (95% CI 84-96%).

Discussion

We could not reproduce the high predictive value of HBImax as observed by Hermida

and co-workers, although we used a similar method for calculation of the reference and

validation values.15-18

We calculated the time-specified tolerance intervals with 90% confidence limits in our

own population using measurements of 90 healthy nulliparous women with an uneventful

pregnancy according to the method as described by Hermida et al.17;18 Group size for

the reference population was similar to that used by Hermida et al. in their first trimester

reference graph (n=78).26 Comparison of our reference graphs with those used by

Hermida et al. demonstrates a slightly higher SBP and DBP in our reference data, while

MAP was comparable. Differences could origin from an earlier gestational age in our

study or from ethnic differences (Mediterranean vs. Northern Europe). It was therefore

more appropriate to use reference values that are specific for our population and not

those published by Hermida et al.26

The HBI cut-off used by Hermida et al.was comparable with the cut-off calculated using

ROC curve analysis (15 vs. 18 mmHg x hour) and using either value offered comparable

predictive characteristics (sensitivity 73 vs. 73 %, specificity 85 vs. 86%). The difference in

study outcome can, therefore, not be explained by differences in calculation of reference

values.

Blood pressure was measured with the same blood pressure device as used by Hermida

et al.16 The Spacelab 90207 is the only device for ABPM that was validated extensively in

pregnant women. In three studies the device passed the Association for the Advancement of

Medical Instrumentation (AAMI) criteria. However, following British Hypertension Society

(BHS) criteria, these studies graded the device as B/B, A/C and B/C respectively.27-29

We decided to use the Spacelab, notwithstanding these conflicting results, as it was used

by the group whose method we wanted to validate and as no other ABPM device was

described with better characteristics for pregnant women at the beginning of our study.

In our study, all nulliparous women and all women with high obstetric risk, who visited

our outpatient clinic for antenatal care early enough for participation, were invited for the

study. However, the method for recruitment in the study of Hermida, the percentage of

64

nulliparous or high-risk women and precise gestational age when women were measured

in the first trimester was not described.

Participants in the study by Hermida et al. were examined at 4-week intervals and part

of the participants contributed twice to the reference graph or the validation study.15;26

We only measured women once between 8 and 11 weeks of pregnancy. Because blood

pressure lowers in the first half of pregnancy a close range of time of the measurements

should only improve predictive characteristics. Although population differences or

selection bias could explain part of the differences in outcome between the study by

Hermida et al 15;26and our study, we have insufficient data to support this supposition.

One other study investigated chronobiological analysis in the first trimester of pregnancy

for prediction of gestational hypertension or preeclampsia.19 This study included 104

women with moderate to high risk for hypertensive complications of pregnancy. Most

women had a history of preeclampsia and also twin pregnancies were allowed. Participants

were examined between 8 and 16 weeks in a hospital-based research setting. The best

predictive parameter in the 8-16 week period was systolic HBI at a cut-off of 5 mmHg/

hour with a sensitivity of 80% and a specificity of 77%. The method for calculating HBI

differed from the method used by Hermida et al.17 and in our study.

Blood pressure is widely used for screening in pregnancy for hypertensive disorders but

the results, as a predictive tool early in pregnancy, are conflicting. A recent systematic

review of literature showed poor predictive accuracy for preeclampsia by measurement

of SBP, DBP, MAP, and increase of blood pressure with the best results for a MAP of at

least 90 mmHg, LR+ of 3.5 (95% CI 2.0-5.0) and LR- of 0.46 (95% CI 0.16-0.75).4

In six of the 34 included studies blood pressure measurement was performed in the first

trimester. The test characteristics of these studies were comparable to the remaining

studies. The two studies targeted at a high risk population had similar characteristics as

the remaining studies. Differences between nulliparous and multiparous women were not

assessed. These findings supported that the combination of high risk women and low risk

nulliparous women in our study is valid. The outcome of our study is comparable with

findings in this meta-analysis. Although a more sophisticated measurement technique as

proposed by Hermida et a.l15 results in slightly better characteristics than obtained in most

studies, the extremely high sensitivity and specificity as reported should be considered as

an overestimation.

A 48-h ambulatory blood pressure measurement is generally considered as a burden.

In our study, 13 % of the registrations were inadequate for analysis, sometimes because

of device failure but mostly because of missing measurements for a couple of hours.

These women discontinued wearing the device because they perceived this as too

much of a strain. Our number of dropouts is comparable to a study which investigated

Chapter 3

65


patient satisfaction with the Spacelab 90207 in pregnant women.30 Sleep disturbance

was the main reason for noncompliance in this study, which was similar in our study.

Sphygmomanometry measurements are easy to perform and take only a couple of

minutes. The predictive efficacy was comparable with the hyperbaric index in our study. It

should be noted that auscultatory blood pressure measurement was performed according

to a standardized procedure, after sufficient rest in supine position and by averaging

three consecutive measurements. This differs from standard outpatient clinic procedure,

explaining better predictive characteristics than observed in most studies.4 DBP is slightly

lower in supine position than in sitting position, while SBP is similar.31 As this difference

is independent of blood pressure level or body mass index, and as we used ROC curve

analysis for determining the optimum cut-off in our population, this difference in position

could not have influenced our results

CONCLUSION

The accuracy of first trimester prediction of preeclampsia or gestational hypertension by

blood pressure measurement alone is insufficient for clinical use. Although blood pressure

is higher early in pregnancy in women who will develop preeclampsia standardized

sphygmomanometry, mean ABPM and HBI calculation from 48-h ABPM had a

comparable, restricted predictive efficacy. We could not confirm the high predictability of

the hyperbaric index using 48-h ABPM for preeclampsia or gestational hypertension as

observed by Hermida et al. 15

66


death: a systematic review. Lancet 2006; 367(9516):1066-1074. (2) Sibai B, Dekker G, Kupferminc M. Pre-eclampsia. Lancet 2005; 365(9461):785-799. (3) Conde-Agudelo A, Villar J, Lindheimer M. World Health Organization systematic review of

screening tests for preeclampsia. Obstet Gynecol 2004; 104(6):1367-1391. (4) Cnossen JS, Vollebregt KC, de VN, Ter RG, Mol BW, Franx A et al. Accuracy of mean arterial

pressure and blood pressure measurements in predicting pre-eclampsia: systematic review and meta-analysis. BMJ 2008; 336(7653):1117-1120.


(6) Brown MA, Bowyer L, McHugh L, Davis GK, Mangos GJ, Jones M. Twenty-four-hour automated blood pressure monitoring as a predictor of preeclampsia. Am J Obstet Gynecol 2001; 185(3):618-622.


(8) Mancia G, De BG, Dominiczak A, Cifkova R, Fagard R, Germano G et al. 2007 Guidelines for the Management of Arterial Hypertension: The Task Force for the Management of Arterial Hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). J Hypertens 2007; 25(6):1105-1187.


(10) Tranquilli AL, Giannubilo SR, Dell’Uomo B, Corradetti A. Prediction of gestational hypertension or intrauterine fetal growth restriction by mid-trimester 24-h ambulatory blood pressure monitoring. Int J Gynaecol Obstet 2004; 85(2):126-131.

(11) Perini R, Fisogni C, Bonera R, Rovetto B, Frusca T, Lojacono A et al. Role of Doppler velocimetry of uterine arteries and ambulatory blood pressure monitoring in detecting pregnancies at risk for preeclampsia. Minerva Ginecol 2004; 56(2):117-123.

(12) Benedetto C, Zonca M, Marozio L, Dolci C, Carandente F, Massobrio M. Blood pressure patterns in normal pregnancy and in pregnancy-induced hypertension, preeclampsia, and chronic hypertension. Obstet Gynecol 1996; 88(4 Pt 1):503-510.

(13) Steyn DW, Odendaal HJ, Hall DR. Diurnal blood pressure variation in the evaluation of early onset severe pre-eclampsia. Eur J Obstet Gynecol Reprod Biol 2008; 138(2):141-146.

(14) Halberg F, Ahlgren A, Haus E. Circadian systolic and diastolic hyperbaric indices of high school and college students. Chronobiologia 1984; 11(3):299-309.

(15) Hermida RC, Ayala DE, Mojon A, Fernandez JR, Silva I, Ucieda R et al. Blood pressure excess for the early identification of gestational hypertension and preeclampsia. Hypertension 1998; 31(1):83-89.

(16) Hermida RC, Ayala DE, Fernandez JR, Mojon A, Iglesias M. Prospective evaluation of the hyperbaric-tolerance test for the diagnosis of gestational hypertension and preeclampsia. Med Clin (Barc ) 2004; 123(5):161-168.

(17) Hermida RC, Fernandez JR. Computation of time-specified tolerance intervals for ambulatorily monitored blood pressure. Biomed Instrum Technol 1996; 30(3):257-266.

(18) Hermida RC, Mojon A, Fernandez JR, Ayala DE. Computer-based medical system for the computation of blood pressure excess in the diagnosis of hypertension. Biomed Instrum Technol 1996; 30(3):267-283.

Chapter 3

67


(19) Benedetto C, Marozio L, Giarola M, Chiarolini L, Maula V, Massobrio M. Twenty-four hour blood pressure monitoring in early pregnancy: is it predictive of pregnancy-induced hypertension and preeclampsia? Acta Obstet Gynecol Scand 1998; 77(1):14-21.


(21) Hermida RC, Ayala DE. Sampling requirements for ambulatory blood pressure monitoring in the diagnosis of hypertension in pregnancy. Hypertension 2003; 42(4):619-624.

(22) Verdecchia P, Angeli F, Sardone M, Borgioni C, Garofoli M, Reboldi G. Is the definition of daytime and nighttime blood pressure prognostically relevant? Blood Press Monit 2008; 13(3):153-155.

(23) Staessen JA, Thijs L, Fagard R, O’Brien ET, Clement D, De Leeuw PW et al. Predicting cardiovascular risk using conventional vs ambulatory blood pressure in older patients with systolic hypertension. Systolic Hypertension in Europe Trial Investigators. JAMA 1999; 282(6):539-546.

(24) Gardosi J, Chang A, Kalyan B, Sahota D, Symonds EM. Customised antenatal growth charts. Lancet 1992; 339(8788):283-287.

(25) Youden WJ. Index for rating diagnostic tests. Cancer 1950; 3(1):32-35. (26) Hermida RC, Ayala DE, Mojon A, Fernandez JR, Silva I, Ucieda R et al. High sensitivity test

for the early diagnosis of gestational hypertension and preeclampsia. IV. Early detection of gestational hypertension and preeclampsia by the computation of a hyperbaric index. J Perinat Med 1997; 25(3):254-273.

(27) Shennan AH, Kissane J, de Swiet M. Validation of the SpaceLabs 90207 ambulatory blood pressure monitor for use in pregnancy. Br J Obstet Gynaecol 1993; 100(10):904-908.

(28) O’Brien E, Darling M, Higgins J. Ambulatory blood pressure measurement in pregnancy. Br J Obstet Gynaecol 1994; 101(5):462-463.

(29) Franx A, van der Post JAM, van Montfrans GA, Bruinse HW. Comparison of an Auscultatory Versus an Oscillometric Ambulatory Blood Pressure Monitor in Normotensive, Hypertensive, and Preeclamptic Pregnancy.

(30) Walker SP, Permezel MJ, Brennecke SP, Tuttle LK, Higgins JR. Patient satisfaction with the SpaceLabs 90207 ambulatory blood pressure monitor in pregnancy. Hypertens Pregnancy 2004; 23(3):295-301.

(31) Netea RT, Smits P, Lenders JW, Thien T. Does it matter whether blood pressure measurements are taken with subjects sitting or supine? J Hypertens 1998; 16(3):263-268.

14Karlijn C. VollebregtKees BoerJoris A.M. van der PostHans Wolf

Submitted

The association of three different techniques to measure blood pressure in the first trimester of pregnancy with later preeclampsia

70

ABSTRACT

Blood pressure (BP) in the first trimester has limited efficacy for the prediction of

preeclampsia. It is not known if automated devices perform better in this respect than

conventional sphygmomanometry. This study compares two different automated devices

(continuous finger arterial pressure waveform registration and ambulatory blood pressure

monitoring (ABPM)) with conventional sphygmomanometry for their association with later

preeclampsia and gestational hypertension. In a cohort of 289 healthy normotensive

women (235 nulliparous and 44 parous women with previous preeclampsia) at 8 0/7-11 0/7 weeks of pregnancy BP was significantly higher in women, who developed

preeclampsia with all three methods. After adjustment for previous preeclampsia the point

estimate of the Odds ratios for association with later preeclampsia of both automated

devices was better than conventional sphygmomanometry although differences were

not statistically significant. Odds ratio (95% confidence intervals) for every 1 mmHg

pressure increase of MAP was 1.08 (1.02-1.15) for sphygmomanometry, 1.17 (1.09-

1.27) for finger arterial pressure waveform registration and 1.17 (1.07-1.27) for ABPM.

Since patient compliance with ABPM was less, we concluded that finger arterial pressure

waveform registration is potentially the most useful device in first pregnancy trimester

screening and needs further evaluation.

Chapter 4

71

Blood pressure measurem

ents and later preeclampsia

INTRODUCTION

Women who develop preeclampsia have a higher blood pressure in the first trimester

of pregnancy compared to women whose pregnancy remains uncomplicated. 1;2 This

association holds when only women with a blood pressure in the normal range are

considered.3;4 Selection in early pregnancy of women at high risk for preeclampsia

could facilitate preventive interventions and could enable targeting intensive care at

those women who need this most. The predictive capacity of blood pressure has been

meta-analysed and 34 studies including 60599 women with 3341 cases of preeclampsia

showed a moderate predictive accuracy.5 Poor reporting of study populations, definitions

of preeclampsia and the use of different methods of blood pressure assessment and

devices were identified as sources of variation in this meta-analysis. No separate analysis

was performed for studies that used automated devices.

In pregnancy blood pressure is generally measured by conventional sphygmomanometry

and therapeutic interventions are based on these measurements. This technique is

questioned because blood pressure is measured only during a single heart beat cycle, is

subject to inter-observer variation and does not account for beat-to-beat variation and

diurnal rhythm in blood pressure. Automated devices with continuous or repeated diurnal

measurement might have a higher precision for the prediction of later preeclampsia.6

In the present study we compared blood pressure measurement at 8 0/7-11 0/7 weeks

of pregnancy using continuous finger arterial waveform registration and ambulatory

blood pressure monitoring with conventional sphygmomanometry. Our hypothesis

was that blood pressure measurement by automated devices in early pregnancy has

a higher association with the development of later preeclampsia than conventional

sphygmomanometry.

SUBJECTS AND METHODS

Study populationThis study is part of a prospective study which investigates cardiovascular parameters in the

beginning of pregnancy for the prediction of preeclampsia.7 Healthy nulliparous women

and healthy parous women with a previous pregnancy complicated by preeclampsia

were included between April 2004 and June 2006. Women with multiple pregnancy or a

pregnancy after hormonal fertility treatment, and women with diabetes, chronic hypertension,

antihypertensive medication or renal disease were not eligible. All women had normal blood

pressure (less than 140 / 90 mmHg, measured by conventional sphygmomanometry) at

72

their first visit. Preventive use of acetylsalicylic acid in women with a previous preeclampsia

was allowed. Women with an infant with major congenital malformations were excluded

from analysis. Women with pregnancy associated complications like gestational diabetes,

preterm birth or fetal growth restriction were not excluded.

Participants were recruited by advertisements in local newspapers, flyers at midwife

practices near our hospital and from our outpatient clinic. The Medical Ethical Committee

of the Academical Medical Center approved the study.

Outcome definitions

Pregnancy outcome was defined as preeclampsia or combined pregnancy outcome

(preeclampsia and gestational hypertension). Two obstetricians (KB and HW) being

unaware of the results of the measurements classified pregnancy outcome according to

the ISSHP guidelines.8 Gestational hypertension was defined as a systolic blood pressure

≥ 140 and / or a diastolic blood pressure ≥ 90 mmHg after 20 weeks in a previously

normotensive woman, measured twice with an interval of at least 6 hours. Preeclampsia

was defined by the combination of gestational hypertension and proteinuria ≥ 0,3 g/ 24

h or dipstick ≥ ++ after 20 weeks gestation.

Study design

After written informed consent, all participants underwent identical study protocols

at 8 0/7-11 0/7 weeks gestational age. Gestational age was defined by first trimester

ultrasonography. Women were asked about medical history, maternal age, ethnic origin

(Caucasian or other) and education (high: university or higher; median: higher or

intermediate vocational training or low: no education, primary school or lower vocational

training). The maternal weight and length were measured and the Body Mass Index (BMI)

was calculated as weight (kg)/ (length (m)) 2. Study measurements were not available for

the clinicians or midwives responsible for prenatal care. All participants received usual

obstetric care, nulliparous women by a midwife and women with previous preeclampsia

by our outpatient clinic according to our local protocol.

Measurements

Blood pressure was measured by finger arterial pressure waveform registration (Finometer,

FMS, Amsterdam, the Netherlands) in supine position after at least 15 minutes of rest

in a quiet room with an ambient temperature between 20-22° Celsius, between 7

and 12 AM. The device was validated in pregnancy and passed the Association for

the Advancement of Medical Instrumentation (AAMI) criteria. According to the British

Hypertension Society (BHS) criteria, the device was graded a B for diastolic and a C for

systolic blood pressure.9 A cuff of appropriate size was placed around the middle finger

Chapter 4

73



of the non-dominant hand with the arm along the body. Hydrostatic height correction was

used to correct for the position of the hand with respect to heart level. At a stable signal

the pressure signal was corrected by the return-to-flow method.10 Data collection was

performed over a period of 10 minutes after the signal was stable for at least 5 minutes.

The brachial arterial pressure was reconstructed from the finger pulse wave.11Offline

beat to beat systolic blood pressure (SBP), diastolic blood pressure (DBP), mean arterial

pressure (MAP) and heart rate were extracted by Beatscope software (TNO Biomedical

Instrumentation, Amsterdam, the Netherlands) and averaged over 10 minutes.

Subsequently blood pressure was measured using conventional sphygmomanometry

(Maxi Stabil 3, an aneroid sphygmomanometer, Welch Allyn, Skaneateles Falls, New

York, USA), in the same position at both arms; Korotkoff V was used to determine DBP. If

there were little differences, further measurements were done at the non-dominant arm

else the woman was excluded. The average of three measurements with an interval of at

least 1 minute was used for calculations. The MAP was calculated as ((2x DBP) + SBP)/3.

Thereafter participants were instructed to measure blood pressure with an ambulatory

blood pressure monitor (ABPM; Spacelab 90207TM, Spacelabs Inc. Redmond,

Washington, USA) using oscillometry. The Spacelab 90207 is the only ABPM that was

validated extensively in pregnant women. In three studies the device passed the Association

for the Advancement of Medical Instrumentation (AAMI) criteria. Following British

Hypertension Society (BHS) criteria, these studies graded the device for systolic / diastolic

blood pressure as B/B, A/C and B/C respectively.12-14 An appropriately sized cuff was

placed on the non-dominant arm. Blood pressure was measured every 30 minutes

between 6:00 AM and 00:00 and every 60 minutes between 00:00 and 6:00 AM. All

women started the measurements between 7:00 AM and 12:00 AM and continued for

48 consecutive hours.15 The participants were instructed to maintain their daily routine.

Statistical analysis

Continuous variables were analyzed with the Kolmogorov-Smirnov test for normality.

Normally distributed data are presented as mean ± standard deviation (SD), not

normally distributed data as median (range). To investigate the differences between

the groups, one-Way ANOVA tests with post-hoc Bonferroni tests, Kruskal-Wallis test or

chi-square test were used whichever was appropriate. Next logistic regression analysis

with adjustment for previous preeclampsia was performed to determine the association

of each blood pressure variable with pregnancy outcome. Discriminative capacity was

assessed with Receiver Operating Characteristics (ROC) analysis. A p value < 0.05 was

considered significant. The statistical software package SPSS 16 (SPSS Inc, Chicago, IL)

was used for all analyses.

74

RESULTS

Study populationA total of 295 women were included, 251 nulliparous women and 44 multiparous women

with a history of preeclampsia. Five women with previous preeclampsia used prophylactic

low-dose acetylsalicylic acid. Five nulliparous women were excluded because of severe

foetal congenital malformations and one woman was lost to follow up. 289 Women were

available for analysis, 20 (6.9 %) developed preeclampsia and 16 (5.5 %) gestational

hypertension. None of them had significant differences in blood pressure by conventional

sphygmomanometry between both arms. Maternal characteristics and pregnancy

outcome are described in table 1.

Women with preeclampsia or gestational hypertension had a higher body mass index

(BMI) compared to women with uncomplicated pregnancies. Women with preeclampsia

delivered earlier and their babies had a lower birth weight compared to women with

unaffected pregnancies. Women with gestational hypertension were comparable to

those with unaffected pregnancy regarding gestational age at delivery and infant birth

weight. Women with preeclampsia had a lower education compared to those without

preeclampsia.

Table 1. Maternal characteristics and pregnancy outcomecharacteristics unaffected

n=253Preeclampsia

n=20Gestational

Hypertension n=16

P-value

Gestational age at measurement(days) 65 ± 5 64 ± 7 64 ± 3 0.78

Age (y) 30.0 ± 4.2 29.6 ± 4.5 31.6 ± 4.1 0.31

BMI (kg/m2) 23.8(17.6-41.8) 26.5(19.6-32.3)* 26.6(21.7-33.4)* <0.001

Ethnicity n (%)

Caucasian 218 (86.2) 15 (75.0) 14(87.5) 0.38

Other 35 (13.8) 5 (25.0) 2(12.5)

Smoking n (%) 27 (10.8) 3 (15.0) 1 (6.2) 0.70

Education n(%)

High 81 (32.0) 2 (10.0) 5 (31.2) 0.009

Median 94 (37.2) 4 (20.0) 7 (43.8)

low 78 (30.8) 14 (70.0) 4 (25.0)

High risk n (%) 30 (11.9) 9 (45) 5 (31.2) <0.001

Gestational age at delivery(days) 279 (207-296) 256(200-286)* # 276(245-294) <0.001

Birth weight(grams) 3338 ± 562 2106 ± 890 * # 3303 ± 639 <0.001

Small for gestational age n (%) 27 (10.7) 13 (65.0) 4 (25.0) <0.001

Data are expressed as mean ± SD, median (range) or n (%). P-value; differences between the groups. *P<0.05 compared to unaffected; #P<0.05 compared to gestational hypertension.

Chapter 4

75



Blood pressure measurements

Finger arterial pressure waveform registrations of 4 (1.4%) women were not available

for analysis because of technical difficulties. ABPM data for 48 hour were available in

220 women. Drop out was partly due to lack of a monitoring device at the scheduled

registration time (62 women). A total of 39 women (17.7%) had no or incomplete

data. Eight women refused to wear the blood pressure monitor and 31 women had

an incomplete registration. Three of these had less then 10 measurements and were

excluded. All other incomplete registrations were used.

Women who developed preeclampsia or gestational hypertension had a significantly

higher blood pressure in the first trimester than women with unaffected pregnancies.

The absolute blood pressure values were different between the devices, p<0.01 for all

measurements. The highest absolute values were measured by finger arterial pressure

waveform registration and the lowest by conventional sphygmomanometry (Table 2).

The point estimate of the Odds ratios of blood pressure in the first trimester for later

preeclampsia or combined pregnancy outcome was higher for the automated blood

pressure measurement techniques, although there was an overlap between 95% CI

intervals. For all devices MAP had the highest odds ratios. Figures 1A and 1B show the

odds ratios with 95% CI intervals for every mmHg pressure increase adjusted for previous

Table 2. First trimester blood pressure measurements in different outcome groups.Blood pressure measurements unaffected Preeclampsia Gestational

Hypertension P-value

Heart rate (beats/min) 73 ± 7 73 ± 7 72 ± 8 0.95

Sphygmomanometry

n 253 20 16

SBP (mmHg) 104 ± 9 113 ± 13** 110 ± 8 * <0.001

DBP (mmHg) 60 (46-88) 67 (50-85)** 67 (58-85)** 0.002

MAP (mmHg) 76 ± 7 82 ± 11** 81 ± 7** <0.001

Finger arterial pressure

n 250 19 16

SBP (mmHg) 121 ± 8 131 ± 13** 125 ± 9 <0.001

DBP (mmHg) 67 ± 6 74 ± 8** 71 ± 7* <0.001

MAP (mmHg) 88 ± 7 98 ± 10** 93± 7** <0.001

ABPM

n 179 19 14

SBP (mmHg) 111 ± 7 120 ± 13** 114 ± 8 <0.001

DBP (mmHg) 66 ± 4 72 ± 9** 69 ± 6* <0.001MAP (mmHg) 80 ± 5 88 ± 10** 84 ± 7* <0.001

Data are expressed as mean ± SD or median (range); P-value; differences between the groups.* P<0.05 compared to unaffected; **P<0.01 compared to unaffected

76

Sphygmo SBP

Sphygmo MAP

Sphygmo DBP

FAP SBP

FAP MAP

FAP DBP

ABPM SBP

ABPM MAP

ABPM DBP

2,01,00,0

1.08 (1.03-1.13)

1.08 (1.02-1.15)

1.06 (1.002-1.13)

1.13 (1.06-1.19)

1.17 (1.09-1.26)

1.15 (1.06-1.23)

1.12 (1.05-1.20)

1.17 (1.07-1.27)

1.17 (1.07-1.27)

Sphygmo SBP

Sphygmo MAP

Sphygmo DBP

FAP SBP

FAP MAP

FAP DBP

ABPM SBP

ABPM MAP

ABPM DBP

2,01,00,0

1.08 (1.04-1.12)

1.10 (1.04-1.15)

1.08 (1.028-1.14)

1.10 (1.05-1.15)

1.15 (1.09-1.21)

1.13 (1.07-1.20)

1.10 (1.04-1.16)

1.16 (1.08-1.25)

1.16 (1.07-1.25)

Figure 1A. Odds ratios (95% confidence intervals) for every 1 mmHg pressure increase, after adjustment for previous preeclampsia, to determine the relationship of SBP, MAP and DBP measured by the different devices and preeclampsia.

Sphygmo; sphygmomanometry, FAP; finger arterial pressure waveform registration. ABPM; ambulatory blood pressure monitoring.

Figure 1B. Odds ratios (95% confidence intervals) for every 1 mmHg pressure increase after adjustment for previous preeclampsia to determine the relationship of SBP, MAP and DBP measured by the different devices and combined pregnancy outcome.

Sphygmo; sphygmomanometry, FAP; finger arterial pressure waveform registration, ABPM; ambulatory blood pressure monitoring.

Table 3. Area under the receiver operating characteristics curves with 95% CI of blood pressure measurements for the subsequent development of preeclampsia or combined pregnancy outcome.Measurement preeclampsia Combined pregnancy outcome

SphygmomanometrySBP 0.73 (95% CI 0.58-0.83) 0.73 (95% CI 0.64-0.83)

DBP 0.68 (95% CI 0.53-0.84) 0.72 (95% CI 0.61-0.82)

MAP 0.70 (95% CI 0.55-0.85) 0.73 (95% CI 0.63-0.83)

Finger arterial pressure

SBP 0.82 (95% CI 0.68-0.95) 0.77 (95% CI 0.68-0.87)

DBP 0.77 (95% CI 0.63-0.91) 0.77 (95% CI 0.68-0.86)

MAP 0.82 (95% CI 0.69-0.96) 0.80 (95% CI 0.71-0.89)

ABPM

SBP 0.78 (95% CI 0.65-0.91) 0.74 (95% CI 0.63-0.84)

DBP 0.77 (95% CI 0.63-0.90) 0.72(95% CI 0.62-0.83)MAP 0.78 (95% CI 0.65-0.91) 0.74 (95% CI 0.64-0.85)

Chapter 4

77



preeclampsia for the three measurement techniques. The largest area under the ROC

curves (AUC) was 0.82 (95% CI 0.69-0.96) for preeclampsia and 0.80 (95% CI 0.71-

0.89) for combined outcome using MAP estimated by finger arterial pressure waveform

registration (Figure 2A and 2B). Table 3 provides the values of all AUCs, with 95%

confidence intervals. Differences were small and did not reach statistical significance.

DISCUSSION

In this prospective cohort of women with normal blood pressure we confirm that blood

pressure has a significant association with hypertensive disorders later in pregnancy.

Assessment by automated devices at 8 0/7-11 0/7 weeks of pregnancy seems superior

Figure 2A. Association of mean arterial pressure measured with the different devices in the first trimester and prevois preeclampsia with subsequent development of preeclampsia, expressed by Receiver Operating Characteristics (ROC) curves.

Figure 2B. Association of mean arterial pressure measured with the different devices in the first trimester and previous preeclampsia with subsequent development of combined pregnancy outcome, expressed by Receiver Operating Characteristics (ROC) curves.

1 - Specificity1,00,80,60,40,20,0

1,0

0,8

0,6

0,4

0,2

0,0

Sens

itivi

ty

1 - Specificity1,00,80,60,40,20,0

Sens

itivi

ty1,0

0,8

0,6

0,4

0,2

0,0

MAP ABPM

MAP finger arterial waveform

MAP sphygmomanometry

MAP ABPM

MAP finger arterial waveform

MAP sphygmomanometry

78

to conventional sphygmomanometry. ABPM measurements failed more often than

finger arterial pressure waveform registration due to technical difficulties or patient

incompliance.

The accuracy of automated blood pressure devices in pregnant women is often

questioned especially in hypertensive and preeclamptic women and it is generally advised

to use conventional sphygmomanometry to establish the diagnosis of preeclampsia.8

Both Spacelab 90207™ and Finometer™ passed the less strict AAMI criteria, but none

received A/A BHS grading in validation studies in pregnant women where the devices

were compared to auscultatory conventional sphygmomanometry.9;12-14

However, it was not our intention to compare absolute blood pressure values of the

automated devices with conventional sphygmomanometry, which is the usual validation

procedure. Clinical relevance of a measurement method is far more important than

absolute values and therefore we chose a clinical endpoint (preeclampsia or combined

gestational hypertension or preeclampsia). Repeated or continuous measurements provide

a more reliable estimate of average blood pressure than a single sphygmomanometry

measurement. This could explain why finger arterial pressure waveform registration and

ABPM measurements tend to have a higher association with later preeclampsia than

sphygmomanometry in our study.

One of the limitations of our study is the relatively low number of included women. This

causes some uncertainty regarding the exact quantity of the association, as reflected by

the large confidence intervals of odds ratios and possibly causes an overestimation of the

effect.16 The point estimate of the Odds ratios for association with later preeclampsia of

both the automated devices was better for all blood pressure parameters but there was

an overlap between 95% CI intervals and significant differences could not be detected.

Women with elevated blood pressure, cardiovascular disease, renal disease or diabetes

were not included because the elevated risk of these conditions for preeclampsia or

pregnancy induced hypertension preclude the usefulness of further risk assessment.17

Recently we published a meta-analysis of the accuracy of blood pressure measurements

in predicting preeclampsia.5 Studies included in this meta-analysis used all types of blood

pressure measurement devices and it was not possible to compare devices in this meta-

analysis Areas under the ROC-curve for SBP, DBP and MAP for low risk women in the first

trimester were respectively 0.66, 0.67 and 0.79 comparable to our results.

Several studies investigated the predictive capacity of ABPM and preeclampsia and

compared this with office blood pressure measurements.18-23 With ABPM information

on diurnal blood pressure can be obtained. Some found comparable predictive capacity

while others showed better results with ABPM in second trimester pregnancy. There

Chapter 4

79



were no differences in predictive capacity of only daytime or night time blood pressure

compared to mean values.22;24

As the origin of the disease is in early pregnancy it is likely that preventive interventions

should start early in pregnancy to be effective.25 Therefore, early prediction is of major

importance. Only one of the studies on the predictive capacity of ABPM was performed

in the first trimester. This study showed an extremely high predictive capacity with 48 hour

ABPM using the hyperbaric index.23 However, using an identical approach as Hermida

et al. we could not confirm this and observed a similar association with preeclampsia of

hyperbaric index, mean pressure of 48 hour ABPM and day/night differences in the first

trimester.7 The study population used in this analysis was included in the present study.

The three devices used in our study measure blood pressure in different ways.

Sphygmomanometry and the Spacelab 90207™ register vascular oscillations during the

gradual relaxation of vascular occlusion at the upper arm. In sphygmomanometry the

appearance and disappearance of vascular (Korotkoff) sounds determines SBP and DBP.

The Spacelab 90207™ is an oscillometric device and the MAP is estimated primarily by

the pressure where oscillations are largest and SBP and DBP are derived from this point.26

Conventional sphygmomanometry registers SBP from one and DBP from another heart

beat cycle. Both techniques are influenced by arm diameter and vascular deformability.

Finger arterial pressure waveform registration is registered by a small cuff around the

middle finger. The cuff contains a diode to measure blood flow in the finger and a server

mechanism connected to the cuff regulates pressure in such a way that flow in the finger

remains constant. Thus the cuff pressure is a copy of the arterial finger pressure wave. This

technique enables continuous non-invasive registration of cardiovascular parameters.27

The algorithm for blood pressure calculation depends on vascular compliance which is

larger in pregnant than in non-pregnant women.28 However, the return to flow method

enables calibration for blood pressure calculation and thus overcomes the disturbing

effect of pregnancy on the algorithm for blood pressure calculation.10

According to international guidelines for the classification of preeclampsia blood

pressure is preferably measured at the right arm while seated.8;29;30 In our study the cuff

for ABPM was placed around the non-dominant arm because we wanted to reduce the

inconvenience of ABPM. Therefore all other measurements were also performed at the

same arm, which usually is the left arm. We also chose supine position for measuring

blood pressure with sphygmomanometry and finger arterial waveform registration in

order to be certain that all subjects were at rest. DBP is slightly lower in supine position

than in sitting position, whereas SBP is similar.31 If blood pressure is included in a first

trimester prediction model for preeclampsia, position needs further evaluation. All

women received usual obstetric care prenatal care and blood pressure used for our

80

clinical outcome was measured by conventional sphygmomanometry according to the

international guidelines.

Both automated devices that are used in this study could help in designing prediction

models for preeclampsia. However, ambulatory blood pressure measurement in

pregnancy, as been shown earlier, is considered a burden.32;33 We could not obtain

a complete registration in 18% of the women because of patient incompliance and

technical difficulties. In contrast finger arterial pressure waveform registration only needs

20 minutes for a complete registration and thus could be used in clinical practice more

easily.

Conclusion and impact for further research

Both automated devices used in this study are considered to be of help in designing

prediction models for preeclampsia. Although our study was too small to find significant

differences between the devices and conventional sphygmomanometry, finger arterial

pressure waveform registration is potentially the most useful device in first trimester

screening in pregnancy and needs further evaluation.

Chapter 4

81



Reference List (1) Easterling TR, Benedetti TJ, Schmucker BC, Millard SP. Maternal hemodynamics in normal and

preeclamptic pregnancies: a longitudinal study. Obstet Gynecol 1990; 76(6):1061-1069. (2) Rang S, Wolf H, Montfrans GA, Karemaker JM. Serial assessment of cardiovascular control

shows early signs of developing pre-eclampsia. J Hypertens 2004; 22(2):369-376. (3) Moutquin JM, Rainville C, Giroux L, Raynauld P, Amyot G, Bilodeau R et al. A prospective

study of blood pressure in pregnancy: prediction of preeclampsia. Am J Obstet Gynecol 1985; 151(2):191-196.


(5) Cnossen JS, Vollebregt KC, de VN, Ter RG, Mol BW, Franx A et al. Accuracy of mean arterial pressure and blood pressure measurements in predicting pre-eclampsia: systematic review and meta-analysis. BMJ 2008; 336(7653):1117-1120.

(6) Ayala DE, Hermida RC, Mojon A, Fernandez JR, Silva I, Ucieda R et al. Blood pressure variability during gestation in healthy and complicated pregnancies. Hypertension 1997; 30(3 Pt 2):611-618.

(7) Vollebregt KC, Gisolf J, Guelen I, Boer K, van MG, Wolf H. Limited accuracy of the hyperbaric index, ambulatory blood pressure and sphygmomanometry measurements in predicting gestational hypertension and preeclampsia. J Hypertens 2010; 28(1):127-134.


(9) Hehenkamp WJ, Rang S, van Goudoever J, Bos WJ, Wolf H, van der Post JA. Comparison of Portapres with standard sphygmomanometry in pregnancy. Hypertens Pregnancy 2002; 21(1):65-76.

(10) Bos WJ, van GJ, van Montfrans GA, van den Meiracker AH, Wesseling KH. Reconstruction of brachial artery pressure from noninvasive finger pressure measurements. Circulation 1996; 94(8):1870-1875.

(11) Guelen I, Westerhof BE, Van Der Sar GL, van Montfrans GA, Kiemeneij F, Wesseling KH et al. Finometer, finger pressure measurements with the possibility to reconstruct brachial pressure. Blood Press Monit 2003; 8(1):27-30.

(12) Shennan AH, Kissane J, de Swiet M. Validation of the SpaceLabs 90207 ambulatory blood pressure monitor for use in pregnancy. Br J Obstet Gynaecol 1993; 100(10):904-908.

(13) O’Brien E, Darling M, Higgins J. Ambulatory blood pressure measurement in pregnancy. Br J Obstet Gynaecol 1994; 101(5):462-463.

(14) Franx A, van der Post JAM, van Montfrans GA, Bruinse HW. Comparison of an Auscultatory Versus an Oscillometric Ambulatory Blood Pressure Monitor in Normotensive, Hypertensive, and Preeclamptic Pregnancy. Hypertens Pregnancy 1997; 16(2):187-202.

(15) Hermida RC, Ayala DE. Sampling requirements for ambulatory blood pressure monitoring in the diagnosis of hypertension in pregnancy. Hypertension 2003; 42(4):619-624.

(16) Leeflang MM, Bossuyt PM, Irwig L. Diagnostic test accuracy may vary with prevalence: implications for evidence-based diagnosis. J Clin Epidemiol 2009; 62(1):5-12.

(17) Caritis S, Sibai B, Hauth J, Lindheimer M, VanDorsten P, Klebanoff M et al. Predictors of pre-eclampsia in women at high risk. National Institute of Child Health and Human Development Network of Maternal-Fetal Medicine Units. Am J Obstet Gynecol 1998; 179(4):946-951.

(18) Peek M, Shennan A, Halligan A, Lambert PC, Taylor DJ, de SM. Hypertension in pregnancy: which method of blood pressure measurement is most predictive of outcome? Obstet Gynecol 1996; 88(6):1030-1033.

82

(19) Halligan AW, Shennan A, Lambert PC, Bell SC, Taylor DJ, de SM. Automated blood pressure measurement as a predictor of proteinuric pre-eclampsia. Br J Obstet Gynaecol 1997; 104(5):559-562.

(20) Penny JA, Halligan AW, Shennan AH, Lambert PC, Jones DR, de SM et al. Automated, ambulatory, or conventional blood pressure measurement in pregnancy: which is the better predictor of severe hypertension? Am J Obstet Gynecol 1998; 178(3):521-526.

(21) Bellomo G, Narducci PL, Rondoni F, Pastorelli G, Stangoni G, Angeli G et al. Prognostic value of 24-hour blood pressure in pregnancy. JAMA 1999; 282(15):1447-1452.

(22) Brown MA, Bowyer L, McHugh L, Davis GK, Mangos GJ, Jones M. Twenty-four-hour automated blood pressure monitoring as a predictor of preeclampsia. Am J Obstet Gynecol 2001; 185(3):618-622.

(23) Hermida RC, Ayala DE, Fernandez JR, Mojon A, Iglesias M. Reproducibility of the tolerance-hyperbaric test for diagnosing hypertension in pregnancy. J Hypertens 2004; 22(3):565-572.


(25) Steegers EA, von DP, Duvekot JJ, Pijnenborg R. Pre-eclampsia. Lancet 2010; 376(9741):631-644.

(26) Ramsey M, III. Noninvasive automatic determination of mean arterial pressure. Med Biol Eng Comput 1979; 17(1):11-18.

(27) Wesseling KH, de Wit B, Settels JJ, Klawer WH. On the indirect registration of finger blood pressure after Penaz. Funkt Biol Med 1982; 1:245-250.

(28) Rang S, de Pablo LB, van Montfrans GA, Bouma BJ, Wesseling KH, Wolf H. Modelflow: a new method for noninvasive assessment of cardiac output in pregnant women. Am J Obstet Gynecol 2007; 196(3):235-238.

(29) Lowe SA, Brown MA, Dekker GA, Gatt S, McLintock CK, McMahon LP et al. Guidelines for the management of hypertensive disorders of pregnancy 2008. Aust N Z J Obstet Gynaecol 2009; 49(3):242-246.

(30) Lindheimer MD, Taler SJ, Cunningham FG. Hypertension in pregnancy. J Am Soc Hypertens 2008; 2(6):484-494.

(31) Netea RT, Smits P, Lenders JW, Thien T. Does it matter whether blood pressure measurements are taken with subjects sitting or supine? J Hypertens 1998; 16(3):263-268.


(33) Taylor RS, Gamble G, McCowan L, North RA. Sleep effects on ambulatory blood pressure measurements in pregnant women. Am J Hypertens 2001; 14(1):38-43.

15Karlijn C. VollebregtLarissa van leijdenKees BoerJoris A.M. van der Post Junichiro HashimotoHans WolfBerend E. Westerhof

Submitted

Arterial stiffness at 8 0/7-110/7 weeks of pregnancy and the association with later preeclampsia

86

ABSTRACT

Objective To study the association of the augmentation index (AIx), augmentation pressure (AP) at

8 0/7-11 0/7 weeks of pregnancy with the later onset of preeclampsia. And if combination

with MAP improves this association.

Study Design

235 nulliparous and 42 former preeclamptic women were included with an incidence

of preeclampsia of 4.3% and 16.6%. AIx, AP and MAP obtained from finger arterial

waveform registration were analysed using logistic regression.

Results

After adjustment for previous preeclampsia the AIx and AP were associated with later

preeclampsia (OR 1.84 (95% CI 1.10-3.0) and 1.90 (1.31-2.76) respectively). AP in

combination with MAP and previous preeclampsia had an area under the ROCcurve of

0.82 (0.66-0.97) compared to 0.74 (0.61-0.87), 0.78 (0.65-0.91) and 0.79 (0.64-

0.93) for the single parameters.

Conclusion

The combination of MAP and AP had a higher association with later preeclampsia

compared to single parameters, but the additive effect of AP was small.

Chapter 5

87

Arterial stiffness and later preeclampsia

INTRODUCTION

Preeclampsia affects 2-8% of all pregnancies and remains a leading cause of maternal

and perinatal morbidity and mortality.1;2 Women who had a pregnancy complicated

by preeclampsia are at increased risk for cardiovascular disease in later live and

preeclampsia and cardiovascular disease share similar risk factors.3;4

Increased arterial stiffness is an independent predictor of adverse outcome in

cardiovascular disease.5 Several studies showed an elevated arterial stiffness in women

with preeclampsia.6-9 One research group recently observed an increased arterial

stiffness, measured by applanation tonometry and calculation of the augmentation

index (AIx) at 11-13 weeks of pregnancy, in women who developed preeclampsia later

compared to women with normal pregnancy.10;11

It is not known if this difference in AIx only occurs in the beginning of pregnancy or if it is pre-

existing. Thus the increased risk of cardiovascular disease in former preeclamptic women

could result from persisting abnormalities in cardiovascular control after preeclampsia

or from pre-existing abnormalities which are associated with both conditions. At present,

literature is not helpful to answer this question. Studies report conflicting results about

arterial stiffness in former preeclamptic women compared to women with uncomplicated

pregnancies.7;8;12

Women predisposed to develop preeclampsia already have a higher blood pressure

early in pregnancy, long before clinical symptoms arise, compared to women with an

uncomplicated pregnancy. Although they have no hypertension yet their blood pressure

measured in the beginning of pregnancy could be used to predict preeclampsia.13;14

A systematic review concluded that mean arterial pressure was the blood pressure

parameter with the highest association with preeclampsia.15 Blood pressure and arterial

stiffness are not independent parameters and it is not well described if combined use of

these parameters improves the association with later preeclampsia.16

The primary aim of this study is to test the hypothesis that the association between early

pregnancy arterial stiffness, measured by augmentation index or augmentation pressure,

is improved if combined with mean arterial blood pressure. Our second hypothesis is that

these parameters differ between nulliparous women with an uncomplicated pregnancy

and women with a previous preeclampsia, which could indicate the presence of pre-

existing differences that might contribute to the elevated risk for cardiovascular disease in

women with previous preeclampsia.

88

METHODS

SubjectsThis study is part of a prospective study which investigated cardiovascular parameters in the

beginning of pregnancy for the prediction of preeclampsia.17 Healthy nulliparous women

and healthy parous women with a previous pregnancy complicated by preeclampsia

were included between April 2004 and June 2006. Women with multiple pregnancy,

and women with diabetes, chronic hypertension, antihypertensive medication or renal

disease were not eligible. Women with an infant with major congenital malformations

were excluded. All women had normal blood pressure (<140 / 90 mmHg, measured

by conventional sphygmomanometry) at their first visit. Preventive use of acetylsalicylic

acid in women with a previous preeclampsia was allowed. Participants were recruited

by advertisements in local newspapers, flyers at midwife practices near our hospital and

from our outpatient clinic. The Medical Ethical Committee of the Academical Medical

Center approved the study. After written informed consent, all participants underwent

identical study protocols.

Study protocol

At 8 0/7-11 0/7 weeks of pregnancy gestational age was confirmed by ultrasound.

All measurements were scheduled between 7 AM and 12 noon and took place in a

quiet room with a temperature between 20 and 22 ºC. Subjects were asked to refrain

from coffee from the night before the measurements. They were informed about the

procedures and instructed to empty their bladder before the start of the measurements.

The measurements were performed in supine position at the non-dominant arm after at

least 15 minutes of rest.

Blood pressure and hemodynamic parameters were measured continuously for 10

minutes using non-invasive finger arterial waveform recording by FinometerTM (FMS,

Amsterdam, the Netherlands).

Measurement of finger arterial pressure is based on the volume clamp method of

Peňáz and the ‘Physiocal’ criteria of Wesseling.18;19 20 The brachial arterial pressure is

reconstructed from the finger pulse wave.21 This device is validated in pregnancy and

passed the Association for the Advancement of Medical Instrumentation (AAMI) criteria.

According to the British Hypertension Society (BHS) criteria, the device was graded a B for

diastolic and a C for systolic blood pressure.22 In pregnancy stroke volume measurement

is comparable with Doppler echocardiography.23 An appropriate cuff was placed at the

middle finger of the non-dominant hand with the arm along the body. Hydrostatic height

correction was used to correct for the position of the hand with respect to heart level.

Chapter 5

89


When a stable signal was attained, the pressure signal was corrected by the return-to-

flow method.24 Data from the FinometerTM were sampled at 200 Hz and analyzed by

Beatscope software (TNO Biomedical Instrumentation, Amsterdam, the Netherlands).

The average of the systolic blood pressure (SBP), diastolic blood pressure (DBP), mean

arterial pressure (MAP) and heart rate (HR) of ten minutes were used for comparisons.

As a measure of arterial stiffness the augmentation index (AIx) was calculated. For the AIx,

a period of 60 seconds, with a stable signal, was selected from the Finometer recording.

The finger arterial waveform of this period was filtered to an aortic waveform by the

Beatscope software. In Mathematica (Wolfram Research, Inc., Mathematica, Version 4.0,

Champaign, IL) the filtered pressure was further analyzed to determine the augmentation

index. 20 Beats were selected and the averaged wave form of these 20 beats was used.

The inflection point was determined with the use of the fourth derivate of the aortic

pressure wave.25 If the inflection point was found before the first pressure peak, A-type

augmentation index was calculated. If the inflection point lay after the first pressure peak

B-type augmentation index was calculated. B-type augmentation index is expressed as

a negative index. AIx was calculated as augmenting pressure (AP, determined by the

inflection point and SBP) divided by pulse pressure (SBP-DBP).26 DAIx was calculated as

AIx postpartum-AIx first trimester.

Outcome

Gestational hypertension was defined as a systolic blood pressure ≥ 140 and/ or a

diastolic blood pressure ≥ 90 mmHg after 20 weeks in a previously normotensive

woman, measured twice with an interval of at least 6 hours. Preeclampsia was defined by

the combination of gestational hypertension and proteinuria ≥ 0,3 g/ 24 h or dipstick ≥

++ after 20 weeks gestation according to the ISSHP guidelines.27

Statistics

To assess normality of the distribution of the numerical data the Kolmogorov-Smirnov test

was used. Normally distributed data are presented as mean ± standard deviation (SD),

not normally distributed data as median (range).

(Paired) Student t-tests, Mann-Whitney U tests and chi-square tests were used when

appropriate.

Logistic regression analysis was used to determine the association of AIx, AP and MAP

in the beginning of pregnancy with later preeclampsia. The continuous variables of

AIx, AP and MAP were categorized per 10 %, 3 and 5 mmHg respectively. Because

the two inclusion groups had highly different a priori risks for preeclampsia, previous

preeclampsia (nulliparous versus parous women with previous preeclampsia) was entered

for adjustment in all models. The probability for entry was set at 0.05, for removal at

90

0.1. For comparison between the models the discriminative capacity was assessed with

Receiver Operating Characteristics (ROC) analysis. The statistical software package SPSS

18 (SPSS Inc, Chicago, IL) was used for all analyses. A p value < 0.05 was considered

statistically significant.

RESULTS

Baseline characteristics283 women participated in the study. After inclusion we excluded 3 women from analysis

because of severe lethal congenital malformations of their infants, 2 women who delivered

an infant with Down syndrome and one woman was lost to follow-up. Thus, 277 women

were available for analysis in the first trimester and 225 of them returned postpartum.

5 of the 42 parous women with previous preeclampsia used prophylactic acetylsalicylic

acid. The mean gestational age at measurement in the first trimester was 9 2/7±0.7

weeks and the mean interval post partum was 28 5/7±7 weeks. Seventeen (6.1 %)

women developed preeclampsia and 16 (5.8 %) other women gestational hypertension.

The women who developed gestational hypertension were allocated to the control group.

Maternal characteristics and pregnancy outcome are described in Table 1.

Women with preeclampsia delivered earlier and their babies had a lower birth weight

compared to women with unaffected pregnancies. Nulliparous women with preeclampsia

were younger compared to the nulliparous unaffected women. There were no differences

between parous women with or without preeclampsia in maternal characteristics.

Hemodynamics

All women who developed preeclampsia (nulliparous and parous women with previous

preeclampsia) had a higher blood pressure at 8-11 weeks and 29 weeks postpartum

Table 1. Baseline characteristics at first visit and pregnancy outcome of the study population. Nulliparous women(n=235) Parous women with previous

preeclampsia(n=42)

No preeclampsian=225

Preeclampsian=10 (4.4%)

No preeclampsian=35

Preeclampsian=7 (16.7%)

Age (y) 30.0 ± 4.2 26.0 ± 3.0‡ 30.7 ± 4.0 33.8 ± 2.2

Caucasian (%) 295 (86.7) 7 (70) 30 (85.7) 5 (71.4)

BMI (m2/kg) 23.3 (17.6-41.8) 24.3 (20.6-31.6) 23.9 (18.8-37.5) 28.5 (19.6-32.2)

Smoking (%) 25 (11.1) 1 (10.0) 3 (8.6) 1 (14.3)

Gestational age at delivery(d) 278 (207-297) 255 (209-286) ‡ 272 (243-289) 224 (200-273) ‡Birth weight (gr) 3475(1170-4585) 2031(870-3232)‡ 3165 (1620-4132) 1300(1000-3025)‡

Data are presented as mean (±standard deviation) or median (range) ‡P<0.01 vs no preeclampsia

Chapter 5

91


Alx>3525-3515-255-15<=5

100%

80%

60%

40%

20%

0%

MAP>9590-9585-9080-8575-80<=75

100%

80%

60%

40%

20%

0%

preeclampsiano preeclampsia

AP>1512-159-126-93- 6<=3

100%

80%

60%

40%

20%

0%

17 100 103 44 13

34 82 75 45 24 17

6 22 59 89 56 45

Figure 1. Percentage of preeclampsia for augmentation index (AIx) classes, augmentation pressure (AP) classes, and mean arterial pressure (MAP) classes with group size above each column.

92

compared to unaffected women, although the postpartum difference did not reach

statistical significance in parous women. Blood pressure in nulliparous and in parous

women with a previous preeclampsia, who did not develop preeclampsia, was similar at

8-11 weeks and 29 weeks postpartum (Table 2).

All women who developed preeclampsia had a higher AIx and AP in the first trimester

compared to women who did not. This difference was not statistically significant in parous

women (Table 2). First trimester values were lower than values 29 weeks postpartum

in women with uncomplicated pregnancy. In women who developed preeclampsia first

Table 2. Hemodynamic measurements at 8-11 weeks gestational age and at 29 weeks postpartum. Nulliparous women (n=235) Multiparous women with previous preeclampsia (n=42)

Outcome No preeclampsia Preeclampsia No preeclampsia Preeclampsia8-11 w 29 w PP 8-11 w 29 w PP 8-11 w 29 w PP 8-11 w 29 w PP

n=225 N=172 n=10 N=7 n=35 n=25 N=7 N=5

Heart rate (b/m) 70±8 66±9* 69±6 69±6 70±8 69±13 71±6 70±11

SBP (mmHg) 121±8 119±11* 129±15† 132±16V 119±8 122±13 129±5† 125±10

DBP (mmHg) 67±6 69±8* 71±9† 73±9 68±7 74±9* 76±5† 75±8

MAP (mmHg) 87±6 89±9* 96±12† 97±12† 89±8 95±8* 98±5† 97±8

AIx (%) 17 (-11-47) 24 (3-45)* 26 (9-36) † 25 (20-38) 17 (-3-39) 26 (11-48)* 23 (17-33) 27 (18-29)

AP (mmHg) 6.5 (0.9-22.5) 9.4 (1.3-22.2)* 12.3(2.7-19.2) † 11.1 (9.0-25.7) 7.1 (1.1-20.1) 9.5 (3.7-31.3)* 9.8 (6.5-16.1) 10.1(7.4-14.1)DAIx - 7.5±10.9 - 2.2±12.4 - 6.7±10.5 - 0.25±10.5

Data are presented as mean (± standard deviation) or median (range) *P<0.05 vs first trimester,†P<0.05 vs. no preeclampsia, PP: postpartum, SBP: systolic blood pressure, DBP : diastolic blood pressure, MAP: mean arterial pressure, AIx : augmentation index, AP : augmentation pressure

Table 3. Results of the logistic regression analyses to assess the association between different parameters and preeclampsia later in pregnancy, with adjustment for previous preeclampsia.Entry parameters Last step parameters OR (95% CI) Area under the ROCcurve (95% CI)

AIx and parity AIx 1.84 (1.10-3.00) 0.74 (0.61-0.87)*previous preeclampsia 4.02(1.41-11.50)

AP and parity AP 1.90 (1.31-2.76) 0.78 (0.65-0.91) *

previous preeclampsia 4.29 (1.47-12.58)

MAP and parity MAP 2.26 (1.36-3.78) 0.79 (0.64-0.93) *


AIx, MAP and parity AIx 1.63 (0.93-2.86) 0.80 (0.65-0.95) *

MAP 2.10 (1.25-3.51)


AP, MAP and parity AP 1.72 (1.06-2.56) 0.82 (0.66-0.97)*

MAP 1.94 (1.15-3.27)previous preeclampsia 3.27(1.06-10.15)

AIx: augmentation index, AP: augmentation pressure, MAP: mean arterial pressure * Area significantly different from 0.5

Chapter 5

93


Table 2. Hemodynamic measurements at 8-11 weeks gestational age and at 29 weeks postpartum. Nulliparous women (n=235) Multiparous women with previous preeclampsia (n=42)

Outcome No preeclampsia Preeclampsia No preeclampsia Preeclampsia8-11 w 29 w PP 8-11 w 29 w PP 8-11 w 29 w PP 8-11 w 29 w PP

n=225 N=172 n=10 N=7 n=35 n=25 N=7 N=5

Heart rate (b/m) 70±8 66±9* 69±6 69±6 70±8 69±13 71±6 70±11

SBP (mmHg) 121±8 119±11* 129±15† 132±16V 119±8 122±13 129±5† 125±10

DBP (mmHg) 67±6 69±8* 71±9† 73±9 68±7 74±9* 76±5† 75±8

MAP (mmHg) 87±6 89±9* 96±12† 97±12† 89±8 95±8* 98±5† 97±8

AIx (%) 17 (-11-47) 24 (3-45)* 26 (9-36) † 25 (20-38) 17 (-3-39) 26 (11-48)* 23 (17-33) 27 (18-29)

AP (mmHg) 6.5 (0.9-22.5) 9.4 (1.3-22.2)* 12.3(2.7-19.2) † 11.1 (9.0-25.7) 7.1 (1.1-20.1) 9.5 (3.7-31.3)* 9.8 (6.5-16.1) 10.1(7.4-14.1)DAIx - 7.5±10.9 - 2.2±12.4 - 6.7±10.5 - 0.25±10.5

Data are presented as mean (± standard deviation) or median (range) *P<0.05 vs first trimester,†P<0.05 vs. no preeclampsia, PP: postpartum, SBP: systolic blood pressure, DBP : diastolic blood pressure, MAP: mean arterial pressure, AIx : augmentation index, AP : augmentation pressure

trimester and 29 weeks postpartum values were comparable. Nulliparous and parous

women with a previous preeclampsia were similar in this respect. DAIx was higher in

women with normal pregnancy compared to women who developed preeclampsia,

although this difference did not reach statistical significance.

Figure 1 shows bar plots of the incidence of preeclampsia for different classes of AIx, AP

and MAP, for the complete population of nulliparous and parous women with previous

preeclampsia. All plots had statistical significance by chi-square analysis except for AIx.

Within the time window of 8-11 weeks, gestational age was not correlated with AIx, AP

or MAP (Pearson correlation coefficient < 0.1).

Table 3 shows the results of the planned logistic regression analyses. AIx, AP and MAP at

8-11 weeks gestational age were significantly associated with later preeclampsia when

adjusted for parity. When entered in combination with MAP AIx and AP remained in the

model. The model with the highest area under the curve of the ROC curve used AP, MAP

and parity.

DISCUSSION

In this prospective cohort of women with normal blood pressure we found positive

associations between higher AIx, AP and mean arterial pressure at 8-11 weeks gestational

age and later development of preeclampsia. Within the gestational age window of

the study these parameters were not affected by gestational age. Although all indices

generally demonstrated similar associations and measures of model fit, the combination

of augmentation pressure, previous preeclampsia and mean arterial pressure showed

94

the strongest association and best model fit. The additive effect of AIx or AP to MAP and

parity was small.

Based on our data we conclude that differences in early pregnancy hemodynamic

measurements between women, who develop preeclampsia later and women who do not,

are not pre-existing and only due to early pregnancy adaptation. We found no differences

between nulliparous women and parous women with previous preeclampsia both in

the first trimester as well as 29 weeks postpartum. Both subgroups had a significantly

lower blood pressure, AIx and AP in the first trimester compared to postpartum values

when pregnancy developed normally, reflecting early pregnancy adaption. Women who

developed preeclampsia, irrespective of parity, had similar values in the first trimester

as 29 weeks postpartum. Postpartum values of AIx and AP were comparable between

women who had an uncomplicated pregnancy and women who had preeclampsia. Blood

pressure was higher 29 weeks postpartum in nulliparous women with preeclampsia;

this seems to contradict our previous statement unless a longer interval is needed for

normalization of blood pressure. A longer recovery interval for blood pressure is supported

by our observation of similar first trimester blood pressure values in nulliparous women

and in parous women with a previous preeclampsia with an uncomplicated pregnancy.

One of the limitations of our study is our limited sample size; we cannot rule out small

differences between the subgroups. Because of the limited sample size we could not

adjust for other known confounders like age and body mass index. We excluded parous

women with previous preeclampsia who had chronic hypertension because the use of

antihypertensive medication could influence arterial stiffness. In addition, these women

were not included because the elevated risk of these conditions for preeclampsia or

pregnancy induced hypertension precludes the usefulness of further risk assessment.28

Despite not selecting this high risk group, the incidence of preeclampsia was high in the

remaining women with previous preeclampsia (4.3 % vs. 16.7%)

We could not confirm the high association of the AIx and AP with later preeclampsia as

found by Khalil et al.10;11 In two studies (one cohort, one case-control) using applanation

tonometry for calculation of AIx, AIx-75 and AP at 110/7-13 6/7 weeks of pregnancy this

group observed an area under the ROC curve of approximately 0.9 for AIx, AIx-75

and AP using univariate analysis. Their first study population was specified as low-

risk, with 43% nulliparous women, but with an incidence of preeclampsia of 6.7% (of

which 50% developed preeclampsia <34 weeks). This is a high incidence of severe

preeclampsia in a low risk population and cannot be compared with our nulliparous

women. The second population was specified as high risk with 10.3 % preeclampsia. A

second difference is that in our study early pregnancy blood pressure differed between

women who developed preeclampsia and women who did not, with a comparable level

Chapter 5

95


of association as described by Cnossen et al15, while in both studies by Khalil et al no

difference in blood pressure was observed between women who developed preeclampsia

and women who did not. Thirdly, their population contained a high percentage of Afro-

Caribbean women (37%), while our population was predominantly Caucasian. These

differences in study population could explain the difference in level of association of AIx

and AP with our study.

Because a negative linear correlation exists between heart rate and AIx some advocate

to standardise AIx calculation for a heart rate of 75 beats per minute. However, studies

regarding arterial stiffness and changing heart rates report conflicting results. Studies

using pacing have different results compared to studies with other methods to adjust

for heart rate.29;30 This makes it impossible to use standard formula’s to adjust the AIx

to heart rate and these formulas are not validated in pregnancy. We therefore decided

not to use a standardised AIx. In our study population heart rate was similar between

subgroups and thus standardisation would not have changed our findings.

We used finger arterial waveform registration for calculation of AIx and AP while most

studies use Sphygmocor® or other applanation tonometry devices. One of the advantages

of finger arterial waveform registration is that it uses a cuff around the finger instead of

a transducer which has to be hold in place. This easier application allows for a sufficient

rest period and stabilization of the signal in advance of the measurement window.

All these devices have in common that the aortic pressure used for calculating AIx and

AP is not measured but reconstructed from a peripheral pulse wave registration. Invasive

methods are not widely useful in populations for screening purposes. This approach has

been validated in a wide variety of healthy and patient populations.31

CONCLUSION

Combined use of MAP and AP at 8-11 weeks in a model had the highest level of association

with later preeclampsia although the differences between the models were small. Our

results do not support the addition of AIx or AP to blood pressure measurement, which is

far easier to apply, for the assessment of risk for preeclampsia.

The differences in blood pressure, AIx or AP in the first trimester between women who will

develop preeclampsia and women whose pregnancy will develop normally seem to be

pregnancy related and not pre-existing. These differences could reflect an inadequate

hemodynamic adaption to pregnancy.

96

Reference List (1) Steegers EA, von DP, Duvekot JJ, Pijnenborg R. Pre-eclampsia. Lancet 2010; 376(9741):631-

644. (2) Khan KS, Wojdyla D, Say L, Gulmezoglu AM, Van Look PF. WHO analysis of causes of maternal

death: a systematic review. Lancet 2006; 367(9516):1066-1074. (3) Sattar N, Greer IA. Pregnancy complications and maternal cardiovascular risk: opportunities for

intervention and screening? BMJ 2002; 325(7356):157-160. (4) Manten GT, Sikkema MJ, Voorbij HA, Visser GH, Bruinse HW, Franx A. Risk factors for

cardiovascular disease in women with a history of pregnancy complicated by preeclampsia or intrauterine growth restriction. Hypertens Pregnancy 2007; 26(1):39-50.

(5) Laurent S, Boutouyrie P, Asmar R, Gautier I, Laloux B, Guize L et al. Aortic stiffness is an independent predictor of all-cause and cardiovascular mortality in hypertensive patients. Hypertension 2001; 37(5):1236-1241.

(6) Spasojevic M, Smith SA, Morris JM, Gallery ED. Peripheral arterial pulse wave analysis in women with pre-eclampsia and gestational hypertension. BJOG 2005; 112(11):1475-1478.

(7) Ronnback M, Lampinen K, Groop PH, Kaaja R. Pulse wave reflection in currently and previously preeclamptic women. Hypertens Pregnancy 2005; 24(2):171-180.

(8) Robb AO, Mills NL, Din JN, Smith IB, Paterson F, Newby DE et al. Influence of the menstrual cycle, pregnancy, and preeclampsia on arterial stiffness. Hypertension 2009; 53(6):952-958.

(9) Avni B, Frenkel G, Shahar L, Golik A, Sherman D, Dishy V. Aortic stiffness in normal and hypertensive pregnancy. Blood Press 2010; 19(1):11-15.

(10) Khalil AA, Cooper DJ, Harrington KF. Pulse wave analysis: a preliminary study of a novel technique for the prediction of pre-eclampsia. BJOG 2009; 116(2):268-276.

(11) Khalil A, Cowans NJ, Spencer K, Goichman S, Meiri H, Harrington K. First-trimester markers for the prediction of pre-eclampsia in women with a-priori high risk. Ultrasound Obstet Gynecol 2010; 35(6):671-679.

(12) Paez O, Alfie J, Gorosito M, Puleio P, de MM, Prieto N et al. Parallel decrease in arterial distensibility and in endothelium-dependent dilatation in young women with a history of pre-eclampsia. Clin Exp Hypertens 2009; 31(7):544-552.




(16) Benetos A, Laurent S, Asmar RG, Lacolley P. Large artery stiffness in hypertension. J Hypertens Suppl 1997; 15(2):S89-S97.


(18) Penaz J. Photoelectric measurement of blood pressure, volume and flow in the finger. Digest 10th Int Conf Med Biol Engp . 1973.

(19) Imholz BP, Wieling W, van Montfrans GA, Wesseling KH. Fifteen years experience with finger arterial pressure monitoring: assessment of the technology. Cardiovasc Res 1998; 38(3):605-616.

(20) Wesseling KH, de Wit B., van der Hoeven GM, van Goudoever J, Settels JJ. Physiocal, calibrating finger vascular physiology for finapres. Homeostasis 1995; 36:67-82.

Chapter 5

97


(21) Guelen I, Westerhof BE, Van Der Sar GL, van Montfrans GA, Kiemeneij F, Wesseling KH et al. Finometer, finger pressure measurements with the possibility to reconstruct brachial pressure. Blood Press Monit 2003; 8(1):27-30.

(22) Hehenkamp WJ, Rang S, van Goudoever J, Bos WJ, Wolf H, van der Post JA. Comparison of Portapres with standard sphygmomanometry in pregnancy. Hypertens Pregnancy 2002; 21(1):65-76.

(23) Rang S, de Pablo LB, van Montfrans GA, Bouma BJ, Wesseling KH, Wolf H. Modelflow: a new method for noninvasive assessment of cardiac output in pregnant women. Am J Obstet Gynecol 2007; 196(3):235-238.

(24) Bos WJ, van GJ, van Montfrans GA, van den Meiracker AH, Wesseling KH. Reconstruction of brachial artery pressure from noninvasive finger pressure measurements. Circulation 1996; 94(8):1870-1875.

(25) Segers P, Rietzschel E, Heireman S, De BM, Gillebert T, Verdonck P et al. Carotid tonometry versus synthesized aorta pressure waves for the estimation of central systolic blood pressure and augmentation index. Am J Hypertens 2005; 18(9 Pt 1):1168-1173.

(26) Hirata K, Kawakami M, O’Rourke MF. Pulse wave analysis and pulse wave velocity: a review of blood pressure interpretation 100 years after Korotkov. Circ J 2006; 70(10):1231-1239.


(28) Caritis S, Sibai B, Hauth J, Lindheimer M, VanDorsten P, Klebanoff M et al. Predictors of pre-eclampsia in women at high risk. National Institute of Child Health and Human Development Network of Maternal-Fetal Medicine Units. Am J Obstet Gynecol 1998; 179(4):946-951.

(29) Wilkinson IB, MacCallum H, Flint L, Cockcroft JR, Newby DE, Webb DJ. The influence of heart rate on augmentation index and central arterial pressure in humans. J Physiol 2000; 525 Pt 1:263-270.

(30) Pannier B, Slama MA, London GM, Safar ME, Cuche JL. Carotid arterial hemodynamics in response to LBNP in normal subjects: methodological aspects. J Appl Physiol 1995; 79(5):1546-1555.

(31) Gallagher D, Adji A, O’Rourke MF. Validation of the transfer function technique for generating central from peripheral upper limb pressure waveform. Am J Hypertens 2004; 17(11 Pt 1):1059-1067.

16Karlijn C. Vollebregt, Marcel F. van der Wal, Hans Wolf, Tanja G.M. Vrijkotte, Kees Boer, Gouke J. Bonsel

BJOG 2008;115:607-615

Is psychosocial stress in first ongoing pregnancies associated with preeclampsia and gestational hypertension?

100

ABSTRACT

Objective

Investigating the association of preeclampsia and gestational hypertension with

psychosocial stress in the first half of pregnancy.

Design Prospective community-based cohort study

Setting Amsterdam, The Netherlands

Population

Between January 2003 and March 2004 all pregnant Amsterdam women (n= 12 377)

were invited to fill out a questionnaire with sociodemographic and psychosocial variables

(response rate 67%). Only nulliparous women with a singleton pregnancy, who completed

the questionnaire before 24 weeks, and delivered after 24 weeks, were included.

Methods

A postpartum questionnaire was used to gather information on hypertension or proteinuria.

If this questionnaire was not available the national obstetric register was used for pregnancy

outcome. Medical files were examined for women with hypertension and/ or proteinuria

to confirm the diagnosis of pre-eclampsia and gestational hypertension according to the

International Society for the Study of Hypertension in Pregnancy guidelines. Psychosocial

stress was defined as work stress (Work Experience and Appreciation Questionnaire partly

based on the Job Content Instrument of Karasek et al.), anxiety (the State-Trait Anxiety

Inventory, STAI), depression (Center for Epidemiological Studies Depression Scale, CES-

D) and pregnancy related anxiety (PRAQ-R). The association of psychosocial stress with

the incidence of preeclampsia and gestational hypertension was explored by multivariate

analysis adjusted for sociodemographic and medical confounders.

Main outcome measures Incidence of preeclampsia and gestational hypertension.

Results

3679 women were included. The incidence of preeclampsia and gestational hypertension

was respectively 3.5% and 4.4%. Work stress, anxiety, pregnancy related anxiety or

depression had no effect on the incidence of preeclampsia or gestational hypertension.

Conclusion

Psychosocial stress in the first half of pregnancy does not influence the incidence of

preeclampsia and gestational hypertension in nulliparous women.

Chapter 6

101

Preeclampsia and psychosocial stress

INTRODUCTION

Hypertensive disorders of pregnancy are a major cause of maternal and fetal mortality

and morbidity, and affect approximately 8% of all pregnancies.1 Although risks for adverse

events are largest in case of pre-eclampsia, gestational hypertension too is associated

with maternal and fetal mortality and morbidity.

The precise pathogenesis of preeclampsia and gestational hypertension is unknown. The

origin is thought to be associated with abnormal placentation early in pregnancy, which

may be caused by immunological, genetic or environmental factors.1 Maternal factors

like increased age, obesity, chronic hypertension, diabetes mellitus, renal and connective

tissue diseases are all known to be associated with both preeclampsia and gestational

hypertension. It is still unclear whether preeclampsia, gestational hypertension and

haemolysis, elevated liver enzymes, low platelet count (HELLP) syndrome are different

stages of the same disorder or related, yet different syndromes.2;3 Here we focus on

the claimed association between psychosocial stress of the mother and hypertensive

disorders of pregnancy.4-8 Stress activates the hypothalamus- pituitary- adrenal cortex

system (HPA) which in turn increases in levels of corticosteroids and catecholamines.

Stress also activates the sympathetic nervous system and affects the immune system.9;10

Increased levels of corticotrophin releasing hormone 11 and increased sympathetic activity

have been observed in women with preeclampsia and gestational hypertension.12-14

Because psychosocial stress is, in part, an external factor, which could be mediated by

intervention, it is relevant to obtain more knowledge on the association with preeclampsia

and gestational hypertension.

This study investigates the association of psychosocial stress early in pregnancy and the

development of hypertensive disorders later in pregnancy in a prospective cohort of

nulliparous women and compares preeclampsia and gestational hypertension to assess

possible dissimilarities in origin.

METHODS

Study population

The present study is part of the Amsterdam Born Children and their Development study

(ABCD-study).15 This prospective community-based study examined the relationship

between various lifestyles during pregnancy and pregnancy outcome in a multicultural

population. Between January 2003 and March 2004, 12 381 Amsterdam women who

attended antenatal care for their first check-up were approached by their obstetric caregiver

102

to participate in the study. Within 2 weeks a questionnaire (pregnancy questionnaire) on

sociodemographic characteristics, medical history, psychosocial stress and lifestyle was

sent to the pregnant women to be filled out at home and returned by prepaid mail. All

nulliparous women with singleton pregnancy, who filled out the pregnancy questionnaire

before 24 weeks gestational age, and who delivered after 24 weeks were included.

Pregnancy outcome, infant gender, birth weight and gestational age were additionally

obtained from a second questionnaire (infant questionnaire) who completed 3-5 months

after delivery and the Youth Health Care of the Municipal Health Service registry in

Amsterdam. In case no infant questionnaire was filled out information about pregnancy

outcome was gathered from the National Obstetric Register. In the women who were

registered as having hypertension and/ or proteinuria we verified from medical records

if the diagnosis was gestational hypertension or preeclampsia according to the criteria

of the International Society for the Study of Hypertension in Pregnancy.16 Five percent of

the women in the control group were randomly selected for verification of data from their

medical records. The study protocol was approved by the medical ethical committees of

all Amsterdam hospitals and the Municipal Privacy Protection Committee of Amsterdam.

Participation and permission for data retrieval rested on written consent.

Definitions of hypertensive disorders

Gestational hypertension was defined as a diastolic pressure ≥ 90 mmHg after 20 weeks

in a previously normotensive woman. Preeclampsia was defined by the combination

of gestational hypertension and proteinuria ≥ 0.3 g/ 24 h or dipstick ≥ ++ after 20

weeks gestation, according to the International Society for the Study of Hypertension in

Pregnancy guidelines.16 Women with chronic hypertension who had a diastolic pressure

≥ 90 mmHg after 20 weeks gestation were included in the gestational hypertension

group. Superimposed preeclampsia was defined as de novo proteinuria ≥ 0.3 g/ 24

h or dipstick ≥ ++ after 20 weeks gestation together with a diastolic pressure ≥ 90

mmHg in women with chronic hypertension. Chronic hypertension was defined as a

systolic pressure ≥140 mmHg and/or diastolic pressure ≥ 90 mmHg or the necessity for

antihypertensive treatment before pregnancy or before 20 weeks gestational age.

Measurement of psychosocial stress

Psychosocial stress was decomposed into four main components; work stress, depression,

anxiety and pregnancy-related anxiety, with all concepts measured by internationally

accepted instruments. Work stress was examined using four variables: total working

hours, workload, work control and job strain. Total working hours was defined as the

weekly hours of paid work which was categorized in ‘no work’ ‘<32 h’ and ‘≥ 32 h’.

Workload was measured using three scales and workcontrol using one scale of the Work

Chapter 6

103


Experience and Appreciation Questionnaire of Van Veldhoven et al.17, partly based on

the Job Content Instrument of Karasek et al.18 The scales used were pace of work (11

items), mental work load (7 items), physical work load (7 items) and job control (11

items). Items were scored on a four-point-scale (0=never, 1=sometimes, 2=often and

3=always). The reported working hours were used as a separate variable. The scores

for workload and workcontrol were trichotomised into three categories, for workload:

low (< 50th percentile) moderate (between 50th and 90th percentile) and high (>90th

percentile), for workcontrol: high (> 50th percentile), moderate (between 10th and 50th

percentile) and low (< 10th percentile). Jobs that are high in workload (>P90) and low

in control (<P10) are presumed to give the most stress (job strain) and negative health

effects.19 Women with high workload and low or moderate workcontrol were defined as

having high job strain, women with low workload and high or moderate control as low

job strain and al remaining women as moderate job strain. The total score of the three

mood scales were all trichotomised round the 50th and 90th percentiles to create three

categories: low, moderate and high.

Depressive symptoms were measured by the Dutch version (validated by Hanewald)20 of

the Center for Epidemiological Studies Depression Scale (CES-D)21. The 20-item CES-D

scale assesses the self-reported frequency of depressive symptoms experienced over the

past week. Items were scored on a four-point scale (0= (almost) never, 1=sometimes,

2=often and 3=very often).

The Dutch version of the State-Trait Anxiety Inventory (STAI) was used to measure anxiety

symptoms22. The state anxiety scale consists of 20 items and asks subjects to indicate

how they felt in the past week. Subjects rated each item on a four-point scale (0= almost

never, 1=sometimes, 2=often and 3=very often).

Anxiety and specific fears related to pregnancy were measured by three scales developed

by Huizink et al.23 and based on the revised version of the Pregnancy Related Anxiety

Questionnaire (PRAQ-R) of Van den Bergh.24 The scales used were ‘fear of giving birth’

(three items), ‘fear of bearing a handicapped child’ (four items) and ‘concern about

one’s appearance’ (three items). Items were scored on a four-point scale (4=very true,

3=not true, 2=true, 1=certainly not true).

In this study internal consistencies (Cronbachs’alphas) were 0.90 for the depression

scale, 0.94 for the anxiety scale, 0.81 for the total pregnancy related anxiety scale, 0.85

for the total work load scale and 0.92 for the control scale.

Characteristics and Covariates

Covariates were decomposed in two categories; sociodemographic and medical

covariates. Sociodemographic covariates are maternal age, education, marital status and

104

ethnicity. Medical covariates included pre-pregnancy body mass index (BMI), smoking

during pregnancy, previous miscarriage/ abortion, chronic hypertension, diabetes and

vaginal haemorrhage during pregnancy. Maternal age, gestational age (based on

routine first trimester ultrasound examination or, if unavailable, on the timing of the last

menstrual period), birth weight and infant gender, were obtained from the Youth Health

Care registration of the Municipal Health Service in Amsterdam. All other variables

were self reported and defined as follows: maternal education (years of education after

primary school; 3 categories: < 5 years; 5-10 years; > 10 years), marital status (married/

cohabiting or single) and ethnicity (Caucasian, black, Turkish/Moroccan and other

countries). Pre-pregnancy BMI was calculated using weight and height. Missing data on

pregravid maternal weight (about 5% compared to 1% general missing) were formally

imputed by models using maternal length and parity data, retaining the variability of

the data. Smoking during pregnancy (yes or no), previous miscarriage/ abortion (yes or

no), vaginal haemorrhage during pregnancy (yes or no). Women were asked if they had

hypertension, diabetes mellitus or other disease during their pregnancy and if they used

medication. This was classified according to the International Classification for Primary

Care. If women used medication for a specific disease but did not fill in this disease (e.g.

insulin) they were classified as having the associated disease (diabetes).

Statistical analysis

We estimated the effects of psychosocial stress for all seven variables of psychosocial

stress separately by predefined explorative logistic regression models, with increasing

levels of adjustment. First univariate analysis was applied for each variable to estimate the

unadjusted effects. Subsequently multivariate logistic regression was performed to adjust

for medical covariates. Those medical covariates that were statistically significant in this

analysis were combined with the sociodemographic covariates in a second multivariate

logistic analysis. The selection of covariates that influenced the model significantly was

performed by back step procedure with the p value for entry set at 0.05 and for removal

at 0.10.

For all work-related stress variables employed women with the least heavy working

condition were taken as the reference category, while unemployed women were taking

as a separate category. For anxiety, depression and pregnancy related anxiety; women

with the least anxiety/depressive feelings were taken as the reference category. The level

of significance used was p<0.05. Statistical calculations were performed with SPSS

12.0.2(SPSS Inc.Chicago, IL).

Chapter 6

105


Figure 1. Inclusion

106

RESULTS

In total, 12 373 women were invited for the ABCD-study and 8266 women returned the

questionnaire (response rate 67%). 3679 women were included in our study according

to our selection criteria (see figure 1). The median gestational age at completing

the pregnancy questionnaire was 15.6 weeks (quartiles 14.0-17.3). The diagnosis

preeclampsia or gestational hypertension could not be confirmed in the medical files in

300 of the 599 women, who were registered as hypertensive disorder in the National

Obstetrics Register, or who indicated this in the second questionnaire. These women were

allocated to the control group. The medical records of the randomly selected women in

the control group (n=190) confirmed correct allocation to the control group. Thirteen

files could not be traced and these women were allocated to the control group (Figure 1).

Table 1. Sociodemographic and medical characteristics of included women.Total group Controls PE GH

N (%) 3679(100) 3390 (92.1) 128 (3.5) 161 (4.4)Birth weight (g) 3381 ± 557 3397 ± 530 2875 ± 979 3380 ± 561

Gestational age (days) 278 ± 14 278 ± 14 264 ± 22 278 ± 13

Age (years) 29.9 ± 5.1 29.8 ± 5.1 31.6 ± 5.0 30.9 ± 4.7

Etnicity

-Caucasian 2461 (100) 2250 (91.4) 91 (3.7) 120 (4.9)

-Black 157 (100) 148 (94.3) 4 (2.5) 5 (3.2)

-Turkish/Moroccan 347 (100) 324 (93.4) 9 (2.6) 14 (4.0)

- Other 712 (100) 666 (93.5) 24 (3.4) 22 (3.1)

Education N (%)

-0-5 years 557 (100) 527 (94.6) 17 (3.1) 13 (2.3)

-6-10 years 1445 (100) 1324 (91.6) 48 (3.3) 73 (5.1)

->10 years 1651 (100) 1514 (91.7) 63 (3.8) 74 (4.5)

Marriage/ co-habitation N (%)

no 497 (100) 471 (94.8) 10 (2.0) 16 (3.2)

BMI (kg/m2) 22.6 ± 3.7 22.4 (3.6) 23.7 (4.4) 24.5 (4.1)

Chronic hypertension N (%)

yes 92 (100) 56 (60.9) 21 (22.8) 15 (16.3)

Diabetes N (%)

yes 19 (100) 16 (84.2) 1 (5.3) 2 (10.2)

Smoking in pregnancy N (%)

yes 343 (100) 329 (94.8) 10 (2.9) 8 (2.3)

Miscarriage/ abortion N (%)

yes 969 (100) 890 (91.8) 30 (3.1) 49 (5.1)

Hemorrhage N (%)yes 431 (100) 404 (93.7) 15 (3.5) 12 (2.8)

Data are expressed as mean ± SD or n (%)

Chapter 6

107


The incidence of preeclampsia was 3.5% and 4.4% of the women had gestational

hypertension. Women in the preeclampsia group were older than those in the control

group and both women in the preeclampsia and gestational hypertension group had a

higher pre-pregnancy BMI and had more often chronic hypertension. Fewer women in

the preeclampsia group were married or lived with a partner (Table 1).

Table 2. Odds ratio’s with 95% confidence limits for preeclampsia (PE) of psychosocial stress, calculated by univariate analysis, by multivariate analysis using Model 1 and Model 2.

Controls(%) PE(%) Univariate effects Multivariate effects model 1

Multivariate effects model 2

Working hoursNo work 24.5 14.4 0.53(0.29-0.97)* 0.49 (0.26-0.89)* 0.58 (0.30-1.11)

< 32 19.5 21.6 1 1 1

≥ 32 h 66.0 64.0 1.04 (0.67-1.63) 1.01 (0.64-1.60) 0.93 (0.59-1.49)

Workload

No work 24.5 14.3 0.61 (0.35-1.06) 0.54 (0.31-0.96)* 0.69 (0.38-1.27)

Low 35.6 34.1 1 1 1

Moderate 31.9 39.7 1.30 (0.86-1.97) 1.17 (0.76-1.80) 1.16 (0.75-1.78)

High 8.0 11.9 1.60 (0.88-2.92) 1.46 (0.78-2.72) 1.76 (0.94-3.29)

Workcontrol

No work 24.6 14.3 0.53 (0.31-0.92)* 0.49 (0.28-0.87)* 0.66 (0.36-1.21)

High 36.5 39.7 1 1 1

Moderate 30.5 32.5 0.98 (0.64-1.49) 0.99 (0.68-1.29) 1.07 (0.69-1.65)

Low 8.4 13.5 1.49 (0.85-2.62) 1.25 (0.68-2.29) 1.51 (0.81-2.82)

Jobstrain

No work 24.6 14.3 0.62 (0.35-1.09) 0.55 (0.31-0.99)* 0.71 (0.38-1.32)

Low 33.1 31.2 1 1 1

Moderate 36.7 48.0 1.39 (0.92-2.91) 1.27 (0.83-1.93) 1.27 (0.83-1.95)

high 5.6 6.5 1.36(0.65-2.85) 1.24 (0.58-2.66) 1.61 (0.75-3.49)

Anxiety

Low 45.9 50.0 1 1 1

Moderate 40.8 37.5 0.82 (0.56-1.20) 0.74 (0.49-1.10) 0.86 (0.57-1.30)

High 13.3 12.5 1.08 (0.61-1.91) 0.95 (0.52-1.75) 1.38 (0.74-2.58)

Depression

Low 46.5 50.4 1 1 1

Moderate 42.7 37.0 0.80 (0.55-1.18) 0.76 (0.51-1.13) 0.86 (0.58-1.29)

High 10.8 12.6 1.06 (0.60-1.85) 0.85 (0.46-1.56) 1.29 (0.68-2.42)

Pregnancy related anxiety

Low

Moderate 46.3 50.0 1 1 1

High 42.5 38.1 0.83 (0.57-1.22) 0.80 (0.54-1.19) 0.82 (0.55-1.22)

11.2 11.9 0.98 (0.55-1.73) 0.91 (0.50-1.65) 1.13 (0.61-2.07)Model 1:Adjusted for BMI, chronic hypertension, diabetes mellitus, smoking in pregnancy, previous miscarriage/ abortion and haemorrhage; Model 2:Model 1 + additional adjusted age, ethnicity, education and marriage/ co-habitation; *significance level <0.05

108

Working, irrespective of the duration, was associated with approximately twice the

risk for preeclampsia and gestational hypertension in comparison to not working in

univariate analysis. If high or moderate work control was present, risks were similar to not

working. Low work control increased the risks considerably (pre-eclampsia about 3 times

Table 3. Odds ratio’s with 95% confidence limits for gestational hypertension (GH) of psychosocial stress, calculated by univariate analysis, by multivariate analysis using Model 1 and Model 2.

Controls (%) GH(%) Univariate effects Multivariate effects model 1


Working hoursNo work 24.5 15.0 0.63 (0.37-1.10) 0.58 (0.33-1.04) 0.72 (0.40-1.31)

< 32 h 19.5 18.8 1 1 1

≥ 32 h 66.0 66.2 1.24 (0.82-1.88) 1.30 (0.84-2.00) 1.22 (0.79-1.89)

Workload

No work 24.5 15.1 0.57 (0.35-0.92)* 0.50 (0.30-0.82)** 0.66 (0.39-1.14)

Low 35.6 38.4 1 1 1

Moderate 31.9 38.4 1.12 (0.78-1.61) 1.09 (0.75-1.59) 1.11 (0.76-1.61)

High 8.0 8.1 0.97 (0.53-1.81) 0.86 (0.44-1.65) 0.97 (0.50-1.88)

Workcontrol

No work 24.6 15.1 0.52 (0.32-0.83)* 0.45 (0.27-0.73)* 0.60 (0.35-1.03)

High 36.5 43.4 1 1 1

Moderate 30.5 33.4 0.92 (0.63-1.32) 0.89 (0.61-1.30) 0.93 (0.64-1.37)

Low 8.5 8.1 0.82 (0.45-1.51) 0.71 (0.37-1.35) 0.83 (0.43-1.59)

Jobstrain

No work 24.6 15.1 0.58 (0.35-0.94)* 0.50 (0.30-0.83)** 0.67 (0.39-1.16)

Low 33.1 35.3 1 1 1

Moderate 36.7 43.3 1.11 (0.77-1.59) 1.08 (0.77-1.57) 1.11 (0.77-1.61)

high 5.6 6.3 1.05 (0.53-2.09) 0.87 (0.41-1.84) 1.03 (0.48-2.20)

Anxiety

Low 45.9 48.5 1 1 1

Moderate 40.8 39.1 0.97 (0.69-1.36) 0.84 (0.59-1.19) 0.95 (0.66-1.36)

High 13.3 12.4 1.21 (0.73-2.01) 1.11 (0.65-1.89) 1.46 (0.84-2.54)

Depression

Low 46.5 49.1 1 1 1

Moderate 42.7 38.5 0.86 (0.61-1.20) 0.78 (0.55-1.11) 0.85 (0.60-1.22)

High 10.8 12.4 1.07 (0.65-1.78) 0.87 (0.51-1.51) 1.19 (0.67-2.11)

Pregnancy related anxiety

Low

Moderate 46.3 44.7 1 1 1

High 42.5 41.6 1.01 (0.72-1.43) 0.96 (0.67-1.36) 0.99 (0.69-1.41)11.2 13.7 1.25 (0.77-2.05) 1.18 (0.71-1.96) 1.15 (0.87-2.47)

Model 1:Adjusted for BMI, chronic hypertension, diabetes mellitus, smoking in pregnancy, previous miscarriage/ abortion and haemorrhage; Model 2:Model 1 + additional adjusted age, ethnicity, education and marriage/ co-habitation; * significance level <0.05; **significance level <0.01

Chapter 6

109


Table 4. Odds ratios of associated factors for preeclampsia and gestational hypertension compared with the controls.

Pre-eclampsia Significance Gestational Hypertension

Significance

Age (year) 1.09 (1.04-1.14) <0.001 1.04(0.99-1.08) 0.08Ethnicity

-Caucasian 1 1

-Black 0.86 (0.25-2.95) 0.81 0.67 (0.20-1.58) 0.43

-Turkish/Moroccan 1.08 (0.49-2.40) 0.85 1.04 (0.53-2.03) 0.91

- Other 0.98 (0.60-1.61) 0.95 0.66 (0.41-1.08) 0.10

Education

-0-5 years 1 1 1

-6-10 years 0.91 (0.48-1.71) 0.612 2.19 (1.12-2.27) 0.22

->10 years 1.00 (0.52-1.93) 0.847 2.01 (0.99-4.06) 0.052

Not married/ co-habitation 0.38 (0.18-0.81) 0.01 0.72 (0.41-1.29) 0.27

BMI (kg/m2) 1.07 (1.03-1.11) 0.002 1.12 (1.09-1.15) <0.001

Chronic hypertension 12.51 (6.95-22.53) <0.001 5.80 (3.00-11.22) <0.001

Diabetes 0.79 (0.91-6.86) 0.83 1.51 (0.29-7.60) 0.62

Smoking in pregnancy 1.08 (0.56-2.21) 0.76 0.60 (0.29-1.26) 0.18

miscarriage/ abortion 0.77 (0.49-1.20) 0.25 1.37 (0.95-1.97) 0.91Haemorrhage 0.94 (0.54-1.67) 0.85 0.55 (0.30-1.01) 0.053

Odds ratio’s adjusted for all the other variables in table

and gestational hypertension about 2 times). Any grade of job strain was associated

with approximately twice the risk for preeclampsia and gestational hypertension in

comparison to not working. No association between the stress components (job strain,

anxiety, depression or pregnancy-related anxiety) and preeclampsia or gestational

hypertension remained after adjustment for medical and socioeconomic covariates by

logistic regression analysis (Table 2 and 3).

In table 4 the odds ratio’s of the covariates that were used in this analysis are demonstrated.

Maternal age, high BMI and chronic hypertension contributed significantly to the model

for preeclampsia and for gestational hypertension. Being single was associated with a

significant reduction of preeclampsia, but did not contribute to the model for gestational

hypertension. None of the other confounders contributed significantly to the models,

although higher education and vaginal bleeding in the first half of pregnancy came close

to statistical significance in the models for gestational hypertension.

110

DISCUSSION

In this study we found no association of work stress, anxiety, depression or pregnancy-

related anxiety early in pregnancy and the development of gestational hypertension or

preeclampsia later in pregnancy.

Four studies observed an association of preeclampsia / gestational hypertension with job

strain in working women.4-6;25 Some of these studies used job title to calculate jobstrain

rather than individually experienced stress which is regarded as more trustworthy. 4;5

Although associations in these studies reached statistical significance, the size of their

effects was small (odds ratio’s between 1.5 and 3) and larger in case-control than in

cohort studies. Most studies adjusted for confounders, although the set of variables was

considerably smaller than used in our study. Kurki et al. found a significant association

of anxiety and depression with preeclampsia.26 They used a questionnaire with one

question for assessment of anxiety and a modified depression questionnaire which was

originally designed for psychiatric patients. They found a prevalence of depression of

30%, while other studies observed a prevalence in pregnant women of approximately 12

%, which suggest that the questionnaire Kurki et al. used was not suitable for population

studies.27 A recent study by Leeners et al. found a significant effect of emotional stress

in pregnancy on the incidence of hypertensive disorders of pregnancy (OR 1.6 CI 1.15-

2.9). In this retrospective case-control study a recall bias can not be excluded. The

emotional stress level experienced during the pregnancy by the women with hypertensive

disorders of pregnancy could be influenced by the disease and the experienced stress

from hospital admissions and treatment. Similar to our study, two other prospective

studies did not observe an association between depression or anxiety and hypertensive

complications.28;29 Sikkema et al. also used the STAI and PRAQ-R while Andersson et al.

used The Primary Care Evaluation of Mental disorders system, a questionnaire validated

for use in primary care settings.

One of the strong points of our study was that we investigated prospectively four main

components of psychosocial stress by use of well-validated questionnaires and in the

beginning of pregnancy, as it is assumed that the preclinical origin of hypertensive

disorders of pregnancy is situated early in pregnancy.1 On the contrary, we were not

informed about events later in pregnancy. However, it has been described that feelings of

anxiety and depression are stable during pregnancy.30

One of the limitations of our study was that all data about psychosocial stress were self

reported without clinical verification. This could cause a report bias underestimating or

exaggerating perceived stress. Possible cultural differences in interpreting questionnaires

and response style could also cause a cross-cultural bias in our multicultural population.

Chapter 6

111


To prevent a report bias in diagnosis we retrieved all medical files of women who reported

hypertension and/ or proteinuria according to the infant questionnaire or national

obstetric register to confirm the diagnosis preeclampsia or gestational hypertension.

There was no pre-eclampsia or gestational hypertension in the randomly selected files of

the control group.

Although our study group was large, due to the low incidence of preeclampsia and

gestational hypertension and due to low numbers in the highest work and anxiety

classification groups, the confidence intervals of the risk estimates are large. Small effects

could have been overlooked therefore. However, all odds ratio’s are close to 1 and

the clinical importance appears negligible, even if in a larger study population some

effects (e.g. no work) might have reached statistically significance. Psychosocial stress

is therefore not suitable as a target for intervention for prevention of preeclampsia and

gestational hypertension.

The covariates that were used in our study to adjust the effect of psychosocial stress on

the incidence of preeclampsia and gestational hypertension have been well described

by others and the size of the effects was comparable to what has been reported in

literature (Table 4).2 In literature both young as well as older mothers are at risk.1;2 The

mean age of the women in our study was 29.9 years and only 1.8 % was ≤ 18 years.

This is comparable with the population of the Netherlands and the low incidence of

teenage pregnancy could explain why only higher maternal age was a risk indicator.

Why women who were living single had a lower incidence of preeclampsia remains

unclear. In general, risk factors were comparable between preeclampsia and gestational

hypertension suggesting a common origin of these two conditions.

The association of psychosocial stress and cardiovascular disease has been well

described and it is assumed that this effect is mediated through a mechanism involving

excessive sympathetic nervous system activation.31-33 In women with preeclampsia

higher levels of cortisol releasing hormone and signs of higher sympathetic activity in

comparison to women with normal pregnancies have been documented.12-14 However,

haemodynamic and sympathetic-nerve responses to reflex tests (e.g. paced breathing,

Valsalva manoeuvre, cold pressor test, isometric hand-grip exercise) did not differ

between women with or without preeclampsia in these studies. This could implicate that

sympathetic hyperactivity is associated with the origin of the disease but that this is not

mediated by external stress.

112

CONCLUSION

This prospective cohort study could not detect any association between work and

four major sources of stress (anxiety, depression, pregnancy-related anxiety and job

strain) in the beginning of pregnancy, and the development with pre-eclampsia or

gestational hypertension. Apparent univariate associations with a lower incidence of

preeclampsia and gestational hypertension in women with no work disappeared after

proper adjustment. The observed covariates were nearly identical for preeclampsia and

gestational hypertension suggesting a common origin.

Chapter 6

113


Reference List 1. Sibai B, Dekker G, Kupferminc M. Pre-eclampsia. Lancet 2005;365:785-99. 2. Villar J, Carroli G, Wojdyla D, Abalos E, Giordano D, Ba'aqeel H et al. Preeclampsia,

gestational hypertension and intrauterine growth restriction, related or independent conditions? Am.J.Obstet.Gynecol. 2006;194:921-31.

3. Roberts J. Preeclampsia: a two stage disorder. In: Critchley H, Maclean A, Poston L, Walter J, editors. Preeclampsia. RCOG Press; 2003.

4. Klonoff-Cohen HS, Cross JL, Pieper CF. Job stress and preeclampsia. Epidemiology 1996;7:245-49. 5. Marcoux S, Berube S, Brisson C, Mondor M. Job strain and pregnancy-induced hypertension.

Epidemiology 1999;10:376-82. 6. Landsbergis PA, Hatch MC. Psychosocial work stress and pregnancy-induced hypertension.

Epidemiology 1996;7:346-51. 7. Paarlberg KM, Vingerhoets AJ, Passchier J, Dekker GA, van Geijn HP. Psychosocial factors

and pregnancy outcome: a review with emphasis on methodological issues. J.Psychosom.Res. 1995;39:563-95.

8. Leeners B, Neumaier-Wagner P, Kuse S, Stiller R, Rath W. Emotional stress and the risk to develop hypertensive diseases in pregnancy. Hypertens.Pregnancy. 2007;26:211-26.

9. Mulder EJ, Robles de Medina PG, Huizink AC, Van den Bergh BR, Buitelaar JK, Visser GH. Prenatal maternal stress: effects on pregnancy and the (unborn) child. Early Hum.Dev. 2002;70:3-14.

10. Coussons-Read ME, Okun ML, Nettles CD. Psychosocial stress increases inflammatory markers and alters cytokine production across pregnancy. Brain Behav.Immun. 2006.

11. Perkins AV, Linton EA, Eben F, Simpson J, Wolfe CD, Redman CW. Corticotrophin-releasing hormone and corticotrophin-releasing hormone binding protein in normal and pre-eclamptic human pregnancies. Br.J.Obstet.Gynaecol. 1995;102:118-22.

12. Greenwood JP, Scott EM, Walker JJ, Stoker JB, Mary DA. The magnitude of sympathetic hyperactivity in pregnancy-induced hypertension and preeclampsia. Am.J.Hypertens. 2003;16:194-99.

13. Schobel HP, Fischer T, Heuszer K, Geiger H, Schmieder RE. Preeclampsia -- a state of sympathetic overactivity. N.Engl.J.Med. 1996;335:1480-85.

14. Rang S, Wolf H, Montfrans GA, Karemaker JM. Serial assessment of cardiovascular control shows early signs of developing pre-eclampsia. J.Hypertens. 2004;22:369-76.

15. van Eijsden M, van der Wal MF, Bonsel GJ. Folic acid knowledge and use in a multi-ethnic pregnancy cohort: the role of language proficiency. BJOG. 2006;113:1446-51.

16. Brown MA, Lindheimer MD, de SM, Van AA, Moutquin JM. The classification and diagnosis of the hypertensive disorders of pregnancy: statement from the International Society for the Study of Hypertension in Pregnancy (ISSHP). Hypertens.Pregnancy. 2001;20:IX-XIV.

17. Van Velthoven M, Meijman T. het meten van psychosociale arbeidsbelasting met een vragenlijst: De Vragenlijst Beleving en Beoordeling van de Arbeid(VBBA). Amsterdam: Nederlands Instituut voor Arbeidsomstandigheden, 1994.

18. Karasek R, Theorell T. Healthy work: Stress, productivity and the reconstruction of working life. New York: Basic Books, 1990.

19. Karasek R, Brisson C, Kawakami N, Houtman I, Bongers P, Amick B. The Job Content Questionnaire (JCQ): an instrument for internationally comparative assessments of psychosocial job characteristics. J.Occup.Health Psychol. 1998;3:322-55.

20. Hanewald G.J.F.P. CES-D: De Nederlandse versie. Een onderzoek naar de betrouwbaarheid en validiteit. 1987. Amsterdam, University of Amsterdam, department of Clinical Psychology.

21. Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Applied Psychological Measurement 1977;1:385-401.

114

22. Spielberger C GRLR. STAI Manual for the State-Trait Anxiety Inventory. Palo Alto CA: Consulting Psychologists Press, 1970.

23. Huizink AC, Mulder EJ, Robles de Medina PG, Visser GH, Buitelaar JK. Is pregnancy anxiety a distinctive syndrome? Early Hum.Dev. 2004;79:81-91.

24. Bergh van de B.R.H. The influence of maternal emotions during pregnancy on fetal and neonatal behavior. Pre- and perinatal psychology 1990;5:119-30.

25. Wergeland E, Strand K. Working conditions and prevalence of pre-eclampsia, Norway 1989. Int.J.Gynaecol.Obstet. 1997;58:189-96.

26. Kurki T, Hiilesmaa V, Raitasalo R, Mattila H, Ylikorkala O. Depression and anxiety in early pregnancy and risk for preeclampsia. Obstet.Gynecol. 2000;95:487-90.

27. Bennett HA, Einarson A, Taddio A, Koren G, Einarson TR. Prevalence of depression during pregnancy: systematic review. Obstet.Gynecol. 2004;103:698-709.

28. Sikkema JM, Robles de Medina PG, Schaad RR, Mulder EJ, Bruinse HW, Buitelaar JK et al. Salivary cortisol levels and anxiety are not increased in women destined to develop preeclampsia. J.Psychosom.Res. 2001;50:45-49.

29. Andersson L, Sundstrom-Poromaa I, Wulff M, Astrom M, Bixo M. Implications of antenatal depression and anxiety for obstetric outcome. Obstet.Gynecol. 2004;104:467-76.

30. Field T, Hernandez-Reif M, Diego M, Schanberg S, Kuhn C. Stability of mood states and biochemistry across pregnancy. Infant Behav.Dev. 2006;29:262-67.

31. Rozanski A, Blumenthal JA, Kaplan J. Impact of psychological factors on the pathogenesis of cardiovascular disease and implications for therapy. Circulation 1999;99:2192-217.

32. Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F et al. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. Lancet 2004;364:937-52.

33. Lucini D, Di FG, Parati G, Pagani M. Impact of chronic psychosocial stress on autonomic cardiovascular regulation in otherwise healthy subjects. Hypertension 2005;46:1201-06.

17Karlijn C. VollebregtHans WolfKees BoerMarcel F. van der WalTanja G.M. VrijkotteGouke J. Bonsel

Acta Obstetricia et Gynecologica Scandinavica 2010; 89: 261–267

Does physical activity in leisure time early in pregnancy reduce the incidence of preeclampsia or gestational hypertension?

118

ABSTRACT

ObjectiveAssessment of the association of physical activity in leisure time with preeclampsia and

gestational hypertension in the first half of pregnancy in nulliparous women.

Design

Population based prospective cohort study.

Setting

Amsterdam, the Netherlands

Population

All Amsterdam pregnant women between January 2003 and March 2004 who were

nulliparous with a singleton pregnancy and delivered after 24 weeks

Design

At their first prenatal care visit women were invited to fill out a questionnaire with

sociodemographic and psychosocial variables. Physical activity in leisure time in the

past week was measured using questions about walking, cycling, doing sports and other

activities in leisure time. The amount of minutes and intensity of each activity was studied

using four categories; no, low, moderate or high activity. The 50th and 90th percentiles

of the scores of the whole group was used to create these categories. Data from the

National Obstetric Register were used for pregnancy outcome. By using multivariate

logistic regression we adjusted for sociodemographic and medical confounders.

Main outcome

Incidence of preeclampsia and gestational hypertension

Results

A total of 12.377 women were invited with a response of 67%. 3679 nulliparous women

were included. The incidence of preeclampsia and gestational hypertension was 3.5%

and 4.4% respectively. The amount of time or intensity of physical activity in leisure time

was not associated with a difference in risk of preeclampsia or gestational hypertension.

Conclusion

Physical activity in leisure time early in pregnancy does not reduce the incidence of

preeclampsia or gestational hypertension in unselected nulliparous women.

Chapter 7

119

Preeclampsia and physical exercise

INTRODUCTION

Hypertensive disorders of pregnancy are a leading cause of maternal and perinatal

morbidity and mortality worldwide.1 The origin is thought to be associated with deficient

placentation early in pregnancy, which in turn may be caused by immunological, genetic

or environmental factors.2 The expression of the disease is mediated through endothelial

dysfunction and increased systemic resistance, resulting in hypertension with a highly

variable degree of organ dysfunction of which proteinuria is the most conspicuous. So

far no effective treatment is available other than antihypertensive drugs and termination

of pregnancy.

Hypertensive disorders in pregnancy and adult cardiovascular diseases share

predisposing conditions, like elevated blood pressure and obesity. Protective or disease

modifying factors of cardiovascular disease might also be applicable for the prevention

of hypertensive disorders in pregnancy. Physical activity is one such factor, known to

prevent cardiovascular disease and to reduce its symptoms. Its precise mechanism is

unclear, but an acceptable hypothesis is that it mediates a reduction of oxidative stress,

which in turn influences the immune response and improves endothelial dysfunction.3-6

These processes are also involved in the development of preeclampsia and gestational

hypertension.2 Because physical activity is, in part, an external factor, which can be

mediated by intervention, its association with preeclampsia and gestational hypertension

is vital for the development of non-medical prevention strategies. A number of studies

explored the association, but most studies were hampered by the low incidence of

preeclampsia, and overall results appear inconsistent.7-11

This study investigates the association between physical activity in leisure time early

in pregnancy and the development of hypertensive disorders later in pregnancy in an

unselected prospective cohort of nulliparous women.

METHODS

The present study is part of the Amsterdam Born Children and their Development study

(ABCD-study).12 This prospective community-based study examined the relationship

between various lifestyles during pregnancy and pregnancy outcome in a multicultural

population. Between January 2003 and March 2004 12,381 Amsterdam women who

attended antenatal care for their first visit were approached by their obstetric caregiver

to participate in the study. Within two weeks a questionnaire (pregnancy questionnaire)

on sociodemographic characteristics and lifestyle was sent to the pregnant women to

120

be filled out at home and returned by prepaid mail. Women who gave permission to

follow-up received a second questionnaire (infant questionnaire) 3 months after birth

concerning pregnancy outcome, lifestyles, and health of the baby. Additional data like

pregnancy outcome, infant gender, birth weight and gestational age were obtained from

the national obstetric register and the Youth Health Care of the Municipal Health Service

in Amsterdam.

Study population

All nulliparous women with singleton pregnancy, who filled out the pregnancy questionnaire

before 24 weeks gestational age, who delivered after 24 weeks and gave permission

for follow-up were included. The study protocol was approved by the medical ethical

committees of all Amsterdam hospitals and the Municipal Privacy Protection Committee

of Amsterdam. Participation and permission for data retrieval rested on written consent.

Measurements

Gestational hypertension was defined as a diastolic pressure ≥ 90 mmHg after 20 weeks

in a previously normotensive woman. Preeclampsia was defined by the combination

of gestational hypertension and proteinuria ≥ 0,3 g/ 24 h or dipstick ≥ ++ after 20

weeks gestation, according to the International Society for the Study of Hypertension in

Pregnancy (ISSHP) guidelines.13 Superimposed preeclampsia was defined as de novo

proteinuria ≥ 0,3 g/ 24 h or dipstick ≥ ++ after 20 weeks gestation together with a

diastolic pressure ≥ 90 mmHg in women with chronic hypertension. Chronic hypertension

was defined as a systolic pressure ≥140 mmHg and/or diastolic pressure ≥ 90 mmHg

or the necessity for antihypertensive treatment before pregnancy or before 20 weeks

gestational age. Women with chronic hypertension who had a diastolic pressure ≥ 90

mmHg after 20 weeks gestation were included in the gestational hypertension group.

The medical files of all women who documented elevated blood pressure in the infant

questionnaire or who had a diastolic blood pressure ≥ 90 mmHg and/ or proteinuria

in The National Obstetric Register were reviewed for confirmation of the diagnosis

preeclampsia or gestational hypertension. If gestational hypertension was diagnosed first

and proteinuria occurred later in pregnancy, preeclampsia was the assigned diagnosis.

From the remaining women a random sample of 190 women was selected to validate the

absence of hypertensive disorders during pregnancy by reviewing their medical records.

Physical activity in leisure time in the past week was questioned using four questions

about walking, cycling, doing sports and other activities like ‘do it yourself’ or gardening

in leisure time. The amount of time spent in minutes and the intensity (low, moderate or

vigorous) was asked. We analyzed the amount of time spent on each activity and the total

time spent on physical activity in leisure time. The scores were trichotomized using the

Chapter 7

121


50th and 90th percentiles of the amount of minutes spent on activity of the total group to

create four categories. The quantity of minutes spent on each activity was multiplied with

the intensity (arbitrarily we coded low by 1, moderate by 2 and vigorous by 3), yielding

a numerical score for each activity question, which were summated into a total score.

These scores were trichotomized using the 50th and 90th percentiles of levels of intensity

of activity of the total group to create four categories;no activity, low activity, moderate

activity and high activity.

Characteristics and covariates

Maternal age, gestational age (based on routine ultrasound examination or, if

unavailable, on the date of the last menstrual period), birth weight and infant gender

were obtained from the Youth Health Care registration of the Municipal Health Service

in Amsterdam. All other variables were self reported and defined as follows: maternal

education (years of education after primary school; 3 categories: < 5 years; 5-10 years;

> 10 years), marital status (married/cohabiting or single), pre-gravid maternal weight

and height, previous miscarriage/ abortion (yes or no), smoking during pregnancy (yes

or no). Women were asked if they had hypertension, diabetes mellitus or any other

disease during their pregnancy, and if they used any medication. Drug utilization was

classified according to the International Classification for Primary Care.14 If women used

medication for a specific disease but did not report the disease (for example insulin) they

were classified as having the disease (diabetes). Pre-pregnancy body mass index (BMI)

was calculated. Missing data on pregravid maternal weight (about 5% compared to 1%

general missings) were formally imputed by models using maternal length and parity

data, retaining the variability of the data.

Statistical analyses

To compare groups one-way ANOVA with post-hoc Bonferoni analysis or Pearson’s

chi-square test was used. We estimated the effects of physical activity in the beginning

of pregnancy on the incidence of preeclampsia or gestational hypertension later in

pregnancy by predefined explorative logistic regression models, with increasing levels of

adjustment. First univariate analysis for each variable was used to calculate unadjusted

odds ratio’s. Subsequently multivariate logistic regression was performed to adjust for

medical covariates: BMI, chronic hypertension, diabetes mellitus, previous miscarriage/

abortion, smoking in pregnancy and haemorrhage in the first half of pregnancy (model

1). Those medical covariates that were statistically significant in this analysis were

combined with sociodemographic covariates in a second multivariate logistic analysis.

Sociodemographic covariates were age, ethnicity, education and marriage/ co-habitation

(model 2). The selection of covariates that influenced the model significantly was

122

performed by back step procedure with the p value for entry set at 0.05 and for removal

at 0.10. Women with no physical activity were taken as the reference category. The

Hosmer and Lemeshowtest was used to test the goodness of fit of the data. The level of

significance used was p<0.05. Statistical calculations were performed with SPSS 14.0.2(

SPSS Inc.Chicago, IL).

RESULTS

In all, 12,373 women were invited for the ABCD-study and 8266 women returned the

questionnaire (response rate 67%). Of these 3679 nulliparous women were included in

our study according to our selection criteria. The median gestational age at completing

the pregnancy questionnaire was 15.6 weeks (quartiles 14.0-17.3). From the 599

women, who reported elevated blood pressure in the questionnaire or who were coded

as such in the national obstetric register, the diagnosis preeclampsia was confirmed

in 128 women (3.5%) by reviewing the medical records and gestational hypertension

in 161 women(4.4%). Ten medical files could not be retrieved and 300 women had

neither preeclampsia nor gestational hypertension. These women were all allocated to

the control group. From the 190 women, who were preselected in the control group

for confirmation by reviewing medical files, 187 files could be assessed. None of these

women had preeclampsia or gestational hypertension during their pregnancy and all

were correctly allocated to the control group. All women whose medical files could not

be retrieved, were all allocated to the control group. The selection of our study group

is described elsewhere using the same cohort for studying the influence of psychosocial

stress on preeclampsia and gestational hypertension.15

From 80 women (2%), information of physical activity was missing. Demographic and

medical characteristics according to degree of physical exercise are presented in Table

1. Women with no or low leisure time physical activities were significantly younger (p <

0.001), less often Caucasian (p < 0.001), had a higher BMI (p = 0.005) and a lower

education (p < 0.001). Women with no physical activity also smoked more often (p <

0.001) and lived more often single (p < 0.001). They delivered earlier and birthweight

was lower compared tomothers who performed any exercise (p < 0.001). Low exercise

level was associated with shorter gestational length and lower birthweights compared to

moderate but not high exercise (p < 0.05). Women with moderate or high total leisure

time physical activities were similar for these parameters.

The incidence of preeclampsia and gestational hypertension did not differ between the

groups with different total physical activities in leisure time based on time and intensity. After

Chapter 7

123


adjustment for possible confounders, walking, bicycling, playing sports, other activities

or performing any physical activity in leisure time (minutes/week) were not associated

with a reduction of preeclampsia (Table 2) or gestational hypertension (Table 3). When

women spent more than 585 minutes on physical activities, the adjusted odds ratio

was 0.62 (95%CI (confidence interval) 0.25–1.53) for preeclampsia and 0.54 (95%CI

0.22–1.36) for gestational hypertension. Likewise, playing sports or total leisure time

physical activities at a high level lacked a significant association with preeclampsia (OR

0.43; 95%CI 0.17–1.10) and gestational hypertension (OR 0.78; 95% CI 0.36–1.69).

Table 1. Sociodemographic and medical characteristics of included women according to level and time spend on physical exercise in leisure time in early pregnancy.

Total group No exercise low moderate high

N (%) 3679 (100) 454 (12.3) 1502 (40.1) 1322 (35.9) 321 (8.7)Age (years) 29.9 ± 5.1 27.4 ± 5.4 29.4 ± 5.3 31.1 ± 4.4 31.6 ± 4.2

Etnicity N(%)

-Caucasian 2461 (66.9) 167 (36.9) 945 (63.0) 1055 (79.8) 263 (81.9)

-Black 157 (4.3) 52 (11.5) 74 (4.9) 22 (1.7) 3 (0.9)

-Turkish/Moroccan 347 (9.4) 93 (20.5) 174 (11.6) 54 (4.1) 9 (2.8)

- Other 712 (19.4) 141 (31.1) 308 (20.5) 191 (14.4) 46 (14.3)

Education N (%)

-0-5 years 557 (15.2) 123 (27.4) 261 (17.5) 120 (9.1) 21 (6.6)

-6-10 years 1445 (39.6) 239 (53.2) 638 (42.8) 442 (33.6) 99 (31.0)

->10 years 1651 (45.2) 87 (19.4) 593 (39.7) 754 (57.3) 199 (62.4)

Marriage/ co-habitation N (%)

No 497 (13.5) 99 (21.8) 197 (13.1) 145 (11.0) 50 (15.6)

BMI (kg/m2) 22.6 ± 3.7 23.3 ± 4.7 22.7 ± 3.8 22.3 ± 3.2 21.8± 2.7

Chronic hypertension N (%)

Yes 92 (2.5) 16 (3.5) 37 (2.5) 31 (2.4) 4 (1.2)

Diabetes N (%)

Yes 19 (0.5) 3 (0.7) 10 (0.7) 4 (0.3) 2 (0.6)

Smoking in pregnancy N (%)

Yes 343 (9.3) 76 (16.7) 141 (9.4) 98 (7.4) 21 (6.5)

Miscarriage/ abortion N (%)

Yes 969 (26.3) 124 (27.3) 370 (24.6) 360 (27.2) 96 (29.9)

Hemorrhage N (%)

Yes 431 (11.7) 59 (13.2) 175 (11.8) 153 (11.6) 32 (10.0)

Preeclampsia N (%) 127 (3.5) 20 (4.3) 49 (3.2) 51(3.8) 7 (2.2)

Gestational hypertension N(%) 159 (4.3) 19 (4.1) 77 (5.0) 49 (3.7) 14 (4.3)

Birth weight (g) 3381 ± 557 3229 ± 610 3370 ± 550 3432 ± 543 3346 ± 520Gestational age (days) 278 ± 14 276 ± 16 278 ± 14 280 ± 13 279 ± 14

Data are expressed as mean ± SD of N (%). BMI: body mass index

124

DISCUSSION

We observed no significant association between physical activity in leisure time in

early pregnancy and the incidence of preeclampsia and gestational hypertension after

appropriate adjustment of covariates. Literature regarding this subject is conflicting. Studies

use different methods and definitions which makes it difficult to compare their results.16

Some studies investigate in particular physical activity at work with or without account for

physical activity in leisure time. Most of the studies are retrospective case-control studies.

In three retrospective case control studies with data collection by questionnaire after

delivery it was observed that leisure time physical activity in the first half of pregnancy was

more frequent in controls than in women with preeclampsia.7;8 More vigorous activity

was associated with a lower incidence of preeclampsia.17 One retrospective cohort study

of enlisted active duty military women observed less preeclampsia in nulliparous women

Table 2. The association between preeclampsia and sports and total leisure time physical activity in minutes/week or based on intensity and time, calculated by univariate analysis, by multivariate analysis using model 1 and model 2, expressed as odds ratios with 95% confidence limits.

Controls(%) PE(%) Univariate effects Multivariate effects model 1


Sports (minutes/week)0 69.8 1 1 1 1

1-70 15.1 16.4 1.09 (0.67–1.77) 1.18 (0.72–1.95) 1.02 (0.61–1.71)

71-180 12.5 10.9 0.88 (0.50–1.56) 0.97 (0.54–1.75) 0.81 (0.45–1.48)

>180 2.6 3.1 1.19 (0.42–3.32) 1.32 (0.46–3.76) 1.04 (0.36–3.00)

Total LTPA ( minutes/week)

0 12.5 15.9 1 1 1

1-235 48.9 48.4 0.78(0.47-1.31) 0.92(0.53-1.60) 0.71(0.40-1.27)

236-585 31.4 29.4 0.74(0.42-1.28) 0.90(0.50-1.64) 0.63(0.33-1.18)

>585 7.2 6.3 0.70(0.30-1.61) 0.88(0.37-2.11) 0.62(0.25-1.53)

Sports

No sports 69.8 69.5 1 1 1

Low 15.2 14.9 0.98 (0.59–1.62) 1.06 (0.63–1.77) 0.91 (0.54–1.55)

Moderate 12.0 13.3 1.11 (0.66–1.89) 1.22 (0.75–2.10) 1.02 (0.58–1.78)

Vigorous 3.0 2.3 0.80 (0.25–2.56) 0.93 (0.29–3.03) 0.72 (0.22–2.37)

Total LTPA

No LTPA 12.5 15.9 1 1 1

Low 41.6 38.1 0.72 (0.42–1.23) 0.85 (0.48–1.51) 0.68 (0.38–1.32)

Moderate 36.9 40.5 0.87 (0.51–1.47) 1.04 (0.59–1.85) 0.72 (0.39–1.32)Vigorous 9.0 5.6 0.48 (0.20–1.16) 0.62 (0.25–1.53) 0.43 (0.17–1.10)

Model 1: BMI, chronic hypertension, diabetes mellitus, previous miscarriage/abortion, smoking in pregnancy and hemorrhage in the first half of pregnancy. Model 2: age, ethnicity, education and marriage/co-habitation. PE, preeclampsia; LTPA, leisure time physical activity; BMI, body mass index

Chapter 7

125


employed in jobs involving high levels of physical activity compared to nulliparous working

at low levels of physical activity.18 Contrary to these studies one other retrospective case-

control study of nulliparous women observed that women with clerical work experienced

less frequently preeclampsia than women with moderate to high physical activity at work,

after adjustment for age, BMI, smoking and social status.19

Three prospective cohort studies with assessment of physical activity in leisure time

early in pregnancy also showed different results.9-11 Saftlas et al. observed a non-

significant reduction of the incidence of preeclampsia. The extent of the risk reduction

was comparable to our findings. Magnus et al. found a significant reduction when there

were ≥ 25 physical activities per month. Osterdal et al. studied the largest cohort and

investigated time, intensity and metabolic equivalents. Although this study did not observe

an association between leisure time activity and the incidence of preeclampsia, those

women who spend ≥ 270 minutes on physical activity per week had an odds ratio of 1.7

Table 3. The association between gestational hypertension and sports and total leisure time physical activity in minutes/week or based on intensity and time, calculated by univariate analysis, by multivariate analysis using model 1 and model 2, expressed as odds ratios with 95% confidence limits.

Controls(%) GH(%) Univariate effects Multivariate effects model 1


Sports (minutes/week)0 69.8 69.4 1 1 1

1-70 15.1 12.5 0.83(0.51-1.35) 0.91(0.55-1.49) 0.77(0.46-1.28)

71-180 12.5 15.6 1.26(0.81-1.97) 1.40(0.89-2.22) 1.18(0.74-1.88)

>180 2.6 2.5 0.96(0.35-2.65) 0.86(0.27-2.80) 0.71(0.22-2.32)

Total LTPA ( minutes/week)

0 12.5 12.0 1 1 1

1-235 48.9 55.7 1.19(0.72-1.97) 1.27(0.75-2.15) 1.01(0.58-1.75)

236-585 31.4 27.2 0.90(0.52-1.57) 1.00(0.56-1.77) 0.71(0.39-1.30)

>585 7.2 5.1 0.73(0.32-1.70) 0.73(0.30-1.79) 0.54(0.22-1.36)

Sports

No sports 69.8 69.4 1 1 1

Low 15.2 14.4 0.95 (0.60–1.50) 1.03(0.64–1.65) 0.87(0.54–1.41)

Moderate 12.0 11.3 0.94 (0.57–1.57) 1.06(0.63–1.79) 0.90(0.53–1.52)

Vigorous 3.0 5.0 1.70 (0.81–3.59) 1.75(0.79–3.90) 1.43(0.64–3.20)

Total LTPA

No LTPA 12.5 12.0 1 1 1

Low 41.6 48.1 1.21 (0.72–2.02) 1.26 (0.74–2.16) 1.02 (0.58–1.78)

Moderate 36.9 31.0 0.88 (0.51–1.51) 0.97 (0.56–1.71) 0.69 (0.38–1.26)Vigorous 9.0 8.9 1.02 (0.50–2.07) 1.12 (0.53–2.34) 0.78 (0.36–1.69)

Model 1: BMI, chronic hypertension, diabetes mellitus, previous miscarriage/abortion, smoking in pregnancy and hemorrhage in the first half of pregnancy. Model 2: age, ethnicity, education and marriage/co-habitation. GH,gestational hypertension; LTPA, leisure time physical activity; BMI, body mass index

126

for severe preeclampsia. The precise definition of severe preeclampsia was not reported.

In our study physical activity levels in women with preeclampsia < 34 weeks were not

different from women with preeclampsia ³ 34 weeks.

A recent Cochrane review could identify two trials (total n=45) randomising women

to participate in an exercise program or not.20 These studies were too small to draw

conclusions regarding the prevention of preeclampsia or gestational hypertension, as

only one case of preeclampsia was encountered. One of the two identified studies

observed that diastolic blood pressure was lower in the exercise group, which supports a

positive effect.21 A literature search using the terms used in this Cochrane review detected

one new study. This study had not enough statistical power to detect a difference in the

incidence of preeclampsia. 22

In practice the low incidence of preeclampsia precludes randomised studies using this

endpoint. Case control studies with retrospective design are prone to recall bias as the

data have to be collected after delivery. Prospective cohort studies, like the present one,

beginning in early pregnancy have the advantage that recall bias is minimised, but the

inclusion of a large number of women is needed. But lack of power, especially in groups

with high physical activity, is a serious threat even of fairly large cohort studies, due to the

low prevalence of (severe) preeclampsia and low frequency of intense physical activity.

A potential bias in case-control and cohort studies is that women who exercise probably

differ from women who do not. In our study 70 % did not practice any kind of sports while

13 % did not practice any physical activity at al. In the Danish cohort of Odenthal 63

% of the women did no sportive activity at all. We adjusted our analysis with a number

of covariates that have been well described by others to be associated with gestational

hypertension or preeclampsia.2;23 However, certainty that adjustment was adequate can

never be obtained.

Assessment of physical activity is complicated because there is no ‘gold standard’ and

unequivocal norm. Self report may suffer from recall bias. Therefore questionnaires focus

on discrete activities like sport and activity at work which are easy to recall, and less

subject to interpretation as ‘physical activity’. Activity monitors are more objective and

could reduce report errors but these are not suitable for prospective studies with a large

sample size like our study.24 Hence our approach was questionnaire-based with self-

reporting. Existing questionnaires were unacceptably long. Therefore we used our own

questionnaire. If anything, our measurement of leisure time may have been subject to

random error, although the current results most likely will not alter with more extensive

measurement.

Even if preventive physical programs are considered, practical constraints can be

expected. It is generally assumed that disturbed placental development in early pregnancy

Chapter 7

127


is an important factor in the origin of preeclampsia and gestational hypertension.2 If the

promotion of physical activity can reduce the incidence of preeclampsia and gestational

hypertension than the first half of pregnancy is likely the most effective period. At this early

stage of pregnancy women are often easily tired and generally abstain from high activity.

CONCLUSION

Physical exercise in leisure time in early pregnancy is not associated with reduced

incidence of preeclampsia and gestational hypertension in an unselected nulliparous

population. An effect of high physical activity can not be excluded because of the low

incidence of high physical activity and the low incidence of preeclampsia.

128


death: a systematic review. Lancet 2006; 367(9516):1066-1074. (2) Sibai B, Dekker G, Kupferminc M. Pre-eclampsia. Lancet 2005; 365(9461):785-799. (3) Mora S, Cook N, Buring JE, Ridker PM, Lee IM. Physical activity and reduced risk of cardiovascular

events: potential mediating mechanisms. Circulation 2007; 116(19):2110-2118. (4) Dickinson HO, Mason JM, Nicolson DJ, Campbell F, Beyer FR, Cook JV et al. Lifestyle

interventions to reduce raised blood pressure: a systematic review of randomized controlled trials. J Hypertens 2006; 24(2):215-233.

(5) Kojda G, Hambrecht R. Molecular mechanisms of vascular adaptations to exercise. Physical activity as an effective antioxidant therapy? Cardiovasc Res 2005; 67(2):187-197.

(6) Moyna NM, Thompson PD. The effect of physical activity on endothelial function in man. Acta Physiol Scand 2004; 180(2):113-123.

(7) Marcoux S, Brisson J, Fabia J. The effect of leisure time physical activity on the risk of pre-eclampsia and gestational hypertension. J Epidemiol Community Health 1989; 43(2):147-152.

(8) Sorensen TK, Williams MA, Lee IM, Dashow EE, Thompson ML, Luthy DA. Recreational physical activity during pregnancy and risk of preeclampsia. Hypertension 2003; 41(6):1273-1280.

(9) Saftlas AF, Logsden-Sackett N, Wang W, Woolson R, Bracken MB. Work, leisure-time physical activity, and risk of preeclampsia and gestational hypertension. Am J Epidemiol 2004; 160(8):758-765.

(10) Magnus P, Trogstad L, Owe KM, Olsen SF, Nystad W. Recreational physical activity and the risk of preeclampsia: a prospective cohort of Norwegian women. Am J Epidemiol 2008; 168(8):952-957.

(11) Osterdal ML, Strom M, Klemmensen AK, Knudsen VK, Juhl M, Halldorsson TI et al. Does leisure time physical activity in early pregnancy protect against pre-eclampsia? Prospective cohort in Danish women. BJOG 2009; 116(1):98-107.

(12) van Eijsden M, van der Wal MF, Bonsel GJ. Folic acid knowledge and use in a multi-ethnic pregnancy cohort: the role of language proficiency. BJOG 2006; 113(12):1446-1451.

(13) Brown MA, Lindheimer MD, de SM, Van AA, Moutquin JM. The classification and diagnosis of the hypertensive disorders of pregnancy: statement from the International Society for the Study of Hypertension in Pregnancy (ISSHP). Hypertens Pregnancy 2001; 20(1):IX-XIV.

(14) Hofmans-Okkes IM, Lamberts H. The International Classification of Primary Care (ICPC): new applications in research and computer-based patient records in family practice. Fam Pract 1996; 13(3):294-302.

(15) Vollebregt KC, van der Wal MF, Wolf H, Vrijkotte TG, Boer K, Bonsel GJ. Is psychosocial stress in first ongoing pregnancies associated with pre-eclampsia and gestational hypertension? BJOG 2008; 115(5):607-615.

(16) Bonzini M, Coggon D, Palmer KT. Risk of prematurity, low birthweight and pre-eclampsia in relation to working hours and physical activities: a systematic review. Occup Environ Med 2007; 64(4):228-243.

(17) Rudra CB, Williams MA, Lee IM, Miller RS, Sorensen TK. Perceived exertion during prepregnancy physical activity and preeclampsia risk. Med Sci Sports Exerc 2005; 37(11):1836-1841.

(18) Irwin DE, Savitz DA, St Andre KA, Hertz-Picciotto I. Study of occupational risk factors for pregnancy-induced hypertension among active duty enlisted Navy personnel. Am J Ind Med 1994; 25(3):349-359.

(19) Spinillo A, Capuzzo E, Colonna L, Piazzi G, Nicola S, Baltaro F. The effect of work activity in pregnancy on the risk of severe preeclampsia. Aust N Z J Obstet Gynaecol 1995; 35(4):380-385.

Chapter 7

129


(20) Meher S, Duley L. Exercise or other physical activity for preventing pre-eclampsia and its complications. Cochrane Database Syst Rev 2006;(2):CD005942.

(21) Yeo S, Steele NM, Chang MC, Leclaire SM, Ronis DL, Hayashi R. Effect of exercise on blood pressure in pregnant women with a high risk of gestational hypertensive disorders. J Reprod Med 2000; 45(4):293-298.

(22) Yeo S, Davidge S, Ronis DL, Antonakos CL, Hayashi R, O’Leary S. A comparison of walking versus stretching exercises to reduce the incidence of preeclampsia: a randomized clinical trial. Hypertens Pregnancy 2008; 27(2):113-130.

(23) Villar J, Carroli G, Wojdyla D, Abalos E, Giordano D, Ba’aqeel H et al. Preeclampsia, gestational hypertension and intrauterine growth restriction, related or independent conditions? Am J Obstet Gynecol 2006; 194(4):921-931.

(24) Janz KF. Physical activity in epidemiology: moving from questionnaire to objective measurement. Br J Sports Med 2006; 40(3):191-192.

18Karlijn C. VollebregtLobke SeesingSaskia RangKees BoerHans Wolf

Hypertension in pregnancy 2006.25;3:159-167

Sensitivity of spontaneous baroreflex conrol of the heart and hemodynamic parameters are not influenced by the menstrual cycle

132

ABSTRACT

ObjectiveThe purpose of this study was to evaluate if hemodynamic parameters and sympathetic

activity vary between the follicular and luteal phase of the menstrual cycle before using

sympathetic activity in pre-pregnancy risk assessment for preeclampsia.

Methods

We studied 39 healthy women at days 5 to 10 and days 18 to 25 of the menstrual cycle.

Blood pressure, heart rate, cardiac output and total peripheral resistance were measured

continuously using noninvasive finger arterial pressure waveform registration (Portapres

Model 2, TNO-BMI, The Netherlands). Baroreflex sensitivity (BRS) and sympathetic

activity by phase angle difference were studied using spectral analysis and xBRS.

Results

There were no differences in hemodynamic parameters, BRS or phase angle difference.

Conclusion

There is no difference in blood pressure, BRS or sympathetic activity between the first and

second half of the menstrual period. We recommend using the first half of the cycle to be

certain no pregnancy exist, as the influence of very early pregnancy is unknown.

Chapter 8

133

Hem

odynamics and the m

enstrual cycle

INTRODUCTION

Preeclampsia is a leading cause of maternal and perinatal morbidity and mortality.1 It is

a multisystem disorder with unknown etiology and is characterized by increased systemic

vascular resistance, endothelial dysfunction and increased sympathetic activity.1;2 Despite

the advances in perinatal care, there is no test for adequate prediction of the disease.3

Early selection of women at high risk could offer the opportunity to target care at those

most likely to benefit and to evaluate or design preventive strategies more effectively.

Earlier research has shown that, before pregnancy, women who will develop preeclampsia

have a slightly higher blood pressure and a larger blood pressure drop after orthostatic

stress when compared with women who have a normal blood pressure throughout

pregnancy.4 From 8 weeks gestational age onward they also had a larger negative phase

difference in the low frequency band with spectral analysis. This indicates increased

sympathetic activity early in pregnancy among women who will develop preeclampsia

later in pregnancy. Recently sympathetic hyperactivity and reduced baroreflex sensitivity

were found in formerly preeclamptic normotensive women.5 These women were formerly

preeclamptic and the sympathetic hyperactivity could be the result of the disease or it

could have been present already before pregnancy. These findings could thus be useful

for early identification for women at risk early in pregnancy and even before pregnancy.

It is well known that ovarian hormones affect the cardiovascular and renal-aldosteron

system.6-8 However, the magnitude of the effect of hormonal fluctuations during the

menstrual cycle on blood pressure regulation is not clear and literature in this respect is

not consistent.9-11 The present study was designed to ascertain if it would be necessary

to account for the menstrual cycle when measuring pre-pregnancy hemodynamic

parameters for the identification of women at risk for preeclampsia.

METHODS

Study PopulationWe recruited participants by advertisement in a local newspaper. All women were over

the age of 18 years, had a regular cycle (median 28 days; range 26 to 32 days), and

had no previous pregnancy. The participants did not use medication, oral contraceptives

or a levonorgestrel releasing intrauterine device. We excluded women with hypertension,

diabetes or cardiovascular diseases. After providing written informed consent, all

participants underwent identical study protocols. The Medical Ethical Committee of the

Academic Medical Center approved the study.

134

Study Protocol

All women were studied in the follicular phase, between the 5th and 10th day, and

in the luteal phase, between the 18th and 25th day of the menstrual cycle. Menstrual

cycle phase was determined according to the first day of the last menstruation. The first

measurement was allocated randomly to the follicular phase (FP) or the luteal phase (LP).

All measurements were doe by two researchers. The participants were advised to abstain

from drinking coffee or smoking cigarettes from the night before the measurements. After

resting in supine position for at least 10 minutes, they were informed about the study

procedure and were allowed to adapt to the measuring devices and the room before

the measurements started. The measurements were preformed with the participants

in supine position under standardized conditions and took place in a quiet room

with a temperature between 20 and 22°C. Blood pressure was measured twice with

conventional sphygmomanometry (Maxi Stabil 3, an aneroid sphygmomanometer, Welch

Allyn, Skaneateles Falls, New York, USA) before the measurement with the Portapres™

started. The mean value was used for comparisons.

Continuous recordings of hemodynamic parameters

Non-invasive finger arterial pressure waveform registration by Portapres, Model 2 (TNO-

BMI, Amsterdam, The Netherlands) was used for continuous monitoring of blood pressure,

heart rate and cardiac output. Portapres is a device for the measurement of finger arterial

pressure on a beat-to-beat basis, according to the volume clamp method of Penaz12;13.

The use of continuous recordings of finger arterial blood pressure by this method has

been validated in comparison to invasive beat-to-beat blood pressure recording in non-

pregnant subjects and in comparison to conventional sphygmomanometry in pregnant

women.14;15 Stroke volume analysis out of continuous finger arterial pressure waveform

registration by the pulse contour method of Wesseling (the Modelflow method) has been

validated extensively.16 An appropriate size finger cuff was applied at the middle finger

of the non-dominate hand. At a stable signal the pressure registration was calibrated to

upper-arm cuff pressure by the Return to Flow method.17 Data collection was started after

a stable signal had been reached for 5 minutes.

Data analysis

Data from the Portapres were sampled at 200 Hz and analyzed by the Beatfast program

(TNO-BMI, Amsterdam, The Netherlands). The average of the systolic blood pressure

(SBP), diastolic blood pressure (DBP), mean arterial pressure (MAP), heart rate (HR),

cardiac output (CO) and total peripheral resistance (TPR) of 10 minutes during supine

rest were used for analysis.

Chapter 8

135

Hem

odynamics and the m

enstrual cycle

Baroreflex sensitivity (BRS)

Beat to beat SBP and interbeat interval (IBI) time series were detrended and Hanning

windowed. Power spectral density and cross-spectra of SBP and IBI in the low-frequency

band (0.06-0.15 Hz) and the high-frequency band (0.15-0.35 Hz) were computed using

discrete Fourier transform as described elsewhere.18 The cross-correlation method xBRS

(BMEYE, Amsterdam, the Netherlands) was used for time-domain analysis of spontaneous

BRS 19 For this purpose the SDP and IBI time series were resampled at 1Hz. In a 10-s

window, the correlation and regression slope between SBP and IBI were computed. These

computations were done for 0 to 5 s delays in IBI with respect to SBP and the delay with

the highest positive coefficient of correlation was selected. The slope between SBP and IBI

was recorded as BRS estimate if the correlation was significant at p=0.01.

Statistics

Results were compared by paired t-test or Wilcoxon signed ranks test as appropriate. A

sample size of 35 participants allowed for the detection of a blood pressure difference

of 2 mmHg between the two measurement periods, assuming a mean blood pressure of

100 mmHg and a standard deviation of 10 mmHg with alpha 0.05 and beta 0.2.

RESULTS

Thirty-nine participants were recruited; all women were tested in both phases; 18 women

entered the study in the follicular phase and 21 women in the luteal phase. They had a

mean age of 29 ( ±4.7) years and a mean body mass index of 22.0 (± 2.3) kg/m2 at the

first measurement. Blood pressure measured by conventional sphygmomanometry and

Table 1. Comparison of hemodynamic parameters by phase of the menstrual cycleDay 5-10 Day18-25 P-value

Sphygmomanometry measurementsSBP (mmHg) 112 ±8 110± 9 0.21

DBP (mmHg) 73± 7 73 ±8 0.98

Portapres measurements

SBP (mmHg) 105 ± 11 105 ± 12 0.93

DBP (mmHg) 59 ± 7 60 ±8 0.56

MAP (mmHg) 78±9 79 ±10 0.74

HR (bpm) 69± 9 69 ±9 0.62

CO ( l/min) 5.9 ±1.4 6.2 ±1.9 0.38TPR (dyne.s-1.cm-5) 0.81±0.2 0.83±0.2 0.70

Data presented as mean ± standard deviation and P values calculated with paired t-test. SBP: systolic blood pressure, DBP: diastolic blood pressure, MAP: mean arterial pressure, HR: heart rate, CO: cardiac output, TPR: total peripheral resistance.

136

the parameters measured by the Portapres (SBP, DBP, HR, CO and TPR) were comparable

between the two phases of the menstrual cycle (table 1). SBP and DBP measured by

Portapres were significantly lower than SBP and DBP measured by sphygmomanometry

in both periods (p<0.01).

There was no difference of BRS and phase angle difference measured with spectral

analysis and xBRS between the two periods (figures 1). None of the measurements were

influenced by age or body mass index.

Figure 1. Baroreflex sensitivity measured with spectral analysis at low frequency (A), with xBRS (B) and phase angle difference with spectral analysis (C) at day 5-10 and day 18-25 of the menstrua; cycle. Data are displayed using box whisker plots. The box indicates the lower and upper quartiles and the central line is the median. The horizontal lines are the 2.5th and 97.5th centiles. There were no significant differences between both phases.

DISCUSSION

To our knowledge this is the first study of the baroreflex during the menstrual period

with an adequate sample size. We observed no differences in blood pressure and heart

rate between days 5 and 10 and days 18 and 25 of the menstrual cycle, which agrees

with previous studies.9;11;20-22 Our observation that hemodynamic parameters, BRS and

sympathetic activity were comparable in the first and second half of the menstrual cycle

agrees with a number of studies. 9;23;24 However, Tanaka and co-workers observed

greater baroreflex sensitivity in the early follicular phase compared to the midluteal phase

in response to a decrease in arterial blood pressure by nitroprusside, using tonometry for

blood pressure registration.10 Minson and co-workers performed a similar experiment

and did not detect such a difference, using finger pulse registration for blood pressure

calculation (Finapres), which is similar to our method.9 Additionally, this study reported

a higher sympathetic outflow of muscle sympathetic nerve activity(MSNA), determined by

microneurography and an increased sympathetic baroreflex sensitivity based on MSNA/

diastolic blood pressure slopes in the midluteal phase.9 They concluded that hormonal

Chapter 8

137

Hem

odynamics and the m

enstrual cycle

fluctuations during the normal menstrual cycle may alter sympathetic outflow but vascular

resistance is not increased. They hypothesized that the increase of sympathetic activity is

balanced by an increase of the vasodilator nitric oxide. A similar interpretation of data

was made in a longitudinal study before and during pregnancy in women who developed

preeclampsia.4

Sato and co-workers demonstrated a higher low frequency/ high frequency ratio of heart

rate variability with spectral analysis in the luteal phase.25 They suggest that sympathetic

nervous activity is predominant in the luteal phase as compared to the follicular phase.

This was not confirmed by others, using spectral analysis of heart rate variability.26;27

Blood pressure measurements with the Portapres were significantly lower than by

sphygmomanometry. The differences in DBP can be explained because measurements

with the Portapres with Return to Flow method intends to reconstruct brachial intra-arterial

pressure in a non-invasive way,17 whereas sphygmomanometry overestimate brachial

intra-arterial diastolic blood pressure.28 Sphygmomanometry always involves a certain

amount of stress due to the painfulness of cuff inflation and blood pressure rises slightly

during measurements with an upper-arm cuff.29 However, with a continuous method like

the Portapres the subject is completely at rest after a period of accustomization.

No hormone levels were measured because subjects were only selected if they had

a regular menstrual cycle between 26 and 32 days for several months. Ninety-eight

percent of women with a regular cycle have an ovulation with changes in estrogen and

progesterone plasma concentration.30 It is unlikely that our observation of similarity

between follicular and luteal phase could have been influenced by one or two subjects

without ovulation.

We chose a wide range (days 5 to10 and days 18 to 25) for the examination day. By

making this selection it might be possible that effects caused by the periovulatory estrogen

peak were not detected. However the differences in estrogen and progesterone levels

between days 5 and 10 and days 18 and 25 are of such a magnitude that significant

effects, if present, should have been detected in our study.31

Because we intend to perform a longitudinal study of the change of cardiovascular

control from the non-pregnant state to early pregnancy in future research, no controlled

stimulation of the cardiovascular system by drug injections was used as these would

not be acceptable in early pregnancy. Spontaneous BRS and drugs-induced BRS both

reflect baroreflex physiology and are generally correlated.32 In this study we measured

spontaneous BRS by averaging data over a time period of 10 minutes to get a reliable

estimation.

We only measured BRS in rest and we did not use cardiovascular reflex tests like orthostatic

stress test, isometric handgrip test, Valsalva’s manoeuvre or deep breathing test. Previous

138

work by our group did not demonstrate differences with these tests between women, who

developed preeclampsia and women who had an uneventful pregnancy. Furthermore,

it appeared difficult to perform such tests in a standardized way in pregnant women.33

CONCLUSION

No significant differences in blood pressure and baroreflex were found between days 5

and 10 and days 18 and 25. This implicates that it is not necessary to take the phase of

the menstrual cycle into account in a study of cardiovascular changes between the non-

pregnant and the pregnant state. Nevertheless it might be sensible to select the first half

of the cycle to be certain that no pregnancy exists as the influence of very early pregnancy

is unknown.

Chapter 8

139

Hem

odynamics and the m

enstrual cycle

Reference List 1. Sibai B, Dekker G, Kupferminc M. Pre-eclampsia. Lancet 2005;365:785-99. 2. Schobel HP, Fischer T, Heuszer K, Geiger H, Schmieder RE. Preeclampsia -- a state of sympathetic

overactivity. N.Engl.J.Med. 1996;335:1480-85. 3. Conde-Agudelo A, Villar J, Lindheimer M. World Health Organization systematic review of

screening tests for preeclampsia. Obstet.Gynecol. 2004;104:1367-91. 4. Rang S, Wolf H, Montfrans GA, Karemaker JM. Serial assessment of cardiovascular control

shows early signs of developing pre-eclampsia. J.Hypertens. 2004;22:369-76. 5. Courtar DA, Spaanderman ME, Aardenburg R, Janssen BJ, Peeters LL. Low plasma volume

coincides with sympathetic hyperactivity and reduced baroreflex sensitivity in formerly preeclamptic patients. J.Soc.Gynecol.Investig. 2006;13:48-52.

6. Oelkers WK. Effects of estrogens and progestogens on the renin-aldosterone system and blood pressure. Steroids 1996;61:166-71.

7. Mendelsohn ME. Protective effects of estrogen on the cardiovascular system. Am.J.Cardiol. 2002;89:12E-7E.

8. Orshal JM, Khalil RA. Gender, sex hormones, and vascular tone. Am.J.Physiol Regul.Integr.Comp Physiol 2004;286:R233-R249.

9. Minson CT, Halliwill JR, Young TM, Joyner MJ. Influence of the menstrual cycle on sympathetic activity, baroreflex sensitivity, and vascular transduction in young women. Circulation 2000;101:862-68.

10. Tanaka M, Sato M, Umehara S, Nishikawa T. Influence of menstrual cycle on baroreflex control of heart rate: comparison with male volunteers. Am.J.Physiol Regul.Integr.Comp Physiol 2003;285:R1091-R1097.

11. Dubey RK, Oparil S, Imthurn B, Jackson EK. Sex hormones and hypertension. Cardiovasc.Res. 2002;53:688-708.

12. Imholz BP, Wieling W, van Montfrans GA, Wesseling KH. Fifteen years experience with finger arterial pressure monitoring: assessment of the technology. Cardiovasc.Res. 1998;38:605-16.

13. Penaz, J. Photoelectric measurement of blood pressure, volume and flow in the finger. Digest 10th Int Conf Med Biol Engp . 1973

14. Hehenkamp WJ, Rang S, van Goudoever J, Bos WJ, Wolf H, van der Post JA. Comparison of Portapres with standard sphygmomanometry in pregnancy. Hypertens.Pregnancy. 2002;21:65-76.

15. Imholz BP, Langewouters GJ, van Montfrans GA, Parati G, van GJ, Wesseling KH et al. Feasibility of ambulatory, continuous 24-hour finger arterial pressure recording. Hypertension 1993;21:65-73.

16. Jellema WT, Wesseling KH, Groeneveld AB, Stoutenbeek CP, Thijs LG, van Lieshout JJ. Continuous cardiac output in septic shock by simulating a model of the aortic input impedance: a comparison with bolus injection thermodilution. Anesthesiology 1999;90:1317-28.

17. Bos WJ, van GJ, van Montfrans GA, van den Meiracker AH, Wesseling KH. Reconstruction of brachial artery pressure from noninvasive finger pressure measurements. Circulation 1996;94:1870-75.

18. deBoer RW, Karemaker JM, Strackee J. Hemodynamic fluctuations and baroreflex sensitivity in humans: a beat-to-beat model. Am.J.Physiol 1987;253:H680-H689.

19. Westerhof BE, Gisolf J, Stok WJ, Wesseling KH, Karemaker JM. Time-domain cross-correlation baroreflex sensitivity: performance on the EUROBAVAR data set. J.Hypertens. 2004;22:1371-80.

20. Van Beek E., Houben AJ, van Es PN, Willekes C, Korten EC, De Leeuw PW et al. Peripheral haemodynamics and renal function in relation to the menstrual cycle. Clin.Sci.(Lond) 1996;91:163-68.

140

21. Giannattasio C, Failla M, Grappiolo A, Stella ML, Del BA, Colombo M et al. Fluctuations of radial artery distensibility throughout the menstrual cycle. Arterioscler.Thromb.Vasc.Biol. 1999;19:1925-29.

22. Williamson PM, Buddle ML, Brown MA, Whitworth JA. Ambulatory blood pressure monitoring (ABPM) in the normal menstrual cycle and in women using oral contraceptives. Comparison with conventional blood pressure measurement. Am.J.Hypertens. 1996;9:953-58.

23. Cooke WH, Ludwig DA, Hogg PS, Eckberg DL, Convertino VA. Does the menstrual cycle influence the sensitivity of vagally mediated baroreflexes? Clin.Sci.(Lond) 2002;102:639-44.

24. Meendering JR, Torgrimson BN, Houghton BL, Halliwill JR, Minson CT. Menstrual cycle and sex affect the hemodynamic responses to combined orthostatic and heat stress. Am.J.Physiol Heart Circ.Physiol 2005.

25. Sato N, Miyake S, Akatsu J, Kumashiro M. Power spectral analysis of heart rate variability in healthy young women during the normal menstrual cycle. Psychosom.Med. 1995;57:331-35.

26. Leicht AS, Hirning DA, Allen GD. Heart rate variability and endogenous sex hormones during the menstrual cycle in young women. Exp.Physiol 2003;88:441-46.

27. Saeki Y, Atogami F, Takahashi K, Yoshizawa T. Reflex control of autonomic function induced by posture change during the menstrual cycle. J.Auton.Nerv.Syst. 1997;66:69-74.

28. Ochiai H, Miyazaki N, Miyata T, Mitake A, Tochikubo O, Ishii M. Assessment of the accuracy of indirect blood pressure measurements. Jpn.Heart J. 1997;38:393-407.

29. van Montfrans GA, van der Hoeven GM, Karemaker JM, Wieling W, Dunning AJ. Accuracy of auscultatory blood pressure measurement with a long cuff. Br.Med.J.(Clin.Res.Ed) 1987;295:354-55.

30. Waller K, Swan SH, Windham GC, Fenster L, Elkin EP, Lasley BL. Use of urine biomarkers to evaluate menstrual function in healthy premenopausal women. Am.J.Epidemiol. 1998;147:1071-80.

31. Speroff L, Fritz M.A. Clinical Gynecologic Endocrinology and Infertility. Philadelphia, USA: 2005:187-232.

32. Parati G, Di RM, Castiglioni P, Bouhaddi M, Cerutti C, Cividjian A et al. Assessing the sensitivity of spontaneous baroreflex control of the heart: deeper insight into complex physiology. Hypertension 2004;43:e32-e34.

33. Rang S, Wolf H, Montfrans GA, Karemaker JM. Non-invasive assessment of autonomic cardiovascular control in normal human pregnancy and pregnancy- associated hypertensive disorders: a review. J.Hypertens. 2002;20:2111-19.

19 Summary, conclusion and recommendations

Chapter 9

145

Summ

ary, conclusion and recomm

endations

SUMMARY

Preeclampsia is a major cause of maternal and neonatal morbidity and mortality. The

precise underlying mechanisms are unknown and are probably related to vascular,

immunological, genetic and environmental factors.1 Despite extensive research the onset

of preeclampsia and disease progression remains unpredictable and there is no real cure

except inducing delivery. Although the clinical symptoms appear in the second half of

pregnancy the origin of the underlying pathophysiologic mechanisms is thought to occur

much earlier in pregnancy at the time of placentation.

Prevention of preeclampsia will have a huge impact on maternal health worldwide.

Because the origin of the disease is early in pregnancy development of methods for

early detection of the preclinical stage and thus prediction of preeclampsia is of major

importance. Though some promising tests seem to arise,2;3 up to now such predictive

test is not available.4

The work presented in this thesis focuses on cardiovascular differences in the beginning

of pregnancy between women with a healthy pregnancy and women who develop

preeclampsia later in pregnancy and in the modifiable risk factors psychosocial stress

and physical activity. Effective risk assessment enables targeting intensified obstetric care

at those who need this most and selection of high risk women for the assessment of

preventive treatment strategies.

This chapter summarizes the findings from this research and evaluates the results and

future research implications. While up to now most predictive tests were designed to

be performed in the second trimester we evaluated tests for the first trimester in the

expectation that any possible preventive treatment should be applied early in pregnancy

before the development of placental vascular supply is finished.5-7 Preeclampsia and

gestational hypertension were defined according to the guidelines of the International

Society for the Study of Hypertension in Pregnancy.8

SYSTEMATIC REVIEW

In chapter 2 we present a systematic review on the use of blood pressure as a predictive

test for preeclampsia. Medline, Embase, Cochrane library, Medion and reference lists

of included articles were searched from inception to February 2007. 34 studies were

included testing 60.599 women with 3341 cases of preeclampsia. Clinical heterogeneity

between studies was found for the method or device used for measuring blood pressure,

146

the definition of preeclampsia used and the risk for preeclampsia in the patients under

study. In women at a low risk for preeclampsia the areas under the summary receiver

operating characteristics curves in the second trimester were 0.68 (95% CI 0.64-0.72),

0.66 (95% CI 0.59-0.72) and 0.76 (95% CI 0.70-0.82) for systolic blood pressure,

diastolic blood pressure and mean arterial pressure, respectively. Only six studies

reported blood pressure measurements in the first trimester. These studies showed similar

results. Second trimester mean arterial pressure of 90 mmHg or more showed a positive

likelihood ratio of 3.5 (95% CI 2.0-5.0) and a negative likelihood ratio of 0.46 (95% CI

0.16-0.75). In women at high risk a diastolic blood pressure of 75 mmHg or more at

13 to 30 weeks of gestational age predicted preeclampsia best: positive likelihood ratio

of 2.8 (95% CI 1.8-3.6) and a negative likelihood ratio of 0.39 (95% CI 0.18-0.71).

Additional subgroup analysis did not show improved predictive accuracy.

We concluded that as a single test blood pressure is insufficient for the prediction

of preeclampsia or selection of women at risk in daily clinical practice. Of all blood

pressure measurements the mean arterial pressure is the most promising parameter.

It was not possible to investigate if automated devices are superior to conventional

sphygmomanometry. This is subject to further study.

Prospective cohort study

We collected a prospective cohort of healthy nulliparous women and multiparous

women with an increased risk for developing hypertensive disorders to investigate if

cardiovascular differences can be observed early in pregnancy between women, who will

develop preeclampsia or gestational hypertension later in pregnancy and women with an

uncomplicated pregnancy. We measured blood pressure early in pregnancy between 8 0/7-11 0/7 weeks gestational age using conventional sphygmomanometry (Maxi Stabil 3,

an aneroid sphygmomanometer, Welch Allyn, Skaneateles Falls, New York, USA) (using

an aneroid sphygmomanometer), continuous finger arterial waveform registration by the

FinometerTM (Finapres Medical Systems, Amsterdam, The Netherlands) and ambulatory

blood pressure monitoring (ABPM) for 48 hours with the Spacelab 90207TM (Redmond,

WA, USA). Further data to be collected were age, ethnic origin, smoking, medical history,

and weight and length.

295 women were included, 251 nulliparous women and 44 multiparous women with a

history of preeclampsia. Five women were excluded because of serious fetal congenital

malformations and one woman was lost to follow up. 289 Women were available for

analysis, 20 (6.9 %) developed preeclampsia and 16 (5.5 %) gestational hypertension.

39 women had an incomplete ABPM registration due to signal loss and patient non-

Chapter 9

147

Summ


endations

compliance. Of these women 28 had at least 10 measurements in 48 hour and were

used for analysis. Study results from this cohort are presented in chapter 3, 4 and 5.

Using 48 hour measurements of the ABPM the hyperbaric index (HBI) can be calculated

as the amount of blood pressure excess during the measurement period above a

90% tolerance limit.9 One research group evaluated the HBI for the prediction of

preeclampsia with extremely good results; they observed a sensitivity of 94% and a

specificity of 100%.10;11 Since the original report by Hermida et al. only one other

group studied HBI in pregnancy in a hospital-based research setting (sensitivity 80% and

specificity 77%).10 Therefore, we wanted to validate the HBI as a test for the prediction

of preeclampsia in our own cohort. In chapter 3 we present the results of this study.

We compared the predictive efficacy of the HBI with mean ABPM measurement for 48

hours and conventional manometry in our research cohort of nulliparous women and

parous women with a previous preeclampsia. We could include 219 of the 289 women

in our cohort in this study. For 42 women no ABPM measurement was available on

inclusion. Data of 28 women could not be analysed because of missing measurements

or failure of the device. The first 90 nulliparous women who had an uneventful pregnancy

formed thereference group for calculation of a time specified tolerance interval with 90%

confidencelimits. In the validation group, consisting of the remaining 63 nulliparous

women and 38 women with previous preeclampsia, the hyperbaric index was calculated

as the time-specified blood pressure excess over this tolerance limit for SBP, DBP and

mean arterial pressure. The maximum HBI was defined as the maximum value of either

the HBI of SBP or DBP or MAP for each individual.

For preeclampsia the maximum HBI had the best predictive capacity, however, the

difference with standard ABPM measurement or conventional manometry was small.

The HBImax had a sensitivity of 73 % and a specificity of 86 % at an optimum cut-off

of 17.6mmHg x hour. SBP with sphygmomanometry had a sensitivity of 60 % and a

specificity of 92 % at a cut-off of 118 mmHg and 24 hours ABPM using DBP had a

sensitivity of 53 % with a specificity of 95% at a cut-off of 74 mmHg. The predictive

efficacy for gestational hypertension was poor with all methods (sensitivity between 54 %

and 77 %, specificity between 41 % and 78 %).

We concluded that conventional manometry, ABPM measurement and the HBI calculated

from 48 hour ABPM had a comparable, restricted predictive efficacy in a group with low

and high risk women. The high predictive value of the HBI as observed in earlier studies

could not be reproduced and therefore this method can not be recommended for use in

daily practice.

148

We observed a significantly higher blood pressure in women who developed preeclampsia

or gestational hypertension in the first trimester compared to women with unaffected

pregnancies with all three devices (Chapter 4). The absolute blood pressure values were

different between the devices. The point estimate of the Odds ratios of blood pressure

(SBP, DBP or MAP) in the first trimester after adjustment for previous preeclampsia for later

preeclampsia or combined pregnancy outcome (preeclampsia or gestational hypertension)

was higher using the automated blood pressure measurement techniques compared to

conventional sphygmomanometry, although there was a considerable overlap of 95%

CI intervals. The area under the ROC curves for conventional sphygmomanometry with

preeclampsia as pregnancy outcome were 0.73 (95% CI 0.58-0.83), 0.68 (95% CI

0.53-0.84) and 0.70 (95% CI 0.55-0.85) for SBP, DBP and MAP respectively. For finger

arterial pressure waveform registration this was 0.82 (95% CI 0.68-0.95), 0.77 (95% CI

0.63-0.91) and 0.82 (95% CI 0.69-0.96) and for ABPM 0.78 (95% CI 0.65-0.91), 0.77

(95% CI 0.63-0.90) and 0.78 (95% CI 0.65-0.91) for SBP, DBP and MAP respectively.

For the combined pregnancy outcome results were comparable.

Since signal loss in ABPM is considerable with sub-optimal patient compliance and

conventional sphygmomanometry is subject to considerable interobserver variation in

daily practice we concluded that early pregnancy blood pressure measurement by finger

arterial pressure waveform registration is the most promising automated blood pressure

measurement technique to study in a prospective large cohort study.

In chapter 5 we use data from our research cohort to evaluate arterial stiffness early

in pregnancy and the association with later preeclampsia. We hypothesised that the

combination with blood pressure (MAP) could improve this association.

235 nulliparous women and 42 parous women were included with an incidence of

preeclampsia of 4.3% and 16.6% respectively. Data collected with finger arterial

waveform registration were used. The finger arterial waveform was filtered to an aortic

waveform and the augmentation indexes (AIx) and augmentation pressure (AP) were

determined. After adjustment for previous preeclampsia the AIx and AP were associated

with later preeclampsia, areas under the ROC-curves 0.74 (95% CI 0.61-0.87) and

0.77 (95% CI0.61-0.87). When MAP was added to the model, the combination with

AP had the highest area under the ROC-curve (0.82 (95% CI 0.66-0.97)). This is not a

considerable improvement compared to MAP and parity alone, area under the ROC-

curve 0.79 (95% CI 0.64-0.93). Our results do not support the addition of AIx or AP to

blood pressure measurement, which is far easier to apply, for the assessment of risk for

preeclampsia.

Chapter 9

149

Summ


endations

Hypertensive disorders in pregnancy and adult cardiovascular diseases share predisposing

conditions, like elevated blood pressure, diabetes and obesity. Protective or disease

modifying factors of cardiovascular disease might also be applicable for the prevention

of hypertensive disorders in pregnancy. One of these factors is psychosocial stress.11;12

Several studies showed a relation with high psychosocial stress in pregnancy and

hypertensive disorders of pregnancy.15-18 However most of these studies are retrospective

or case-control studies.

In chapter 6 we investigated the association of preeclampsia and gestational hypertension

with psychosocial stress in the first half of pregnancy in a prospective cohort study. For

this study we used the data of the Amsterdam Born Children and their Development

study (ABCD-study). This is a prospective community-based study which examines the

relationship between various lifestyles during pregnancy and pregnancy outcome in a

multicultural population. For this assessment we included only the nulliparous women

with a singleton pregnancy in this cohort, who completed the questionnaire before 24

weeks, and delivered after 24 weeks. A postpartum questionnaire was used to gather

information on preeclampsia, hypertension or proteinuria. Data were linked with the

Dutch national obstetric register and all women who were coded as having hypertensive

complications in pregnancy were selected. Medical files were examined for all selected

women, either from the questionnaires, or from the Dutch national obstetric register

to confirm the diagnosis of preeclampsia and gestational hypertension according

to the International Society for the Study of Hypertension in Pregnancy guidelines.

From the non-selected women a random sample of 5% was drawn for confirmation

of absence of hypertensive complications by examination of medical files. The study

population consisted of 128 women with preeclampsia, 161 women with gestational

hypertension and 3390 controls. Psychosocial stress was defined as work stress (Work

Experience and Appreciation Questionnaire of van Velthoven et al,13 partly based on

the Job Content Instrument of Karasek et al.14), anxiety (the State-Trait Anxiety Inventory,

STAI15), depression (Center for Epidemiological Studies Depression Scale, CES-D16;17)

and pregnancy related anxiety (PRAQ-R18). After correction for sociodemographic and

medical confounders, we could not observe an association of psychosocial stress in the

first half of pregnancy with later preeclampsia or gestational hypertension.

Using the same cohort as described in chapter 6 we studied another disease modifying

factor for adult cardiovascular disease i.e. physical exercise. Physical exercise is known to

prevent cardiovascular disease and to reduce its symptoms. In chapter 7 we assessed if

physical activity in leisure time early in pregnancy reduces the incidence of preeclampsia

150

or gestational hypertension. In the questionnaire of the ABCD-study questions about

physical activity in leisure time in the past week were asked using questions about

walking, cycling, doing sports and other activities like “Do It Yourself” or gardening in

leisure time. As well the amount of time spent in minutes as the intensity of the physical

activity was asked.

After correction for sociodemographic and medical confounders, the amount of time or

intensity of physical activity in leisure time in the first half of pregnancy was not associated

with preeclampsia or gestational hypertension.

Many researchers focus on risk assessment for preeclampsia in (early) pregnancy

while developing prediction models which can be used before pregnancy are equally

interesting. It is well known that ovarian hormones affect the cardiovascular and renin-

aldosteron system.19-21 However, the magnitude of the effect of hormonal fluctuations

during the menstrual cycle on blood pressure regulation is not clear and literature in this

respect is not consistent.22-24 Moreover it is unclear if it is important to take into account

the menstrual cycle while performing measurements before conception. The purpose

of the study described in chapter 8 was to evaluate if hemodynamic parameters and

cardiovascular control vary between the follicular and luteal phase of the menstrual

cycle. Healthy women who were never pregnant were studied at days 5 to 10 and days

18 to 25 of the menstrual cycle. Blood pressure, heart rate, cardiac output and total

peripheral resistance were measured continuously using non-invasive finger arterial

pressure waveform registration. Baroreflex sensitivity (BRS) and sympathetic activity by

phase angle difference were studied using spectral analysis and xBRS. There were no

differences in hemodynamic parameters, BRS and sympathetic activity between the two

phases of the menstrual cycle.

We concluded that, although there is no difference in blood pressure, BRS and sympathetic

activity between the first and second half of the menstrual period, measurements are

preferably performed in the first half of the menstrual cycle as one is certain that no

pregnancy exists, since the magnitude of the influence on blood pressure of very early

pregnancy is unknown.

GENERAL DISCUSSION

This thesis reports on hemodynamic parameters early in pregnancy associated with the

development of preeclampsia or gestational hypertension later in pregnancy. These

Chapter 9

151

Summ


endations

variables could be used for development of models for the prediction of gestational

hypertensive disease.

For a model for the prediction of preeclampsia the nature and magnitude of risk indicators

is essential. This model might in addition help to better understand the pathophysiological

mechanism of the disease and to design preventive therapies. Selection of high risk women

facilitates the establishment of effectiveness of these preventive strategies, because the

low incidence of preeclampsia in an unselected population necessitates large patient

cohorts. Strategies for prevention will have to start early in pregnancy because the origin

of the disease is thought to occur at the time of placentation.25 Consequently, a test for

the selection of women at high risk should be effective in the first trimester of pregnancy.

Furthermore, the selection of women at high risk for obstetric complications enables

targeting of intensive obstetric care at those who will benefit most. A predictive test needs

a high sensitivity to reduce the number of false-negatives since the consequences of

preeclampsia are severe for mother and child. Severe preeclampsia is a rare disease

(0,5 %) which means that test characteristics must also show high specificity. Until now

such test in the first trimester is not available for screening in the general population.26

Studies in this thesis confirmed that blood pressure is increased already in the first

trimester in women with normal blood pressure according to standard criteria, who will

develop preeclampsia or gestational hypertension. This difference of blood pressure can

potentially be used for prediction (chapter 3 and 4). This association was independent

of the device used. The mean arterial pressure (MAP) had the best predictive capacity

according to our meta-analysis of the literature.27 This we could only partially confirm in

our prospective cohort study in first trimester pregnancy. We found a higher association

with later preeclampsia (higher point estimate of Odds ratios) of ABPM and of finger

arterial waveform registration than of conventional sphygmomanometry. Unfortunately

our sample size was not large enough for detecting significant differences. Wearing an

ABPM device for 48 hours was considered as a burden and this resulted in 18% signal

loss. 28 women discontinued wearing the device because it impeded them in their daily

activities or disturbed their sleep. Especially sleep disturbance is reported as a major

cause of noncompliance.28

By using finger arterial waveform registration not only blood pressure is measured.

From the peripheral pulse wave the aortic pressure wave can be reconstructed and

parameters of arterial compliance can be calculated. This can be used to gain more

knowledge about preeclampsia and the development of the disease during pregnancy.

Increased arterial stiffness is an independent predictor of adverse outcome in adult

cardiovascular disease.29 Several studies showed an elevated arterial stiffness in women

with preeclampsia.30-33 Although in our prospective cohort study (chapter 5), women

152

who developed later preeclampsia had a higher AIx and AP compared to women with

an uncomplicated pregnancy, we found, in contrast to the results of an other research

group, comparable Odds ratios with mean arterial blood pressure.34 We concluded

that there are no advantages of adding the AIx or AP to automated blood pressure in

predicting preeclampsia.

Several other factors early in pregnancy are associated with later preeclampsia and their

predictive capacity has been studied by others. Maternal factors like age, BMI, a previous

pregnancy with preeclampsia and medical history are associated with the development

of preeclampsia.35 Even when these maternal factors are combined in a multivariate

analysis and a prediction algorithm is developed the detection rate of preeclampsia is

low. At a 5% false positive rate the detection rate of preeclampsia was approximately 30%

and 20 % for gestational hypertension.36 Other potential predictors like abnormal uterine

artery Doppler findings or biochemical markers have also been investigated. Uterine

artery Doppler assessment showed to be of limited value as a screening test. In a meta-

analysis Cnossen et al. found a sensitivity of 43% and a specificity of 93% for bilateral

notching of the uterine arteries in the second trimester in a low risk population.37 Serum

markers for Down’s syndrome are also not suitable as a single test.38 Prediction using

angiogenic factors is only reported in small studies and have insufficient discriminative

capacity.39 Combination of biochemical markers and abnormal uterine artery Doppler

findings improved the performance of early prediction of preeclampsia but in low risk

population the sensitivity was not higher than 60-80%.40 Recent developments using

metabolomics for predicting of preeclampsia need confirmation in large cohorts with

selected variables.3

A clinically useful prediction test will probably integrate a combination of multiple known

factors associated with preeclampsia. Earlier in this thesis we considered that preeclampsia

and cardiovascular disease had similar risk indicators. The most frequently used

prediction model for cardiovascular disease is the Framingham score (using sex, age, total

cholesterol, high-density-lipoprotein cholesterol, systolic blood pressure and smoking-

status) for estimating 10-year risk for adverse cardiovascular outcome. Until now this

model could not be improved by adding new biomarkers associated with cardiovascular

risk obtained by large scale genomic analysis of patients.41 These prediction models

including the Framingham score are not useful for prediction of preeclampsia so we have

to search for at more pregnancy specific markers. Using logistic regression analysis Poon

et al. developed an algorithm for the early prediction of preeclampsia combining the

logs of the uterine artery pulsatility index, mean arterial pressure, pregnancy-associated

plasma protein-A, and placental growth factor and maternal factor-derived a-priori

risk.42 Blood pressure was measured with an automated device using oscillometry. Using

Chapter 9

153

Summ


endations

the same variables they developed three prediction models, for preeclampsia before 34

weeks of pregnancy (early preeclampsia), preeclampsia after 34 weeks of pregnancy (late

preeclampsia) and gestational hypertension. At 110/7-136/7 weeks of pregnancy 93.1%

of the women with early preeclampsia, 35.7% of the women with late preeclampsia,

and 18.3% of the women who developed gestational hypertension could be detected

with a 5% false-positive rate. This model seems promising although 476 women tested

positive for early preeclampsia, of whom 32 actually developed early disease and 444

did not. Two cases with early preeclampsia were missed. Only 35.7% of the women with

preeclampsia after 34 weeks of pregnancy were identified. Because preeclampsia at

term is associated with severe maternal morbidity and mortality this model still must be

improved before it is suitable for daily clinical practice.

The authors made different models for early and late preeclampsia with different predictive

capacity. While early preeclampsia is often complicated by fetal growth retardation

because of poor placentation, in late preeclampsia the placenta and fetus often have

an appropriate size for the gestational age. It is hypothesized that early preeclampsia is

caused by excessive or atypical maternal immune response to poor placentation while

in late preeclampsia predisposing cardiovascular and metabolic disorders are more

prominent factors that lead to a cascade of placental and systemic inflammation and

oxidative stress resulting in preeclampsia.43 This indeed can explain the necessity for

different prediction models.

Women with obesity, chronic hypertension, vascular disease, diabetes, or a previous

preeclampsia have an increased risk for developing preeclampsia. Seed et al. found an

incidence of preeclampsia 23% in nulliparous women with chronic hypertension and a

body mass index >30 m/kg2. In women with chronic hypertension and preeclampsia in a

previous preeclampsia 30% developed again preeclampsia.44 If prevention for preeclampsia

will be available it is probably more cost-effective to treat these women without testing. It

will depend on screening test efficacy and the benefit of a potential preventive treatment

to determine at which a priori risk it is cost effective to use screening for the selection of

women for preventive therapy in stead of advising preventive treatment to all.

The development of prediction models is hampered by the low incidence of preeclampsia

as a low number of events (the occurrence of preeclampsia) and many suspected

confounders cause an overestimation of regression coefficients.45 For the same reasons

we could not develop a prediction model in our cohort.

All reported models have to be validated in an external population. They must be tested

by other investigators and in populations that differ from those used to develop them, in

order to validate their use in daily clinical practice. This is underlined by the present thesis

in which we show the importance of external validation since we could not reproduce the

154

high predictive capacity of the HBI and of the AIx as found by others.34;46;47 Even if this

could be explained by small differences in population or gestational age this makes these

markers questionable for prediction of preeclampsia in the general population.

IMPLICATIONS FOR FUTURE RESEARCH

1. Development of prediction models

Prediction of preeclampsia early in pregnancy or even before pregnancy remains

important to facilitate timely preventive strategies. Prevention will have to start early

because the origin of the disease is thought to occur at the time of placentation. Large

population-based studies evaluating algorithms combining maternal characteristics

with mean arterial pressure and other parameters measured in the first trimester are

needed. As women with hypertension, renal disease, diabetes or a previous preterm

preeclampsia are at high risk for developing preeclampsia, further selection is

questionable. Because 2/3 of all cases of preeclampsia occur in nulliparous women

and in most of these cases the disease starts unexpectedly and acutely, research

should focus on this group. Possibly different prediction models are needed for early

and late preeclampsia.

2. Use of automated blood pressure devices in prediction of hypertensive disease

Conventional sphygmomanometry is sufficient to diagnose preeclampsia since

most data for intervention and anti hypertensive treatment are based on this type

of measurement. Several automated devices are validated in pregnancy and

measure blood pressure securely and in addition can measure other cardiovascular

characteristics. Data on these additional parameters to predict preeclampsia and its

complications might be worth while to obtain. Based on our findings, finger arterial

waveform registration is the preferred method for further research.

3. Prospective collection of clinical data and biomaterials of pregnant women

Prospectively collecting and storing clinical data and biomaterials of pregnant women

according to standardized and professional protocols is essential for developing

prediction models. As shown by us and by other research groups blood pressure

alone measured by conventional sphygmomanometry or other devices is not sufficient

to predict preeclampsia in daily clinical practice.27;48 A useful model will probably

contain maternal characteristics, blood pressure and biomarkers. Large cohorts are

being studied at this time and the first data are emerging. The Scope study collects such

a cohort and is an example of a high quality biobank in New Zealand, Australia, the

Chapter 9

155

Summ


endations

UK and the USA.3;49 Since Dutch obstetricians and midwives are very well organized

in research consortia the same effort is recommended within for example the String of

Pearls Initiative (www.parelsnoer.org). By collecting data together we can enhance the

development and validation of such a prediction model.

156

Reference List (1) Sibai B, Dekker G, Kupferminc M. Pre-eclampsia. Lancet 2005; 365(9461):785-799. (2) Poon LC, Karagiannis G, Leal A, Romero XC, Nicolaides KH. Hypertensive disorders in

pregnancy: screening by uterine artery Doppler imaging and blood pressure at 11-13 weeks. Ultrasound Obstet Gynecol 2009; 34(5):497-502.

(3) Kenny LC, Broadhurst DI, Dunn W, Brown M, North RA, McCowan L et al. Robust early pregnancy prediction of later preeclampsia using metabolomic biomarkers. Hypertension 2010; 56(4):741-749.

(4) Meads CA, Cnossen JS, Meher S, Juarez-Garcia A, Ter RG, Duley L et al. Methods of prediction and prevention of pre-eclampsia: systematic reviews of accuracy and effectiveness literature with economic modelling. Health Technol Assess 2008; 12(6):1-270.

(5) Yu CK, Smith GC, Papageorghiou AT, Cacho AM, Nicolaides KH. An integrated model for the prediction of preeclampsia using maternal factors and uterine artery Doppler velocimetry in unselected low-risk women. Am J Obstet Gynecol 2005; 193(2):429-436.

(6) Yu J, Shixia CZ, Wu Y, Duan T. The study of inhibin A, activin A, placental growth factor and uterine artery Doppler pulsatility index to predict pre-eclampsia. Ultrasound Obstet Gynecol 2010.

(7) Thilaganathan B, Wormald B, Zanardini C, Sheldon J, Ralph E, Papageorghiou AT. Early-pregnancy multiple serum markers and second-trimester uterine artery Doppler in predicting preeclampsia. Obstet Gynecol 2010; 115(6):1233-1238.




(11) Rozanski A, Blumenthal JA, Kaplan J. Impact of psychological factors on the pathogenesis of cardiovascular disease and implications for therapy. Circulation 1999; 99(16):2192-2217.

(12) Anand SS, Islam S, Rosengren A, Franzosi MG, Steyn K, Yusufali AH et al. Risk factors for myocardial infarction in women and men: insights from the INTERHEART study. Eur Heart J 2008; 29(7):932-940.

(13) Van Velthoven M, Meijman T. het meten van psychosociale arbeidsbelasting met een vragenlijst: De Vragenlijst Beleving en Beoordeling van de Arbeid(VBBA). Amsterdam: Nederlands Instituut voor Arbeidsomstandigheden; 1994.

(14) Karasek R, Theorell T. Healthy work: Stress, productivity and the reconstruction of working life. New York: Basic Books; 1990.

(15) Spielberger C GRLR. STAI Manual for the State-Trait Anxiety Inventory. Palo Alto CA: Consulting Psychologists Press; 1970.

(16) Hanewald G.J.F.P. CES-D: De Nederlandse versie. Een onderzoek naar de betrouwbaarheid en validiteit. 1987. Amsterdam, University of Amsterdam, department of Clinical Psychology.

(17) Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Applied Psychological Measurement 1977; 1:385-401.

(18) Huizink AC, Mulder EJ, Robles de Medina PG, Visser GH, Buitelaar JK. Is pregnancy anxiety a distinctive syndrome? Early Hum Dev 2004; 79(2):81-91.

(19) Oelkers WK. Effects of estrogens and progestogens on the renin-aldosterone system and blood pressure. Steroids 1996; 61(4):166-171.

(20) Mendelsohn ME. Protective effects of estrogen on the cardiovascular system. Am J Cardiol 2002; 89(12A):12E-17E.

Chapter 9

157

Summ


endations

(21) Orshal JM, Khalil RA. Gender, sex hormones, and vascular tone. Am J Physiol Regul Integr Comp Physiol 2004; 286(2):R233-R249.

(22) Minson CT, Halliwill JR, Young TM, Joyner MJ. Influence of the menstrual cycle on sympathetic activity, baroreflex sensitivity, and vascular transduction in young women. Circulation 2000; 101(8):862-868.

(23) Tanaka M, Sato M, Umehara S, Nishikawa T. Influence of menstrual cycle on baroreflex control of heart rate: comparison with male volunteers. Am J Physiol Regul Integr Comp Physiol 2003; 285(5):R1091-R1097.

(24) Dubey RK, Oparil S, Imthurn B, Jackson EK. Sex hormones and hypertension. Cardiovasc Res 2002; 53(3):688-708.

(25) Lykke JA, Langhoff-Roos J, Sibai BM, Funai EF, Triche EW, Paidas MJ. Hypertensive pregnancy disorders and subsequent cardiovascular morbidity and type 2 diabetes mellitus in the mother. Hypertension 2009; 53(6):944-951.

(26) Cnossen JS, Ter RG, Mol BW, van der Post JA, Leeflang MM, Meads CA et al. Are tests for predicting pre-eclampsia good enough to make screening viable? A review of reviews and critical appraisal. Acta Obstet Gynecol Scand 2009; 88(7):758-765.










(36) Poon LC, Kametas NA, Chelemen T, Leal A, Nicolaides KH. Maternal risk factors for hypertensive disorders in pregnancy: a multivariate approach. J Hum Hypertens 2009.

(37) Cnossen JS, Morris RK, Ter RG, Mol BW, van der Post JA, Coomarasamy A et al. Use of uterine artery Doppler ultrasonography to predict pre-eclampsia and intrauterine growth restriction: a systematic review and bivariable meta-analysis. CMAJ 2008; 178(6):701-711.

(38) Morris RK, Cnossen JS, Langejans M, Robson SC, Kleijnen J, Ter RG et al. Serum screening with Down’s syndrome markers to predict pre-eclampsia and small for gestational age: systematic review and meta-analysis. BMC Pregnancy Childbirth 2008; 8:33.

(39) Lim JH, Kim SY, Park SY, Yang JH, Kim MY, Ryu HM. Effective prediction of preeclampsia by a combined ratio of angiogenesis-related factors. Obstet Gynecol 2008; 111(6):1403-1409.

(40) Giguere Y, Charland M, Bujold E, Bernard N, Grenier S, Rousseau F et al. Combining Biochemical and Ultrasonographic Markers in Predicting Preeclampsia: A Systematic Review. Clin Chem 2009.

158

(41) Lloyd-Jones DM. Cardiovascular risk prediction: basic concepts, current status, and future directions. Circulation 2010; 121(15):1768-1777.

(42) Poon LC, Kametas NA, Maiz N, Akolekar R, Nicolaides KH. First-trimester prediction of hypertensive disorders in pregnancy. Hypertension 2009; 53(5):812-818.

(43) Roberts JM, Hubel CA. The two stage model of preeclampsia: variations on the theme. Placenta 2009; 30 Suppl A:S32-S37.

(44) Seed PT, Chappell LC, Black MA, Poppe KK, Hwang YC, Kasabov N et al. Prediction of Preeclampsia and Delivery of Small for Gestational Age Babies Based on a Combination of Clinical Risk Factors in High-Risk Women. Hypertens Pregnancy 2010.

(45) Steyerberg EW, Eijkemans MJ, Habbema JD. Stepwise selection in small data sets: a simulation study of bias in logistic regression analysis. J Clin Epidemiol 1999; 52(10):935-942.



(48) Poon LC, Kametas NA, Pandeva I, Valencia C, Nicolaides KH. Mean arterial pressure at 11(+0) to 13(+6) weeks in the prediction of preeclampsia. Hypertension 2008; 51(4):1027-1033.

(49) Groom KM, North RA, Stone PR, Chan EH, Taylor RS, Dekker GA et al. Patterns of change in uterine artery Doppler studies between 20 and 24 weeks of gestation and pregnancy outcomes. Obstet Gynecol 2009; 113(2 Pt 1):332-338.

19 Nederlandse samenvatting, conclusies en aanbevelingen

Chapter 9

163

Nederlandse sam

envatting, conclusies en aanbevelingen

SAMENVATTING

Preëclampsie is een belangrijke oorzaak van zowel moederlijke als neonatale morbiditeit

en mortaliteit.1;2 Ondanks dat er al lang onderzoek naar wordt gedaan, is het ontstaan

en het verloop van de ziekte onvoorspelbaar. Ook bestaat er geen behandeling

behalve het beëindigen van de zwangerschap. De symptomen worden pas in de

tweede helft van de zwangerschap zichtbaar, terwijl het waarschijnlijk ten tijde van de

ontwikkeling van de placenta al mis gaat. Het exacte onderliggende pathofysiologische

mechanisme is onduidelijk en waarschijnlijk spelen vasculaire, immunologische,

genetische en omgevingsgerelateerde factoren een rol.3 Preventie van preëclampsie zou

wereldwijd een belangrijke bijdrage leveren aan het verbeteren van de uitkomst van

de zwangerschap voor moeder en kind. Omdat de ziekte waarschijnlijk al in het begin

van de zwangerschap ontstaat, is het vroeg kunnen voorspellen van het ontstaan van

preëclampsie erg belangrijk.3 Hoewel er op dit moment een aantal veelbelovende testen

ontwikkeld worden is een dergelijke test niet beschikbaar in de dagelijkse praktijk.4;5

Het onderzoek, beschreven in dit proefschrift, richt zich op cardiovasculaire verschillen

in het begin van de zwangerschap tussen vrouwen met een ongecompliceerde

zwangerschap en vrouwen die later in de zwangerschap preëclampsie hebben gekregen.

En ook op mogelijke risicofactoren die beïnvloedbaar zijn zoals psychosociale stress

en lichaamsbeweging. Deze verschillen en risicofactoren zouden gebruikt kunnen

worden bij het voorspellen van de kans op preëclampsie en bij eventuele preventie

van de ziekte. Wij hebben ons onderzoek voornamelijk beperkt tot vrouwen met een

eerste doorgaande zwangerschap; zij hebben namelijk een twee keer zo hoog risico op

hypertensieve aandoeningen van de zwangerschap, vergeleken met vrouwen die al een

keer een kind hebben gekregen. Bij vrouwen, die al eerder een kind hebben gekregen,

is de obstetrische voorgeschiedenis een dusdanig belangrijke voorspeller voor het al of

niet optreden van complicaties dat aanvullende testen waarschijnlijk niet nuttig meer

zijn. In dit hoofdstuk zullen de resultaten van ons onderzoek worden samen gevat en in

de discussie zal deze kennis in perspectief worden geplaatst. Ook zullen er suggesties

gedaan worden voor toekomstig onderzoek.

SYSTEMATISCHE REVIEW

In hoofdstuk 2 worden de resultaten besproken van een meta-analyse van de literatuur

naar de accuratesse van de systolische (SBP) en diastolische bloeddruk (DBP), de

gemiddelde arteriële bloeddruk (MAP) en van een stijging in de bloeddruk als voorspeller

164

van preëclampsie. Er voldeden 34 studies aan de inclusiecriteria. In totaal zijn in die

studies 60.599 vrouwen getest, waarvan 3341 met preëclampsie. Er is geen rekening

gehouden met het gebruik van verschillende methoden en apparaten om de bloeddruk te

meten. De oppervlakten onder de samenvattende receiver operating characteristic (ROC)

curven met 95% betrouwbaarheidsinterval voor bloeddrukmeting in het tweede trimester

bij vrouwenmet een laag risico zijn: 0,68 (0,64-0,72) voor de systolische bloeddruk; 0,66

(0,59-0,72) voor de diastolische bloeddruk; en 0,76 (0,70-0,82) voor de gemiddelde

arteriële bloeddruk. Bloeddrukmeting in het eerste trimester van de zwangerschap toont

vergelijkbare voorspellende waarden voor de verschillende parameters. Een gemiddelde

arteriële bloeddruk ≥ 90 mmHg in het tweede trimester toont een likelihood ratio voor een

positief test resultaat van 4,0 (95% betrouwbaarheidsinterval 2,4-5,6) en een likelihood

ratio voor een negatief test resultaat van 0,43 (95% BI 0,10-0,78). Bij vrouwen met een

hoog risico voorspelt een diastolische bloeddruk ≥ 75 mmHg gemeten tussen de 13 en

20 weken zwangerschapsduur preëclampsie het beste met een positieve likelihood ratio

2,8 (95% BI 1,8-3,6) en een negatieve likelihood ratio van 0,39 (95% BI 0,18-0,71).

De voorspellende waarde verbetert niet in de subgroep analyses. We concluderen dat de

hoogte van de bloeddruk weliswaar geassocieerd is met het ontstaan van preëclampsie

maar dat bloeddruk als enige test onvoldoende goed voorspelt om bruikbaar te zijn in

de dagelijkse praktijk.

Prospectieve cohort studie

Om te onderzoeken of er verschillen zijn in cardiovasculaire parameters vroeg in

de zwangerschap tussen vrouwen die later een hypertensieve aandoening van de

zwangerschap zullen krijgen en vrouwen met een ongecompliceerde zwangerschap,

hebben we prospectief een cohort vrouwen verzameld bestaand uit gezonde nullipara

vrouwen en gezonde multipara vrouwen met een preëclampsie in de voorgeschiedenis.

Alle vrouwen hadden een normale bloeddruk bij inclusie. Bij deze vrouwen werd vroeg

in de zwangerschap, tussen 8 0/7-11 0/7 weken zwangerschapsduur, de bloeddruk

gemeten met behulp van conventionele sfygmomanometrie (Maxi Stabil 3, Welch Allyn,

Skaneateles Falls, New York, USA), continue registratie van de arteriële drukgolf van de

vinger met de Finometer(Finapres Medical Systems, Amsterdam, Nederland) en met een

ambulante bloeddrukmeter (ABPM) gedurende 48 uur (Spacelab 90207, Redmond, WA,

USA). Ook werd informatie verzameld over leeftijd, etnische origine, roken, medische

voorgeschiedenis, opleiding en lengte en gewicht.

295 Vrouwen konden geïncludeerd worden waarvan 251 nullipara waren en 44 multipara

met een preëclampsie in de voorgeschiedenis. Vijf vrouwen werden geëxcludeerd in

verband met ernstige congenitale afwijkingen van de foetus en van één vrouw konden de

Chapter 9

165

Nederlandse sam


gegevens over de afloop van de zwangerschap niet worden achterhaald. De gegevens van

289 vrouwen waren beschikbaar voor analyse, 20 (6.9 %) kregen preëclampsie en 16 (5.5

%) zwangerschapshypertensie. 39 vrouwen hadden een incomplete ABPM registratie als

gevolg van technische problemen of omdat zij het apparaat niet konden verdragen. Van

deze vrouwen hadden 28 tenminste 10 metingen in 48 uur. Deze registraties werden wel

gebruikt bij de analyses. De resultaten van al deze metingen zijn beschreven in hoofdstuk

3, 4 en 5.

Met behulp van de 48-uurs registratie met ABPM kan de hyperbare index (HBI) berekend

worden. De HBI geeft aan hoeveel en hoe lang de bloeddruk boven de 90ste percentiel

van de normaalwaarden is geweest per tijdseenheid (mmHg x uur).6 Door één studiegroep

is de HBI gebruikt om in het eerste trimester latere preëclampsie te voorspellen met zeer

goede resultaten; zij vonden een sensitiviteit van 94% en een specificiteit van 100%.10;11

Sinds de publicatie van Hermida en collega’s is de HBI nog door één andere studiegroep

bestudeerd, waarbij de deelnemers allemaal in het ziekenhuis waren opgenomen.

(sensitiviteit 80%, specificiteit 77%).7 We wilden daarom de HBI als voorspellende test

voor preëclampsie onderzoeken in ons eigen cohort. In hoofdstuk 3 beschrijven

we de resultaten van dit onderzoek. We hebben de voorspellende waarde van de HBI

vergeleken met de voorspellende waarde van de gemiddelde bloeddruk over 48 uur

en conventionele sfygmomanometrie. 219 van de 289 vrouwen konden geïncludeerd

worden. De registraties van 28 vrouwen konden niet gebruikt worden in verband met

ontbrekende metingen of technische problemen. De eerste 90 nullipara vrouwen,

met een normale bloeddruk gedurende de hele zwangerschap, zijn gebruikt voor het

berekenen van de normaalwaarden over 24 uur. Bij de resterende 63 nullipara vrouwen

en 38 multipara vrouwen met een eerdere preëclampsie werd de HBI berekend voor de

SBP, DBP en MAP. De HBImax was gelijk aan de hoogste HBI van of SBP of DBP of MAP

voor iedere vrouw. Voor preëclampsie had de HBImax de best voorspellende waarde,

echter het verschil met de gemiddelde bloeddruk gemeten met ABPM of conventionele

sfygmomanometrie was klein. De HBImax had een sensitiviteit van 73 % en een specificiteit

van of 86 % bij een afkapwaarde van 17,6 mmHg x uur. SBP met sfygmomanometrie

had een sensitiviteit van 60 % en een specificiteit van 92 % bij een afkapwaarde van 118

mmHg. De gemiddelde DBP over 24 uur met ABPM had een sensitiviteit van 53 % met

een specificiteit van 95% bij een afkapwaarde van 74 mmHg. De voorspellende waarde

voor zwangerschapshypertensie was slecht met alle meetmethoden (sensitiviteit tussen 54

% en 77 %, specificiteit tussen 41 % and 78 %).

Onze conclusie is dan ook dat conventionele sfygmomanometry, ABPM en de HBI een

vergelijkbare maar beperkte voorspellende waarde hebben in een groep met een laag

en hoog risico. De hoge voorspellende waarde van de HBI zoals eerder beschreven

166

konden wij niet reproduceren en deze test kan daarom niet aanbevolen worden als

voorspellende test in de dagelijkse praktijk.

Vrouwen die later in de zwangerschap preëclampsie of zwangerschapshypertensie

kregen, hadden in het eerste trimester een hogere bloeddruk vergeleken met

vrouwen met een onaangedane zwangerschap, gemeten met alle drie de apparaten,

conventionele sfygmomanometrie, continue registratie van de arteriële drukgolf van de

vVinger en ABPM (hoofdstuk 4). De absolute waarden waren verschillend tussen de

apparaten. De puntschatters van de Odds ratio’s voor de associatie met preëclampsie

of gecombineerde zwangerschapsuitkomst (preëclampsie of zwangerschapshypertensie)

voor de verschillende bloeddrukparameters (SBP, DBP of MAP) in het eerste trimester

waren, na correctie voor een eerdere preëclampsie, hoger bij het gebruik van de

automatische bloeddrukmeters vergeleken met conventionele sfygmomanometry. Er was

wel een aanzienlijke overlap van de 95% betrouwbaarheidsintervallen. De oppervlakte

onder de ROCcurven voor conventionele sphygmomanometry met preëclampsie als

uitkomst waren 0.73 (95% BI 0.58-0.83), 0.68 (95% BI 0.53-0.84) en 0.70 (95% BI

0.55-0.85) voor respectievelijk SBP, DBP en MAP. Voor de Finometer was dit 0.82 (95%

BI 0.68-0.95), 0.77 (95% BI 0.63-0.91) en 0.82 (95% BI 0.69-0.96) en voor ABPM

0.78 (95% BI 0.65-0.91), 0.77 (95% BI 0.63-0.90) and 0.78 (95% BI 0.65-0.91) voor

respectievelijk SBP, DBP en MAP. Voor de gecombineerde uitkomst (preëclampsie of

zwangerschapshypertensie) waren de resultaten vergelijkbaar.

Omdat het bij een aanzienlijk deel van de vrouwen niet mogelijk was een volledige

ABPM registratie te verkrijgen en er bij conventionele sfygmomanometrie in de dagelijkse

praktijk een interobserver variatie is, concluderen wij dat bloeddrukmeting met behulp van

de arteriële drukgolf van de vinger de meest belovende automatische bloeddrukmeting is

om nader te bestuderen te worden in een groter cohort.

In hoofdstuk 5 hebben we de metingen van ons cohort gebruikt om de associatie

van arteriële stijfheid vroeg in de zwangerschap en het later krijgen van preëclampsie te

bestuderen. Onze hypothese was dat in combinatie met bloeddruk (MAP) de associatie

hoger zou zijn dan van de enkele parameters.

235 nullipara vrouwen en 42 multipara vrouwen konden geïncludeerd worden met een

respectievelijke incidentie van preëclampsie van 4.3% en 16.6%. De registraties met de

Finometer zijn gebruikt voor de analyses. Van de arteriële drukgolf van de vinger werd de

aortadrukgolf gereconstrueerd en werden de augmentatieindex (AIx) en augmentatiedruk

(AP) berekend. Na correctie voor een eerdere preëclampsie waren de AIx en AP in het

eerste trimester geassocieerd met latere preëclampsie, oppervlakten onder de ROCcurven

van 0.74 (95% BI 0.61-0.87) en 0.77 (95% BI 0.61-0.87). Wanneer de MAP werd

toegevoegd aan het model gaf de combinatie met de AP de grootste oppervlakte onder

Chapter 9

167

Nederlandse sam


de ROCcurve (0.82 95% BI 0.66-0.97). Vergeleken met MAP in combinatie met eerdere

preëclampsie is dit geen grote verbetering (oppervlakte onder de ROCcurve 0.79 (95%

BI 0.64-0.93)). Onze resultaten pleiten niet voor het toevoegen van de AIx of AP aan

een eventueel predictiemodel. De bloeddruk geeft vergelijkbare resultaten en is veel

makkelijker te meten.

Hypertensieve aandoeningen van de zwangerschap en hart- en vaatziekten delen een

aantal risicofactoren zoals hypertensie, diabetes en obesitas. Factoren die beïnvloedbaar

zijn en hart- en vaatziekten zouden kunnen voorkomen of een (on)gunstige invloed hebben

zouden misschien ook bruikbaar zijn bij de preventie van hypertensieve aandoeningen

van de zwangerschap. Een van deze factoren is psychosociale stress.8;9 Verscheidene

studies laten een verband zien tussen veel psychosociale stress in de zwangerschap en

het ontwikkelen van preëclampsie of zwangerschapshypertensie.15-18 Echter de meeste

van deze studies zijn retrospectief of case-control studies.

In hoofdstuk 6 hebben we de associatie van preëclampsie en zwangerschapshypertensie

met psychosociale stress in de eerste helft van de zwangerschap onderzocht in een

prospectief cohort. Voor dit onderzoek hebben we gebruik gemaakt van de data

verzameld in de ABCD-studie (Amsterdam Born Children and their Development study).

Dit is een prospectief community-based onderzoek dat de relatie tussen verschillende

levensstijlen tijdens de zwangerschap en zwangerschapuitkomsten onderzoekt in een

multiculturele populatie. In het kader van dit onderzoek werden tussen januari 2003

en maart 2004 alle zwangere vrouwen in Amsterdam bij eerste controle verzocht

een vragenlijst in te vullen. Voor dit onderzoek hebben we de gegevens gebruikt van

alle nullipara vrouwen met een eenlingzwangerschap, die de vragenlijst voor de 24e

zwangerschapsweek ingevuld hadden en na 24 weken zwangerschapsduur zijn bevallen.

De vragenlijst die postpartum was ingevuld is gebruikt om informatie te krijgen over

preëclampsie, hypertensie of eiwitverlies in de urine tijdens de zwangerschap. Ook

werden gegevens gebruikt van de Landelijke Verloskundige Registratie (LVR) om

informatie te verkrijgen over de zwangerschapsuitkomst. Van alle vrouwen die aangaven

tijdens de zwangerschap hypertensie of eiwitverlies in de urine te hebben gehad werden

de medische dossiers opgevraagd en bestudeerd. Net als van die vrouwen waarbij in de

de LVR was aangegeven dat de zwangerschap gecompliceerd was door hypertensie. De

diagnose preëclampsie of zwangerschapshypertensie werd gesteld volgens de richtlijnen

van de International Society for the Study of Hypertension in Pregnancy. Van alle vrouwen,

die niet aangegeven hadden hypertensie te hebben gehad, werd van willekeurig 5% de

medische dossiers bestudeerd om er zeker van te zijn dat er geen vrouwen met een

hypertensieve aandoening gemist werden. De studiepopulatie bestond uit 128 vrouwen

met preëclampsie, 161 vrouwen met zwangerschapshypertensie en een controlegroep

168

van 3390 vrouwen. Psychosociale stress was gedefinieerd als werk stress (gemeten met

Work Experience and Appreciation Questionnaire van van Velthoven en collega’s,10

gedeeltelijk gebaseerd op the Job Content Instrument of Karasek et al.11), angst ( gemeten

met the State-Trait Anxiety Inventory, STAI12), depressie (Center for Epidemiological

Studies Depression Scale, CES-D13;14) en zwangerschapsgerelateerde angst (PRAQ-R15).

Na correctie voor sociodemografische en medische factoren vonden wij geen associatie

tussen psychosociale stress in de eerste helft van de zwangerschap en het ontwikkelen

van preëclampsie of zwangerschapshypertensie later in de zwangerschap.

Gebruik makend van hetzelfde cohort als in hoofdstuk 6 hebben we de associatie

tussen lichaamsbeweging en hypertensieve aandoeningen in de zwangerschap

onderzocht. Lichaamsbeweging voorkomt hart- en vaatziekten en kan de symptomen

verminderen.16;17 De hypothese was dat vrouwen met meer lichaamsbeweging

minder kans op een hypertensieve aandoening in de zwangerschap zouden hebben.

In hoofdstuk 7 werd deze hypothese onderzocht. In de vragenlijst gebruikt door de

ABCD-studie werden vragen gesteld over lichaamsbeweging in vrije tijd in de week

voor het invullen. Er werden vragen gesteld over wandelen, fietsen, sporten en andere

activiteiten zoals klussen en tuinieren. Ook werd er gevraagd naar het aantal minuten

en de intensiteit van de activiteit. Na correctie voor sociodemografische en medische

factoren werd er geen verband waargenomen tussen de incidentie van preëclampsie of

zwangerschapshypertensie met de duur en de mate van inspanning van lichaamsbeweging

in vrije tijd.

Veel onderzoekers richten zich op het ontwikkelen van een risicoschatting op hypertensieve

aandoeningen in de (vroege) zwangerschap. Maar ook voor de zwangerschap zou een

risicoschatting interessant zijn. Het is bekend dat de ovariële hormonen het cardiovasculaire

en renine-angiotensine-aldosteron systeem beinvloeden.18-20 Echter hoe groot het effect

van de hormonale fluctuaties tijdens de menstruatiecyclus is op de bloeddrukregulatie is

niet duidelijk en de literatuur over dit onderwerp is inconsistent.21-23 Het is dan ook niet

zeker of het noodzakelijk is rekening te houden met de menstruatiecyclus bij het doen

van metingen voor de conceptie. Het doel van de studie beschreven in hoofdstuk

8 was om te bestuderen of hemodynamische parameters en cardiovasculaire controle

anders zijn in de folliculaire fase van de menstruatiecyclus dan in de luteale fase. 39

gezonde vrouwen, die nog nooit zwanger waren geweest, werden geïncludeerd voor

deze studie. Op dag 5-10 en op dag 18-25 van de menstruatiecyclus werden bloeddruk,

hartfrequentie, cardiac output en de totale perifere weerstand gemeten met behulp van

continue registratie van de arteriële drukgolf van de vinger. De baroreflex sensitivity (BRS)

en sympatische activiteit werden bestudeerd met behulp van spectraal analyse en xBRS.

Chapter 9

169

Nederlandse sam


Er waren geen verschillen in hemodynamische parameters, BRS en sympatische activiteit

tussen de twee fases van de menstruatiecyclus. Hoewel er geen verschil is tussen de eerste

en de tweede helft van de menstruatiecyclus, worden metingen bij voorkeur gedaan in de

eerste helft van de cyclus omdat het dan zeker is dat er geen zwangerschap bestaat. Dit

omdat de invloed van een zeer jonge zwangerschap op de bloeddrukregulatie onbekend

is.

DISCUSSIE

Dit proefschrift rapporteert over hemodynamische parameters vroeg in de zwangerschap

die geassocieerd zijn met het ontwikkelen van preëclampsie of zwangerschapshypertensie

later in de zwangerschap. Deze variabelen zouden gebruikt kunnen worden in een

predictiemodel voor deze aandoeningen.

Een predictiemodel zou van groot nut kunnen zijn bij het ontwikkelen van een preventieve

therapie. Omdat preëclampsie in een ongeselecteerde populatie weinig voorkomt zijn

nu grote cohorten nodig voor onderzoek. Dit onderzoek zou gemakkelijker worden

als vrouwen met een verhoogd risico geselecteerd zouden kunnen worden. Ook zal

preventie waarschijnlijk al vroeg in de zwangerschap moeten starten omdat de oorzaak

voor de ziekte vroeg in de zwangerschap ligt ten tijde van de ontwikkeling van de

placenta.24 Daarom zou een test die vrouwen met een hoog risico kan selecteren al in

het eerste trimester van de zwangerschap bruikbaar moeten zijn. Een tweede mogelijke

toepassing is om op basis van een risicoschatting op hypertensieve complicaties in de

zwangerschap de intensiteit van de prenatale zorg te bepalen. Een voorspellende test

zal een hoge sensitiviteit moeten hebben om het aantal fout negatieven te verkleinen

aangezien preëclampsie ernstige gevolgen voor moeder en kind kan hebben. Omdat

vooral ernstige preëclampsie niet veel voorkomt (0.5%) moet de voorspellende test ook

een hoge specificiteit hebben. Tot nu toe is een dergelijke test niet beschikbaar om te

gebruiken in de algemene populatie zwangere vrouwen.25

Het onderzoek beschreven in deze thesis bevestigt dat al in het eerste trimester de

bloeddruk bij vrouwen, zonder preëxistente hypertensie, die later preëclampsie

of zwangerschapshypertensie krijgen, verhoogd is. Dit verschil in bloeddruk zou

gebruikt kunnen worden bij het voorspellen van hypertensieve aandoeningen van de

zwangerschap (hoofdstuk 3 en 4). De gemiddelde arteriële bloeddruk (MAP) heeft

het meest onderscheidende vermogen volgens onze meta-analyse van de literatuur

(hoofdstuk 2).26 Dit hebben we gedeeltelijk kunnen bevestigen in onze prospectieve

cohortstudie. Ook vonden we een hogere associatie met later preëclampsie van

170

bloeddruk gemeten met een ambulante bloeddrukmeter of met de Finometer dan met

conventionele sfygmomanometrie. Helaas was het aantal vrouwen in ons cohort te

klein om significante verschillen te kunnen vinden. Het dragen van een ambulante

bloeddrukmeter is belastend, in ons onderzoek was het niet mogelijk om bij 18% van de

vrouwen een complete registratie te maken. Met de Finometer kan meer dan alleen de

bloeddruk worden gemeten. Van de arteriële drukgolf van de vinger kan de aortodrukgolf

gereconstrueerd worden en zo kan de compliantie van de vaatwand bestudeerd worden.

Dit kan worden gebruikt om meer kennis over preëclampsie te verkrijgen. Verhoogde

arteriële stijfheid is een onafhankelijke voorspeller van een slechte uitkomst bij mensen

met hart- en vaatziekten.27 Verscheidene studies hebben een verhoogde arteriële stijfheid

aangetoond bij vrouwen met preëclampsie.28-31 Ook in ons cohort hadden vrouwen

in het eerste trimester van de zwangerschap die later preëclampsie kregen een hogere

augmentatieindex en hogere augmentatiedruk (hoofdstuk 5). In tegenstelling tot anderen

vonden wij een vergelijkbare associatie van AIx, AP en MAP. Wij concluderen dan ook dat

het toevoegen van AIx of AP aan MAP in een predictiemodel geen toegevoegde waarde

heeft.32

Verscheidene andere factoren zijn al vroeg in de zwangerschap geassocieerd met

preëclampsie en hun voorspellend vermogen is bestudeerd door anderen. Maternale

factoren zoals body mass index, medische voorgeschiedenis en een eerdere zwangerschap

gecompliceerd met preëclampsie zijn geassocieerd met het krijgen van preëclampsie.33

Wanneer deze factoren gecombineerd worden in een multivariate analyse en een

predictiemodel ontwikkeld is, is het aantal vrouwen, dat preëclampsie zal krijgen, die

met dit model opgespoord worden laag. Met een fout positief percentage van 5% werd

30% van de vrouwen die een preëclampsie zou krijgen en 20% van de vrouwen die

zwangerschapshypertensie zou krijgen opgespoord.34 Andere potentiële voorspellende

factoren zoals een abnormale doorbloeding van de arteria uterina of biochemische

markers zijn ook onderzocht. In een meta-analyse vonden Cnossen en collegae een

sensitiviteit van 43% en een specificiteit van 93% bij dubbelzijdige notches in de arteria

uterina in het tweede trimester van de zwangerschap in een populatie met een laag risico

op preëclampsie.35 De eiwitten die gebruikt worden bij de screening op Down syndroom

zijn niet bruikbaar als voorspellende test.36 De voorspellende waarde van eiwitten

betrokken bij de angiogenese is alleen onderzocht in kleine studies en deze hebben

onvoldoende discriminerend vermogen.37 Het combineren van biochemische markers

met een abnormale doorbloeding van de arteria uterina verbetert het voorspellend

vermogen maar in een populatie met een laag risico op het krijgen van preëclampsie is

de sensitiviteit niet groter dan 60-80%.38

Chapter 9

171

Nederlandse sam


Met behulp van logistische regressie analyse ontwikkelde Poon en collegae een

predictiemodel voor preëclampsie. Zij combineren de logs van de pulsatility index van de

arteria uterina, gemiddelde bloeddruk, 2 eiwitten gemeten in het serum van de moeder

(PAPP-A en PGF) en maternale karakteristieken.4 De bloeddruk werd gemeten met een

automatische bloeddrukmeter die gebruik maakte van oscillometrie. Met deze variabelen

ontwikkelden zij drie modellen, een voor het krijgen van preëclampsie vóór 34 weken

zwangerschap (vroege preëclampsie), een voor het krijgen van preëclampsie na 34

weken (late preëclampsie) en een voor het ontwikkelen van zwangerschapshypertensie.

Bij een zwangerschapsduur van 110/7-136/7 weken konden zij met deze modellen, bij een

percentage foutpositieven van 5%, 93.1% van de vrouwen met een vroege preëclampsie,

35.7% van de vrouwen met een late preëclampsie en 18.3% van de vrouwen met een

zwangerschapshypertensie detecteren. Dit model lijkt veel belovend. Maar van de 476

vrouwen met een positieve test kregen er 32 dadwerkelijk preëclampsie en 444 niet.

Twee vrouwen met een vroege preëclampsie werden gemist. En van de vrouwen die een

late preëclampsie kregen had maar 35.7% een positieve test.

De auteurs hebben verschillende modellen ontwikkeld voor het voorspellen van vroege

en late preëclampsie met verschillend discriminerend vermogen. Vroege preëclampsie

wordt vaak gecompliceerd door intra-uteriene groeivertraging door een inadequate

placentatie terwijl bij een late preëclampsie er vaak een normale placenta is en een

normaal gegroeid kind. Men denkt dat dit komt omdat een vroege preëclampsie

veroorzaakt wordt door een abnormale reactie van het maternale immuunsysteem op

de ontwikkeling van de placenta en dat een late preëclampsie veroorzaakt wordt door

aanleg van de moeder voor hart- en vaatziekten en metabole stoornissen. 39 Dit kan

verklaren waarom er meerdere predictiemodellen nodig zijn.

Vrouwen met obesitas, hoge bloeddruk, hart- en vaatziekten, diabetes of een eerder

doorgemaakte preëclampsie hebben een verhoogde kans om in de zwangerschap

preëclampsie te krijgen. Seed en collegae vonden een incidentie van 23 % in nullipara

vrouwen met hoge bloeddruk en obesitas. Bij vrouwen met een eerdere preëclampsie en

hoge bloeddruk was de incidentie 30 %.40 Als het voorkomen van preëclampsie in de

toekomst mogelijk is, is het waarschijnlijk goedkoper om deze vrouwen te behandelen

zonder ze eerst te testen. Het zal afhangen van de kwaliteit van het predictiemodel en

de voordelen van een preventieve behandeling of het kosteneffectief is om eerst de kans

te berekenen op het krijgen van preëclampsie of meteen alle vrouwen te behandelen.

De ontwikkeling van predictiemodellen wordt gehinderd door de lage incidentie van

preëclampsie. In een cohort zijn vaak weinig ziektegevallen en veel factoren waar voor

gecorrigeerd moet worden. Dit kan zorgen voor een overschatting van het model.41 Dit

is ook de reden dat wij geen predictiemodel konden ontwikkelen.

172

Alle gepubliceerde modellen moeten worden gevalideerd in een andere populatie dan

waarin ze ontwikkeld zijn. Dit wordt nog eens benadrukt door de resultaten van ons

onderzoek aangezien we de goede voorspellende waarde van de HBI of de AIx niet

konden reproduceren. 32;42;43 Ook al kan dit verklaard worden door kleine verschillen

tussen de populaties en zwangerschapsduur. Dit maakt deze parameters ongeschikt om

de kans op preëclampsie in de algemene populatie te voorspellen.

IMPLICATIES VOOR TOEKOMSTIG ONDERZOEK

1. Ontwikkeling van predictiemodellenHet voorspellen van de kans op preëclampsie vroeg in de zwangerschap is belangrijk

om preventie van de ziekte mogelijk te maken. Preventie zal vroeg in de zwangerschap

moeten starten omdat de oorzaak van de ziekte zeer waarschijnlijk gevonden kan

worden ten tijde van de ontwikkeling van de placenta. Het is noodzakelijk om grote

studies te doen waarin de voorspellende waarde van de combinatie van maternale

karakteristieken met de gemiddelde bloeddruk in het eerste trimester en andere

parameters geëvalueerd worden. De eerste selectie zal gemaakt moeten worden

op basis van medische en obstetrische voorgeschiedenis. Vrouwen met preëxistente

hypertensie, nierziekten, diabetes of een preëclampsie vóór 34 weken zwangerschap in

de anamnese hebben een verhoogde kans op het krijgen van preëclampsie. Het is de

vraag of er voor deze vrouwen een predictiemodel ontwikkeld moet worden. Omdat

2/3 van alle ziektegevallen optreden bij vrouwen die hun eerste kind verwachten

zal verder onderzoek zich moeten richten op deze groep. Zeer waarschijnlijk zijn er

verschillende predictiemodellen nodig voor vroege en aterme preëclampsie.

2. Het gebruik van automatische bloeddrukmeters bij het voorspellen van hypertensieve aandoeningen van de zwangerschap

Conventionele sfygmomanometrie wordt gebruikt om de diagnose preëclampsie

te stellen omdat met deze methode het meeste onderzoek is gedaan naar de

gevolgen van preëclampsie en antihypertensieve therapie. Meerdere automatische

bloeddrukmeters zijn gevalideerd voor het gebruik in de zwangerschap en meten zeer

nauwkeurig de bloeddruk. Ook kunnen er vaak andere parameters met betrekking

tot het cardiovasculaire systeem gemeten worden. Deze parameters zouden bruikbaar

kunnen zijn bij het voorspellen van het optreden van preëclampsie en eventuele

complicaties. Gebaseerd op de resultaten van ons onderzoek adviseren wij de

Chapter 9

173

Nederlandse sam


bloeddruk te meten met behulp van de arteriële drukgolf van de vinger in toekomstig

onderzoek.

3. Prospectieve verzameling van klinische gegevens en biologisch materiaal van zwangere vrouwen

Het prospectief verzamelen en opslaan van klinische gegevens en biologisch

materiaal van zwangere vrouwen is van essentieel belang voor het ontwikkelen

van predictiemodellen. Zoals blijkt uit de resultaten van ons onderzoek en dat van

anderen, is bloeddruk alleen niet voldoende om het voorkomen van preëclampsie te

voorspellen in de dagelijkse praktijk.26;44 Een bruikbaar model zal zeer waarschijnlijk

bestaan uit een combinatie van maternale karakteristieken, bloeddruk en biomarkers.

Op dit moment worden grote cohorten vrouwen bestudeerd en de eerste resultaten

gepubliceerd. De Scope studie is hier een voorbeeld van. Binnen deze studie wordt

biologisch materiaal afkomstig van vrouwen uit Nieuw Zeeland, Australië, het

Verenigd koninkrijk en de Verenigde Staten van Amerika verzameld.45;46 In Nederland

zijn gynaecologen en verloskundigen verenigd binnen onderzoeksconsortia en zijn er

dus goede mogelijkheden voor het verzamelen van een gelijksoortig cohort binnen

bijvoorbeeld het Parelsnoerinitiatief (www.parelsnoer.org). Zo kan de ontwikkeling en

validatie van een predictiemodel gestimuleerd worden.

174

Referenties (1) Khan KS, Wojdyla D, Say L, Gulmezoglu AM, Van Look PF. WHO analysis of causes of maternal

death: a systematic review. Lancet 2006; 367(9516):1066-1074. (2) Schutte JM, Schuitemaker NW, van RJ, Steegers EA. Substandard care in maternal mortality due

to hypertensive disease in pregnancy in the Netherlands. BJOG 2008; 115(6):732-736. (3) Steegers EA, von DP, Duvekot JJ, Pijnenborg R. Pre-eclampsia. Lancet 2010; 376(9741):631-

644. (4) Poon LC, Kametas NA, Maiz N, Akolekar R, Nicolaides KH. First-trimester prediction of

hypertensive disorders in pregnancy. Hypertension 2009; 53(5):812-818. (5) Meads CA, Cnossen JS, Meher S, Juarez-Garcia A, Ter RG, Duley L et al. Methods of prediction

and prevention of pre-eclampsia: systematic reviews of accuracy and effectiveness literature with economic modelling. Health Technol Assess 2008; 12(6):1-270.



(8) Rozanski A, Blumenthal JA, Kaplan J. Impact of psychological factors on the pathogenesis of cardiovascular disease and implications for therapy. Circulation 1999; 99(16):2192-2217.

(9) Anand SS, Islam S, Rosengren A, Franzosi MG, Steyn K, Yusufali AH et al. Risk factors for myocardial infarction in women and men: insights from the INTERHEART study. Eur Heart J 2008; 29(7):932-940.

(10) Van Velthoven M, Meijman T. het meten van psychosociale arbeidsbelasting met een vragenlijst: De Vragenlijst Beleving en Beoordeling van de Arbeid(VBBA). Amsterdam: Nederlands Instituut voor Arbeidsomstandigheden; 1994.

(11) Karasek R, Theorell T. Healthy work: Stress, productivity and the reconstruction of working life. New York: Basic Books; 1990.

(12) Spielberger C GRLR. STAI Manual for the State-Trait Anxiety Inventory. Palo Alto CA: Consulting Psychologists Press; 1970.

(13) Hanewald G.J.F.P. CES-D: De Nederlandse versie. Een onderzoek naar de betrouwbaarheid en validiteit. 1987. Amsterdam, University of Amsterdam, department of Clinical Psychology.

(14) Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Applied Psychological Measurement 1977; 1:385-401.

(15) Huizink AC, Mulder EJ, Robles de Medina PG, Visser GH, Buitelaar JK. Is pregnancy anxiety a distinctive syndrome? Early Hum Dev 2004; 79(2):81-91.

(16) Mora S, Cook N, Buring JE, Ridker PM, Lee IM. Physical activity and reduced risk of cardiovascular events: potential mediating mechanisms. Circulation 2007; 116(19):2110-2118.

(17) Dickinson HO, Mason JM, Nicolson DJ, Campbell F, Beyer FR, Cook JV et al. Lifestyle interventions to reduce raised blood pressure: a systematic review of randomized controlled trials. J Hypertens 2006; 24(2):215-233.

(18) Oelkers WK. Effects of estrogens and progestogens on the renin-aldosterone system and blood pressure. Steroids 1996; 61(4):166-171.

(19) Mendelsohn ME. Protective effects of estrogen on the cardiovascular system. Am J Cardiol 2002; 89(12A):12E-17E.

(20) Orshal JM, Khalil RA. Gender, sex hormones, and vascular tone. Am J Physiol Regul Integr Comp Physiol 2004; 286(2):R233-R249.

(21) Minson CT, Halliwill JR, Young TM, Joyner MJ. Influence of the menstrual cycle on sympathetic activity, baroreflex sensitivity, and vascular transduction in young women. Circulation 2000; 101(8):862-868.

Chapter 9

175

Nederlandse sam


(22) Tanaka M, Sato M, Umehara S, Nishikawa T. Influence of menstrual cycle on baroreflex control of heart rate: comparison with male volunteers. Am J Physiol Regul Integr Comp Physiol 2003; 285(5):R1091-R1097.

(23) Dubey RK, Oparil S, Imthurn B, Jackson EK. Sex hormones and hypertension. Cardiovasc Res 2002; 53(3):688-708.

(24) Lykke JA, Langhoff-Roos J, Sibai BM, Funai EF, Triche EW, Paidas MJ. Hypertensive pregnancy disorders and subsequent cardiovascular morbidity and type 2 diabetes mellitus in the mother. Hypertension 2009; 53(6):944-951.

(25) Cnossen JS, Ter RG, Mol BW, van der Post JA, Leeflang MM, Meads CA et al. Are tests for predicting pre-eclampsia good enough to make screening viable? A review of reviews and critical appraisal. Acta Obstet Gynecol Scand 2009; 88(7):758-765.









(34) Poon LC, Kametas NA, Chelemen T, Leal A, Nicolaides KH. Maternal risk factors for hypertensive disorders in pregnancy: a multivariate approach. J Hum Hypertens 2009.

(35) Cnossen JS, Morris RK, Ter RG, Mol BW, van der Post JA, Coomarasamy A et al. Use of uterine artery Doppler ultrasonography to predict pre-eclampsia and intrauterine growth restriction: a systematic review and bivariable meta-analysis. CMAJ 2008; 178(6):701-711.

(36) Morris RK, Cnossen JS, Langejans M, Robson SC, Kleijnen J, Ter RG et al. Serum screening with Down’s syndrome markers to predict pre-eclampsia and small for gestational age: systematic review and meta-analysis. BMC Pregnancy Childbirth 2008; 8:33.

(37) Lim JH, Kim SY, Park SY, Yang JH, Kim MY, Ryu HM. Effective prediction of preeclampsia by a combined ratio of angiogenesis-related factors. Obstet Gynecol 2008; 111(6):1403-1409.

(38) Giguere Y, Charland M, Bujold E, Bernard N, Grenier S, Rousseau F et al. Combining Biochemical and Ultrasonographic Markers in Predicting Preeclampsia: A Systematic Review. Clin Chem 2009.

(39) Roberts JM, Hubel CA. The two stage model of preeclampsia: variations on the theme. Placenta 2009; 30 Suppl A:S32-S37.

(40) Seed PT, Chappell LC, Black MA, Poppe KK, Hwang YC, Kasabov N et al. Prediction of Preeclampsia and Delivery of Small for Gestational Age Babies Based on a Combination of Clinical Risk Factors in High-Risk Women. Hypertens Pregnancy 2010.

(41) Steyerberg EW, Eijkemans MJ, Habbema JD. Stepwise selection in small data sets: a simulation study of bias in logistic regression analysis. J Clin Epidemiol 1999; 52(10):935-942.

176



(44) Poon LC, Kametas NA, Pandeva I, Valencia C, Nicolaides KH. Mean arterial pressure at 11(+0) to 13(+6) weeks in the prediction of preeclampsia. Hypertension 2008; 51(4):1027-1033.

(45) Kenny LC, Broadhurst DI, Dunn W, Brown M, North RA, McCowan L et al. Robust early pregnancy prediction of later preeclampsia using metabolomic biomarkers. Hypertension 2010; 56(4):741-749.

(46) Groom KM, North RA, Stone PR, Chan EH, Taylor RS, Dekker GA et al. Patterns of change in uterine artery Doppler studies between 20 and 24 weeks of gestation and pregnancy outcomes. Obstet Gynecol 2009; 113(2 Pt 1):332-338.

Appendices

Authors’ affiliationsBibliography

Appendices

181

Authors’ affiliations

AUTHORS’AFFILIATIONS

K. Boer Department of Obstetrics and Gynecology, Academic Medical Center, Amsterdam

G.J. Bonsel Department of Social Medicine, Academic Medical Center, Amsterdam

J.S. Cnossen Department of General Practice, Academic Medical Center, Amsterdam

A.Franx Department of Obstetrics and Gynecology, St Elisabeth Hospital, Tilburg

J. Gisolf Department of Physiology, Academic Medical Center, Amsterdam

I. Guelen BMYE, Amsterdam

J. Hashimoto Department of Blood Pressure Research, Tohoku University, Japan

K.S. Khan Department of Obstetrics and Gynaecology, Birmingham Womens Hospital, Birmingham, United Kingdom

L. van Leijden Department of Obstetrics and Gynecology, Academic Medical Center, Amsterdam

B.W.J. Mol Department of Obstetrics and Gynecology, Academic Medical Center, Amsterdam

G.A. van Montfrans Department of Internal Medicine, Academic Medical Center, Amsterdam

J.A.M. van der Post Department of Obstetrics and Gynecology, Academic Medical Center, Amsterdam

S. Rang Department of Obstetrics and Gynecology, Academic Medical Center, Amsterdam

G. ter Riet Department of General Practice, Academic Medical Center, Amsterdam

L.Seesing Department of Obstetrics and Gynecology, Academic Medical Center, Amsterdam

N. de Vrieze Department of General Practice, Academic Medical Center, Amsterdam

T.G.M. Vrijkotte Department of Social Medicine, Academic Medical Center, Amsterdam

M.F. van der Wal Municipal Health Services, GGD Amsterdam, Amsterdam

B.E. Westerhof BMEYE, Amsterdam and Laboratory for Clinical Cardiovascular Physiology, AMC Centre for Heart Failure Research, Amsterdam

H. Wolf Department of Obstetrics and Gynecology, Academic Medical Center, Amsterdam

Appendices

183

Bibliography

BIBLIOGRAPHY

Vollebregt KC, Wolf H, Boer K, van der Wal MF, Vrijkotte TG, Bonsel GJ. Does physical

activity in leisure time early in pregnancy reduce the incidence of preeclampsia or gestational

hypertension? Acta Obsteticia et Gynecologica Scandinavica. 2010;89(2):261-7

Vollebregt KC, Boer K, van der Post JAM, Wolf H. A critical view on the use of

ambulatory blood pressure monitoring in hypertensive disorders of pregnancy Journal of

Hypertension. 2010;28(5): 1111-3

Vollebregt KC, Gisolf J, Guelen I, Boer K, van Montfrans G, Wolf H. Limited

accuracy of the hyperbaric index, ambulatory blood pressure and sphygmomanometry

measurements in predicting gestational hypertension and preeclampsia. Journal of

Hypertension. 2010;28(1):127-34

Vollebregt KC, Walter A.W., Omtzigt A.W.J., Kleiverda G.Terugblik op een ernstig

verloop en perimortem sectio bij een patiënte met pre-eclampsie/HELLP-syndroom

NTOG 2008;121(9):299-02

Cnossen JS, Vollebregt KC, de Vrieze N, ter Riet G, Mol BW, Franx A, Khan KS, van der

Post JA. Accuracy of mean arterial pressure and blood pressure measurements in predicting

pre-eclampsia: systematic review and meta-analysis. BMJ. 2008;336(7653):1117-20.

Vollebregt KC, van der Wal MF, Wolf H, Vrijkotte TG, Boer K, Bonsel GJ.Is psychosocial

stress in first ongoing pregnancies associated with pre-eclampsia and gestational

hypertension? BJOG. 2008;115(5):607-15

Siegemund M, Van Bommel J, Sinaasappel M, Schwarte LA, Studer W, Girard T,

Vollebregt K, Ince C.The NO donor SIN-1 improves intestinal-arterial P(CO(2)) gap

in experimental endotoxemia: an animal study. Acta Anaesthesiologica Scandinavica

2007;51(6):693-700.

Vollebregt KC, Seesing L, Rang S, Boer K, Wolf H. Sensitivity of spontaneous baroreflex

control of the heart and hemodynamic parameters are not influenced by the menstrual

cycle. Hypertension in pregnancy. 2006;25(3):159-67

Vollebregt KC, Boer K, Mathura KR, de Graaff JC, Ubbink DT, Ince C.Impaired

vascular function in women with pre-eclampsia observed with orthogonal polarisation

spectral imaging. BJOG. 2001;108(11):1148-53

Mathura KR, Vollebregt KC, Boer K, De Graaff JC, Ubbink DT, Ince C. Comparison of

OPS imaging and conventional capillary microscopy to study the human microcirculation.

Journal of Applied Physiology. 2001;91(1):74-8

Dankwoord

Appendices

187

Dankw

oord

DANKWOORD

Mijn proefschrift is af! Wetenschappelijk onderzoek doe je niet alleen. Dat is maar goed

ook. Als ik het even niet meer zag zitten was er altijd wel iemand die mij, door zijn of

haar enthousiasme, weer kon motiveren om door te gaan. Ik wil dan ook iedereen, met

wie ik de afgelopen jaren heb samengewerkt, heel hartelijk bedanken en een aantal in

het bijzonder.

Als eerste wil ik alle vrouwen bedanken die belangeloos hebben meegewerkt aan de

verschillende projecten. Zonder de informatie over hun zwangerschappen en bloeddruk

had dit proefschrift niet tot stand kunnen komen.

Mijn promotor Prof. Dr. J.A.M. van der Post.

Beste Joris, je hebt altijd een grote betrokkenheid getoond voor mij en mijn onderzoek.

Ook al zat je agenda helemaal vol, altijd was er nog een klein gaatje te vinden dat

vervolgens goed benut werd. Dank daarvoor. Ik ging altijd met nieuwe inspiratie naar

buiten, hopelijk is alles nu rond.

Mijn co-promotoren Dr. H. Wolf en Dr. K. Boer.

Beste Hans, altijd stond jij voor mij klaar en was er tijd voor een discussie. Het maakte

niet uit waar en wanneer; in het trappenhuis bij H3, tussen klinische bezigheden door,

zondagavond laat of per mail vanuit de Syrische woestijn. Het zou mij niets verbazen

als jij bijna net zo veel tijd als ik hebt besteed aan de manuscripten. Mede dankzij jouw

tomeloze inzet ligt dit proefschrift hier, dankjewel.

Beste Kees, eigenlijk is alles bij jou begonnen. Ik heb jou leren kennen tijdens mijn

wetenschappelijke stage onder supervisie van Prof. Dr. Çan Ince. Via dat onderzoek

ben ik in het project gerold dat is uitgemond in dit proefschrift. Al die jaren was het fijn

samenwerken. Ik vind het dan ook erg bijzonder dat je mijn voorkant gemaakt hebt,

dankjewel hiervoor. En hoe komt het toch dat ondanks de vele keren lezen van een

manuscript en de spellingcontrole jij er altijd weer een spelfout uit haalde?

De leden van de commissie: Prof. dr. E.T. van Bavel, Dr. J.M. Karemaker, Prof. dr. B.W.J.

Mol, Prof. dr. B.J.M. Mulder, Dr. M.E.A. Spaanderman en Prof. dr. E.A.P. Steegers wil ik

bedanken voor het beoordelen van mijn proefschrift.

De medeauteurs van alle manuscripten:

188

Beste Jeltsje Cnossen, Gerben ter Riet en Ben Willem Mol, het was fijn en erg leerzaam

om met jullie te werken.

Beste Gert van Montfrans, Ilja Guelen en Janneke Gisolf, jullie waren net zo enthousiast

als ik om de methode van de hyperbare index te ontcijferen en te reproduceren. Het was

eigenlijk te mooi om waar te zijn. Dat onze resultaten minder goed waren was dan ook

geen verassing. Het was erg leuk om met jullie te werken.

Beste Berend Westerhof en Junichiro Hashimoto, jullie kwamen met het idee van de

augmentatieindex en toevallig had ik de data liggen. Het is een leuk project geworden.

Beste Gouke Bonsel, Marcel van der Wal en Tanja Vrijkotte van de ABCD studiegroep,

jullie hebben een zeer waardevolle database met informatie over Amsterdamse zwangeren

en hun kinderen. Dank dat ik daar gebruik van heb kunnen maken.

Beste Saskia Rang, ik kan mij helemaal vinden in één van jouw stellingen: “Een golf is

eindeloos te analyseren”.

De studenten die bij mij hun wetenschappelijk stage hebben gedaan, Lobke, Eva, Kim,

Lize, Nicole, Maud, Eva, Larissa en Lara. In het kader van jullie wetenschappelijke stage

hebben jullie allemaal een steentje bijgedragen aan mijn onderzoek. Hopelijk hebben

jullie net zo veel geleerd van mij als ik van jullie. En bedenk met chocoladekoekjes wordt

SPSS een stuk makkelijker.

Alle medewerkers van de afdeling Verloskunde en Gynaecologie van het AMC, ik voel

mij thuis in het AMC en heb de afgelopen jaren enorm veel geleerd. Dank aan alle

stafleden en arts-assistenten voor de leuke opleidingstijd en aan alle verloskundigen,

echoscopisten, verpleegkundigen, doktersassistenten en secretaresses voor de prettige

samenwerking en niet te vergeten de gezelligheid. Ina Vink, fijn dat jij er was om, als

ik niet beschikbaar was, al die telefoontjes te beantwoorden. Peter Goossens, je hebt

meerdere malen mijn computer met kostbare data succesvol gereanimeerd en bent

daarom een van mijn helden. Debbie Bax, dank voor je hulp de laatste maanden.

Alle medewerkers van de afdeling Verloskunde en Gynaecologie van het Flevoziekenhuis.

Toen ik mijn verpleeghulpstage op jullie afdeling deed had ik nooit gedacht nog eens

terug te komen als AIOS. Ik heb dan ook veel zin in het laatste jaar van mijn opleiding.

Appendices

189

Dankw

oord

Mijn paranimfen.

Lieve Mariëlle, toen ik hoorde dat jij als verpleegkundige mij kwam helpen had ik eerst zo

mijn twijfels. Je had geen ervaring met wetenschappelijk onderzoek en had niet veel met

diverse computerprogramma’s gewerkt. Maar voordat ik het wist had jij orde gebracht

in mijn chaos, kon je metingen verrichten en analyseren en had je de opleiding tot

researchnurse gedaan. Daarom vind ik dat jij bij mijn promotie moet zijn, dankjewel dat

jij mijn paranimf wil zijn.

Lieve Marjolein, samen in een hok zonder ramen dat schept een band. En die band is

er ook in de kliniek en daarbuiten. Inmiddels ben ik bezig met het laatste deel van mijn

opleiding in het Flevoziekenhuis en ben jij weer terug in het AMC voor een fellowship.

Ik ben benieuwd hoe ons verdere leven eruit gaat zien, zowel professioneel als privé.

Hopelijk zullen er nog vele etentjes met of zonder partners en kinderen volgen (Janneke

kookt wel erg lekker). Ik heb er wel langer over gedaan dan jij, maar nu ook ik het afsluit

hoor jij erbij.

Alle collega onderzoekers en dat zijn er in de loop van de jaren heel wat geworden;

Anna, Arianne, Bas, Christianne, Elisabeth, Emily, Emmy, Esther, Etelka, Evelien,

Femke,Floortje, Inge, Jan Willem, Jolande, Katrien, Laura, Liesbeth van Leeuwen,

Liesbeth Visser, Madelon, Margreet, Marscha, Menke, Moniek, Pieternel, Saskia, Stef,

Wessel, Wouter Hehenkamp en Wouter Wegdam. Als parttime onderzoeker was ik niet

overal bij maar dank voor jullie steun en gezelligheid.

In het bijzonder wil ik my roomies noemen, eerst Marjolein en daarna Jiska. En vervolgens

als oudgediende tussen de jonkies; Noortje, Lobke, Marleen en Marjet. Regelmatig

moesten wij onszelf dwingen om terug te draaien naar de beeldschermen en door te

werken in plaats van thee te leuten. Hopelijk worden jullie boekjes minstens zo mooi.

Lieve vriendinnen en dan vooral Anoek, Ilja, Ilse, Karen, Roos en Sandra. Wat is het fijn

om mensen om je heen te hebben die je al heel lang kent en waarbij weinig woorden

genoeg kunnen zijn. Of bij wie je juist heel je hart kan uitstorten. Hopelijk is er nu meer

tijd voor avondjes Eijlders of gewoon een middag op de bank.

Lieve familie de Boer-Bruch, dank voor jullie steun en interesse in mijn werk.

Lieve Ada en Jan, in bijna alles wat wij kinderen wilden hebben jullie ons gesteund en

volgens mij komt daar het doorzettingsvermogen van ons alle vier vandaan. Het is fijn

om te weten dat jullie er altijd voor ons zullen zijn.

190

Lieve Bram, Eva en Marit. Onze familie wordt steeds groter en gezelliger, eerst met jullie

partners Jennifer, Maurits en Pim en nu ook met Kyran en Jurre erbij. Fijn dat jullie altijd

naar mijn verhalen wilden luisteren.

Special thanks to Jennifer for correcting my English grammar.

Lieve D, de afgelopen jaren heb je niet alleen achter mij gestaan maar ook naast mij. Je

moest eens weten hoe belangrijk dat is geweest, ik hou van jou!

Lieve Meinske en Stella, ieder mensenkind is weer een wonder, jullie zijn mijn

wereldwonderen.

Appendices

191

Curriculum

Vitae

CURRICULUM VITAE

Karlijn Vollebregt werd op 22 april 1976 als oudste van vier kinderen geboren in Bussum

en groeide op in Weesp. Nadat zij in 1994 haar VWO diploma behaalde op het Goois

Lyceum in Bussum werd zij uitgeloot voor de studie Geneeskunde. Zij ging Geneeskunde

studeren aan de Universiteit Gent in Gent (België). Na behalen van het eerste jaar

werd zij in 1995 ingeloot en ging zij terug naar Nederland om Geneeskunde te gaan

studeren aan de Universiteit van Amsterdam. Op aanraden van haar moeder deed zij

haar verpleeghulpstage op de afdeling verloskunde van het Flevoziekenhuis te Almere en

werd zij enthousiast voor het vak. In het kader van haar wetenschappelijke stage tijdens

haar studie deed zij onderzoek onder leiding van Professor Ç. Ince bij de experimentele

anaesthesiologie. Daar ontmoette zij Kees Boer en werd de basis gelegd voor haar verder

onderzoek binnen de verloskunde. In 2002 ging zij, na haar afstuderen, werken als

ANIOS gynaecologie in het Academisch Medisch Centrum. Vanaf januari 2004 heeft zij

haar klinisch werk gecombineerd met het wetenschappelijk onderzoek wat zou uitgroeien

tot dit proefschrift. In januari 2006 startte zij met haar opleiding tot gynaecoloog in

het Flevoziekenhuis (opleider dr. G. Kleiverda) en een jaar later keerde zij terug naar

het AMC (opleider prof. dr. M.J. Heineman). Sinds juli 2010 is zij weer terug in het

Flevoziekenhuis voor het laatste deel van haar opleiding.

Karlijn woont samen met Derk Jan de Boer, zij hebben twee dochters; Meinske en Stella.

UvA-DARE (Digital Academic Repository) … · Hemodynamics and maternal characteristics prior to...

Documents

Transcript of UvA-DARE (Digital Academic Repository) … · Hemodynamics and maternal characteristics prior to...