Recurrence and long-term maternal health risks of hypertensive disorders of pregnancy: a population-based study

Research www.AJOG.org

OBSTETRICS

Recurrence and long-term maternal health risksof hypertensive disorders of pregnancy:a population-based studyAlice B. Andersgaard, MD; Ganesh Acharya, MD, PhD; Ellisiv B. Mathiesen, MD, PhD;Stein H. Johnsen, MD, PhD; Bjørn Straume, MD; Pål Øian, MD, PhD

OBJECTIVE: The purpose of this study was to investigate the recur-rence risk of hypertensive disorders in subsequent pregnancies and toexplore the associations among hypertensive disorders of pregnancyand maternal cardiovascular risk factor profile and the development ofcardiovascular diseases later in life.

STUDY DESIGN: We used population-based, cross-sectional data fromthe fourth survey of the Tromsø Study.

RESULTS: Preeclampsia in the first pregnancy increased the risk of re-currence in later pregnancies (relative risk, 6.6; 95% confidence inter-
val, 5.5–7.9) compared with a normotensive first pregnancy. Women
population-based study. Am J Obstet Gynecol 2012;206:143.e1-8.

imsctmebspsv

preeclampsia. Furthermdoi: 10.1016/j.ajog.2011.09.032

with a history of preeclampsia or nonproteinuric hypertension had anunfavorable cardiovascular risk profile. Hypertension was prevalent in25% and 28% of the women, respectively. The carotid artery intima-media thickness and total carotid plaque area were significantly largerin women with previous preeclampsia.

CONCLUSION: A strong association between hypertensive disorders of preg-nancyandan increasedriskofatherosclerosisandcardiovasculardiseaseswasdemonstratedby theassessmentof risk factors thatcanbepotentiallymodified.

Key words: carotid artery intima-media thickness (IMT), hypertensive
disorders of pregnancy, preeclampsia, recurrence
Cite this article as: Andersgaard AB, Acharya G, Mathiesen EB, et al. Recurrence and long-term maternal health risks of hypertensive disorders of pregnancy: a

Hypertensive disorders of preg-nancy affect approximately 10% of

gestations. Defective placentation andmaternal endothelial cell dysfunction

From Department of Obstetrics andGynecology, Innlandet Hospital Trust,Gjøvik (Dr Andersgaard), and theDepartments of Clinical Medicine (DrsAndersgaard, Acharya, Mathiesen,Johnsen, and Øian) and CommunityMedicine (Dr Straume), University ofTromsø, and the Departments ofObstetrics and Gynecology (Drs Acharyaand Øian) and Neurology andNeurophysiology (Drs Mathiesen andJohnsen), University Hospital of NorthNorway, Tromsø, Norway.

Received May 28, 2011; revised July 27, 2011;accepted Sept. 29, 2011.

Supported in part by a research grant from theResearch Council of Norway (A.B.A.).

The authors report no conflict of interest.

Reprints: Alice Beathe Andersgaard, MD,Department of Obstetrics and Gynecology,Innlandet Hospital Trust, 2819 Gjøvik, [email protected].

0002-9378/$36.00© 2012 Published by Mosby, Inc.

have been hypothesized as the anteced-ents of clinical presentation.1 However,t is unclear whether the underlying

echanism is a constitutional predispo-ition or defective placentation. Thehanges in uteroplacental circulationhat is associated with incomplete re-

odeling of the spiral arteries and thextent of placental vascular lesions varyetween different clinical conditions,uch as gestational hypertension andreeclampsia. Preeclampsia has been as-ociated with an increased risk of cardio-ascular disease (CVD) in later life.2-5

However, the relative risk that is associatedwith different types of hypertensive disor-ders of pregnancy has not been clarified.Despite the epidemiologic evidence of in-creased risk for hypertension, stroke, cor-onary artery disease, and end-stage renaldisease,6-8 a clear pathophysiologic expla-nation remains elusive and mostly hypo-thetic. Although preeclampsia and CVDshare many of the same constitutional riskfactors9 and endothelial dysfunction maypersist after the pregnancy,4,10,11 no studyso farhasdemonstratedunequivocally thatthe risk profile is altered in association with

ore, whether and

FEBRUARY 2012 Americ

for how long the impaired vascular func-tion persists after preeclampsia remainscontroversial. A recent study showed thatthe vascular dysfunction persists 6-24months after delivery only in women withearly-onset preeclampsia, but not inwomen who had late-onset disease,11

whereas the risk of CVD is increased inboth.

Ultrasound-assessed carotid plaquesand intima-media thickness (IMT) areuseful markers of early atherosclerosisand are predictive of future CVD.12-15

Information on associations between hy-pertensive disorders of pregnancy and ca-rotid atherosclerosis is limited, however,to smaller case-control studies and hasshown partly conflicting results.

In this population-based study, weaimed to evaluate the recurrence risk ofhypertensive disorders in subsequentpregnancies in women who report this intheir first pregnancy. Additionally, wewanted to investigate the associationsamong hypertensive disorders of preg-nancy and a maternal cardiovascular riskfactor profile, atherosclerosis of the ca-rotid arteries, and the risk of maternal
CVD later in life.
an Journal of Obstetrics & Gynecology 143.e1

mailto:[email protected]

�spaici

csyinupbbwttwuse

(ptcpSupdguctipfe

es7tt6wwpi

p(mcwcsaWsJfLpFpcwe

Research Obstetrics www.AJOG.org

MATERIALS AND METHODSThe Tromsø Study is a population-basedmultipurpose, single-center study with amain focus on cardiovascular risk factorsand disease. The University of Tromsø,in cooperation with the National HealthScreening Service, conducted the sur-veys. The Tromsø Study consists of dif-ferent surveys that were conducted at 6-to 7-year intervals from 1974. The data inthe present study are from the fourthsurvey (1994-95), which was the firstsurvey that included questions on parityand hypertensive disorders of preg-nancy. The survey had 2 different parts: afirst visit to which everybody �25 yearsold (born before 1970) were invited anda second visit with more extensive exam-inations to which all men who were55-74 years old (born 1920-1939), allwomen who were 50-74 years old (born1920-1944) and smaller random samples(5-8%) were invited.

Study populationA total of 12,862 men and 14,293 womenparticipated (77% of those who were eligi-ble). The female participants were asked toanswer questionnaires on parity and hy-pertensive complications in pregnancies.Women who did not answer a question onparity (n � 1749), nulliparous women (n

1964), parous women who did not an-wer the questions on hypertensive com-lications in their pregnancies (n � 434),nd women who did not consent to partic-pate in the research (n � 172) were ex-luded, which left 9974 women who werencluded in the study (Figure).

The occurrence of hypertensiveomplications in pregnancy was as-essed by the following questions: “Ifou have had high blood pressure dur-ng pregnancy, was it your first preg-ancy?” and “If you have had protein-ria during pregnancy, was it your firstregnancy?” Four groups were definedased on the questionnaire responses;ased on self-reporting by women; womenho had preeclampsia (group I), nonpro-

einuric hypertension (group II), normo-ensive proteinuria (group III), andomen without hypertension or protein-ria (group IV, control group). Compari-ons are made between women with pre-
clampsia (group I) and the control group
143.e2 American Journal of Obstetrics & Gynecolo

group IV) and between women with non-roteinuric hypertension (group II) andhe control group (group IV). The majorauses of proteinuria during pregnancy arereeclampsia and urinary tract infection.creening for proteinuria depends on these of urine dipsticks that may give false-ositive results; contamination by vaginalischarge must be excluded. Women inroup III reported normotensive protein-ria, which is encountered commonly inlinical practice, but the cause of gesta-ional proteinuria could not be ascertainedn this group. It might have been a false-ositive test result or contamination, in-

ection, primary renal disease, or renal dis-ase that resulted from systemic disorders.

The second visit with more extensivexaminations included an ultrasoundcan of the right carotid artery. A total of965 persons (76% of the eligible) at-ended the second visit, and carotid ul-rasound examination was performed in727 men and women. Of the 9974omen who attended the first visit andho fulfilled the inclusion criteria for theresent study, 2524 women were exam-

ned with carotid ultrasound (Figure).

Examination methodsand measurementsThe screening consisted of self-adminis-tered questionnaires, clinical physicalexaminations, and laboratory tests. In-formation about smoking habits, historyof diabetes mellitus, angina pectoris,previous myocardial infarction, familyhistory, and use of medication was col-lected from self-administered question-naires. CVD was defined as previous strokeand/or myocardial infarction and/or pres-ence of angina pectoris. Diabetes mellituswas defined as self-reported diabetes mel-litus and/or regular use of insulin or oralantidiabetic drugs.

Weight and height were measuredwith light clothing without shoes. Bodymass index was calculated as weight inkilograms divided by the square of theheight in meters. Blood pressure (BP)was measured with an automatic device(Dinamap Vital Signs Monitor 1846;Critikon Inc, Tampa, FL) after a 2-min-utes rest in the sitting position and 3readings were taken at 1-minute inter-
vals. The mean of the second and third
gy FEBRUARY 2012

measurements is used in our analyses.Current hypertension was defined as sys-tolic BP �140 mm Hg and/or diastolicBP �90 mm Hg and/or current use ofantihypertensive medication.

A nonfasting blood sample was takenat the first visit and analyzed for serumtotal cholesterol, high-density lipopro-tein cholesterol, and triglycerides. Gly-cosylated hemoglobin was measured atthe second visit in 3088 women.

Carotid ultrasound examinationMethodologic details regarding carotid ar-tery ultrasound scanning and its reproduc-ibility have been published previously.16-18

Briefly, high-resolution B-mode ultra-sonography was performed with a duplexscanner (Acuson Xp10 128, ART-upgraded;Acuson, Mountain View, CA) that wasequipped with a 7.5-MHz linear arraytransducer. Automated measurements ofIMT were performed in 10-mm segmentsin the near and far walls of the distal com-mon carotid artery and the far wall of thebifurcation. The mean IMT from 3 prese-lected images was calculated for each loca-tion, and the mean of these averages wasused in the analyses. The far and near wallsof the right common carotid artery, bifur-cation, and internal carotid artery (6 loca-tions) were scanned for the presence ofplaques. Plaque was defined as a localized

rotrusion of the vessel wall into the lumen�50% compared with the adjacent inti-a-media layer). Still images were re-

orded for each plaque and were digitizedith the Matrox Meteor II frame grabber

ard and Matrox Intellicam software (ver-ion 2.07; Matrox Imaging, Quebec, Can-da) at a resolution of 768 � 576 pixels.

ith the software Adobe Photoshop (ver-ion 7.0; Adobe Systems Incorporated, Sanose, CA), the following steps were per-ormed: each plaque was outlined with theasso tool (Adobe Photoshop), and thelaque area was calculated as pixel values.or the resolution that was used in theresent study, a plaque area of 167 pixelsorresponded to 1 mm2. Total plaque areaas calculated as the sum of all plaque ar-

as from each location.

Statistical analysesAnalyses were performed with STATA
software (Stata Corporation, College

www.AJOG.org Obstetrics Research

FIGUREFlow chart of the study population: the Tromsø Study

The fourth survey of the Tromsø Study included inhabitants of the municipality of Tromsø

n = 27,158

Did not consent to medical research n = 172

Women included

n = 9974

Nulliparous women n = 1964

Had not answered the question on parity

n = 1749

Parous women who had not answered the question on

hypertension and/or proteinuria during pregnancy

n = 434

Group I Women with preeclampsia

n = 901

Group II Women with

nonproteinuric hypertension

n = 607

Group III Women with normotensive

proteinuria

n = 1279

Group IV Women without

hypertension and proteiuria

n = 7187

Women reporting ≥1 child birthn = 10,408

Women

n = 14,293

Ultrasound scanning of the carotid arteries

n = 138


n = 358


n = 1778


n = 250

Andersgaard. Preeclampsia and long-term maternal health risk. Am J Obstet Gynecol 2012.

FEBRUARY 2012 American Journal of Obstetrics & Gynecology 143.e3

TgtIgr

wBg

th ris


Station, TX) and SPSS software (version15.0 for Windows; SPSS Inc, Chicago,IL). Comparisons between groupswere performed with 1-way analysis ofvariance for continuous variables andwith the �2 test for categoric variables.

he statistical analyses included all 4roups. Contrast tests were performedo evaluate differences between groups

and group (control group) androups II and IV (control group),

TABLE 1General characteristics of the stud

General characteristics

Grouppreecl(n � 9

Age, ya 48.8...................................................................................................................

Years from first child born untilthe year of the studyd

25.4

...................................................................................................................

Pulse, beats/mind 75 (7...................................................................................................................

Mean arterial pressure, mm Hgd 100 (9...................................................................................................................

Systolic blood pressure, mm Hgd 138 (1..........................................................................................................

�140 mm Hg, n (%) 374 (4..........................................................................................................

�160 mm Hg, n (%) 171 (1...................................................................................................................

Diastolic blood pressure, mm Hgd 80 (7..........................................................................................................

�90 mm Hg (n) 195 (2..........................................................................................................

�100 mm Hg, n (%) 72 (8...................................................................................................................

Blood samples..........................................................................................................

Total cholesterol, mmol/Ld 6.12..........................................................................................................

High-density lipoproteincholesterol, mmol/Ld

1.61

..........................................................................................................

Triglycerides, mmol/Ld 1.43..........................................................................................................

Glycosylated hemoglobin, %d,f 5.48..........................................................................................................

Glycosylated hemoglobin�6.1%, n (%)f

18 (6

...................................................................................................................

Body mass index, kg/m2d 26 (2..........................................................................................................

�30 kg/m2, n (%) 155 (1...................................................................................................................

Waist circumference, cmd 87 (8...................................................................................................................

Daily cigarette smoking, n (%) 284 (3...................................................................................................................

Family history..........................................................................................................

First-degree relative(s) withcardiovascular disease, n (%)

585 (6

..........................................................................................................

First-degree relative(s) withdiabetes mellitus, n (%)

173 (1

...................................................................................................................

Means of groups I and II are compared with the control groupa Data are given as mean (range); b P � .01 � P � .001; c P

Andersgaard. Preeclampsia and long-term maternal heal

espectively.


EthicsThe Norwegian Data Inspectorate licensedall data. The Regional Committee for Medi-cal Research Ethics approved the study. Allparticipants gave informed written consent.

RESULTSOf the 9974 parous women, 901 women(9.0%) reported preeclampsia in 1 of theirpregnancies (group I); 607 women (6.0%)

roups: the Tromsø Study

reviouspsia

Group II:nonproteinurichypertension inpreviouspregnancy(n � 607)

Group Inormotproteinprevioupregna(n � 12

87)b 45.2 (25–87)c 48.0 (2.........................................................................................................................

4–26.4)e 21.0 (19.9–22.2)c 25.8 (2

.........................................................................................................................

76.1)e 76 (74.5–76.6)b 74 (73.........................................................................................................................

100.7)c 103 (102.0–104.1)c 95 (94.........................................................................................................................

–139.4)c 142 (140.6–143.3)c 131 (13.........................................................................................................................

246 (41) 341 (27.........................................................................................................................

125 (21) 141 (11.........................................................................................................................

80.6)c 83 (81.6–83.4)c 76 (75.........................................................................................................................

144 (24%) 173 (14.........................................................................................................................

59 (10%) 54 (4%.........................................................................................................................

.........................................................................................................................

04–6.19)e 6.13 (6.04–6.22) 6.17.........................................................................................................................

58–1.64)b 1.57 (1.54–1.60)c 1.61

.........................................................................................................................

37–1.48)c 1.46 (1.39–1.52)c 1.41.........................................................................................................................

37–5.59) 5.47 (5.38–5.56) 5.48.........................................................................................................................

17 (9.9) 28 (6.5

.........................................................................................................................

26.3)c 27 (26.2–26.8)c 25 (24.........................................................................................................................

123 (21) 165 (13.........................................................................................................................

9)c 90 (88–92)c 86 (85.........................................................................................................................

184 (31) 536 (42.........................................................................................................................

.........................................................................................................................

374 (61.6) 760 (59

.........................................................................................................................

106 (17.5) 238 (18

.........................................................................................................................

r by contrasts in analysis of variance.

.001; d Data are given as mean (95% CI); e P � .05 � P � .01;

k. Am J Obstet Gynecol 2012.

reported nonproteinuric hypertension s

gy FEBRUARY 2012

(group II); 1279 women (12.8%) reportednormotensive proteinuria (group III), and7187 women (72.2%) reported neither hy-pertension nor proteinuria (group IV[control group]; Figure).

General characteristics of the studypopulation and the results of physical ex-aminations are summarized in Table 1,

here all comparisons are age-adjusted.oth the preeclamptic group and theroup with nonproteinuric hyperten-

ivein

Group IV: nohypertension andno proteinuria inpreviouspregnancies(n � 7187) P value

90) 47.4 (25–94) � .001..................................................................................................................

–26.6) 24.0 (23.6–24.3) � .001

..................................................................................................................

74.6) 74 (73.8–74.7) .001..................................................................................................................

95.4) 94 (94.0–94.6) � .001..................................................................................................................

132.0) 130 (129.8–130.6) � .001..................................................................................................................

1817 (25)..................................................................................................................

762 (11)..................................................................................................................

76.5) 75 (75.1–75.7) � .001..................................................................................................................

864 (12%)..................................................................................................................

286 (4%)..................................................................................................................

..................................................................................................................

1–6.23) 6.04 (6.01–6.07) � .001..................................................................................................................

9–1.64) 1.65 (1.64–1.66) � .001

..................................................................................................................

4–1.46) 1.33 (1.31–1.34) � .001..................................................................................................................

2–5.54) 5.43 (5.40–5.45) .074..................................................................................................................

115 (5.2) .603

..................................................................................................................

25.3) 25 (24.4–24.6) � .001..................................................................................................................

699 (10)..................................................................................................................

) 84 (84–85) � .001..................................................................................................................

2710 (38) � .001..................................................................................................................

..................................................................................................................

3938 (54.8) � .001

..................................................................................................................

1161 (16.2) .034

..................................................................................................................

asured in 3088 women.

y g

I: pam01)

II:ensuriasncy79)

(25– 5–......... .........

(24. 5.0

......... .........

4.5– .4–......... .........

8.0– .0–......... .........

37.2 0.1–......... .........

2) )......... .........

9) )......... .........

9.2– .3–......... .........

2%) %)......... .........

%) )......... .........

......... .........

(6. (6.1......... .........

(1. (1.5

......... .........

(1. (1.3......... .........

(5. (5.4......... .........

.5) )

......... .........

5.8– .8–......... .........

7) )......... .........

6–8 –87......... .........

2) )......... .........

......... .........

4.9) .4)

......... .........

9.2) .6)

......... .........

teste

� f Me

ion had a significantly higher mean ar-

dcifgwlsgGs

tarf

wT1tdpttp

tiHwsti

fgg

epptapsw

td1owpA

th ris


terial, systolic, and diastolic BP thanthe control group. The proportion ofwomen with systolic BP �160 mm Hg ordiastolic BP �100 mm Hg was almost

ouble in both groups with hypertensiveomplications in pregnancy. Triglycer-des, body mass index, and waist circum-erence were significantly higher inroups I and II than in the control group,hereas high-density lipoprotein cho-

esterol was lower. Total cholesterol wasignificantly higher in the preeclampticroup compared with the control group.lycosylated hemoglobin showed no

ignificant difference among the groups.Women with preeclampsia and gesta-

ional hypertension more often reportedfamily history of stroke or brain hemor-

hage, angina pectoris, and myocardial in-arction. A family history of diabetes mel-

TABLE 2The health status of the study popu

Current health situation

Group I: previoupreeclampsia(n � 901)

Blood pressure �140/90mm Hg or currentlyantihypertensive therapy,n (%)

223 (25)

...................................................................................................................

Cardiovascular disease:angina pectoris and/ormyocardial infarctionand/or stroke, n (%)

69 (7.7)

...................................................................................................................

Diabetes mellitus, n (%) 17 (1.9)...................................................................................................................


TABLE 3Age-adjusted results of the carotid

Variable

Group I: previoupreeclampsia(n � 250)

Presence of carotidplaques, n (%)

127 (51)

...................................................................................................................

Total carotid plaque area,mm2b

10.00a (8.24–11

...................................................................................................................

Mean intima-media-thickness, mm

0.86 (0.84–0.8

...................................................................................................................

Means of groups I and II are compared with the control groupa P � .05 � P � .01; b Data in parentheses are 95% confid

Andersgaard. Preeclampsia and long-term maternal health ris

litus showed only a higher prevalence ingroup I.

Sixty-two percent (n � 6138 women)ere current smokers or had smoked.he mean age of starting smoking was9.2 years; 80% started to smoke beforeheir first delivery. There were feweraily cigarette smokers in group I, com-ared with the control group, beforehe first delivery. We have no informa-ion on the habits of smoking duringregnancy.Table 2 presents data on health status of

he study population at the time of screen-ng. The prevalence of BP �140/90 mm

g or use of antihypertensive medicationere 25% and 28% in groups I and II, re-

pectively, and was significantly higherhan in the control group (13%). CVD thatncluded angina pectoris, myocardial in-

tion at the screening: the Tromsø Stu

Group II: nonproteinurichypertension inprevious pregnancy(n � 607)

Group III: normoteproteinuria in prepregnancy(n � 1279)

167 (28) 190 (15)

.........................................................................................................................

43 (7.1) 73 (5.7)

.........................................................................................................................

13 (2.1) 33 (2.6).........................................................................................................................


trasound examinations: the Tromsø S

Group II: nonproteinurichypertension inprevious pregnancy(n � 138)

Group III: normoproteinuria in prpregnancy(n � 358)

74 (53) 154 (43)

.........................................................................................................................

) 10.70c (8.05–13.36) 8.18 (6.67–9.6

.........................................................................................................................

0.84 (0.81–0.88)a 0.82 (0.80–0.8

.........................................................................................................................

r by contrasts in analysis of variance.

interval; c P � .001; d P � .01 � P � .001.



arction, and cerebral stroke was higher inroups I and II compared with controlroup (group IV).

The results of the carotid ultrasoundxaminations are shown in Table 3. Bothlaque prevalence and the total carotidlaque area were significantly larger inhe groups with previous preeclampsiand nonproteinuric hypertension com-ared with the control group. IMT wasignificantly larger in group I comparedith group IV.Of the women who were included in

he study, 21% had given birth to 2 chil-ren, and 79% had �2 children (range,–13 children). The reported incidencef preeclampsia in the first pregnancyas 6.7% (n � 672 women) and in laterregnancies was 4.7% (n � 471 women).total of 1.3 % (n � 131 women) re-

ives

Group IV: no hypertensionand no proteinuria inprevious pregnancies(n � 7187) P value

913 (13) � .001

..................................................................................................................

305 (4.2) � .001

..................................................................................................................

107 (1.5) .029..................................................................................................................

dy

iveus

Group IV: no hypertensionand no proteinuria inprevious pregnancies(n � 1778) P value

751 (42) .018

..................................................................................................................

7.09 (6.50–7.67) .001

..................................................................................................................

0.82 (0.81–0.83) .001

..................................................................................................................

la dy

sns

viou

......... .........

......... .........

......... .........

ul tu

stensevio

......... .........

.77 9)

......... .........

9)d 4)

......... .........

teste

ence


hhrsetcvtodeghpud

ascdebrtetCbneci2cnnoecccnseimwes

mpItrbpit

th ris


ported preeclampsia in both first andlater pregnancies. The risk of having hy-pertensive complications in later preg-nancies, according to the outcome in thefirst pregnancy, is given in Table 4. Pre-eclampsia in the first pregnancy in-creased the risk of recurrence in laterpregnancies by 6.6-fold compared with anormotensive first pregnancy.

COMMENTOur main findings are that the womenwith a history of hypertensive disorder ofpregnancy have an unfavorable cardio-vascular risk profile, a higher frequencyof CVD and hypertension, and more ex-tensive carotid atherosclerosis. The in-creased risk of CVD among women witha history of preeclampsia is well-de-scribed.2,3,6,7 We show that these women

ave an unfavorable risk profile based onistory, physical examination, blood testesults, and carotid artery ultrasoundcanning. Our findings indicate that pre-clampsia alters the CVD risk profile ei-her by altering the maternal physiologicondition or by unmasking preexistingulnerability that is related to constitu-ional predisposition. In our study, thebjectively assessed CVD risk profile wasifferent among women who had pre-clampsia, gestational hypertension, orestational proteinuria that supports theypothesis that differences in the clinicalresentation might be due to differentnderlying mechanisms that may haveifferent implications for later CVD.19

Women with a history of preeclampsiahad double the risk of hypertension andcoronary artery disease compared withcontrol subjects. They had carotid plaquesmore often, had larger total carotid plaque

TABLE 4The recurrence risk of hypertensiv

Outcome in later pregnancies

Preeclampsia in later pregnancies...................................................................................................................

Nonproteinuric hypertension in later pregnanc...................................................................................................................

Data are given as relative risk (95% confidence interval).


area and thicker intima-media layer com-


pared with control subjects. Family historyof CVD was more common among thesewomen, which suggested that familial riskmay be associated with underlying geneticpredisposition to vascular dysfunction orother related factors (such as familial foodhabits, life style).

We found significantly higher plaqueprevalence and larger total carotid plaquerea in women with previous preeclamp-ia and nonproteinuric hypertension,ompared with the control group. In ad-ition, the IMT was increased in pre-clamptic group. The linear relationshipoth between traditional cardiovascularisk factors and IMT and plaque and be-ween IMT, plaque, and cardiovascularvents found in several studies supportshe use of IMT as a surrogate marker ofVD.13,20,21 Knowledge of associationsetween hypertensive disorders of preg-ancy and objective measurements of ath-rosclerosis is scarce and conflicting. In aross-sectional study, Blaauw et al22 foundncreased femoral, but not carotid, IMT in2 primiparous preeclamptic womenompared with similar-sized groups ofulliparous and primiparous women withormal pregnancies. A case-control studyf 20 women with severe early-onset pre-clampsia and 20 age- and parity-matchedontrol subjects, carotid IMT was in-reased in preeclamptic women who re-eived antihypertensive medication, butot in the total preeclamptic group.23 In atudy of 35 women with previous pre-clampsia and 61 women with pregnancy-nduced hypertension without microalbu-

inuria, plaque presence, but not IMT,as associated significantly with pre-

clampsia, compared with healthy controlubjects.24 Early intima-media thickening

omplications in later pregnancies: th

First pregnancy

Preeclampsia(n � 464)

Nonproteinurichypertension(n � 346)

6.6 (5.5–7.9) 1.2 (0.8–1.9).........................................................................................................................

4.1 (3.1–5.5) 7.5 (5.8–9.6).........................................................................................................................


is thought primarily to reflect a hyperten-

gy FEBRUARY 2012

sive hypertrophic response,25,26 whichay explain the relationship between hy-

ertensive disorders of pregnancy andMT. Interestingly, we show, for the firstime, a clear association between pregnancy-elated hypertensive disorders and plaqueurden. Compared with early IMT changes,laqueformationmayrepresenta laterman-

fest stage of atherosclerosis and a closer rela-ionship to clinical vascular disease.15

Family history is a significant indepen-dent risk factor of coronary heart disease.In our study, women who previously hadhypertensive disorders of pregnancy re-ported a higher frequency of family his-tory of stroke, hypertension, angina pec-toris, and myocardial infarction, whichsupports that there are similarities in riskprofiles between preeclampsia and CVD.However, the women with nonprotein-uric hypertension had a higher risk scorefor most parameters compared with thepreeclamptic group. Genetic predisposi-tion to preeclampsia is suggested throughobservational studies27; Sattar and Greer28

reported the probability that predisposi-tion to vascular disease underlies pre-eclampsia, rather than being the cause ofCVD.Inastudyof217womenwhounder-went coronary angiography 36.8 � 9.9years after their first pregnancy, Valdes etal8 concluded that the women who had hy-pertension during pregnancy experiencedcoronary artery disease earlier perhaps be-cause of gestational modification of preex-isting intermediate risk.

In our study, both the preeclamptic andthe nonproteinuric hypertension groupshad an unfavorable cardiovascular riskprofile; 25% and 28% of them, respec-tively, had current hypertension. The BPwas higher in women who had nonpro-

romsø Study

Normotensiveproteinuria(n � 533)

Without hypertensionand proteinuria(n � 6545)

1.1 (0.7–1.6) Reference..................................................................................................................

2.2 (1.5–3.2) Reference..................................................................................................................

e c e T

......... .........

ies......... .........

teinuric hypertension in their pregnancy

vied

(Tcpehgd

tCm


compared with those who had preeclamp-sia, which is similar to the observations ofPouta et al.29 This is interesting becauseascular dysfunction seems to persist onlyn women with severe early-onset pre-clampsia, but not in those with late-onsetisease.11 This emphasizes the importance

of including women with nonproteinurichypertension in the follow-up evaluationsand in future studies that will evaluate therisk of CVD.

The results of this study support thefindings that smoking reduces the risk ofpreeclampsia. This protective effect ofsmoking is not fully understood, but 1 the-ory is that smoking lowers the risk throughan effect on angiogenic factors, endothelialfunction, and the immune system.30 An-other possibility is that induction of anti-oxidant genes and the hypoxemia-relatedpromotion of trophoblast invasion.31

We found that the risk of recurrence ofhypertensive disorders of pregnancy issubstantial in subsequent pregnancies,which is in line with previous reports.23

The relative risk of preeclampsia in a sec-ond pregnancy was 6.6 (95% confidenceinterval [CI], 5.5–7.9) after preeclampsiaand 1.2 (95% CI, 0.8–1.9) after nonpro-teinuric hypertension in the first preg-nancy. This is lower than reported in aNorwegian study by Lie et al32 who found a12-fold increase in the risk of preeclampsiain a second pregnancy when the womanhad preeclampsia in her first pregnancy.This discrepancy might be explained bythe differences in study design.

The main limitations of our study arethat (1) there was no strict definition of thelevel of the BP considered pathologic dur-ing the pregnancy, (2) the occurrence ofhypertensive complications was based onself-reporting by women years after theirpregnancy (36% were �50 years old), and3) the clinical diagnosis was not validated.herefore, recall bias might be a valid con-ern. This might also be reflected in the ap-arently higher reported incidence of pre-clampsia (9.0%) than of nonproteinuricypertension (6.0%), which suggests areater likelihood of recall of more seriousiseases. A recent study by Coolman et al33

reported that 76 of 152 women who self-reported preeclampsia a mere 2 months af-ter delivery appeared not to have had the
disease according to the definitions of pre-
eclampsia; they concluded that the validityof maternal-recall self-reported pre-eclampsia is moderate. The reliability ofself-reported information on CVD hasbeen found to be valid; in a populationstudy in Norway, 84% of the myocardialinfarction and 73% of the angina cases ac-cording to the questionnaire were verifiedby medical records.34 Another limitation ishat the information on family history ofVD is based on questionnaires. The Fra-ingham Offspring study35 found some

value in obtaining family history, but theinformation might not be accurate.

Several epidemiologic studies have linkedpreeclampsia with later cardiovascularevents. This study provides further objec-tive evidence for an unfavorable cardiovas-cular risk profile in women who had hy-pertensive disorders of pregnancy that wasshown by anthropometry, physical exam-ination, serum lipid profile, and carotid ar-tery ultrasound scanning. Recognition ofthe association between hypertensive dis-orders of pregnancy and future CVDshould promote strategies to reduce therisk of CVD. Women with preeclampsiaand with nonproteinuric hypertensionshould be informed about recurrence riskin future pregnancies and advised to con-trol their BP after pregnancy and receivelifestyle counseling. f

REFERENCES1. Redman CW, Sacks GP, Sargent IL. Pre-eclampsia: an excessive maternal inflammatoryresponse to pregnancy. Am J Obstet Gynecol1999;180:499-506.2. Bellamy L, Casas JP, Hingorani AD, WilliamsDJ. Pre-eclampsia and risk of cardiovasculardisease and cancer in later life: systematic re-view and meta-analysis. BMJ 2007;335:974.3. McDonald SD, Malinowski A, Zhou Q, YusufS, Devereaux PJ. Cardiovascular sequelae ofpreeclampsia/eclampsia: a systematic reviewand meta-analyses. Am Heart J 2008;156:918-30.4. Ramsay JE, Stewart F, Greer IA, Sattar N.Microvascular dysfunction: a link between pre-eclampsia and maternal coronary heart disease.BJOG 2003;110:1029-31.5. Jonsdottir LS, Arngrimsson R, Geirsson RT,Sigvaldason H, Sigfusson N. Death rates fromischemic heart disease in women with a historyof hypertension in pregnancy. Acta Obstet Gy-necol Scand 1995;74:772-6.6. Smith GC, Pell JP, Walsh D. Pregnancy com-plications and maternal risk of ischaemic heartdisease: a retrospective cohort study of 129,290
births. Lancet 2001;357:2002-6.

7. Wilson BJ, Watson MS, Prescott GJ, et al.Hypertensive diseases of pregnancy and risk ofhypertension and stroke in later life: results fromcohort study. BMJ 2003;326:845.8. Valdes G, Quezada F, Marchant E, et al. As-sociation of remote hypertension in pregnancywith coronary artery disease: a case-controlstudy. Hypertension 2009;53:733-8.9. Berends AL, de Groot CJ, Sijbrands EJ, et al.Shared constitutional risks for maternal vascu-lar-related pregnancy complications and futurecardiovascular disease. Hypertension 2008;51:1034-41.10. Agatisa PK, Ness RB, Roberts JM, Costan-tino JP, Kuller LH, McLaughlin MK. Impairmentof endothelial function in women with a historyof preeclampsia: an indicator of cardiovascularrisk. Am J Physiol Heart Circ Physiol 2004;286:H1389-93.11. Yinon Y, Kingdom JC, Odutayo A, et al.Vascular dysfunction in women with a history ofpreeclampsia and intrauterine growth restric-tion: insights into future vascular risk. Circula-tion 2010;122:1846-53.12. Johnsen SH, Mathiesen EB, Joakimsen O,et al. Carotid atherosclerosis is a stronger pre-dictor of myocardial infarction in women than inmen: a 6-year follow-up study of 6226 persons:the Tromsø Study. Stroke 2007;38:2873-80.13. Lorenz MW, Markus HS, Bots ML, RosvallM, Sitzer M. Prediction of clinical cardiovascularevents with carotid intima-media thickness: asystematic review and meta-analysis. Circula-tion 2007;115:459-67.14. Mathiesen EB, Johnsen SH, Wilsgaard T,Bønaa KH, Løchen ML, Njølstad I. Carotidplaque area and intima-media thickness in pre-diction of first-ever ischemic stroke: a 10-yearfollow-up of 6584 men and women: the TromsøStudy. Stroke 2011;42:972-8.15. Spence JD, Hackam DG. Treating arteriesinstead of risk factors: a paradigm change inmanagement of atherosclerosis. Stroke 2010;41:1193-9.16. Fosse E, Johnsen SH, Stensland-BuggeE, et al. Repeated visual and computer-assistedcarotid plaque characterization in a longitudinalpopulation-based ultrasound study: the Tromsøstudy. Ultrasound Med Biol 2006;32:3-11.17. Joakimsen O, Bønaa KH, Stensland-BuggeE. Reproducibility of ultrasound assessment ofcarotid plaque occurrence, thickness, and mor-phology: the Tromsø Study. Stroke 1997;28:2201-7.18. Stensland-Bugge E, Bønaa KH, JoakimsenO. Reproducibility of ultrasonographically deter-mined intima-media thickness is dependent onarterial wall thickness: the Tromsø Study.Stroke 1997;28:1972-80.19. Craici IM, Wagner SJ, Hayman SR, GarovicVD. Pre-eclamptic pregnancies: an opportunityto identify women at risk for future cardiovascu-lar disease. Womens Health (Lond Engl)2008;4:133-5.20. O’Leary DH, Polak JF, Kronmal RA, ManolioTA, Burke GL, Wolfson SK Jr. Carotid-artery
intima and media thickness as a risk factor for

https://www.researchgate.net/publication/14666054_Death_Rates_From_Ischemic_Heart_Disease_in_Women_With_a_History_of_Hypertension_in_Pregnancy?el=1_x_8&enrichId=rgreq-05cf466e-eaf1-4791-993f-e1e33bd01056&enrichSource=Y292ZXJQYWdlOzUxNzU2MDE5O0FTOjEwMzEyMjUwNDUxOTY5OUAxNDAxNTk3NzMwMDQy





https://www.researchgate.net/publication/23628775_Cardiovascular_sequleae_of_preeclampsiaeclampsia_a_systematic_review_and_meta-analyses_Am_Heart_J?el=1_x_8&enrichId=rgreq-05cf466e-eaf1-4791-993f-e1e33bd01056&enrichSource=Y292ZXJQYWdlOzUxNzU2MDE5O0FTOjEwMzEyMjUwNDUxOTY5OUAxNDAxNTk3NzMwMDQy





https://www.researchgate.net/publication/11903244_Pregnancy_complications_and_maternal_risk_of_ischemic_heart_disease_A_retrospective_cohort_study_of_129290_births?el=1_x_8&enrichId=rgreq-05cf466e-eaf1-4791-993f-e1e33bd01056&enrichSource=Y292ZXJQYWdlOzUxNzU2MDE5O0FTOjEwMzEyMjUwNDUxOTY5OUAxNDAxNTk3NzMwMDQy




https://www.researchgate.net/publication/43299339_Regarding_Article_Treating_Arteries_Instead_of_Risk_Factors_A_Paradigm_Change_in_Management_of_Atherosclerosis_Response?el=1_x_8&enrichId=rgreq-05cf466e-eaf1-4791-993f-e1e33bd01056&enrichSource=Y292ZXJQYWdlOzUxNzU2MDE5O0FTOjEwMzEyMjUwNDUxOTY5OUAxNDAxNTk3NzMwMDQy




https://www.researchgate.net/publication/291190853_Pre-eclampsia_and_risk_of_cardiovascular_disease_and_cancer_in_later_life_Systematic_review_and_meta-analysis?el=1_x_8&enrichId=rgreq-05cf466e-eaf1-4791-993f-e1e33bd01056&enrichSource=Y292ZXJQYWdlOzUxNzU2MDE5O0FTOjEwMzEyMjUwNDUxOTY5OUAxNDAxNTk3NzMwMDQy





myocardial infarction and stroke in older adults:Cardiovascular Health Study CollaborativeResearch Group. N Engl J Med 1999;340:14-22.21. Stensland-Bugge E, Bønaa KH, JoakimsenO, Njølstad I. Sex differences in the relationshipof risk factors to subclinical carotid atheroscle-rosis measured 15 years later: the Tromsøstudy. Stroke 2000;31:574-81.22. Blaauw J, van Pampus MG, van DoormaalJJ, et al. Increased intima-media thickness afterearly-onset preeclampsia. Obstet Gynecol2006;107:1345-51.23. Gaugler-Senden IP, Berends AL, de GrootCJ, Steegers EA. Severe, very early onset pre-eclampsia: subsequent pregnancies and futureparental cardiovascular health. Eur J ObstetGynecol Reprod Biol 2008;140:171-7.24. Haukkamaa L, Moilanen L, Kattainen A, etal. Pre-eclampsia is a risk factor of carotid arteryatherosclerosis. Cerebrovasc Dis 2009;27:599-607.25. Chapman CM, Palmer LJ, McQuillan BM, et
al. Polymorphisms in the angiotensinogen gene

are associated with carotid intimal-medialthickening in females from a community-basedpopulation. Atherosclerosis 2001;159:209-17.26. Tabara Y, Kohara K, Nakura J, Miki T. Riskfactor-gene interaction in carotid atherosclero-sis: effect of gene polymorphisms of renin-ang-iotensin system. J Hum Genet 2001;46:278-84.27. Arngrimsson R, Bjørnsson S, Geirsson RT,Bjørnsson H, Walker JJ, Snaedal G. Geneticand familial predisposition to eclampsia andpre-eclampsia in a defined population. BJOG1990;97:762-9.28. Sattar N, Greer IA. Pregnancy complica-tions and maternal cardiovascular risk: oppor-tunities for intervention and screening? BMJ2002;325:157-60.29. Pouta A, Hartikainen AL, Sovio U, et al.Manifestations of metabolic syndrome after hy-pertensive pregnancy. Hypertension 2004;43:825-31.30. England L, Zhang J. Smoking and risk ofpreeclampsia: a systematic review. Front Biosci
2007;12:2471-83.
gy FEBRUARY 2012

31. Genbacev O, Zhou Y, Ludlow JW, FisherSJ. Regulation of human placental develop-ment by oxygen tension. Science 1997;277:1669-72.32. Lie RT, Rasmussen S, Brunborg H, Gjess-ing HK, Lie-Nielsen E, Irgens LM. Fetal and ma-ternal contributions to risk of pre-eclampsia:population based study. BMJ 1998;316:1343-7.33. Coolman M, de Groot CJ, Jaddoe VW, Hof-man A, Raat H, Steegers EA. Medical recordvalidation of maternally reported history of pre-eclampsia. J Clin Epidemiol 2010;63:932-7.34. Tretli S, Lund-Larsen PG, Foss OP. Reliabil-ity of questionnaire information on cardiovascu-lar disease and diabetes: cardiovascular dis-ease study in Finnmark county. J EpidemiolCommunity Health 1982;36:269-73.35. Murabito JM, Nam BH, D’Agostino RB Sr,Lloyd-Jones DM, O’Donnell CJ, Wilson PW.Accuracy of offspring reports of parental cardio-vascular disease history: the Framingham Off-spring Study. Ann Intern Med 2004;140:
434-40.
https://www.researchgate.net/publication/6671490_Smoking_and_risk_of_preeclampsia_A_systematic_review?el=1_x_8&enrichId=rgreq-05cf466e-eaf1-4791-993f-e1e33bd01056&enrichSource=Y292ZXJQYWdlOzUxNzU2MDE5O0FTOjEwMzEyMjUwNDUxOTY5OUAxNDAxNTk3NzMwMDQy



https://www.researchgate.net/publication/5284316_Severe_very_early_onset_preeclampsia_Subsequent_pregnancies_and_future_parental_cardiovascular_health?el=1_x_8&enrichId=rgreq-05cf466e-eaf1-4791-993f-e1e33bd01056&enrichSource=Y292ZXJQYWdlOzUxNzU2MDE5O0FTOjEwMzEyMjUwNDUxOTY5OUAxNDAxNTk3NzMwMDQy





https://www.researchgate.net/publication/13934213_Regulation_of_human_placental_development_by_oxygen_tension_Science?el=1_x_8&enrichId=rgreq-05cf466e-eaf1-4791-993f-e1e33bd01056&enrichSource=Y292ZXJQYWdlOzUxNzU2MDE5O0FTOjEwMzEyMjUwNDUxOTY5OUAxNDAxNTk3NzMwMDQy




https://www.researchgate.net/publication/16031761_Reliability_of_questionnaire_information_on_cardiovascular_disease_and_diabetes_Cardiovascular_disease_study_in_Finnmark_County?el=1_x_8&enrichId=rgreq-05cf466e-eaf1-4791-993f-e1e33bd01056&enrichSource=Y292ZXJQYWdlOzUxNzU2MDE5O0FTOjEwMzEyMjUwNDUxOTY5OUAxNDAxNTk3NzMwMDQy





https://www.researchgate.net/publication/12614193_Sex_Differences_in_the_Relationship_of_Risk_Factors_to_Subclinical_Carotid_Atherosclerosis_Measured_15_Years_Later_The_Tromso_Study?el=1_x_8&enrichId=rgreq-05cf466e-eaf1-4791-993f-e1e33bd01056&enrichSource=Y292ZXJQYWdlOzUxNzU2MDE5O0FTOjEwMzEyMjUwNDUxOTY5OUAxNDAxNTk3NzMwMDQy





https://www.researchgate.net/publication/298978139_Accuracy_of_offspring_reports_of_parental_cardiovascular_disease_history_The_Framingham_offspring_study?el=1_x_8&enrichId=rgreq-05cf466e-eaf1-4791-993f-e1e33bd01056&enrichSource=Y292ZXJQYWdlOzUxNzU2MDE5O0FTOjEwMzEyMjUwNDUxOTY5OUAxNDAxNTk3NzMwMDQy






Recurrence and long-term maternal health risks of hypertensive disorders of pregnancy: a population-based study

Documents