Pregnancy Hypertension 14 (2018) 9–14

Contents lists available at ScienceDirect

Pregnancy Hypertension
journal homepage: www.elsevier.com/locate/preghy

Long-term follow-up of women with early onset pre-eclampsia shows T

subclinical impairment of the left ventricular function by two-dimensional
speckle tracking echocardiography
⁎
Tor Skibsted Clemmensena, ,1, Martin Christensenb,c,1, Camilla Jensenius Skovhus Kronborgd,
Ulla Breth Knudsenb,e, Brian Bridal Løgstrupa
a
Department of Cardiology, Aarhus University Hospital, Skejby, Denmark
b
Institute of Clinical Medicine, Aarhus University, Denmark
c
Department of Gynecology and Obstetrics, Randers Regional Hospital, Randers, Denmark
d
Department of Oncology, Aarhus University Hospital, Aarhus, Denmark
e
Department of Gynecology and Obstetrics, Horsens Regional Hospital, Denmark

A R T I C LE I N FO A B S T R A C T

Keywords: Objectives: This study compares differences in the long-term myocardial function between women with early
Pregnancy (EOPE) and late onset preeclampsia (LOPE) and age matched normotensive controls using two-dimensional
Preeclampsia speckle tracking echocardiography.
Heart failure Methods: The study population comprised 93 women who gave birth at Department of Gynecology and
Global longitudinal systolic function
Obstetrics, Randers Regional Hospital between 1998 and 2008. The women were grouped as EOPE (n = 31),
Speckle-tracking echocardiography
LOPE (n = 22), and women with previous normotensive pregnancies (n = 40). All women underwent com-
prehensive blinded echocardiographic assessment of myocardial function.
Results: The median time since delivery was 12 years [9;15]. Left ventricular (LV) ejection fraction did not differ
between groups. In contrast, LV longitudinal systolic myocardial function by LV global longitudinal strain
(LVGLS) magnitude was significantly lower in EOPE women than controls (−18 ± 3% versus −21 ± 2%,
p < 0.001) and LOPE women (−18 ± 3% versus −21 ± 2%, p < 0.01). In alignment with systolic para-
meters, the diastolic filling pattern indicated more restrictive filling pattern in EOPE women than controls and
LOPE women. Thus, EOPE women had lower septal e′ velocities leading to lower mean e′ and subsequently
higher E/e′ ratio (p < 0.01) than controls and LOPE women. LVGLS was the echocardiographic parameter with
the strongest association with EOPE in ROC curves.
Conclusions: Women with a history of EOPE are more likely to have subclinical impairment of left ventricular
function 12 years after PE than are those with a history of LOPE and controls. LVGLS was the echocardiographic
parameter with the strongest association with EOPE.

1. Introduction cardiovascular risk profile to be different in two subgroups i.e. early

Preeclampsia (PE) affects 5–8% of all pregnancies [1]. Still, the
clinical severity of PE varies immensely, which furthermore leads to
guideline differences in PE severity definitions [2]. It is well established
that women with a history of PE have increased cardiovascular disease
(CVD) risk later in life [3–5]. However, studies suggest the
onset PE (EOPE: onset before 34 weeks) and late onset PE (LOPE: onset
at or after 34 weeks) [6]. Systematic reviews support the epidemiolo-
gical findings and demonstrate an approximately doubled risk of is-
chemic heart disease, cerebrovascular incidents and mortality of car-
diovascular disease after PE [1]. The gestational age at PE onset has
been shown to be negatively associated with markers of subclinical

Abbreviations: BMI, Body mass index; CVD, Cardiovascular disease; DM, Diabetes mellitus; EOPE, Early onset PE; PWT, Infero-lateral wall thickness; IVS,
Interventricular septum; LOPE, Late onset PE; LV, Left ventricular; LVEF, Left ventricular ejection fraction; LVGLS, Left ventricular global longitudinal strain; LVID,
Left ventricular internal diameter; e′, Peak diastolic mitral annular velocities; S′, Peak systolic mitral annular velocities; PE, Preeclampsia
⁎
Corresponding author.
E-mail address: torclemm@rm.dk (T.S. Clemmensen).
1
The authors has contributed equally to the paper and share the first authorship.

https://doi.org/10.1016/j.preghy.2018.07.001
Received 14 February 2018; Received in revised form 21 June 2018; Accepted 5 July 2018
Available online 06 July 2018
2210-7789/ © 2018 International Society for the Study of Hypertension in Pregnancy. Published by Elsevier B.V. All rights reserved.
T.S. Clemmensen et al.
atherosclerosis [7]. However, the association between PE and the pre-
mature CVD is incompletely understood. Studies have shown that PE is
a vascular disease related to traditional CVD risk factors such as hy-
pertension, dyslipidemia and obesity. Nonetheless, how early the CVD
develop and the underlying hemodynamic explanation and ways to
evaluate these still remains unresolved.
Solid evidence supports echocardiography as a valuable tool to
stratify risk and to guide management and counseling in the preclinical
and clinical phases of gestational hypertension and PE. Changes in
cardiac function and morphology are recognizable at an asymptomatic
early stage and correlate with disease severity and adverse outcomes
[8]. PE is associated with left ventricular (LV) hypertrophy and LV
diastolic dysfunction when compared to normotensive individuals
[9,10].
However, traditional LV systolic parameters have shown limited
value in the evaluation of myocardial dysfunction long-term post PE
despite abnormal loading conditions [11]. Therefore, there is a need for
novel and sensitive parameters of myocardial dysfunction in the long
term follow-up of women with previous PE. The introduction of two-
dimensional (2D) speckle-tracking echocardiography (STE) has enabled
LV global longitudinal strain (GLS) assessment as a useful and robust
marker of systolic myocardial function. Notably, LV GLS proves a va-
luable tool in detecting subclinical myocardial dysfunction despite
normal LV ejection fraction (EF) [12,13]. Therefore, recent echo-
cardiographic guidelines recommends routine LV GLS assessment while
evaluating systolic myocardial performance [14].
Our objective was to compare the myocardial function between
groups (EOPE vs. LOPE vs. normotensive controls) 12 years after index
pregnancy using STE to assess the LV function.
2. Methods
2.1. Study population
Present study was based on an observational cohort study recruiting
women giving birth at Department of Gynecology and Obstetrics,
Randers Regional Hospital 1998–2008 [7]. Originally, a total of 185
(75 controls, 55 LOPE and 55 EOPE) eligible participants were identi-
fied in The National Patient Registry and The Medical Birth Registry
using relevant ICD-10 diagnosis codes resulting 98 participating
women. Subsequently, 93 women were included in present study after
having provided written informed consent in pursuance of the Helsinki
Declaration principles. Of the 5 non-participating women, one had
undergone breast cancer surgery, one was pregnant and 3 refused to
participate. These women were grouped as early onset PE (EOPE,
n = 31), late onset PE (LOPE, n = 22), and controls (n = 40).
Exclusion criteria at follow-up were pregnancy, breastfeeding, nat-
ural menopause or address more than 100 km away from the study site.
All eligible women with late-onset preeclampsia and with normotensive
pregnancies were identified to match eligible early-onset women
(age ± 2 years) and time since delivery ( ± 1 year) and then randomly
selected for participation. The women with normotensive pregnancies
were randomly selected from eligible matched unexposed women in the
source population. Women with normotensive pregnancies were ex-
cluded if they before index pregnancy experienced gestational hy-
pertension, preeclampsia, HELLP syndrome or eclampsia. No women
with EOPE or LOPE had chronic hypertension during pregnancy. All
eligible women were invited to participate in the current study by mail.
The study protocol was approved by the local Medical Ethics
Committee and The Danish Data Protection Agency. The study was
registered at: www.clinicaltrials.gov as NCT02337049.
2.2. Preeclampsia definitions
During 1998–2008, preeclampsia diagnosis was defined as new-
onset hypertension (systolic blood pressure ≥ 140 mmHg and/or
10
Pregnancy Hypertension 14 (2018) 9–14
diastolic blood pressure ≥ 90 mmHg) and proteinuria (≥300 mg/24 h
or ≥30 mg/mmol albumin: creatinine random urine or ≥1 + on repeat
dipstick). Early-onset preeclampsia was defined as symptoms occurring
before 34 weeks of gestation and late-onset preeclampsia at or after
34 weeks of gestation [15].
2.3. Echocardiography
We used a commercially available ultrasound system (Vivid E9
XDclear, GE Healthcare Horten, Norway) with a 3.5-MHz-phased array
transducer (M5Sc).
All women underwent a comprehensive echocardiographic assess-
ment according to current guidelines [14] and volumetric measure-
ments were indexed to body surface area when appropriate. We ob-
tained M-mode or 2D-guided linear measurement of LV interventricular
septum (IVS), infero-lateral wall thickness (PWT), and LV internal
diameter (LVID) at end-diastole. LV mass was estimated using the linear
method by Cube formula as
LV mass = 0.8 × 1.04 × [(IVS + LVID + PWT)3−LVID3] + 0.6 g.
2D LV ejection fraction (LVEF) was calculated using Simpson’s bi-
plane method of discs. Peak systolic mitral annular velocities (S′) were
estimated from the tissue Doppler velocity images as an average of
septal, lateral, anterior, and posterior velocities. Peak diastolic mitral
annular velocities (e′) was estimated both septal, lateral, and as the
average of septal and lateral velocities. In the assessment of E/e′ ratio
we used the average e′.
The magnitude of peak systolic left ventricular global longitudinal
strain (LVGLS) [16] was obtained in standard 2D cine-loops using
frame-by-frame tracking of speckle patterns throughout the left-sided
myocardium with a frame rate > 60 frames/s. We manually adjusted
the speckle area of interest for optimal tracking results. We excluded
segments with an unacceptably low tracking quality. LVGLS was cal-
culated using a 17-myocardial segment model [17]. Fig. 1 shows ex-
amples of LVGLS plots in a EOPE patient, a LOPE patient, and a control
subject. The LVGLS magnitude is expressed as a negative value. Thus,
the more negative LVGLS value the better myocardial deformation.
The echocardiographic assessment and data analysis was performed
blinded to PE status. The data were analyzed offline using dedicated
software by a single observed BBL (EchoPAC PC SW-Only, 201, GE-
Healthcare, Milwaukee, Wisconsin, USA).
2.4. Blood sampling and blood pressure measurement
At follow-up both fasting and non-fasting blood samples were col-
lected and analyzed according to standard laboratory protocols. A fully
automatic oscillometric device (Omron MIT Elite; Omron Healthcare
Co; Kyoto; Japan) with appropriate cuff size was used to arterial blood
pressure measurements following 10 min rest in upright position.
Measurements were performed three times with 3-minutes interval and
repeated if values differed more than 5 mmHg. The mean of the last two
blood pressure measurements is the reported value.
2.5. Statistical analysis
Normally distributed data are presented as mean ± standard de-
viation; non-normally distributed data are presented as median and
interquartile range. Categorical data are presented as absolute values
with percentages. Histograms and Q-Q plots were used to check con-
tinuous values for normality. Between-group differences were assessed
using the t-test and ANOVA for normally distributed data, the Mann-
Whitney U test and Kruskal-Wallis test for non-normally distributed
data, and the chi2 test for dichotomized data. A linear regression model
was used to compare continuous variables. In a multivariable regression
model we adjusted for blood pressure and body mass index. Area under
ROC curve was used to compare the predictive ability of the parameters
T.S. Clemmensen et al. Pregnancy Hypertension 14 (2018) 9–14

Fig. 1. Examples of bulls eye 17 segments model of global longitudinal strain. The strain value is reported for each myocardial segment. Furthermore, deformation
magnitude is expressed in color. The greater deformation the darker red color. A: Healthy control: Left ventricular global longitudinal strain (LVGLS = −21.4%); B:
Patient with late onset preeclampsia: LVGLS = −20.2%; C: Patient with early onset preeclampsia: LVGLS = -16.8%. (For interpretation of the references to colour in
this figure legend, the reader is referred to the web version of this article.)

Table 1
Patient characteristics.
EOPE (31) LOPE (22) Controls (40) P value

PE data at index pregnancy

Time since delivery (years) 13 ± 4 12 ± 3 12 ±3 0.47
Age at index pregnancy (years) 30 ± 6 30 ± 4 30 ±5 0.95
Primiparous at index pregnancy, n (%) 13 (42) 14 (64) 20 (50) 0.30
GA at PE diagnosis, weeks 31 ± 3 38 ± 2 – < 0.0001
GA at delivery, weeks 34 ± 4 39 ± 2 39 ±2 < 0.0001
Still birth n (%) 4 (13) 1 (5) 0 0.06
Pregnancy outcome < 0.05
Single n (%) 26 (84) 22 39 (98)
Gemelli n (%) 5 (16) 0 1 (3)
Three or more n (%) 0 0 0
Birth weight (g) 1851 ± 853 3077 ± 720 3441 ± 642 < 0.0001

Pregnancy history at follow-up

Total pregnancies (no) 3 [2–4] 2.5 [2–3] 3 [2–4] 0.23
Completed pregnancies (no) 2 [2–3] 2 [2–3] 2 [2–3] 0.84
Number of spontaneous abortions (no) 0 [0–1] 0 [0–0] 0 [0–1] 0.18

CVD risk factors at follow-up

Visit age (years) 41 ± 7 42 ± 5 41 ± 6 0.91
Body mass index (kg/m2) 30 ± 7 28 ± 5 27 ± 5 0.07
Diabetes n (%) 2 (6) 1 (5) 1 (3) 0.40
Hypertension n (%) 12 (39) 1 (5) 6 (15) < 0.01
Hypercholesterolemia n (%) 8 (26) 1 (5) 4 (10) 0.06
Smoking 0.16
Current smoker n (%) 9 (29) 1 (5) 8 (20)
Former smoker n (%) 6 (19) 3 (14) 9 (23)
Exercise (hours per week) 2.5 [1–4] 2 [1–3] 3 [2–5] < 0.05

Medication at follow-up
Antihypertensive drugs n (%) 10 (32) 0 4 (10) < 0.01
Lipid lowering drugs n (%) 4 (13) 0 1 (3) 0.07
Antidiabetic drugs n (%) 2 (6) 1 (5) 0 0.29
Biochemistry at follow-up
Total cholesterol 4.9 ± 1.1 5.0 ± 0.8 4.9 ± 0.7 0.82
LDL cholesterol 3.1 ± 1.1 3.1 ± 0.7 2.9 ± 0.7 0.47
HbA1c 36 ± 4 33 ± 3 34 ± 3 < 0.05
Creatinine (µmol/l) 63 ± 11 65 ± 14 68 ± 12 0.21
Hemoglobin (mmol/l) 8.4 ± 0.7 8.3 ± 0.5 8.3 ± 0.6 0.31

Data are presented as present or mean ± standard deviation or median and [IQR].
EOPE = early onset preeclampsia, LOPE = late onset preeclampsia, PE = preeclampsia, CVD = cardio-vascular disease, GA = gestational age.

of interest.
All tests were two-sided, and a p < 0.05 was considered statisti-
cally significant. Analyses were performed using STATA (STATA/IC 13,
StataCorp LP, Texas, College Station, USA).
3. Results
3.1. Patient demographics
Table 1 presents the clinical characteristics of the PE and control
groups. The groups were comparable regarding age at visit, age at index

11
T.S. Clemmensen et al. Pregnancy Hypertension 14 (2018) 9–14

Table 2
(a)

−10
Left ventricular global longitudinal strain (%)
Echocardiography.
EOPE (31) LOPE (22) Controls (40) P value

Non-invasive hemodynamics
Mean arterial blood 111 ± 15 107 ± 15 104 ± 12 0.14

−15
pressure
Heart rate 71 ± 7 71 ± 8 67 ± 13 0.14

LV dimensions
IVS (mm) 0.8 ± 0.2 0.8 ± 0.1 0.8 ± 0.2 0.78
PWT (mm) 0.9 ± 0.2 0.9 ± 0.1 0.9 ± 0.4 0.97

−20
LV EDD (cm/m2) 2.7 ± 0.3 2.5 ± 0.3 2.6 ± 0.5 0.41
LV ESD (cm/m2) 1.6 ± 0.3 1.5 ± 0.2 1.6 ± 0.3 0.33
LV EDV (ml/m2) 53 ± 7 54 ± 9 54 ± 10 0.93
LV ESV (ml/m2) 23 ± 6 22 ± 6 23 ± 6 0.68
LV-mass (g/m2) 81 ± 22 73 ± 12 76 ± 13 0.20

LV systolic function

−25
2D LVEF (%) 57 ± 9 61 ± 7 59 ± 6 0.27 Controls LOPE EOPE
S′ mean (cm/s) 7.9 ± 1.7 8.3 ± 1.8 8.4 ± 1.7 0.54
LVGLS (%) −18 ± 3 −21 ± 2 −21 ± 2 < 0.001 (b)
LV diastolic function
LA-volume (ml/m2)

70
29 ± 8 28 ± 9 28 ± 6 0.95

Left ventricular ejection fraction (%)

E velocity (m/s) 0.9 ± 0.2 0.8 ± 0.2 0.8 ± 0.1 0.57
A velocity (m/s) 0.6 ± 0.2 0.6 ± 0.2 0.5 ± 0.1 0.54
e′ lateral (cm/s) 12.8 ± 3.7 14.3 ± 3.8 13.9 ± 3.7 0.38

60
e′ septal (cm/s) 8.5 ± 2.6 11.0 ± 3.3 10.9 ± 2.8 < 0.01
e′ mean 10.6 ± 2.9 12.6 ± 3.1 12.4 ± 2.9 < 0.05
E/A 1.6 ± 0.6 1.6 ± 0.5 1.6 ± 0.5 0.82
E/e′ mean 9.1 ± 3.8 6.8 ± 1.9 7.1 ± 2.0 < 0.01
50
E-dec (ms) 174 ± 39 175 ± 45 160 ± 47 0.31

Data are presented as mean ± SD.

EOPE = early onset preeclampsia, LOPE = late onset preeclampsia, LV = left
40

ventricle IVS = interventricular septum, PWT = posterior wall thickness,

EDD = end-diastolic diameter, ESD = end-systolic diameter, EDV = end-dia-
stolic volume, ESV = end-systolic volume, LVEF = left ventricular ejection
fraction, LVGLS = left ventricular global longitudinal strain, LA = left atrium,
30

Controls LOPE EOPE

E-dec = E deceleration time.

pregnancy, time since delivery, number of fetuses (singleton/multiple Fig. 2. A: Box plots showing left ventricular global longitudinal strain (LVGLS)
in controls, women with late onset preeclampsia (LOPE), and women with early
pregnancy), and the prevalence of primiparous women. As expected,
onset preeclampsia (EOPE). We found LVGLS significantly lower in EOPE than
gestational age at delivery was lower in the EOPE group than in the
LOPE (p < 0.01) and controls (p < 0.001). B: Box plots showing left ven-
control and LOPE group (p < 0.0001) and subsequently birth weight
tricular ejection fraction (LVEF) in controls, women with late onset pre-
and length were reduced (both p < 0.0001). None of the controls eclampsia (LOPE), and women with early onset preeclampsia (EOPE). We found
previously had PE, whereas 4 women (13%) in the EOPE group and 2 no difference in LVEF between EOPE, LOPE, and controls (ANOVA p = 0.27).
women in the LOPE (9%) previously had PE (p = 0.07). Six women in
the EOPE (19%) and 3 women in the LOPE group had PE during a later
difference in LVGLS was observed between LOPE and controls, Fig. 2A.
pregnancy (p = 0.07).
LVGLS remained significantly lower in the EOPE groups than the con-
The body mass index (BMI, p = 0.07) tended to be higher in the
trols and LOPE groups in women with LVEF > 45% (p < 0.01).
EOPE group than in the LOPE group and the control group. No sig-
LVGLS remained significantly lower in the EOPE group after adjustment
nificant between groups difference was observed in the smoking pre-
of mean arterial blood pressure and body mass index (p < 0.01). We
valence (p = 0.17). Furthermore, hypertensions was significantly more
observed no inter group difference in LVEF (Fig. 2B) or systolic long-
prevalent in the EOPE group than controls (p < 0.05), whereas no
itudinal myocardial tissue Doppler velocities.
significant difference in hypertension prevalence was observed between
Similar to the systolic parameters, the diastolic filling pattern
controls and LOPE (p = 0.21). Diabetes mellitus (DM) was rare in all
parameters were within normal range in both EOPE and LOPE women.
groups, but women in the EOPE group had significantly higher HbA1C
However, EOPE women had lower septal e′ velocities leading to lower
than controls and LOPE women (EOPE versus controls: p < 0.01; EOPE
mean e′ and subsequently higher E/e′ ratio than controls and LOPE
versus LOPE: p < 0.05). We observed no difference in cholesterol le-
women (EOPE versus controls: p < 0.05; EOPE versus LOPE:
vels and renal function between the groups.
p < 0.05).

3.2. Echocardiography
Table 2 shows non-invasive hemodynamic and echocardiographic
markers in the PE and control groups. We observed no significant dif-
ference in blood pressure and heart rate between groups. Likewise, we
observed no difference in LV and left atrial volumes and dimensions
between groups.
The mean longitudinal myocardial deformation measure by LVGLS
was within normal range in all three groups. However, women with
EOPE had significantly lower LVGLS than controls. In contrast, no
3.3. Factors associated with myocardial longitudinal deformation
LVGLS was the echocardiographic parameter with the strongest
association with EOPE, Fig. 3. We found LVGLS significantly correlated
with BMI (r = 0.30, p < 0.01), HbA1C levels (r = 0.21, p < 0.05),
hypertension (r = 0.33, p < 0.01), gestation weeks at delivery
(r = −0.25, p < 0.05), and birth weight (r = −0.27, p < 0.05).
Among women with PE, LVGLS was borderline significantly associated
with gestation weeks at PE diagnosis (r = −0.28, p = 0.05). LVGLS did

12
T.S. Clemmensen et al.
1.00
0.75
Sensitivity
0.50
0.25
0.00
0.00 0.25 0.50 0.75 1.00
1−Specificity
AUC GLS = 0.76 (0.64−0.88)
AUC LVEF = 0.51 (0.37−0.66)
AUC S’ = 0.59 (0.43−0.74)
AUC E/e’ = 0.64 (0.49−0.80)
Fig. 3. Receiver-operating characteristic curves for prediction of previous early
onset preeclampsia (EOPE). Left ventricular global longitudinal strain (LVGLS)
area under curve (AUC) AUC versus left ventricular ejection fraction (LVEF)
AUC: p < 0.01; LVGLS versus peak systolic mitral annular velocity (S’):
p = 0.08, LVGLS versus E/e′: p = 0.27.
not correlate with exercise hours per week (r = −0.18, p = 0.10).
We found a moderate relation between LVGLS and LVEF
(r = −0.40, p < 0.001) and between LVGLS and E/e′ (r = 0.42,
p < 0.001).
4. Discussion
This is the first published study to evaluate LVGLS in women long-
term post PE. The main findings in this study were: (1) mean long-
itudinal myocardial deformation measure by LVGLS was within normal
range in all three groups; (2) women with EOPE had significantly lower
LVGLS than controls; (3) no difference in LVGLS was observed between
LOPE and controls; (4) LVGLS was the echocardiographic parameter
with the strongest association with previous EOPE.
Earlier reports in previously PE women reported a permanent sub-
clinical LV dysfunction [18–21]. Furthermore, it has been shown that in
normal pregnancy, an increased preload and a decreased afterload
favor an improved emptying of the left ventricle during systole and a
reduction of the end-systolic pressure [22]. In PE, the elevated afterload
is linked with a reduced emptying of the left ventricle and elevated end-
systolic pressure. Moreover, it has previously been reported that the
mean LV end-systolic volume is markedly increased in PE compared to
normotensive women [23]. However, in our long-term follow-up we did
not observe any differences in the LV end-systolic volume. It is note-
worthy that only one third of the EOPE women in the present study
received antihypertensive drugs and the blood pressure was compar-
able between groups. Furthermore, the LVGLS impairment remained
significantly lower in EOPE patients even after adjustment for BMI and
blood pressure. Therefore, LVGLS seems reduced in long-term EOPE
despite optimized afterload conditions suggesting other mechanisms to
be involved.
We observed LVEF measured by conventional 2D echocardiography
to be preserved in all 3 groups. This was also observed recently by
Vaddamani S et al. between EOPE and LOPE [11]. However, PE women
in that study had severely elevated blood pressure and myocardial
hypertrophy leading to a high degree of diastolic dysfunction in con-
trast to the results in our follow-up study showing well controlled mean
13
Pregnancy Hypertension 14 (2018) 9–14
arterial blood pressure in all groups. Furthermore, we observed sig-
nificantly less LV hypertrophy emphasizing the importance of con-
tinued afterload optimization post PE.
A higher proportion of cardiac diastolic dysfunction one year after
delivery has been observed in women who had EOPE compared to
normotensive women with a history of low-risk pregnancy [24]. We
extent this observation to a long-term observation showing persistent
evidence of subclinical diastolic cardiac dysfunction in EOPE women.
In a recently published cohort of previously preeclamptic women
observed between 6 months up to 4 years (mean 2.3 years) post gesta-
tion a significant difference between LVEF in EOPE and healthy controls
was observed. The same authors reported highly significant differences
between groups for LVGLS [25] with rather high “normal” values for
LVGLS in both LOPE and healthy controls using same echocardio-
graphic equipment as we did. In alignment with the present results,
LVGLS in EOPE was within normal range suggesting subclinical im-
pairment. Our results indicate that the reduction in LVGLS magnitude is
preserved over time suggesting chronic myocardial damage.
Different pathways are proposed to link the increased cardiovas-
cular risk to PE. The lower LVGLS in EOPE could be attributes to the
presence of LV fibrosis at short to medium follow-up after delivery
compared to LOPE and controls. LV fibrosis is known to be linked with
adverse clinical outcomes in adults [25]. Increased LV fibrosis can be
attributed by different causes such as dysglycemia and overt DM. In
patients with type 1 or type 2 DM the risk of PE is increased two to four-
fold. Insulin resistance is thought of being a contributor to the patho-
physiology of PE [26]. Furthermore, an increased risk of late onset DM
among women with an eclampsia history has been reported [27]. Even
though DM was only diagnosed in 4 subject in the present study, HbA1C
and BMI was significantly higher among EOPE women than controls
and LOPE women, indicating a higher proportion of dysglycemia in this
group. However, we did not investigate the remaining population for
dysglycemia (e.g. by oral glucose tolerance test). DM and dysglycemia
can result in structural cardiac changes [28,29] and subsequently po-
tentially reduced longitudinal myocardial deformation. In our previous
work, we observed an association between dysglycemia and decreased
coronary microcirculation [30] and furthermore demonstrated a cor-
relation between coronary flow velocity reserve and LVGLS in different
patient populations [31,32]. In the present study, we observed a weak
but significant correlation between LVGLS and HbA1C levels. Thus, we
cannot rule out a potential common etiological pathway for PE induced
myocardial dysfunction and dysglycemia.
The use of LVGLS gives a more comprehensive evaluation of LV
systolic function. LVGLS has a better prognostic value for predicting
major adverse cardio vascular events than LVEF in various cardiac
conditions including heart failure, acute myocardial infarction, valvular
heart disease, and miscellaneous cardiac diseases [33]. Even though the
prognostic value of LVGLS in PE women is unknown, the present study
provides encouraging evidence to use LVGLS in the assessment of
myocardial damage and dysfunction both early and late after PE.
5. Limitations
We acknowledge a number of limitations to this study. It reflects the
experience of a single center with a relatively small cohort of women.
Exposure was defined by ICD-10 diagnoses in two different registries
possibly resulting in preeclampsia underestimation [34]. Participants
themselves confirmed obstetric and medical history at follow-up as
neither complete obstetric nor medical records were available. In ad-
dition, we do not have information on changes over time due to the
cross-sectional study design. Forty-seven (87/185 women) percent of
eligible women declined participation. Yet, the participation rate was
similar in each group.
T.S. Clemmensen et al.
6. Conclusion
Women with a history of EOPE are more likely to have subclinical
impairment of left ventricular function than are those with a history of
LOPE and controls. Thus, LVGLS magnitude was lower and E/e′ higher
in EOPE women than LOPE and controls. LVGLS was the echocardio-
graphic parameter with the strongest association with previous EOPE.
7. Funding sources
None.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the
online version, at https://doi.org/10.1016/j.preghy.2018.07.001.
References
[1] S.D. McDonald, A. Malinowski, Q. Zhou, S. Yusuf, P.J. Devereaux, Cardiovascular
sequelae of preeclampsia/eclampsia: a systematic review and meta-analyses, Am.
Heart J. 156 (5) (2008) 918–930.
[2] G. Bro Schmidt, M. Christensen, Knudsen U. Breth, Preeclampsia and later cardio-
vascular disease – what do national guidelines recommend? Pregnancy Hypertens.
10 (2017) 14–17.
[3] J.A. Lykke, J. Langhoff-Roos, B.M. Sibai, E.F. Funai, E.W. Triche, M.J. Paidas,
Hypertensive pregnancy disorders and subsequent cardiovascular morbidity and
type 2 diabetes mellitus in the mother, Hypertension 53 (6) (2009) 944–951.
[4] J.G. Ray, M.J. Vermeulen, M.J. Schull, D.A. Redelmeier, Cardiovascular health after
maternal placental syndromes (CHAMPS): population-based retrospective cohort
study, Lancet 366 (9499) (2005) 1797–1803.
[5] L. Bellamy, J.P. Casas, A.D. Hingorani, D.J. Williams, Pre-eclampsia and risk of
cardiovascular disease and cancer in later life: systematic review and meta-analysis,
BMJ 335 (7627) (2007) 974.
[6] D. Raymond, E. Peterson, A critical review of early-onset and late-onset pre-
eclampsia, Obstet. Gynecol. Surv. 66 (8) (2011) 497–506.
[7] M. Christensen, C.S. Kronborg, R.K. Carlsen, N. Eldrup, U.B. Knudsen, Early ge-
stational age at preeclampsia onset is associated with subclinical atherosclerosis 12
years after delivery, Acta Obstet. Gynecol. Scand. 96 (9) (2017) 1084–1092.
[8] J.S. Castleman, R. Ganapathy, F. Taki, G.Y. Lip, R.P. Steeds, D. Kotecha,
Echocardiographic, Structure and function in hypertensive disorders of pregnancy:
a systematic review, Circ. Cardiovasc. Imaging 9 (9) (2016), https://doi.org/10.
1161/CIRCIMAGING.116.004888.
[9] K. Melchiorre, B. Thilaganathan, Maternal cardiac function in preeclampsia, Curr.
Opin. Obstet. Gynecol. 23 (6) (2011) 440–447.
[10] E.V. Tyldum, B. Backe, A. Stoylen, S.A. Slordahl, Maternal left ventricular and
endothelial functions in preeclampsia, Acta Obstet. Gynecol. Scand. 91 (5) (2012)
566–573.
[11] S. Vaddamani, A. Keepanasseril, A.A. Pillai, B. Kumar, Maternal cardiovascular
dysfunction in women with early onset preeclampsia and late onset preeclampsia: a
cross-sectional study, Pregnancy Hypertens. 10 (2017) 247–250.
[12] O.A. Smiseth, H. Torp, A. Opdahl, K.H. Haugaa, S. Urheim, Myocardial strain
imaging: how useful is it in clinical decision making? Eur. Heart J. 37 (15) (2016)
1196–1207.
[13] L.F. Tops, V. Delgado, N.A. Marsan, J.J. Bax, Myocardial strain to detect subtle left
ventricular systolic dysfunction, Eur. J. Heart Fail. 19 (3) (2017) 307–313.
[14] R.M. Lang, L.P. Badano, V. Mor-Avi, J. Afilalo, A. Armstrong, L. Ernande, et al.,
Recommendations for cardiac chamber quantification by echocardiography in
14
Pregnancy Hypertension 14 (2018) 9–14
adults: an update from the American Society of Echocardiography and the European
Association of Cardiovascular Imaging, J. Am. Soc. Echocardiogr. 28 (1) (2015)
1–39.e14.
[15] A.L. Tranquilli, M.A. Brown, G.G. Zeeman, G. Dekker, B.M. Sibai, The definition of
severe and early-onset preeclampsia. Statements from the International Society for
the Study of Hypertension in Pregnancy (ISSHP), Pregnancy Hypertens. 3 (1)
(2013) 44–47.
[16] S.A. Reisner, P. Lysyansky, Y. Agmon, D. Mutlak, J. Lessick, Z. Friedman, Global
longitudinal strain: a novel index of left ventricular systolic function, J. Am. Soc.
Echocardiogr. 17 (6) (2004) 630–633.
[17] M.D. Cerqueira, N.J. Weissman, V. Dilsizian, A.K. Jacobs, S. Kaul, W.K. Laskey,
et al., Standardized myocardial segmentation and nomenclature for tomographic
imaging of the heart. A statement for healthcare professionals from the Cardiac
Imaging Committee of the Council on Clinical Cardiology of the American Heart
Association, Circulation 105 (4) (2002) 539–542.
[18] T. Ghi, D. Degli Esposti, E. Montaguti, M. Rosticci, F. De Musso, A. Youssef, et al.,
Post-partum evaluation of maternal cardiac function after severe preeclampsia, J.
Matern. Fetal Neonatal Med. 27 (7) (2014) 696–701.
[19] C.S. Evans, L. Gooch, D. Flotta, D. Lykins, R.W. Powers, D. Landsittel, et al.,
Cardiovascular system during the postpartum state in women with a history of
preeclampsia, Hypertension 58 (1) (2011) 57–62.
[20] I. Strobl, G. Windbichler, A. Strasak, V. Weiskopf-Schwendinger, U. Schweigmann,
A. Ramoni, et al., Left ventricular function many years after recovery from pre-
eclampsia, BJOG 118 (1) (2011) 76–83.
[21] L.A. Simmons, A.G. Gillin, R.W. Jeremy, Structural and functional changes in left
ventricle during normotensive and preeclamptic pregnancy, Am. J. Physiol. Heart
Circ. Physiol. 283 (4) (2002) H1627–33.
[22] H. Valensise, G.P. Novelli, B. Vasapollo, G. Di Ruzza, M.E. Romanini, M. Marchei,
et al., Maternal diastolic dysfunction and left ventricular geometry in gestational
hypertension, Hypertension 37 (5) (2001) 1209–1215.
[23] C. Shivananjiah, A. Nayak, A. Swarup, Echo changes in hypertensive disorder of
pregnancy, J. Cardiovasc. Echogr. 26 (3) (2016) 94–96.
[24] P. Soma-Pillay, M.C. Louw, A.O. Adeyemo, J. Makin, R.C. Pattinson, Cardiac dia-
stolic function after recovery from pre-eclampsia, Cardiovasc. J. Afr. 28 (2017) 1–6.
[25] R. Orabona, E. Vizzardi, E. Sciatti, I. Bonadei, A. Valcamonico, M. Metra, et al.,
Insights into cardiac alterations after pre-eclampsia: an echocardiographic study,
Ultrasound Obstet. Gynecol. 49 (1) (2017) 124–133.
[26] T.L. Weissgerber, L.M. Mudd, Preeclampsia and diabetes, Curr. Diab. Rep. 15 (3)
(2015) 9-015-0579-4.
[27] L.C. Chesley, Remote prognosis after eclampsia, Perspect. Nephrol. Hypertens. 5
(1976) 31–40.
[28] A. Aneja, W.H. Tang, S. Bansilal, M.J. Garcia, M.E. Farkouh, Diabetic cardiomyo-
pathy: insights into pathogenesis, diagnostic challenges, and therapeutic options,
Am. J. Med. 121 (9) (2008) 748–757.
[29] L. Maya, F.J. Villarreal, Diagnostic approaches for diabetic cardiomyopathy and
myocardial fibrosis, J. Mol. Cell. Cardiol. 48 (3) (2010) 524–529.
[30] B.B. Logstrup, D.E. Hofsten, T.B. Christophersen, J.E. Moller, H.E. Botker,
P.A. Pellikka, et al., Influence of abnormal glucose metabolism on coronary mi-
crovascular function after a recent myocardial infarction, JACC Cardiovasc.
Imaging 2 (10) (2009) 1159–1166.
[31] B.B. Logstrup, D.E. Hofsten, T.B. Christophersen, J.E. Moller, H.E. Botker,
P.A. Pellikka, et al., Correlation between left ventricular global and regional long-
itudinal systolic strain and impaired microcirculation in patients with acute myo-
cardial infarction, Echocardiography 29 (10) (2012) 1181–1190.
[32] T.S. Clemmensen, B.B. Logstrup, H. Eiskjaer, S.H. Poulsen, Coronary flow reserve
predicts longitudinal myocardial deformation capacity in heart-transplanted pa-
tients, Echocardiography (2015).
[33] K. Kalam, P. Otahal, T.H. Marwick, Prognostic implications of global LV dysfunc-
tion: a systematic review and meta-analysis of global longitudinal strain and ejec-
tion fraction, Heart 100 (21) (2014) 1673–1680.
[34] B.M. Luef, L.B. Andersen, K.M. Renault, E.A. Nohr, J.S. Jorgensen, H.T. Christesen,
Validation of hospital discharge diagnoses for hypertensive disorders of pregnancy,
Acta Obstet. Gynecol. Scand. 95 (11) (2016) 1288–1294.