Title: Cardiac Output In Pre eclampsia Is Determined By The Presence Of

Fetal Growth Restriction, Not Gestation At Onset: A prospective cohort study

Authors: Jasmine TAY 1, 3, Lin FOO 1, 3, Giulia MASINI 1, Phillip R BENNETT3, Carmel M

MCENIERY 2, Ian B WILKINSON 2, Christoph C LEES 1, 3, 4,*

Affiliations:

1Centre for Fetal Care, Queen Charlotte's and Chelsea Hospital, Imperial College Healthcare

NHS Trust, London, United Kingdom.

2Experimental Medicine and Immunotherapeutics, ACCI, Level 3 Addenbrooke’s, Hills Road,

CB2 0QQ, Cambridge.

3 Institute for Reproductive and Developmental Biology, Department of Surgery and Cancer,

Imperial College London, London, United Kingdom

4Department of Development and Regeneration, KU Leuven, Leuven, Belgium

* Dr Christoph C. Lees, christoph.lees@nhs.net

Conflict of interest statement: The authors report no conflict of interests

Abstract word count –

Main text word count -

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

2

Abstract:

Background and Objectives

Pre-eclampsia (PE) and fetal growth restriction (FGR) are considered to be placentally-mediated

disorders. The clinical manifestations are widely held to relate to gestation age at onset with

early- and late-onset PE considered to be phenotypically distinct. Recent studies have reported

conflicting findings in relation to cardiovascular function, and in particular cardiac output, in PE

and FGR.

Study Design

We investigated maternal cardiovascular function in relation to clinical subtype in 45

pathological pregnancies (14 ‘PE only’, 16 ‘FGR only’, 15 ‘PE and FGR’) and compared these

with 107 healthy person observations. Cardiac output (CO) was the primary outcome measure,

and was assessed using an inert gas rebreathing method (Innocor®), from which peripheral

vascular resistance was derived (PVR); arterial function was assessed by Vicorder ®, a cuff-

based oscillometric device. Cardiovascular parameters were normalised for gestational age in

relation to healthy pregnancies using Z scores, thus allowing for comparison across the

gestational range 24-40 weeks.

Results

Compared with healthy control pregnancies, women with PE had higher CO Z scores (1.87 ±

1.35; p=0.0001) and lower PVR Z scores (-0.76± 0.89; p=0.025); those with FGR had higher

PVR Z scores (0.57± 1.18; p=0.04) and those with both PE and FGR had lower CO Z scores (-

0.80 ± 1.3; p= 0.007) and higher PVR Z scores (2.16 ± 1.96; p=0.0001). These changes were not

related to gestational age of onset. All those affected by PE and/or FGR had abnormally raised

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

3

augmentation index (AIx) and pulse wave velocity (PWV). Furthermore, in PE, low CO was

associated with low birthweight and high CO with high birthweight. (r=0.42, p=0.03).

Conclusions

PE is associated with high CO, but if PE presents with FGR, the opposite is true; both conditions

are, nevertheless, defined by hypertension. FGR without PE is associated with high PVR.

Though 'early' and 'late' gestation PE are considered to be different diseases, we show that the

haemodynamic characteristics of PE were unrelated to gestation age at onset, but strongly

associated with the presence or absence of FGR. FGR more commonly co-exists with PE at early

gestation, thus explaining the conflicting results of previous studies. Furthermore, anti-

hypertensive agents act by reducing CO or PVR and are administered without reference to

cardiovascular function in PE. The underlying pathology (PE, FGR, PE and FGR) defines

cardiovascular phenotype, providing a rational basis for choice of therapy where high or low CO

or PVR are the predominant features.

Key Words: Cardiac output, vascular resistance, arterial function, pregnancy, hypertension

Condensation: ‘Pure’ PE is characterized by high cardiac output, FGR alone with high vascular

resistance; the co-existence of FGR with PE alters the haemodynamic phenotype of PE to low

cardiac output, high vascular resistance.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

4

[Main Text:]

Introduction

Pre-eclampsia (PE) is not simply a pregnancy-specific syndrome: its implications on later life

cardiovascular1 and cognitive function2 and healthcare cost3 are only now being understood. Fetal

growth restriction (FGR) has a close but poorly understood relationship with PE. Although PE

and FGR frequently present in isolation, they may occur together, particularly at early gestation

4. The underlying pathophysiology of PE has never been fully understood, but the cause has often

been attributed to the placenta as PE resolves completely after delivery. Inadequate trophoblast

invasion leading to uteroplacental malperfusion is thought to underlie both PE and FGR5,6.

However, this placental theory does not explain why women who had PE in their pregnancies

have a higher cardiovascular risk in later life7-9 or why women with pre-pregnancy

cardiovascular risk factors have a higher risk of PE and FGR in pregnancy10. Emerging data

suggests that maternal cardiovascular function is impaired post-delivery in women with PE11,12

and high blood pressure prior to pregnancy increases the risk of PE developing10. Nevertheless,

PE and FGR are commonly referred to as ‘placenta mediated disorders’, suggested to arise

through defective placentation.

Studies of cardiovascular function in pregnancy have shown inconsistent, and in some cases,

contradictory results. The classic studies of Easterling suggested that PE was associated with a

high CO state13, though this relationship was thought to be explained, at least in part, by the

increased body surface area 14. Other studies have suggested that PE should be subdivided into

those where low or high CO is predominant.15 FGR with and without PE have a different

cardiovascular profile 16,17 though the relationship of these changes to healthy pregnancy or PE

without FGR has not been studied. A recent systematic review concluded that studies of

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

5

cardiovascular function in gestational hypertension and PE show conflicting results. The authors

concluded that “increased peripheral vascular resistance correlates with disease severity” but of

note, the co-existence of FGR was not considered in majority of the studies. 18

Previous studies on cardiovascular function in PE have incompletely characterized FGR 19,

recruited only within a particular gestation range 16 and/or not compared findings to a healthy

reference pregnant population. 20 Existing studies provide interesting insights to cardiovascular

dysfunction in pathological pregnancies but leave important questions unanswered in relation to

gestational effects and pregnancies where PE and FGR are combined.

Over the last two decades it has become customary to subdivide PE into “early” and “late”

variants with an arbitrary 34 week gestational age ‘cut-off’. This distinction arose not because of

a pathophysiological difference between the conditions, but because of the inability of screening

by uterine artery Doppler to effectively identify cases of late onset disease.21-23 It is suggested

that early and late variants have different underlying vascular characteristics, though in most

studies, late PE is rarely or never associated with FGR. 20 Adding further complexity, recent

studies have suggested that PE at term is associated with babies with larger birth weight.14

We sought to investigate the relationship between cardiovascular function and clinical phenotype

of PE and FGR alone and in combination across a gestation range, with cardiac output being the

primary outcome measure. Cardiovascular function changes with gestational age, therefore, in

order to allow comparison we transformed all data in relation to that obtained from women with

healthy pregnancies using the statistical technique of z-scoring. This removed the need for a

gestation matched cohort design and allowed all data points to be considered. 19 In doing so, we

performed comprehensive haemodynamic assessments from 24 weeks gestation onwards in

women with healthy pregnancies, with PE in combination with FGR (PE and FGR), PE without

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

6

FGR (PE only), and FGR without PE (FGR only), managed and monitored in a single maternity

unit.

Materials and Methods

Study protocol

In this prospective cross-sectional study, women between weeks 24-40 of pregnancy were

recruited from the antenatal clinic, labour ward, day assessment unit and antenatal ward of a

tertiary level London teaching hospital between January 2015 – May 2017. The study protocol

was approved by NHS National Research Ethics Committee, London Riverside and all

participants gave written informed consent. Gestation of pregnancy was determined by

measurement of crown-rump length between 11-13 weeks in early pregnancy.

Women were recruited at presentation or diagnosis where it was feasible to conduct a detailed

cardiovascular examination and were not included if they were already in labour, undergoing

induction of labour or imminent delivery was planned. (Figure 1) We describe the diagnosis as

those with PE alone ‘PE only’, FGR alone ‘FGR only’ and PE together with FGR ‘PE and FGR’.

Women’s assessments are reported at the time of initial diagnosis of one of these conditions, and

the diagnosis assigned at the time of study inclusion. One hundred and seven healthy person

observations acted as control cases with no control subject assessed twice within the same epoch.

PE was defined as blood pressure at diagnosis of >140/90 mmHg and urine protein creatinine

ratio of >30. FGR was defined as fetal abdominal circumference or estimated fetal weight <10th

centile24 and umbilical Doppler PI >95th centile on ultrasound scan. 25 Those women with known

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

7

underlying cardiovascular conditions, multiple pregnancies and fetal anomalies were excluded

from the study.

Participants underwent peripheral blood pressure measurement, comprehensive cardiovascular

assessments and fetal ultrasound scans as detailed below.

Cardiovascular Assessments

Participants were requested to abstain from caffeinated drinks for 4 hours before the visit. In

brief, cardiovascular measurements were performed with the patient standing upright for the

assessment of cardiac output and in left lateral lying position for assessment of arterial function.

Cardiac output was measured using Innocor TM , a non-invasive inert gas rebreathing technique as

previously described 26,27. Brachial blood pressure was measured using Omron TM M-7 (OMRON

Healthcare Europe B.V.) which has been validated in pregnancy 28. Blood pressure was measured

on the right arm in the seated position following recommendations from the European Society of

Hypertension 29. Mean arterial pressure (MAP) was calculated by [diastolic pressure + (systolic

pressure - diastolic pressure)/ 3]. Peripheral vascular resistance (PVR) was derived with the

following formula, PVR = MAP x 80 / cardiac output.

Arterial analysis was performed using the Vicorder TM (Skidmore Medical, Bristol, UK), an

oscillometric device validated for the measurements of AIx and pulse wave velocity (PWV).30

With the patient lying in the left lateral position, a brachial cuff was applied to the right upper

arm, a leg cuff applied to the right upper thigh and a neck sensor positioned over the carotid

artery. A standard measuring tape was used to measure the linear distance between the

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

8

suprasternal notch and a defined point on the femoral cuff. Augmentation Index (AIX), PWV

and maternal heart rate (beats/minute) were recorded.

Ultrasound

Obstetric ultrasound assessment was performed on Samsung WS80 equipment (Samsung

Medison, Korea). Fetal growth parameters (head circumference, biparietal diameter, abdominal

circumference and femur length) and Doppler studies (umbilical, middle cerebral artery, ductus

venosus and uterine artery) were recorded in all pathological pregnancy cases.

Statistical Analysis

Statistical analyses were performed using SPSS (Version 24_0_0, 2016; SPSS Inc., Chicago,

Illinois, USA). The Kruskal-Wallis test was used to compare the demographic characteristics

between the four groups. The normality of distribution of the data was examined with the

Shapiro-Wilk test and histograms and data are expressed as means ± SD or medians

(interquartile range) for normally- and non-normally distributed data, respectively. We analysed

the data using BMI, age and ethnicity as co-variates. Since cardiovascular function changes with

gestational age, and in order to compare haemodynamic characteristics across groups with

different gestational ages, we transformed all data in relation to that obtained from women with

healthy pregnancies using the statistical technique of z-scoring. This removed the need for a

gestation-matched cohort design and allowed all data points to be considered.16 Briefly,

measurements were transformed to the corresponding Z scores with reference to means and SD

values derived from Controls in four-weekly gestational epochs (24-27+6; 28-31+6; 32-35+6 and

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

9

36-39+6 weeks) for CO, PVR, AIx and PWV. The Z scores of each cardiovascular parameter

were then compared using unpaired t-test. In our cohort, heart rate did not change with gestation

and comparisons were made with untransformed data.

Results

Forty-five pathological pregnancies (14 ‘PE only’, 16 ‘FGR only’, 15 ‘PE and FGR’) were

recruited, of whom three had a prior history of PE and or FGR. All women remained within the

category to which they were first assigned at recruitment. A further sixty four women with

healthy pregnancies and normal pregnancy outcomes were studied across the gestation, yielding

one hundred and seven healthy pregnancy observations. Kruskal- Wallis test showed no

statistically significant differences between age and ethnicity within the 4 groups. The group

with ‘PE’ had a higher booking BMI when compared to the control group (P =0.001) [Table 1].

Cardiovascular parameters in pathological and control groups are presented in Table 2. The

cardiovascular parameters in pathological pregnancies are presented in Table 3; Z scores were

calculated in relation to control cases from healthy pregnancies which were subdivided into four

weekly gestational epochs.

FGR only

In women with FGR only, PVR Z score was significantly higher (0.57 ± 1.18, p= 0.04) [Figure

3], and CO Z score was no different to healthy controls (-0.35 ± 0.99, p= 0.19) [Figure 4]. The Z

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

10

scores of AIx (1.18 ± 1.44, p= 0.0001) [Figure 5] and PWV (0.87± 1.34, p= 0.002) [Figure 6]

were higher than controls.

PE only

Women with PE only had a higher CO Z score than controls 1.87 ± 1.35, p= 0.0001) [Figure 4],

a lower PVR Z score (-0.76 ± 0.89, p= 0.025) [Figure 2], higher AIx Z score (1.52 ± 1.39, p=

0.0001) [Figure 5] and higher PWV (0.71 ± 1.47, p=0.05) [Figure 6].

Of the 13 women with PE only, 7 women were examined before starting anti-hypertensive

medication, 3 women were on labetalol and 1 on nifedipine. The Z scores for CO and PVR of

those examined on treatment versus those without were not different (p= 0.55 and p = 0.57,

respectively).

PE and FGR

In PE and FGR cases, a lower CO Z score (-0.80 ± 1.3, p=0.007) [Figure 4] and higher PVR Z

score (2.16 ± 1.95, p= 0.0001) [Figure 3] was observed compared to controls. There was higher

AIx Z score (1.45 ± 1.09, p= 0.0001) [Figure 5] and higher PWV Z score (2.01 ± 1.55, p=

0.0001) [Figure 6].

Of the 14 women with ‘PE and FGR’; 3 were examined before starting anti-hypertensive

medication, 4 were taking labetalol, 2 nifedipine and 3 both labetalol and nifedipine. Comparing

the Z scores for CO and PVR of those examined on treatment versus those without, there were

no differences (p=0.52 and p=0.99, respectively).

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

11

Heart Rate

There was no difference in observed heart rate between the four groups. (p=0.26) [Figure 7].

Gestation vs CO/PVR

When compared across gestation (decimal weeks by estimated due date derived from dating

scan), there was no relationship between gestation and either CO Z score (Pearson correlation =

0.16, p = 0.35) or PVR Z score (Pearson correlation = -0.21, p = 0.2) for PE only, FGR only and

PE and FGR.

There was no difference between CO Z score (p=0.41) or PVR Z Score (p=0.18) in ‘PE only’

prior to 34 weeks and those after 34 weeks. In ‘PE and FGR’ cases, there was again no

difference between CO Z score (p=0.34) or PVR Z score (p=0.76) in cases prior to 34 weeks and

those after.

CO vs Z score birth weight

Within the group of women with PE (‘PE only’ and ‘PE and FGR’), CO was positively

associated with Z score birth weight (r= 0.42, p=0.03). Within the group of women with ‘FGR

only’, CO was not associated with Z score birth weight (r=0.27, p=0.31).

Post Hoc power calculation

Since this was novel work, a formal sample size calculation prior to undertaking the study was

not possible and the enrolment period was guided by our previous pre-pregnancy studies in

healthy women27,31,32. However, a formal post hoc power calculation, assuming four unequal

groups and a standard deviation for the CO z-score equivalent to the highest SD obtained in the

three pathological pregnancy groups, gave >98% power to detect a significant overall difference

in CO z-score, at an alpha level of 0.05.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

12

Discussion

We report that women with ‘PE only’, ‘FGR only’ and ‘PE and FGR’ have distinctly different,

and, in some cases, opposing cardiovascular characteristics. The predominant abnormalities in

‘PE only’ were a higher CO and lower PVR, while ‘FGR only’ was characterized by higher PVR

than healthy pregnancies. When both PE and FGR occur together, the effect was a cardiovascular

phenotype characterised by a markedly higher PVR and lower CO than FGR only. Previous

studies of PE and FGR have reported inconsistent findings however none have studied

haemodynamic findings across the whole gestation range in relation to healthy pregnancies

recruited concurrently, nor have considered FGR, PE and PEFGR as separate entities.

Recently, a meta-analysis by Castleman18 found that in gestational hypertensive

diseases “severity of disease corresponded with increasing vascular resistance”. We suggest that

this is an oversimplification in that according to our data it holds for PE with FGR; the converse

being true for PE not affected by FGR. 18 Irrespective of differences in CO and PVR, all three

pregnancy complications were associated with abnormal arterial function assessed by increased

augmentation index and increased pulse wave velocity.

The implications for understanding of the disease process is important, particularly in respect of

the deeply ingrained, but empirical and arbitrary distinction between ‘early’ and ‘late’ PE with a

“cut off” at 34 weeks.22,33 It is true that the cardiovascular features of early and late PE were

different, but this was purely by virtue of the association with FGR rather than because of the

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

13

gestational age at onset of the disease. ‘PE only’ had the same cardiovascular findings (high CO

and low PVR) at early and late gestational ages, but if PE was associated with FGR, CO was low

and PVR high, irrespective of gestational age. Importantly, in our study, pathological cases were

similarly represented in numerical terms before and after 34 weeks; 24 and 21 cases respectively.

The contradictory findings of previous studies, namely that gestational age is the determinant of

haemodynamic profile in PE, may be explained because PE and FGR commonly co-exist at early

gestational ages 34 and FGR is uncommonly encountered in late PE 14. PE is therefore more

precisely defined by the presence or absence of FGR rather than by virtue of its gestational age at

onset. We trace the origins of the categorization of ‘early’ and ‘late’ PE to a study on the

sensitivity of second trimester uterine artery Doppler for these complications which was high for

‘early’ but low for ‘late’ disease. Interestingly, early PE was frequently found with FGR in that

study.

In women with PE, with and without FGR, we found a significant positive association between

CO Z score and birthweight normalized for gestation which was no different before or after 34

weeks.. We did not specifically select women with PE, with and without FGR, but instead

recruited 'all comers' and categorized FGR based on ultrasound and Doppler findings. Hence,

there was no selection bias within women with PE as it might be argued that a larger fetus and

placenta will 'drive' a larger CO. We previously reported that the incremental increase in

maternal CO from prior to pregnancy to the mid-second trimester was positively associated with

larger birthweight in normally grown babies 35 irrespective of the size of the fetus in the mid

trimester. This suggests that fetal size on its own did not determine CO. A recent study in women

with PE found that those with the lowest CO had the smallest babies 15. There is strong-albeit

circumstantial-evidence that CO may be at least in part responsible for birthweight, likely

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

14

mediated through an effect on uteroplacental perfusion. Though this is tempting to speculate, we

cannot infer causality from our results.

Of direct clinical relevance is that the current management of hypertension in pregnancy is

exclusively based on a ‘one size fits all’ philosophy where BP alone is the target of therapy

irrespective of the underlying aetiology. 36,37 This approach does not distinguish between PE with

FGR (low CO and high PVR) from PE without FGR (high CO). There is, however, a clear

rationale in targeting therapy according to the underlying cardiovascular phenotype. Drugs with

beta blocking activity are negatively chronotropic and reducing CO may be harmful in women

with FGR where the uteroplacental circulation is already impaired; previous studies have

reported an association with small for gestational age babies 38-40. We report that PE only was

typically associated with high CO: for these women, a beta blocker might be preferable whereas

a drug leading to vasodilation primarily, such as a calcium channel antagonist, may be less

effective. Conversely calcium antagonists are likely to be more effective in the high PVR state

that characterizes PE with FGR. Though a model using haemodynamic parameters to determine

response to labetalol in gestational hypertension has been reported 41, the effectiveness of anti-

hypertensive agents with different modes of action has not. Indeed, if impaired utero-placental

function is the by-product of a low CO-high PVR state such as that found in PE with FGR, then

this should be amenable to targeted manipulation. Recent approaches have included expanding

the intravascular compartment 42, with and without nitrovasodilators 43 and calcium antagonists 44

with reported improvement in maternal cardiovascular and/or fetal parameters.

Arterial function was abnormal in ‘PE only’, ‘FGR only’ and the combination of both ‘PE and

FGR’ as assessed by AIx and PWV, irrespective of whether CO and PVR were increased or

decreased. Aortic AIx is an index of wave reflection and may be considered as a measure of

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

15

arterial function 45-48. Raised AIx is an important predictor of future cardiovascular risk 49 and in

hypertensive patients an independent predictor of future myocardial infarction 50. Although AIx

is influenced by endothelial function, PWV describes aortic stiffness and is determined largely

by the elastic component of large vessels. It is also known to be a reliable marker of

cardiovascular risk in patients with hypertension 51.

The strength of this study is that detailed cardiovascular assessments were undertaken in women

who were phenotyped according to accepted definitions for both PE and FGR. Importantly, our

definition of FGR did not rely simply on fetal smallness52, but required abnormal fetal umbilical

artery Doppler. Our sample size allowed a high power to detect small differences in CO between

the groups. A weakness is that as our protocol required time consuming examinations, not all

women wished to or could take part, in particular where delivery or further therapy was being

planned. This limited the number of women eligible for recruitment, but there was no attempt at

selection other than through women consenting or otherwise for logistical and therapeutic

reasons. Furthermore, though we intended to recruit women prior to starting anti-hypertensive

treatment, these drugs were often started prior to hospital admission. Though we found no

difference in CO and PVR within the ‘PE only’ and ‘PE and FGR’ groups for those on

antihypertensive treatment, the modest numbers within these groups preclude discrimination

between small effect sizes in anti-hypertensive agents and the duration of therapy.

Whilst not being validated specifically in pregnancy, Innocor has been validated extensively in

healthy populations and those with diseases, including chronic heart failure and during

haemodynamic perturbation caused by exercise as a method of assessing cardiac output. 53-56 In

addition, our group has previously shown that this method has good reproducibility in pregnant

women and is sensitive enough to detect longitudinal changes in CO both in very early

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

16

pregnancy and throughout the gestation period.27,35,57 All our participants were examined in a

similar position which limits any potential bias. Moreover as the gas re-breathing method is not

operator dependent, there is no-inter observer bias compared to echocardiography, which is the

other widely used method. The Vicorder has been validated for use against SphygmoCor which

is the most widely used technique to assess pulse wave velocity. There is no agreed gold

standard non-invasive technique to assess pulse wave velocity in pregnancy.58

In summary, we found no evidence to support the categorization of PE of “early” and “late” as

different disease processes, though early onset PE was much more likely to be associated with

FGR. The significance of this is that PE which is not associated with FGR has the same

cardiovascular phenotype, namely high CO whether it is diagnosed at 24 or 36 weeks. The

presence of FGR was associated with high PVR whether or not in combination with PE. Whether

the abnormal haemodynamic profiles that we have observed in the distinct subtypes of PE, FGR

and PE combined with FGR are the cause of the clinical manifestation of hypertension and FGR,

or the effect of an inherent underlying maternal cardiovascular dysfunction is unclear. This study

raises an intriguing question as to whether the cardiovascular dysfunction seen in pathological

pregnancies is a result of maladaptation to the increased cardiovascular demands of pregnancy in

a woman with normal cardiovascular function, or is predisposed by an inherently abnormal pre-

pregnancy cardiovascular status and its expression modulated through placental release of

vasoactive substances.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

17

Acknowledgments: We wish to thank the participants for taking part at what for many was a

critical and difficult point in their pregnancies

Author contributions: JT was involved in data collection, statistical analysis and drafting the

manuscript. LF, GM, CMM contributed to analysis and interpretation of the manuscript.

PB, CMM, IBW, CCL were involved in project supervision and manuscript editing. CCL

conceived the idea for this study and is the principle investigator.

Funding: JT is funded by Imperial College NHS Trust with support from the Imperial

Healthcare Charities. CCL and PB are supported by the UK National Institute for Health

Research Biomedical Research Centre (BRC) based at Imperial College Healthcare

National Health Service Trust and Imperial College London. CMM is funded, in part by

the NIHR Cambridge BRC. LF is supported by Action Medical Research, Tommy’s and

the Genesis Research Trust.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

18

References:

1. Bokslag A, Teunissen PW, Franssen C, et al. Effect of early-onset preeclampsia on cardiovascular risk in the fifth decade of life. Am J Obstet Gynecol. 2017;216(5):523.e521-523.e527.

2. Fields JA, Garovic VD, Mielke MM, et al. Preeclampsia and cognitive impairment later in life. Am J Obstet Gynecol. 2017;217(1):74.e71-74.e11.

3. Stevens W, Shih T, Incerti D, et al. Short-term costs of preeclampsia to the United States health care system. Am J Obstet Gynecol. 2017;217(3):237-248.e216.

4. Lees C, Marlow N, Arabin B, et al. Perinatal morbidity and mortality in early-onset fetal growth restriction: cohort outcomes of the trial of randomized umbilical and fetal flow in Europe (TRUFFLE). Ultrasound Obstet Gynecol. 2013;42(4):400-408.

5. Khong TY, De Wolf F, Robertson WB, Brosens I. Inadequate maternal vascular response to placentation in pregnancies complicated by pre-eclampsia and by small-for-gestational age infants. Br J Obstet Gynaecol. 1986;93(10):1049-1059.

6. Genbacev O, Joslin R, Damsky CH, Polliotti BM, Fisher SJ. Hypoxia alters early gestation human cytotrophoblast differentiation/invasion in vitro and models the placental defects that occur in preeclampsia. J Clin Invest. 1996;97(2):540-550.

7. Sattar N, Greer IA. Pregnancy complications and maternal cardiovascular risk: opportunities for intervention and screening? BMJ. 2002;325(7356):157-160.

8. Brown MC, Best KE, Pearce MS, Waugh J, Robson SC, Bell R. Cardiovascular disease risk in women with pre-eclampsia: systematic review and meta-analysis. Eur J Epidemiol. 2013;28(1):1-19.

9. Ray JG, Vermeulen MJ, Schull MJ, Redelmeier DA. Cardiovascular health after maternal placental syndromes (CHAMPS): population-based retrospective cohort study. The Lancet. 2005;366(9499):1797-1803.

10. Magnussen EB, Vatten LJ, Lund-Nilsen TI, Salvesen KA, Davey Smith G, Romundstad PR. Prepregnancy cardiovascular risk factors as predictors of pre-eclampsia: population based cohort study. BMJ. 2007;335(7627):978.

11. Melchiorre K, Thilaganathan B. Maternal cardiac function in preeclampsia. Curr Opin Obstet Gynecol. 2011;23(6):440-447.

12. Ghossein-Doha C, van Neer J, Wissink B, et al. Pre-eclampsia: an important risk factor for asymptomatic heart failure. Ultrasound Obstet Gynecol. 2017;49(1):143-149.

13. 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.

14. Verlohren S, Melchiorre K, Khalil A, Thilaganathan B. Uterine artery Doppler, birth weight and timing of onset of pre-eclampsia: providing insights into the dual etiology of late-onset pre-eclampsia. Ultrasound Obstet Gynecol. 2014;44(3):293-298.

15. Tomsin K, Mesens T, Molenberghs G, Peeters L, Gyselaers W. Characteristics of heart, arteries, and veins in low and high cardiac output preeclampsia. Eur J Obstet Gynecol Reprod Biol. 2013;169(2):218-222.

16. Bamfo JE, Kametas NA, Chambers JB, Nicolaides KH. Maternal cardiac function in normotensive and pre-eclamptic intrauterine growth restriction. Ultrasound Obstet Gynecol. 2008;32(5):682-686.

17. Ferrazzi E, Stampalija T, Monasta L, Di Martino D, Vonck S, Gyselaers W. Maternal hemodynamics: a method to classify hypertensive disorders of pregnancy. Am J Obstet Gynecol. 2017.

18. Castleman JS, Ganapathy R, Taki F, Lip GY, Steeds RP, Kotecha D. Echocardiographic Structure and Function in Hypertensive Disorders of Pregnancy: A Systematic Review. Circ Cardiovasc Imaging. 2016;9(9).

19. Bamfo JE, Kametas NA, Chambers JB, Nicolaides KH. Maternal cardiac function in fetal growth-restricted and non-growth-restricted small-for-gestational age pregnancies. Ultrasound Obstet Gynecol. 2007;29(1):51-57.

20. Valensise H, Vasapollo B, Gagliardi G, Novelli GP. Early and late preeclampsia: two different maternal hemodynamic states in the latent phase of the disease. Hypertension. 2008;52(5):873-880.

1

2

3

456789

101112131415161718192021222324252627282930313233343536373839404142434445464748495051

19

21. Harrington K, Cooper D, Lees C, Hecher K, Campbell S. Doppler ultrasound of the uterine arteries: the importance of bilateral notching in the prediction of pre-eclampsia, placental abruption or delivery of a small-for-gestational-age baby. Ultrasound Obstet Gynecol. 1996;7(3):182-188.

22. Tranquilli AL. Early and late-onset pre-eclampsia. Pregnancy Hypertens. 2014;4(3):241.23. Parry S, Sciscione A, Haas DM, et al. Role of early second-trimester uterine artery Doppler screening to

predict small-for-gestational-age babies in nulliparous women. Am J Obstet Gynecol. 2017;217(5):594.e591-594.e510.

24. Chitty LS, Altman DG, Henderson A, Campbell S. Charts of fetal size: 3. Abdominal measurements. Br J Obstet Gynaecol. 1994;101(2):125-131.

25. Parra-Cordero M, Lees C, Missfelder-Lobos H, Seed P, Harris C. Fetal arterial and venous Doppler pulsatility index and time averaged velocity ranges. Prenat Diagn. 2007;27(13):1251-1257.

26. Gabrielsen A, Videbaek R, Schou M, Damgaard M, Kastrup J, Norsk P. Non-invasive measurement of cardiac output in heart failure patients using a new foreign gas rebreathing technique. Clin Sci (Lond). 2002;102(2):247-252.

27. Mahendru AA, Everett TR, Wilkinson IB, Lees CC, McEniery CM. A longitudinal study of maternal cardiovascular function from preconception to the postpartum period. J Hypertens. 2014;32(4):849-856.

28. de Greeff A, Beg Z, Gangji Z, Dorney E, Shennan AH. Accuracy of inflationary versus deflationary oscillometry in pregnancy and preeclampsia: OMRON-MIT versus OMRON-M7. Blood Press Monit. 2009;14(1):37-40.

29. O'Brien E, Asmar R, Beilin L, et al. European Society of Hypertension recommendations for conventional, ambulatory and home blood pressure measurement. J Hypertens. 2003;21(5):821-848.

30. Hickson SS, Butlin M, Broad J, Avolio AP, Wilkinson IB, McEniery CM. Validity and repeatability of the Vicorder apparatus: a comparison with the SphygmoCor device. Hypertens Res. 2009;32(12):1079-1085.

31. Mahendru AA, Everett TR, McEniery CM, Wilkinson IB, Lees CC. The feasibility of prospectively studying maternal cardiovascular changes from before conception. Hypertens Res. 2013;36(8):698-704.

32. Foo FL, Collins A, McEniery CM, Bennett PR, Wilkinson IB, Lees CC. Preconception and early pregnancy maternal haemodynamic changes in healthy women in relation to pregnancy viability. Hum Reprod. 2017;32(5):985-992.

33. Sonek J, Krantz D, Carmichael J, et al. First-trimester screening for early and late preeclampsia using maternal characteristics, biomarkers, and estimated placental volume. Am J Obstet Gynecol. 2018;218(1):126.e121-126.e113.

34. Lees CC, Marlow N, van Wassenaer-Leemhuis A, et al. 2 year neurodevelopmental and intermediate perinatal outcomes in infants with very preterm fetal growth restriction (TRUFFLE): a randomised trial. Lancet. 2015;385(9983):2162-2172.

35. Mahendru AA, Foo FL, McEniery CM, Everett TR, Wilkinson IB, Lees CC. Change in maternal cardiac output from preconception to mid-pregnancy is associated with birth weight in healthy pregnancies. Ultrasound Obstet Gynecol. 2017;49(1):78-84.

36. Magee LA, von Dadelszen P, Singer J, et al. The CHIPS Randomized Controlled Trial (Control of Hypertension in Pregnancy Study): Is Severe Hypertension Just an Elevated Blood Pressure? Hypertension. 2016;68(5):1153-1159.

37. Sharma C, Soni A, Gupta A, Verma A, Verma S. Hydralazine vs nifedipine for acute hypertensive emergency in pregnancy: a randomized controlled trial. Am J Obstet Gynecol. 2017;217(6):687.e681-687.e686.

38. Butters L, Kennedy S, Rubin PC. Atenolol in essential hypertension during pregnancy. BMJ. 1990;301(6752):587-589.

39. Lip GY, Beevers M, Churchill D, Shaffer LM, Beevers DG. Effect of atenolol on birth weight. Am J Cardiol. 1997;79(10):1436-1438.

40. Meidahl Petersen K, Jimenez-Solem E, Andersen JT, et al. beta-Blocker treatment during pregnancy and adverse pregnancy outcomes: a nationwide population-based cohort study. BMJ open. 2012;2(4).

41. Stott D, Bolten M, Paraschiv D, Papastefanou I, Chambers JB, Kametas NA. Longitudinal hemodynamics in acute phase of treatment with labetalol in hypertensive pregnant women to predict need for vasodilatory therapy. Ultrasound Obstet Gynecol. 2017;49(1):85-94.

42. Valensise H, Vasapollo B, Novelli GP, et al. Maternal and fetal hemodynamic effects induced by nitric oxide donors and plasma volume expansion in pregnancies with gestational hypertension complicated by intrauterine growth restriction with absent end-diastolic flow in the umbilical artery. Ultrasound Obstet Gynecol. 2008;31(1):55-64.

123456789

1011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556

20

43. Oyelese KO, Black RS, Lees CC, Campbell S. A novel approach to the management of pregnancies complicated by uteroplacental insufficiency and previous stillbirth. Aust N Z J Obstet Gynaecol. 1998;38(4):391-395.

44. Valensise H, Larciprete G, Vasapollo B, et al. Nifedipine-induced changes in body composition in hypertensive patients at term. Eur J Obstet Gynecol Reprod Biol. 2003;106(2):139-143.

45. 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.

46. Kaihura C, Savvidou MD, Anderson JM, McEniery CM, Nicolaides KH. Maternal arterial stiffness in pregnancies affected by preeclampsia. Am J Physiol Heart Circ Physiol. 2009;297(2):H759-764.

47. Khalil A, Jauniaux E, Harrington K. Antihypertensive therapy and central hemodynamics in women with hypertensive disorders in pregnancy. Obstet Gynecol. 2009;113(3):646-654.

48. Hausvater A, Giannone T, Sandoval YH, et al. The association between preeclampsia and arterial stiffness. J Hypertens. 2012;30(1):17-33.

49. Nurnberger J, Keflioglu-Scheiber A, Opazo Saez AM, Wenzel RR, Philipp T, Schafers RF. Augmentation index is associated with cardiovascular risk. J Hypertens. 2002;20(12):2407-2414.

50. Kingwell BA, Waddell TK, Medley TL, Cameron JD, Dart AM. Large artery stiffness predicts ischemic threshold in patients with coronary artery disease. Journal of the American College of Cardiology. 2002;40(4):773-779.

51. Blacher J, Asmar R, Djane S, London GM, Safar ME. Aortic pulse wave velocity as a marker of cardiovascular risk in hypertensive patients. Hypertension. 1999;33(5):1111-1117.

52. Mlynarczyk M, Chauhan SP, Baydoun HA, et al. The clinical significance of an estimated fetal weight below the 10th percentile: a comparison of outcomes of <5th vs 5th-9th percentile. Am J Obstet Gynecol. 2017;217(2):198.e191-198.e111.

53. Sackner MA, Greeneltch D, Heiman MS, Epstein S, Atkins N. Diffusing capacity, membrane diffusing capacity, capillary blood volume, pulmonary tissue volume, and cardiac output measured by a rebreathing technique. Am Rev Respir Dis. 1975;111(2):157-165.

54. Christensen P, Clemensen P, Andersen PK, Henneberg SW. Thermodilution versus inert gas rebreathing for estimation of effective pulmonary blood flow. Crit Care Med. 2000;28(1):51-56.

55. Peyton PJ, Thompson B. Agreement of an inert gas rebreathing device with thermodilution and the direct oxygen Fick method in measurement of pulmonary blood flow. J Clin Monit Comput. 2004;18(5-6):373-378.

56. Jakovljevic DG, Nunan D, Donovan G, Hodges LD, Sandercock GR, Brodie DA. Comparison of cardiac output determined by different rebreathing methods at rest and at peak exercise. Eur J Appl Physiol. 2008;102(5):593-599.

57. Mahendru AA, Everett TR, Wilkinson IB, Lees CC, McEniery CM. Maternal cardiovascular changes from pre-pregnancy to very early pregnancy. J Hypertens. 2012;30(11):2168-2172.

58. Foo FL, McEniery CM, Lees C, Khalil A. Assessment of arterial function in pregnancy: recommendations of the International Working Group on Maternal Hemodynamics. Ultrasound Obstet Gynecol. 2017;50(3):324-331.

123456789

101112131415161718192021222324252627282930313233343536373839

40

41424344