Journal of Perinatology (2017) 00, 1–8

© 2017 Nature America, Inc., part of Springer Nature. All rights reserved 0743-8346/17
www.nature.com/jp

ORIGINAL ARTICLE
Pregnancy-induced hypertension and neonatal outcomes:
a systematic review and meta-analysis
A Razak1, A Florendo-Chin1, L Banfield2, MG Abdul Wahab1, S McDonald3, PS Shah4 and A Mukerji1

OBJECTIVE: Pregnancy-induced hypertension (PIH) is associated with preterm delivery but its independent impact on neonatal
outcomes remains unclear. We sought to systematically review and meta-analyze clinical outcomes of preterm infants o 37 weeks’
gestation born to mothers with and without PIH.
STUDY DESIGN: Medline, Embase, PsychINFO and CINAHL were searched from January 2000 to October 2016. Studies with low-
moderate risk of bias reporting neonatal outcomes based on PIH as primary exposure variable were included. Data were extracted
independently by two co-authors.
RESULTS: PIH was associated with lower mortality (3 studies; adjusted odds ratio (aOR) 0.65; 95% confidence interval (CI) 0.54 to
0.79), lower severe retinopathy of prematurity (ROP) (2 studies; aOR 0.83; 0.72 to 0.96) and lower severe brain injury (2 studies;
unadjusted OR (uOR) 0.57; 0.49 to 0.66). No association between PIH and short-term respiratory outcomes, bronchopulmonary
dysplasia (BPD) or necrotizing enterocolitis (NEC) was identified. In subgroup analysis among infants o 29 weeks’ gestation, BPD
odds were higher (3 studies; aOR 1.15; 1.06 to 1.26), whereas mortality lower (2 studies; aOR 0.73; 0.69 to 0.77). In subgroup analysis
limited to severe PIH, odds of mortality (3 studies; uOR 2.36; 1.07 to 5.22) and invasive ventilation (3 studies; uOR 3.26; 1.11 to 9.61)
were higher. In subgroup analysis limited to preeclampsia, odds of BPD (3 studies; uOR 1.21; 95% CI:1.03 to 1.43) and NEC were
higher (3 studies; uOR 2.79; 95% CI:1.57 to 4.96).
CONCLUSION: PIH was associated with reduced odds of mortality and ROP (all infants), but higher odds for BPD (o 29 weeks’
gestation). The paradoxical reduction in mortality may be due to survival bias and deserves further exploration in future studies.
Journal of Perinatology advance online publication, 2 November 2017; doi:10.1038/jp.2017.162

INTRODUCTION
Pregnancy-induced hypertension (PIH) complicates 5% of all
pregnancies and is commonly associated with preterm birth.1,2
The auxiliary effects of PIH in relation to neonatal outcomes
remain unclear. It is postulated that it adapts fetus to survive by
accelerating organ maturation;3 however, there is conflicting data
on its association with mortality,4–6 bronchopulmonary dysplasia
(BPD),5,7 intraventricular hemorrhage (IVH)5,8 and necrotizing
enterocolitis (NEC).5,9 A recent systematic review found no
association of PIH with retinopathy of prematurity (ROP)10 but a
subsequent large collaborative study found lowered association.5
In addition, it remains unclear whether its impact on neonatal
outcomes is influenced by gestational age (GA) or severity of
PIH.4,7,8
This systematic review and meta-analysis was conducted to
summarize the impact of PIH on neonatal outcomes based on
existing literature, while understanding its limitations. There was a
particular focus on delineating the degree of impact across GA
and the influence of severity of PIH. Summative outcomes data
from this comprehensive review will help clinicians develop an
understanding of the current state of the literature along with its
limitations and develop an appreciation of the challenges in
inferring the results, whereas it will help researchers by high-
lighting the limitations and current knowledge gaps, and help in
devising future studies.
METHODS
This meta-analysis was conducted and reported as per the guidelines from
PRISMA (preferred reporting items for systematic reviews and meta-
analyses).11
Search strategy
A systematic review of the literature was conducted using appropriate pre-
specified search terms (Supplementary File 1) within the following
databases: Medline, Embase, PsychINFO and CINAHL from January 2000
through 5 October 2016. The search was limited to articles published since
2000, to ensure relevance to contemporary neonatal practices.
Study selection
Cohort and case–control studies published in English in a peer-reviewed
journal, which evaluated the neonatal outcomes based primarily on the
presence or absence of exposure variable (PIH) only were eligible for
inclusion in the study. Studies reporting on preterm infants o37 weeks’
GA (including studies with mixed infant population where data could be
extracted for preterm infants) were included. Studies where PIH was just a
covariate (as only studies where PIH as a covariate influenced outcomes
may have selectively reported this, leading to potential bias), descriptive
studies, studies that enrolled patients before 1990 or were published
before 2000, narrative reviews, dissertations, case reports, case series,
letters to the editor, conference abstracts, cross-sectional studies and
studies with a high risk of bias were deemed ineligible.

1
Department of Pediatrics, McMaster University, Hamilton, ON, Canada; 2Faculty of Health Science, Health Sciences Library, McMaster University, Hamilton, ON, Canada;
3
Department of Obstetrics and Gynecology, McMaster University, Hamilton, ON, Canada and 4Department of Pediatrics, University of Toronto, Toronto, ON, Canada.
Correspondence: Dr A Mukerji, Department of Pediatrics, McMaster University, 1280 Main Street West, HSC-4F1E, Hamilton, ON, Canada L8S 4K1.
E-mail: mukerji@mcmaster.ca
Received 11 May 2017; revised 8 August 2017; accepted 29 August 2017
 Neonatal outcomes in PIH: a meta-analysis
A Razak et al
2
Exposure and comparison
The exposure of interest was PIH with or without proteinuria, including PIH
superimposed on previous chronic hypertension. Studies where maternal
hypertension was not sub-classified were included. Studies reporting only
on isolated chronic hypertension were excluded. PIH was defined as new-
onset hypertension (systolic blood pressure ⩾ 140 mm Hg or diastolic
blood pressure ⩾ 90 mm Hg) after 20 weeks’ gestation with or without
proteinuria. Proteinuria was defined as the excretion of 300 mg or more of
protein in a 24 h urine collection or a protein/creatinine ratio of at least 0.3
or a determination of 1+ in dipstick test.12 PIH with proteinuria was
considered as preeclampsia. PIH was considered severe if any one of the
following were present: a systolic blood pressure ⩾ 160 mm Hg or a
diastolic blood pressure ⩾ 110 mm Hg on two occasions ⩾ 4 h apart in a
pregnant woman who is on bed rest; progressive renal insufficiency (serum
creatinine concentration > 1.1 mg dl − 1 or a doubling of the serum
creatinine concentration in the absence of other renal disease); pulmonary
edema; impairment of liver function; new onset visual or cerebral
disturbances; pain in the epigastric area or right upper quadrant; platelet
count below 100 000 μl − 1; or HELPP syndrome (hemolysis, elevated liver
enzymes and low platelet count).12
Outcomes
Studies were included as long as they reported at least one of the
following outcomes of interest:
1) Mortality: before 28 to 30 days or before discharge
2) BPD defined as oxygen or positive pressure ventilation dependence at
36 weeks post-menstrual age13
3) Short-term respiratory morbidity with each of following evaluated:
a) Any invasive ventilation support or surfactant therapy
b) Duration of invasive ventilation support
c) Duration of oxygen therapy
4) Severe brain injury (SBI) defined as grade III or IV IVH Papile grading,14
and/or grade II or higher periventricular leukomalacia (PVL) de Vries’
grading.15 In addition, each of these components was also evaluated
individually
a) Grade III or IV IVH
b) Grade II or higher PVL
5) Severe ROP defined as stage 3 or above, or any stage requiring
treatment16
6) NEC defined as stage 2A or above using modified Bell’s criteria17
Data extraction
Author 1 (AR) searched the aforementioned databases with help of a
reference librarian (LB). A final list of literature after the search was
assimilated using a reference management software (EndNote version 7.7,
Thomson Reuters, New York, NY, USA). AR screened the titles and abstracts
to determine relevant reports and the shortlisted literature was evaluated
to determine eligibility for inclusion based on predefined inclusion and
exclusion criteria independently by two authors (AR and AC)
(Supplementary File 2). Of the studies deemed eligible for final inclusion,
each study was examined for information including study design,
inclusion, key characteristics and outcomes, using a standardized data
collection form. No blinding strategies were employed. Any discrepancies
were resolved by consensus and involving a third author (AM).
Assessment of risk of bias
The risk of bias for included studies was assessed using a modified
Newcastle–Ottawa scale (Supplementary File 3) and the following domains
were evaluated: selection, comparability and outcome. A priori, a score of
⩾ 7/9 was deemed low risk, a score 4 to 6/9 was deemed moderate risk and
a score of ⩽ 3/9 was deemed as high risk of bias. Selection bias as defined
by inclusion of a very restricted or specific neonatal population cohort not
representative of a typical neonatal patient delivered at a perinatal centre
(Supplementary File 3) was automatically deemed high risk of bias,
regardless of the score.18 Assessment of risk of bias was performed by two
authors (AR and AC) and any conflicts were resolved through consensus
and involvement of third author (AM) as needed. Publication bias was
assessed using a funnel plot if > 10 studies were included for any individual
meta-analysis.
Journal of Perinatology (2017), 1 – 8
Assessment of heterogeneity
Variability in the study design, participants, exposures, outcomes assessed
and biases was evaluated qualitatively, to determine clinical and
methodological heterogeneity and to assess the appropriateness of
pooling studies together by two authors (AR and AC) independently and
discrepancies were resolved by involving a third author (AM). Studies
deemed not comparable due to clinical or methodological heterogeneity
were not meta-analyzed. Furthermore, pooled forest plots were assessed
qualitatively to assess for heterogeneity. Once qualitative assessment of
heterogeneity was completed as described, statistical heterogeneity was
determined for studies grouped together using I2 values (derived from the
chi-squared Q-statistic). Statistical heterogeneity was considered significant
if I2 value was > 75%.
Data synthesis and statistical analysis
The effect measure (odds ratio (OR) or mean) and variance were calculated
for each included study, and each study was assigned a weight based on
variance and statistical heterogeneity using the formula weight = (1/
(variance+statistical heterogeneity)). The s.e. of the pooled measure effect
estimate (pooled ORs and mean differences) was used to generate 95%
confidence intervals (CIs) and P-value. The Hozo method was used to
convert medians and ranges to means and standard deviations,
respectively.19 Meta-analyses were performed using Review Manager 5.3
(Cochrane Collaboration, Nordic Cochrane centre, Copenhagen, Denmark)
when appropriate using random-effects model to yield pooled ORs and
mean differences. Subgroup analyses for all outcomes using aforemen-
tioned methodology, contingent on availability of data, were performed
for (a) extreme prematurity ( o29 weeks GA), (b) severe PIH (all preterm
infants and o29 weeks GA) and (c) preeclampsia (all preterm infants and
o29 weeks GA).
RESULTS
The results of the database searches and the study selection log is
as shown in Figure 1. Of 4757 records identified through database
searches, 26 studies were included in the meta-analysis (details of
included studies outlined in Supplementary File 4). The list of
studies excluded following detailed review, including the reasons
for exclusion, are provided in Supplementary File 5. Of 26 included
studies, 18 had low risk of bias5–7,20–34 and remainder (8 studies)
had moderate risk of bias.4,8,9,35–39 The risk of bias for included
studies is as shown in Table 1. Pooled data for each pre-defined
outcome from these comparisons are described below and also
delineated in further detail in Table 2, with select forest plots
shown in Figure 2 and Figure 3. Funnel plots for two meta-
analyses with 10 or more studies did not show evidence of
publication bias.
Association between PIH and mortality
The unadjusted odds were not different in neonates born to
mothers with and without PIH.4–6,8,20,21,24,25,29–32,38,39 However, the
adjusted odds were significantly lower with maternal PIH (aOR:
0.65; 95% CI: 0.54 to 0.79; I2 = 93%; 3 studies; 1 804 382
subjects).5,21,25 Similarly, the aOR was lower in subgroup analysis
for o 29 weeks GA (aOR: 0.73; 95% CI: 0.69 to 0.77; I2 = 0%; 2
studies).5,25 On the contrary, the unadjusted odds were higher in
subgroup analysis for severe PIH (uOR: 2.36; 95% CI: 1.07 to 5.22;
I2 = 43%; 3 studies; 493 subjects)4,6,8 and preeclampsia
(o 29 weeks GA only) (uOR: 1.39; 95% CI: 1.099 to 1.77; I2 = 0%;
2 studies; 4634 subjects).6,31 No association was found for
unadjusted data in subgroup analysis for o 29 weeks GA5,6,25,31
and for preeclampsia (all preterm infants).4,6,8,20,29–31,39
Association between PIH and BPD
There was no association found for both unadjusted5,7,20,26,28,31,34,37,38
and adjusted5,7,26,34 comparison for all preterm infants. However, the
association was significant in subgroup analysis for o29 weeks’
gestation, for both unadjusted (uOR: 1.34; 95% CI: 1.22 to 1.48;
I2 = 16%; 4 studies)5,7,26,31 and adjusted (aOR: 1.15; 95% CI: 1.06 to
© 2017 Nature America, Inc., part of Springer Nature.

Figure 1. Study flow diagram outlining stages of search results and filtering process (as per Preferred Reporting Items for Systematic Reviews
and Meta-Analyses (PRISMA) guidelines).
1.26, I2 = 0%; 3 studies)5,7,26 comparisons. Similarly, a significant
association was found in subgroup analysis for preeclampsia in less
than 29 weeks GA (uOR: 1.21; 95% CI: 1.03 to 1.43; I2 = 0%; 3
studies).7,31,37 No association was found between preeclampsia and
BPD for all preterm infants (both adjusted5,7,26,34 and unadjusted
data,5,7,20,26,28,31,34,37,38 and adjusted data for preeclampsia in less
than 29 weeks GA5,7,26).
Association between PIH and short term respiratory
The odds for invasive ventilation or surfactant therapy,4,6–8,23,
28,29,35,36,39,40
duration of invasive ventilation6,7,20,27,35 and duration
of oxygen therapy7,20,27 were not different for all preterm infants.
Similarly, no association was found between invasive ventilation and/
or surfactant therapy,4,6–8,28,29,35,36,39 duration of invasive
ventilation6,7,20,27,35 and duration of oxygen therapy,7,20,27 in pree-
clampsia subgroup. However, a significant association was found
between severe PIH and invasive ventilation and/or surfactant
therapy (uOR: 3.26; 95% CI: 1.11 to 9.61; I2 = 80%; 3 studies; 493
subjects).4,6,8
Association between PIH and brain injury
The odds for SBI were lower in PIH (uOR: 0.57; 95% CI: 0.49 to 0.66;
I2 = 0%; 2 studies; 28 603 subjects).5,29 However, no independent
association was found for IVH (all preterm infants4,6,8,28,29,31,35,38
and o 29weeks GA6,31) and cystic PVL.29,31
No association was found in subgroup analysis for severe PIH
and IVH.4,6,8 Similarly, no association was found in subgroup
analysis between preeclampsia and IVH (all patients4,6,8,28,29,31,35
and o 29 weeks GA6,31) or PVL.29,31
Association between PIH and severe ROP
No difference was found on unadjusted comparison.5,27,33,38
However, the adjusted odds were lower in neonates born to PIH
(aOR: 0.83; 95% CI: 0.72 to 0.96; I2 = 0%; 2 studies; 33 564
subjects).5,22 No association was found in subgroup analysis for
preeclampsia.27,33
© 2017 Nature America, Inc., part of Springer Nature.
Neonatal outcomes in PIH: a meta-analysis
A Razak et al
3
Association between PIH and NEC
Both uOR4,5,8,9 and aOR5,9 were not different. Similarly, no
association was found between severe PIH and NEC.4,8 However,
when the analysis was restricted to preeclampsia, a significant
association was found on unadjusted comparison (uOR: 2.79; 95%
CI: 1.57 to 4.96; I2 = 0%; 3 studies; 878 subjects).4,8,9
DISCUSSION
In this comprehensive systematic review and meta-analysis, we
identified that among all preterm infants PIH is associated with
lower odds of mortality, ROP and SBI (unadjusted data only for
latter), whereas no association was found for BPD or NEC.
However, the subgroup analysis for o29 weeks’ gestation, there
were higher odds of BPD and lower odds of mortality. In the
subgroup analysis restricted to preeclampsia, there were higher
odds of BPD ( o29 weeks’ GA only) and NEC (unadjusted data for
both). In the subgroup analysis for severe PIH, the odds of
mortality and invasive ventilation or surfactant therapy were
higher (unadjusted data).
This is the first review to summarize the association of spectrum
of maternal PIH with a wide range of neonatal outcomes that
highlights the heterogeneity and limitations of the current body of
literature. There have been a number of purported biological
rationales for why neonates born to mothers with PIH may have
superior outcomes. A commonly proposed argument favouring
PIH is fetal stress in utero accelerating organ maturity, thereby
adapting the fetus better for survival.3 It is important to note that
other causes of preterm birth (representing the normotensive
comparison group) include chorioamnionitis, placental abruption
and various maternal medical conditions are linked with
inflammation, ischemia and poor outcomes.41 In addition, mothers
with PIH receive more stringent monitoring with augmented
prenatal surveillance, with potentially higher rates of antenatal
corticosteroid and magnesium sulfate administration, and earlier
deliveries in cases of recognized fetal intolerance of labour. In fact,
obstetrical interventions leading to earlier deliveries (and there-
fore higher survival) may in fact be confounding the true effect of
PIH by leading to a survival bias. This has been delineated in detail,
whereby a model using the denominator of the ‘fetus at risk’ has
been suggested to eliminate such a confounding bias.42
Journal of Perinatology (2017), 1 – 8
 Neonatal outcomes in PIH: a meta-analysis
A Razak et al
4
Table 1. Assessment of risk of bias of included studies using a modified Newcastle–Ottawa scale

Study (author, Study type GA in weeks Sample size Selection Comparable Outcome Total score Overall
year) (n) (max 4) (max 2) (max 3) (max 9) risk

Aslan et al.4 Retrospective o 36 256 4 0 2 6 Moderate

cohort
Cetinkaya et al.20 Retrospective o 36 84 4 1 2 7 Low
cohort
Cetinkaya et al.9 Prospective cohort o 37 501 4 0 2 6 Moderate
Chen et al.21 Retrospective 24-36 1.688 Million 4 1 3 7 Low
cohort
Cheng et al.35 Retrospective o 32 89 3 0 2 5 Moderate
cohort
5
Gemmel et al. Retrospective 24–29 27846 4 0 3 7 Low
cohort
Hansen et al.34 Prospective cohort 23–32 107 4 0 3 7 Low
Hiett et al.6 Retrospective 24–28 116 4 1 3 8 Low
cohort
Huang et al.22 Retrospective o 36 5718 4 0 3 7 Low
cohort
Kalay et al.36 Retrospective o 32 54 3 0 2 5 Moderate
cohort
Kim et al. 8
Retrospective o 37 121 4 0 2 6 Moderate
cohort
Liu et al.23 Retrospective 24–36 3367 4 0 3 7 Low
cohort
24
Luo et al. Retrospective 20–42 0.502 Million 4 0 3 7 Low
cohort
McBride et al.25 Retrospective 22–30 a
88275 4 0 3 7 Low
cohort
O Shea et al. 26
Retrospective o 37 749 4 0 3 7 Low
cohort
Ozkan et al.27 Prospective cohort o 32 501 4 0 3 7 Low
Ozkan et al.37 Prospective cohort o 32 332 4 0 2 6 Moderate
Park et al.38 Retrospective 24-32 191 3 0 3 6 Moderate
cohort
Procianoy et al.39 Prospective cohort o 37 911 4 0 2 6 Moderate
Silveira et al.28 Prospective cohort o 37 86 4 0 3 7 Low
Spinillo et al.29 Retrospective 24-35 757 4 0 3 7 Low
cohort
Tastekin et al.30 Prospective cohort o 36 27 4 1 2 7 Low
Tokumasu et al.31 Retrospective 23-27 4518 4 0 3 7 Low
cohort
Wang et al.32 Nested case control o 37 439 4 0 3 7 Low
Yen et al.7 Retrospective o 37 5753 4 0 3 7 Low
cohort
Zayed et al.33 Retrospective o 37 4996 4 0 3 7 Low
cohort
Abbreviations: GA, gestational age. aIn subgroup analyses of o29 weeks, only GA up to 27 weeks was included from this study.

The reduction in mortality for all infants including o 29 weeks’
gestation in PIH may be explained by the aforementioned reasons.
However, the association appears inverse in infants with severe
PIH, albeit based on unadjusted data only. This could be explained
due to increased inflammation occurring as a result of increased
oxidative stress43 and increased pro-inflammatory cytokines in
HELPP syndrome. In addition, it is conceivable that mothers with
severe PIH may present acutely in suboptimal conditions for
delivery, including delivery at a non-tertiary centre, as well as
lower rates of complete course of antenatal corticosteroid
administration due to the urgency of delivering such fetuses.
Finally, mothers with severe PIH may deliver at a lower GA, which
itself could act as an effect modifier. Despite these aforemen-
tioned limitations and uncertainties, it may be important to
consider PIH in risk adjustment models of mortality. However,
based on the heterogeneity identified in our review, it would be
prudent that the severity (and type) of PIH, as well as GA of infants
are taken into consideration, due to the differential association of
PIH and mortality based on such factors.
The association of BPD in infants born to mothers with PIH was
restricted to preterm infants o 29 weeks’ gestation, which may be
explained simply by the relatively low prevalence of BPD in
neonates born at higher GAs. The increased odds of BPD in PIH
could be attributed to impaired angiogenesis and lung growth.
The elevated soluble vascular growth factor (s-flt-1) in PIH, an anti-
angiogenic factor, may prevent vascular endothelial growth factor
signaling, which is one of the critical components of lung
maturation.32–34 Conversely, this anti-angiogenic milieu may
confer protection against development of ROP, as identified in
the present review. In contrast to our study, a recent meta-analysis
did not find association between any stage of ROP/severe ROP
and PIH.10 The contrasting results may be due to the influence of
large collaborative study published subsequent to the other
review.5 The previous meta-analysis included all studies providing
ORs based on PIH as an exposure variable or a covariate, which in
contrast to our review where only studies where PIH was analysed
as primary exposure variable were included. In addition, our meta-
analysis evaluated specifically the association between PIH and
severe type of ROP rather any stage of ROP.

Journal of Perinatology (2017), 1 – 8 © 2017 Nature America, Inc., part of Springer Nature.
 Neonatal outcomes in PIH: a meta-analysis
A Razak et al
5
Table 2. Pooled analysis of outcomes including subgroup analysis

ALL PIH Severe PIH Preeclampsia

Mortality (all patients)
Mortality (o 29 weeks GA)
BPD (all patients)
BPD ( o 29 weeks GA)
Invasive ventilation and/or
surfactant therapy (o 29 weeks)
Days of invasive ventilation (All
patients)
Days of invasive ventilation
(o29 weeks)
Days of oxygen therapy (all
patients)
Days of oxygen therapy
(o29 weeks)
SBI (all patients)
SBI ( o 29 weeks GA)
Severe IVH (all patients)
Severe IVH ( o 29 weeks GA)
Cystic PVL (all patients)
Cystic PVL ( o 29 weeks GA)
ROP (all patients)
ROP (o 29 weeks GA)
NEC (all patients)
NEC ( o 29 weeks GA)
Abbreviations: aOR, adjusted odds ratio; BPD, bronchopulmonary dysplasia; CI, confidence interval; IVH, intraventricular hemorrhage; MD, mean difference;
NA, not applicable; NEC, necrotizing enterocolitis; PVL, periventricular leukomalacia; ROP, retinopathy of prematurity; SBI, severe brain injury; uOR, unadjusted
odds ratio Note: (a) Number of subjects only provided when available from original studies; (b) confounders adjusted for aOR in the included studies provided
in Supplementary File 4 aCI was used to generate point estimate due to discrepancy between the point estimate and CI provided in the study.5,9
uOR:0.87; 95% CI: 0.66–1.14;
I2 = 93%; 14 studies; 2 313 915
subjects
aOR: 0.65; 95% CI: 0.54–0.79;
I2 = 93%; 3 studies; 1 804 382
subjects
uOR: 1.10; 95% CI: 0.86–1.40; I2
= 90%; 4 studies
aOR: 0.73; 95% CI: 0.69–0.77;
I2 = 0%; 2 studies
uOR: 1.14; 95% CI: 0.76–1.70;
I2 = 94%; 9 studies; 38 366 subjects
aOR: 1.10; 95% CI: 0.76–1.59;
I2 = 83%; 4 studies; 34 455 subjects
uOR: 1.34; 95% CI: 1.22–1.48;
I2 = 16%; 4 studies
aOR: 1.15; 95% CI: 1.06–1.26;
I2 = 0%; 3 studies
NA
MD: 6.33; 95% CI: − 3.09 to 15.74;
I2 = 97%; 5 studies; 6427 subjects
NA
MD: − 1.79; 95% CI: − 17.59 to 14.02;
I2 = 97%; 3 studies; 6222 subjects
NA
uOR: 0.57; 95% CI: 0.49–0.66;
I2 = 0%; 2 studies; 28 603 subjects
NA
uOR: 0.90; 95% CI: 0.43–1.91;
I2 = 59%; 8 studies; 4834 subjects I2 = 83%; 3 studies; 493 subjects
uOR: 0.71; 95% CI: 0.42–1.18;
I2 = 0%; 2 studies; 3334 subjects
uOR:0.77; 95% CI:0.45-1.34; I = 0%;
2
2 studies; 3975 subjects
N/A
uOR: 1.12; 95% CI: 0.68–1.85;
I = 58%; 4 studies; 33 418 subjects
2
aOR: 0.83; 95% CI: 0.72–0.96;
I2 = 0%; 2 studies; 33 564 subjects
NA
uOR: 1.89; 95% CI: 0.82–4.37;
I2 = 76%; 4 studies; 28 724 subjects I2 = 0%; 2 studies; 377 subjects
aOR: 0.96; 95% CI: 0.80–1.14;
I2 = 0%; 2 studies; 28 347 subjectsa
NA
uOR: 2.36; 95% CI: 1.07–5.22; uOR: 1.12; 95% CI: 0.72–1.76; I2 = 72%;
I2 = 43%; 3 studies; 493 subjects 8 studies; 6790 subjects
NA uOR: 1.39; 95% CI: 1.099–1.77; I2 = 0%;
2 studies; 4634 subjects
NA uOR: 1.17; 95% CI: 0.70–1.96; I2 = 90%;
7 studies; 10 329 subjects
aOR: 1.28; 95% CI: 0.61–2.69; I2 = 82%;
3 studies; 6609 subjects
NA uOR: 1.21; 95% CI: 1.03–1.43;
I2 = 0%; 3 studies
aOR: 1.09; 95% CI: 0.83–1.45; I2 = 0%; 2
studies
NA NA
NA MD: 6.33; 95% CI: − 3.09 to 15.74;
I2 = 97%; 5 studies; 6427 subjects
NA NA
NA MD: − 1.79; 95% CI: − 17.59 to 14.02;
I2 = 97%; 3 studies; 6222 subjects
NA NA
NA NA
NA NA
uOR: 2.05; 95% CI: 0.22–18.85; uOR: 0.94; 95% CI:0.40-2.20; I2 = 65%;
7 studies; 4643 subjects
NA uOR:0.71; 95% CI: 0.42–1.18; I2 = 0%; 2
studies; 3334 subjects
NA uOR: 0.77; 95% CI: 0.45–1.34; I2 = 0%; 2
studies; 3975 subjects
NA NA
NA uOR: 1.60; 95% CI: 0.66–3.87; I2 = 64%;
2 studies; 5381 subjects
NA NA
uOR: 3.21; 95% CI: 0.78–13.23; uOR: 2.79; 95% CI: 1.57–4.96;
I2 = 0%; 3 studies; 878 subjects
NA NA

The lower odds of SBI associated with PIH could be attributed to
selection intervention (higher antenatal steroids) as mentioned
above, administration of magnesium sulfate in PIH44 and
increased inflammation linked with other causes of preterm birth
(chorioamnionitis). Interestingly, when evaluating the individual
components comprising SBI—that is, severe IVH and cystic PVL, no
significant association was found. This may be a result of much
smaller numbers of subjects from whom this individualized data
was available and as such may be underpowered. The collabora-
tive study by Gemmel et al.5 have reported SBI but not reported
data on IVH and PVL individually, and this study is the major
contribution for pooled analysis in SBI (study population 27 846).
As such, non-inclusion of this study in individual meta-analysis for
IVH or PVL may have resulted in lack of association. In addition,
this study also provided adjusted comparison which showed
lower SBI in PIH group but no other studies have provided the
adjusted comparison hence no pooled analysis was performed.
The higher odds of NEC were found only when the exposure
was restricted to preeclampsia (unadjusted data). In PIH, the
placental circulation is compromised leading to fetal hypoxia and
oxidative stress, which is an essential element linked with NEC.
However, if this were to be the hypothesis, the association should
have been found with severe PIH(HELPP/severe preeclampsia),
which in fact is associated with higher oxidative stress43 and
increased pro-inflammatory cytokines. It is possible that the lack of
association could be due to smaller number of subjects in severe
PIH. The occurrence of IUGR is higher is severe PIH compared with
preeclampsia and IUGR is an important risk factor in development

© 2017 Nature America, Inc., part of Springer Nature. Journal of Perinatology (2017), 1 – 8
 Neonatal outcomes in PIH: a meta-analysis
A Razak et al
6

Figure 2. Top: forest plot for adjusted data of studies examining the association of PIH and mortality in all preterm infants. Bottom: forest plot
for adjusted data of studies examining the association of PIH and mortality in preterm infants o 29 weeks GA.

Figure 3. Top: forest plot for adjusted data of studies examining the association of PIH and BPD in all preterm infants. Bottom: forest plot for
adjusted data of studies examining the association of PIH and BPD in preterm infants o29 weeks GA.

of NEC. IUGR infants postnatally are on cautious feeding regimen
practiced in many intensive care units and it is possible that these
practices may act as an effect modifier and henceforth masking
the association of NEC in severe PIH. In addition, there are many
other factors that might have a role in development of NEC in
preterm infants. One such important factor is feeding policies
including formula or donor milk, feeding increment and feeding
initiation. These factors were not adjusted in many studies and
hence limiting the interpretation of the results.
This review did not find any association of PIH with short term
respiratory outcomes, which contradicts the potential explanation
of accelerated lung maturity in PIH infants. On the other hand,
infants born to mothers with severe PIH have increased odds of
invasive ventilation and/or surfactant therapy (unadjusted data),
which is likely due to lack of beneficial effect of antenatal steroids
and increased oxidative stress in severe PIH.43 However, studies
primarily evaluating short-term respiratory outcomes were few
and the comparison was unadjusted for all short-term respiratory
outcomes. These factors limit us to draw any definite conclusions
on association of short-term respiratory outcomes with PIH.
The following are important strengths of this review: first, it
comprehensively summarizes the association of maternal PIH with
a broad range of neonatal outcomes; second, it provides the
differences in association of outcomes for the extremely preterm
infants (o29 weeks’ gestation), as well as by severity of PIH and
preeclampsia. It is particularly important to know that there are
differences between gestational hypertension and preeclampsia
reflecting different pathophysiology, which is why this subgroup
was considered.45 Similarly, there are differences in outcomes
depending on the onset of preeclampsia (early versus late);
however, this subgroup was not considered as none of these
studies would have qualified for.46 Finally, only studies with a low
or moderate risk of bias, and studies evaluating outcomes based
on PIH as the primary exposure variable were included, helping to
generalize the pooled outcome data with a substantial degree of
confidence. However, a number of limitations must be acknowl-
edged. First, despite efforts to limit the clinical heterogeneity by
appropriate study selection, the statistical heterogeneity was often
high. The significant heterogeneity may exist due to differences
across gestation, difference in outcome definition for mortality
(neonatal mortality or mortality before discharge), ethnic factors,
diverse neonatal and obstetric practices across centers and
multifactorial disease natures of certain outcomes (BPD and
ROP). These factors are particularly germane to the overall analysis
of all preterm infants with exposure to any and all PIH, whereby
differences in GAs included along with varying severity of PIH
among included studies may limit generalizability to individual
scenarios. Acknowledging this limitation of high heterogeneity,

Journal of Perinatology (2017), 1 – 8 © 2017 Nature America, Inc., part of Springer Nature.

we used the random effects model for all of our meta-analyses.
Furthermore, we performed subgroup analyses of infants
o29 weeks’ GA, as well as exposure severe PIH and pre-
eclampsia to delineate the impact of these factors on outcomes.
Second, it is important to note that there is a degree of overlap
among included studies in the groups. This may have influenced
some outcomes in the all PIH group where an association could be
obscured, due to the resulting aforementioned heterogeneity of
studies in that group. Another limitation was that studies were
also included where the distinction between PIH and chronic
hypertension was not provided. As such, some studies may have
included patients with both PIH and chronic hypertension.
However, given the relatively low prevalence of chronic hyperten-
sion in women of child-bearing age, it was felt this may not
confound the data to an appreciable degree. Finally, we
acknowledge that there are some inconsistencies among some
of the reported outcomes (for example, mortality) when compar-
ing the adjusted and unadjusted pooled data. However, we felt it
important to provide pooled estimates for all available data to
provide a detailed summary of the current literature, but advise
caution when interpreting results of unadjusted analyses.
In conclusion, our comprehensive review and quantitative
analysis demonstrates that maternal PIH is associated with a
reduced odds of mortality, ROP and SBI among all preterm infants,
increased odds of BPD in preterm infants o 29 weeks GA and no
association with NEC. In the subgroup of severe PIH, the odds of
mortality were increased, although only unadjusted were avail-
able. There is very little, albeit conflicting, data on the impact of
PIH on short term respiratory outcomes. Importantly, this review
highlights the various limitations and knowledge gaps in the
current body of literature for both clinicians and researchers.
CONFLICT OF INTEREST
The authors declare no conflict of interest.
AUTHOR CONTRIBUTIONS
Abdul Razak co-conceptualized and designed the study, performed the initial
screening of the articles, performed initial analyses including risk of bias
assessment, drafted the manuscript and approved the final version. Anna
Florendo-Chin double checked the initial screening of articles and double
checked all the data extraction from included studies. Laura Banfield devised
the search strategy for the systematic review based on study protocol and
approved the final manuscript version. Muzafar Gani Abdul Wahab assisted in
the conduct of the systematic review, provided logistical support, reviewed the
study protocol and helped revise the manuscript and approved the final
version. Sarah McDonald and Prakesh S Shah provided methodological and
statistical expertise in data analyses and interpretation, helped revise the
manuscript and approved the final version. Amit Mukerji co-conceptualized and
designed the study, provided guidance in the conduct of the review and meta-
analysis, oversaw the conduct of all meta-analyses, revised the initial
manuscript draft and approved the final version.
REFERENCES
1 Manuck TA, Rice MM, Bailit JL, Grobman WA, Reddy UM, Wapner RJ et al. Preterm
neonatal morbidity and mortality by gestational age: a contemporary cohort. Am
J Obstet Gynecol 2016; 215(1): 103.e1–103.e14.
2 Villar J, Say L, Shennan A, Lindheimer M, Duley L, Conde-Agudelo A et al.
Methodological and technical issues related to the diagnosis, screening, pre-
vention, and treatment of pre-eclampsia and eclampsia. Int J Gynaecol Obstet
2004; 85: S28–S41.
3 Haig D. Genetic conflicts in human pregnancy. Q Rev Biol 1993; 495–532.
4 Aslan H, Gul A, Cebeci A. Neonatal outcome in pregnancies after preterm delivery
for HELLP syndrome. Gynecol Obstet Invest 2004; 58(2): 96–99.
5 Gemmell L, Martin L, Murphy KE, Modi N, Hakansson S, Reichman B et al.
Hypertensive disorders of pregnancy and outcomes of preterm infants of 24 to
28 weeks’ gestation. J Perinatol 2016; 36(12): 1067–1072.
© 2017 Nature America, Inc., part of Springer Nature.
Neonatal outcomes in PIH: a meta-analysis
A Razak et al
7
6 Hiett AK, Brown HL, Britton KA. Outcome of infants delivered between 24 and
28 weeks' gestation in women with severe pre-eclampsia. J Matern Fetal Med
2001; 10(5): 301–304.
7 Yen TA, Yang HI, Hsieh WS, Chou HC, Chen CY, Tsou KI et al. Preeclampsia and the
risk of bronchopulmonary dysplasia in VLBW infants: a population based study.
PLoS ONE 2013; 8(9): e75168.
8 Kim HY, Sohn YS, Lim JH, Kim EH, Kwon JY, Park YW et al. Neonatal outcome after
preterm delivery in HELLP syndrome. Yonsei Med J 2006; 47(3): 393–398.
9 Cetinkaya M, Ozkan H, Koksal N. Maternal preeclampsia is associated with
increased risk of necrotizing enterocolitis in preterm infants. Early Hum Dev 2012;
88(11): 893–898.
10 Chan PY, Tang SM, Au SC, Rong SS, Lau HH, Ko ST et al. Association of gestational
hypertensive disorders with retinopathy of prematurity: a systematic review and
meta-analysis. Sci Rep 2016; 6: 30732.
11 Welch V, Petticrew M, Petkovic J, Moher D, Waters E, White H et al. Extending the
PRISMA statement to equity-focused systematic reviews (PRISMA-E 2012):
explanation and elaboration. Int J Equity Health 2015; 14: 92.
12 American College of Obstetricians and Gynecologists, Task Force on Hypertension
in Pregnancy. Hypertension in pregnancy. Report of the American college of
obstetricians and gynecologists’ task force on hypertension in pregnancy. Obstet
Gynecol 2013; 122(5): 1122–1131.
13 Jobe AH, Bancalari E. Bronchopulmonary dysplasia. Am J Respir Crit Care Med
2001; 163(7): 1723–1729.
14 Papile L-A, Burstein J, Burstein R, Koffler H. Incidence and evolution of sub-
ependymal and intraventricular hemorrhage: a study of infants with birth weights
less than 1,500 gm. J Pediatr 1978; 92(4): 529–534.
15 de Vries LS, Eken P, Dubowitz LM. The spectrum of leukomalacia using cranial
ultrasound. Behav Brain Res 1992; 49(1): 1–6.
16 Gole GA, Ells AL, Katz X, Holmstrom G, Fielder AR, Capone A Jr et al. The inter-
national classification of retinopathy of prematurity revisited. Arch Ophthalmol
2005; 123(7): 991–999.
17 Walsh MC, Kliegman R. Necrotizing enterocolitis: treatment based on staging
criteria. Pediatr Clin North Am 1986; 33(1): 179–201.
18 The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised
studies in meta-analyses. http://www.ohri.ca/programs/clinical_epidemiology/
oxford.asp. Access Date: October 2016.
19 Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the
median, range, and the size of a sample. BMC Med Res Methodol 2005; 5: 13.
20 Cetinkaya M, Ozkan H, Köksal N, Karali Z, Ozgür T. Neonatal outcomes of pre-
mature infants born to preeclamptic mothers. J Matern Fetal Neonatal Med 2010;
23(5): 425–430.
21 Chen XK, Wen SW, Smith G, Yang Q, Walker M. Pregnancy-induced hypertension
is associated with lower infant mortality in preterm singletons. BJOG 2006; 113(5):
544–551.
22 Huang HC, Yang HI, Chou HC, Chen CY, Hsieh WS, Tsou KI et al. Preeclampsia and
retinopathy of prematurity in very-low-birth-weight infants: a population-
based study. PLoS ONE 2015; 10(11): e0143248.
23 Liu A, Carlsson E, Nilsson S, Oei J, Bajuk B, Peek M et al. Hypertensive disease of
pregnancy is associated with decreased risk for respiratory distress syndrome in
moderate preterm neonates. Hypertens Pregnancy 2013; 32(2): 169–177.
24 Luo ZC, Simonet F, An N, Bao FY, Audibert F, Fraser WD. Effect on neonatal
outcomes in gestational hypertension in twin compared with singleton preg-
nancies. Obstet Gynecol 2006; 108(5): 1138–1144.
25 McBride CA, Bernstein IM, Badger GJ, Horbar JD, Soll RF. The effect of maternal
hypertension on mortality in infants 22, 29weeks gestation. Pregnancy Hypertens
2015; 5(4): 362–366.
26 O'Shea JE, Davis PG, Doyle LWVictorian Infant Collaborative Study Group.
Maternal preeclampsia and risk of bronchopulmonary dysplasia in preterm
infants. Pediatr Res 2012; 71(2): 210–214.
27 Ozkan H, Cetinkaya M, Koksal N, Ozmen A, Yildiz M. Maternal preeclampsia is
associated with an increased risk of retinopathy of prematurity. J Perinat Med
2011; 39(5): 523–527.
28 Silveira RC, Procianoy RS, Koch MS, Benjamin ACW, Schlindwein CF. Growth and
neurodevelopment outcome of very low birth weight infants delivered by pre-
eclamptic mothers. Acta Paediatr 2007; 96(12): 1738–1742.
29 Spinillo A, Gardella B, Preti E, Zanchi S, Tzialla C, Stronati M. Preeclampsia and
brain damage among preterm infants: a changed panorama in a 20-year analysis.
Am J Perinatol 2007; 24(2): 101–106.
30 Tastekin A, Ors R, Demircan B, Saricam Z, Ingec M, Akcay F. Oxidative stress in
infants born to preeclamptic mothers. Pediatr Int 2005; 47(6): 658–662.
31 Tokumasu H, Tokumasu S, Kawakami K. Impact of pre-eclampsia in extremely
premature infants: Population-based study. Pediatr Int 2016; 58(7): 578–583.
32 Wang A, Holston AM, Yu KF, Zhang J, Toporsian M, Karumanchi SA et al. Circu-
lating anti-angiogenic factors during hypertensive pregnancy and increased risk
Journal of Perinatology (2017), 1 – 8
 Neonatal outcomes in PIH: a meta-analysis
A Razak et al
8
of respiratory distress syndrome in preterm neonates. J Matern Fetal Neonatal
Med 2012; 25(8): 1447–1452.
33 Zayed MA, Uppal A, Hartnett ME. New-onset maternal gestational hypertension and
risk of retinopathy of prematurity. Invest Ophthalmol Vis Sci 2010; 51(10): 4983–4988.
34 Hansen AR, Barnés CM, Folkman J, McElrath TF. Maternal preeclampsia predicts the
development of bronchopulmonary dysplasia. J Pediatr 2010; 156(4): 532–536.
35 Cheng S-W, Chou H-C, Tsou K-I, Fang L-J, Tsao P-N. Delivery before 32 weeks of
gestation for maternal pre-eclampsia: Neonatal outcome and 2-year develop-
mental outcome. Early Hum Dev 2004; 76(1): 39–46.
36 Kalay S, Cakcak B, Oztekin O, Tezel G, Tosun O, Akcakus M et al. The role of VEGF
and its soluble receptor VEGFR-1 in preterm newborns of preeclamptic mothers
with RDS. J Matern Fetal Neonatal Med 2013; 26(10): 978–983.
37 Ozkan H, Cetinkaya M, Koksal N. Increased incidence of bronchopulmonary dys-
plasia in preterm infants exposed to preeclampsia. J Matern Fetal Neonatal Med
2012; 25(12): 2681–2685.
38 Park YH, Lee GM, Yoon JM, Cheon EJ, Ko KO, Lee YH et al. Effect of early postnatal
neutropenia in very low birth weight infants born to mothers with pregnancy-
induced hypertension. Korean J Pediatr 2012; 55(12): 462–469.
39 Procianoy RS, Silveira RC, Mussi-Pinhata MM, Souza Rugolo LM, Leone CR, de
Andrade Lopes JM et al. Sepsis and neutropenia in very low birth weight infants
delivered of mothers with preeclampsia. J Pediatr 2010; 157(3): 434–438 438.e431.
Supplementary Information accompanies the paper on the Journal of Perinatology website (http://www.nature.com/jp)
Journal of Perinatology (2017), 1 – 8
40 Masoura S, Kalogiannidis I, Margioula-Siarkou C, Diamanti E, Papouli M, Drossou-
Agakidou V et al. Neonatal outcomes of late preterm deliveries with pre--
eclampsia. Minerva Ginecol 2012; 64(2): 109–115.
41 McElrath TF, Fichorova RN, Allred EN, Hecht JL, Ismail MA, Yuan H et al. Blood
protein profiles of infants born before 28 weeks differ by pregnancy complication.
Am J Obstet Gynecol. 2011; 204(5): 418.e1–418.e12.
42 Joseph KS. Theory of obstetrics: an epidemiologic framework for justifying
medically indicated early delivery. BMC Pregnancy Childbirth 2007; 7: 4.
43 Diedrich F, Renner A, Rath W, Kuhn W, Wieland E. Lipid hydroperoxides and free
radical scavenging enzyme activities in preeclampsia and HELLP (hemolysis,
elevated liver enzymes, and low platelet count) syndrome: no evidence for cir-
culating primary products of lipid peroxidation. Am J Obstet Gynecol 2001; 185(1):
166–172.
44 Bain E, Middleton P, Crowther CA. Different magnesium sulphate regimens for
neuroprotection of the fetus for women at risk of preterm birth. Cochrane
Database Syst Rev 2012; 15(2): CD009302.
45 Shen M, Smith GN, Rodger M, White RR, Walker MC, Wen SW. Comparison of risk
factors and outcomes of gestational hypertension and pre-eclampsia. PLoS ONE
2017; 12(4): e0175914.
46 Raymond D, Peterson E. A critical review of early-onset and late-onset pre-
eclampsia. Obstet Gynecol Surv 2011; 66(8): 497–506.
© 2017 Nature America, Inc., part of Springer Nature.