See

discussions, stats, and author profiles for this publication at:

https://www.researchgate.net/publication/51072024

Impact of fetal sex in pregnancy-induced

hypertension and preeclampsia in Japan
Article in Journal of Reproductive Immunology May 2011
Impact Factor: 2.82 DOI: 10.1016/j.jri.2010.12.011 Source: PubMed

CITATIONS

READS

13

110

4 authors, including:
Arihiro Shiozaki

Yoshio Matsuda

University of Toyama

Tokyo Women's Medical University

55 PUBLICATIONS 1,451 CITATIONS

67 PUBLICATIONS 539 CITATIONS

SEE PROFILE

SEE PROFILE

Shoji Satoh
Oita prefectural hospital, japan
53 PUBLICATIONS 655 CITATIONS
SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate,

letting you access and read them immediately.

Available from: Arihiro Shiozaki

Retrieved on: 07 May 2016

Journal of Reproductive Immunology 89 (2011) 133139

Contents lists available at ScienceDirect

Journal of Reproductive Immunology

journal homepage: www.elsevier.com/locate/jreprimm

Impact of fetal sex in pregnancy-induced hypertension and

preeclampsia in Japan
Arihiro Shiozaki a , Yoshio Matsuda b , Shoji Satoh c , Shigeru Saito a,
a
b
c

Department of Obstetrics and Gynecology, University of Toyama, 2630 Sugitani, Toyama 930-0194, Japan
Department of Obstetrics and Gynecology, Tokyo Womens Medical University, 8-1 Kawada-cho, Shinjyuku-ku, Tokyo 162-8666, Japan
Maternal and Perinatal Care Center, Oita Prefectural Hospital, Bunyo 476, Oita 870-8511, Japan

a r t i c l e

i n f o

Article history:
Received 19 October 2010
Received in revised form
17 December 2010
Accepted 23 December 2010

Keywords:
Fetal gender
HY antigen
Preeclampsia
MD twins
DD twins
Twin pregnancy

a b s t r a c t
The male antigen (HY), the elevated level of fetal antigen in twin pregnancies, and the
increased number of MHC mismatches in dizygotic twin pregnancies might affect immunological tolerance during pregnancy. Using the Perinatal Database of the Japanese Society
for Obstetrics and Gynecology, we studied the occurrence of pregnancy-induced hypertension (PIH) and preeclampsia in mothers delivering singleton babies and in those delivering
monochorionic diamniotic (MD) twin pregnancies and dichorionic diamniotic (DD) twin
pregnancies at 125 centers of the perinatal network in Japan from 2001 through 2005.
In singleton pregnancies, pregnant women carrying female fetuses had a signicantly
higher incidence of PIH and preeclampsia compared with those carrying male fetuses. In
MD twin pregnancies, compared with mothers carrying malemale fetuses, those carrying
femalefemale fetuses had signicantly higher incidences of PIH and preeclampsia and a
marked difference was observed in primiparous cases. In DD twin pregnancies, the incidences of PIH and preeclampsia were signicantly higher in mothers with femalefemale
fetuses than those with malemale fetuses, while those with malefemale fetuses had
intermediate values. The incidence of PIH and preeclampsia in MD twin pregnancies was
similar to that in DD twin pregnancies with malemale fetuses or femalefemale fetuses.
The male antigen and the increased number of MHC mismatches in DD twin pregnancies
were not a risk factor for PIH and preeclampsia. Female fetal sex was a risk factor for PIH
and preeclampsia.
2011 Elsevier Ireland Ltd. All rights reserved.

1. Introduction
Preeclampsia (preeclampsia) is a pregnancy complication affecting pregnant women and one of the major
causes of maternal mortality and morbidity, perinatal
death, preterm birth, and fetal growth restriction. Several risk factors for preeclampsia have been identied
such as nulliparity, prolonged interpregnancy interval,
short cohabitation, condom user, and use of donated

Corresponding author. Tel.: +81 76 434 7355; fax: +81 76 434 5036.
E-mail address: s30saito@med.u-toyama.ac.jp (S. Saito).
0165-0378/$ see front matter 2011 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.jri.2010.12.011

embryos (Robillard et al., 2003; Dekker, 2002; Saito et al.,

2007).
The embryo (fetus) and placenta are a semi-allograft to
the maternal immune system because half of the embryonal (fetal) genes are paternally derived. In general, the
risk of preeclampsia is greatest in primiparous women.
Pathogenesis of preeclampsia in primiparous women may
differ from that in multiparous women, multifetal gestation, or previous preeclampsia. Subsequent pregnancy with
the same partner reduces the risk of preeclampsia (Trupin
et al., 1996). Moreover, the risk of preeclampsia seems to
be partner-dependent. Subsequent pregnancy with a new
partner increases the risk of preeclampsia (Robillard et al.,

134

A. Shiozaki et al. / Journal of Reproductive Immunology 89 (2011) 133139

2003; Dekker, 2002). Conversely, prolonged exposure to

a partners semen may reduce the risk of preeclampsia
(Einarsson et al., 2003; Robillard et al., 2003). These phenomena suggest that immunological mechanisms such as
induction of tolerance may contribute to the pathogenesis
of preeclampsia.
Several studies have identied the association of fetal
gender with PIH and preeclampsia, while their results
are contradictory. In singleton pregnancies, Toivanen and
Hirvonen (1970) reported that the ratio of males to
females in babies born to mothers with PIH was 1.24
and the ratio increased up to 1.72 according to the
severity of the disease. On the other hand, Hsu et al.
(1994) found a predominance of female fetuses in preterm
preeclamptic pregnancies compared with preterm normotensive pregnancies (p = 0.043), but not those in term
preeclamptic and normotensive pregnancies (p = 0.989).
In multifetal pregnancy, there is no difference in the
male/female ratio between normotensive mothers and
preeclamptic mothers (Makhseed et al., 1998). However, the sample size in this twin study was very small
(<70 cases). Caution is needed when evaluating this
result.
There are three hypothesized types of pathogenesis for
the risk of preeclampsia in malemale and femalefemale
twins (Tables 1A and 1B). The rst hypothesis speculates that immune-incompatibility between mother and
fetus (major histocompatibility complex [MHC] mismatch) contributes to the pathogenesis of preeclampsia
(Stevenson et al., 1971). If this were the case, the incidence
of preeclampsia should be higher in dichorionic diamniotic (DD) twins compared with monochorionic diamniotic
(MD) twins and should be similar in MD twins and in
singletons, because all the MD twins are derived from
monozygotic twins and 8090% of the DD twins are derived
from dizygotic twins in Japan.
The second hypothesis suggests that increased levels
of fetal antigen lead to the pathogenesis of preeclampsia. If this were the case, the incidence of preeclampsia
should be twice as high in twins compared with singletons and should be similar in DD twins and MD
twins.
The third hypothesis is related to the HY antigen. Recent
data suggested that the chance of a subsequent live birth
in secondary recurrent miscarriage patients with rst-born
boys compared with rst-born girls was signicantly lower
in women with HY-restricting HLA class II alleles (Nielsen
et al., 2009). Most patients with recurrent placental abruptions had rst-born boys and signicantly more of these
patients carried HLA haplotypes with HY-restricting class II
alleles compared with controls (Christiansen et al., 2010). A
maternal immune reaction against fetal HY antigens might
break the maternal tolerance to semiallograftic fetuses. If
this were the case, the malemale twins should have the
highest rate of preeclampsia, the femalefemale twins and
female singletons should have the lowest rate, and the
malefemale twins and male singletons should have an
intermediate rate.
The aim of the present study was to evaluate the effects
of fetal sex on the pathogenesis of PIH/preeclampsia. To
demonstrate that any of the hypotheses above contribute

to preeclampsia, we examined the incidence of preeclampsia in twin pregnancies as well as singleton pregnancies and
analyzed the relationship among fetal gender, fetal number, and PIH/preeclampsia.
2. Materials and methods
This study was approved by the Tokyo Womens Medical
University Ethics Committee.
Detailed descriptions of the database have been published elsewhere (Matsuda et al., in press). Briey, the
attendant physicians at 125 tertiary perinatal centers of
Perinatal Research Network in Japan collected yearly data
for women in an off-line clinical database with a common
format. Data were sent to the Perinatal Committee of the
Japanese Society of Obstetrics and Gynecology, and quality
control was assessed for the database.
There were 241,672 singleton births and 20,050 twin
births (10,025 mothers) that resulted in live birth or fetal
death. Fetal death was dened as follows: fetal death before
complete expulsion or extraction from the mother of a
product of conception with a gestation of at least 22 weeks.
All measurements reported in the database were obtained
as the usual care provided to high-risk obstetric patients
at tertiary perinatal centers. Determination of chorionicity
was performed non-invasively during the rst trimester of
pregnancy by ultrasound examination of the base of the
inter-twin membrane for the presence or absence of the
lambda sign (Sepulveda et al., 1996).
Gestational age was determined based on the menstrual
history, prenatal examination and ultrasound ndings during early pregnancy (gestational sac diameter, crown rump
length, and biparietal diameter).
Women were classied as having pregnancy-induced
hypertension when they had hypertension (systolic blood
pressure 140 mmHg or diastolic 90 mmHg) on two
occasions. Women were considered to have preeclampsia when they had hypertension (systolic blood pressure 140 mmHg or diastolic 90 mmHg) on two occasions and proteinuria dened as either 300 mg/24 h
urine collection or 1+ on a dipstick on at least
two separate occasions without urinary tract infection.
Women were stratied to have severe preeclampsia
when they had hypertension (systolic blood pressure 160 mmHg and 100 mmHg) on two and proteinuria
dened as either 2 g/24 h urine collection or 3+ on
a dipstick on at least two separate occasions without
urinary tract infection. Women with chronic hypertension were excluded. Variables considered to be of
potential importance in the analysis included maternal
age, parity, gestational age at delivery, maternal smoking.
We compared numbers and rates of PIH, preeclampsia, preeclampsia with fetal death, severe preeclampsia,
and severe preeclampsia with fetal death among mothers (mothers carrying male fetuses; mothers carrying
female fetuses; mothers carrying malemale MD
fetuses; mothers carrying femalefemale MD fetuses;
mothers carrying malemale DD fetuses; mothers
carrying malefemale DD fetuses; mothers carrying
femalefemale DD fetuses) and compared background

A. Shiozaki et al. / Journal of Reproductive Immunology 89 (2011) 133139

135

Table 1A
Hypothesis (1): estimated risks of preeclampsia.
Immunological pathogeneses of preeclampsia

Estimated risks of preeclampsia

I. Increased number of mismatch MHC

Malemale DD twins = malefemale DD twins = femalefemale DD twins

>Malemale MD twins = femalefemale MD twins
=Male singleton = female singleton
Malemale DD twins = malefemale DD twins = femalefemale DD twins
=Malemale MD twins = femalefemale MD twins
>Male singleton = female singleton
Malemale DD twins = malemale MD twins
>Malefemale DD twins = male singleton
>Femalefemale DD twins = femalefemale MD twins = female singleton

II. Elevated level of fetal antigens

III. Elevated level of HY antigens

MHC, major histocompatibility; HY, male-specic minor histocompatibility; DD, dichorionic diamniotic; MD, monochorionic diamniotic.

factors (maternal age, gestational age, and parity) among

them.
2.1. Statistical analysis
The data were analyzed statistically by the Chi-squared
test using a statistical software package (SAS version 9.1;
SAS Institute, Cary, NC, USA). The criterion for statistical
signicance was a level of 0.05.
3. Results
3.1. Effect of fetal gender on maternal background
Maternal background, such as maternal age, especially
40 or older, essential hypertension, diabetes mellitus, body
mass index (BMI), gestational age, were analyzed for a possible correlation with fetal gender. However, no correlation
between any of the clinical parameters and fetal gender
was observed (Table 2).
3.2. Effect of fetal gender on PIH/preeclampsia
3.2.1. Singleton: male fetus vs. female fetus
In primipara the frequency of PIH mothers with a female
fetus was signicantly higher than those with a male fetus

(5.2% vs. 4.6%, p < 0.0001; Table 3). Of 6199 babies born to
PIH mothers, 3009 were male fetuses and 3109 were female
fetuses, i.e. 3.3% more female fetuses. And the incidence of
primiparous and multiparous preeclampsia mothers with a
female fetus was also signicantly higher than those with a
male fetus (3.8% vs. 3.2%, p < 0.0001). Of 4439 babies delivered from preeclampsia mothers, 2333 were female fetuses
and 2106 were male fetuses, i.e. 10.8% more female fetuses.
When preeclamptic pregnancies were further stratied
into mothers with severe preeclampsia, we showed that
the rate of female fetuses was signicantly higher than that
of male fetuses (1.5% vs. 1.3%, p < 0.0001). Surprisingly, of
1753 babies from mothers with severe preeclampsia mothers, 929 were female fetuses and 824 were male fetuses, i.e.
12.7% more female fetuses.

3.2.2. MD twins: malemale vs. femalefemale

From Table 3 it will be seen that the frequency of
PIH and preeclampsia in primiparous mothers carrying
femalefemale twins was signicantly higher than those
carrying malemale twins (9.1% vs. 5.6%, p = 0.006; 7.6% vs.
3.6%, p = 0.0005), while there was no differences in multiparous mothers.

Table 1B
Hypothesis (2): impact of fetal sex on risk of preeclampsia.

I. MD twins (monozygote) vs. singleton

Number of fetal MHC mismatches
Amount of fetal MHC antigen
Amount of fetal HY antigen

Malemale MD twins

Femalefemale MD twins

Twice
Same
Twice

Twice
Same
Not evaluated

Malemale DD twins
II. DD twin (dizygote [90%] or monozygote [10%]) vs. singleton
Number of fetal MHC mismatches
Twice
Amount of fetal MHC antigen
Twice (90%)Same (10%)
Amount of fetal HY antigen
Twice

III. DD twins vs. MD twins

Number of fetal MHC mismatches
Amount of fetal MHC antigen
Amount of fetal HY antigen

Malefemale DD twins

Femalefemale DD twins

Twice
Twice
Same

Twice
Twice (90%) Same (10%)
Not evaluated

Malemale

Femalefemale

DD > MD
DD = MD
DD = MD

DD > MD
DD = MD
Not evaluated

MD, monochorionic diamniotic; DD, dichorionic diamniotic; MHC, major histocompatibility; HY, male-specic minor histocompatibility; twice means
twice as much as singletons, same means same as singletons.

136

Table 2
Fetal gender and maternal background.
Singleton
Fetal gender

MD Twins

DD Twins

Female

Malemale

Femalefemale

Malemale

Malefemale

Femalefemale

29.4 5.1

29.4 5.1

0.3133

28.7 4.6

29.1 4.7

0.0672

31.2 4.7

32.0 4.7

31.4 4.7

2.6% (1675/65,351)

2.7% (1633/61,136)

0.229

1.4% (12/828)

1.8% (16/891)

0.7067

3.1% (43/1376)

4.6% (87/1883)

4.0% (50/1263)

38.1 2.9

38.4 2.8

<0.001

34.1 3.7

34.3 3.7

0.2074

35.2 3.0

35.2 3.2

35.3 3.1

24.9 3.5

24.8 3.5

0.5917

24.7 2.5

25.1 3.6

0.9657

25.2 3.2

25.3 3.7

25.3 3.4

1.8% (1170/65,268)

1.7% (1066/61,045)

0.1315

0.8% (7/833)

1.1% (10/899)

0.7415

0.9% (13/1383)

1.4% (26/1909)

0.9% (16/1296)

Renal disease

1.1% (734/65,280)

1.2% (707/61,071)

0.5777

0.1% (1/834)

0.9% (8/899)

0.0583

0.8% (11/1388)

0.5% (10/1912)

0.7% (9/1276)

Smoking

5.3% (2574/48,407)

5.3% (2432/45,501)

0.8626

4.0% (24/607)

3.7% (25/678)

0.917806

3.1% (31/996)

2.4% (34/1.393)

2.4% (22/924)

<0.005*1 , <0.001*2 ,
0.408*3
0.292*1 , 0.039*2 ,
0.423*3
0.6541*1 , 0.9875*2 ,
0.7060*3
0.9997*1 , 0.9999*2 ,
0.9999*3
0.5440*1 , 0.3465*2 ,
0.9315*3
0.9715*1 , 0.4597*2 ,
0.6742*3
0.4020*1 , 0.3857*2 ,
0.9631*3

31.8 4.6

31.8 4.6

0.6064

31.2 4.4

30.8 4.6

0.1068

32.2 4.3

32.5 4.1

31.9 4.3

4.2% (2942/58,847)

4.3% (2413/56,200)

0.6212

2.6% (18/696)

2.6% (18/694)

0.993

3.7% (26/707)

3.6% (30/840)

2.9% (19/647)

37.7 2.9

37.9 2.9

< 0.001

34.4 3.7

34.6 3.4

0.1558

35.0 3.4

35.6 2.8

35.3 3.1

25.1 3.5

25.1 3.7

0.9561

25.3 3.5

25.7 3.1

0.9679

25.4 3.3

25.5 3.9

25.4 3.1

1.9% (1134/58,769)

1.9% (1043/56,158)

0.3686

0.8% (6/710)

1.8% (13/703)

0.1592

2.4% (17/704)

2.0% (17/840)

1.1% (7/646)

Renal Disease

0.9% (504/58,782)

0.9% (491/56,128)

0.7506

0.0% (0/709)

1.0% (7/704)

0.0224

0.8% (6/707)

0.8% (7/839)

0.6% (4/651)

Smoking

6.4% (2829/43,965)

6.4% 2674/41,999)

0.6949

4.0% (21/520)

5.3% (27/507)

0.4071

6.0% (31/517)

6.4% (40/624)

5.5% (26/474)

30.5 5.0

30.5 5.0

0.6338

29.8 4.7

29.8 4.7

0.9368

31.5 4.6

32.1 4.5

31.6 4.6

3.4% (4167/124,198)

3.4% (4046/117,336)

0.2071

2.0% (30/1524)

2.1% (34/1585)

0.8256

3.3% (69/2079)

4.3% (117/2715)

3.6% (69/1910)

38.0 2.9

38.2 2.9

<0.001

34.2 3.7

34.5 3.6

0.0669

35.1 3.1

35.3 3.1

35.3 3.1

25.0 3.5

25.0 3.6

0.9981

25.0 3.0

25.4 3.4

0.8317

25.2 3.2

25.3 3.7

25.3 3.3

1.9% (2304/124,037)

1.8% (2109/117,203)

0.2875

0.8% (13/1543)

1.4% (23/1602)

0.1628

1.4% (30/2087)

1.6% (43/2749)

1.2% (23/1915)

Renal Disease

1.0% (1238/124,062)

1.0% (1198/117199)

0.5506

0.1% (1/1543)

0.9% (15/1603)

0.0015

0.8% (17/2095)

0.6% (17/2751)

0.7% (13/1927)

Smoking

5.8% (5403/92,372)

5.8% (5106/87,500)

0.9091

4.0% (45/1127)

4.4% (52/1185)

0.7113

4.1% (62/1513)

3.7% (74/2017)

3.4% (48/1398)

I. Primipara
Maternal age
(mean SD) (years)
Maternal age: 40 or
older
Gestational age
(mean SD; weeks)
BMI (mean SD)
(kg/m2 )
Diabetes mellitus

II. Multipara
Maternal age
(mean SD; years)
Maternal age: 40 or
older
Gestational age
(mean SD; weeks)
BMI (mean SD)
(kg/m2 )
Diabetes Mellitus

III. Primipara + multipara

Maternal age
(mean SD; years)
Maternal age: 40 or
older
Gestational age
(mean SD; weeks)
BMI (mean SD;
kg/m2 )
Diabetes Mellitus

*1 Malemale vs. Femalefemale; *2 Malemale vs. Malefemale; *3 Malefemale vs. Femalefemale.

MD: monochorionic diamniotic, DD: dichorionic diamniotic, BM1: body mass index.

0.2869*1 , 0.2999*2 ,
0.0073*3
0.5305*1 , 0.9680*2 ,
0.5708*3
0.3410*1 , 0.001 *2 ,
0.1229*3
1*1 , 0.9997*2 ,
0.9998*3
0.1004*1 , 0.7284*2 ,
0.2233*3
0.8519*1 , 0.8035*2 ,
0.8519*3
0.8348*1 , 0.8688*2 ,
0.3989*3
0.9636*1 , <0.001*2 ,
<0.001*3
0.6743*1 , 0.0921*2 ,
0.2663*3
0.2801*1 , 0.1046*2 ,
0.9291*3
0.9999*1 , 0.9998*2 ,
1*3
0.6064*1 , 0.8111*2 ,
0.3644*3
0.7487*1 , 0.5315*2 ,
0.9578*3
0.3999*1 , 0.5707*2 ,
0.7867*3

A. Shiozaki et al. / Journal of Reproductive Immunology 89 (2011) 133139

Male

Table 3
Fetal gender and PIH/PE.
Singleton
Fetal gender
I. Primipara
PIH

PE with fetal death

Severe PE
Severe PE with fetal
death
II. Multipara
PIH
PE
PE with fetal death
Severe PE
Severe PE with fetal
death
III. Primipara + Multipara
PIH
PE
PE with fetal death
Severe PE
Severe PE with fetal
death

DD Twin

Male

Female

MaleMale

FemaleFemale

MaleMale

MaleFemale

FemaleFemale

4.6%
(3,009/65,380)
3.2%
(2,106/65,380)
6.0%
(62/1034)
1.3%
(824/65,371)
2.8%
(29/1034)

5.2%
(3,190/61,158)
3.8%
(2,333/61,158)
6.0%
(53/897)
1.5%
(929/61,134)
3.2%
(28/879)

<0.0001

5.6%
(47/834)
3.6%
(30/834)
0.0%
(0/35)
1.2%
(10/834)
0.0%
(0/35)

9.1%
(82/899)
7.6%
(68/899)
0.0%
(0/29)
2.3%
(21/899)
0.0%
(0/29)

0.0058

7.1%
(98/1,389)
4.7%
(65/1,389)
4.2%
(1/24)
1.2%
(16/1,388)
4.2%
(1/24)

7.7%
(148/1,914)
5.4%
(104/1,914)
0.0%
(0/33)
1.6%
(31/1,913)
0.0%
(0/33)

8.8%
(112/1,276)
6.0%
(77/1,276)
12.0%
(3/25)
1.8%
(23/1,276)
4.0%
(1/25)

0.0993 *1

3.0%
(1,771/58,892)
1.9%
(1,127/58,892)
4.0%
(35/873)
0.7%
(434/58,877)
1.8%
(16/873)

3.3%
(1,875/56,242)
2.2%
(1,224/56,242)
4.1%
(35/860)
0.9%
(480/56,229)
2.3%
(20/858)

0.0017

4.8%
(34/710)
2.7%
(19/710)
0.0%
(0/33)
0.4%
(3/710)
0.0%
(0/33)

4.3%
(30/704)
2.1%
(15/704)
0.0%
(0/33)
0.6%
(4/704)
0.0%
(0/17)

0.6334

4.0%
(28/707)
2.3%
(16/707)
7.7%
(1/13)
0.8%
(6/706)
7.7%
(1/12)

3.8%
(32/840)
2.7%
(23/840)
0.0%
(0/4)
1.1%
(9/840)
0.0%
(0/4)

5.2%
(34/651)
3.5%
(23/651)
0.0%
(0/7)
0.9%
(6/651)
0.0%
(0/7)

3.8%
(4,780/124,272)
2.7%
(3,233/124,272)
5.1%
(97/1907)
1.0%
(1,258/124,248)
2.4%
(45/1907)

4.3%
(5,065/117,400)
3.0%
(3,557/117,400)
5.1%
(88/1739)
1.2%
(1,409/117,363)
2.8%
(48/1737)

< 0.0001

5.2%
(51/1,544)
3.2%
(49/1,544)
0.0%
(0/68)
0.8%
(13/1,544)
0.0%
(0/68)

7.0%
(112/1,603)
5.2%
(83/1,603)
0.0%
(0/46)
1.6%
(25/1,603)
0.0%
(0/46)

0.0419

6.0%
(126/2,096)
3.9%
(81/2,096)
5.4%
(2/37)
1.1%
(22/2,094)
5.4%
(2/37)

6.5%
(180/2,754)
4.6%
(127/2,754)
0.0%
(0/37)
1.5%
(40/2,753)
0.0%
(0/37)

7.6%
(146/1,927)
5.2%
(100/1,927)
9.4%
(3/32)
1.5%
(29/1,927)
3.1%
(1/32)

<0.0001
0.9475
< 0.0001
0.7239

0.0018
0.9537
0.0283
0.577

< 0.0001
0.9684
< 0.0001
0.5051

0.0005
0.11
-

0.6201
0.9912
-

0.0066
0.0931
-

0.1198 *1
0.6317 *1
0.2174 *1
0.4885 *1

0.2656 *1
0.1198 *1
0.7470 *1
0.2174 *1
0.7470 *1

0.0483 *1
0.0428 *1
0.8660 *1

A. Shiozaki et al. / Journal of Reproductive Immunology 89 (2011) 133139

PE

MD Twin

0.2522 *1
0.8976 *1

*1 MaleMale vs. FemaleFemale.

PIH: pregnancy-induced hypertension; PE: pre-eclampsia; MD: monochorionic diamniotic; DD: dichorionic diamniotic.

137

138

A. Shiozaki et al. / Journal of Reproductive Immunology 89 (2011) 133139

3.2.3. DD twins: malemale vs. malefemale vs.

femalefemale
Also from Table 3 it will be seen that there was no signicant difference in the occurrence of PIH/preeclampsia
among primiparous mothers with malemale, those with
malefemale, and those with femalefemale twins. There
was also no signicant difference among multiparous
mothers. In total (primiparous and multiparous) mothers, however, there was a signicant difference in the
occurrence of PIH and preeclampsia between mothers with
femalefemale and those with malemale twins (7.6% vs.
6.0%, p = 0.0483 and 5.2% vs. 3.9%, p = 0.043, respectively),
while mothers with malefemale twins have intermediate
values (6.5% vs. 4.6%).
3.2.4. MD twins vs. DD twins
There was no signicant difference in the occurrence
of PIH/preeclampsia between primiparous mothers with
MD malemale twins and those with DD malemale twins
(5.6% vs. 7.1%, p = 0.189 and 3.6% vs. 4.7%, p = 0.266 respectively) (Table 3). There was also no signicant difference in
the occurrence of PIH/preeclampsia between primiparous
mothers with MD femalefemale twins and those with DD
femalefemale twins (9.1% vs. 8.8%, p = 0.782 and 7.6% vs.
6.0%, p = 0.159 respectively; Table 3). No differences were
observed in the occurrence of PIH/preeclampsia between
multiparous mothers carrying MD malemale twins and
those carrying DD malemale twins (4.8% vs. 4.0%, p = 0.446
and 2.7% vs. 2.3%, p = 0.742 respectively; Table 3), or
between total (primiparous + multiparous) mothers with
MD malemale twins and those with DD malemale twins
(5.2% vs. 6.0%, p = 0.324 and 3.2% vs. 3.9%, p = 0.267 respectively; Table 3).
3.3. Effect of fetal gender on preeclampsia/severe
preeclampsia with fetal death
Irrespective of parity and fetal number, no signicant
differences in the frequencies of male and female fetuses
were found among mothers with preeclampsia and fetal
death, and among those with severe preeclampsia and fetal
death (Table 3).
4. Discussion
The goal of this study was to ascertain whether fetal
sex affects the frequency of PIH/preeclampsia or whether
DD twin and MD twin pregnancies affect the incidence of
PIH/preeclampsia. Of 9845 singleton babies born to PIH
mothers, 4780 were boys and 5065 were girls (6.0% more
girls) and of 6790 singleton babies born to preeclampsia
mothers, 3233 were boys and 3557 were girls (10.0% more
girls). This female preponderance in primiparous mothers with PIH/preeclampsia was also seen in multiparous
mothers. In MD twin pregnancy, compared with mothers
carrying malemale fetuses, those carrying femalefemale
fetuses had signicantly higher incidences of PIH and
preeclampsia. In DD twin pregnancies, the incidences of
PIH and preeclampsia were signicantly higher in mothers
with femalefemale fetuses than in those with malemale
fetuses, while those with malefemale fetuses had inter-

mediate values. This was precisely the opposite of the

report, which showed that male gender is associated with
an increased risk of preeclampsia in singleton pregnancies
(Toivanen and Hirvonen, 1970). But caution is necessary
because the sample size of this study is very small. Moreover, two Danish Birth Registries (Basso and Olsen, 2001;
Nielsen et al., 2010) also showed that male gender is a
risk factor for preeclampsia. In a Danish study carried out
between 1980 and 1994, the male-to-female sex ratio in
birth among mothers with preeclampsia was 1.10 (95%
CI = 1.071.12; Basso and Olsen, 2001), while the risk for
of was equal in malemale twins and in femalefemale
twins, although they did not classify the twin pregnancies between MD twins and DD twins. In another Danish
Birth Registry (19822005), the male-to-female sex ratio
was 1.05 and the overall frequencies of preeclampsia in
births among nulliparous mothers with male fetuses and
female fetuses were 4.3% and 4.1% respectively (p = 0.0001)
and those among multiparous mothers were 1.9% and 1.8%
respectively (p = 0.44; Nielsen et al., 2010). These previous
ndings were the opposite result of our study.
The reason why the results in Denmark are different from those in Japan is not clear. In our study, the
male/female ratio in singletons was 1.06, which was similar
to that in Australia, 1.06 (19911998), and that in all cases
in Japan, 1.05 (2008). This male-to-female ratio was rather
low compared with that in Denmark (1.10). Furthermore,
as shown in Table 3, there were no signicant differences in
the frequencies of male and female fetuses among mothers
with preeclampsia and fetal death and among those with
severe preeclampsia and fetal death, irrespective of parity and fetal number. Therefore, the possibility that more
male babies had already been rejected in miscarriage during early pregnancy and in stillbirth during late pregnancy
is not taken into account in our study. Also, the higher proportion of male fetuses does not explain the preponderance
of female fetuses in mothers with preeclampsia. It is still
unclear why mothers with preeclampsia have slightly more
female babies in Japan and why mothers with preeclampsia have slightly more male babies in Denmark. This result
could be due to differences in racial and ethnic backgrounds
or racial differences in immunological response to HY antigens.
To test the hypotheses discussed above as a possible
cause of preeclampsia, we categorized PIH/preeclampsia
mothers into seven groups (male singletons, female
singletons, malemale MD twins, femalefemale MD
twins, malemale DD twins, malefemale DD twins, and
femalefemale DD twins) according to fetal gender and
zygosity, and compared the rate of PIH/preeclampsia
among those groups.
As a rst hypothesis, Stevenson et al. (1971) speculated
that immune-incompatibility between mother and fetus
(mismatching of MHC) contributes to the pathogenesis of
preeclampsia. In HLA-C mismatch cases, decidual CD8+ T
cells are activated, functional decidual regulatory T cells
regulate these CD8+ T cell, resulting in normal pregnancy
(Tilburgs et al., 2009). The combination of maternal KIR AA
genotype and fetal HLA-C2 genotype may be a risk factor
for preeclampsia (Hiby et al., 2004). The frequency of fetal
HLA-C2 genotype is twice as high in DD twin pregnancies

A. Shiozaki et al. / Journal of Reproductive Immunology 89 (2011) 133139

as that in MD twin pregnancies. If these were the case, the

incidence of preeclampsia should be higher in DD twins
than in MD twins and should be similar in MD twins and
in singletons. The present study has demonstrated that the
incidence of preeclampsia is similar in both types of twins
(DD twins, 4.5% vs. MD twins, 4.2%) and consequently does
not support the MHC mismatching theory in the pathogenesis of PIH/preeclampsia.
The second hypothesis suggests that the difference in
the amount of fetal antigens leads to the pathogenesis
of preeclampsia. If this were the case, the incidence of
preeclampsia should be twice as high in twins as in singletons and should be similar in DD twins and MD twins.
In this study we found that the occurrence of preeclampsia is similar in both types of twins (DD twins, 4.5% vs. MD
twins, 4.2%) and these twins are at 1.51.6 times increased
risk of preeclampsia. Our result supports the fetal antigen
theory in the pathogenesis of PIH/preeclampsia.
The third hypothesis is related to the HY antigen.
Recent reports suggest that HY antigens are recognized
by maternal lymphocytes, and may induce abortion or
placental abruption (Nielsen et al., 2009; Christiansen
et al., 2010). If this were the case, the malemale twins
should have highest preeclampsia rate, the femalefemale
twins and female singletons should have the lowest
rate, and the malefemale twins and male singletons
should have the intermediate rate. We observed that the
femalefemale DD twins had the highest preeclampsia rate
(5.2%), malefemale (opposite-sex) DD twins had the intermediate rate (4.6%), and the malemale DD twins had the
lowest rate (3.9%). Additionally, we identied that the frequency of preeclampsia was higher in femalefemale MD
twins than in malemale MD twins. Finally, this nding
does not support the HY antigen theory in the pathogenesis of PIH/preeclampsia. Hiby et al. (2004) reported
the prevalence of preeclampsia decrease accompanying the
increase in the activation of receptors on KIR. Therefore,
maternal immune cell activation against fetal HY antigens
might reduce the risk of PIH or preeclampsia. Further studies are needed to clarify these points.
There is a limitation of this study. The fetal gender
of previous pregnancies in multiparous women was not
checked in the study. As a result, we could not compare the rates of preeclampsia in rst pregnancies with
those in second pregnancies and we could not evaluate
the independent effect of fetal gender on the preponderance of preeclampsia. Additional studies are needed
to clarify the role of fetal gender in the rst pregnancy
on the risk of preeclampsia in subsequent pregnancies.
This is a rst attempt to study the paradox of maternal
tolerance based on fetal gender and zygosity in large samples. The present study suggests that fetal HY antigen and
numbers of mismatched MHC antigens are not risk factors
in PIH and preeclampsia. Further studies are needed to clarify the pathogenesis of PIH and preeclampsia with regard
to the immunological aspect.

139

Funding
This study was supported by a grant from the Japanese
Ministry of Health, Labor and Welfare, H20-KodomoIppan-003, and a grant from the Ministry of Education,
Culture, Sports, Science and Technology, Japan (Grand-inAid for Scientic Research (B)-20390431).
Acknowledgment
We wish to thank Mr. Norio Sugimoto for statistical help.
References
Basso, O., Olsen, J., 2001. Sex ratio and twinning in women with hyperemesis or preeclampsia. Epidemiology 12, 747749.
Christiansen, O.B., Steffensen, R., Nielsen, H.S., 2010. The impact of antiHY, responses on outcome in current and subsequent pregnancies
of patients with recurrent pregnancy losses. J. Reprod. Immunol. 85,
914.
Dekker, G., 2002. The partners role in the etiology of preeclampsia. J.
Reprod. Immunol. 57, 203215.
Einarsson, J.I., Sangi-Haghpeykar, H., Gardner, M.O., 2003. Sperm exposure and development of preeclampsia. Am. J. Obstet. Gynaecol. 188,
12411243.
Hiby, S.E., Walker, J.J., OShaughnessy, K.M., Redman, C.W., Carrington, M.,
Trowsdale, J., Moffett, A., 2004. Combinations of maternal KIR and
fetal HLA-C genes inuence the risk of preeclampsia and reproductive
success. J. Exp. Med. 200, 957965.
Hsu, D.-D., Johnson, T.R.B., Witter, F.R., 1994. Fetal gender effect on
preeclampsia: a retrospective study of Baltimore area in the United
States of America. Int. J. Gynaecol. Obstet. 45, 160161.
Makhseed, M., Musini, V.M., Ahmed, M.A., 1998. Association of fetal gender with pregnancy-induced hypertension and preeclampsia. Int. J.
Gynaecol. Obstet. 63, 5556.
Matsuda, Y., Hayashi, K., Shiozaki, A., Kawamichi, Y., Satoh, S., Saito, S.
Comparison of risk factors for placental abruption and placenta previa,
case-cohort study. J. Obstet. Gynaecol. Res., doi:10.1111/j.14470756.2010.01408, in press.
Nielsen, H.S., Steffensen, R., Varming, K., Van Halteren, A.G., Spierings, E.,
Ryder, L.P., Goulmy, E., Christiansen, O.B., 2009. Association of HYrestricting HLA class II alleles with pregnancy outcome in patients
with recurrent miscarriage subsequent to a rst-born boy. Hum. Mol.
Genet. 18, 16841691.
Nielsen, H.S., Steffensen, R., Lund, M., Egestad, L., Mortensen, L.H.,
Andersen, A.M., Lidegaard, ., Christiansen, O.B., 2010. Frequency
and impact of obstetric complications prior and subsequent to
unexplained secondary recurrent miscarriage. Hum. Reprod. 25,
15431552.
Robillard, P.Y., Hulsey, T.C., Dekker, G.A., Chaouat, G., 2003. Preeclampsia
and human reproduction. An essay of a long term reection. J. Reprod.
Immunol. 59, 93100.
Saito, S., Shiozaki, A., Nakashima, A., Sakai, M., Sasaki, Y., 2007. The
role of the immune system in preeclampsia. Mol. Aspects Med. 28,
192209.
Sepulveda, W., Sebire, N.J., Souka, A., Nicholaides, K.H., 1996. The lambda
sign at 1014 weeks of gestation as a predictor of chorionicity in twin
pregnancies. Ultrasound Obstet. Gynaecol. 7, 421423.
Stevenson, A.C., Davison, B.C.C., Say, B., Ustouplu, S., Liya, D., Abul-Einen,
M., Toppozada, H.K., 1971. Contribution of fetal/maternal imcompatibility to aetiology of preeclamptic toxaemia. Lancet ii, 12861289.
Tilburgs, T., Scherjon, S.A., van der Mast, B.J., Haasnoot, G.W., VoortMaarschalk, Versteeg-V.D., Roelen, M., van Rood, D.L., Claas, J.J.F.H.,
2009. Fetal-maternal HLA-C mismatch is associated with decidual T
cell activation and induction of functional T regulatory cells. J. Reprod.
Immunol. 82, 148157.
Toivanen, P., Hirvonen, T., 1970. Sex ratio of newborns: preponderance of
males in toxemia of pregnancy. Science 170, 187188.
Trupin, L.S., Simon, L.P., Eskenazi, B., 1996. Change in paternity: a risk
factor for preeclampsia. Epidemiology 7, 240244.