Birth defects after a history of subfertility: a registry ...scripties.umcg.eldoc.ub.rug.nl/FILES/root/geneeskunde/2013/BosME/BosM.pdf · Birth defects after a history of subfertility:

Birth defects after a history of subfertility:

a registry-based case-control study in the

Northern Netherlands

M.E. Bos s1735187

EUROCAT Northern Netherlands, UMCG

Supervisors: dr. H.E.K. de Walle dept GENETICS

dr. A. Hoek dept OBGYN, dr. M.L. Haadsma dept GENETICS

Research period: August 2012 – December 2012

1

1 Abstract

Background: In children conceived via assisted reproduction, an increased overall

risk for birth defects has been described in literature. In these newborn, a specific

increase has been reported for cardiovascular, musculoskeletal, urogenital and

gastrointestinal abnormalities. It remains uncertain, however, whether the rise in

occurrence of birth defects can be attributed to patient characteristics related to

subfertility, or to the treatment itself. The primary aim of the study is to evaluate

whether specific types of birth defects occur more often in offspring of subfertile

couples rather than in the offspring of fertile couples. The secondary aim is to

evaluate when birth defects are more likely to develop: after a history of subfertility

compared to the offspring of couples who received in vitro fertilization (IVF) or

intracytoplasmatic sperm injection (ICSI). The tertiary aim is to evaluate whether

prolonged time to pregnancy (TTP) in the group of subfertile couples shows a positive

correlation with regard to the prevalence of specific types of birth defects.

Methods: For performing this case-control study we used data from EUROCAT –

Northern Netherlands, a registry of birth defects. 4392 cases and 1456 (malformed)

controls were identified. These subjects were born between 1997 and 2010, and were

registered by EUROCAT. The cases were children and fetuses with major birth

defects of a non-genetic origin. Controls were defined as children and fetuses with a

chromosomal anomaly, a microdeletion or a monogenic anomaly. In the statistical

analysis, three groups of children and fetuses were compared: Sub-NC (n=296),

IVF/ICSI (n=174) and children and fetuses of fertile couples (n=5378).

Results: Compared with fertile women, Sub-NC women were at an increased risk of

having a child with a penoscrotal hypospadia (adjusted odds ratio [aOR]: 6.80; 95%

confidence interval [CI]: 2.35 – 19.63). Subfertile women conceiving via IVF/ICSI

were at an increased risk of having a child with a limb anomaly (aOR: 2.80; CI: 1.51 –

5.20), especially polydactyly (aOR: 5.26; CI: 2.47 – 11.20). Compared to fertile

women, the entire group of subfertile women (Sub-NC and IVF/ICSI) was at an

additional increased risk of having a child with an epispadia (aOR: 4.28; CI: 1.04 –

17.57) and cystic kidneys (aOR: 2.10; CI: 1.02 – 4.34). Surprisingly, Sub-NC women

showed a significant decreased risk for having a child with a congenital anomaly in

general, considering congenital anomalies occurring in all groups. Overall, a

significantly decreased risk of having a child with a congenital hip dysplasia was seen

in subfertile women, as was also described for the Sub-NC women alone.

Conclusions: In this study, Sub-NC women were at an increased risk of having a

child with specific major congenital anomalies, especially penoscrotal hypospadias.

The results of this study may be of great importance for public health, since

subfertility is a growing health issue since women are conceiving at older ages.

Keywords: assisted reproductive techniques, IVF, ICSI, congenital abnormalities,

history of subfertility

2

2 Preface

This paper is intended as the final report of the science project for the Master in

Medicine. The emphasis lies on whether there is a correlation between specific major

congenital anomalies and subfertility. After the study was conducted, which greatly

expanded my interest in this topic, I have the ambition to do more research within this

field.

The main objective of the science project is to improve different skills: scientific

writing, critical reading and perform statistical analyses. This research is based on the

analysis of data from fertile records of fourteen hospitals in the Dutch provinces

Friesland, Groningen and Drenthe. Performing this research improved my ability to

read and understand fertility- and gynaecology- records. Moreover, I became more

experienced in communicating with gynaecologists.

Performing this science project and writing this report have been made

possible with the help of my supervisors, Hermien de Walle, Annemieke Hoek and

Maaike Haadsma.

Marly Bos

18-01-2013

3

Contents

1 Abstract 1

2 Preface 2

3 Introduction 4

3.1 Congenital anomalies 4

3.2 Assisted reproductive technology 5

3.3 IVF and ICSI procedures 5

3.4 Perinatal risks and congenital anomalies associated with ART 6

3.5 The role of subfertility in increasing birth defects rates 7

3.6 The aim of the present study 8

4 Methods 9

4.1 EUROCAT 9

4.2 Study population 10

4.3 Definitions and recruitment 10

4.4 Methods 11

4.5 Statistical analysis 11

5 Results 13

5.1 General 13

5.2 Characteristics 14

5.3 Fertility status in two and three categories 15

5.4 Time to pregnancy 18

6 Discussion and Conclusion 20

6.1 Main results 20

6.2 Strengths and weaknesses 21

6.3 Potential underlying mechanisms 22

6.4 Further research 23

6.5 Public health implications 24

6.6 Conclusion 24

7 Acknowledgments 25

8 References 26

9 Nederlandse samenvatting (Dutch summary) 30

Appendix 32

4

3 Introduction

3.1 Congenital anomalies

The term congenital anomaly refers to a broad spectrum of defects in body structures.

Congenital anomalies can be divided into anomalies of a genetic nature or anomalies

of a non-genetic nature. Genetic anomalies include chromosomal disorders (e.g.,

Down syndrome); single gene (monogenic) disorders, including those that are

autosomal recessive (e.g., cystic fibrosis), automosal dominant (e.g., Marfan

syndrome), or X-linked (e.g., hemophilia); as well as multifactorial disorders which

result from the interaction of multiple genetic and environmental factors. The latter

include birth defects such as a cleft lip/palate, congenital heart disease, and neural

tube defects. Non-genetic etiologies include environmental factors. Examples of non-

genetic defects are maternal phenylketonuria (PKU) or diabetes, teratogens (e.g.,

alcohol consumption) and infections. An important cause of perinatal mortality are

congenital anomalies, where 6-12% of stillbirths are attributed to a congenital

anomaly(1).

Congenital anomalies can be classified into four different categories. They can

be subdivided according to their presumed pathogenesis into malformations,

deformations, disruptions and dysplasias. Malformations are caused by an intrinsically

abnormal developmental process which leads to defects of organs or certain body

parts. Generally, malformations already occur in the first eight weeks of gestation and

are the result of a defect in embryonic development. Common occurring

malformations are neural tube defects and congenital heart defects. Deformations,

however, are not caused by intrinsically abnormal development, but are rather

produced by abnormal mechanical forces which distort otherwise normal structures

and could develop at any time in gestation. Oligohydramnion, for instance, may lead

to a child being born with an abnormal foot position. Unlike malformations or

deformations, disruptions results from an extrinsic disturbance leading to a

morphologic anomaly to an organ or body part. Anomalies caused by extrinsic

disturbance are typically caused by adverse environmental factors such as teratogenic

drugs, compression or strangulation. Anomalies can also appear as the result of

abnormal organization or function of cells into tissues, called dysplasias. When cells

are not correctly organized, this may lead to abnormal growth, resulting, for instance,

in skeletal dysplasia(2).

While congenital anomalies can be classified by the suspected underlying

cause, they may also be subdivided according to severity; as are major or minor

anomalies. The difference between major- and minor anomalies is that the former

have medical and/or social implications and often require surgery to be corrected,

while the latter generally at most have a cosmetic impact. Though minor anomalies

may be associated with a disorder or disability, they generally do not impede leading a

normal life and rarely require surgical intervention. Between 1981 and 2009, 2.8% of

all pregnancies in the Netherlands were affected by a major congenital anomaly(3).

5

3.2 Assisted reproductive technology

An assisted reproductive technique (ART) which is well known to the general public

is in vitro fertilization (IVF). The first IVF-baby, Louise Brown, was born in 1978.

Subsequently, over three million IVF children were born(5) and in Europe alone up to

3.9% of all pregnancies are conceived via IVF(6). Another ART is intracytoplasmatic

sperm injection (ICSI)(7). Ten to twenty percent of couples are subfertile: which

means they have an active waiting time to pregnancy of over 12 months(4). A

considerable number of these subfertile couples request the use of an assisted

reproductive technology in order to conceive. One of the reasons for the increasing

amount of ART requests is the fact that women start trying to conceive at increased

age. Not only do older woman have diminished fecundity, they also have an increased

risk for gestational diabetes, placenta previa, breech presentation, operative vaginal

delivery, Caesarean section, postpartum haemorrhage, delivery before 32 weeks

gestation, birth weight below the 5th

centile and stillbirth(8).

In addition, a Swedish population-based study reports a higher risk of

pregnancy complications after ART itself, regardless of increased parental age(9).

While women who conceive via IVF generally show lower smoking rates in early

pregnancy and have higher level of education, the increased odds of pregnancy

complications nevertheless persist(10).

Previous research has shown that children born after ART have more perinatal

health problems compared to naturally conceived children. Prematurity and low birth

weight, for example, are seen more often in ART-children, as well as perinatal

mortality and neonatal intensive care committal(11,12).

ART is also associated with higher rates of birth defects in offspring, although

contributing mechanisms are still controversial. Nonetheless, it is conceivable that

ART-related procedures interfere with developmental processes. Factors which ought

to be taken into account are, for instance, bypassing the natural selection of gametes

and artificially maturing the ovum. Additionally, the early embryonic development

takes place in vitro and the embryo is transferred into an altered uterus as the result of

ovarian hyperstimulation. Besides ART-related procedures, subfertility itself should

also be considered as a contributing mechanism, rather than just the ART procedures

themselves.

In the next sections, the IVF and ICSI procedures will be reviewed first, followed by a

discussion of the results of pregnancies after ART and the history of subfertility.

3.3 IVF and ICSI procedures

IVF refers to a procedure designed to overcome infertility. This procedure aims at

achieving a pregnancy as a direct result of the intervention. First, the ovaries are

stimulated by a combination of fertility medication, followed by the aspiration of

multiple oocytes from ovarian follicles. The oocytes are, then, fertilized by sperm in

the laboratory (hence "in vitro", meaning “in glass” in Latin), after which one or more

embryos are transferred into the uterine cavity. These steps occur over about a two-

week interval, which is called an IVF cycle.

6

The number of embryo’s which are transferred into the uterus is positively

associated with pregnancy rates. On the other hand, transferring multiple embryos has

a positive correlation with twin pregnancies. Transferring the embryos is successful

most of the times, but the biggest problem of the IVF cycle is embryo-implantation.

Of the IVF procedures started only 25-30% lead to a pregnancy. Of these pregnancies

another 20-25% is lost in abortion or as extrauterine gravidity. In total, only 20-25%

of IVF cycles results in the delivery of a child. Most couples undergo more than one

cycle, and the cumulative success rate of three cycles is 50%(13,14).

In 1992, ICSI was for the first time successfully used in humans. This was the

next step in assisted reproductive techniques(15). The technique of ICSI depends on

bringing a single spermatozoa directly into the cytoplasma of an oocyte by injection.

This procedure is performed as part of an IVF cycle, and is particularly useful when

semen parameters are suboptimal or when fertilization rates after conventional IVF

treatment were low or nil. Next to bypassing the natural selection of the fertilizing

sperm, ICSI may appear as a traumatic event which could damage the egg by the

operative procedure. Nevertheless, fertilization is generally successful and is usually

followed by normal embryonic development(16).

When embryos of good quality remain that are not transferred to the uterus,

they can be cryopreserved for future use. Around 10-20% of embryos do not survive

the thawing process. The freezing and thawing process is thought to cause subtle

damage to the embryos. Systematic reviews of observational studies have found that

children born after transfer of frozen-thawed embryos have better perinatal outcomes

than those born after transfer of fresh embryos (i.e., lower rates of preterm birth, low

birth weight, growth restriction, perinatal mortality). Malformation rates are

comparable, while data on growth, childhood morbidity, and mental development

were limited though with few differences appeared between the groups(17,18). The

reason for the slightly higher number of healthy children born after cryopreservation

compared to children born after fresh transfer (in most studies) is not known. It may,

however be related to differences in endometrial receptivity between women

undergoing fresh versus cryopreserved embryo transfer (e.g., adverse effects of

ovarian stimulation in ‘fresh’ IVF cycles). Another hypothesis is that freezing and

thawing serves as a selection mechanism, since only the most viable embryo’s survive

this procedure.

3.4 Perinatal risks and congenital anomalies associated with ART

Over the years, assisted reproductive techniques have become widely accepted in

society. However, removing and handling gametes outside the body has always raised

concerns what the short- and long-term safety outcomes may be. For couples

considering ART request it is important to know the risks of ART techniques. Next to

that, society as a whole will benefit from the evaluation of perinatal and

neurodevelopmental risks associated with ART, since adverse outcomes would place

a heavy burden on the health care system. Negative short-term outcomes are

repeatedly described. Singletons born after ART are more often preterm and of low

birth weight rather than naturally conceived singletons(12,19). Among the former

7

group, there are higher numbers of perinatal mortality and they need to be admitted to

neonatal intensive care more often (11,12). The low birth weight may not only

indicate preterm birth but also a growth retarding effect. This effect was demonstrated

by Kallen et al., and they found that 5.1% of IVF singleton infants were small for

gestational age, whereas no more than 2.8% among all singleton infants delivered

were born with a low birth weight. This study also reported a doubled over-all

hospitalization rate for IVF children(20).

Additionally, another current discussion in the field concerns the safety

outcome of ART with regard to the prevalence of birth defects. In a meta-analysis,

Hansen et al. (2005) showed an 40% increased risk for having a child with birth

defects after conceiving via ART(21). Previous research by Hansen et al. in 2002

suggested already that children born after ART were at higher risk of having multiple

major birth defects, especially musculoskeletal defects(22).

A study by Reefhuis et al. found a relationship between ART and several

specific anomalies. They showed an increase in the occurrence of septal heart defects,

cleft lip (with or without cleft palate), esophageal atresia and anorectal atresia in

singletons born after ART(7). Another study confirmed the findings by Reefhuis et al.

reported a higher prevalence of neural tube defects(23). Davies et al. (2012) reported

a higher prevalence of cardiovascular, musculoskeletal, urogenital and gastro-

intestinal congenital anomalies, as well as cerebral palsy in children conceived after

ART(24). Multiple studies showed a correlation between ART and the VACTERL-

association, which stands for multiple associated defects: Vertebral defects, Anal

atresia, Cardiac defects, Tracheo-Esophageal fistula, Renal malformations and Limb

defects(7,23).

Hypospadia is the only congenital anomaly which shows a disparity between

IVF and ICSI in terms of occurrence: the risk increases slightly when ICSI is

compared to IVF. This discrepancy may, however, be the result of paternal subfertility

rather than the ICSI procedure per se(22,25-27).

Summarizing, there appears to be a relationship between ART and congenital

malformations. However, the underlying mechanisms remain unclear. Whether the

ART procedure itself is responsible for the higher risk, or the underlying subfertility,

is an issue which still needs to be addressed

3.5 The role of subfertility in increasing birth defects rates

When the health condition in children born after ART is studied, it is imperative to

know the difference between the effects of the treatment techniques and the

underlying subfertility. A premature birth after a history of subfertility is known to be

associated with an increased risk of perinatal death(28).

Evidence is accumulating indicating that spontaneous conception after a

history of subfertility and its determinants are involved in increasing birth defects,

rather than ART-related procedures. These results were to be expected, since

subfertile couples are generally older, and age is a factor which is typically associated

with a higher risk of birth defects in offspring. Nevertheless, after correction for

parental age, it is argued in multiple studies that a history of subfertility was still

8

associated with higher rates of birth defects. A role of the underlying subfertility in

increasing birth defect rates was suggested for the first time by Ghazi et al. They

reported higher rates of congenital anomalies in a subgroup of couples who had

experienced subfertility for at least four years(29). Zhu et al. confirmed that a history

of subfertility without assisted conception is associated significantly with birth defects

and they described a high hazard ratio (hazard ratio [HR]: 1.20; 95% confidence

interval [CI]: 1.07 to 1.35) for congenital anomalies in subfertile couples who

conceived naturally, compared to fertile couples. These associations did not disappear

after correction for maternal age, maternal BMI at conception, smoking, alcohol

and/or coffee consumption during pregnancy and the level of education(30).

Zhu et al. and Seggers et al. found that increased time to pregnancy (TTP) was

associated with a higher prevalence of birth defects(30,31). Another study reported

that the increasing odds for birth defects after ART disappeared after correction for

time to pregnancy(23).

3.6 The aim of the present study

An increase of several specific types of birth defects has been repeatedly described

after ART: cardiovascular defects, musculoskeletal defects, neural tube defects,

urogenital defects, gastrointestinal abnormalities, cleft lip/palate and cerebral palsy.

Whether a history of subfertility is also associated with specific types of birth defects

remains unclear. This information would be quite useful in counseling subfertile

couples and may be applied in (extended) prenatal screening.

Therefore, the primary aim of this case-control study was to examine whether

specific types of birth defects occur more often in offspring of subfertile couples

compared to offspring of fertile couples. The secondary aim was to evaluate whether

specific types of birth defects occur more (or less) often after a history of subfertility

compared to the offspring of couples who received IVF/ICSI. The tertiary aim was to

evaluate whether prolonged time to pregnancy in the group of subfertile couples

shows a positive association with the prevalence of specific types of birth defects.

9

4 Methods

4.1 EUROCAT

In order to accurately describe and analyze birth defects among couples using ART or

suffering from subfertility, a massive amount of data needs to be available for

improving the power of the research. The EUROCAT Northern Netherlands

(EUROCAT NNL) database proved indispensible in this research, and offered a

wealth of information. EUROCAT is a population-based birth defects registry and

since 1981 it registers children and fetuses suffering from congenital anomalies

diagnosed before or after birth, born in the provinces Groningen, Friesland or

Drenthe. In this database, however, no information is collected on non-malformed

infants. Approximately 18.500 births yearly are covered by the registration. Among

those births which are registered, all types of pregnancy outcomes are represented and

there is no discrimination for gestational age. Deliveries that are registered are divided

into five categories: live birth, live born but deceased, stillbirth, terminations of

pregnancy because of a fetal anomaly and miscarriage. For a fetus’ death to be

considered a stillbirth the gestation needs to be at least 24 weeks and the fetus needs

to have either died in utero or during birth. For miscarriages the upper age limit is 24

weeks, there is no weight limit.

Although EUROCAT provides ample date on major anomalies detected in

children, it has far less information to offer on minor anomalies, since it had been

decided only to include those anomalies under certain conditions according to the

EUROCAT Central Registry Guidelines(32). As to what exactly constitutes an

anomaly in infants, EUROCAT relies on the classification of the 9th

and 10th

revisions

of the WHO International Classification of Diseases (ICD). Births up to 2001 were

classified according the ICD-9 classification; from 2002 onward, the ICD-10

classification is used. Information is gathered in a number of ways. Anomalies can be

reported by different medical institutions or sometimes by the parents themselves. In

addition, EUROCAT has an active and systematic search for children and fetuses with

a congenital anomaly in hospital records. The upper age limit for children to be

included up is 10 years. Parents are to provide written informed consent for

registration; the participation rate is over 80%. Of those parents who agree with

registration, around 80% returned the EUROCAT questionnaire. In this questionnaire,

parents were asked to report characteristics such as parental age, height, pre-

pregnancy weight, chronic illnesses, education, smoking, alcohol consumption, the

use of medication, socioeconomic status and information on mode of conception.

Subsequently, information on the use of medication dispended up to approximately

three months before conception and during pregnancy, is collected via community

pharmacies. A telephone interview with the mother is scheduled once the pharmacies

reported on the medication which was dispensed, to verify whether the mother

actually has taken the medication and if she has taken any self-medication before

conceiving or during the pregnancy. In case the questionnaire was not fully

10

completed, the telephone interview granted the opportunity to fill in any missing

answers.

4.2 Study population

The subjects that were included in the study were born between 1997 and 2010 and

were defined as follows: children and fetuses with a major congenital anomaly.

Major congenital anomalies were divided into different diagnostic groups

according to the EUROCAT international guidelines(32). For example subjects with

congenital heart defects were divided into subgroups based on scientific

epidemiological insights(33).

In order to test the accuracy of conclusions drawn from the EUROCAT data, we

choose for a case-control design. Extensive literature research on the topic of

subfertility and birth defects, showed no evidence of subfertility causing chromosomal

anomalies, microdeletion anomalies or monogenic anomalies. Therefore children and

fetuses born between 1997 and 2010 with a chromosomal anomaly, a microdeletion or

a monogenic anomaly were selected to act as a control group. Additionally, in studies

on risk factors for congenital anomalies the use malformed controls is widely

accepted(34).

For each group of cases, a different control group was selected carefully, as it

would be counterproductive to the aim of the study to allow the controls to have the

same congenital anomaly as the cases to which they were compared. The selection

procedure thus entailed, for example, that a child suffering from a chromosomal

anomaly, such as trisomy 18, and also a neural tube defect was used as a control for

all subgroups of anomalies, except for the group affected by neural tube defects. An

overview of the number of controls used for each congenital anomaly can be found in

TABLE 1.

4.3 Definitions and recruitment

From the initial 7720 participants born between 1997 and 2010, 246 participants were

excluded as they were coded into the diagnostic group ‘other syndromes’.

In the EUROCAT questionnaire parents were asked whether they had

problems conceiving with regard to this specific pregnancy. If participants did not

answer the question or when they filled in ‘unknown’, they were excluded.

Participants who answered ‘no’ were included as confirmed fertile participants. When

the answer given was ‘yes’, they were asked to specify whether they eventually

conceived spontaneously or if they received treatment; IVF, ICSI, IUI, KID, hormonal

therapy or other therapy. In case IVF, ICSI or IUI was used, parents were asked to

specify whether donor semen and/or donor oocytes were used. As our research

focuses on subfertile participants who conceived naturally and via IVF/ICSI, all other

fertility therapies were excluded.

To objectify the problems in becoming pregnant, a fertility record search was

performed. Lists of the self-reported subfertile participants were sent to the

EUROCAT contact persons in 15 hospitals in the provinces Groningen, Friesland and

Drenthe. The contacts were requested to let us perform a search for fertility (or

11

gynaecology) records of the participants in question. Medical records provided

information on fertility status, information on IVF/ICSI treatment procedures, fertility

diagnosis and TTP in terms of years. TTP was defined as the time between the active

child wish of the couple and conception. In case of a miscarriage, TTP-onset had to be

reset and ended at conception of the child included in the present study. Participants

were confirmed to be subfertile when they had a fertility record in which at least an

orientating fertility research was reported and/or a TTP of at least one year. The

confirmed subfertile participants were included. FIGURE 1 shows the flowchart.

EUROCAT NNL obtained permission from the registered participants to

request for medical information for scientific purposes. EUROCAT NNL also

obtained permission from the ethics committee of the University Medical Center

Groningen to use their data for scientific purposes.

4.4 Methods

Essentially, the fertility status was divided into two categories in the analysis; fertile

versus subfertile. In case there were sufficient numbers, i.e. more than 3 exposed

cases per category, the fertility status was divided into three categories rather than

two; fertile versus subfertile, as well as naturally conceived (sub-NC) versus

IVF/ICSI.

The main focus lies on subfertile couples who conceived naturally,

Additionally, ‘IVF/ICSI’ was analyzed to see whether ART was associated with

specific congenital anomalies. Furthermore, an attempt was made to analyze a

possible association between TTP and specific congenital anomalies. To this end, TTP

was divided into TTP < 2 years and TTP ≥ 2 years.

4.5 Statistical analysis

Since the controls that were used are mainly comprised of chromosomal anomalies,

which are known to be overrepresented in women of increased age at conception,

maternal age at conception is used as a confounder, even in the ‘crude’ analysis. To

estimate the risk for congenital anomalies, the ‘crude’ odds ratios (ORs) were

calculated first by using a ‘multivariate’ logistic regression with only maternal age at

conception as confounder. In the model, the fertility status in two or three categories

was the main independent variable and the different subgroups of congenital

anomalies were the dependent variables. Confirmed fertile participants were used as

reference category in the fertility status analyses. In the TTP analyses TTP < 2 years

was used as reference category.

To explore the contribution of the potential confounders, multivariate logistic

regression to calculate adjusted ORs was used. Apart from maternal age at

conception, potential confounders included pregnancy outcome, infant sex, education

of the mother, smoking, folic acid use, body mass index (BMI), diabetes mellitus,

parity, number of children in this pregnancy and a history of another pregnancy

affected by congenital anomalies.

To allow the level of education of the mother in question to be taken into

consideration as a possible factor influencing the outcome of the pregnancy, the level

12

of education was subdivided into a number of categories. Subsequently, the maternal

level of education was noted as either low, middle or high. The definition for low

level of education was primary school, lower general secondary education and lower

vocational education. Middle level education was defined as higher general secondary

education and intermediate vocational education. Those mothers who had a higher

vocational education, university degree or studied at a further tertiary college were

grouped among those with a high level of education.

Certain habits, such as smoking, could have an impact on the odds of giving

birth to a child suffering from birth defects. The mothers, who were included in this

study, were divided into the groups smoking or non-smoking during pregnancy. When

a woman was placed in the smoking category, this implies either smoking during

pregnancy, the mother has smoked but stopped at the moment she knew she was

pregnant, smoked only at the end of pregnancy or smoked but the exact period in

which the mother has smoked is unknown. In contrast, to be placed in the non-

smoking category implies that the mother either smoked but stopped before she

became pregnant or was a non-smoker.

The BMI classification of the World Health Organization (WHO) was used:

underweight (BMI < 18.50 kg/m2), normal weight (BMI 18.50 – 24.99 kg/m

2) and

overweight (BMI > 25 kg/m2)(35).

For counteracting the problem of multiple comparisons in statistical analyses,

the Bonferroni correction could have been used. Since this study is an explorative

research it was decided to not correct for multiple testing. Not correcting for multiple

testing in explorative research is widely accepted(36).Therefore, a p-value of less than

0.05 and 95% confidence interval (CI) excluding 1.00 was considered to be

statistically significant. Statistical analyses were performed using IBM SPSS Statistics

20.

13

5 Results

5.1 General

The study period ranged from 1997 to 2010 and within this timeframe, 7720 of the

pregnancies in the EUROCAT-database were affected by a major congenital anomaly.

FIGURE 1, the flowchart, illustrates the inclusion and exclusion of participants in the

study (see Methods, page 9). Out of those 7720 pregnancies, 5848 participants with

known fertility status could be included in the present analyses. As indicated by

FIGURE 1, 470 participants were confirmed subfertile, of whom 296 conceived

naturally, while 174 were conceived through IVF/ICSI.

Eurocat participants 1997-2010

n = 7720

Eurocat questionnaire (n = 7474):

Having problems getting

pregnant?

Other syndromes (n = 246)

Unknown or Missing (n = 1077) No confirmed fertile

participants

n = 5378

Yes subfertile participants

n = 1019

IUI, KID, hormone therapy and

other therapy (n = 329)

Subfertile according to

fertility report research?

(n = 690)

No not confirmed

subfertile (n = 220)

Yes confirmed subfertile

participants

n = 470

Naturally conceived

n = 296

IVF/ICSI

n = 174

Excluded participants Included participants

FIGURE 1. Flowchart.

14

Of those 5848 participants with known fertility status, 4392 were included as

cases (75.1%) and 1456 as controls (24.9%). Of the cases, congenital heart defects

were the most represented defects (1160 (26.4%)), followed by musculoskeletal

anomalies (950 (21.6%)) and digestive anomalies (502 (11.4%)). Genital anomalies,

including hypospadias, were present in 392 cases (8.9%) (TABLE 1).

n (%) n (%) n (%) n n

All anomalies 4065 (92.6) 200 (4.6) 127 (2.9) 4392 1456

Anomalies CNS 288 (94.4) 11 (3.6) 6 (2.0) 305 1345

Neural tube defects 150 (95.5) 6 (3.8) 1 (0.6) 157 1439

Congenital heart defects 1060 (91.4) 60 (5.2) 40 (3.4) 1160 1222

Respiratory anomalies 77 (91.7) 6 (7.1) 1 (1.2) 84 1426

Clefts 351 (93.1) 16 (4.2) 10 (2.7) 377 1391

Digestive anomalies 458 (91.2) 28 (5.6) 16 (3.2) 502 1397

Genital anomalies 362 (92.3) 25 (6.4) 5 (1.3) 392 1421

Urinary anomalies 401 (92.0) 25 (5.7) 10 (2.3) 436 1372

Musculoskelatal anomalies 890 (93.7) 35 (3.7) 25 (2.6) 950 1401

Limb anomalies 249 (89.2) 12 (4.3) 18 (6.5) 279 1334

Total of cases and total of controls per main diagnostic group are displayed. Sub-NC = subfertile, but naturally conceived;

CNS = central nervous system.

TABLE 1. Main groups of congenital anomalies according to fertilitystatus in three categories.

Total casesTotal

controlsCongenital anomaly

Fertile

n = 5378

Sub-NC

n = 296

IVF/ICSI

n = 174

5.2 Characteristics

In TABLE 2, the characteristics of the cases and controls, with a confirmed parental

fertility status, are listed. Additional information on these cases and controls can be

found in Appendix 1. Cases did significantly less often have a mother with a history

of subfertility (4.6%) compared to controls (6.6%) (p-value = 0.006). This outcome

was to be expected, since mothers of controls were significantly older, and controls

are therefore suggested to be less fertile. In addition, mothers of cases were more

often nulliparous and their pregnancies ended less frequently in stillbirth, miscarriage

or termination. Less often did mothers of cases have a previous child with a

congenital anomaly, yet they were more likely to smoke during pregnancy. The level

of education differed slightly between mothers of cases and controls; 33.7% of the

mothers of controls and 30.3% of the mothers of cases had a high level of education.

15

Cases Controls P-value(n=4392) (n=1456)

<0.001

< 25 448 (10.2) 105 (7.2)

25-34 3164 (72.0) 905 (62.2)

> 34 736 (16.8) 432 (29.7)

Missing 44 (1.0) 14 (1.0)

Maternal education 0.015

Low 773 (17,6) 254 (17.4)

Middle 2082 (47.4) 629 (43.2)

High 1329 (30.3) 491 (33.7)

Missing 208 (4.7) 82 (5.6)

Smoking during pregnancy 0.033

Yes 1071 (24.4) 312 (21.4)

No 3232 (73.6) 1100 (75.5)

Missing 89 (2.0) 44 (3.0)

Body mass index (kg/m2

) 0.128

< 18.5 141 (3.2) 39 (2.7)

18.5 - 25 2619 (59.6) 900 (61.8)

> 25 1382 (31.5) 421 (28.9)

Missing 250 (5.7) 96 (6.6)

Diabetes mellitus 0.922

Yes 16 (0.4) 5 (0.3)

No 4305 (98.0) 1415 (97.2)

Missing 71 (1.6) 36 (2.5)

Parity < 0.001

0 1717 (39.1) 462 (31.7)

1 1438 (32.7) 497 (34.1)

2 716 (16.3) 261 (17.9)

3 or more 482 (11.0) 222 (15.2)

Missing 39 (0.9) 14 (1.0)

Fertility status 0.006

Fertile 4065 (92.6) 1313 (90.2)

Sub-NC 200 (4.6) 96 (6.6)

IVF/ICSI 127 (2.9) 47 (3.2)

0.569

< 2 years 129 (39.4) 60 (41.9)

≥ 2 years 165 (50.5) 68 (47.6)

Missing, but with OFO 33 (10.1) 15 (10.5)

Time to pregnancy within subfertile patients n=470

Sub-NC= subfertile, but naturally conceived; OFO = oriëntating fertility research.

TABLE 2. Distribution of maternal and infant characteristics for cases and

controls of subjects with known fertilitystatus.

Maternal age at conception (years)

5.3 Fertility status in two and three categories

Investigated confounders include maternal age at conception, smoking, BMI,

education, parity and another pregnancy affected by congenital anomalies. For

16

anomalies of the central nervous system (CNS), neural tube defects and congenital

heart defects an additional adjustment for maternal pre-gestational diabetes has been

done, since it is a known risk factor for these anomalies(37). Smoking, education,

parity, a previous pregnancy affected by congenital anomalies and diabetes were

considered as categorical variables and maternal age and BMI as continuous

variables. Other variables that we did investigate, but were found to have no

confounding effect included pregnancy outcome, infant sex, folic acid use and number

of children in this pregnancy.

TABLE 3 and 4 present the ‘crude’ and adjusted associations of congenital

anomalies and fertility status in two and three categories. The bold printed values

indicate a statistically significant difference between groups. The main diagnostic

groups are shown here, as well as specific congenital anomalies which show a

significant association. For most congenital anomalies, there were small differences

between the ‘crude’ and adjusted associations, except for genital anomalies. For

description of the results, we only used the adjusted associations as presented in

TABLE 4.

Reference

categoryCrude OR's (95% CI) Crude OR's (95% CI) Crude OR's (95% CI) n n

All anomalies 1.00 0.81 (0.63 - 1.05) 1.08 (0.76 - 1.52) 0.91 (0.74 - 1.13) 4392 1456

Anomalies CNS 1.00 0.59 (0.31 - 1.13) 0.63 (0.27 - 1.50) 0.62 (0.37 - 1.05) 305 1345

Neural tube defects 1.00 0.62 (0.27 - 1.45) . 0.50 (0.23 - 1.09) 157 1439

Congenital heart defects 1.00 0.90 (0.63 - 1.27) 1.40 (0.88 - 2.24) 1.08 (0.81 - 1.44) 1160 1222

Respiratory anomalies 1.00 1.23 (0.52 - 2.92) 0.42 (0.06 - 3.06) 0.99 (0.44 - 2.20) 84 1426

Clefts 1.00 0.68 (0.40 - 1.18) 0.91 (0.45 - 1.83) 0.75 (0.48 - 1.17) 377 1391

Digestive anomalies 1.00 1.00 (0.64 - 1.56) 1.13 (0.63 - 2.04) 1.05 (0.73 - 1.52) 502 1397

Genital anomalies 1.00 1.10 (0.69 - 1.74) 0.45 (0.18 - 1.14) 0.91 (0.60 - 1.38) 392 1421

Urinary anomalies 1.00 0.93 (0.59 - 1.48) 0.84 (0.42 - 1.69) 0.90 (0.61 - 1.34) 436 1372

Musculoskelatal anomalies 1.00 0.61 (0.41 - 0.91) 0.97 (0.58 - 1.60) 0.74 (0.54 - 1.02) 950 1401

Limb anomalies 1.00 0.73 (0.39 - 1.37) 2.28 (1.28 - 4.04) 1.28 (0.83 - 1.96) 279 1334

Adjusted only for maternal age at conception. Total of cases and total of controls per main diagnostic group are displayed. Sub-NC = subfertile, but

naturally conceived; CNS = central nervous system; . = n less than 3.

Total casesTotal

controls

TABLE 3. 'Crude' associations of main groups of congenital anomalies by fertility status categories.

Sub-NC + IVF/ICSI

n = 470Congenital anomaly

Fertile

n = 5378

Sub-NC

n = 296

IVF/ICSI

n = 174

17

Sub-NC

Compared to fertile women, Sub-NC women were at an increased risk of having a

child with a penoscrotal hypospadia. On the other hand, Sub-NC women show a

significant decreased risk of having a child with a musculoskeletal anomaly,

especially congenital hip dysplasia. Surprisingly, Sub-NC women show a nearly

statistically significant decreased risk when looking at the group of all congenital

anomalies together.

IVF/ICSI

Women who conceived via IVF or ICSI were at an increased risk of having a child

with a limb anomaly, especially polydactyly occurred more often in IVF/ICSI

children compared to children of fertile women.

Sub-NC + IVF/ICSI

Compared to fertile women, the entire group of subfertile women (including both the

women who conceived naturally or conceived via IVF or ICSI) were at an increased

risk of having a child with a penoscrotal hypospadia or an epispadia. In addition, an

increased risk of having a child with cystic kidneys was seen. In the entire group of

subfertile women, a significantly decreased risk of having a child with a congenital

hip dysplasia was seen, as was also described for the Sub-NC women alone.

Reference

categoryAdjusted OR's (95% CI)

aAdjusted OR's (95% CI)

aAdjusted OR's (95% CI)

a n n

All anomalies 1.00 0.77 (0.59 - 1.01) 1.08 (0.74 - 1.57) 0.86 (0.69 - 1.08) 4392 1456

Anomalies CNS 1.00 0.57 (0.28 - 1.12)b

0.78 (0.32 - 1.90)b

0.63 (0.36 - 1.10)b 305 1345

Neural tube defects 1.00 0.53 (0.21 - 1.36)b . 0.46 (0.19 - 1.08)

b 157 1439

Congenital heart defects 1.00 0.91 (0.63 - 1.31)b

1.58 (0.95 - 2.63)b

1.09 (0.81 - 1.48)b 1160 1222

Respiratory anomalies 1.00 1.04 (0.40 - 2.68) . 0.74 (0.29 - 1.90) 84 1426

Clefts 1.00 0.68 (0.38 - 1.21) 0.74 (0.32 - 1.71) 0.70 (0.43 - 1.13) 377 1391

Digestive anomalies 1.00 0.89 (0.57 - 1.44) 1.17 (0.62 - 2.19) 0.98 (0.66 - 1.45) 502 1397

Genital anomalies 1.00 1.04 (0.64 - 1.67) 0.27 (0.08 - 0.89) 0.79 (0.51 - 1.24) 392 1421

Penoscrotal hypospadias 1.00 6.80 (2.35 - 19.63) . 4.76 (1.64 - 13.80) 21 1454

Epispadias 1.00 . . 4.28 (1.04 - 17.57) 11 1456

Urinary anomalies 1.00 0.91 (0.56 - 1.48) 0.94 (0.44 - 2.01) 0.92 (0.61 - 1.40) 436 1372

Cystic kidney 1.00 2.23 (1.00 - 4.94) 1.71 (0.38 - 7.65) 2.10 (1.02 - 4.34) 75 1434

Musculoskelatal anomalies 1.00 0.59 (0.39 - 0.90) 0.90 (0.53 - 1.54) 0.69 (0.49 - 0.96) 950 1401

Congenital hip dysplasia 1.00 0.36 (0.20 - 0.65) 0.97 (0.54 - 1.75) 0.55 (0.37 - 0.84) 652 1444

Limb anomalies 1.00 0.51 (0.24 - 1.09) 2.80 (1.51 - 5.20) 1.15 (0.72 - 1.84) 279 1334

Polydactyly 1.00 0.64 (0.23 - 1.82) 5.26 (2.47 - 11.20) 1.79 (0.98 - 3.26) 120 1400

TABLE 4. Adjusted associations of main groups of congenital anomalies by fertility status in categories.

Congenital anomalyFertile

n = 5378

Sub-NC

n = 296

IVF/ICSI

n = 174

a Adjusted for maternal age at conception, smoking, body mass index, education, parity and another pregnancy affected by congenital anomalies.

b Additional adjustment for diabetes.

Total of cases and total of controls per main diagnostic group are displayed.

Sub-NC = subfertile, but naturally conceived; CNS = central nervous system; . = n less than 3.

Sub-NC + IVF/ICSI

n = 470Total cases

Total

controls

18

5.4 Time to pregnancy

By searching through fertility records, TTP of 422 confirmed subfertile participants

was found. The effect of TTP on the prevalence of congenital disorders was analyzed,

by comparing those 422 participants with 5378 confirmed fertile participants. Results

are shown in TABLE 5.

Compared to women who had a TTP less than two years, women with a longer

TTP were at an increased risk of having a child with a polydactyly. A TTP more than

two years shows a decreased risk of having a child with a cleft, compared to a TTP of

less than two years.

Reference

categoryAdjusted OR's (95% CI)

a n n

All anomalies 1.00 0.91 (0.66 - 1.24) 4359 1441

Anomalies CNS 1.00 0.66 (0.31 - 1.43)b

303 1330

Neural tube defects 1.00 0.48 (0.14 - 1.59)b

156 1424

Congenital heart defects 1.00 1.21 (0.79 - 1.84)b

1154 1210

Respiratory anomalies 1.00 0.57 (0.13 - 2.42) 84 1413

Clefts 1.00 0.44 (0.19 - 0.97) 372 1377

Digestive anomalies 1.00 1.14 (0.68 - 1.91) 499 1383

Genital anomalies 1.00 0.91 (0.50 - 1.65) 390 1406

Penoscrotal hypospadias 1.00 2.87 (0.59 - 14.04) 20 1439

Epispadias 1.00 . 11 1441

Urinary anomalies 1.00 0.99 (0.55 - 1.81) 429 1357

Cystic kidney 1.00 2.44 (0.89 - 6.71) 71 1419

Musculoskelatal anomalies 1.00 0.70 (0.44 - 1.12) 946 1388

Congenital hip dysplasia 1.00 0.69 (0.40 - 1.19) 650 1430

Limb anomalies 1.00 1.10 (0.56 - 2.16) 277 1319

Polydactyly 1.00 2.40 (1.12 - 5.13) 120 1385

a Adjusted for maternal age at conception, smoking, body mass index, education, parity and another

pregnancy affected by congenital anomalies. b

Additional adjustment for diabetes.

Total of cases and total of controls per main diagnostic group are displayed.

CNS = central nervous system; . = n less than 3.

Total casesTotal

controls

TABLE 5. Adjusted associations of main groups of congenital anomalies by time to pregnancy

(TTP) in categories within the subfertile participants with known TTP.

Congenital anomalyTTP < 2 year

n = 5567

TTP ≥ 2 years

n = 233

In the examination of TTP in four categories, a positive trend was seen

between increasing TTP and an increased risk of having a child with a penoscrotal

hypospadia, a polydactyly and a cystic kidney (FIGURE 2). This positive trend was

not statistically significant, but adjusted ORs for penoscrotal hypospadias were most

striking. The adjusted OR increased from 4.29 (95% confidence interval [CI]: 0.88 –

20.4) for women with TTP between one and two years, up to 5.21 (CI: 1.04 – 26.12)

for women with a TTP between two and four years.

19

FIGURE 2. Trend of the risk of a penoscrotal hypospadia-affected pregnancy, a polydactyly-

affected pregnancy and a cystic kidney-affected pregnancy in adjusted ORs, with 95% CI,

compared to TTP in years.

0

5

10

15

20

25

30

<1 1-2 2-4 >4

Ad

just

ed

OR

's w

ith

95

% C

I

TTP in years

Penoscrotal hypospadias Cystic kidney Polydactyly

20

6 Discussion

6.1 Main results

In this study, the question was addressed whether a history of subfertility, with or

without IVF/ICSI treatment, affects the risk of having a child with a specific

congenital anomaly. It was shown that subfertility is associated with an increased risk

of several congenital anomalies in naturally conceived offspring in subfertile women,

as well as in offspring conceived via IVF or ICSI. The most striking result was the

increased risk found in subfertile women who conceived naturally: they had a higher

risk of having a child with a penoscrotal hypospadia. Sub-NC women also were at

increased risk of having a child with an epispadia. In women who conceived via IVF

or ICSI we found an increased risk of having a child with a limb anomaly, especially

polydactyly occurred more often in this group. The total group of subfertile

participants (Sub-NC + IVF/ICSI) showed an increased risk of having a child with a

cystic kidney, compared to fertile women. We also found an association between

increased TTP (>2 years) and polydactyly. In addition, increased TTP showed a

positive trend for the risk of a pregnancy affected with a penoscrotal hypospadia, a

polydactyly, and a cystic kidney, but these trends were not statistically significant.

On the other hand, we found a decreased risk of having a child with a

congenital hip dysplasia in Sub-NC women. We also found a decreased risk of having

a child with a cleft in women with a TTP of more than two years, compared to women

with a TTP less than two years.

Comparison with the literature

Some of the results are in line with the literature, whereas others are not. First, in

accordance with Zhu et al., Ghazi et al. and Seggers et al., the present study showed

some evidence for an adverse effect of subfertility. That underlying subfertility played

a part in increasing birth defect rates was suggested by Ghazi et al. in a large Swedish

case-control study. They reported higher rates of congenital anomalies after a history

of subfertility of at least four years(29). Zhu et al. (2006) confirmed that a history of

subfertility without assisted conception is strongly associated with birth defects and

they described a high hazard ratio for congenital anomalies in subfertile couples who

conceived naturally, compared to fertile couples, in a large Danish national birth

cohort(30). Zhu et al. also described an increased prevalence of genital organ

malformations in babies born after ART compared with babies conceived naturally.

Others also repeatedly described that singletons born following IVF/ICSI have a

higher prevalence of congenital malformations, including hypospadias, compared to

singletons conceived naturally. Hypospadia is the only congenital anomaly which

consistently shows a different risk rate among different ART procedures in different

studies: the risk is slightly increased in children conceived via ICSI compared to IVF.

The authors of the studies suggested that this discrepancy could be the result of a

genetic link with paternal subfertility, rather than with the ICSI procedure per

21

se(22,25-27). Within the present study this comparison between IVF and ICSI was not

made, since numbers were too low. A higher risk for penoscrotal hypospadias in the

Sub-NC group was found, suggesting a role of the underlying subfertility instead of

the ART procedures.

A case-control study in the United States of America conducted by Reefhuis et

al. (2009) found associations between ART and several specific anomalies. They

reported a higher occurrence of septal heart defects, cleft lip (with or without cleft

palate), esophageal atresia and anorectal atresia in singletons born after ART(5,38).

Another study showed a higher risk of having a child with a neural tube defect after

conceiving via ART(23). In an Australian study, published in the New England

Journal of Medicine. Davies et al. (2012) reported a higher prevalence of

cardiovascular, musculoskeletal, urogenital and gastro-intestinal congenital anomalies

and of cerebral palsy in children conceived after ART(24). These findings, however,

do not corresponds with our results, since only an increased risk for limb anomalies

after IVF/ICSI was found and not for any of the other birth defects listed above. The

absence of a correlation between ART and for example cardiovascular defects, may

be the result of the choice of cases and controls, in which controls were children with

a chromosomal anomaly, a microdeletion or a monogenic anomaly, though the use of

malformed controls is a widely accepted in medical research. While chromosomal

anomalies are not associated with subfertility in academic literature, it is plausible that

there may indeed be an such a relationship. After having repeated the analyses of the

data by using only monogenic anomalies as controls, the results yielded no important

differences compared to the original controls.

Zhu et al. also found that increased TTP was associated with a higher

prevalence of birth defects(30,31). Unlike this study, they did not investigate whether

specific congenital anomalies are associated with a history of subfertility. In the data

used in this study, TTP was not associated with a higher rate of all congenital

anomalies combined. We did find an association between increased TTP (>2 years)

and polydactyly, and a trend for an association with penoscrotal hypospadia and cystic

kidney. Surprisingly though, clefts were found to occur less often in women with a

TTP of more than 2 years compared to women with a TTP less than 2 years.

6.2 Strengths and weaknesses

Strengths

An important strength of this study is its thoroughness, since all types of births

regardless of gestational age were included. As a result, the coverage of all major

congenital anomalies in the studied area is close to 100%. In this study, terminations

due to fetal anomalies are also included, unlike most studies in this field. In total

almost 25% of pregnancies where terminated, emphasizing the value of including

terminated pregnancies in this study.

EUROCAT’s long-standing registration can definitely be considered a

strength, since the database allows a large number of cases and controls to be used.

The data which had been collected over an extended period of time, combined with a

22

parental questionnaire which is quite detailed, helped reveal a large collection of

variables which necessitated the adjustment of the analyses.

Likewise, another strength is the inclusion of children with congenital

anomalies up to the age of 10 years, since the vast majority of congenital anomalies is

detected before this age, meaning that the study covered most of the occurring

congenital anomalies.

The chance of differential recall of information on confounders between

mothers of malformed and non-malformed children is greatly reduced by selecting

malformed controls rather than seemingly healthy controls. Therefore, the use of

malformed controls can also be considered one of the strengths of this study.

Furthermore, this study is the first case-control study investigating whether

specific congenital anomalies are associated with a history of subfertility. Finally, we

were able to objectify fertility status by searching fertility records.

Weaknesses

A possible limitation is that data on fertility status and other determinants were

collected retrospectively, which means that recall bias may pose a problem, though of

the parental questionnaires 60% was returned before the child was aged two years.

The need to objectify the fertility status of participants by a fertility record search

resulted in loss of data, as participants with unconfirmed subfertility were excluded.

The loss of these data did reduce study power, but it is unlikely that a reduced data

input introduced bias. Parents who indicated that they did not have trouble conceiving

this pregnancy could however be subfertile according to our definition (TTP>1 year),

yet felt they conceived without a problem. These couples were included in the

analyses as ‘fertile couples’. Subfertile couples could subsequently be placed in the

‘fertile couples’ group. Due to this classification, there could have been an

underestimation of the effect of subfertility may have on the presence of congenital

anomalies occurring in offspring.

Finally, we only had relatively small numbers for the analysis on TTP, despite

the large size of the original study population. As a consequence, our study had

limited statistical power to evaluate an effect of increased TTP on specific congenital

abnormalities.

6.3 Potential underlying mechanisms

We found a large disparity between the characteristics of the mothers and infant of

both the cases and controls, particularly with regard to pregnancy outcomes, maternal

age, maternal education, parity, smoking and a history of pregnancies affected by a

congenital anomaly. An explanation of these differences in maternal characteristics is

the fact that prevalence of chromosomal anomalies, the largest number of our

controls, is higher in elderly women(39,40).

One matter in particular still remains unclear, which is the exact mechanism

behind subfertility leading to congenital anomalies, particularly hypospadias, in

fetuses. In academic literature an association between the use of diethylstilbestrol

(DES) in mothers and hypospadia in their offspring is described(41). Human exposure

23

to DES occurred through diverse sources, such as dietary ingestion from

supplemented cattle feed as well as medical treatment for certain conditions, including

breast and prostate cancers. From around 1940 to 1970, DES was given to pregnant

women in the mistaken belief it would reduce the risk of pregnancy complications and

miscarriages. DES use in mothers is known to affect fertility negatively in their

offspring. The association between DES use in mothers and fertility problems in DES-

daughters might be a clarifying mechanism for the possible association between

subfertile DES-daugthers and hypospadia in their sons.

The data show a decreased risk for the entire group consisting of all congenital

anomalies in women with a history of subfertility compared to fertile women.

Although the association is not statistically significant, these findings do, in fact,

contradict academic sources. A possible underlying mechanism explaining the

findings is the change of lifestyle in couples with subfertility problems. It can be

assumed that couples who are trying to conceive for a longer period are trying to

change their lifestyle, such as using folic acid. Another possible explanation for the

findings is the diversity in underlying causes of subfertility in our data. Examples of

underlying causes of subfertility in our data are sperm abnormalities (genetic quality

of the sperm, as well as its volume and motility), hormonal dysregulation,

endometriosis, uterine problems and subfertility with unknown cause. Since all

different causes were grouped together, it was impossible to specify which underlying

cause is responsible for a particular congenital anomaly. As such, there is no clear

insight into the relationship between subfertility and congenital anomalies .

6.4 Further research

More research is necessary to investigate the possible biological mechanisms

clarifying the association between subfertility and hypospadias and other congenital

anomalies. It will be important to determine which mechanisms may cause a higher

risk of having a child with a polydactyly after conceiving via IVF or ICSI.

To investigate whether there is a possible association between a history of

subfertility and the occurrence of chromosomal anomalies, microdeletion anomalies

or monogenic anomalies it is highly recommended to have this study repeated with

non-malformed controls. It will likewise be advisable to replicate this study with

larger numbers in the TTP>2 years category, since a positive but non-significant trend

was observed between increased TTP and a higher risk of a penoscrotal hypospadia-

affected pregnancy, a polydactyly-affected pregnancy and a cystic kidney-affected

pregnancy.

Another recommendation for further research is the possibility that DES use

in grandmothers could act as a mechanism, indicating an association between

subfertile DES-daughters and their offspring with a hypospadia.

Finally, it would be interesting to conduct further research into whether

anorectal malformations, especially VACTERL association, is due to ART procedures

or caused by the underlying subfertility instead(42).

24

6.5 Public health implications

Before this study was conducted, it was unclear whether a history of subfertility is

associated with specific types of birth defects. Only the risk for a penoscrotal

hypospadia appears to be higher in the Sub-NC group. This is a reassuring finding

which is of importance for correct counseling of subfertile couples and may be useful

for (extended) prenatal screening in the future.

6.6 Conclusion

From all the research that had been collected, this study appears to be the first to

examine associations between specific congenital anomalies and a history of

subfertility. Our study partly confirm the hypothesis that women with a history of

subfertility are at an increased risk of giving birth to a child with a specific major

congenital anomaly, namely a penoscrotal hypospadia. For identifying possible

mechanisms which cause the increased risk of hypospadias, further research is

recommended. The results of this study may be of importance for public health, since

subfertility is a growing health issue since women are conceiving at older ages.

25

7 Acknowledgments

In conducting my research, analyzing the data, drafting and finally writing this thesis,

I am greatly indebted to a number of people, for whom I would like to use this

opportunity to show them my deep gratitude.

I wish to thank Hermien de Walle, for her help in planning this research and

for her limitless support. Without her warm support, I doubt whether I would have

been able to finish this research and write my thesis. I would also like to thank

Annemieke Hoek and Maaike Haadsma for helping me to brainstorm on the contents

of this study. My thanks also go to Jorien Seggers for her contributions to the content

of the study, I also wish to thank her for helping me to collect the data. I am much

obliged to Nicole Siemensma, Lies ter Beek and Jorieke van Kammen, for their help

finding answers to some difficult questions, their advice in drafting a proper letter to

send to hospitals and for the classification of some cases. For enabling this research, I

am indebted to Marian Bakker, registry leader of EUROCAT Northern Netherlands

department. Finally, my special thanks go to the whole EUROCAT team for the warm

welcome.

Marly Bos

18-01-2013

26

8 References

(1) Flenady V, Middleton P, Smith GC, Duke W, Erwich JJ, Khong TY, et al. Stillbirths: the

way forward in high-income countries. Lancet 2011 May 14;377(9778):1703-1717.

(2) Up To Date Approach to congenital malformations. Available at:

http://www.uptodate.com/contents/approach-to-congenital-

malformations?source=search_result&search=congenital+anomalies&selectedTitle=1%7E15

0.

(3) Eurocat Northern Netherlands. Available at:

http://www.rug.nl/umcg/faculteit/disciplinegroepen/medischegenetica/eurocat/AlgemeneCijfe

rsTabel1tm4.pdf.

(4) Juul S, Karmaus W, Olsen J. Regional differences in waiting time to pregnancy:

pregnancy-based surveys from Denmark, France, Germany, Italy and Sweden. The European

Infertility and Subfecundity Study Group. Hum Reprod 1999 May;14(5):1250-1254.

(5) International Committee for Monitoring Assisted Reproductive Technology, Adamson

GD, de Mouzon J, Lancaster P, Nygren KG, Sullivan E, et al. World collaborative report on

in vitro fertilization, 2000. Fertil Steril 2006 Jun;85(6):1586-1622.

(6) Nyboe Andersen A, Goossens V, Bhattacharya S, Ferraretti AP, Kupka MS, de Mouzon J,

et al. Assisted reproductive technology and intrauterine inseminations in Europe, 2005: results

generated from European registers by ESHRE: ESHRE. The European IVF Monitoring

Programme (EIM), for the European Society of Human Reproduction and Embryology

(ESHRE). Hum Reprod 2009 Jun;24(6):1267-1287.

(7) Reefhuis J, Honein MA, Schieve LA, Correa A, Hobbs CA, Rasmussen SA, et al. Assisted

reproductive technology and major structural birth defects in the United States. Hum Reprod

2009 Feb;24(2):360-366.

(8) Jolly M, Sebire N, Harris J, Robinson S, Regan L. The risks associated with pregnancy in

women aged 35 years or older. Hum Reprod 2000 Nov;15(11):2433-2437.

(9) Nygren KG, Finnstrom O, Kallen B, Olausson PO. Population-based Swedish studies of

outcomes after in vitro fertilisation. Acta Obstet Gynecol Scand 2007;86(7):774-782.

(10) Kallen B, Finnstrom O, Nygren KG, Otterblad Olausson P. In vitro fertilization in

Sweden: maternal characteristics. Acta Obstet Gynecol Scand 2005 Dec;84(12):1185-1191.

(11) Helmerhorst FM, Perquin DA, Donker D, Keirse MJ. Perinatal outcome of singletons

and twins after assisted conception: a systematic review of controlled studies. BMJ 2004 Jan

31;328(7434):261.

(12) Jackson RA, Gibson KA, Wu YW, Croughan MS. Perinatal outcomes in singletons

following in vitro fertilization: a meta-analysis. Obstet Gynecol 2004 Mar;103(3):551-563.

27

(13) Up To Date In vitro fertilization. Available at: http://www.uptodate.com.proxy-

ub.rug.nl/contents/in-vitro-

fertilization?source=search_result&search=IVF+and+ICSI&selectedTitle=2~150.

(14) Heineman MJ, Evers JLH, Massuger LFAG, Steegers EAP. Vruchtbaarheidsstoornissen.

Obstetrie en Gynaecologie: Elversier Gezondheidszorg; 2007. p. 647-682.

(15) Palermo G, Joris H, Devroey P, Van Steirteghem AC. Pregnancies after intracytoplasmic

injection of single spermatozoon into an oocyte. Lancet 1992 Jul 4;340(8810):17-18.

(16) Lanzendorf SE, Maloney MK, Veeck LL, Slusser J, Hodgen GD, Rosenwaks Z. A

preclinical evaluation of pronuclear formation by microinjection of human spermatozoa into

human oocytes. Fertil Steril 1988 May;49(5):835-842.

(17) Wennerholm UB, Soderstrom-Anttila V, Bergh C, Aittomaki K, Hazekamp J, Nygren

KG, et al. Children born after cryopreservation of embryos or oocytes: a systematic review of

outcome data. Hum Reprod 2009 Sep;24(9):2158-2172.

(18) Maheshwari A, Pandey S, Shetty A, Hamilton M, Bhattacharya S. Obstetric and perinatal

outcomes in singleton pregnancies resulting from the transfer of frozen thawed versus fresh

embryos generated through in vitro fertilization treatment: a systematic review and meta-

analysis. Fertil Steril 2012 Aug;98(2):368-77.e1-9.

(19) Schieve LA, Meikle SF, Ferre C, Peterson HB, Jeng G, Wilcox LS. Low and very low

birth weight in infants conceived with use of assisted reproductive technology. N Engl J Med

2002 Mar 7;346(10):731-737.

(20) Kallen B, Finnstrom O, Nygren KG, Olausson PO. In vitro fertilization in Sweden: child

morbidity including cancer risk. Fertil Steril 2005 Sep;84(3):605-610.

(21) Hansen M, Bower C, Milne E, de Klerk N, Kurinczuk JJ. Assisted reproductive

technologies and the risk of birth defects--a systematic review. Hum Reprod 2005

Feb;20(2):328-338.

(22) Hansen M, Kurinczuk JJ, Bower C, Webb S. The risk of major birth defects after

intracytoplasmic sperm injection and in vitro fertilization. N Engl J Med 2002 Mar

7;346(10):725-730.

(23) Kallen B, Finnstrom O, Nygren KG, Olausson PO. In vitro fertilization (IVF) in Sweden:

risk for congenital malformations after different IVF methods. Birth Defects Res A Clin Mol

Teratol 2005 Mar;73(3):162-169.

(24) Davies MJ, Moore VM, Willson KJ, Van Essen P, Priest K, Scott H, et al. Reproductive

technologies and the risk of birth defects. N Engl J Med 2012 May 10;366(19):1803-1813.

(25) Ericson A, Kallen B. Congenital malformations in infants born after IVF: a population-

based study. Hum Reprod 2001 Mar;16(3):504-509.

28

(26) Silver RI, Rodriguez R, Chang TS, Gearhart JP. In vitro fertilization is associated with an

increased risk of hypospadias. J Urol 1999 Jun;161(6):1954-1957.

(27) Wennerholm UB, Bergh C, Hamberger L, Lundin K, Nilsson L, Wikland M, et al.

Incidence of congenital malformations in children born after ICSI. Hum Reprod 2000

Apr;15(4):944-948.

(28) Draper ES, Kurinczuk JJ, Abrams KR, Clarke M. Assessment of separate contributions

to perinatal mortality of infertility history and treatment: a case-control analysis. Lancet 1999

May 22;353(9166):1746-1749.

(29) Ghazi HA, Spielberger C, Kallen B. Delivery outcome after infertility--a registry study.

Fertil Steril 1991 Apr;55(4):726-732.

(30) Zhu JL, Basso O, Obel C, Bille C, Olsen J. Infertility, infertility treatment, and

congenital malformations: Danish national birth cohort. BMJ 2006 Sep 30;333(7570):679.

(31) Seggers J, Haadsma ML, Bos AF, Heineman MJ, Keating P, Middelburg KJ, et al.

Dysmorphic features in 2-year-old IVF/ICSI offspring. Early Hum Dev 2012 Oct;88(10):823-

829.

(32) Eurocat Central Registry guidelines. Available at: http://www.eurocat-netwerk.eu/.

(33) Botto LD, Lin AE, Riehle-Colarusso T, Malik S, Correa A, National Birth Defects

Prevention Study. Seeking causes: Classifying and evaluating congenital heart defects in

etiologic studies. Birth Defects Res A Clin Mol Teratol 2007 Oct;79(10):714-727.

(34) Jentink J, Loane MA, Dolk H, Barisic I, Garne E, Morris JK, et al. Valproic acid

monotherapy in pregnancy and major congenital malformations. N Engl J Med 2010 Jun

10;362(23):2185-2193.

(35) WHO BMI. Available at: http://apps.who.int/bmi/index.jsp?introPage=intro_3.html.

(36) Streiner DL, Norman GR. Correction for multiple testing: is there a resolution? Chest

2011 Jul;140(1):16-18.

(37) Garne E, Loane M, Dolk H, Barisic I, Addor MC, Arriola L, et al. Spectrum of

congenital anomalies in pregnancies with pregestational diabetes. Birth Defects Res A Clin

Mol Teratol 2012 Mar;94(3):134-140.

(38) Reefhuis J, Honein MA, Schieve LA, Rasmussen SA, National Birth Defects Prevention

Study. Use of clomiphene citrate and birth defects, National Birth Defects Prevention Study,

1997-2005. Hum Reprod 2011 Feb;26(2):451-457.

(39) Hook EB. Rates of chromosome abnormalities at different maternal ages. Obstet Gynecol

1981 Sep;58(3):282-285.

29

(40) Up To Date Effect of advanced age on fertility and pregnancy in women. Available at:

http://www.uptodate.com/contents/effect-of-advanced-age-on-fertility-and-pregnancy-in-

women?source=search_result&search=Effect+of+advanced+age+on+fertility+and+pregnancy

+in+women+47&selectedTitle=1%7E150.

(41) Klip H, Verloop J, van Gool JD, Koster ME, Burger CW, van Leeuwen FE, et al.

Hypospadias in sons of women exposed to diethylstilbestrol in utero: a cohort study. Lancet

2002 Mar 30;359(9312):1102-1107.

(42) Zwink N, Jenetzky E, Schmiedeke E, Schmidt D, Marzheuser S, Grasshoff-Derr S, et al.

Assisted reproductive techniques and the risk of anorectal malformations: a German case-

control study. Orphanet J Rare Dis 2012 Sep 15;7(1):65.

30

9 Nederlandse samenvatting

Inleiding: In de literatuur is een toegenomen risico op het voorkomen van aangeboren

afwijkingen bij in vitro fertilisatie (IVF)- en intracytoplasmatische sperma injectie

(ICSI)- kinderen beschreven. Specifiek worden een toename van cardiovasculaire,

musculoskeletale, urogenitale en gastrointestinale afwijkingen gezien. Het is nog niet

bekend of deze toegenomen risico’s kunnen worden verklaard door patiënt

karakteristieken gerelateerd aan subfertiliteit, of aan de behandelingstechnieken op

zichzelf. Het primaire doel van de studie was om te evalueren of specifieke

aangeboren afwijkingen vaker voorkomen bij de kinderen van subfertiele paren, die

natuurlijk zwanger werden (Sub-NC) in vergelijking met de kinderen van fertiele

paren. Het secundaire doel was het evalueren of specifieke aangeboren afwijkingen

vaker (of minder vaak) voorkomen na natuurlijke conceptie bij subfertiliteit in

vergelijking met kinderen van paren die middels IVF/ICSI zwanger zijn geworden.

Het tertiaire doel was om te kijken of de ‘time to pregnancy’ (TTP) bij subfertiele

paren een positieve associatie laat zien met het voorkomen van specifieke aangeboren

afwijkingen.

Methode: 4392 cases en 1456 (aangedane) controles tussen 1997 en 2010 geboren,

waarvan informatie over de fertiliteitstatus van de moeder bekend was, werden uit de

EUROCAT registratie geïncludeerd. Cases waren kinderen en foetussen met major

aangeboren afwijkingen met een niet-genetische oorzaak. Controles waren

gedefinieerd als kinderen en foetussen met chromosomale afwijking, een microdeletie

of een monogene afwijking. Multivariate logistische regressie met de vergelijking

tussen drie groepen van kinderen en foetussen vond plaats: Sub-NC (n=296),

IVF/ICSI (n=174) en kinderen en foetussen van fertiele paren (n=5378). Mogelijke

confounders waren leeftijd van de moeder bij de conceptie, roken tijdens de

zwangerschap, ‘body mass index’ voor de zwangerschap, opleidingsniveau van de

moeder, pariteit, ander kind met een aangeboren afwijking en maternale diabetes.

Resultaten: We vonden een hoger risico voor Sub-NC vrouwen in vergelijking met

fertiele vrouwen op het krijgen van een kind met een penoscrotale hypospadie

(gecorrigeerde odds ratio [aOR]: 6.80; 95% confidence interval [CI]: 2.35 – 19.63).

Subfertiele vrouwen die zwanger werden middels IVF/ICSI hebben een verhoogd

risico op het krijgen van een kind met een afwijking aan de ledematen (aOR: 2.80; CI:

1.51 – 5.20), in het bijzonder polydactylie (aOR: 5.26; CI: 2.47 – 11.20). In

vergelijking met fertiele vrouwen, heeft de totale groep van subfertiele vrouwen (Sub-

NC en IVF/ICSI) een verhoogd risico op het krijgen van een kind met een epispadie

(aOR: 4.28; CI: 1.04 – 17.57) en cysteuze nieren (aOR: 2.10; CI: 1.02 – 4.34).

Verrassend genoeg is er een bijna significante afname van de totale groep van

aangeboren afwijkingen te zien bij de Sub-NC vrouwen in vergelijking met fertiele

vrouwen. Aanvullend is er in de totale groep van subfertiele vrouwen een significant

lager risico op het krijgen van een kind met een congenitale heupdysplasie in

31

vergelijking met de fertiele vrouwen, zoals tevens te zien is bij groep Sub-NC

vrouwen alleen.

Conclusie: Deze studie heeft laten zien dat subfertiele vrouwen, die vervolgens wel

op de natuurlijke manier zwanger werden, een verhoogd risico hebben op het krijgen

van een kind met specifieke aangeboren afwijkingen, in het bijzonder penoscrotale

hypospadie. Gezien het feit dat subfertiliteit een groeiend probleem is, aangezien

vrouwen op steeds latere leeftijd zwanger proberen te raken, zijn onze bevindingen

van belang voor de volksgezondheid.

32

Appendix Maternal and infant characteristics

Cases Controls P-value(n=4392) (n=1456)

Characteristic n (%) n (%)

Pregnancy outcome <0.001

Live birth 3915 (89.1) 933 (64.1)

Live birth but deceased 240 (5.5) 139 (9.5)

Miscarriage 27 (0.6) 41 (2.8)

Termination of pregnancy 156 (3.6) 291 (20.0)

Stillbirth 54 (1.2) 52 (3.6)

0.013

Male 2445 (55.7) 757 (52.0)

Female 1941 (44.2) 699 (48.0)

Unknown 6 (0.1) 0 (0.0)

<0.001

< 25 448 (10.2) 105 (7.2)

25-34 3164 (72.0) 905 (62.2)

> 34 736 (16.8) 432 (29.7)

Missing 44 (1.0) 14 (1.0)

Paternal age at delivery (years) <0.001

< 25 141 (3.2) 34 (2.3)

25-34 2673 (60.9) 742 (51.0)

> 34 1454 (33.1) 638 (43.8)

Missing 124 (2.8) 42 (2.9)

Maternal education 0.015

Low 773 (17,6) 254 (17.4)

Middle 2082 (47.4) 629 (43.2)

High 1329 (30.3) 491 (33.7)

Missing 208 (4.7) 82 (5.6)

Smoking during pregnancy 0.033

Yes 1071 (24.4) 312 (21.4)

No 3232 (73.6) 1100 (75.5)

Missing 89 (2.0) 44 (3.0)

Folic acid use 0.865

Not or in wrong period 1219 (27.8) 400 (27.5)

Periconceptional 2929 (66.7) 950 (65.2)

Missing 244 (5.6) 106 (7.3)

Body mass index (kg/m2

) 0.128

< 18.5 141 (3.2) 39 (2.7)

Infant sex

Maternal age at conception (years)

Distribution of maternal and infant characteristics for cases and controls of

subjects with known fertilitystatus.

33

18.5 - 25 2619 (59.6) 900 (61.8)

> 25 1382 (31.5) 421 (28.9)

Missing 250 (5.7) 96 (6.6)

Diabetes mellitus 0.922

Yes 16 (0.4) 5 (0.3)

No 4305 (98.0) 1415 (97.2)

Missing 71 (1.6) 36 (2.5)

Parity < 0.001

0 1717 (39.1) 462 (31.7)

1 1438 (32.7) 497 (34.1)

2 716 (16.3) 261 (17.9)

3 or more 482 (11.0) 222 (15.2)

Missing 39 (0.9) 14 (1.0)

Number of children this pregnancy 0.149

1 4167 (94.9) 1391 (95.5)

2 193 (4.4) 52 (3.6)

3 6 (0.1) 0 (0.0)

Missing 26 (0.6) 13 (0.9)

Another child with congenital anomaly < 0.001

Yes 513 (11.7) 250 (17.2)

No 3845 (87.5) 1188 (81.6)

Missing 34 (0.8) 18 (1.2)

Fertility status 0.006

Fertile 4065 (92.6) 1313 (90.2)

Sub-NC 200 (4.6) 96 (6.6)

IVF/ICSI 127 (2.9) 47 (3.2)

0.569

< 2 years 129 (39.4) 60 (41.9)

≥ 2 years 165 (50.5) 68 (47.6)

Missing, but with OFO 33 (10.1) 15 (10.5)

Time to pregnancy within subfertile patients n=470

Sub-NC = subfertile, but naturally conceived; OFO = oriëntating fertility research.